University of Zurich. Chlamydial abortion in ruminants. Serological, epidemiological and diagnostic investigations. Zurich Open Repository and Archive

Transcription

1 University of Zurich Zurich Open Repository and Archive Winterthurerstr. 190 CH-8057 Zurich Year: 2008 Chlamydial abortion in ruminants. Serological, epidemiological and diagnostic investigations Borel, N Borel, N. Chlamydial abortion in ruminants. Serological, epidemiological and diagnostic investigations. 2008, University of Zurich, Vetsuisse Faculty. Postprint available at: Posted at the Zurich Open Repository and Archive, University of Zurich. Originally published at: University of Zurich, Vetsuisse Faculty, 2008.

2 University of Zurich Zurich Open Repository and Archive Winterthurerstr. 190 CH-8057 Zurich Year: 2008 Chlamydial Abortion in Ruminants; Serological, Epidemiological and Diagnostic Investigations Borel, N Borel, N. Chlamydial Abortion in Ruminants; Serological, Epidemiological and Diagnostic Investigations. 2008, University of Zurich, Vetsuisse Faculty. Postprint available at: Posted at the Zurich Open Repository and Archive, University of Zurich. Originally published at: University of Zurich, Vetsuisse Faculty, 2008.

3 Chlamydial Abortion in Ruminants Serological, Epidemiological and Diagnostic Investigations Habilitationsschrift zur Erlangung der Venia legendi der Vetsuisse-Fakultät der Universität Zürich vorgelegt von Nicole Simona Borel Dr. med. vet. Von Couvet und Neuchâtel/NE Zürich 2008

4

5 Institut für Veterinärpathologie der Universität Zürich, Vetsuisse-Fakultät (Direktor: Prof. Dr. A. Pospischil) CHLAMYDIAL ABORTION IN RUMINANTS: SEROLOGICAL, EPIDEMIOLOGICAL AND DIAGNOSTIC INVESTIGATIONS HABILITATIONSSCHRIFT zur Erlangung der Venia legendi für Veterinärpathologie an der Vetsuisse-Fakultät der Universität Zürich vorgelegt von Nicole Simona Borel Dr. med. vet. Von Couvet und Neuchâtel/NE Zürich 2008

6 Habilitationsschrift Nicole Simona Borel 1. Auflage 2008 Nicole Simona Borel Institut für Veterinärpathologie der Universität Zürich Winterthurerstrasse Zürich Abgedruckte Artikel: Jeweiliger Verlag Satz: FocusedPublishing GmbH Berggasse Russikon info@focusedpublishing.com Verwendete Schriften: Frutiger light condensed, Frutiger bold condensed und Berthold Garamond BE Regular 2

7 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Table of contents 1. Introduction: The Chlamydiales 2. Serological Investigations in Small Ruminants 2.1. Introduction and Conclusion 2.2. Seroprevalences for Ovine Enzootic Abortion in Switzerland 2.3. Ovine Enzootic Abortion (OEA): antibody response in vaccinated sheep compared to naturally infected sheep 2.4. Ovine Enzootic Abortion (OEA): a comparison of antibody responses of vaccinated and naturally-infected Swiss sheep over a two year period 3. Epidemiological Investigations in Ruminants 3.1. Introduction and Conclusion 3.2. Prevalence of chlamydiae in semen and genital tracts of bulls, rams and bucks 3.3. Chlamydia-related abortions in cattle from Graubunden, Switzerland 3.4. Parachlamydia spp. and related Chlamydia-like organisms and bovine abortion 3.5. Evidence for Parachlamydia spp. in bovine abortion 4. Development and Evaluation of New Chlamydial Diagnostic Tools 4.1. Introduction and Conclusion 4.2. Direct identification of chlamydiae from clinical samples using a DNA microarray assay - a validation study 4.3. Tissue MicroArray (TMA) analysis of normal and persistent Chlamydophila pneumoniae infection 5. Summary and Future Work 6. Literature 7. Acknowledgments 3

8 Habilitationsschrift Nicole Simona Borel 4

9 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 1. Introduction: The Chlamydiales Members of the order Chlamydiales are a unique and closely related group of Gramnegative obligate intracellular bacteria. Despite the evolutionary maintenance of a complex developmental cycle and a relatively small genome size (about 1000 kb), chlamydiae are ubiquitous in the environment and are able to infect an amazingly wide range of hosts and an equally wider range of tissue sites in these hosts. Until recently, one family Chlamydiaceae and one genus Chlamydia was classified in the order Chlamydiales. Under this classification, four recognized chlamydial species (Chlamydia trachomatis, Chlamydia pneumoniae, Chlamydia pecorum and Chlamydia psittaci) were grouped in the family Chlamydiaceae. The recent development and application of broad-range ribosomal RNA (rrna) polymerase chain reaction (PCR) strategies for detection of members of the order Chlamydiales has demonstrated the restrictions of this earlier classification model with the identification of novel members sharing greater than 80% 16S rrna relatedness and a biphasic developmental cycle (Ossewaarde and Meijer, 1999; Everett, 2000). Prompted by an awareness of the increased diversity in the order Chlamydiales, a new classification system was proposed based on extensive analysis of signature sequences in the 16S and 23S rrna genes (Everett et al., 1999). This new official classification system separates the family Chlamydiaceae into two lineages of nine species separating the predominantly animal pathogens Chlamydia psittaci, Chlamydia pecorum and Chlamydia pneumoniae into a new genus Chlamydophila (Cp.). Three new families (Parachlamydiaceae, Waddliaceae and Simkaniaceae) were also proposed to classify novel Chlamydia -like organisms. Many more new species of these novel Chlamydia-like organisms have been described recently (Corsaro and Greub, 2006). Chlamydiae can cause a wide range of diseases in non-human mammals, birds and reptiles (reviewed in Longbottom and Coulter, 2003). In veterinary medicine, chlamydial infections are a common problem mostly leading to keratoconjunctivitis, abortion, mastitis, polyarthritis, enteritis and respiratory infections in affected mammals (Longbottom and Coulter, 2003). In birds, Chlamydiaceae cause infections of the respiratory and the gastrointestinal tract known as rhinoconjunctivitis, sacculitis, hepatitis and pneumoenteritis (Vanrompay et al., 1995). Current data suggest that there is an accumulation of chlamydial infections in certain dairy farms (Hässig and Lubsen, 1998; Cavirani et al., 2001; Twomey et al., 2006), but precise data on the worldwide or regional spread and distribution of chlamydial infections in farm livestock is not available (Longbottom, 2004). Interestingly, potentially zoonotic chlamydiae, such as Chlamydophila (Cp.) abortus and Cp. psittaci have been found in calves and pigs - beside the presence of host-adapted chlamydiae (e.g. Cp. pecorum in cattle) (Jee et al., 2004). The route of transmission between animals is mostly fecaloral (Sachse and Grossmann, 2002). Clinical detection of individual species of Chlamydiaceae has advanced by introduction of new PCR techniques, immunofluorescence and Fluorescent in situ Hybridization (FisH), thus allowing highly sensitive and specific confirmation of infection (Poppert et al., 2002; Sachse and Grossmann, 2002; Kaltenboeck et al., 2005). Since some chlamydiae are pathogenic for humans (Peeling et al., 1996), e.g. Chlamydophila (Cp.) psittaci and Cp. abortus, there are occupational hazards to persons in direct contact with infected animals and subsequent high-level exposure (Hinton et al., 1993). Cp. abortus (formerly Chlamydia psittaci serotype 1) is targeting the placenta of ewes and goats, causing tissue damage and inflammation leading to acute placentitis and abortion (Aitken, 1993). Cp. abortus infection is 5

10 Habilitationsschrift Nicole Simona Borel asymptomatic or at the best oligosymptomatic in non-pregnant animals and during pregnancy. Severe disease resulting in abortion with expulsion of dead lambs occurs 2 to 3 weeks before the end of gestation. Ovine enzootic abortion (OEA) caused by Cp. abortus, is also a serious and potentially life-threatening zoonosis, affecting pregnant women after contact with infected sheep and goats, leading to severe febrile illness in pregnancy. Transmissions to humans have been repeatedly reported throughout Europe (Buxton, 1986; Kampinga et al., 2000; Pospischil et al., 2002a). Bovine chlamydial abortion due to Cp. abortus is similar to OEA but more sporadic and less common (Longbottom and Coulter, 2003). Infections by Cp. pecorum include a series of severe clinical manifestations such as bovine encephalomyelitis and purulent endometritis among others. Infections by Cp. psittaci have also been reported in two cases (Piercy et al., 1999). Recent data have demonstrated that subclinical chlamydial infections by both species, Cp. abortus and Cp. pecorum, are ubiquitous in cattle and often not detected due to the insensitivity of diagnostic techniques (Jee et al., 2004; Kaltenboeck et al., 2005). Such subclinical infections do occasionally manifest into clinical disease. offer cheaper, safer and more stable alternatives however appropriate candidate vaccine antigens have yet to be defined (Longbottom and Livingstone, 2006). Against this background, the aim of the studies combined in this work was to provide scientific information that will contribute to the general knowledge about chlamydial abortions in ruminants. The following study is subdivided into three parts: Serological investigations in small ruminants Epidemiological investigations in ruminants Development and evaluation of new chlamydial diagnostic tool. Sequence data is now available for Cp. abortus strain S26/3 (Thomson et al., 2007). The genome of Cp. abortus (1.14 million base pairs) contains 961 predicted coding sequences, the vast majority of which are conserved with those of other members of the Chlamydiaceae. Within this conserved Cp. abortus core, highly variable protein families, including transmembrane head (TMH)/Inc and polymorphic membrane protein D (PmpD) have been suggested to determine diversity of host tropism and disease causation (Thomson et al., 2007). At present, a whole-organism based killed or live attenuated vaccine against Cp. abortus is available for veterinary use. Development of subunit or multi-component vaccines may 6

11 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 2. Serological Investigations in Small Ruminants 2.1. Introduction and Conclusion The detection of antibodies in animal chlamydial infections has multiple purposes. Confirmation of clinical disease or confirmation of the presence or absence of infection, performance of epidemiological surveys to estimate the prevalence of infection, or the determination of immune status after vaccination to name some. In general, antibodies are identified by binding to inclusions, to elementary bodies (EBs) or to individual chlamydial antigens. Detection of the bound antibody is achieved by fluorescent (indirect immunofluorescence, IF, and microimmunofluorescence test, MIF) or otherwise tagged secondary antibodies (indirect enzyme-linked immunosorbent assay, ELISA) or by estimating the consumption or fixation of complement (direct and indirect complement fixation test, CFT). Infected animals are often asymptomatic, possibly shedding infectious chlamydiae from the intestine, with low or borderline levels of antibodies (Storz, 1971 and 1988; Longbottom and Coulter, 2003). Furthermore, multiple chlamydial species can infect the same host, a situation often encountered in ruminants. The selection of the suitable antibody test depends therefore on the particular epidemiological situation in the infected host species. The serological diagnosis of Ovine Enzootic Abortion (OEA) has been complicated by multiple factors, the first being the latent nature of the infection itself which remains undetected until day 90 of gestation (Buxton et al., 1990). Furthermore, ruminants are also often infected with Chlamydophila pecorum (Cp. pecorum), which causes a variety of clinical manifestations depending on the subspecies (Fukushi and Hirai, 1992; Kaltenboeck et al., 1993; Philips and Clarkson, 1995; Anderson et al., 1996). In addition, subclinical enteric infections due to Cp. abortus have been known for a long time and Cp. abortus strains have been isolated from the feces of sheep with no previous history of abortion but with borderline levels of antibodies (Storz, 1971; Rodolakis and Souriau, 1989; Salti-Montesanto et al., 1997; Gut-Zangger et al., 1999). For more than 50 years the CFT has been the most widely used assay for the serological diagnosis of OEA and the one recommended by the Office Internationale des Epizooties (OIE) ( The CFT is the first described LPS-based assay and the most used in veterinary laboratories (Stamp et al., 1952). In OEA, CFT-serum titers rise at the time of abortion and remain high for at least 6 weeks (Storz, 1971). However, the CFT lacks specificity due to its antigen, consisting mainly of the heat-resistant lipopolysaccharide (LPS), common in all Chlamydiaceae species (Brade et al., 1987). The lack of both an appropriate specific reference method for Cp. abortus and a confirmatory assay for the presence of Cp. pecorum infection, other than the cumbersome isolation of the organism, have made the evaluation of novel assays for the improvement of the serological diagnosis of OEA difficult. Nevertheless, several laboratory-based assays have been developed including indirect ELISAs using whole elementary body (EB) or extracts thereof (Cevenini et al., 1989; Markey et al., 1993; Anderson et al., 1995; Donn et al., 1997). Semi-quantitative methods have been proposed to distinguish between infections due to Cp. abortus and to Cp. pecorum such as an indirect IF assay and immunoblotting (Markey et al., 1993; Jones et al., 1997). ELISAs with purified LPS have been published (Sting and Hafez, 1992; Jones et al., 1997). With the introduction of molecular techniques more defined assays were developed based on recombinant LPSantigen (Griffiths et al., 1996), on the major outer membrane protein (MOMP; Kaltenboeck et al., 1997; Salti-Montesanto et al., 1997; Gut-Zangger et al., 1999; Borel et al., 2002; Hoelzle et al., 2004) and the polymorphic outer membrane proteins (POMPs) at 80-90kDa (Buendia et al., 2001; Longbottom et al., 2001 and 2002). A major step that has facilitated 7

12 Habilitationsschrift Nicole Simona Borel the evaluation of novel assays was the implementation of large panels of documented reference sera which have been shared among laboratories (Jones et al., 1997; Longbottom et al., 2001 and 2002; McCauley et al., 2007; Vretou et al., 2007). Assays based on the MOMP The diversity of the variable segments VS1-4 in the ompa genes of Cp. abortus and Cp. pecorum strains has provided the molecular basis for the development of a specific test capable of discrimination between the two species (Kaltenboeck et al., 1993). The enhanced seroreactivity of VS1 and VS2 compared to the other VS was confirmed when testing sera from experimental infections with recombinant variable domains expressed as fusion proteins (Livingstone et al., 2005). Another test based on the VS1 and VS2 of MOMP is the competitive ELISA (celisa) developed by Salti-Montesanto et al. in The test is based on the binding of specific monoclonal antibodies (mabs) against the MOMP variable segment 1 (VS1) or variable segment 2 (VS2) that is inhibited by the presence of serum antibodies. The inhibition depends largely on the quantity and quality (affinity) of the competing antibodies, which might recognize linear or conformational epitopes on the native MOMP (Vretou et al., 2001). The original test had two cut-offs, at 30% and at 55% inhibition; the gray zone encompassed sera originating from either enteric Cp. pecorum or enteric Cp. abortus infections (Salti-Montesanto et al., 1997; Gut-Zangger et al., 1999). A recent evaluation of the assay with documented reference sera and a cut-off at 50% inhibition reported 98.1% specificity and 77.7% sensitivity for the celisa (Vretou et al., 2007). The first study applying the celisa on field sera on 19 flocks in Switzerland was undertaken at the Institute of Veterinary Pathology, University of Zürich and gave a first insight into the epidemiologic situation of OEA in Switzerland: Gut-Zangger P, Vretou E, Psarrou E, Pospischil A, Thoma R: Chlamydienabort beim Schaf: Möglichkeiten der serologischen Diagnostik mit einem kompetitiven ELISA und Einblick in die epidemiologische Situation in der Schweiz, Schweiz Arch Tierheilk. 141, , Chlamydial abortion in sheep: possibilities of serological diagnosis with a competitive ELISA and insight into the epidemiological situation in Switzerland 466 sheep sera out of 19 flocks in Switzerland were examined by a competitive enzyme linked immunosorbent assay (celisa) for antibodies against Chlamydia psittaci serotype 1 ( ovine enzootic abortion ). Since numerous positive reactors were found in flocks without abortion history, 30 fecal samples out of two of these flocks were examined by PCR for evidence of chlamydial DNA. One of these samples turned out to contain DNA of Chlamydia psittaci serotype 1. These results suggest, that in Switzerland serotype 1 of Chlamydia psittaci is widespread not only as cause of chlamydial abortion but also as latent intestinal infection in sheep. The resulting difficulties for serological diagnosis of chlamydial abortion and possible solutions based on the celisa are discussed. The complement fixation test (CFT), still considered as standard method for serological examination for chlamydiae, has additionally been applied. In a further study, the pool approach was established and an insight into the seroprevalence of OEA in Switzerland was given: Borel N, Doherr MG, Vretou E, Psarrou E, Thoma R, Pospischil A: Chlamydienabort beim Schaf: Untersuchung der Seroprävalenz in der Schweiz mittels eines kompetitiven ELISA (celisa). Schweiz Arch Tierheilk. 144, , Ovine enzootic abortion: seroprevalence in Switzerland using a enzyme linked immunosorbent assay (celisa) The present study gives an overview over the seroprevalence of ovine enzootic abortion in Switzerland. 639 sheep flocks out of eight cantons in Switzerland were examined by a competitive enzyme linked immunosorbent assay (celisa) for antibodies against Chlamydophila abortus (Chlamydia psittaci serotype 1), the agent causing ovine enzootic abortion. The eight cantons included Aargau, Bern, Zürich, Appenzell-Ausserrhoden, Appenzell- Innerrhoden and Fribourg, the Vallais and the Graubünden. They were representative for 57% of the Swiss sheep flocks and for 60% of Swiss sheep population. In total, almost 19% (118) of the examined flocks were seropositive. Seroprevalence was the highest in Graubünden with 41%; this requires further examination and the evaluation of the need for a monitoring and controlling program. The examination of pooled sera made it possible to test a large number of samples with a reasonable amount of work. Higher sensitivity (92.9%) and specifity (97.6%) than the complement fixation test (CFT) in combination with testing 8

13 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations of pooled sera makes the celisa to be an usuable tool for serological screening on flock level. The successful evaluation of the pooled serum approach made it possible to test large number of samples with a reasonable amount of work. Higher sensitivity and specificity than the CFT in combination with testing of pooled sera made the celisa a useful tool for serological screening on flock level. We thus initiated a national seroepidemiological study in Switzerland: 2.2. Borel N, Doherr MG, Vretou E, Psarrou E, Thoma R, Pospischil A. Seroprevalences for ovine enzootic abortion in Switzerland. Prev Vet Med. 2004, 65: enzootic character of animal chlamydiosis however, that is the latency, persistence or chronic infection within the animal is far from being understood. It is possible that novel specific assays, when applied together, might help differentiate between the different conditions and thereby contribute to our understanding of the pathogenesis of the disease. The sheep sera investigated in that study originated from a population survey for Brucella melitensis in At that time, an inactivated, egg-grown preparation of Cp. abortus (Ovax Clamdia, Fatro, Italy) has been licensed in Switzerland by the Federal Veterinary Office (FVO). The possible influence of vaccination against OEA on the immunological responses in sheep was still unknown and thus falsepositive results due to vaccination would have been possible. Therefore, two further studies were initiated: 2.3. Borel N, Sachse K, Rassbach A, Bruckner L, Vretou E, Psarrou E, Pospischil A. Ovine enzootic abortion (OEA): antibody response in vaccinated compared to naturally infected sheep. Vet Res Commun. 2005, 29 Suppl 1: Gerber A, Thoma, R, Vretou E, Psarrou E, Kaiser C, Doherr MG, Zimmermann DR, Polkinghorne A, Pospischil A, Borel N. Ovine Enzootic Abortion (OEA): a comparison of antibody responses of vaccinated and naturally-infected Swiss sheep over a two year period. BMC Vet Res Sep 28;3:24. Conclusions Over the last years novel assays have been developed either in house or commercially available, that have met the requirements for a specific serological diagnosis of Cp. abortus. Novel needs have emerged, i.e. the specific detection of Cp. pecorum antibodies and the serological distinction between vaccinated and infected animals. The understanding of the 9

14 Habilitationsschrift Nicole Simona Borel 10

15 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Preventive Veterinary Medicine 65 (2004) Seroprevalences for ovine enzootic abortion in Switzerland N. Borel a, *, M.G. Doherr b, E. Vretou c, E. Psarrou c, R. Thoma d, A. Pospischil a a Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Winterthurerstrasse 268, CH-8057 Zurich, Switzerland b Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Berne, Bremgartenstrasse 109A, PO Box 8466, CH-3001 Bern, Switzerland c Laboratory of Biotechnology, Department of Microbiology, Hellenic Pasteur Institute, Athens , Greece d Cantonal Laboratory of Veterinary Bacteriology, Hofgraben 11, CH-7001 Chur, Switzerland Received 18 December 2003; received in revised form 13 July 2004; accepted 7 August 2004 Abstract Our aim was to assess the seroprevalence of Chlamydophila (Cd) abortus (Chlamydia psittaci serotype 1), denoted ovine enzootic abortion (OEA), in the Swiss sheep population. A competitive enzyme-linked immunosorbent assay (celisa) was adapted for the investigation of pooled serum samples (pool approach) and receiver-operator characteristic (ROC) analysis was applied to define the cut-off of the pool approach. At a cut-off value of 30% inhibition, the flock-level pooled sensitivity and specificity were 92.9% and 97.6% when compared to classifying the flock based on individual-animal samples. Subsequently, sera from 775 randomly selected flocks out of 11 cantons of Switzerland were investigated using the pool approach. The cantons included in the study represented 72% of the Swiss sheep flocks and 76% of Swiss sheep population. Antibodies against Cd. abortus were found in almost 19% (144) of the 775 examined sheep flocks. Test prevalences were adjusted for the imperfect test characteristics using the Rogan Gladen estimator and Bayesian inference. Seroprevalence was highest (43%) in the canton Graubünden. In the remaining 10 cantons the seroprevalence ranged from 2 to 29%. The celisa in combination with testing pooled sera and statistical methods for true * Corresponding author. Tel.: ; fax: address: n.borel@access.unizh.ch (N. Borel) /$ see front matter # 2004 Elsevier B.V. All rights reserved. doi: /j.prevetmed

16 Habilitationsschrift Nicole Simona Borel 206 N. Borel et al. / Preventive Veterinary Medicine 65 (2004) prevalence estimation provided a good survey tool at lower costs and time when compared to other approaches. # 2004 Elsevier B.V. All rights reserved. Keywords: Abortion; Sheep; Chlamydophila abortus (Chlamydia psittaci serotype 1); celisa; Seroprevalence; Pooling; Rogan Gladen estimator; Bayesian inference; OEA 1. Introduction Ovine enzootic abortion (OEA) induced by Chlamydophila (Cd) abortus (formerly Chlamydia psittaci serotype 1) is a major cause of reproductive failures in most sheepproducing countries (Aitken, 1993). In UK, it has been calculated that OEA causes economic losses of around million pounds per year (Aitken et al., 1990). In addition to the economic importance in the sheep industry, Cd. abortus induces abortions in humans as a result of contact with aborting sheep or goats (Buxton, 1986; Jorgensen, 1997; Rodolakis et al., 1998; Everett et al., 1999; Aitken, 2000; Pospischil et al., 2002a). Aborting ewes shed large amounts of infectious chlamydiae with placental and vaginal discharges (Aitken, 1993). In an epidemic of OEA, up to 30% of ewes abort in the last 3 weeks of gestation or give premature birth to weak or dead lambs. After abortion, the ewes develop a protective immunity and, in endemically infected flocks, 5 10% of the ewes abort annually (Rodolakis et al., 1998; Aitken, 2000). Despite the economically important losses due to late-term abortions and weak lambs, the prevalence of Cd. abortus is not well known in most European countries, including Switzerland. An earlier study examining the cause of abortions of sheep and goats in this country indicated that Cd. abortus is the leading cause of infectious abortions, inducing 39% of the abortions in sheep and 23% in goats (Chanton-Greutmann et al., 2002). Though direct evidence of the infectious agent is the ultimate diagnosis, sero-assays are more suitable for screening large numbers of samples. The complement-fixation test (CFT) is the most-widely used test and the one recommended by the Office International des Epizooties ( However, sensitivity of the CFT is 71% (Buendia et al., 2001). The specificity is even lower due to a common antigen shared by all members of the family Chlamydiaceae and also to some Gram-negative bacteria (Markey et al., 1993a; Donn et al., 1997; Jones et al., 1997; Rodolakis et al., 1998). Especially, Chlamydophila (Cd) pecorum (formerly Chlamydia pecorum) a widely distributed chlamydial agent in small ruminants causing diseases like the arthritis/conjunctivitis and pneumonia syndrome in lambs and also subclinical intestinal infections is responsible for false-positive reactions in the CFT (Fukushi and Hirai, 1992; Storz and Kaltenböck, 1993). To overcome these shortcomings, further tests have been developed. The indirect immunofluorescence test and immunoblotting were much more sensitive and specific but too laborious and expensive for testing large numbers of sera (Wang, 1971; Jones et al., 1997). Several ELISAs based on different antigens of Cd. abortus have been developed (Anderson et al., 1995; Donn et al., 1997; Kaltenböck et al., 1997; Salti-Montesanto et al., 1997; De Sa et al., 1999; Buendia et al., 2001; Longbottom et al., 2001). In 1997, Salti-Montesanto et al. described a competitive ELISA (celisa) that was based on Cd. abortus-specific 12

17 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations N. Borel et al. / Preventive Veterinary Medicine 65 (2004) monoclonal antibodies directed against the variable segments (VS) 1 and 2 of the major outer-membrane protein (MOMP). This test could distinguish between Cd. abortus and Cd. pecorum infections and its individual-animal sensitivity (Se) and specificity (Sp) was superior to the CFT. On this basis, those authors concluded that the test was acceptable both for routine diagnostics and for screening large numbers of samples (Salti-Montesanto et al., 1997). Using the same celisa, Gut-Zangger et al. (1999) confirmed these observations. In addition, these authors compared tests of pooled sera with those of individual-animal samples and suggested that pooled samples of a flock might provide a more efficient and less laborious herd-level assessment (Gut-Zangger et al., 1999; Borel et al., 2002). Our objectives were: (a) to evaluate the suitability of the celisa for pooled sera (referred to as the pool approach ), (b) to estimate the herd-level seroprevalence of OEA in different regions (cantons) of Switzerland and (c) to estimate the canton-specific true prevalences of seropositive flocks using the Rogan Gladen estimator and Bayesian inference. 2. Materials and methods 2.1. Serum samples The Swiss sheep population is composed of 0.42 million sheep in 13,155 registered sheep flocks. The tested sera were collected in 1998 for a population survey on Brucella melitensis (Anon., 1999). For that study, the Swiss veterinary authorities randomly selected >2000 sheep flocks, and serum samples from all adult animals in those flocks were taken and stored in a sample bank at 80 8C. Hemolytic sera were excluded from the present study because hemolysis interferes with the celisa (Gut-Zangger et al., 1999) Competitive enzyme-linked immunosorbent assay (celisa) The assay using the monoclonal antibody mab 188 (directed against MOMP-VS1 of Cd. abortus) was performed according to Salti-Montesanto et al. (1997). Briefly, 96-well ELISA plates were coated with semi-purified elementary bodies (EB) of the strain ME/ 4004, treated with sodium periodate, blocked with non-fat dried milk, air-dried, and stored at 20 8C. Hybridoma 188 was propagated in a Mini-perm apparatus (Heraus, Osterode, Germany) according to the instructions of the manufacturer. A working dilution of this mab in phosphate-buffered saline (PBS), selected to give an optical density (OD) between 0.8 and 1.0, was mixed with the test serum and the mixture was incubated on the ELISA plates for 1 h at 37 8C. After several washes, the bound competitor was detected with peroxidase-conjugated anti-mouse immunoglobulin and 3,3 0,5,5 0 -tetramethylbenzidine (TMB) as the substrate. Optical density was measured with an ELISA reader, and the results were expressed as % of inhibition according to the previous report. Inhibition values >55% were positive for OEA infection (positive cut-off); inhibition between 30 55% was considered questionable, attributed to either Cd. abortus or Cd. pecorum (Salti-Montesanto et al., 1997). Salti-Montesanto et al. in that work concluded that when using these cut-offs, the Sp of the celisa was close to 100% and the Se was somewhat lower. Comparison of 13

18 Habilitationsschrift Nicole Simona Borel 208 N. Borel et al. / Preventive Veterinary Medicine 65 (2004) the celisa with a second ELISA and the complement-fixation test showed agreement of between 66% and 92% (Longbottom et al., 2001). Our preliminary Bayesian analysis of these comparison data provided estimates for the celisa individual-animal Se and Sp of 70% and 91%, respectively (analysis not shown) Pilot study for the validation of the pool approach Ninety-eight of the 102 flocks (1195 sheep sera) collected in the canton Zurich in 1998 were tested. Four flocks were excluded from the pilot study because they consisted only of rams (one flock), or they had been vaccinated against OEA (one flock) or all sera from the flock were haemolytic (two flocks). One or more pools containing 10 sera were prepared for each flock (depending on flock size). This pooling process was based on a small pilot study showing that positive individual-animal sera, diluted 1/10 in negative serum, were still positive by the celisa (Gut-Zangger et al., 1999). Each pool of the 98 flocks, as well as all 1195 individual-animal sera, was tested in the celisa as described above. For the flock-level validation, a flock was classified as gold-standard positive if at least one individual serum sample from that flock had >55% inhibition, and gold-standard negative otherwise. The single pool per flock with the highest inhibition value was used in a receiver-operator characteristic (ROC) analysis to derive a flock-level test classification (Greiner et al., 2000). Within this analysis, the optimal cut-off (% inhibition) for a given pool to achieve maximum flock-level sensitivity and specificity of the pool testing (FPSe; FPSp) when compared to herd classification based on individual-animal testing (used here as the gold-standard) was identified. As a descriptive measure of the ROC approach, the area under the curve (AUC) (that is maximum at 1.0 when both Se and Sp are 100%), was calculated Population survey and prevalence estimation Target sample sizes for all cantons originally were derived after grouping them into six distinct geographic regions. For each region, a required sample size was calculated assuming the Zurich prevalence of positive herds in the respective sheep flock population, 5% accepted absolute error, and 95% level of confidence (Cannon and Roe, 1982). After calculating the sample sizes for each region, the sample sizes for each canton were derived proportional to the respective sheep population size of each canton within the selected region. Due to limited resources, however, it subsequently was decided to include only the central-valley (midland) cantons Appenzell-Ausserrhoden and Appenzell-Innerrhoden (combined), Aargau, Bern, Fribourg, St. Gallen and Thurgau and the larger Alpine southeastern cantons Graubünden, Ticino, Valais (i.e. those with large sheep populations) in the study. The originally calculated cantonal sample sizes based on the six regions were retained for these nine cantons. In those instances, when fewer flocks were represented in the 1998 serum sample bank than calculated for a canton, all represented flocks from that canton were included in the study. Otherwise, all available flocks per canton were assigned random numbers between 0 and 1 using the random-number function within MS-Excel 1, ordered by these numbers and selected from top down until the desired number of flocks had been reached. 14

19 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations N. Borel et al. / Preventive Veterinary Medicine 65 (2004) As for the pilot study, all available sera from the selected flocks were distributed into one or more pools of approximately equal size, with a maximum of 10 sera per pool. The pools were tested by the celisa and their percentage inhibition values determined. Flocks were classified test positive when the percentage inhibition of at least one pool was above the cut-off, and when subsequently at least one individual sample from such a pool was classified as positive (>55% inhibition). Flock-level true prevalences (FTP) for each canton were derived from the flock-level apparent (test) prevalences (FAP) using the Rogan Gladen estimator and information about the flock-level sensitivity (FPSe) and specificity (FPSp): FTP RG ¼ FAP þ FPSp 1 FPSe þ FPSp 1 (1) The calculations per region were done with the following distributions for FAP, FPSe and FPSp: FAP Betaðd c þ 1; n c d c þ 1Þ; FPSe Betaðd þ 1; n d þ 1Þ; FPSp Betaðd þ 1; n d þ 1Þ where d is the number of desired (positive or negative) outcomes, n the number of samples tested per region (Vose, 1996), and the subscript c indicates a canton-specific value. The MS-Excel Professional 4.0 with 10,000 iterations was used to derive medians and 2.5 and 97.5 percentiles (95% prediction interval) for the Rogan Gladen FTP estimate for each canton. Lower prediction limits (2.5 percentiles) were truncated at zero (0) whenever negative. In a second approach, a Bayesian model was used to derive posterior Bayesian estimates (denoted FTP B, FPSE B and FPSP B ) from prior distributions and the data from the 10 (11) cantons in the study (the WinBugs code is in Appendix A). The same prior beta distributions for FPSe and FPSp as in the Rogan Gladen approach were used. The Markov chain Monte Carlo (MCMC) simulation was run for 100,000 iterations of which the first 20,000 iterations were discarded (burn-in phase). Medians and 2.5 and 97.5 percentiles of the posterior distributions (Bayesian prediction interval) where recorded for the cantonspecific FTP B estimates and for posterior estimates of the test characteristics, FPSe B and FPSp B Statistical analysis Sample-size calculations were done in WinEpiscope 2.0 ( Descriptive statistics and ROC curves were calculated using NCSS 2001 ( For estimation of true flock-level prevalence with the Rogan Gladen Professional 4.0 ( was used. The Bayesian model was run in WinBugs 1.4 ( The source code for WinBugs to estimate the herd-level prevalence with one test in multiple (sub-) populations was made available in 2003 by Adam Branscum et al. (see diagnostictests/abherdlevelprev.htm). 15

20 Habilitationsschrift Nicole Simona Borel 210 N. Borel et al. / Preventive Veterinary Medicine 65 (2004) Results 3.1. Flock-level pool approach A total of 144 pools (consisting of 10 individual-animal sera each) from the pilot canton (Zurich) representing 98 flocks and 1195 sheep were examined with the celisa. All individual-animal samples in the pools also were tested with that assay. Fourteen flocks were classified as gold-standard positive, because they had at least one sheep classified as positive. In this pilot study, no flocks were considered as questionable based on the individual-animal results. Based on the ROC analysis and practical considerations, a pool cut-off value of 30% inhibition was chosen. At this cut-off value, the FPSe was estimated to be 92.9% ( %), and the estimate for the FPSp was 97.6% ( %). The total cumulative area under the ROC curve (AUC) was 94.9%. These values for FPSp and the AUC were only marginally lower than the maximal values to be achieved with a cut-off value between 31.3% and 32.7% inhibition (same FPSe, FPSp = 98.8% and AUC = 96%) Population survey The selected 11 cantons in 1998 contained 9411 (72%) of the 13,155 registered Swiss sheep flocks and 76% of the 0.42 million registered sheep. With the exception of the central as well as some western regions, the Swiss sheep population was well represented in the study. Overall, 144 (18.6%) of the examined 775 sheep flocks were classified as testpositive for Cd. abortus-specific antibodies. The overall FTP RG was estimated to be 18.1% ( %). With the exception of one canton, Bayesian FTP estimates were consistently lower than the Rogan Gladen estimates. However, all Rogan Gladen point estimates were well within the Bayesian prediction intervals (Table 1). Within positive flocks, the median percentage of seropositive animals was 33.7% (1 100%). The Bayesian posterior sampling means and 95% prediction intervals for the FPSe and FPSp, estimated from the survey, were 87.9% (65.3; 98.1%) and 96.0% (91.0; 99.0%), respectively. 4. Discussion This study represents the first large-scale seroepidemiological investigation on Cd. abortus in Swiss sheep flocks and we found that many sheep flocks in which animals had seroconverted after contact to the agent. Previous Swiss studies, performed between 1996 and 1998, had focussed on abortion cases of small ruminants that were submitted to the laboratories for further examination. In those studies, Cd. abortus was the most-commonly involved infectious agent in sheep and goat abortion cases (39% and 23%, respectively). The next-most-common agents were Toxoplasma gondii (sheep 19%, goats 15%) and Coxiella burnetii (sheep 1%, goats 10%) (Chanton-Greutmann et al., 2002). The included cantons hold over 70% of all Swiss registered sheep flocks and sheep. They represent the most important sheep-producing regions and different management systems, and included the Alpine mountainous (south-eastern part) as well as the 16

21 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Table 1 Seroprevalences for Chlamydophila (Cd) abortus (Chlamydia psittaci serotype 1) in sheep flocks from 10 of the 26 Swiss regions (cantons) Swiss canton Sheep in population Flocks in population Target flock sample Flocks tested Positive flocks Apparent flock-level prevalence (%) True flock-level prevalence (95% CI) Rogan Gladen (%) Bayesian (%) Bern (BE) (0 8.9) 1.9 (0 8.7) Appenzell (AI/AR) (0 22) 4.6 (0 19) Thurgau (TG) (0 28) 7.6 (0 25) Valais (VS) (1.5 17) 7.8 (0 16) Aargau (AG) (0.7 30) 9.5 (0 27) Fribourg (FR) (1.2 33) 10 (0 29) Zürich (ZH) (4.6 24) 12 (0.8 23) St. Gallen (SG) (7.4 32) 16 (3.4 30) Ticino (TI) (17 56) 29 (13 52) Graubünden (GR) (34 61) 43 (33 61) Total 322, (12 25) n/a Population data and serum samples represent the situation in True flock-level prevalences were estimated using the Rogan Gladen estimator as well as by Bayesianmodeling with distributions for flock sensitivity Beta(14, 2) and flock specificity Beta(83, 3). N. Borel et al. / Preventive Veterinary Medicine 65 (2004)

22 Habilitationsschrift Nicole Simona Borel 212 N. Borel et al. / Preventive Veterinary Medicine 65 (2004) more-hilly middle land (pastures in the north-western part). We thus consider it likely that the Cd. abortus seroprevalence in the cantons that were not sampled within this study is comparable to those of the surrounding cantons. However, the prevalence range observed for the cantons included in our study is rather wide. The highest seroprevalence estimates were derived for the cantons Graubünden (43%) and Ticino (29%). This, in connection with numerous abortion cases positive for Cd. abortus reported previously from Graubünden (Chanton-Greutmann et al., 2002) shows the importance of this agent in the etiology of sheep abortions in those areas. The importance of these findings was highlighted by two further occurrences in Graubünden: (i) the first Swiss case of human abortion caused by a zoonotic infection of a pregnant woman with Cd. abortus from a caprine abortion, and (ii) the first confirmed case of a bovine abortion due to Cd. abortus (Pospischil et al., 2002a,b). These observations certainly require further investigation. In previously reported sero-surveys, the less-sensitive and -specific CFT had been used. Positive CFT titers can be caused by Cd. abortus as well as by Cd. pecorum. Bostedt and Dedié (1996) reported antibodies to Chlamydiae (CFT) in 70% of sheep flocks (Germany). In Spain, almost 51% seroprevalence in sheep was found with the CFT (Mainar-Jaime et al., 1998). In these reports, however, no differentiation between Cd. abortus and Cd. pecorum antibodies was made. In surveys based on differentiating ELISA test systems, seroprevalences of 11% (Ireland; Markey et al., 1993b) and 32% (Tyrol, Austria; Khaschabi and Brandstätter, 1994) were reported. In our study, using the celisa, only antibodies against Cd. abortus were detected thus providing better specificity. Use of the celisa in combination with testing pooled sera in our opinion resulted in lower costs and was less time consuming when compared to other approaches. The pool approach initially provided information on the prevalence of OEA at flock-level. Information on individual-animal seroprevalence had to be investigated in a second step by testing all sera in the positive pools. Depending on the objective of a study, this second step could be considered unnecessary. However, defining flocks as positive on pool test results and the selected pool cut-off value could result in false-positive flock classifications, and would not provide the information on the within-flock prevalence. We therefore recommend in a second step examining individual animal samples from positive pools whenever feasible. If a chlamydial infection is present in a flock, then vaccination and/or the use of oxytetracycline often are considered for treatment (Buxton and Henderson, 1999). Currently, there are two chlamydial vaccines registered in Switzerland. One is an egggrown inactivated vaccine (Ovax Clamidia, Berna) and has been in use for several years. It is therefore possible that sera from vaccinated sheep were tested in the context of this survey. The second vaccine is an attenuated strain of Cd. abortus (Ovilis Enzovax, Intervet) and only recently has been registered in Switzerland (December 2002). Thus, sheep tested in our study could not have been vaccinated with this second vaccine before the time of sampling (1998). In a pilot study, we observed that antibodies against the inactivated vaccine (Ovax Clamidia) were not detected by the celisa therefore, we consider falsepositive results due to vaccination with this inactivated vaccine to be unlikely. In contrast, a parallel study with the attenuated vaccine (Ovilis Enzovax) indicated that this vaccine induced an antibody response that was detected by the celisa (unpublished data). As a 18

23 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations N. Borel et al. / Preventive Veterinary Medicine 65 (2004) result, false-positive results due to vaccination are possible, and the vaccination status of flocks under investigation needs to be taken into account when interpreting survey results. The celisa used in this study fulfilled the main criteria (i.e. simplicity, speed, low cost, no specific equipment needed, and relatively high sensitivity and high specificity when assessed at the flock-level) needed for screening large numbers of samples in epidemiological studies or surveillance programs. It nevertheless remains important to adjust the apparent prevalences for the imperfect test characteristics. We used both a modified (stochastic) Rogan Gladen approach and Bayesian inference. The approaches yielded comparable true-prevalence estimates, with those of the Bayesian model being slightly lower and having wider prediction intervals. The Rogan Gladen approach has the advantage that it is more-widely known and also can be used as a simple deterministic function (entering fixed values for AP, Se and Sp). One disadvantage is that that estimator (for certain combinations of AP, Se and Sp) can yield negative results. The Bayesian model approach is more complex but relatively easily can be implemented in the freely available software WinBugs. Its advantage is that, in addition to providing posterior distributions for the true prevalence, it also provides posterior distributions (estimates) for Se and Sp. However, knowledge and assumptions on the prior shape, value range and starting values of the model inputs are required. Acknowledgement This study was kindly supported by the Federal Veterinary Office (FVO), Bern (Switzerland). Appendix A WinBugs 1.4 program code to derive posterior MCMC distributions for true seroprevalence of Cd. abortus antibodies in Swiss sheep flocks and the sensitivity and specificity of the celisa from prior distributions and sampling data for 10 Swiss cantons. A template for this source code was provided by Wes Johnson, University of California, Davis, during a Bayesian-modeling course in November Abbreviations and comments ## loop for k = 10 cantons ## ap = test prevalence in sample ## tp = estimated true prevalence with beta(a, b) distribution ## Se = sensitivity of the test with prior distribution beta(14, 2) based on validation data ## Sp = specificity of the test with prior distribution beta(83, 3) based on validation data model {for (i in 1:k { y[i] dbin(ap[i], n[i]) 19

24 Habilitationsschrift Nicole Simona Borel 214 N. Borel et al. / Preventive Veterinary Medicine 65 (2004) ap[i] < tp[i] Se + (1 tp[i]) (1 Sp) tp[i] dbeta(alpha, beta) } alpha < mu gamma beta < (1 mu) gamma mu dbeta(1, 1) gamma dgamma(0.1, 0.1) Se dbeta(14, 2) Sp dbeta(83, 3) } ## end of model ## data for 10 cantons list(k = 10, n = c(113, 31, 29, 161, 32, 30, 98, 73, 35, 173), y = c(5, 2, 3, 17, 4, 4, 14, 13, 11, 71)) ## starting values list(tp = c(0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5), Se = 0.5, Sp = 0.5, mu = 0.5, gamma = 1) References Aitken, I.D., Ovine chlamydial abortion. In: Woldehiwet, Z., Ristic, M. (Eds.), Rickettsial and Chlamydial Diseases of Domestic Animals. Pergamon Press, Oxford, pp Aitken, I.D., Chlamydial abortion, Diseases of Sheep, 3rd ed. Blackwell Science, Oxford. Aitken, I.D., Clarkson, M.J., Linklater, K., Enzootic abortion of ewes. Vet. Rec. 10, Anderson, I.E., Herring, A.J., Jones, G.E., Low, J.C., Greig, A., Development and evaluation of an indirect ELISA to detect antibodies to abortion strains of Chlamydia psittaci in sheep sera. Vet. Microbiol. 43, Anon., Stichprobenuntersuchungen 1998 und In: Berichte zur Gesundheit von Mensch und, Tier. Bundesamt für Veterinärwesen, April 1999, pp Borel, N., Doherr, M.G., Vretou, E., Psarrou, E., Thoma, R., Pospischil, A., Ovine enzootic abortion: seroprevalence in Switzerland using a competitive enzyme-linked immunosorbent assay (celisa). Schweiz. Arch. Tierheilk. 144, (in German). Bostedt, H., Dedié, K., Chlamydien abort. Schafkrankheiten. Hrsg. H. Bostedt, Eugen-Ulmer-Verlag, Stuttgart. 20

25 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations N. Borel et al. / Preventive Veterinary Medicine 65 (2004) Buendia, A.J., Cuello, F., Del Rio, L., Gallego, M.C., Caro, M.R., Salinas, J., Field evaluation of a new commercially available ELISA based on recombinant antigen or diagnosing Chlamydophila abortus (Chlamydia psittaci serotype 1) infection. Vet. Microbiol. 78, Buxton, D., Potential danger to pregnant woman of Chlamydia psittaci from sheep. Vet. Rec. 118, Buxton, D., Henderson, D., Infectious abortion in sheep. In Practice 21, Cannon, R.M., Roe, R.T., Livestock Disease Surveys: A Field Manual for Veterinarians. Australian Bureau of Animal-health, Canberra, 35 pp. Chanton-Greutmann, H., Thoma, R., Corboz, L., Borel, N., Pospischil, A., Abortion in small ruminants in Switzerland: Investigations during two lambing seasons with special regard to Chlamydiae. Schweiz. Arch. Tierheilk. 144, (in German). De Sa, C., Souriau, A., Nussbaum, D., Salord, J., Rodolakis, A., Evaluation of a recombinant enzyme-linked immunosorbent assay for the serological of ovine chlamydial abortion. In: Proceedings of the IX International Symposium of the World Association of the Veterinary Laboratory Diagnostican. College Station, TX, p. 52. Donn, A., Jones, G.E., Ruiu, A., Ladu, M., Machell, J., Stancanelli, A., Serological diagnosis of chlamydial abortion in sheep and goats: comparison of the complement-fixation test and an enzyme-linked immunosorbent assay employing solubilised proteins as antigen. Vet. Microbiol. 59, Everett, K.D.E., Bush, R.M., Andersen, A.A., Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam.nov. and Simkaniaceae fam.nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int. J. Syst. Bacteriol. 49, Fukushi, H., Hirai, K., Proposal of Chlamydia pecorum sp. nov. for Chlamydia strains derived from ruminants. Int. J. Syst. Bacteriol. 42, Greiner, M., Pfeiffer, D., Smith, R.D., Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests. Prev. Vet. Med. 45, Gut-Zangger, P., Vretou, E., Psarrou, E., Pospischil, A., Thoma, R., Chlamydial abortion in sheep: possibilities of serological diagnosis with a competitive ELISA and insight into the epidemiological situation in Switzerland. Schweiz. Arch. Tierheilk. 141, (in German). Jones, G.E., Low, J.C., Machell, J., Armstrong, K., Comparison of five tests for the detection of antibodies against chlamydia (enzootic) abortion of ewes. Vet. Rec. 141, Jorgensen, D.M., Gestational psittacosis in a Montana sheep rancher. Emerg. Infect. Dis. 3, Kaltenböck, B., Heard, D., DeGraves, F.J., Schmeer, N., Use of syntethic antigens improves detection by enzyme-linked immunosorbent assay of antibodies against abortigenic Chlamydia psittaci in ruminants. J. Clin. Microbiol. 35, Khaschabi, D., Brandstätter, A., Seroepidemiologische Untersuchungen zum Nachweis von Antikörpern gegen Coxiella burnetii und Chlamydia psittaci bei Schafen in Tirol. Wien. Tierärztl. Mschr. 81, Longbottom, D., Psarrou, E., Livingstone, M., Vretou, E., Diagnosis of ovine enzootic abortion using an indirect ELISA (romp91b ielisa) based on a recombinant protein fragment of the polymorphic outer membrane protein POMP91B of Chlamydophila abortus. FEMS Microbiol. Lett. 195, Mainar-Jaime, R.C., de la Cruz, C., Vàzquez-Boland, J.A., Epidemiologic study of chlamydial infection in sheep farms in Madrid, Spain. Small Rumin. Res. 28, Markey, B.K., McNulty, M.S., Todd, D., 1993a. Comparison of serological tests for the diagnosis of Chlamydia psittaci infection of sheep. Vet. Microbiol. 36, Markey, B.K., McNulty, M.S., Burns, K., 1993b. Chlamydia psittaci infection in sheep in Northern Ireland. Vet. Rec. 132, 389. Pospischil, A., Thoma, R., Hilbe, M., Grest, P., Zimmermann, D., Gebbers, J.O., 2002a. Abortion in humans by Chlamydophila abortus (Chlamydia psittaci serovar 1). Schweiz. Arch. Tierheilk. 144, (in German). Pospischil, A., Thoma, R., von Bomhard, W., Reitt, K., Cantieni, J., Zimmermann, D., Polkinghorne, A., 2002b. Abortion in the cow by Chlamydia psittaci. Schweiz. Arch. Tierheilk. 144, (in German). Rodolakis, A., Salinas, J., Papp, J., Recent advances on ovine chlamydial abortion. Vet. Rec. 29,

26 Habilitationsschrift Nicole Simona Borel 216 N. Borel et al. / Preventive Veterinary Medicine 65 (2004) Salti-Montesanto, V., Tsoli, E., Papavassiliou, P., Psarrou, E., Markey, B.K., Jones, G.E., Vretou, E., Diagnosis of ovine enzootic abortion, using a competitive ELISA based on monoclonal antibodies against variable segments 1 and 2 of the major outer membrane protein of Chlamydia psittaci serotype 1. Am. Res. 58, Storz, J., Kaltenböck, B., Diversity of chlamydial-induced diseases. In: Woldehiwet, Z., Ristic, M. (Eds.), Rickettsial and Chlamydial Diseases of Domestic Animals. Pergamon Press, Oxford, pp Wang, S.P., A micro immunofluorescence method. Study of antibody response to TRIC organisms in mice. Excerpta Medica Vose, D., Quantitative Risk Analysis A Guide to Monte Carlo Simulation Modelling. Wiley, New York. 22

27 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 23

28 Habilitationsschrift Nicole Simona Borel 24

29 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 25

30 Habilitationsschrift Nicole Simona Borel 26

31 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 27

32 Habilitationsschrift Nicole Simona Borel 28

33 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Veterinary Research BioMed Central Research article Ovine Enzootic Abortion (OEA): a comparison of antibody responses in vaccinated and naturally-infected swiss sheep over a two year period Andrea Gerber 1, Ruedi Thoma 2, Evangelia Vretou 3, Evgenia Psarrou 3, Carmen Kaiser 1, Marcus G Doherr 4, Dieter R Zimmermann 5, Adam Polkinghorne 1, Andreas Pospischil 1 and Nicole Borel* 1 Open Access Address: 1 Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Switzerland, 2 Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland, 3 Department of Microbiology, Hellenic Pasteur Institute, Athens, Greece, 4 Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Berne, Switzerland and 5 Institute of Clinical Pathology, University Hospital, Zurich, Switzerland Andrea Gerber - gerber_andrea@access.unizh.ch; Ruedi Thoma - Rudolf.Thoma@alt.gr.ch; Evangelia Vretou-vretou@mail.pasteur.gr; Evgenia Psarrou - eugeniaps@mail.pasteur.gr; Carmen Kaiser - C.Kaiser@access.uzh.ch; Marcus G Doherr - marcus.doherr@itn.unibe.ch; Dieter R Zimmermann - dieter.zimmermann@usz.ch; Adam Polkinghorne - adampolkinghorne@hotmail.com; Andreas Pospischil - apos@vetpath.uzh.ch; Nicole Borel* - n.borel@access.uzh.ch * Corresponding author Published: 28 September 2007 BMC Veterinary Research 2007, 3:24 doi: / This article is available from: Received: 5 June 2007 Accepted: 28 September Gerber et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: Prevention and control of ovine enzootic abortion (OEA) can be achieved by application of a live vaccine. In this study, five sheep flocks with different vaccination and infection status were serologically tested using a competitive enzyme-linked immunosorbent assay (celisa) specific for Chlamydophila (Cp.) abortus over a two-year time period. Results: Sheep in Flock A with recent OEA history had high antibody values after vaccination similar to Flock C with natural Cp. abortus infections. In contrast, OEA serology negative sheep (Flock E) showed individual animal-specific immunoreactions after vaccination. Antibody levels of vaccinated ewes in Flock B ranged from negative to positive two and three years after vaccination, respectively. Positive antibody values in the negative control Flock D (without OEA or vaccination) are probably due to asymptomatic intestinal infections with Cp. abortus. Excretion of the attenuated strain of Cp. abortus used in the live vaccine through the eye was not observed in vaccinated animals of Flock E. Conclusion: The findings of our study indicate that, using serology, no distinction can be made between vaccinated and naturally infected sheep. As a result, confirmation of a negative OEA status in vaccinated animals by serology cannot be determined. Background Chlamydophila abortus (formerly Chlamydia psittaci serotype 1) is the most common infectious bacteria causing abortion in small ruminants in Switzerland with a previous study demonstrating that 39% of the examined abortions in sheep and 23% in goats were caused by this agent [1]. In the Swiss canton of Graubünden, a mountainous region in the countries' east, the economic losses associ- Page 1 of 8 (page number not for citation purposes) 29

34 Habilitationsschrift Nicole Simona Borel BMC Veterinary Research 2007, 3:24 ated with ovine enzootic abortion (OEA) are significantly higher than in other cantons [2]. Cp. abortus is generally introduced into immunologically naïve flocks by a latently infected animal with the agent being subsequently transmitted from aborting ewes via shedding of large amounts of infectious Chlamydia in the foetal membranes and in vaginal discharges [3]. In newly infected flocks, up to 30% of ewes may abort in the last trimester of gestation or give birth to weak or dead lambs. After abortion, ewes in these flocks may develop a protective immunity. Subsequent yearly losses in endemically infected flocks may decrease to a lower level (eg. 5 10%) with sheep either born into the flock or newly introduced animals likely to suffer abortions during their initial pregnancies [4,5]. Prevention and control of OEA is achieved by vaccination and/or treatment with oxytetracyclines [6]. Two vaccines against chlamydial abortion are licensed in Switzerland by the Federal Veterinary Office (FVO) in Berne. The first of these available was an egg-grown, formalin-inactivated, whole-organism vaccine (Ovax Clamidia, Fatro, Italy) which reduces the incidence of abortion in vaccinated herds but not completely [7-10]. Since December 2002, an avirulent, temperature-sensitive, live chlamydia vaccine (Ovilis Enzovax, Intervet, The Netherlands), which is marketed to induce strong long-lasting protection, has been made commercially available in Switzerland. The attenuated strain 1B, which forms the basis of this vaccine, was obtained from the virulent Cp. abortus strain AB7 by nitrosoguanidine mutagenesis [11-13]. In 2005, a small pilot study was undertaken to determine if administration of vaccines to protect sheep flocks from OEA would result in antibody levels in the complementfixation test (CFT) and in the competitive enzyme-linked immunosorbent assay (celisa) tests similar to those following natural infection [14]. After vaccination with the inactivated vaccine (Ovax Clamidia) only one sheep developed a detectable antibody response. In contrast, vaccination with the attenuated live vaccine (Ovilis Enzovax) resulted in detectable antibody titers in all tested sheep. The aim of this study is to investigate a larger number of sheep over a two-year period in the field to compare flocklevel ELISA responses between (a) vaccinated (live vaccine), (b) naturally infected and (c) non-infected sheep flocks. It was anticipated that the follow up study of the humoral responses could possibly discriminate between vaccinated and naturally OEA-infected sheep. An additional objective of the study was to attempt to detect chlamydiae and/or the attenuated strain of Cp. abortus used in the live vaccine in conjunctival swabs of sheep. Results Serological results and abortion cases celisa classifications (frequency and proportion positive), median titer and respective range of positive classified sheep in flocks A, B, C, D and E over the four different investigation dates are shown in Table 1. The comparison between vaccinated and non-vaccinated animals in Flock B and E is shown in Table 2. Figure 1 shows the titer ranges (box plots) of all examined sheep in the five flocks over the four investigation dates. All ewes (n = 15) of Flock A were serologically positive after vaccination showing a high median antibody value of 91.7%. The median antibody level of positive sheep (n = 13) decreased marginally to 86.6% in autumn In spring 2006 and autumn 2006, the seroprevalence in the flock was 73% (n = 11). The median antibody value of the positive sheep was 81.3% (spring 2006) and 82.3% (autumn 2006). % Inhibition % Inhibition % Inhibition A C D A1v A2v A3v A4v C1 C2 C3 C4 D1 D2 D3 D4 % Inhibition Box over Figure plots the 1four of celisa investigation antibody dates values of all examined sheep Box plots of celisa antibody values of all examined sheep over the four investigation dates. Some or all animals in flocks A, B and E were vaccinated at given times (gray boxes). % Inhibition B E B1 B1v B2 B2v B3 B3v B4 B4v Vaccination E1 E1v E2 E2v E3 E3v E4 E4v Symbol legends Box plot with 25th, 50th and 75th percentiles, and whiskers representing ~ 95% of the range Moderate and severe outliers A1v Flock A, examination 1, vaccinated Page 2 of 8 (page number not for citation purposes) 30

35 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Veterinary Research 2007, 3:24 Table 1: Serological results A, B, C, D and E. celisa positive (above cutoff) sheep with frequency, respective proportion (%), median titers and titer range. Flock (n) Parameter Spring 2005 A No. positive 15 sheep Prop. Pos. (%) Median titer Titer range B No. positive 26 sheep Prop. Pos. (%) Median titer Titer range Autumn 2005 Spring Autum C No positive 17 sheep Prop. Pos (%) 1 Median titer Titer range D No. positive 63 sheep Prop. Pos. (%) Median titer Titer range E No. positive 63 sheep Prop. Pos. (%) Median titer Titer range significant difference in % positive (Fishers Exact Test, p = 0.024) In spring 2005, two years after the first vaccination, six out of 14 vaccinated sheep in Flock B had a positive serological result (median antibody value 62.9%), whereas two out of 12 non-vaccinated sheep in the same flock were positive. The number of positive sheep decreased to three and two in the vaccinated group (n = 14) in autumn 2005 and spring 2006, respectively. In the non-vaccinated group, one sheep tested positive in autumn 2005, but none in spring In autumn 2006, the number of positive sheep increased in the vaccinated (n = 6) and nonvaccinated (n = 2) group, although abortions were not reported. The mean antibody values in the two groups were comparable, both being slightly greater than 60%. Flock C (naturally infected flock) served as the positive control. The seroprevalence in sheep in spring 2005 was high at 82% (n = 14). The median antibody value in the Table 2: Serological results vaccinated vs. non-vaccinated (Flock B and E). Comparison of celisa positive (above cutoff) vaccinated and naturally exposed sheep with frequency, respective proportion (%), median titers and titer range. Flock (n) Parameter Spring 2005 Autumn 2005 Spring 2006 Autum 2006 B 1 No positive 14 sheep Prop. Pos (%) Median titer Titer range B 2 No positive 12 sheep Prop. Pos (%) Median titer Titer range E 1 No positive 13 sheep Prop. Pos (%) Median titer Titer range E 2 No positive 50 sheep Prop. Pos (%) Median titer Titer range vaccinated group 2 non-vaccinated group positive group was 82.9%. The seroprevalence remained continuously high (76% 88%) during the whole study period and median antibody values in positive sheep were above 70%. In autumn 2005 newborn lambs were largely negative and had a significantly lower median antibody value than older ewes (Kruskal Wallis test, p < 0.05) (data not shown). The ewe with the confirmed chlamydial abortion in spring 2005 had positive antibody levels for the remaining sampling period comparable to the other animals in the flock (50.5% 77%). The seroprevalence in goats after confirmed chlamydial abortion in all four animals in spring 2005 was 100% (n = 4) with a high median antibody value of 91.6% (data not shown). In contrast to the sheep, all goats remained serologically positive with very high antibody values (71.2% 97.5%) over the whole testing period (data not shown). Flock D served as the negative control for this study. Despite this, 21% (n = 13) of the ewes showed positive results in spring 2005, whereas 44% (n = 28) of the ewes had negative serological results and 35% (n = 22) of animals showed questionable readings. The median antibody values of the positive animals were 69.5%. Half a year later, in autumn 2005, 21 animals continued to be Page 3 of 8 (page number not for citation purposes) 31

36 Habilitationsschrift Nicole Simona Borel BMC Veterinary Research 2007, 3:24 serologically positive. In spring 2006, the seroprevalence increased to 46%, whereas the mean antibody values of the positive animals were comparable to spring and autum 2005 (around 69%). In autumn 2006, the number of serologically positive ewes decreased to 30% (n = 19), whereas the mean antibody value of positive sheep increased to 74.3%. Prior to vaccination in spring 2005, only one animal in Flock E was positive in the vaccination group (antibody value 61.4%), whereas 25 sheep (50%) were positive in the non-vaccinated group (n = 50). All 13 sheep of the vaccinated group were serologically negative in autumn 2005 and therefore selected for vaccination in winter The non-vaccinated group showed seroprevalences between 38 48% from autumn 2005 to 2006 and median antibody values of positive animals were consistently between %. In comparison to vaccinated sheep in Flock A, none of the animals vaccinated in winter 2005 were serologically positive in spring In autumn 2006, one ewe had a positive antibody value of 73.2%, whereas the other 12 vaccinated sheep had negative (n = 6) or questionable values (n = 6). Statistical comparison of mean titers In flocks A (all animals vaccinated), C and D (no animals in both flocks vaccinated, Figure 1), differences in titer values between the sampling periods were always highly significant in the RM ANOVA model (p < 0.01). In Flock B, with a vaccination date between sampling periods 2 and 3, both vaccination status and an interaction term between vaccination and visit were statistically significant (p < 0.05). In Flock E, in which vaccination took place before the first sampling, both main effects were significant (time: p < 0.05; vaccination status: p < 0.01), while the interaction term was not. PCR results of eye swabs In Flock E, 118 conjunctival swabs were collected before application of the live vaccine in autumn No obvious signs of ocular surface diseases such as conjunctivitis and keratitis were observed in any animal. IGS-S PCR screening detected 22 samples that were positive for chlamydial DNA. Sequencing of these PCR products identified 18 samples that shared greater than 98% sequence similarity to Cp. abortus [GenBank: CR ]. One sample each was revealed to be positive for Cp. pecorum [GenBank: CPU68434] and Cp. felis [GenBank: AP ]. The identity of two samples could not be determined. Five months after vaccination, in spring 2006, 118 eye swabs were sampled in the same flock. 12 samples were tested positive by the IGS-S PCR but all were from nonvaccinated ewes. Of these samples, 5/12 were positive for Cp. abortus [GenBank: CR ] while three were positive for Cp. pecorum [GenBank: CPU68434]. The identity of four samples could not be determined. None of the vaccinated sheep showed a positive IGS-S PCR result and it was concluded that no excretion of the vaccine strain had occurred. Discussion This study represents the first longterm chlamydial serological study comparing vaccinated and non-vaccinated flocks in Switzerland. The investigations were undertaken in the canton Graubünden, where numerous chlamydial abortion cases in sheep were previously reported [1] and the highest seroprevalence (43%) for Cp. abortus in Swiss cantons was observed [2]. The results obtained from this study confirm the previous observations of the pilot study [14] that serology (celisa) cannot be used to distinguish between sheep vaccinated with the live attenuated vaccine and naturally-infected sheep. The antibody value range in the recently vaccinated Flock A was comparable to Flock C in which acute infections of Cp. abortus occured at the same time. In Flock A, very high antibody levels (around 90%) were visible in every vaccinated sheep (n = 15), whereas antibody levels of sheep in the previous pilot study were somewhat lower (around 60%) 21 days post vaccination [14]. As chlamydial abortion was reported in Flock A in the past, sheep could have been already serologically positive before vaccination and the very high antibody levels could represent an overlay of both abortion and vaccine-associated antibody values. The mean antibody value of positive animals decreased in both flocks (A and C) from spring 2005 to spring A chlamydial abortion was diagnosed in one goat from Flock C in autumn 2006 explaining the increasing seroprevalence and antibody value in this group of animals at that time. The antibody values in the goats of Flock C after an acute infection with Cp. abortus were higher and persisted at a very high level (80 to 90%) over the observation period compared to the situation in sheep. No correlation with protection was seen however, as a chlamydial abortion occurred in a seropositive goat which had previously aborted. This observation was also made in other goat flocks in canton Graubünden (R. Thoma, personal communication). Goats treated with the live vaccine also aborted. In general, it is known that if Chlamydiae are introduced in a naive flock, the losses are much higher in goats (60%) than in sheep (30%). The differences between goats and sheep are consistent with previous records and to date remain unexplained [15,16]. Antibody levels of vaccinated ewes of Flock B ranged from negative to positive two and three years after vaccination, respectively. Questionable antibody levels are either attributed to undiagnosed Cp. pecorum infections [17] or Page 4 of 8 (page number not for citation purposes) 32

37 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Veterinary Research 2007, 3:24 are possibly due to the vaccination in spring In a similar situation to the naturally infected sheep (Flock C), a slow decrease of antibody values was observed over the sampling period. This observation strongly suggests that serology (celisa) cannot be used to distinguish between sheep vaccinated with the live attenuated vaccine and naturally-infected sheep as anticipated in the previous pilot study [14]. As a direct consequence to this, the confirmation of negative OEA status in vaccinated animals by serology cannot be made. This is unfortunate as reliable confirmation is important if an abatement of OEA through assembly of OEA-free flocks is to be performed as undertaken by the Sheep and Goat Health Schemes in England and Wales and the Premium Health Scheme in Scotland. Positive antibody values have been observed in the negative control flock (Flock D), which had not been vaccinated and was free from chlamydial abortion. An explanation for the observations of an increasing antibody value amongst this flock is that the animals may have asymptomatic intestinal infections with Cp. abortus as presumed in previous studies [17,19]. An alternative scenario is that the ewes were infected with a less virulent strain of Cp. abortus, which provokes seroconversion but no abortion [17,20]. Fluctuations in the antibody levels could be the result of bacterial shedding during oestrus which provokes an induction of antibody levels without causing abortion [21,22]. Unfortunately, little is still known at this time about the ability of Cp. abortus to persist in animals (and the anatomical location of this persistent infection) compared to other chlamydial species, which require more investigations. In Flock E, the serological reaction of 13 selected vaccinated sheep and the 50 non-vaccinated sheep in the flock was evaluated. Surprisingly and in contrast to the observations in the previous pilot study [14] and in the two vaccinated flocks A and B, six of 13 vaccinated sheep of Flock E showed no seroconversion eight months after vaccination. Only one ewe had a positive serological result (73.2%), comparable to the vaccinated sheep of Flock A and the naturally OEA-infected sheep of Flock C. The remaining six ewes had questionable antibody levels. The primary difference between animals in flocks A and E was the high variability of antibody levels in vaccinated animals. These results suggest that individual immunoreactions between sheep can vary considerably. Sampling of conjunctival swabs from sheep in Flock E was performed to detect and compare the presence of chlamydial DNA before and after vaccination. Furthermore, a possible excretion of the vaccine through the eye could be screened with this approach. Although chlamydiae were frequently detected by PCR in conjunctival swabs of sheep, the attenuated strain of Cp. abortus used in the live vaccine was not detected in swabs collected from vaccinated sheep. The incidence of Cp. abortus and Cp. pecorum and even C. suis in clinically healthy non-vaccinated sheep was previously observed in a recent study [23]. The significance of this possible new mode of transmission for OEA needs further investigation. Conclusion The findings in our study strongly suggest that serology (celisa) cannot be used to distinguish between sheep vaccinated with the live attenuated vaccine and naturallyinfected sheep. The course of antibody levels, nevertheless, can vary between individual animals and flocks. Compared to sheep, goats displayed higher antibody levels, which persist over a longer time period but do not correlate with protection. The attenuated strain of Cp. abortus used in the live vaccine was not detected in eye swabs collected from vaccinated sheep. Methods Flock details Five different sheep flocks in the canton Graubünden were followed over a two-year period with four flock visits. These five flocks were available for the study in spring 2005 through an established collaboration with veterinary authorities in the canton Graubunden. Due to constant turnover in each flock (i.e. slaughtering of old or sick ewes, birth of lambs, introduction of new animals) the number of animals tested all four times was much lower than the number of individual sheep in the flock. Details on the five tested flocks (A, B, C, D and E) over the four investigation dates (spring 2005/06, autumn 2005/06) are provided in Table 3. Briefly, animals of Flock A were available for serological testing after vaccination of 15 sheep in spring History of chlamydial abortion in autumn 2004 was reported, but none of the 15 sheep in the study suffered an abortion during the examination period. Ewes (n = 14) of Flock B were vaccinated in spring 2003 with the live vaccine because of a chlamydial abortion outbreak in the vicinity of this flock. Before and after vaccination, no abortions due to Cp. abortus occurred and, as a result, the owner abandoned a vaccination booster two years later. Access to this flock was possible in spring Flock C had an average of 11 goats over the four investigation dates, of which four were available for repeated testing during the four sampling periods, but the results were not included in the overall statistical calculations of Flock C. Flock C had confirmed chlamydial abortions in autumn 2004 (unknown number of animals) and spring 2005 (one ewe and four goats). No further chlamydial abortions occurred in this flock after spring Animals suffering from abortions were tested four times during the study. Flock D represented the neg- Page 5 of 8 (page number not for citation purposes) 33

38 Habilitationsschrift Nicole Simona Borel BMC Veterinary Research 2007, 3:24 Table 3: Flock details Flock A B C D E Examinat ion dates spring & autumn 2005/ 2006 spring & autumn 2005/ 2006 spring & autumn 2005/ 2006 spring & autumn 2005/ 2006 spring & autumn 2005/ 2006 Average no. sheep 1 OEA = ovine enzootic abortion 2Average no. goats: 11 3Goats tested all 4 times: 4 Sheep tested all 4 times Flock history chlamydia l abortions in autumn chlamydia l abortion outbreak nearby in chlamydia l abortions (positive control) no abortions (negative control) chlamydia l abortions in the past OEA status 1 positive negative positive negative positive Vaccinati on with live vaccine 15 sheep (spring 2005) 14 sheep (spring 2003), no vaccinati on booster no ative control flock as no abortion or vaccination occured during the study period. Sheep in this flock spent summer together with a flock that had reports of chlamydial abortion in the past. Nevertheless, no abortions in Flock D were observed during that time. The ewes from Flock E (n = 63) suffered from chlamydial abortions for many years. The last confirmed case of chlamydial abortion was documented in spring In this flock, selected sheep (n = 13) that were negative by a celisa screen in autumn 2005 were vaccinated with the live vaccine according to the instructions of the manufacturer in winter 2005 and tested two times after vaccination (spring/autumn 2006). The serology of the non-vaccinated sheep in this flock (n = 50) was also followed. During the investigations, no abortion due to Cp. abortus was observed in this flock. Blood samples were collected from each flock during spring and autumn of 2005 and 2006 using Vacutainer tubes Becton Dickinson, Heidelberg, Germany). Four hours after collection, blood samples were centrifuged at 3000 g for 10 minutes and stored in Nunc CryoTubes (Nalge Nunc International, Roskilde, Denmark) at -20 C until further processing. no 13 sheep (winter 2005) celisa Serum samples were tested by the competitive enzymelinked immunosorbent assay (celisa) using the monoclonal antibody mab 188 directed against the variable segments 1 (VS1) and 2 (VS2) of the major outer membrane protein (MOMP) of Cp. abortus, according to the protocol of Salti-Montesanto et al. [17]. The results of the celisa were expressed as 'percentage of inhibition' corresponding to the antibody concentration in the sample. Inhibition values above 55 per cent were considered positive for infection with Cp. abortus (positive cut-off) whereas inhibition values between per cent were classified as questionable, attributable to either Cp. abortus or Cp. pecorum, a widely distributed chlamydial agent in small ruminants causing diseases such as arthritis/conjunctivitis and pneumonia syndrome in lambs and also subclinical intestinal infections [18,19]. Inhibition values below 30 per cent were assumed to be negative [17,24]. PCR of eye swabs Conjunctival swabs (Cytobrushes, Berdat Charles, Bourroux, Switzerland) were collected from Flock E before and after vaccination to investigate possible excretion of chlamydiae and/or the Cp. abortus vaccine strain through the eye. Before application of the vaccine, conjunctival swabs from every sheep in the flock (n = 118) were collected in autumn Five months following vaccination (spring 2006), the second conjunctival swab samples were taken from every sheep in the flock (n = 118). Cytobrushes were each placed in a 1.5-ml Eppendorf tube and stored at -80 C until further processing. DNA extraction from all swabs was performed as described previously [25] using a commercial DNA extraction kit (DNeasy Tissue Kit, Qiagen, Hombrechtikon, Switzerland). The conjunctival swabs were investigated for the presence of chlamydial DNA by a Chlamydiales-order specific PCR targeting the intergenic spacer region (IGS) between chlamydial 16S and the 23S rrna genes [26] and using primers cigs1f (5'-CAA GGT GAG GCT GAT GAC-3') and cigs2r (5'-TCG CCT KTC AAT GCC AAG-3'). PCR conditions are described elsewhere [26]. The identity of all positively tested IGS PCR products was determined by direct sequencing of the PCR product from both strands. Sequencing was performed with an ABI Prism 377 DNA sequencer (Applied Biosystems) or Applied Biosystems 3100 (Synergene Biotech). The obtained sequences were compared with the sequences available in GenBank using the BLAST server from the National Center for Biotechnology Information [27]. Investigation of abortion cases Abortion cases in the flocks were further investigated for the presence of chlamydiae by routine bacteriology and immunohistochemistry of the placenta and the fetal organs (lung, liver, kidney) as described elsewhere [28]. Statistical analysis Ewe ELISA antibody values were initially categorized into positive, questionable or negative as described previously Page 6 of 8 (page number not for citation purposes) 34

39 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Veterinary Research 2007, 3:24 [17,24]. For the analysis, questionable and negative results were both interpreted as negative. Whole flock response patterns over time were visualized using box plots. For those sheep that were tested all four times, the proportion of positive ewes at each time point was compared within each flock using a Fishers Exact Test with exact p-values. In addition, the mean titers of those sheep were compared using a repeated measures ANOVA with animal ID, time (within animal repetition factor), vaccination status (flocks B and E only), and the interaction between time and vaccination (again only for flocks B and E). Data were stored and handled in MS Excel, and analysed using the statistical software packages NCSS 2004 [29] and SPSS 14 [30]. The overall level of statistical significance was set to Competing interests The author(s) declares that there are no competing interests. Authors' contributions AG carried out the serum sampling and the serological investigations and drafted the manuscript. RT performed the investigation of the abortion cases and contacted the flock owners. EV and EP prepared the celisa plates. CK investigated the eye swabs by PCR. MGD performed the statistical analysis. DRZ performed the DNA sequencing. AP assisted in the drafting and editing of the manuscript. APOS and NB participated in the design and coordination of the study. All authors read and approved the final manuscript. Acknowledgements We are grateful to the laboratory staff of the Institute of Veterinary Pathology and of the Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland. We also thank the sheep and goat owners of the canton Graubünden and their attending veterinarians. This study was kindly supported by the Doerenkamp-Stiftung. References 1. Chanton-Greutmann H, Thoma R, Corboz L, Borel N, Pospischil A: Abortion in small ruminants in Switzerland: Investigations during two lambing seasons ( ) with special regard to chlamydial abortions. Schweiz Arch Tierheilkd 2002, 144: (in German). 2. Borel N, Doherr MG, Vretou E, Psarrou E, Thoma R, Pospischil A: Seroprevalences for ovine enzootic abortion in Switzerland. Prev Vet Med 2004, 65: Aitken ID: Ovine chlamydial abortion. In Rickettsial and Chlamydial Diseases of Domestic Animals Edited by: Woldehiwet Z, Ristic M. Oxford: Pergamon Press; 1993: Rodolakis A, Salinas J, Papp P: Recent advances on ovine chlamydial abortion. Vet Res 1998, 29: Aitken ID: Chlamydial abortion. In Diseases of Sheep 3rd edition. Oxford: Blackwell Science; 2000: Buxton D, Henderson D: Infectious abortion in sheep. In Pract 1999, 21: McEwen AD, Foggie A: Enzootic abortion of ewes. Comparative studies of different vaccines. Vet Rec 1954, 66: Aitken ID, Clarkson MJ, Linklater K: Enzootic abortion in ewes. Vet Rec 1990, 126: Jones GE, Jones KA, Machell J, Brebner J, Anderson IE, How S: Efficacy trials with tissue-culture-grown, inactivated vaccines against chlamydial abortion in sheep. Vaccine 1995, 13: Garcia de la Fuente JN, Gutierrez-Martin CB, Ortega N, Rodriguez- Ferri EF, del Rio ML, Gonzalez OR, Salinas J: Efficacy of different commercial and new inactivated vaccines against ovine enzootic abortion. Vet Microbiol 2004, 100: Rodolakis A: In vitro and in vivo properties of chemically induced temperature-sensitive mutants of Chlamydia psittaci var. ovis: screening in a murine model. Infect Immun 1983, 42: Rodolakis A, Souriau A: Response of ewes to temperature-sensitive mutants of Chlamydia psittaci (var. ovis) obtained by NTG mutagenesis. Ann Rech Vet 1983, 14: Rodolakis A, Bernhard F: Vaccination with temperature-sensitive mutant of Chlamydia psittaci against enzootic abortion of ewes. Vet Rec 1984, 114: Borel N, Sachse K, Rassbach A, Bruckner L, Vretou E, Psarrou E, Pospischil A: Ovine enzootic abortion (OEA): Antibody response in vaccinated sheep compared to naturally infected sheep. Vet Res Commun 2005, 29(Suppl 1): Stamp JT, McEwen AD, Watt JA, Nisbet DI: Enzootic abortion in ewes; transmission of the disease. Vet Rec 1950, 62: Dawson M, Zaghloul A, Wilsmore AJ: Ovine enzootic abortion: experimental studies of immune responses. Res Vet Sci 1986, 40: Salti-Montesanto V, Tsoli E, Papavassiliou P, Psarrou E, Markey BK, Jones GE, Vretou E: Diagnosis of ovine enzootic abortion, using a competitive ELISA based on monoclonal antibodies against variable segments 1 and 2 of the major outer membrane protein of Chlamydia psittaci serotype 1. Am J Vet Res 1997, 58: Fukushi H, Hirai K: Proposal of Chlamydia pecorum sp. nov. for Chlamydia strains derived from ruminants. Int J Syst Bacteriol 1992, 42: Jones GE: Chlamydia disease-more than just abortion. Vet J 1997, 153: Tsakos P, Siarkou V, Guscetti F, Chowdhury H, Papaioannou N, Vretou E, Papadopoulos O: Experimental infection of pregnant ewes with enteric and abortion-source Chlamydophila abortus. Vet Microbiol 2001, 82: Papp JR, Shewen PE, Gartly CJ: Abortion and subsequent excretion of chlamydiae from the reproductive tract of sheep during estrus. Infect Immun 1994, 62: Papp JR, Shewen PE: Pregnancy failure following vaginal infection of sheep with Chlamydia psittaci prior to breeding. Infect Immun 1996, 64: Becker A, Wohlgroth L, Vaughan L, Brugnera E, Zimmermann D, Kaps S, Spiess B, Pospischil A: Chlamydial conjunctivitis in domesticated animals: preliminary results in guinea pigs, pigs and sheep. In Proceedings of the Fifth Meeting of the European Society for Chlamydia Research: 1 4 September 2004 Edited by: Judith Deak. Budapest, Hungary; 2004: Gut-Zangger P, Vretou E, Psarrou E, Pospischil A, Thoma R: Chlamydia abortion in sheep: possibilities for serological diagnosis using a competitive ELISA and insight into the epidemiologic situation in Switzerland. Schweiz Arch Tierheilkd 1999, 141: (in German). 25. Vögtlin A, Fraefel C, Albini S, Leutenegger CM, Schraner E, Spiess B, Lutz H, Ackermann M: Quantification of feline herpesvirus 1 DNA in ocular fluid samples of clinically diseased cats by real-time TaqMan PCR. J Clin Microbiol 2002, 40: Teankum K, Pospischil A, Janett F, Bürgi E, Brugnera E, Hoelzle K, Polkinghorne A, Weilenmann R, Zimmermann DR, Borel N: Detection of chlamydiae in boar semen and genital tracts. Vet Microbiol 2006, 116: BLAST: Basic Local Alignment and Search Tool from the National Center for Biotechnology Information [ Borel N, Thoma R, Spaeni P, Weilenmann R, Teankum K, Brugnera E, Zimmermann DR, Vaughan L, Pospischil A: Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet Pathol 2006, 43: Page 7 of 8 (page number not for citation purposes) 35

40 Habilitationsschrift Nicole Simona Borel BMC Veterinary Research 2007, 3: NCSS, PASS & GESS: Statistics, Graphics, Power Analysis, Sample Size, & Microarray Analysis [ 30. SPSS, Data Mining, Statistical Analysis Software, Predictive Analysis, Predictive Analytics, Decision Support Systems [ Publish with BioMed Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours you keep the copyright Submit your manuscript here: BioMedcentral Page 8 of 8 (page number not for citation purposes) 36

41 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 3. Epidemiological Investigations in Ruminants 3.1. Introduction and Conclusion Ovine Enzootic Abortion (OEA), caused by Chlamydophila (Cp). abortus, is recognized as the major infectious abortion cause in sheep and goat in Switzerland: Chanton-Greutmann H, Thoma R, Corboz L, Borel N, Pospischil A: Aborte beim kleinen Wiederkäuer in der Schweiz: Untersuchungen während zwei Ablammperioden ( ) unter besonderer Beachtung des Chlamydienabortes. Schweiz Arch Tierheilkd. 2002;144, Abortion in small ruminants in Switzerland: Investigations during two lambing seasons with special regard to chlamydiae. Abortion cases of 144 goats und 86 sheep were investigated etiologically during 2 lambing seasons (1996/1997, 1997/1998). Macroscopic inspection of fetus and placenta was completed by histopathology and bacteriological isolation of agents. In addition, immunohistologically the following antigens were labeled in formalin-fixed and paraffin-embedded tissue sections: Toxoplasma gondii, Neospora caninum, Chlamydophila abortus (formerly Chlamydia psittaci serovar 1) and Border Disease Virus. From farms with abortions caused by Chlamydophila abortus specific data were recorded. In 75% of abortion cases in sheep and in 59% of cases in goats an etiologic diagnosis could be substantiated. Chlamydophila abortus is the most commonly involved agent in the etiology of caprine and ovine abortion (sheep 39%, goats 23%), followed by Toxoplasma gondii (sheep 19%, goats 15%) and Coxiella burnetii (sheep 1%, goats 10%). All other agents are of minor importance. An infectious cause of abortion based on histopathologic findings without isolation of agents was observed in sheep (10%) and goats (21%). Malformation occured in sheep (2%) and goats (3%) and lesions suggestive for Vitamin E/ Selenium deficiency were seen in goats only (2%). Transmission of the disease takes place during abortion or lambing, when extremely large numbers of chlamydiae are shed in the fluid discharges, in the placenta, and on the coats of the lambs; these are the main sources of environmental contamination and oral transmission to other animals, as well as to humans (Aitken, 1993). There is limited evidence to suggest that OEA can be transmitted venereally. The presence of chlamydiae in the genital organs of male animals (bulls) displaying epididymitis, orchitis and seminal vesiculitis was first described in the 1960s (Ball et al., 1964; Storz et al., 1968). Chlamydia was also later isolated from lesions in the genital tract of a boar (Sarma et al., 1983) and more recently, the presence of Chlamydia in semen of boars, bulls and men was confirmed by immunofluorescence (Veznik et al., 1996). Moreover, several studies have revealed that chlamydiae could survive in cryopreserved semen of bulls (Storz et al., 1968) and men (Sherman and Jordan, 1985). This finding may have significant repercussions to the common practice of artificial insemination (AI) worldwide. Prior to our investigations in bulls, rams and bucks, we evaluated suitable PCR assays for detection of Chlamydia in semen of boars: Teankum K, Pospischil A, Janett F, Bürgi E, Brugnera E, Hoelzle K, Polkinghorne A, Weilenmann R, Zimmermann DR, Borel N: Detection of chlamydiae in boar semen and genital organs. Vet Microbiol. 2006, 116: Chlamydiae cause abortion and reproductive disorders in sows. Although organisms can infect the male genital tract, little is known about the disease situation in boars. Hence, we examined the prevalence of chlamydial infection in semen and genital tracts of boars. Samples collected from Swiss boars (group A: n = 42), and boars from Germany (group B: n = 39) were examined by bacteriology, LPS-ELISA, immunohistochemistry (IHC) and polymerase chain reaction (PCR). The latter methodology involved use of three PCR assays including 16Sig rdna, IGS-S (intergenic spacer 16S/23S-Short) and IGS-L (intergenic spacer 16S/23S-Long) PCR for comparison methods. PCR sensitivity and the presence of potential PCR inhibitors were determined by spiking semen with Chlamydophila (Cp.) abortus DNA. Detection limits of the 16Sig and IGS-S PCR were 10 templates, while the IGS-L PCR was less sensitive (100 templates). Of 25 semen samples that were collected from group A, one semen sample was positive for Cp. psittaci and two were positive for Chlamydia-like organisms by 16Sig PCR. Screening of sera from Swiss boars revealed three animals with positive reactions in the LPS-ELISA, although we failed to detect chlamydiae within organs of these or sera-negative animals by IHC or IGS-S PCR. In group B, 10 ejaculates were positive for Chlamydia (C.) suis and two were positive for Chlamydialike organisms by 16S PCR. The identification of DNA from Chlamydia-like organisms in semen from both groups 37

42 Habilitationsschrift Nicole Simona Borel of boars was surprising and a role for these bacteria in reproductive diseases requires further assessment. In conclusion, the prevalence of chlamydial infection was low in group A animals indicating that venereal transmission may not be significant for Chlamydia-associated reproductive diseases in pigs, although rare cases may occur. After successful evaluation of PCR detecting chlamydiae on semen samples of boars, we evaluated the same PCR assays to investigate the prevalence of these organisms in ruminants: 3.2. Teankum K, Pospischil A, Janett F, Brugnera E, Hoelzle LE, Hoelzle K, Weilenmann R, Zimmermann DR, Gerber A, Polkinghorne A, Borel N: Prevalence of chlamydiae in semen and genital tracts of bulls, rams and bucks. Theriogenology. 2007, 67: The second part of the epidemiological investigations in ruminants focussed on bovine abortion: Abortion in bovines is of worldwide economic importance. In Switzerland, 14,000 to 28,000 bovine abortion cases are reported every year resulting in a loss of 22 to 45 million Swiss Francs per year (Hässig et al., 2000). Neospora caninum and Bovine Viral Diarrhea Virus (BVDV) were shown to be important abortigenic agents. However, most abortion cases (67.7%) remain of unknown etiology (Reitt et al., 2007). The first case of bovine abortion due to Cp. abortus in Switzerland was reported in the canton Graubünden (Pospischil et al., 2002b). Economic losses due to OEA in this canton are significantly higher than in the other cantons (43% seroprevalence in sheep for Cp. abortus, Borel et al., 2002; R. Thoma, personal communication). In our study, we adressed whether the occurence of Chlamydia-related abortion in cattle is comparable to the situation in small ruminants in the canton Graubünden: 3.3. Borel N, Thoma R, Spaeni P, Weilenmann R, Teankum K, Brugnera E, Zimmermann DR, Vaughan L, Pospischil A. Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet Pathol. 2006, 43: An outcome of this study was the detection of Chlamydia-like organisms by PCR. During the last decade, new Chlamydialike organisms have been discovered e.g. new species such as Waddlia chondrophila, Parachlamydia acanthamoebae and Simkania negevensis (Greub and Raoult, 2002). Waddlia chondrophila has been twice isolated from an aborted bovine fetus, and a growing evidence supports the role of Parachlamydia and Simkania as agents of pneumonia in humans (Dilbeck et al., 1990; Henning et al., 2002; Baud et al., 2007). First results of the study mentioned above describe Parachlamydia and Chlamydia-like isolates other than Waddlia chondrophila for the first time in the setting of bovine abortion (bovine abortion cases from canton Graubünden, Borel et al., 2006). In collaboration with the University of Lausanne (Dr. Gilbert Greub), we established an immunhistochemistry protocol using an antibody specifically directed against Parachlamydia. This study has been published recently: 3.4. Borel N, Ruhl S, Casson N, Kaiser C, Pospischil A, Greub G. Parachlamydia spp. and related Chlamydia-like organisms and bovine abortion. Emerg Infect Dis. 2007, 12: To date, diagnostic tools for these novel Chlamydia-like organisms were not available. Given this first evidence of a role for Parachlamydia in bovine abortion (Borel et al., 2007), we further studied these 235 bovine placenta samples, using recently developed TaqMan real-time PCR for Parachlamydia and Waddlia and performed immunohistochemistry on positive samples using specific polyclonal mouse antibodies directed against Parachlamydia and Waddlia to demonstrate the agent directly in the placenta: 3.5. Ruhl S, Casson N, Kaiser C, Thoma R, Pospischil A, Greub G, Borel N. Evidence for Parachlamydia spp. in bovine abortion, submitted to Veterinary Microbiology. 38

43 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Conclusion In contrast to the situation in small ruminants in the canton Graubünden, bovine abortion due to Cp. abortus seems not to play an important role. Bovine abortion due to placentitis of unknown etiology may be caused by novel chlamydiae, which remain unrecognized using routine diagnostic approaches. As Parachlamydia is involved in bronchitis and pneumonia in humans, caution should be taken when handling bovine abortion material because of the potential zoonotic risk. Overall, the significance of this finding as well as the host reservoir of the Chlamydia-like organisms remain an open question and need further investigation. 39

44 Habilitationsschrift Nicole Simona Borel 40

45 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Theriogenology 67 (2007) Prevalence of chlamydiae in semen and genital tracts of bulls, rams and bucks K. Teankum a,e, A. Pospischil a, F. Janett b, E. Brugnera a, L.E. Hoelzle c, K. Hoelzle c, R. Weilenmann a, D.R. Zimmermann d, A. Gerber a, A. Polkinghorne a, N. Borel a, * a Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zürich, Winterthurerstrasse 268, CH-8057 Zürich, Switzerland b Clinic of Reproduction, Vetsuisse Faculty, University of Zürich, Winterthurerstrasse 268, CH-8057 Zürich, Switzerland c Institute of Veterinary Bacteriology, Vetsuisse Faculty, University of Zürich, Winterthurerstrasse 268, CH-8057 Zürich, Switzerland d Institute of Clinical Pathology, University Hospital, Schmelzbergstrasse 12, CH-8091 Zürich, Switzerland e Department of Pathology, Faculty of Veterinary Science, Chulalongkorn University, Henri Dunant Road, Pathumwan, Bangkok, Thailand Received 28 April 2006; received in revised form 29 June 2006; accepted 25 July 2006 Abstract Chlamydiae infect male genital organs of ruminants. However, little is known about their prevalence. Hence, we investigated fresh and cryopreserved semen (bulls: n = 304; rams: n = 78; bucks: n = 44) by polymerase chain reaction (PCR), as well as genital organs (bulls: n = 13; rams: n = 10; bucks: n = 6) by immunohistochemistry (IHC) and PCR. Sera from bulls (n = 104) and small ruminants (n = 61) were tested by LPS and rmomp (recombinant major outer membrane protein) ELISA and competitive ELISA (celisa), respectively. Three PCR assays were compared in this study for detection of chlamydial DNA in semen: 16S rrna, IGS- S (intergenic spacer 16S/23S-short), and IGS-L (intergenic spacer 16S/23S-long) PCRs. PCR sensitivity and inhibitory effects were determined by spiking semen with Chlamydophila (Cp.) abortus DNA. In bull semen, detection limits of the 16S, IGS-S and IGS-L PCRs were 10, 10, 100 templates, respectively. However, PCR sensitivity was reduced in ram and buck semen suggesting the presence of potential PCR inhibitors. Of 304 bull semen samples, the 16S PCR revealed DNA of chlamydiae in 20 samples (6.6%), including Cp. abortus (n = 2), Cp. psittaci (n = 1), Chlamydia suis (n = 2), and Chlamydia-like organisms (n = 15). In rams, one semen sample was positive for Chlamydia-like organism. All investigated male genital organs were negative for Chlamydia. Serology revealed 47.1% (49/104) positive bulls by LPS ELISA. Of these, 30 samples were positive by rmomp ELISA, predominantly for Cp. pecorum. In small ruminants, celisa displayed 34.8% (16/46) and 60% (9/15) positivity for Cp. abortus in rams and bucks, respectively. There was no correlation between serology and PCR of semen. The presence of chlamydiae in semen suggests the possibility of venereal transmission, although risk may be low in Switzerland. # 2006 Elsevier Inc. All rights reserved. Keywords: Ruminants; Semen; Chlamydia; PCR; Prevalence 1. Introduction Chlamydial infections in ruminants cause a wide range of diseases including polyarthritis, conjunctivitis, * Corresponding author. Tel.: ; fax: address: n.borel@access.unizh.ch (N. Borel). pneumonia and abortion [1 4]. According to the reclassification [5], chlamydiae responsible for the diseases in ruminants are Chlamydophila (Cp.) abortus, Cp. pecorum (members of the family Chlamydiaceae), and Waddlia chondrophila (member of the family Waddliaceae). Of these, Cp. abortus is the leading infectious cause of abortion in sheep and goats [6,7]. However, abortion in cattle due to this species is less X/$ see front matter # 2006 Elsevier Inc. All rights reserved. doi: /j.theriogenology

46 Habilitationsschrift Nicole Simona Borel 304 K. Teankum et al. / Theriogenology 67 (2007) common [8,9]. Cp. pecorum also causes several diseases in ruminants such as enteritis, polyarthritis, and endometritis [4,8]. Recently, the novel species Waddlia chondrophila has been reported from cases of bovine abortion [10,11]. Natural infections of chlamydiae have been reported in bulls and rams since the 1960s [12,13]. Subsequently, Chlamydia was isolated from the testes, epididymides and semen of bulls with seminal vesiculitis [13]. The organism was also recognised as a cause of epididymitis in rams [14]. Organisms from experimental infections were successfully re-isolated from genital organs, confirming the capability of the organism to cause an infection of the male genital tract [14,15]. These observations are of great importance particularly for bovine industry as artificial insemination (AI) is widely used and especially since it was shown that Chlamydia can survive in cryopreserved semen [13]. The evidence of chlamydial infections in male genital tracts of ruminants indicates the possibility of venereal transmission. The significance of this event under natural conditions is, however, controversial and information concerning prevalence of chlamydial infection in male ruminants is scarce. The aim of the current study was to investigate the prevalence of chlamydial infection in semen and male genital tracts of bulls, rams, and bucks. Several studies have developed and evaluated PCR techniques to investigate the presence of chlamydial DNA in various clinical samples including semen samples [16 18]. Serological methods, in addition, have also been used to investigate the prevalence of chlamydial infection in ruminants [19 21]. In the current study, we used both serological techniques and several different PCR methods to investigate the prevalence of chlamydial infection in male ruminants. 2. Materials and methods 2.1. Specimens: semen, sera, and male genital organs Fresh semen and blood samples were collected from three animal species: bulls (n = 104, case no ) aged between 8 months and 7 years, rams (n = 46) aged between 8 months and 5 years, and bucks (n = 15) aged between 7 and 32 months. All animals were clinically sound, and used in natural breeding, and the semen samples were collected from farms in different regions of Switzerland. All fresh semen samples were aliquotted, and stored at 20 8C prior to testing. Blood samples were centrifuged (3000 g, 10 min) and sera were stored at 20 8C prior to testing. Cryopreserved semen (bulls: n = 200, case no ; rams: n = 32; bucks: n = 29) was obtained from different AI centres. Cryopreserved bull semen samples were derived from four AI centres (AI-1: n = 23; AI-2: n = 34; AI-3: n = 19; AI-4: n = 124), and cryopreserved semen of small ruminants were obtained from three AI centres (AI-3: n = 15; AI-4: n = 7; AI-5: n = 39). Male genital organs (bulls: n = 13; rams: n = 10; bucks: n = 6), aged between 4 months and 4 years, were collected from necropsy submissions at the Veterinary Pathology Institute, University of Zurich. Most animals were free from reproductive problems, except a ram with evidence of orchitis, a ram with epididymitis, two bulls suspected for epididymitis, and two bulls with evidence of hypogonadism. Three sections (proximal, middle, and distal parts) were collected from each side of the testis. Other parts of the genital organs including three sections of the epididymides (head, body, and tail), ampulla, seminal vesicles, bulbourethral glands, and four parts of the urethra (proximal, sigmoid flexure, distal and glans penis) were sampled. All organ samples were fixed in 4% formalin, embedded in paraffin and further processed for immunohistochemistry DNA extraction Extraction of DNA from semen samples was performed using a commercial DNA extraction kit (DNeasy TM Tissue kit, Qiagen, Hilden, Germany), according to the body fluid protocol, using 200 ml of bull semen and 50 ml of semen samples from rams and bucks. DNA was finally eluted in 200 ml of elution buffer (Qiagen). For genital organs, DNA was extracted from paraffin blocks using the protocol as previously described [22] Polymerase chain reaction (PCR) and DNA sequencing To detect chlamydial DNA in semen samples, three PCR assays were evaluated: (i) 16S rrna PCR: The primer pair 16S-IGF (5 0 - GATGAGGCATGCAAGTCGAACG-3 0 ) and 16S- IGR (5 0 -CCAGTGTTGGCGGTCAATCTCTC-3 0 ) [5] was designed to amplify 278-bp product of the 16S rrna gene specific for the order Chlamydiales. (ii) IGS-short (IGS-S) PCR: The primer pair cigs-1f (5 0 -CAAGGTGAGGCTGATGAC-3 0 ) and cigs-2r (5 0 -TCGCCTKTCAATGCCAAG-3 0 ) was designed to target a variable region of chlamydial DNA 42

47 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations K. Teankum et al. / Theriogenology 67 (2007) approximately 370-bp PCR product including 80-bp of the 16S rrna gene, 240-bp of the rrna intergenic spacer (IGS) region (depending on chlamydial strain), and 50-bp of the 23S rrna gene. (iii) IGS-long (IGS-L) PCR: The primer pair cigs-1f (5 0 -CAAGGTGAGGCTGATGAC-3 0 ) and IGS-1r (5 0 -AGTGGTCTCCCCAGATTC-3 0 ) were designed to amplify approximately 750-bp PCR product containing 80-bp of the 16S rrna gene, 240-bp of the IGS region (of variable length between chlamydial species), and 440-bp of the 23S rrna gene. The conditions of 16S, IGS-S and IGS-L PCRs are described elsewhere [9,22,23]. Negative and positive control for PCR reactions were performed as previously described [22]. All PCR reactions were carried out in a TGRADIENT thermal cycler (Biometra, Göttingen, Germany). PCR products (5 ml) from all PCR assays were separated by electrophoresis in 1.5% agarose gels and visualised with a UV transilluminator. PCR products were purified by a DNA purification kit (Qiagen), and directly sequenced at the sequencing service of the University of Zurich. The obtained sequences were compared with sequences available in GenBank using the BLAST server from the National Center for Biotechnology Information ( Sensitivity test of PCR To assess the sensitivity of PCR assays, semen samples were spiked with a known amount of Cp. abortus S26/3 genomic DNA. The genomic DNA of Cp. abortus, prepared as previously described [9,22] was 10-fold diluted in dh 2 O, and the diluted DNA was used in spiking experiments as follows: (i) Bull semen samples (200-ml aliquots of selected fresh and cryopreserved semen) were spiked with a known amount of Cp. abortus DNA to obtain the final concentration of 10 3, 10 2, 10, and 1 template/ ml of semen samples, respectively. (ii) Ram semen samples (40-ml aliquots of selected fresh and cryopreserved semen) were spiked with a known amount of Cp. abortus DNA to obtain the final concentration of 10 3, 10 2, 10, and 1 template/ ml of semen samples, respectively. (iii) Buck semen samples (40-ml aliquots of selected fresh and cryopreserved semen) were spiked with a known amount of Cp. abortus DNA to obtain the final concentration as in Experiment (ii). DNA was extracted from all spiked semen samples using a commercial DNA extraction kit (DNeasy TM Tissue kit, Qiagen), according to the body fluid protocol, and was finally eluted in 200 ml of elution buffer (Qiagen). The DNA extraction products from all dilution series were compared between all three PCR assays: 16S, IGS-S, and IGS-L PCRs Immunohistochemistry (IHC) Paraffin sections from all parts of the genital organs were cut and labelled for the presence of chlamydial antigen using a Chlamydiaceae family-specific mouse monoclonal antibody (Ab) directed against the chlamydial lipopolysaccharide (clps; clone AC-1, Progen, Heidelberg, Germany) and the streptavidin peroxidase method (DAKO, ChemMate TM, Glostrup, Denmark) as previously described [9,22] Serological tests Lipopolysaccharide (LPS) ELISA Bulls sera (n = 104) were examined by LPS-ELISA. The method was performed as previously described [24]. Briefly, microtiter plates were coated with 40 ng per well of antigen (LPS prepared from EBs of Cp. abortus strain OCLH196). Incubations with serum dilutions (1:100) and horseradish peroxidase-conjugated goat anti-bovine IgG (H + L chain specific, Sigma, Buchs, Switzerland) were performed for 1 h. Antigen-antibody reactions were visualised with ABTS (Roche, Rotkreuz, Switzerland) according to the manufacturer s recommendations. Optical densities (OD) were measured at 405 nm by a computer-assisted microplate reader (Tecan, Maennedorf, Switzerland). Cut-off values were calculated for each microtiter plate from mean OD values of seven negative serum samples according to the method of Tijssen [25] Recombinant major outer membrane protein (rmomp) ELISA All LPS-ELISA-positive bull sera (n = 49) were examined by rmomp ELISA using purified recombinant major outer membrane protein (rmomp) of Cp. abortus, Cp. pecorum, and Cp. psittaci as antigens as described elsewhere [26,27]. All sera were diluted 1:100 and pre-adsorbed with Escherichia (E.) coli cells overnight at 4 8C to remove antibodies against E. coli residues in the rmomp antigens. To control the preadsorption, all sera were tested against purified E. coli. OD values were calculated as netto OD values (OD rmomp OD E. coli ). 43

48 Habilitationsschrift Nicole Simona Borel 306 K. Teankum et al. / Theriogenology 67 (2007) Competitive ELISA (celisa) Sera of rams (n = 46) and bucks (n = 15) were investigated by the competitive enzyme-linked immunosorbent assay (celisa) using the monoclonal antibody mab 188 directed against the variable segment 1 of the major outer membrane protein of Cp. abortus according to the protocol of Salti-Montesanto et al. [28]. The results of the celisa were expressed as percentage of inhibition corresponding to the antibody concentration in the serum. Inhibition values above 55% were considered positive for infection with Cp. abortus (positive cut-off) whereas inhibition values between 30 and 55% were considered questionable, attributable to either Cp. abortus or Cp. pecorum. Inhibition values below 30% were considered negative [21,28,29] Bacterial culture of non-chlamydia bacteria An aliquot of fresh semen samples from all animals and one piece of each genital organ were submitted for routine bacteriological examination. 3. Results 3.1. Sensitivity of PCR The results of sensitivity tests are summarised in Table 1. In fresh and cryopreserved bull semen spiked with Cp. abortus DNA, we detected approximately 100, 10 and 10 chlamydial DNA templates/ml of semen samples by the IGS-L, IGS-S, and 16S PCRs, respectively. In fresh and cryopreserved ram semen, we detected approximately 5000, 500 and 500 chlamydial DNA - templates/ml by the IGS-L, IGS-S, and 16S PCRs, respectively. In fresh buck semen, the detection limits of the IGS-L, IGS-S, and 16S PCRs in semen were 500, 50, 50 templates/ml of semen, respectively. Based on the results of these PCR sensitivity tests, the 16S and IGS-S PCRs were selected to test semen samples, and organ samples were examined by the IGS-S PCR method only PCR detection of chlamydiae in semen samples Of 104 fresh semen samples from bulls, the 16S PCR detected DNA from a variety of chlamydial species. These included: Cp. abortus in one sample (1/104, 0.96%), sharing 100% identity to Cp. abortus type sequences; Cp. psittaci in one sample (1/104, 0.96%) shared 96% identity to available Cp. psittaci 16S sequences and; two samples (2/104, 1.9%) shared 98% identity to the 16S rrna sequence of Chlamydia (C.) suis. DNA from diverse uncultured Chlamydia-like organisms in the order Chlamydiales was detected in eight samples (8/104, 7.7%), shared a range of similarities to previously identified Chlamydia-like 16S rrna sequences (89 100%). Comparison of the Chlamydia-positive semen and serological findings of these particular animals is shown in Table 2. Of 200 cryopreserved semen samples from bulls, the 16S PCR revealed DNA of Cp. abortus in only one sample (1/200, 0.5%) which was confirmed by DNA sequencing. Chlamydia-like organisms sharing a range of sequence similarities (86 100%) to previously described uncultured Chlamydia-like organisms were detected in seven samples (7/200, 3.5%, Table 3). In the semen of small ruminants, the 16S PCR revealed the presence of a Chlamydia-like DNA sharing 86% identity to a previously described sequence (GenBank accession no. AY ) in one cryopreserved semen sample from a ram (1/32, 3.1%) originating from AI-5. The remaining ram semen and all semen samples from bucks were negative. In all semen samples (fresh and cryopreserved) tested in this study, no semen samples were positive by IGS-S PCR PCR and IHC detection of chlamydiae in genital organs Detection of chlamydiae in the genital organs using IGS-S PCR yielded negative results in all animals. IHC Table 1 Sensitivity of Chlamydia-specific PCRs performed on semen samples of ruminants Spiking experiments Detection limits of PCR assays IGS-long (templates/ml) IGS-short (templates/ml) 16S (templates/ml) (i) Bull semen (fresh/cryopreserved) a (ii) Ram semen (fresh/cryopreserved) a (iii) Buck semen (fresh) a (iii) Buck semen (cryopreserved) a a Spiked with Chlamydophila abortus DNA followed by DNA extraction using kits (Qiagen). 44

49 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations K. Teankum et al. / Theriogenology 67 (2007) Table 2 Summary of 16S PCR results of fresh semen from bulls (n = 104) used in natural breeding and comparison with serological results (LPS/rMOMP ELISA) Case no. Age (yrs) 16S PCR semen GenBank accession no. Identity (%) ELISA LPS rmomp 25 5 Cp. abortus CR n.d C. suis CTU n.d Cp. psittaci AY n.d Uncultured chlamydiales AY n.d Uncultured chlamydiales AY Cp. abortus, Cp. pecorum C. suis CTU n.d Uncultured chlamydiales AY n.d Uncultured chlamydiales AY n.d Uncultured chlamydiales AY Cp. pecorum Uncultured chlamydiales AY Cp. pecorum Uncultured chlamydiales AY n.d Candidatus rhabdochlamydia AY n.d.: not done; : negative; +: positive; yrs: years. for chlamydial antigen of organ samples was negative in all cases Serological results of bulls (LPS and rmomp ELISA) Antibodies against the LPS antigen of chlamydiae were found in 49 (47.1%) of 104 bulls. The LPSpositive bulls were further examined by rmomp ELISA and we detected antibody against chlamydial species in 39 cases as shown in Table Serological results of small ruminants (celisa) Antibodies against Cp. abortus were detected in 16 (34.8%) of 46 rams and in 9 (60%) of 15 bucks (Table 5). Semen samples from these animals were negative by the 16S and IGS-S PCRs Bacteriology of semen and organ samples Bacteriological examination of fresh bull semen (n = 104) revealed Arcanobacterium pyogenes in four samples, Haemophilus somnus in two samples, Proteus mirabilis in three samples, and Corynebacterium spp. in two samples. In fresh ram semen (n = 46), bacteriology revealed Arcanobacterium pyogenes in four samples, Escherichia coli in one sample, and Pasteurella spp. in one sample. In fresh buck semen (n = 15), we found E. coli in one samples, Staphylococcus spp. in one sample, and mixed culture of Staphylococcus spp. with alphahemolytic Streptococcus spp. in one sample. Bacteriological examination of the organ samples revealed Staphylococcus aureus from orchitis in a ram, and E. coli from a lesion of the epididymitis in a ram. The remaining organ samples from all other animals were negative by bacteriological examination. Two bulls suspected for epididymitis were diagnosed as Table 3 Summary of 16S PCR results from cryopreserved semen samples (bulls: n = 200) originating from different AI centres Case no. 16S PCR semen GenBank accession No. Identity (%) AI centres 135 Uncultured chlamydiales AY AI Uncultured chlamydiales AY AI Uncultured chlamydiales AY AI Cp. abortus CR AI Uncultured chlamydiales AF AI Uncultured chlamydiales AF AI Uncultured chlamydiales AY AI Uncultured chlamydiales AY AI-1 AI: artificial insemination. 45

50 Habilitationsschrift Nicole Simona Borel 308 K. Teankum et al. / Theriogenology 67 (2007) Table 4 Serological results of bulls (n = 104) examined by LPS and rmomp ELISA No. of LPS (+) bulls 49 No. of bulls positive for rmomp ELISA (n = 39 a ) Single infection Cp. abortus 6 Cp. pecorum 26 Mixed infections Cp. abortus/cp. pecorum 5 Cp. abortus/cp. psittaci 1 Cp. abortus/cp. pecorum/cp. psittaci 1 +: positive. a Only LPS positive bulls were tested by MOMP ELISA. spermatocele and were negative for bacteriology. In other two bulls, hypogonadism was confirmed as clinically expected. The prevalence of bacterial orchitis or epididymitis in all animals examined was 6.9% (2/29 cases). 4. Discussion In this report, we evaluated three PCR assays (IGSlong, IGS-short, and 16S PCRs) for use in investigating the prevalence of chlamydiae in semen samples from ruminants. Spiking experiments (except Experiment (iii)), designed to assess PCR sensitivity and the presence of potential inhibitors in semen revealed that the sensitivity of the IGS-L PCR was lower than 16S and IGS-S PCRs by at least 10-fold. In semen of small ruminants, sensitivity of all PCR methods used was reduced compared to semen from bulls (in this study) and boars [22] indicating the presence of potential inhibitory factors. The high concentration of spermatozoa in ram and buck semen may contribute to the inhibition of PCR that we observed. Comparison of fresh and cryopreserved semen from bulls and rams revealed identical detection limits suggesting different preservation methods do not affect PCR detection. In bucks, however, PCR sensitivity was higher in cryopreserved semen compared to fresh semen. As described in another recent study [9], we observed that the 16S was more sensitive than IGS-S PCR (as in Experiment (iii)) for detection of chlamydial DNA in semen samples. Despite the latter result, both 16S and IGS-S PCRs were used to test semen samples from all animals. Consistently with our recent investigations into the prevalence of Chlamydia in semen from boars [22], the results from our 16S rrna PCR screening revealed a rather low prevalence (6.6%, 20/304) of chlamydial DNA in semen from bulls. It is also worth noting that IGS-S PCR yielded only negative results from these semen samples suggesting that chlamydial DNA is only present in very few copies (i.e. 16S rrna PCR sensitivity > IGS-S PCR). The results of the current study are contrary to another study reporting a high proportion of bulls positive for Cp. psittaci in semen [16], however, their study was limited to analysis of three different herds. Such discrepancy could be due to different sources of animals with different management conditions. DNA from a number of different chlamydial species including Cp. abortus (n = 2), Cp. psittaci (n = 1), C. suis (n = 2), and Chlamydia-like organisms (n = 15) were detected by the 16S PCR. Cp. abortus is known to be the main infectious cause of abortion in small ruminants, but to a lesser extent in cattle [9]. In this study, DNA of Cp. abortus was detected in one fresh semen sample from an asymptomatic bull used in natural breeding and one cryopreserved semen sample of a bull from AI station. Although the prevalence of Cp. abortus is very low in this study, its presence in semen samples from bulls indicates the feasibility of venereal transmission. Such cases as they occur would be highly significant since a single Chlamydia-positive bull could spread the organism widely via AI. In contrast to Cp. abortus, the significance of Cp. psittaci in bovine reproductive diseases is still controversial. However, Cp. psittaci was also found to be associated with bovine abortion [9] and the presence of Cp. psittaci in bull semen in this study suggests an involvement of this chlamydial species in reproductive diseases. Interestingly, we detected DNA of chlamydiae sharing similarity to C. suis in semen of two bulls. The significance of this finding is difficult to interpret since this is the first report of this kind. Serological findings of these Cp. abortus-, Cp. psittaci-, and C. suis-positive bulls displayed negative Table 5 Summary of serological results (celisa) for Cp. abortus infection in rams and bucks Animals (n) No. of positive animals (%) No. of negative animals (%) No. of questionable animals (%) Rams (46) 16 (34.8) 15 (32.6) 15 (32.6) Bucks (15) 9 (60) 2 (13.3) 4 (26.6) 46

51 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations K. Teankum et al. / Theriogenology 67 (2007) results by LPS ELISA. This may be due to the local infection of male genital tracts and semen not inducing systemic immune responses. Data retrieved from the records of breading soundness examination of these bulls showed that they were clinically sound and had normal semen quality without abortion history in the herds. The fact that these particular bulls yielded positive results only by the 16S PCR may indicate that these bulls harboured only small amount of chlamydiae. In addition to these positive samples, we identified other semen samples containing Chlamydia-like organisms, other than traditional members of the family Chlamydiaceae. Detection of DNA from these novel organisms is consistent with a number of other recent studies screening cervical swab from sows [30] and bovine placenta [9]. However, the significance of Chlamydia-like organisms in reproductive diseases of animals requires further study as a causal role in the manifestation of clinical diseases remains unproven. Of a total of 122 semen samples from small ruminants that were screened via PCR, we could detect DNA from a Chlamydia-like organism in only one cryopreserved sample from a ram originating from AI station. As previously discussed, PCR sensitivity tests revealed that potential PCR inhibition factors were present in semen from rams. It is not clear whether this low prevalence reflected the actual disease situation in these animals or whether detection sensitivity was reduced by the latter factors. Serological investigation of chlamydial infections in bulls and small ruminants revealed contrasting results to the PCR analysis. The overall serological prevalence of chlamydial infection in bulls, accessed by LPS ELISA, was 47.1% (49/104). In comparison to other studies, the seroprevalence in bulls in this study is in the same range reported in cows with reproductive problems, despite the fact that different serological methods were employed [31,32]. Among LPS-ELISA-positive bulls, Cp. pecorum was most commonly detected (26/49, 53.1%) than Cp. abortus (6/49, 12.2%) as determined by rmomp ELISA. We were also able to detect mixed infections of different chlamydial species occurred in the same bulls (Table 4). Mixed infections between Cp. abortus and Cp. pecorum (5/49, 10.2%) were more often detected than mixed infections between Cp. abortus and Cp. psittaci (1/49, 2%) and mixed infections of Cp. abortus, Cp. psittaci and Cp. pecorum (1/49, 2%). A discrepancy between the results of LPS and rmomp ELISAs was found in 10 bulls which were positive for LPS but negative for rmomp ELISA. Most of these cases had low OD values close to the positive cut-off OD value and, therefore false positive results of the LPS ELISA were probably due to cross reaction with other Gram-negative bacteria. Similar discrepancies between serological and PCR results could also be observed by serological investigation (celisa) of chlamydial infections in small ruminants in which 34.8% of rams (16/46) and 60% of bucks (9/15) were positive for Cp. abortus, indicating a previous exposure to this chlamydial species. In this situation, it is most likely that the chlamydial seroprevalence resulted from infection in other anatomical sites of these animals. A previous survey in Switzerland revealed 19% seroprevalence of Cp. abortus infection in sheep with higher prevalence in specific regional areas [21]. In the current study, the seroprevalence of Cp. abortus infection in rams was higher than the previous survey, however as has been noted, the number of animals investigated between these studies was different. To further address the question of disease situation in ruminants, we attempted to investigate the distribution of chlamydiae in the genital organs. Considering the low prevalence found in semen samples, we were not surprised to obtain negative results from all genital organs examined by IHC and IGS-S PCR. In conclusion, we identified the presence of Cp. abortus and Cp. psittaci in bull semen, the former of which is a known cause of pregnancy complications and abortion in cows. Although this study revealed a low prevalence of chlamydial DNA in semen samples from asymptomatic bulls, wide dissemination of the organisms can still occur if even a single Chlamydia-positive bull is used for AI purposes. Acknowledgements We are grateful to Swissgenetics (Zollikofen), Select Star SA (Puplinge), Triple-Genetics-Service AG (Selzach) Switzerland, Dr. Enzo Fuschini (Swissgenetics, Bütschwil) and Dr. Catherine Marguerat (Schweizerischer Ziegenzuchtverband, Bern) for providing us cryopreserved semen of ruminants. We thank also the laboratory staff of the Institute of Veterinary Pathology and the Institute of Veterinary Bacteriology, University of Zürich, as well as Anja Hamburger for her assistance in serological test. References [1] Storz J, Shupe JL, Smart RA, Thornley RW. Polyarthritis of calves: experimental induction by a psittacosis agent. Am J Vet Res 1966;27:

52 Habilitationsschrift Nicole Simona Borel 310 K. Teankum et al. / Theriogenology 67 (2007) [2] White G. Experimental production of pneumonia in calves by infection with an organism of the psittacosislymphogranuloma group. Vet Rec 1965;77: [3] Cox HU, Hoyt PG, Poston RP, Snider TG, Lemarchand TX, O Reilly KL. Isolation of an avian serovar of Chlamydia psittaci from a case of bovine abortion. J Vet Diagn Invest 1998;10: [4] Shewen PE. Chlamydial infection in animals. Can Vet J 1980;21:2 11. [5] Everett KD, Bush RM, Andersen AA. Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int J Syst Bacteriol 1999;4: [6] Nietfeld JC. Chlamydial infections in small ruminants. Vet Clin N Am Food Anim Pract 2001;17: [7] Chanton-Greutmann H, Thoma R, Corboz L, Borel N, Pospischil A. Aborte beim kleinen Wiederkäuer in der Schweiz: Untersuchungen während zwei Ablammperioden ( ) unter besonderer Beachtung des Chlamydienabortes. Schweiz Arch Tierheilkd 2002;144: [8] Everett KD. Chlamydia and Chlamydiales: more than meets the eye. Vet Microbiol 2000;75: [9] Borel N, Thoma R, Spaeni P, Weilenmann R, Teankum K, Brugnera E, et al. Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet Pathol, in press. [10] Rurangirwa FR, Dilbeck PM, Crawford TB, McGuire TC, McElwain TF. Analysis of the 16S rrna gene of micro-organism WSU from an aborted bovine foetus reveals that it is a member of the order Chlamydiales: proposal of Waddliaceae fam. nov., Waddlia chondrophila gen. nov., sp. nov. Int J Syst Bacteriol 1999;49: [11] Henning K, Schares G, Granzow H, Polster U, Hartmann M, Hotzel H, et al. Neospora caninum and Waddlia chondrophila strain 2032/99 in a septic stillborn calf. Vet Microbiol 2002;85: [12] Ball L, Young S, Carroll EJ. Seminal vesiculitis syndrome: lesions in genital organs of young bulls. Am J Vet Res 1968; 29: [13] Storz J, Carroll EJ, Ball L, Faulkner LC. Isolation of a psittacosis agent (Chlamydia) from semen and epididymis of bulls with seminal vesiculitis syndrome. Am J Vet Res 1968;29: [14] Lozano EA. Etiologic significance of bacterial isolates from rams with palpable epididymitis. Am J Vet Res 1986;47: [15] Storz J, Carroll EJ, Stephenson EH, Ball L, Eugster AK. Urogenital infection and seminal excretion after inoculation of bulls and rams with chlamydiae. Am J Vet Res 1976;37: [16] Domeika M, Ganusauskas A, Bassiri M, Froman G, Mardh PA. Comparison of polymerase chain reaction, direct immunofluorescence, cell culture and enzyme immunoassay for the detection of Chlamydia psittaci in bull semen. Vet Microbiol 1994;42: [17] Hartley JC, Kaye S, Stevenson S, Bennett J, Ridgway G. PCR detection and molecular identification of Chlamydiaceae species. J Clin Microbiol 2001;39: [18] Amin AS. Comparison of polymerase chain reaction and cell culture for the detection of Chlamydophila species in the semen of bulls, buffalo bulls, and rams. Vet J 2003;166: [19] Kaltenboeck B, Heard D, DeGraves FJ, Schmeer N. Use of synthetic antigens improves detection by enzyme-linked immunosorbent assay of antibodies against abortigenic Chlamydia psittaci in ruminants. J Clin Microbiol 1997;35: [20] Buendia AJ, Cuello F, Del Rio L, Gallego MC, Caro MR, Salinas J. Field evaluation of a new commercially available ELISA based on a recombinant antigen for diagnosing Chlamydophila abortus (Chlamydia psittaci serotype 1) infection. Vet Microbiol 2001;78: [21] Borel N, Doherr MG, Vretou E, Psarrou E, Thoma R, Pospischil A. Seroprevalences for ovine enzootic abortion in Switzerland. Prev Vet Med 2004;65: [22] Teankum K, Pospischil A, Janett F, Bürgi E, Brugnera E, Hoelzle K, et al. Detection of chlamydiae in boar semen and genital tracts. Vet Microbiol 2006;116: [23] Lutz-Wohlgroth L, Becker A, Brugnera E, Huat ZL, Zimmermann D, Grimm F, et al. Chlamydiales in guinea pigs and their potential zoonotic risk. J Vet Med A 2006;53: [24] Knitz JC, Hoelzle LE, Affolter P, Hamburger A, Zimmermann K, Heinritzi K, et al. Humorale Immunantwort von Zuchtsauen nach Impfung mit einer stallspezifischen Chlamydophila abortus-vakzine. Dtsch Tierarztl Wochenschr 2003;110: [25] Tijssen P. Practice and theory of enzyme immunoassays. In: Burdon RH, van Knippenberg PH, editors. Laboratory techniques in biochemistry and molecular biology. Amsterdam: Elsevier Publishing; p [26] Hoelzle LE, Hoelzle K, Wittenbrink MM. Expression of the major outer membrane protein (MOMP) of Chlamydophila abortus, Chlamydophila pecorum, and Chlamydia suis in Escherichia coli using an arabinose-inducible plasmid vector. J Vet Med B 2003;50: [27] Hoelzle LE, Hoelzle K, Wittenbrink MM. Recombinant major outer membrane protein (MOMP) of Chlamydophila abortus, Chlamydophila pecorum, and Chlamydia suis as antigens to distinguish chlamydial species-specific antibodies in animal sera. Vet Microbiol 2004;103: [28] Salti-Montesanto V, Tsoli E, Papavassiliou P, Psarrou E, Markey BK, Jones GE, et al. Diagnosis of ovine enzootic abortion, using a competitive ELISA based on monoclonal antibodies against variable segments 1 and 2 of the major outer membrane protein of Chlamydia psittaci serotype 1. Am J Vet Res 1997;58: [29] Gut-Zangger P, Vretou E, Psarrou E, Pospischil A, Thoma R. Chlamydienabort beim Schaf: Möglichkeiten der serologischen Diagnostik mit einem kompetitiven ELISA und Einblick in die epidemiologische Situation in der Schweiz. Schweiz Arch Tierheilkd 1999;141: [30] Camenisch U, Lu ZH, Vaughan L, Corboz L, Zimmermann DR, Wittenbrink MM, et al. Diagnostic investigation into the role of Chlamydiae in cases of increased rates of return to oestrus in pigs. Vet Rec 2004;155: [31] Cavirani S, Cabassi CS, Donofrio G, De Iaco B, Taddei S, Flammini CF. Association between Chlamydia psittaci seropositivity and abortion in Italian dairy cows. Prev Vet Med 2001;50: [32] Wehrend A, Failing K, Hauser B, Jager C, Bostedt H. Production, reproductive, and metabolic factors associated with chlamydial seropositivity and reproductive tract antigens in dairy herds with fertility disorders. Theriogenology 2005;63:

53 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Vet Pathol 43: (2006) Chlamydia-related abortions in Cattle from Graubunden, Switzerland N. BOREL, R. THOMA, P. SPAENI, R. WEILENMANN, K. TEANKUM, E. BRUGNERA, D. R. ZIMMERMANN, L. VAUGHAN, AND A. POSPISCHIL Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Switzerland (NB, PS, RW, KT, EB, LV, AP); Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland (RT); and Clinical Pathology, University Hospital, Zurich, Switzerland (DRZ) Abstract. In 2001, the first case of bovine chlamydial abortion was reported in canton Graubunden, Switzerland. In this region, Chlamydophila (Cp.) abortus is endemic in small ruminants. Hence, we aimed to investigate the incidence of chlamydia-related abortions in cattle from Graubunden. During breeding seasons of , formalin-fixed and paraffin-embedded placenta specimens (n 5 235) from late-term abortions in cattle were analyzed by histopathology, immunohistochemistry with a Chlamydiaceae-specific monoclonal antibody against chlamydial lipopolysaccharide (LPS), and 2 different polymerase chain reaction (PCR) methods (16 S ribosomal ribonucleic acid [rrna] PCR, intergenic spacer [IGS-S] PCR), followed by PCR product sequencing. In 149 of 235 cases (63.4%), histopathologic lesions such as purulent and/or necrotizing placentitis were observed. Chlamydial antigen was clearly demonstrated in immunohistochemistry in only 1 of 235 cases (0.4%). Cp. abortus or Cp. psittaci was found in 12 of 235 (5.1%) and 10 of 235 cases (4.2%) by 16 S rrna PCR and IGS-S PCR, respectively. However, we detected, by 16 S rrna PCR, 43 of 235 cases (18.3%) to be positive for chlamydia-like organisms. In contrast to the situation in small ruminants in the canton Graubunden, bovine abortion from Cp. abortus seems not to play an important role. Nevertheless, zoonotic potential should be taken into account when handling abortion material from cattle. The significance of chlamydia-like isolates other than Waddlia chondrophila remains an open question in abortion and needs further investigation. Key words: Abortion; cattle; chlamydia-like organisms; Chlamydophila abortus; Chlamydophila psittaci; incidence; placentitis; zoonosis. Chlamydiae are implicated in a wide variety of clinically and economically important diseases in livestock and companion animals. 6 In cattle, several disease syndromes caused by chlamydiae infection, including abortion and other urogenital tract infections, pneumonia, conjunctivitis, enteritis, polyarthritis, encephalomyelitis, and mastitis, as well as subclinical infections, have been reported around the world. 10,11,17,21,23 Three chlamydial species that cause these conditions are known to infect cattle: Chlamydophila (Cp.) abortus, Cp. pecorum and the chlamydia-like Waddlia chondrophila. 13 Chlamydial abortion in cattle because of Cp. abortus occurs during the sixth to eighth months of gestation, particularly among heifers in their first pregnancy. Some cows give birth to chlamydiaeinfected, weak, premature calves. Among experimentally induced abortions, placentitis is the most consistent and striking pathologic feature. 11,17 Cp. abortus is also known to cause zoonotic infection in humans, where the greatest threat is to pregnant women and results in spontaneous abortion Cp. pecorum has been associated with conjunctivitis, encephalomyelitis, enteritis, pneumonia, and polyarthritis in cattle. 6,13,14 In contrast to Cp. abortus, the zoonotic potential of Cp. pecorum is unknown. 13 W. chondrophila type strain WSU was implicated as an abortigenic agent in 1986 when it was detected in lung, liver, and other tissues of an aborted bovine fetus in the USA. 5,6,12 The role of Waddlia in abortion is not clear, nor is its range of hosts; the diseases that it causes are not known either. 4 In Switzerland, 14,000 to 28,000 bovine abortion cases are reported every year, resulting in a loss of 22 to 45 million Swiss Francs per year. 9 The first case of bovine abortion in Switzerland from Cp. psittaci was reported in the canton Graubunden in The first human zoonotic infection and abortion caused by Cp. abortus from caprine abortion material occurred in the same region concurrently. 18 An earlier study that examined the causative situation of sheep and goat abortions in Switzer- 49

54 Habilitationsschrift Nicole Simona Borel Vet Pathol 43:5, 2006 Chlamydial Abortion in Swiss Cattle 703 land indicated that Cp. abortus is the leading cause of infectious abortion, inducing 39% of sheep abortions and 23% of goat abortions. 3 Seroprevalence in the Swiss sheep population was highest in Graubunden (43%) compared with 10 other Swiss cantons (2 29%). 1 In this study, we addressed whether the occurrence of chlamydia-related abortions in cattle was comparable with small ruminants aborting in the canton Graubunden. Materials and Methods Abortion cases Placental tissue (1 4 cotyledons per case) of 235 randomly selected late-term cattle abortions, as well as 235 serum samples from dams were collected during the breeding seasons and were submitted to the Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland. Placental smears were prepared and stained by Koester and Stamp to determine the presence of Coxiella burnetii. Serum samples from dams were serologically investigated for the presence of antibodies against Brucella abortus and bovine herpesvirus infection (IBR-IPV). Subsequently, placenta samples were fixed in 4% formalin and embedded in paraffin. Histopathology Hematoxylin and eosin (HE) stained histologic sections of all placenta specimens (235) were examined for the type and the degree of placentitis and/or vasculitis. In cases, where fungal etiology was suspected, additional adjacent sections were stained by periodic acid Schiff (PAS) and Grocott s methenamine silver. Immunohistochemistry for chlamydiae Paraffin sections were investigated for the presence of chlamydial antigen by using a Chlamydiaceae family specific mouse monoclonal antibody (Ab) directed against the chlamydial lipopolysaccharide (LPS; Clone ACI-P, Progen, Heidelberg, Germany). Detection was performed with a detection kit (Dako ChemMate; Dako, Glostrup, Denmark) according to the manufacturer s instructions. Briefly, paraffin sections were deparaffinated in xylene and rehydrated through graded ethanol to water. Antigen retrieval was performed by 10-minute enzyme digestion (Pronase; Dako). To inhibit the endogenous peroxidase activity, the slides were immersed in peroxidase-blocking solution for 5 minutes at room temperature (RT) and incubated with the primary antibody diluted in Ab diluent with background-reducing components for 60 minutes at RT. The sections were incubated for 30 minutes at RT with the link-ab, developed in 3-amino, 9 ethyl-carbazole (AEC) substrate solution for 10 minutes at RT, and counterstained with hematoxylin. A negative control of each section was performed by using the Ab diluent instead of the primary Ab. Intestinal tissue from gnotobiotic piglets experimentally infected with porcine Chlamydia suis strain S45 was used as a positive control. 8 Deoxyribonucleic acid extraction for polymerase chain reaction screening Sections (30 60 mm) were cut from each paraffin block and were placed in a sterile microcentrifuge tube. Paraffin was removed by extraction with 1.2 ml of xylene. After centrifugation at 13,000 3g for 5 minutes, residual xylene was removed by twice extraction with 1.2 ml ethanol. Samples were centrifuged (13,000 3g, 5 minutes), and ethanol was carefully removed. Deoxyribonucleic acid (DNA) for polymerase chain reaction (PCR) analysis was extracted from the tissue pellet by using a commercial DNA extraction kit (DNeasy Tissue kit; Qiagen, Hilden, Germany). PCR detection of chlamydial DNA Each sample was investigated for the presence of chlamydial DNA by using 2 different primer sets: (i) Primer set 16SIGF/16SIGR, targeting the 16 S ribosomal ribonucleic acid (rrna) gene described by Everett and others, modified. 7 (ii) Primer set cigs1f/cigs2r, targeting the intergenic spacer region (IGS-S) between the 16 S and the 23 S rrna gene. (i) A 278-base pair (bp) fragment was amplified with the primer pair (16SIGF: 59-GATGAGGCATGCAAG- TCGAACG-39; 16SIGR: 59-CCAGTGTTGGCGGT- CAATCTCTC-39) complementary to a conserved region within the 16 S gene (16 S rrna PCR). Briefly, 1 ml of sample DNA was tested in 3.5 mm MgCl 2, 0.4 mm of deoxyribonucleotides (dntp) (Roche Diagnostics, Basel, Switzerland), 1 mm of each forward (16SIGF) and reverse (16SIGR) primer (Microsynth, Balgach, Switzerland), and 2 U DNA polymerase (AmpliTaq Gold; Applied Biosystems, Rotkreuz, Switzerland) and, with distilled water reaction, was brought to a final volume of 50 ml. PCR cycling conditions consisted of initial denaturation (95uC, 15 minutes), followed by 45 cycles of denaturation at 94uC for 30 seconds, primer annealing at 70uC for 30 seconds, and extension at 72uC for 45 seconds, with a final extension step for 5 minutes at 72uC. The reactions were carried out on a TGRADIENT thermocycler by Biometra, Göttingen, Germany. Identities of all positively tested 16 S PCR products were determined by direct sequencing of the 16 S rrna gene PCR product from both strands with primer 16SIGF and 16SIGR, respectively. (ii) A 352-bp fragment of the IGS-S rrna gene region was amplified by using forward primer cigs1f (59-CAAGGTGAGGCTGATGAC-39) and reverse primer cigs2r (59- TCGCCTKTCAATGCCAAG-39). One ml of extracted DNA was added to a final concentration of 3.5 mm MgCl 2, 0.4 mm of dntp (Roche Diagnostics), 0.25 mm of forward (cigs1f) and 0.5 mm reverse (cigs2r) primer (Microsynth), 2.5 U DNA polymerase (AmpliTaq Gold; Applied Biosystems), and distilled water, to a final volume of 50 ml, was added. 50

55 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 704 Borel, Thoma, Spaeni, Weilenmann, Teankum, Brugnera, Zimmermann, Vaughan, and Pospischil Vet Pathol 43:5, 2006 PCR product amplification was achieved in 40 cycles (TGRADIENT thermocycler; Biometra) with initial denaturation (94uC, 10 minutes), denaturation at 94uC for 30 seconds, annealing at 48uC for 30 seconds, and extension at 72uC for 45 seconds. Sensitivity of the IGS-S PCR compared with the 16 S rrna PCR was determined by spiking water with 10- fold dilutions of genomic Cp. abortus DNA (10,000 to 0,1 template/ml). Genomic DNA of Cp. abortus strain S26/3 was prepared according to the following protocol: confluent HEp-2 cell layers were infected with Cp. abortus S26/3 multiplicity of infection (MOI) for 3 consecutive days. Thereafter, every third day supernatant was removed and was exchanged with fresh growth medium. Supernatant was collected 216 hours after infection and centrifuged at 250 3g to remove remaining cell debris. Dimethyl sulfoxide (DMSO; 10%) was added to cleared supernatant, and 1-ml aliquots were stored at 280uC. For preparation of genomic chlamydial DNA, 2 ml of frozen aliquots were used. For protein digestion, supernatant was adjusted to 200 mm Tris/Cl (ph 8.0), 25 mm ethylenediamine-tetraacetic acid (EDTA; ph 8.0), 300 mm NaCl, 2% sodium dodecyl sulfate (SDS), and a final concentration of 50 mg/ml proteinase K was added, and proteins were digested overnight at 55uC. The remaining protein parts were removed by 3 repetitive phenol/chloroform extractions followed by an ethanol precipitation. We obtained, from these 2 aliquots, a total of 200 ng chlamydial genomic DNA measured by ultraviolet (UV)-spectrometry (in a Spectrophotometer DU 800, Beckmann Coulter, Fullerton, CA). Accordingly, genomic DNA templates per volume were calculated, and DNA was used for PCR titration experiments. Negative controls were performed by using a reaction mixture with water (H 2 O) instead of the template DNA. Ten ml of each PCR product were electrophoresed in a 1.5% agarose gel, stained with ethidium bromide and observed under UV illumination. The desired fragments were excised and further purified for sequencing with the MinElute Gel Extraction Kit or MinElute PCR Purification Kit (Qiagen). Sequencing Sequencing was performed in collaboration with the sequencing service of the University of Zurich, with an ABI Prism 377 DNA sequencer (Applied Biosystems) or Applied Biosystems 3100 (Synergene Biotech). The obtained sequences were compared with the sequences available in GenBank by using the BLAST server from the National Center for Biotechnology Information ( Results Abortion cases Examinations of placental smears for Coxiella burnetii and of sera from the dams for Brucella abortus and IBR-IPV were negative in all 235 cases. Histopathologic findings Necrosis of trophoblastic epithelium of various degrees and/or abundant infiltration by neutrophils was present in 149 of 235 HE-stained placental specimens (63.4%). Multifocal to coalescing or diffuse, purulent to necrotizing, or necrotizing placentitis was diagnosed in these 149 cases. Most of the cases (104/149) showed both purulent and necrotizing changes, whereas 37 of 149 cases had only necrotizing placentitis, and 8 of 149 cases displayed purulent inflammation of epithelial parts without necrosis of trophoblastic epithelium. Vasculitis was diagnosed in 26 of 235 cases (11.1%). Mainly, arterioles within the intercotyledonary area displayed cellular infiltrates within their walls and consisted of neutrophilic granulocytes, as well as mild to moderate fibrinoid vessel-wall necrosis. Severe, diffuse purulent to necrotizing placentitis and vasculitis with myriad, intralesional fungal hyphae was present in 3 of 26 cases. The examination of the placental tissue was limited in 44 of 235 cases (18.7%) because of severe autolysis. In the remaining 42/235 cases (17.9%), no pathologic changes of the placenta were observed. Immunohistochemistry Positive antigen labeling was present in 9 of 235 cases (3.8%); among these 9 samples, only 1 placenta showed obvious multiple positive labeling (0.4%; Fig. 1). The remaining 8 cases (3.4%) displayed only single positive cells and, therefore, were judged as questionable. PCR detection Comparison of 16 S rrna and IGS-S PCR sensitivity. Amplification of the Cp. abortus S26/3 DNA template by using 16SIGF/16SIGR and cigs1f/cigs2r primer pairs produced roughly 278-bp and 352-bp fragments, respectively. The PCR sensitivity was compared in serial dilution steps (10,000 to 0.1 templates) of Cp. abortus DNA in distilled water. By using the 16 S primer set, we were able to detect as few as 1 to 10 DNA templates, whereas the detection limit of the IGS-S primer set was 10 DNA templates. The detection limit of both assays (16 S rrna and IGS-S PCR) tested on the serial dilutions after the DNA extraction process was 10 DNA templates. PCR detection of chlamydial DNA in placental samples. In the 16 S rrna PCR assay, 55 of 235 cases (22.1%) were positive, 6 samples (2.55%) of these were 98 to 100% identical to Cp. abortus, and another 6 cases (2.55%) were 96 to 100% identical to Cp. psittaci (case Nos. 1 12). Forty-three cases 51

56 Habilitationsschrift Nicole Simona Borel Vet Pathol 43:5, 2006 Chlamydial Abortion in Swiss Cattle 705 Table 1. Positive results of the chlamydial screen by LPS IHC, 16 S rrna PCR, IGS-S PCR, and sequence analysis related to Cp. abortus and Cp. psittaci (cases Nos. 1 17). Case No. Placentitis Vasculitis IHC 16 S PCR 16 S Sequencing IGS-S PCR IGS-S Sequencing 1 Purulent/necrotizing Yes + + Cp. abortus + Cp. abortus 2 Purulent/necrotizing Yes Quest + Cp. abortus + Cp. abortus 3 Purulent/necrotizing No + Cp. abortus + Cp. abortus 4 Purulent/necrotizing No + Cp. psittaci + Cp. psittaci 5 Purulent/necrotizing Yes + Cp. psittaci + Cp. abortus 6 Purulent/necrotizing No + Cp. psittaci 2 7 Necrotizing No + Cp. abortus 2 8 Necrotizing No + Cp. abortus 2 9 Autolysis No + Cp. psittaci 2 10 Autolysis No + Cp. psittaci 2 11 None No + Cp. psittaci 2 12 Autolysis No + Cp. abortus 2 13 Purulent/necrotizing Yes Quest + Cp. abortus 14 None No + Cp. abortus 15 Autolysis No + Cp. abortus 16 None No + Chlamydia-like + Cp. abortus 17 None No + Chlamydia-like + Cp. abortus showed 82 to 100% sequence similarity to the newly described chlamydia-like organisms. From these 43 cases, 9 cases revealed high sequence similarity to Parachlamydia acanthamoebae (91 99%), whereas, in the other 34 cases, a definitive identification was not possible (case Nos ). Thus, these isolates were referred to as chlamydia-like organisms. By IGS-S PCR, 10 of 235 cases (4.2%) were positive for Cp. abortus or Cp. psittaci. Nine of these cases showed high similarity values (98 99%) to Cp. abortus, whereas 1 case was 99% identical to Cp. psittaci (case Nos. 1 5, 13 17). Histologic diagnosis, immunohistochemical, PCR (16 S rrna PCR and IGS-S), and sequencing results of the cases positive for Cp. abortus and/ or Cp. psittaci (cases Nos. 1 17) are displayed in Table 1. Cases Nos positive for chlamydia-like organisms are described in Table 2. Nine of 235 cases (3.8%) displayed placental lesions and were positive by PCR for Cp. abortus or Cp. psittaci (16 S rrna and/or IGS-S) and hence were definitively diagnosed as chlamydial abortion (cases Nos. 1 8, 13; Figs. 1 and 2). Discussion Cattle abortion of unknown infectious cause still remains a major economic problem. Hence, in this study, we focussed on new or not well-defined abortigenic agents, such as on the incidence of chlamydial abortion in cattle. We concentrated our examination on placenta specimens from late-term abortions, because it was shown experimentally that chlamydiae multiply primary in the cotyledons, where they cause severe inflammation and necrosis that leads to abortion. 11,17 The Swiss cattle population is composed of 1.58 million cattle in 65,000 registered cattle herds, from which 78,000 cattle (2,350 herds) are located in the canton Graubunden, which belongs to a mountain region of Switzerland. About 1,500 late-term abortions (more than 6 months of gestation) are reported annually in Graubunden. We selected this mountainous region of Switzerland because of its high seroprevalence in sheep (43%), suggesting this canton is endemic for chlamydial abortion. 1 In 9 of 235 cases (3.8%), placentitis was present and chlamydial antigen (IHC) and/or Cp. abortus/ Cp. psittaci DNA could be found within the lesions (cases Nos. 1 8, 13). Therefore, we assumed that bovine abortion from Cp. abortus/cp. psittaci was not of great importance in the canton Graubunden. Previous studies reported high seroprevalence of chlamydiae in cattle by complement-fixation test (CFT) or enzyme-linked immunosorbent assay (ELISA). 2,15 These observations were partly seen because many cattle harbor chlamydiae in the intestinal tract, which stimulates a low CFT titer, leading to false-positive results. Thus, detection of seropositivity in cattle is considered an indication of chlamydial infection but not necessarily of chlamydial disease unless a rise in titer can be demonstrated. 52

57 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 706 Borel, Thoma, Spaeni, Weilenmann, Teankum, Brugnera, Zimmermann, Vaughan, and Pospischil Vet Pathol 43:5, 2006 Table 2. Positive results of the chlamydial screen by LPS IHC, 16 S rrna PCR, IGS-S PCR, and sequence analysis related to chlamydia-like isolates (cases Nos ). Case No. Placentitis Vasculitis IHC 16 S PCR 16 S Sequencing IGS-S PCR Purulent/necrotizing No 2 + P. acanthamoebae 2 21 Necrotizing Yes 2 + P. acanthamoebae 2 22 Necrotizing No 2 + P. acanthamoebae 2 23, 24 Autolysis No 2 + P. acanthamoebae 2 25, 26 None No 2 + P. acanthamoebae Purulent/necrotizing Yes 2 + Chlamydia-like isolate Purulent/necrotizing No 2 + Chlamydia-like isolate 2 47, 48 Purulent No 2 + Chlamydia-like isolate Necrotizing No 2 + Chlamydia-like isolate Autolysis No 2 + Chlamydia-like isolate 2 60, 61 None No 2 + Chlamydia-like isolate 2 Nevertheless, among the chlamydial-positive cases, the detected histopathologic placental lesions (purulent to necrotizing placentits, and vasculitis) in cattle were similar to those found in sheep. However, placentitis (purulent, purulent to necrotizing or necrotizing) was very often observed (149/ 235 cases), the combination of placentitis and vasculitis was only prevalent in 26 of 235 cases (11.1%). From these 26 cases, 3 had a fungal etiology. Regarding the etiologic agent, based on our sequence result, we found similarities to Cp. abortus but also to Cp. psittaci, which is closely related to Cp. abortus. In case No. 5, Cp. abortus was detected by IGS-S PCR, whereas 16 S rrna PCR detected Cp. psittaci. PCR-induced mutations during the amplification process leading to sequence variation might be unlikely and did not skew the results. There were not many options for mutations because (i) we sequenced directly PCR products and hence averaged our PCR sequences and (ii) short PCR products do not leave many options for mutations. Hence, it is most likely that case No. 5 is a mixed infection of Cp. abortus and Cp. psittaci. The natural method of spread of the infection in cattle is not known, but the disease can be reproduced by parenteral inoculation of the agent. 11,17 In general, infectious elementary bodies can be shed in feces, nasal, ocular or vaginal discharges, uterine fluids, urine, or semen, depending on the particular syndrome. 17 In sheep flocks, the aborting ewe is shedding masses of infectious elementary bodies in uterine fluid, the aborted fetus, and the placenta, resulting in infection of other ewes. 21 It is assumed that transmission in cattle occurs similarly; however, this has not been verified. In contrast to the endemic situation of newly infected sheep and goat flocks, of which up to 60% of the animals suffer from abortion, the 7 cases found in cattle occurred in different herds. Close contact between ruminant species on farms suggests that chlamydiae might have been transmitted from infected sheep to cattle, but this could not be confirmed. The amount of chlamydial antigen within the bovine placenta estimated in the IHC seemed to be much lower than in sheep placenta affected by ovine enzootic abortion (OEA). Similar observations were already published by McKercher 16 in 1969, who reported that, in contrast to OEA, which can be diagnosed by the demonstration of elementary bodies in smear preparations of the placentas from aborting ewes, elementary bodies are found only rarely in aborted bovine placentas and then in very low numbers. To date, the reason for this finding remains unclear, but it could be hypothesized that cattle might be less susceptible to chlamydial abortion than sheep and goats. In previous studies, often only PCR was used to diagnose chlamydial abortion in cattle. Because it is known that many cattle harbor chlamydiae in their intestinal tract and almost all placentas collected after fetal delivery are contaminated with fecal material, a high rate of false-positive PCR results can occur. The combined methods of IHC, 16 S rrna PCR, and IGS-S PCR used in our study enabled us to confirm that the pathogen was actually present within the typical lesions. Placental specimens with PCR assays (16 S rrna, IGS-S) positive for chlamydial DNA but without evidence of placentitis were not definitively considered as bovine chlamydial abortion. Overall, we concluded that a diagnosis of bovine chlamydial abortion should include the presence of 53

58 Habilitationsschrift Nicole Simona Borel Vet Pathol 43:5, 2006 Chlamydial Abortion in Swiss Cattle 707 placentitis and the demonstration of the etiologic agent within the lesion by PCR and/or immunohistochemistry. Chlamydia-like organisms were detected in 43 of 235 cases (18.3%) by 16 S rrna PCR. Failure to detect chlamydia-like organisms by IGS-S PCR is most likely because of the lower sensitivity of IGS- S PCR method compared with 16 S rrna PCR. We tended to assume that only low genome numbers of these organisms were prevalent in our placental specimens. Sequencing of PCR products surprisingly revealed no similarities to W. chondrophila but to P. acanthamoebae. In 5 of 9 cases positive for P. acanthomoebae, placentitis could be found, in 1/9 case vasculitis was detectable. In 28 of 34 cases (82.3%) positive for chlamydia-like purulent and/or necrotizing placentitis was obvious; 4 of 28 cases displayed additional vasculitis. Although WSU , Simkania negevensis, and P. acanthomoebae formed a cluster, the 3 species are as far apart from one another as they are from the chlamydiae. 20 Chlamydia-like organisms could just be detected by 16 S rrna PCR assay and not in the IHC. To date, it is known that some new species do not have a LPS at all or possess the LPS in a truncated form. Therefore, we are not able to detect them in IHC when using an antibody directed against the LPS. Production of antibodies directed against the new species and their evaluation for IHC is in process. Acknowledgements This study was kindly supported by the Federal Veterinary Office, Bern, Switzerland (project number ). We are grateful to the laboratory technical staff of the Institute of Veterinary Pathology, University of Zurich and of the Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland. Fig. 1. Placenta; cow No. 1. Purulent to necrotizing placentitis and vasculitis. Positive IHC and PCR for Chlamydiae. HE. Bar mm. Fig. 2. Placenta; cow No. 1. Positive granular reaction within trophoblastic epithelium. Chlamydial LPS IHC; AEC/peroxidase method, hematoxylin counterstain. Bar 5 20 mm. References 1 Borel N, Doherr MG, Vretou E, Psarrou E, Thoma R, Pospischil A: Seroprevalences for ovine enzootic abortion in Switzerland. Prev Vet Med 65: , Cavirani S, Cabassi CS, Donofrio G, De Iaco B, Taddei S, Flammini CF: Association between Chlamydia psittaci seropositivity and abortion in Italian dairy cows. Prev Vet Med 50: , Chanton-Greutmann H, Thoma R, Corboz L, Borel N, Pospischil A: Abortion in small ruminants in Switzerland: Investigations during two lambing seasons with special regard to chlamydiae. Schweiz Arch Tierheilkd 144: , Corsaro D, Venditti D: Emerging chlamydial infections. Crit Rev Microbiol 30:75 106, Dilbeck PM, Evermann JF, Crawford TB, Ward ACS, Leathers CW, Holland CJ, Mebus CA, Logan LL, Rurangirwa FR, McGuire TC: Isolation of previously undescribed Rickettsia from an aborted bovine fetus. J Clin Microbiol 28: , Everett KDE: Chlamydia and Chlamydiales: more than meets the eye. Vet Microbiol 75: , Everett KDE, Bush RM, Andersen AA: Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int J Syst Bacteriol 49: , Guscetti F, Hoop R, Schiller I, Corboz L, Sydler T, Pospischil A: Experimental enteric infection of gnotobiotic piglets with a Chlamydia psittaci strain 54

59 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 708 Borel, Thoma, Spaeni, Weilenmann, Teankum, Brugnera, Zimmermann, Vaughan, and Pospischil Vet Pathol 43:5, 2006 of avian origin. J Vet Med B Infect Dis Vet Public Health 47: , Hässig M, Eggenberger E, Künzle S, Rüsch P: Reassessment of the herd consultation in facilities with accumulated abortions in cattle. Schweiz Arch Tierheilkd 142:55 64, Holliman A, Daniel RG, Parr JG, Griffiths PC, Bevan BJ, Martin TC, Hewinson RG, Dawson M, Munro R: Chlamydiosis and abortion in a dairy herd. Vet Rec 134: , Idtse FS: Chlamydia and chlamydial diseases of cattle: a review of the literature. Vet Med 4: , Kocan KM, Crawford TR, Dilbeck PM, Evermann JF, McGuire TC: Development of a Rickettsia isolated from an aborted bovine fetus. J Bacteriol 172: , Longbottom D: Chlamydial infections of domestic ruminants and swine: new nomenclature and new knowledge. Vet J 168:9 11, Longbottom D, Coulter LJ: Animal chlamydioses and zoonotic implications. J Comp Pathol 128: , Martinov S: Chlamydial infection in herds of cattle with abortions. Vet Med Nauki 21:81 88, McKercher DG: Cause and prevention of epizootic bovine abortion. J Am Vet Med Assoc 154: , Perez-Martinez JA, Storz J: Chlamydial infections in cattle Part 1, Part 2. Mod Vet Pract 66: , , Pospischil A, Thoma R, Hilbe M, Grest P, Zimmermann D, Gebbers JO: Abortion in humans by Chlamydophila abortus (Chlamydia psittaci serovar 1). Schweiz Arch Tierheilkd 144: , Pospischil A, Thoma R, von Bomhard W, Reitt K, Cantieni J, Zimmermann D, Polkinghorne A: Abortion in cattle caused by Chlamydia psittaci. Schweiz Arch Tierheilkd 144: , Rurangirwa FR, Dilbeck PM, Crawford TB, McGuire TC, McElwain TF: Analysis of the 16 S rrna gene of microorganism WSU from an aborted bovine fetus reveals that it is a member of the order Chlamydiales: proposal of Waddliaceae fam. nov., Waddlia chondrophila gen. nov., sp. nov. Int J Syst Bacteriol 49: , Shewen PE: Chlamydial infection in animals: a review. Can Vet J 21:2 11, Storz J: Overview of animal diseases induced by chlamydial infections. In: Microbiology of Chlamydia, ed. Barron AL, pp CRC Press, Boca Raton, FL, Storz J, Kaltenboeck B: The Chlamydiales. In: Rickettsial and Chlamydial Diseases of Domestic Animals, ed. Woldehiwet Z and Ristic M, pp Pergamon Press, Oxford, UK, 1993 Request reprints from Nicole Borel, DVM, FVH, Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Winterthurerstrasse 268, CH-8057 (Switzerland). n.borel@access.unizh.ch. 55

60 Habilitationsschrift Nicole Simona Borel 56

61 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations DISPATCHES Parachlamydia spp. and Related Chlamydia-like Organisms and Bovine Abortion Nicole Borel,* Silke Ruhl,* Nicola Casson, Carmen Kaiser,* Andreas Pospischil,* and Gilbert Greub Chlamydophila abortus and Waddlia chondrophila cause abortion in ruminants. We investigated the role of Parachlamydia acanthamoebae in bovine abortion. Results of immunohistochemical analyses were positive in 30 (70%) of 43 placentas from which Chlamydia-like DNA was ampli- ed, which supports the role of Parachlamydia spp. in bovine abortion. Chlamydiae are implicated in a wide variety of clinically and economically important diseases in livestock and companion animals. Chlamydophila pecorum has been associated with abortion, conjunctivitis, encephalomyelitis, enteritis, pneumonia, and polyarthritis in ruminants, and Cp. abortus infection is the most common cause of abortion in sheep and goats (1). Cp. abortus also causes zoonotic infection in humans, which in pregnant women, can result in spontaneous abortion (2,3). During the past decade, new Chlamydia-like organisms have been discovered and now emerge as possible public health threats. Simkania negevensis is considered a possible emerging agent of pneumonia (4), and evidence supports the role of Parachlamydia acanthamoebae as an agent of pneumonia (5,6). Waddlia chondrophila is another Chlamydia-like organism initially isolated from lung, liver, and other tissues of an aborted bovine fetus in the United States (7). This organism is now considered an abortigenic agent with a worldwide distribution in cattle, as shown by a recent report of Waddlia-related abortion in Germany (8). The role of Chlamydia-like organisms in bovine abortion is further supported by results of a study of abortion in cattle in Graubünden, Switzerland (9). Analysis of placental specimens by PCR showed that 43 (18.3%) of 235 placentas contained DNA from Chlamydia-like organisms (9). Of these 43 specimens, 8 showed sequence similarity to P. acanthamoebae (95% 99%). Identi cation was not possible in the remaining 35 specimens because of their strong sequence similarity with uncultured chlamydial *University of Zurich, Zurich, Switzerland; and University of Lausanne, Lausanne, Switzerland DNA sequences (Table). These 35 specimens were referred to as Chlamydia-like organisms. None of these 35 specimens was positive by immunohistochemical analysis with antibodies against Chlamydiaceae. This nding indicates that routine diagnostic approaches based on chlamydial lipopolysaccharide would not detect most Chlamydia-like infections (9). To con rm the role of these novel chlamydiae in bovine abortion, we analyzed these placental samples from cattle in Switzerland by using a new speci c immunohistochemical protocol and transmission electron microscopy. The Study Formalin- xed and paraf n-embedded placenta specimens were analyzed by using histopathologic and immunohistochemical techniques. Hematoxylin and eosin stained histologic sections of all placenta specimens (n = 235) were examined for the type and degree of placentitis or vasculitis. Paraf n-embedded sections of specimens positive for Chlamydia-like organisms by 16S rrna PCR (n = 43) were analyzed for Parachlamydia spp. and Waddlia by using speci c mouse polyclonal antibodies as described (10). Optimization experiments for immunohistochemical analysis were performed by using infected amebal and infected HEp-2 cell pellets. Brie y, Acanthamoeba castellanii cultures were infected with P. acanthamoebae strain Hall coccus and W. chondrophila strain ATCC HEp-2 cell monolayers were infected with Cp. abortus strain S26/3. Uninfected cells were used as negative controls. Amebal and cell pellets were prepared as described (11). Optimization of the immunohistochemical protocol for experimentally infected amebal pellets showed the species speci city of mouse antibodies to P. acanthamoebae and W. chondrophila. We did not observe cross-reactivity of both antibodies with Cp. abortus infected HEp-2 cell pellet (data not shown). To test placental specimens, we used mouse polyclonal antibody against P. acanthamoebae and W. chondrophila at dilutions of 1:1,000 and 1:2,000, respectively. Antigen detection was performed with the ChemMate Detection Kit (Dako, Glostrup, Denmark) according to the manufacturer s instructions. Brie y, paraf n-embedded sections were deparaf nated in xylene and rehydrated through graded ethanol to water. Antigen was detected by using repeated microwave heating (750 W for 10 min) in citrate buffer, ph 6.0 (Target Retrieval Solution, Dako). Specimens (slides) and primary antibodies were incubated for 1 hour. Negative and positive controls of each section were included as described (9). Histopathologic lesions such as purulent or necrotizing placentitis were observed in 149 (63.4%) of 235 specimens. Placentitis was observed in 5 of 8 specimens positive for P. acanthamoebae, and vasculitis was observed in 1 of 8 specimens (Table). Positive antigen labeling was observed 1904 Emerging Infectious Diseases Vol. 13, No. 12, December

62 Habilitationsschrift Nicole Simona Borel Chlamydia-like Organisms and Bovine Abortion in 6 of 8 specimens for Parachlamydia spp., but antigen labeling was negative in all specimens for Waddlia (Table). The Figure, panel A shows positive immunohistochemical labeling in 1 of these specimens. Among the 35 placentas positive by PCR for Chlamydia-like organisms other than P. acanthamoebae, 28 (82.3%) showed obvious purulent or necrotizing placentitis by histologic analysis. Four of the 28 specimens with placentitis also had vasculitis. A total of 24 (68.6%) of 35 specimens were positive when tested with antibody against P. acanthamoebae, and all 35 specimens were negative when tested with antibody against W. chondrophila. Two placental specimens positive for Parachlamydia spp. by immunohistochemical analysis and 16S rrna Table. Results of histologic, 16S rrna sequence, and immunohistochemical analyses for 43 placentas positive for Chlamydia-like DNA by a 16S rrna PCR* Specimen Histology 16S rrna sequence Immunohistochemistry no. Placentitis Vasculitis Species % Similarity Parachlamydia spp. Waddlia 1 N Yes Parachlamydia N No Parachlamydia P/N No Parachlamydia P/N No Parachlamydia 97 5 P/N No Parachlamydia 97 6 A No Parachlamydia A No Parachlamydia A No Parachlamydia P/N Yes Chlamydia-like P/N Yes Chlamydia-like P/N Yes Chlamydia-like P/N Yes Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like P/N No Chlamydia-like N No Chlamydia-like N No Chlamydia-like N No Chlamydia-like N No Chlamydia-like N No Chlamydia-like N No Chlamydia-like P No Chlamydia-like P No Chlamydia-like A No Chlamydia-like A No Chlamydia-like A No Chlamydia-like A No Chlamydia-like A No Chlamydia-like A No Chlamydia-like A No Chlamydia-like 95 *When partial 16S rrna sequence showed a similarity >95% with a recognized species (i.e., Parachlamydia acanthamoebae), the corresponding genus was reported (i.e., Parachlamydia spp.). Conversely, when the sequence showed a best BLAST ( hit with uncultured or uncharacterized Chlamydia-related organisms, the sequence was designated as being similar to a Chlamydia-like organism. N, necrotizing; +, positive;, negative; P, purulent; A, autolysis. A 278-bp fragment was amplified and sequenced (9). Emerging Infectious Diseases Vol. 13, No. 12, December

63 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations DISPATCHES PCR were further investigated by transmission electron microscopy for ultrastructural evidence of Chlamydia-like organisms. Brie y, placental tissue specimens were xed with glutaraldehyde and osmium tetroxide and embedded in Epon resin. Ultrathin sections (80 nm) were mounted on gold grids (Merck Eurolab, Dietlikon, Switzerland), contrasted with uranyl acetate dihydrate (Fluka, Buchs, Switzerland) and lead citrate (lead nitrate and tri-natrium dehydrate, Merck Eurolab), and analyzed with a Philips (Eindhoven, the Netherlands) CM10 electron microscope. Both placentas showed Chlamydia-like structures (Figure, panel B). Conclusions To our knowledge, this is the rst description of Parachlamydia spp. in bovine abortion. The organism was detected by PCR (9) and within placental lesions by immunohistochemical analysis by using an antibody speci c for Parachlamydia spp. and electron microscopy. All specimens were negative for Waddlia by immunohistochemical analysis. Isolation of Parachlamydia spp. from aborted bovines is necessary to con rm that this agent causes bovine abortion. Parachlamydia spp. may be involved in lower respiratory tract infections in humans (5,6) and may replicate within both pneumocytes (13) and human macrophages (14). Thus, caution should be taken when handling bovine abortion material because of the potential zoonotic risk. Acknowledgments We thank Adam Polkinghorne for reviewing the manuscript; Ruedi Thoma for providing sample material; and the laboratory technical staff of the Institute of Veterinary Pathology, University of Zurich, and the Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland, for assistance. This study was supported by the State Secretary for Education and Research, Berne, Switzerland (project no. C ) as part of the European Cooperation in the Field of Scienti c and Technical Research Action 855. G.G. is supported by the Leenards Foundation through a career award, Bourse Leenards pour la relève académique en médecine clinique à Lausanne. Dr Borel is a researcher and pathologist at the Institute of Veterinary Pathology of the University of Zurich. Her research interests include the epidemiology and pathology of animal chlamydiosis and the role of obligate intracellular chlamydiae in ruminant abortions and their zoonotic potential to humans. References Figure. A) Immunohistochemical analysis of a bovine placenta positive by PCR for Parachlamydia acanthamoebae, showing a positive brown-red granular reaction within trophoblastic epithelium. Antigen detection was conducted with a polyclonal antibody against Parachlamydia spp. (3-amino-9-ethylcarbazole/peroxidase method, hematoxylin counterstain). B) Transmission electron micrograph of bovine placenta positive by PCR and immunohistochemical analysis for P. acanthamoebae, showing 7 cocci-shaped bacteria in an inclusion with morphologic features similar to those of Chlamydialike organisms (12). 1. Aitken ID, Clarkson MJ, Linklater K. Enzootic abortion of ewes. Vet Rec. 1990;126: Pospischil A, Thoma R, Hilbe M, Grest P, Gebbers JO. Abortion in woman caused by caprine Chlamydophila abortus (Chlamydia psittaci serovar 1). Swiss Med Wkly. 2002;132: Longbottom D, Coulter LJ. Animal chlamydioses and zoonotic implications. J Comp Pathol. 2003;128: Friedman MG, Dvoskin B, Kahane S. Infections with the Chlamydia-like microorganism Simkania negevensis, a possible emerging pathogen. Microbes Infect. 2003;5: Corsaro D, Greub G. Pathogenic potential of novel chlamydiae and diagnostic approaches to infections due to these obligate intracellular bacteria. Clin Microbiol Rev. 2006;19: Greub G, Raoult D. Parachlamydiaceae: potential emerging pathogens. Emerg Infect Dis. 2002;8: Dilbeck PM, Evermann JF, Crawford TB, Ward AC, Leathers CW, Holland CJ, et al. Isolation of a previously undescribed rickettsia from an aborted bovine fetus. J Clin Microbiol. 1990;28: Henning K, Schares G, Granzow H, Polster U, Hartmann M, Hotzel H, et al. Neospora caninum and Waddlia chondrophila strain 2032/99 in a septic stillborn calf. Vet Microbiol. 2002;85: Borel N, Thoma R, Spaeni P, Weilenmann R, Teankum K, Brugnera E, et al. Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet Pathol. 2006;43: Emerging Infectious Diseases Vol. 13, No. 12, December

64 Habilitationsschrift Nicole Simona Borel Chlamydia-like Organisms and Bovine Abortion 10. Casson N, Entenza JM, Greub G. Serological cross-reactivity between different Chlamydia-like organisms. J Clin Microbiol. 2007;45: Borel N, Mukhopadhyay S, Kaiser C, Sullivan ED, Miller RD, Timms P, et al. Tissue MicroArray (TMA) analysis of normal and persistent Chlamydophila pneumoniae infection. BMC Infect Dis. 2006;6: Greub G, Raoult D. Crescent bodies of Parachlamydia acanthamoeba and its life cycle within Acanthamoeba polyphaga: an electron micrograph study. Appl Environ Microbiol. 2002;68: Casson N, Medico N, Bille J, Greub G. Parachlamydia acanthamoebae enters and multiplies within pneumocytes and lung broblasts. Microbes Infect. 2006;8: Greub G, Mege JL, Raoult D. Parachlamydia acanthamoebae enters and multiplies within human macrophages and induces their apoptosis. Infect Immun. 2003;71: Address for correspondence: Gilbert Greub, Institute of Microbiology, University of Lausanne, 1011 Lausanne, Switzerland; gilbert. greub@chuv.ch Emerging Infectious Diseases Vol. 13, No. 12, December

65 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Evidence for Parachlamydia in bovine abortion SILKE RUHL 1, NICOLA CASSON 2, CARMEN KAISER 1, RUEDI THOMA 3, ANDREAS POSPISCHIL 1, GILBERT GREUB 2, NICOLE BOREL 1 1 Institute of Veterinary Pathology, University of Zurich, Vetsuisse Faculty, Switzerland 2 Center for Research on Intracellular Bacteria (CRIB), Institute of Microbiology, University Hospital Center and University of Lausanne, Lausanne, Switzerland 3 Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland Corresponding author: Nicole Borel, DVM, FVH Institute of Veterinary Pathology, Vetsuisse Faculty University of Zurich, Winterthurerstrasse 268 CH-8057 Zurich, Switzerland Tel.: Fax.: n.borel@access.uzh.ch 1 61

66 Habilitationsschrift Nicole Simona Borel Abstract Cattle abortion of unknown infectious etiology still remains a major economic problem. In this study, we thus focused on new possible abortigenic agents such as Parachlamydia acanthamoebae and Waddlia chondrophila. Retrospective samples (n = 235) taken from late-term abortions in cattle were investigated by real-time diagnostic PCR for Chlamydiaceae, Parachlamydia spp. and Waddlia spp., respectively. Paraffin sections of cases positive by real-time PCR for any Chlamydiarelated agent were further examined by immunohistochemistry using specific antibodies. Chlamydophila abortus was detected in only three cases (1.3%) by real-time PCR and ArrayTube Microarray playing obviously a less important role in bovine abortion compared to the situation in small ruminants in Switzerland. By real-time PCR as many as 43 of 235 (18.3%) cases turned out to be positive for Parachlamydia. The presence of Parachlamydia within placental lesions was confirmed in 35 cases (81.4%) by immunohistochemistry. The main histopathological feature in parachlamydial abortion was purulent to necrotizing placentitis (25/43). All 235 cases were negative for Waddlia by real-time PCR. Parachlamydia should be considered as a new abortigenic agent in Swiss cattle. Since Parachlamydia may be involved in lower respiratory tract infections in humans, bovine abortion material should be handled with care given the possible zoonotic risk. Keywords: abortion; cattle; Parachlamydia; Waddlia; real-time PCR; immunohistochemistry 2 62

67 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 1. Introduction Abortion in bovines is of worldwide economic importance. In Switzerland, 14,000 to 28,000 bovine abortion cases are reported every year resulting in a loss of 22 to 45 million Swiss Francs per year (Hässig et al., 2000). Neospora caninum and Bovine Viral Diarrhea Virus (BVDV) were shown to be important abortigenic agents. However, most abortion cases (67.7%) remain of unknown etiology (Reitt et al., 2007). In cattle, several disease syndromes due to Chlamydia infection have been reported (Shewen, 1980; Idtse, 1984; Perez-Martinez and Storz, 1985; Storz and Kaltenboeck, 1993). Chlamydophila (Cp.) abortus and Cp. pecorum (Longbottom, 2004) are known to infect cattle and cause abortion and other urogenital tract infections, pneumonia and conjunctivitis. Cattle abortion due to Cp. abortus occurs during the sixth to eight months of gestation particularly among heifers in their first pregnancy. Among experimentally induced abortions, placentitis is the most consistent and striking pathological feature (Idtse, 1984; Perez-Martinez and Storz, 1985). In 2001, the first case of bovine chlamydial abortion was reported in the canton Graubunden, Switzerland (Pospischil et al., 2002b). In this region, Cp. abortus is endemic in small ruminants (Borel et al., 2004). Moreover, Cp. abortus is known to cause zoonotic infection in humans what, in the case of pregnant women, can result in spontaneous abortion (Pospischil et al., 2002a; Longbottom and Coulter, 2003). Cp. pecorum has also been associated with abortion and conjunctivitis in ruminants (Aitken et al., 1990). However, the zoonotic potential of Cp. pecorum is unknown (Longbottom, 2004). Waddlia chondrophila, another Chlamydiales, is a new abortigenic agent in bovines (Rurangirwa et al., 1999; Dilbeck-Robertson et al., 2003). This obligate intracellular bacterium was isolated from an aborted fetus in the 3 63

68 Habilitationsschrift Nicole Simona Borel United States (Dilbeck et al., 1990) and in Germany (Henning et al., 2002). A serological study further supported the abortigenic role of Waddlia in bovine species (Dilbeck-Robertson et al., 2003). Recently, the zoonotic potential of Waddlia chondrophila was suggested by an association of anti-waddlia antibodies and sustained contact with animals (Baud et al., 2007). In that work, Waddlia seroprevalence was higher in women who had had sporadic and recurrent miscarriages than in control women (p<0.001) (Baud et al., 2007). Other Chlamydia-related organisms, such as Parachlamydia acanthamoebae might also be involved in bovine abortion (Borel et al., 2007b). Thus, among 235 bovine placenta specimens, 43 were positive for Chlamydia-related organisms by 16S rrna PCR, of which 70% (30/43) were confirmed by immunohistochemistry performed using anti-parachlamydia antibodies (Borel et al., 2007b). Given this first evidence of a role for Parachlamydia in bovine abortion (Borel et al., 2007b), we further studied these 235 bovine placenta samples, using recently developed TaqMan real-time PCR for Parachlamydia and Waddlia, and performed immunohistochemistry on positive samples using specific polyclonal mouse antibodies directed against the corresponding Chlamydia-related organism to demonstrate the agent directly in the placenta. 2. Material and Methods 2.1. Abortion cases Placental tissue (1-4 cotyledons per case) of 235 randomly selected late-term cattle abortions were collected during the breeding seasons 2003/2004 and submitted to the Cantonal Laboratory of Veterinary Bacteriology, Chur, Switzerland. 4 64

69 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Subsequently, placenta samples were fixed in 4% formalin and embedded in paraffin and were frozen at -20 C, respectively DNA extraction and real-time PCR DNA was extracted using the AquaPure Genomic Tissue Kit (BioRad, Rheinach, Switzerland) according to the manufacturer s instructions. Overnight proteinase K treatment was performed and DNA was resuspended in 100 µl of DNA hydration solution provided in the kit. All cases were examined by real-time PCR for the presence of Chlamydiacaeae-, Parachlamydia and Waddlia DNA Chlamydiaceae PCR Real-time PCR was conducted on a ABI 7500 using a modified version of the procedure of Everett et al., Primers Ch23S-F (5 - CTGAAACCAGTAGCTTATAAGCGGT-3 ), Ch23S-R (5 - ACCTCGCCGTTTAACTTAACTCC-3 ), and probe Ch23S-p (FAM-CTCATCA TGCAAAAGGCACGCCG-TAMRA) were selected. It was used to amplify a 111-bp product specific for members of the family Chlamydiaceae. 2.5 µl of extracted DNA was added to a commercial Master Mix (TaqMan! Fast Universal PCR Master Mix, Applied Biosystems, Foster City, CA, USA) with final concentration of 5 pmol/µl of each primer and the probe (Microsynth, Balgach, Switzerland) and µl of 2x buffer to yield a final volume of 25 µl. The amplification was performed with initial denaturation (95 C, 10 min), followed by 45 cycles of denaturation at 94 C for 15 s, 60 C for 60 s (Ehricht et al., 2006). Cycle threshold (Ct value) of < was considered as positive. 5 65

70 Habilitationsschrift Nicole Simona Borel Parachlamydia and Waddlia PCR Amplification and PCR product detection of Parachlamydia were performed with the ABI Prism 7000 sequence Detection system (Applied Biosystems), using recently described primers (PacF + PacR) and probe (PacS) (Casson et al., in press). Cycling conditions were 2 min at 50 C, 10 min at 95 C, followed by 45 cycles of 15 s at 95 C and 1 min at 60 C (Casson et al., in press). For Waddlia detection, Primer WadF and WadR as well as probe WadS were used (Goy et al., submitted). Samples were considered negative if no amplification was observed during all 45 cycles. The mean Ct values were compared using unpaired Student t-test ArrayTube (AT) Microarray for species identification of Chlamydiaceae Design and print pattern of the microarray version were as described (Borel et al., 2007a). Briefly, the chip carried 28 probes for species identification (four probes each for C. trachomatis, Cp. pneumoniae and Cp. psittaci, three probes for C. muridarum, Cp. pecorum, Cp. caviae and C. suis, two probes for Cp. abortus and Cp. felis), three genus-specific probes (two for Chlamydophila and one for Chlamydia), five probes for the closest relatives, i.e. Simkania negevensis and Waddlia chondrophila, as well as four positive controls (consensus probes), and one internal staining control (biotin marker). The AT assay was performed as described elsewhere (Borel et al., 2007a) Histology Hematoxylin and eosin (HE) stained histological sections of all placenta specimens (n = 235) were examined for the type and degree of placentitis and/or vasculitis. 6 66

71 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 2.5. Immunohistochemistry Chlamydiaceae Paraffin sections were investigated for the presence of chlamydial antigen by using a Chlamydiaceae family specific mouse monoclonal antibody directed against the chlamydial lipopolysaccharide (LPS, Clone ACI-P, Progen, Heidelberg, Germany). Detection was performed with a detection kit (Dako ChemMate, Dako, Glostrup, Denmark) according to the manufacturer s instructions. Briefly, paraffin sections were deparaffinated in xylene and rehydrated through graded ethanol to water. Antigen retrieval was performed by 10 min enzyme digestion (Pronase, Dako). To inhibit the endogenous peroxidase activity, the slides were immersed in peroxidase-blocking solution for 5 min at room temperature (RT) and incubated with the primary antibody diluted in antibody diluent for 60 min at RT. The sections were incubated for 30 min at RT with the link-antibody, developed in 3-amino, 9-ethylcarbazole (AEC) substrate solution for 10 min at RT, and counterstained with hematoxylin. By using the antibody diluent instead of the primary antibody a negative control of each section was performed. Intestinal tissue from gnotobiotic piglets experimentally infected with porcine Chlamydia suis strain S45 was used as a positive control (Guscetti et al., 2000) Parachlamydia Paraffin sections of cases positive for Parachlamydia by real-time PCR were investigated for the presence of Parachlamydia using a specific mouse polyclonal antibody produced as described elsewhere (Casson et al., 2007). Optimisation experiments for immunohistochemistry protocols are described elsewhere (Borel et al., 2007b). To test the placental specimens, we used the mouse polyclonal antibody 7 67

72 Habilitationsschrift Nicole Simona Borel directed against Parachlamydia at dilution of 1:1,000. Detection was performed with the detection Kit (Dako) according to the manufacturer's instructions. Antigen retrieval was performed by repeated microwave heating (750W for 10 min) in citrate buffer (ph 6,0, Target Retrieval Solution, Dako ChemMate). By immersing the slides in peroxidase-blocking solution for 5 min at RT endogenous peroxidase activity was blocked. Primary antibody was incubated for 1h. Incubation with the link-antibody and substrate solution and counterstaining with hematoxylin was preformed as described above. Negative and positive controls of each section were included as described previously (Borel et al., 2007b) Waddlia Paraffin sections of cases positive for Waddlia (if any) and Parachlamydia by real-time PCR were investigated with the same immunohistochemistry protocol as described in The primary mouse antibody directed against Waddlia was used at dilution of 1:2,000 (60 min at RT). 3. Results 3.1. Abortion cases As already reported earlier (Borel et al., 2007b), purulent and/or necrotizing placentitis were observed in 149 of 235 cases (63.4%). Placentitis and an additional vasculitis were present in 26 of 235 cases (11.1%). Examinations of placental smears for Coxiella burnetii and sera from the dams for Brucella abortus and IBR-IPV are described elsewhere (all cases were negative for the respective agents, Borel et al., 2006). 8 68

73 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 3.2. Chlamydiaceae Three of 235 cases (1.3%) were positive for Chlamydiaceae by real-time PCR. The three positive samples were identified as Cp. abortus using the ArrayTube Microarray. All three Cp. abortus positive cases showed typical histopathological lesions of a chlamydial abortion, such as purulent and/or necrotizing placentitis and vasculitis. Immunohistochemistry for Chlamydiaceae on the three positive cases was positive in two cases and questionable in one case (Table 1) Parachlamydia 43 out of 235 (18.3%) turned out to be positive with the Parachlamydia specific real-time PCR. Among them, only nine had already been identified by the 16SigF-16SigR broad-range PCR we used in our previous work (Borel et al. 2007b) to detect the presence of any Chlamydiales. Overall, 25 of the 43 positive cases (58.1%) showed a purulent and or necrotizing placentitis and three (7.0%) an additional vasculitis (Figure 1). Eight (18.6%) cases showed no alteration and 10 (23.2%) were autolytic. By immunohistochemistry (IHC), 35 of the 43 cases positive by PCR (81.4%) were confirmed positive with the antibody elicited against Parachlamydia (Table 2, Figure 2). Conversely, each of the 43 PCR positive placenta samples was negative with the antibody directed against Waddlia. Interestingly, the mean Ct value of samples with a positive IHC result for Parachlamydia was significantly lower than that of samples with a negative IHC result (p=0.011, see Figure 3). Moreover, among the 34 cases positive by real-time PCR and negative with the broad-range 16SigF-16SigR PCR (which sensitivity is about 100x lower when compared to the new real-time PCR), 28 (82.3%) were confirmed by immunohistochemistry. This is in contrast with the only two (6.8%) out of 29 cases 9 69

74 Habilitationsschrift Nicole Simona Borel with a negative real-time PCR and with a broad-range PCR identifying a Chlamydiales exhibiting less than 95% homology with Parachlamydia acanthamoebae Waddlia All 235 cases were negative for Waddlia by real-time PCR. 4. Discussion Although purulent and/or necrotizing placentitis in cattle abortion is often observed (Borel et al., 2006), most of these cases remain of unknown infectious etiology. We therefore focused on new emerging abortigenic agents including Chlamydiaceae and two Chlamydia-related organisms: Waddlia and Parachlamydia. We selected placental specimens of late term abortions since chlamydiae were experimentally shown to multiply primarily in the cotyledons, causing severe inflammation and necrosis leading to abortion (Idtse, 1984; Perez-Martinez and Storz, 1985). Cp. abortus and Cp. psittaci do not seem to play an important role in bovine abortion in Switzerland since only three samples out of 235 (1.8%) were positive by real-time PCR for Chlamydiaceae. All three were classified as Cp. abortus using the ArrayTube Microarray method. This confirms the result of a previous study (Borel et al., 2006). In contrast, a large number of bovine late-term abortion cases (43/235) were found positive for Parachlamydia by real-time PCR. As mentioned in the results section, only 21% of them had previously been identified using the 16SigF-16SigR broad-range PCR, demonstrating that the Parachlamydia specific real-time PCR is more sensitive than the conventional broad-range PCR, that had been developed for 10 70

75 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations taxonomic purposes and not for diagnostic purposes. More important, the positive real-time PCR results were unlikely to be false positive since most (81%) of the 34 positive cases, which were negative with the 16SigF-16SigR broad-range PCR were confirmed by immunohistochemistry. Conversely, the fact that only 6.9% of cases with a negative Parachlamydia PCR were found positive by immunohistochemistry performed with home - made polyclonal anti-parachlamydia antibodies, confirmed the specificity of this immunohistochemistry. Finally, the strong correlation between the DNA copies (i.e. Ct values) obtained by using the real-time PCR and the presence of a positive immunohistochemistry further validate the results of the realtime PCR Overall, these findings confirm thus the results of a previous study in which Parachlamydia was detected in bovine abortion cases by 16S rrna PCR (Borel et al., 2007b). All 43 cases positive for Parachlamydia real-time PCR were negative for all other abortigenic agents investigated in a previous study (Borel et al., 2006). This further supports the role of Parachlamydia in bovine abortion, and since it was the exclusively identified abortigenic organism, it allowed us to state more precisely the histopathological characteristics of parachlamydial abortion. Purulent to necrotizing placentitis was mainly seen. Concurrent vasculitis in the placenta was only present in three (7.0%) cases positive for Parachlamydia. The combination of placentitis and vasculitis does not seem to be a typical pathomorphological feature of parachlamydial abortion since similar pathological changes are also observed in Cp. abortus-associated abortion. Inflammation and necrosis of the placenta was present in almost 60% of the Parachlamydia-positive cases. In chlamydial abortion, severe placentitis is resulting in placental insufficiency and abortion. A similar mechanism could trigger parachlamydial abortion. However, corresponding fetal tissues of the 11 71

76 Habilitationsschrift Nicole Simona Borel investigated cases were not available, and possible lesions and antigen distribution in the fetus should be further investigated. IHC was performed on the 43 real-time positive cases, 35 of them (81.4%) were positive for Parachlamydia. The amount and distribution of parachlamydial antigen within the bovine placenta estimated by IHC was in general comparable to the situation in ovine enzootic abortion caused by Cp. abortus (data not shown). Abundant parachlamydial positive granular elements were mainly seen intracytoplasmic within trophoblastic epithelium congruent with the growth of Parachlamydia within mammalian cell lines (Corsaro and Greub, 2006). This is the first study showing the entry of Parachlamydia in trophoblastic epithelium cells. However, in vitro experiments should further investigate the significance of these findings. Positive parachlamydial antigen labeling was often as prevalent as known from ovine enzootic abortion (Longbottom and Coulter, 2003), but more frequent than described in chlamydial abortion in cattle (Borel et al., 2006). Negative IHC results in eight cases could be explained by lower sensitivity of the IHC compared to the realtime PCR. Indeed, Ct values were in five out of these eight cases higher than 41 ( ), suggesting low amount of parachlamydial genome copy numbers in the sample, likely explaining the negative IHC results. For the other three cases with Ct values ranging from 37.1 to 39.5, the negative IHC results may have been due to the different sampling localization of the fresh placenta samples submitted for realtime PCR and of the formalin-fixed and paraffin-embedded specimens investigated by IHC, respectively. Suprisingly, Waddlia was not detected in any of the 235 bovine abortion cases. This finding was in strong contrast to our expectations since Waddlia has been described in a bovine abortion in Germany (Henning et al., 2002) and in USA (Dilbeck et al., 1990). Waddlia was neither detected by sensitive and specific real

77 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations time PCR nor by IHC in the cases positive by real-time PCR for Parachlamydia. The absence of cross-reactivity between these two agents was expected given the results of a previous investigation that showed little cross-reaction between members of different Chlamydiales families (Casson et al., 2007). Waddlia has been cultivated from the fetal heart in the bovine abortion case from Germany (Henning et al., 2002) and from fetal lung and liver in the case in the USA (Dilbeck et al., 1990). In our study we examined only placental specimens since fetal organs were not available. It is, however, unknown if Waddlia colonizes the placenta. Further studies on the significance of Waddlia in ruminant abortion should include the examination of both, the placenta and the fetal organs. As demonstrated in this study, bovine abortion due to placentitis of unknown etiology may be caused by novel chlamydiae, which remain unrecognized using routine diagnostic methods such as specific PCR for Chlamydiaceae and immunohistochemistry with anti-chlamydiaceae antibodies. Therefore we applied recently developed tools such as real-time PCR and IHC for Parachlamydia and Waddlia. These new methods were used on field samples in this study and proved to be sensitive and specific for detection of Parachlamydia. Parachlamydia was not only detected by PCR but also with species-specific immunohistochemistry protocols, demonstrating the agent within the placental lesions (see above). The applied protocols are suitable for routine diagnostics and should complete the current examination procedure on bovine abortion cases. The advantage of species-specific real-time PCR methods is its higher sensitivity compared to broad-range PCR, likely due to use of lasers and fluorescent probes to detect the PCR products and to the much smaller amplicon targeted. The length of the amplicon is especially important in determining PCR sensitivity on formalin-fixed samples, given the fragmentation of the DNA during preparation of histologic samples. An another advantage of the real-time 13 73

78 Habilitationsschrift Nicole Simona Borel PCR developed by Casson et al. and by Goy et al. is their possible use for large epidemiological studies, since these PCR are compatible with 384-well format. However, with such specific PCRs, new or closely related Chlamydia-related organisms are not detected. Broad-range PCR methods targeting the 16S gene or other conserved regions of the chlamydial genome are more suitable for this purpose and have been applied in our previous study (Borel et al., 2007b) to screen for the range of Chlamydiales occuring in bovine abortion. Having now first evidences of Parachlamydia as a potential new abortigenic agent in cattle, its isolation from bovine abortion cases will be important to definitively prove its impact as an abortion cause in cattle. Our study revealed Parachlamydia in bovine abortion, another potential zoonotic agent likely causing bronchitis and pneumonia in humans (Greub et al., 2002; Greub et al., 2003a, 2003b). The general zoonotic risk of handling aborted bovine material should be emphasized since additionally there are many already known zoonotic agents, such as Coxiella burnetii, Brucella abortus and Chlamydophila abortus causing severe disease in humans. 5. Conclusion Whereas Waddlia seems not play an important role in bovine abortions in Switzerland, Parachlamydia should be considered as a new abortigenic agent in cattle. The main histopathological feature in parachlamydial abortion was purulent to necrotizing placentitis possibly resulting in placental insufficiency. New sensitive methods such as real-time PCR and IHC specific for Parachlamydia are suitable for routine diagnostics. Since Parachlamydia may be involved in lower respiratory tract infections in humans, a potential zoonotic risk arising from bovine abortion material should be taken in consideration

79 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Acknowledgements We are grateful to the laboratory staff of the Institute of Veterinary Pathology, University of Zurich. We also thank Sebastien Aeby from the Microbiology Institute of the University of Lausanne for technical help. This work was supported by COST Action 855, Animal Chlamydiosis and Zoonotic Implications, Switzerland (SBF Nr.: C ). Silke Ruhl was granted as part of this grant (SBF Nr.: C ). This work represents part of the requirement for Silke Ruhl to obtain the degree of Dr. med. vet. at the Vetsuisse Faculty, University of Zurich. Gilbert Greub (Lausanne, Switzerland) is supported by the Leenards Foundation through a career award entitled Bourse Leenards pour la relève académique en médecine clinique à Lausanne. Reference Aitken, I.D., Clarkson, M.J., Linklater, K., Enzootic abortion of ewes. Vet. Rec. 126, Baud, D., Thomas, V., Arafa, A., Regan, l., Greub, G., Waddlia chondrophila, a potential agent of human fetal death. Emerg. Infect. Dis. 13, Borel, N., Doherr, M.G., Vretou, E., Psarrou, E., Thoma, R., Pospischil, A., Seroprevalences for ovine enzootic abortion in Switzerland. Prev. Vet. Med. 65,

80 Habilitationsschrift Nicole Simona Borel Borel, N., Thoma, R., Spaeni, P., Weilenmann, R., Teankum, K., Brugnera, E., Zimmermann, D. R., Vaughan, L., Pospischil, A., Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet. Pathol. 43, Borel, N., Kempf, E., Hotzel, H., Schubert, E., Torgerson, P., Slickers, P., Ehricht, R., Tasara, T., Pospischil, A., Sachse, K., 2007a. Direct identification of chlamydiae from clinical samples using a DNA microarray - A validation study. Mol. Cell. probes Jun 28; [Epub ahead of print]. Borel, N., Ruhl, S., Casson, N., Kaiser, C., Pospischil, A., Greub, G, 2007b. Parachlamydia spp. and related Chlamydia-like organisms and bovine abortion. Emerg. Infect. Dis., Volume 13, Number 12 December Casson, N., Entenza, J.M., Greub, G., Serological cross-reactivity between different Chlamydia-like organisms. J. Clin. Microbiol. 45, Corsaro, D., Greub, G., Pathogenic potential of novel chlamydiae and diagnostic approaches to infections due to these obligate intercellular bacteria. Clin. Microbiol. Rev. 19, Dilbeck, P.M., Evermann, J.F., Crawford, T.B., Ward, A.C.S., Leathers, C.W., Holland, C.J., Mebus, C.A., Logan, L.L., Rurangirwa, F.R., McGuire, T.C., Isolation of previously undescribed Rickettsia from an aborted bovine fetus. J. Clin. Microbiol. 28, Dilbeck-Robertson, P., McAllister, M.M., Bradway, D., Evermann, J.F., Results of a new serologic test suggest an association of Waddlia chondrophila with bovine abortion. J. Vet. Diagn. Invest. 15,

81 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Ehricht, R., Slickers, P., Goellner, S., Hotzel, H., Sachse, K., Optimized DNA microarray assay allows detection and genotyping of single PCR-amplifiable target copies. Mol. Cell. Probes. 20, Everett, K.D.E., Bush. R.M., Andersen, A.A., Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int J. Syst. Bacteriol. 49, Greub, G., Raoult, D., Parachlamydia acanthamoeba, a potential emerging pathogen? Emerg. Infect. Dis. 8, Greub, G., Berger, P., Papazian, L., Raoult, D., 2003a. Parachlamydiaceae as rare agents of pneumonia. Emerg. Infect. Dis. 9, Greub, G., Boyadjiev, I., La Scola, D., Raoult, D., Martin, C., 2003b. Serological hint suggesting that Parachlamydiaceae are agents of pneumonia in polytraumatized intensive-care patients. Ann. NY Acad. Sci. 990, Guscetti, F., Hoop, R., Schiller, I., Corboz, L., Sydler, T., Pospischil, A., Experimental enteric infection of gnotobiotic piglets with a Chlamydia psittaci strain of avian origin. J. Vet. Med. B. 47, Hässig, M., Eggenberger, E., Künzle, S., Rüsch, P., Reassessment of the herd consultation in facilities with accumulated abortions in cattle. Schweiz. Arch. Tierheilk. 142,

82 Habilitationsschrift Nicole Simona Borel Henning, K., Schares, G., Granzow, H., Polster, U., Hartmann, M., Hotzel, H., Neospora caninum and Waddlia chondrophila strain 2032/99 in a septic stillborn calf. Vet. Microbiol. 85, Idtse, F.S., Chlamydia and chlamydial diseases of cattle: a review of the literature. Vet. Med. 4, Longbottom, D., Chlamydial infections of domestic ruminants and swine: new nomenclature and new knowledge. Vet. J. 168, Longbottom, D., Coulter, L.J., Animal chlamydioses and zoonotic implications. J. Comp. Path. 128, Perez-Martinez, J.A., Storz, J., Chlamydial infections in cattle-part 1, Part 2. Mod. Vet. Pract. 66, , Pospischil, A., Thoma, R., Hilbe, M., Grest, P., Gebbers, JO., 2002a. Abortion in woman caused by caprine Chlamydophila abortus (Chlamydia psittaci serovar 1). Swiss. Med. Wkly. 132, Pospischil, A., Thoma, R., von Bomhard, W., Reitt, K., Cantieni, J., Zimmermann, D., Polkinghorne, A. 2002b. Abortion in cattle caused by Chlamydia psittaci. Schweiz. Arch. Tierheilk. 144, Reitt, K., Hilbe, M., Voegtlin, A., Corboz, L., Haessig, M., Pospischil, A., Aetiology of bovine abortions in switzerland from 1986 to 1995 A retrospective study with emphasis on detection of Neospora caninum and Toxoplasma gondii by 18 78

83 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations PCR. J. Vet. Med. A. 54, Rurangirwa, F.R., Dilbeck, P.M., Crawford, T.B., McGuire, T.C., McElwain, T.F., Analysis of the 16S rrna gene of microorganism WSU from an aborted bovine fetus reveals that it is a member of the order Chlamydiales: proposal of Waddliaceae fam. nov., Waddlia chondrophila gen. nov., sp. nov. Int. J. Syst. Bacteriol. 49, Shewen, P.E., Chlamydial infection in animals: a review. Can. Vet. J. 21, Storz, J., Kaltenboeck, B., The Chlamydiales. In: Rickettsial and Chlamydial Diseases of Domestic Animals, ed. Woldehiwet and Z., Ristic, M. Pergamon Press, Oxford, UK,

84 Habilitationsschrift Nicole Simona Borel Tables Table 1: Results for Chlamydiaceae screen by LPS IHC, real-time PCR and ArrayTube Microarray Table 2: Results for Parachlamydia screen by parachlamydial IHC and real-time PCR for Parachlamydia Figures Figure 1: Placenta; cow, case no. 9. Histopathology of a case positive for Parachlamydia by real-time PCR (Ct value 37.29) and immunohistochemistry showing purulent to necrotizing placentitis. Hematoxylin and eosin staining

85 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Figure 2: Placenta; cow, case no. 9. Immunohistochemistry with the anti- Parachlamydia antibody. Presence of positive granular reaction within trophoblastic epithelium. AEC/peroxidase method, hematoxylin counterstain. Figure 3: correlation of mean Ct value of samples with a positive immunohistochemistry (IHC) result for Parachlamydia and of samples with a negative IHC result (p=0.011)

86 Habilitationsschrift Nicole Simona Borel Table 1 Results for Chlamydiaceae screen by LPS IHC, real-time PCR and ArrayTube Microarray Case No. Placentitis Vasculitis IHC real-time PCR Arraytube Ct-Values Microarray (mean values) 1 purulent/necrotizing yes quest + Cp. abortus necrotizing yes + + Cp. abortus purulent/necrotizing yes + + Cp. abortus Table 2 Results for Parachlamydia screen by parachlamydial IHC and real-time PCR for Parachlamydia Case No. Placentitis Vasculitis IHC real-time PCR Ct-Values (mean values) 4-6 purulent/necrotizing yes , 40.59, purulent/necrotizing no , 35.23, 37.29, 37.37, 37.43, 38.19, 38.37, 39.02, 42.46, purulent/necrotizing no , 39.48, 41.02, 41.27, necrotizing no , 37.60, 37.73, 38.24, 38.73, 39.95, autolysis no , 38.11, 38.34, 38.57, 38.84, 39.15, 39.72, 39.95, autolysis no none no , 38.27, 38.32, 39.37, 40.10, , 46 none no ,

87 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 4. Development and Evaluation of New Chlamydial Diagnostic Tools 4.1. Introduction and Conclusion Rapid detection of bacterial pathogens remains an important issue in diagnostic laboratories of both human and veterinary medicine, particularly when slow-growing or nonculturable microorganisms are involved. In the last decade, the most significant progress in this area was created by the broad application of nucleic acid amplification techniques, notably the polymerase chain reaction (PCR). Nowadays PCR detection assays are available for virtually all important microbial pathogens (Sachse et al., 2002), and efforts to standardize protocols are well underway (Hoorfar et al., 2002). While demands on quality, quantity and rapidity of microbiological diagnosis are steadily increasing, the inherent limitations of PCR in its classical version are becoming evident. The presence of specific bands on agarose gels just indicates the presence of a particular agent, or rather a certain gene of it. The same refers to real-time PCR, although specificity can be increased by selection of welldefined probes, and genome copy numbers can be quantified. However, as soon as one has to deal with more subtle differences between species or strains, such as single-nucleotide polymorphisms and other minor sequence variations, PCR alone cannot provide the necessary resolving power. In this situation, a DNA microarraybased approach, which is capable of verifying the exact nucleotide sequence of a genomic target region, appears a promising alternative. Moreover, the DNA microarray assay represents a very accurate way of discriminating between true and false positive amplification products since, in essence, it amounts to sequencing the DNA target. The potential of microarray-based identification and typing of bacterial pathogens has been demonstrated for mycobacteria (Troesch et al., 1999), staphylococci (Monecke et al., 2003), Escherichia coli O157:H7 (Call et al., 2001; Chizhikov et al., 2001), Listeria spp. (Borucki et al., 2003), as well as agents of bacteraemia (Anthony et al., 2000) and germs relevant in biological warfare (Wilson et al., 2002). The most useful feature of this approach is the strictly nucleotide sequence-based identification of the microorganism, which allows a significant increase in specificity of detection. Most of the currently available DNA array systems are designed for gene transcription and expression analysis. They include customized high- or low-density arrays on glass slides or nylon membranes, and sophisticated multichannel fluorescence readers are used for detection. Additionally, purpose-built hybridization devices are required to ensure defined and reproducible conditions for the formation of target-probe DNA duplexes. From the viewpoint of economy and practicability, the existing systems may be less attractive to small and medium-size diagnostic laboratories and research groups. The recently developed ArrayTube (AT) platform represents a less expensive system for processing low- and highdensity DNA arrays (Monecke et al., 2003; It involves spotted or in situ synthesized DNA chips of 3x3 mm size with an active area of 2.4x2.4 mm, which are assembled to the bottom of 1.5-ml plastic micro-reaction tubes. Hybridization and analysis can be conducted in an easy, rapid and highly parallel fashion on standard laboratory equipment without additional devices, and hybridization signals are amplified by an enzyme-catalyzed precipitation reaction (Taton et al., 2000). A microarray hybridization 83

88 Habilitationsschrift Nicole Simona Borel assay for identification of chlamydiae has been described recently using the ArrayTube platform (Sachse et al., 2005). Hybridization probes were designed on the basis of the most variable window approach, which identified species-specific nucleotide polymorphisms in a region of generally high sequence similarity. The selected 26-nt probe sequences were used on two different series of customized microarrays, i.e. combinatorial high-density in situ synthesized arrays and low-density spotted arrays. Target DNA was prepared by consensus PCR amplifying a 1-kbp segment of the ribosomal RNA operon. Unique species-specific hybridization patterns were obtained for all nine species of the family Chlamydiaceae on both microarray types. The present assay proved suitable for unambiguous species identification of chlamydial cell cultures and showed a potential for direct detection of these bacteria from clinical tissue (Sachse et al., 2005). A subsequently developed low-density spotted AT assay has been evaluated for diagnostic purposes in a validation study on field samples: 4.2. Borel N, Kempf E, Hotzel H, Schubert E, Torgerson P, Slickers P, Ehricht R, Tasara T, Pospischil A, Sachse K. Direct identification of chlamydiae from clinical samples using DNA microarray assay a validation study. Mol Cell Probes. 2007, Jun 28. In the second part of this study, the tissue microarray (TMA) technology was established as a chlamydial antigen detection method for evaluating new antibodies directed against various chlamydial proteins. The TMA technology, developed by Kononen et al., 1998, represents one of the most promising approaches in the field of proteomics and drug discovery. TMAs are constructed by obtaining cylindrical tissue specimens from regular paraffin blocks, assembling them into a single block, and preparing sections containing multiple tissue specimens (Kononen et al., 1998; Bubendorf, 2001; Bubendorf et al., 2001). TMA sections can be analyzed using standard pathology methods, such as hematoxylin/eosin staining, immunohistochemistry (IHC) or in situ hybridization (ISH) (Kononen et al., 1998; Bubendorf, 2001; Moch et al., 2001; Simon et al., 2004; Van de Rijn and Gilks, 2004). The TMA technology enables researchers to perform high throughput microscopic studies on a variety of tissues. The first step of the project included the adaption of the TMA technology for antichlamydial antibodies. The results of this study were presented on the 2nd Workshop of COST Action 855 Animal Chlamydiosis and Zoonotic Implications, Budapest, Ungary, 2004: The tissue and cell culture microarray: a diagnostic tool for evaluation of anti-chlamydial antibodies N. Borel* 1, A. Schifferli 1, C. Kaiser 1, L.Vaughan 1, H. Matile 2, U. Ziegler 3, C. Dumrese 3, A. Pospischil 1 1 Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Winterthurerstrasse 268, CH-8057 Zurich, 2 F. Hoffmann-La Roche Ltd, Pharma Research Basel, Discovery Technology, CH-4075 Basel, Switzerland, 3 Institute of Anatomy, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland Aim of the study: The tissue microarray (TMA) technology represents one of the most promising approaches in the fields of proteomics for analysis in situ. TMAs are constructed by obtaining cylindrical tissue specimens from regular paraffin blocks, assembling them into a single block, and preparing sections containing multiple tissue specimens. TMA sections can be analyzed using standard pathology methods, such as hematoxylin and eosin (HE) staining, immunhistochemistry (IHC) or in situ hybridization (ISH). The TMA system is already widely used in human cancer research on formalinfixed and paraffin-embedded biopsy specimens. Our aim was to adapt the TMA technology for evaluation of commercially available and experimentally produced antichlamydial-antibodies (Abs) tested on Chlamydia infected animal and human tissues as well as on cell pellets infected with different Chlamydia strains. Materials and methods: The tissue array block was created with the Tissue Arrayer from Beecher Instruments. Different animal tissues with naturally acquired or experimentally induced chlamydial infections were included: placentas from sheep, goat and cattle (necropsy material, chlamydial abortion due to Chlamydophila (C.) abortus), human placenta (necropsy material, zoonotic infection due to C. abortus from goat abortion), liver from birds (necropsy material, chlamydial 84

89 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations infection due to C. psittaci), kidney and pericardium from reptiles (necropsy material, granulomatous inflammations caused by C. pneumoniae) and intestine from gnotobiotic piglets (experimentally infected with Chlamydia suis and C. pecorum). Uninfected tissues served as negative controls (placentas from sheep, goat and cattle, human placenta, liver from birds, kidney and pericardium from reptiles and intestine from gnotobiotic piglets). The cell culture array block was created with the Tissue MicroArray Builder from Histopathology Ltd, Hungary (P.H. Stehelin & Cie AG). Vero, Caco-2 and Hep-2 cells were infected with different chlamydial strains: 1710S (C. pecorum), S26/3 (C. abortus), S45/6 (Chlamydia suis), T49/90 (C. psittaci) and Kajaaini 6 (C. pneumoniae). After various stages of infection, cells were scraped from the cell culture flasks and the resulting cell pellets were embedded in paraffin. Uninfected cell cultures served as negative controls. The tissue and the cell culture array blocks were cut using a special adhesive-coated tape sectioning aid system. The sections were stained with HE and IHC was performed using a Chlamydiaceae family-specific mouse monoclonal antibody (Ab) directed against the chlamydial lipopolysaccharide (clps; Clone ACI-P, Progen, Heidelberg, Germany) or a monoclonal Ab directed against the chlamydial heat shock protein 60. Detection was performed with the Detection Kit (Dako ChemMateTM, Glostrup, Denmark). Additionally, 19 monoclonal antibodies derived from mice intravenously infected with C. pneumoniae strain AR-39 were tested on the tissue and cell culture array blocks. Results and conclusion: The tissue and cell culture array system is obviously not only a tool for high-throughput microscopic studies on tumor specimens but also a technology for testing commercially available or newly produced anti-chlamydial Abs on different animals tissues or cell cultures on a single microscope slide. The various animal tissues in our study needed different IHC protocols (i.e. microwave pretreatment for tissues of reptiles and birds) and therefore drawbacks can arise. The anti-lps-chlamydial Ab as well as to a lesser extent the chlamydial anti-hsp 60 Ab gave positive results in all naturally and experimentally infected animal tissues and cell pellets. From the 19 monoclonal antibodies, two showed crossreactivity to C. pecorum, C. abortus, Chlamydia suis and C. psittaci. Chlamydophila pneumoniae infection. BMC Infec Dis Oct 19;6:152. Conclusion Two useful methods for chlamydial antigen and chlamydial DNA detection have been evaluated during this study part. The DNA microarray and the TMA are suitable assays for a pathology laboratory dealing with formalinfixed and paraffin-embedded material. Due to its highly parallel approach, DNA microarray technology opens up new possibilities that may be particularly beneficial for laboratory diagnosis of infectious diseases. The AT assay s advantages is becoming most obvious in diagnosis of animal chlamydioses, where one has to deal with up to nine different species of the family Chlamydiaceae, some of which possess a broad host range. For instance, cattle can be infected by Cp. abortus, Cp. pecorum or Cp. psittaci and sheep are known to be host of Cp. abortus or Cp. psittaci. The TMA is an appropiate tool for evaluation of new anti-chlamydial antibodies as positive and negative control pellets or tissues are included in the TMA block and are tested simultaneously to diminish inter-test variability. The TMA has been applied in the meantime for establishing immunohistochemistry protocols for new antibodies against Chlamydia-like organisms. Based on the successful evaluation of the TMA technique, we applied this method on the particular topic of persistent Cp. pneumoniae infection: 4.3. Borel N, Mukhopadhyay S, Kaiser C, Sullivan ED, Miller RD, Timms P, Summersgill JT, Ramirez JA, Pospischil A. Tissue MicroArray (TMA) analysis of normal and persistent 85

90 Habilitationsschrift Nicole Simona Borel 86

91 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations ARTICLE IN PRESS Molecular and Cellular Probes ] (]]]]) ]]] ]]] Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study Nicole Borel a,1, Evelyne Kempf a,1, Helmut Hotzel b, Evelyn Schubert c, Paul Torgerson d, Peter Slickers e, Ralf Ehricht e, Taurai Tasara f, Andreas Pospischil a, Konrad Sachse c, a Institute of Veterinary Pathology, University of Zurich, Switzerland b Institute of Bacterial Infections and Zoonoses, Jena, Germany c Institute of Molecular Pathogenesis, Friedrich-Loeffler-Institut (Federal Research Institute for Animal Health), Naumburger Str. 96a, Jena, Germany d Institute of Parasitology, University of Zurich, Switzerland e Clondiag Chip Technologies GmbH, Jena, Germany f Institute for Food Safety and Hygiene, Vetsuisse Faculty, University of Zurich, Switzerland Received 29 January 2007; accepted 21 June 2007 Abstract While DNA microarrays have become a widely accepted tool for mrna expression monitoring, their use in rapid diagnosis of bacterial and viral pathogens is only emerging. So far, insufficient sensitivity and high costs have been the major limiting factors preventing more widespread use of microarray platforms in direct testing of clinical samples. In the present study, a total of 339 samples, among them 293 clinical specimens from animals and humans, were examined by the ArrayTube TM (AT) DNA microarray assay to detect chlamydial DNA and identify the species of Chlamydia and Chlamydophila involved. Samples included nasal and conjunctival swabs, formalin-fixed, paraffin-embedded and fresh organ tissue, milk, feces and cell culture. Notably, the AT test was shown to detect mixed infections in clinical samples. The calculated median sensitivity of 0.81 over the entire panel of clinical samples was comparable to conventional 16S PCR, but slightly lower than real-time PCR and other PCR assays. However, when a panel of long-time stored swab samples was excluded from the calculation, the sensitivity was clearly higher (0.87) and equivalent to that of real-time PCR. Altogether, the data demonstrate the suitability of this DNA microarray assay for routine diagnosis. r 2007 Published by Elsevier Ltd. Keywords: Chlamydia spp.; Chlamydophila spp.; Species identification; Clinical samples; Routine diagnosis; DNA microarray testing 1. Introduction Although gene expression monitoring remains the predominant area of DNA microarray use [1], diagnostic applications have been on the increase for several years. The highly parallel approach of this technology provides a diagnostic potential surpassing that of currently used PCR methods because precise sequence information on a variety of genomic loci can be obtained simultaneously, whereas even in multiplex PCR only a few target regions can be examined in a single assay. Furthermore, detection of Corresponding author. Tel.: ; fax: address: konrad.sachse@fli.bund.de (K. Sachse). 1 These authors equally contributed to this study. subtle differences among microbial species, subspecies or strains, such as single-nucleotide polymorphisms and other minute sequence variations, is problematic, if not impossible, with existing amplification assays. So far, DNA microarray assays have been used to identify and genotype bacteria and viruses from culture. Recent applications include detection of antibiotic resistance genes in gram-positive bacteria [2,3], toxin typing of Clostridium perfringens [4], species differentiation among mixed bacterial communities [5], as well as identification of chlamydial species [6], respiratory pathogens [7] and viruses [8]. Inadequate sensitivity of detection and high costs have been the major factors limiting more widespread use of microarray platforms in direct testing of clinical samples /$ - see front matter r 2007 Published by Elsevier Ltd. doi: /j.mcp Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

92 Habilitationsschrift Nicole Simona Borel 2 ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] The ArrayTube TM (AT) system represents a less expensive platform involving chips of mm active area size, which are assembled to the bottom of 1.5-ml plastic microreaction tubes. Due to this unique geometry, all steps of the hybridization reaction can be conducted in the AT vessel without the need for a separate hybridization chamber or other additional laboratory equipment. As visualization of specific duplexes is achieved by enzyme-catalyzed precipitation rather than fluorescence, a simple transmission reader can be used to measure signal intensities. This renders the system particularly user friendly and economical. Furthermore, a recent study revealed that specific hybridization patterns can be obtained from a single PCRamplifiable target copy [9]. This high sensitivity, combined with the high specificity of the microarray technology, appeared to be promising for use in diagnosis and prompted us to conduct a validation study on clinical samples from cases of chlamydial infection. The agents of the genera Chlamydia (C.) and Chlamydophila (Cp.) are associated with a large variety of diseases, such as human urogenital disease and trachoma (C. trachomatis), pneumonia and cardiovascular disease (Cp. pneumoniae), as well as psittacosis in birds (Cp. psittaci), abortion in sheep, goats and cows (Cp. abortus), respiratory and intestinal infection in cattle (Cp. pecorum) and swine (C. suis), as well as infections of reptiles and amphibians (Cp. pneumoniae). Three of the agents, Cp. psittaci, Cp. abortus and Cp. felis, are also known for their zoonotic potential causing severe pneumonia (psittacosis), abortion in pregnant women or keratoconjunctivitis in humans, respectively. The question whether other chlamydiae are also involved in animal-to-man transmission is being addressed by current research. The broad host range of some chlamydial species usually requires more than one test to identify all possible species involved, particularly in veterinary medicine. In this paper, we present the data of DNA microarray testing of 282 animal and 11 human samples and compare them to the findings of real-time PCR, conventional PCR assays, immunohistochemistry, DNA sequencing and cell culture. The sample matrixes examined included nasal, conjunctival and cloacal swabs, formalin-fixed and paraffin-embedded organ tissues and cell pellets, bronchoalveolar lavage, milk, and feces. A Bayesian approach was utilized in defining the AT test sensitivity as there is no available gold standard to compare the new diagnostic test. A variety of PCR tests were utilized on the same tissue samples with a proportion also having immunohistochemistry to detect chlamydial antigens. 2. Materials and methods 2.1. Clinical samples The following panels of samples were examined: biological fluids and feces from zoonotic outbreaks of chlamydioses in birds and guinea pigs [10] (n ¼ 39); formalin-fixed and paraffin-embedded tissues from cases of abortion (n ¼ 27; details in Refs. [11 13]); tissues (n ¼ 68) from diseased frogs [14], snakes [15], birds, and piglets [16]; bovine milk samples (n ¼ 21); conjunctival and nasal swabs from calves (n ¼ 59) and sheep (n ¼ 22, Ref. [17]). For specificity testing, 8 samples of cultured chlamydiae from feces of infected poultry flocks (non-permanent cultures) and cell cultures of 18 chlamydial isolates from birds, cattle and swine in buffalo green monkey (BGM) cells were also included in this study. Furthermore, a panel of 31 samples of formalin-fixed and paraffin-embedded cell pellets was examined (Vero, Caco-2, HEp-2 cells infected with different chlamydial strains, monoinfections, double and mixed infections). Finally, 46 samples of chromosomal DNA from 42 different bacterial species and Candida spp. were used as negative controls in specificity testing (Supplement 1) Cell culture of chlamydial isolates Chlamydiae were cultured in BGM cells using a previously described protocol [18] DNA extraction Preparation of bacterial DNA from swabs and bronchoalveolar lavage was described previously [19]. Briefly, nasal and conjunctival swabs were soaked in 500 ml lysis buffer [100 mm Tris-base, ph 8.5, 0.05% (v/v) Tween 20] placed into a 2-ml SafeLock tube (Eppendorf, Cologne, Germany) and vortexed for 1 min. Having centrifuged all the liquid out of the cotton swab (12,000g, 1 min), the complete lysis buffer extract was centrifuged at 12,000g for 15 min. The pellet was resuspended in 50 ml lysis buffer and incubated at 60 1C for 2 h after addition of 20 ml proteinase K (10 mg/ml). Finally, proteinase K was inactivated by heating at 97 1C for 15 min and cell debris was removed by centrifugation at 12,000g for 5 min. One microliter of the supernatant was used as template in biotinylation PCR prior to AT hybridization. From formalin-fixed and paraffin-embedded material, sections of mm were cut from each paraffin block and placed into a sterile microcentrifuge tube. Paraffin was removed by extraction with 1.2 ml of xylene. After centrifugation at 13,000g for 5 min, residual xylene was removed by two-fold extraction with 1.2 ml ethanol. Samples were centrifuged (13,000g, 5 min), and ethanol was carefully removed. DNA for PCR analysis was extracted from the tissue pellet using the DNeasy Tissue Kit (Qiagen, Hilden, Germany). Milk and feces samples were extracted using the QIAamp DNA Stool Mini Kit (QIAGEN) according to the instructions of the manufacturer. DNA from cell culture and from (non-chlamydial) bacterial cultures was prepared using the High Pure PCR Template Preparation Kit (Roche Diagnostics, Mannheim, Germany). Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

93 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] Description of the DNA microarray Design and print pattern of the microarray version used here were described previously [9]. Briefly, the chip carried 28 probes for species identification (four probes each for C. trachomatis, Cp. pneumoniae and Cp. psittaci, three probes for C. muridarum, Cp. pecorum, Cp. caviae and C. suis, two probes for Cp. abortus and Cp. felis), 3 genusspecific probes (two for Chlamydophila and one for Chlamydia), 5 probes for the closest relatives, i.e., Simkania negevensis and Waddlia chondrophila, as well as 4 positive controls (consensus probes), and 1 internal staining control (biotin marker) AT assay Biotinylation PCR Sample DNA was amplified and biotin labeled for the AT microarray assay in 40 cycles of 94 1C/30 s, 55 1C/30 s, and 72 1C/30 s, using primers U23F-19 (5 0 -ATT GAM AGG CGA WGA AGG A-3 0 ) and 23R-22 (5 0 -biotin-gcy TAC TAA GAT GTT TCA GTT C-3 0 ). Each reaction mix contained 1 ml DNA template, 0.2 ml (1 U) Taq DNA Polymerase (Fermentas, St. Leon-Rot, Germany), 2 ml 10 Taq Buffer with KCl (Fermentas), 1 ml MgCl 2 (25 mm), 2 ml dntp mix (1 mm each, from Roth, Karlsruhe, Germany), 1 ml of each primer (0.5 mm), and 12.8 ml deionized water Hybridization The AT vessel was conditioned by washing twice with 500 ml of Hybridization buffer 1 (Clondiag Chip Technologies, Jena, Germany) at 30 1C for 5 min. All incubations were carried out upon slight shaking (550 rpm) on a heatable horizontal tube shaker (Thermomixer comfort, Eppendorf). For denaturation, 1 ml of the biotinylated PCR product was diluted with 99 ml hybridization buffer in a separate tube, heated at 95 1C for 5 min and put on ice. After transfer into the AT, hybridization was allowed to proceed at 58 1C for 60 min. The supernatant was then discarded, and the tube was washed with 500 ml 2xSSC/ 0.01% Triton X-100 (401C, 5 min), 500 ml 2xSSC (301C, 5 min) and 500 ml 0.2 SSC (20 1C, 5 min). Vacant binding sites of the microarray were blocked by incubation with a 2% solution of Blocking Reagent (Roche) in hybridization buffer at 30 1C for 15 min. Subsequently, the AT was incubated with 100 ml of a 1:10,000 dilution of streptavidine-conjugated horseradish peroxidase (AT Staining Kit, Clondiag) followed by three-step washing as above. Finally, 100 ml of the peroxidase substrate (AT Staining Kit), a 3,3, 0 5,5 0 -tetramethyl benzidine derivative, were added and measurement of the hybridization signal was started Measurement and data processing Hybridization signals were measured at 25 1C using the ATR-01 array tube reader (Clondiag) and recording the final image after 10 min of continuous precipitation. Signal intensity data with local background correction were obtained for each spot of each picture using the Iconoclust software, version 2.3 (Clondiag). Background-corrected signal intensities were given as: NI ¼ 1 M=BG, with NI being normalized intensity, M average (mean) spot intensity, and BG local background intensity. Spot intensities were measured as light transmission, with M values ranging from 1 for complete transmission (background, weak spots) to 0 for complete absorption (dark spots). Thus, normalized signal intensities range between 0 and Criteria for assignment of hybridization patterns The species identity of a test sample was derived from the specificity of the probe showing the most intense hybridization signal on the processed array, provided that all other probes of the same chlamydial species were among the 10 most intense signals, and signal intensities of the genusspecific probes (Chlamydia vs. Chlamydophila) were in line with this assignment. A sample was considered chlamydia-negative when all signal intensities except the internal staining control (biotin marker) were below NI ¼ Real-time PCR A Chlamydiaceae family-specific assay targeting the 23S rrna gene [9] was conducted in 96-well microtiter plates on a Mx 3000 thermal cycler (Stratagene, La Jolla, CA) using primers Ch23S-F (5 0 -CTG AAA CCA GTA GCT TAT AAG CGG T-3 0 ), Ch23S-R (5 0 -ACC TCG CCG TTT AAC TTA ACT CC-3 0 ), and probe Ch23S-p (FAM-CTC ATC ATG CAA AAG GCA CGC CG-TAMRA). Each reaction mix contained 1 ml sample DNA template, 12.5 ml qpcr Mastermix Plus (Eurogentec, Seraing, Belgium), 4.5 ml of each primer and 1 ml of the probe (of 5 mm solutions), and 1.5 ml deionized water. The temperature time profile was 95 1C 10 min, 45 cycles of 95 1C 15 s, 60 1C 60 s. The samples were examined at least in duplicate S PCR-RFLP This two-step test was used to confirm the species identity of chlamydiae from cell culture according to the protocol of Everett and Andersen [20] Nested ompa PCR This additional confirmatory test for chlamydial species identification from tissue samples was done as described previously [19]. Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

94 Habilitationsschrift Nicole Simona Borel 4 ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] S PCR This test used primer pair 16S-IGF (5 0 -GAT GAG GCA TGC AAG TCG AAC G-3 0 ) and 16S-IGR (5 0 -CCA GTG TTG GCG GTC AAT CTC TC-3 0 ) (modified acc. to [21]) to amplify a 278-bp product of the 16S rrna gene specific for the order Chlamydiales. One microliter of extracted DNA was added to a PCR mix containing 3.5 mm MgCl 2, 0.2 mm dntps, 1 mm of each primer (Microsynth, Balgach, Switzerland), 5 ml of 10 reaction buffer, 2 U of AmpliTaq Gold DNA polymerase (Applied Biosystems, Darmstadt, Germany) and water to make a final volume of 50 ml. Amplification was performed in 45 cycles with initial denaturation (95 1C, 15 min), followed by denaturation at 94 1C for 30 s, annealing at 51 1C for 30 s, and extension at 72 1C for 45 s IGS-S PCR Primers cigs-1f (5 0 -CAA GGT GAG GCT GAT GAC-3 0 ) and cigs-2r (5 0 -TCG CCT KTC AAT GCC AAG-3 0 ) (Microsynth) are located in the intergenic spacer region of the rrna operon of Chlamydiales and define an amplicon of approximately 350 bp length. PCR was performed in 50-ml reaction volume containing 1 ml of DNA template, 0.25 mm of forward primer (cigs-1f), 0.5 mm reverse primer (cigs-2r), 0.4 mm dntps, 3.5 mm MgCl 2,5ml of 10x reaction buffer, 2.5 of AmpliTaq Gold DNA polymerase and water to make a final volume of 50 ml. The temperature-time profile was as in 16S PCR, except that the annealing temperature was 48 1C, and a final extension at 72 1C for 10 min was added IGS-L PCR This assay involves primer pair cigs-1f and IGS-1r (5 0 -AGT GGT CTC CCC AGA TTC-3 0 ) to amplify a 750-bp segment of the intergenic spacer as described previously [10]. The reaction mix contained 1.5 mm MgCl 2, mm of each dntp, 0.25 mm of each primer, 5 ml of 10 reaction buffer, 2.5 U of AmpliTaq Gold DNA polymerase and water to make a final volume of 50 ml. Amplification was conducted with initial denaturation (94 1C, 10 min), followed by 40 cycles of denaturation at 94 1C for 30 s, annealing at 54 1C for 30 s and extension at 72 1C for 45 s Immunohistochemistry for chlamydiae Paraffin sections were examined for the presence of chlamydial antigen using a Chlamydiaceae family-specific mouse monoclonal antibody (Ab) directed against the chlamydial lipopolysaccharide (LPS; Clone ACI-P, Progen, Heidelberg, Germany). Detection was performed with the Detection Kit (Dako ChemMate TM, Glostrup, Denmark) according to the manufacturer s instructions. Briefly, paraffin sections were deparaffinated in xylene and rehydrated through graded ethanol to water. Antigen demasking was performed by 10-min enzyme digestion (Pronase, Dako). To inhibit the endogenous peroxidase activity, slides were immersed in peroxidase-blocking solution at room temperature (RT) for 5 min and incubated with the primary antibody diluted in Ab diluent with background-reducing components for 60 min. The sections were incubated at RT for 30 min with the link-ab, developed in 3-amino, 9-ethyl-carbazole (AEC) substrate solution for 10 min, and counterstained with hematoxylin. A negative control of each section was performed using the Ab diluent instead of the primary Ab. Intestinal tissue from gnotobiotic piglets experimentally infected with porcine Chlamydia suis strain S45 was used as positive control [22] Statistical calculations Bayesian techniques (Markov Chain Monte Carlo) were used to estimate the diagnostic parameters of each procedure. These techniques have been known for some time and can be exploited to estimate the diagnostic performance of tests and the underlying true prevalence of infection providing certain conditions are met. These include the numbers of tests required on a population of unknown prevalence in the absence of prior information and is reviewed by Brunscum et al. [23]. In total, results from 293 clinical samples were included in the analysis, among them 40 samples from experimentally infected animals and cell culture. The latter were assumed to have a different probability of infection than the other samples. Likewise, because different DNA extraction techniques were used, the results were calculated using different probabilities of infection (or effectively of successful DNA extraction) for the different techniques. For each sample, the true infection status was not observed, but a series of binary test results. Likewise, the true sensitivity of each test was unknown. For 73 samples, 4 tests were undertaken, 39 samples had 3 tests, 141 samples had 2 tests, whilst 40 samples had only the array test, but were taken from animals experimentally infected or from tissue cultures and hence had different probabilities of infection compared to the other groups. The specificity of each of the DNA tests was assumed to be 1. Consequently, the model was identifiable even with the possibility of multiple covariances between test sensitivities [24]. The sensitivity and specificity of the immunohistochemistry samples were also unknown. Additional parameters were introduced to model covariance of test sensitivities. In the Bayesian analysis, prior knowledge of parameters must be given. In the case where there is no prior knowledge available this can be given as uniform or non-informative distributions. In the case of the unknown sensitivities, for example, this is easily achieved by using the beta distribution with a and b parameters of 1 and 1, respectively. Likewise for the unknown covariance of the sensitivities between tests, priors of a distribution were used with upper and lower Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

95 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] 5 limits estimated from the possible range [25]. Because each tissue sample had a varying number of tests done on them, a macro written in Excel was used to implement the Markov chain [26]. The chain was burnt in using 1000 iterations and then run for a further 10,000 iterations to sample the posterior distribution. Covariances between tests were examined in a pairwise fashion and if the upper (for negative covariance) 95% posterior probability distribution or lower (for positive covariance) included 0, the covariances were eliminated from the model. Likewise, three-way covariances were examined for various combinations of tests. The final model included significant covariance estimates of test sensitivities and was used to calculate the probability distribution of all tests used on the tissues. 3. Results A total of 339 samples were included in the study. Clinical samples (n ¼ 293) were examined by both the AT assay and either at least one reference test for chlamydiae, or were from experimentally infected animals. Chlamydiaceae-specific real-time PCR targeting the 23S rrna gene was chosen as the reference test for swabs, milk, feces and biological fluids. Due to the wide variety of sample matrixes among the present panel of specimens, 16S PCR, IGS-S and IGS-L PCRs, nested ompa PCR and partial ompa gene sequencing, as well as immunohistochemistry were conducted as additional confirmatory tests to verify the accuracy of the microarray findings in selected cases, particularly in the subset of formalin-fixed, paraffinembedded, and fresh tissue samples (n ¼ 95). The AT test was also used to establish the species identity of chlamydiae from cell culture (n ¼ 57). Finally, 46 non-chlamydial microbial strains were tested to check the specificity of the AT assay. No evidence of lack of specificity was detected for the AT assay and hence justified a specificity of 1 in the latent class models Samples from outbreaks of chlamydioses Lines 1 14 of Table 1 show the test results of 39 samples, among them 10 from humans. DNA microarray testing revealed the presence of Cp. psittaci in 6 samples from infected poultry flocks and a severely diseased contact person (Patient K), all of them related to an outbreak of avian chlamydiosis in central Germany in As an example, the test results of a cloacal swab from an infected laying hen are illustrated in Fig. 1A. Samples from two other patients, one of them also ill and Cp. psittaci positive (Patient R, see Fig. 1B), were sent in as suspected cases of psittacosis from different outbreaks. Interestingly, sample 1/13, a cloacal swab from a symptomless chicken, tested clearly positive in real-time PCR, but yielded an atypical hybridization pattern consisting only of the genus-specific Chlamydophila probes, but no species-specific signals. Sequencing of the ompa gene revealed 88% similarity to its closest relative, an unspecified avian strain of Chlamydophila, but no homology to any of the currently defined species of Chlamydiaceae (data not shown). Furthermore, Cp. caviae was identified in 8 samples from an infected guinea pig flock in Switzerland in An example is shown in Fig. 1(C). Examination by real-time PCR and AT assay of swabs from conjunctiva and contact lenses of the owner of this flock, who reported clinical symptoms of ocular surface disease such as conjunctivitis and keratitis, as well as from healthy guinea pigs living in close contact with infected animals, failed to confirm the presence of Chlamydiaceae spp. (Fig. 1D). However, further testing of the sample by 16S PCR indicated the presence of Parachlamydia acantamoebae. The significance of this finding is yet unclear, and further investigation is needed [10]. Comparison of the microarray results with real-time PCR (Table 1) shows 18 concordant positive and 20 concordant negative findings, and one discrepant result (sample ]18; 1/39 ¼ 2.6%) Formalin-fixed and paraffin-embedded organ tissue A total of 95 samples of this category were examined. The results of those 57 samples that previously tested positive in 16S PCR and/or other tests are given in Table 1, lines Six different species were identified, i.e. Cp. abortus in tissue from sheep, cattle and a human case, Cp. psittaci in birds, cattle and swine, Cp. pneumoniae in frogs and snakes, as well as Cp. pecorum, C. trachomatis and C. suis in experimentally infected gnotobiotic piglets. While the findings of the microarray and the confirmatory assay were identical in 44 cases, the AT assay failed to produce a measurable hybridization pattern in 9 instances (9.5%). Two PCR-negative samples tested positive on the microarray. In those two cases with discrepant results (samples 2a/31 and 42) the AT results appear more plausible than 16S PCR in view of the known host specificities of Cp. pneumoniae and Cp. psittaci. Additionally, 38 tissue samples found to be chlamydia negative in 16S PCR and other tests also proved negative in the AT assay (Table 1, line 35) Milk samples A panel of 21 milk samples from three dairy herds reportedly having problems with milk quality (elevated somatic cell counts) and repeated abortions was examined (Table 1, lines 36 41). Twelve samples that tested positive in real-time PCR were also found positive with the microarray assay, and 5 were negative in both assays, whereas 3 (14.3%) real-time PCR-positives did not react in the microarray assay and one sample (4.8%) was only positive in the latter. The most frequently detected species was Cp. abortus (n ¼ 9), and two samples were found to contain both Cp. abortus and Cp. psittaci. Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

96 Habilitationsschrift Nicole Simona Borel 6 ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] Table 1 Examination of 236 clinical samples by the AT assay and confirmatory tests (summarized data) Line no. Sample description (Sample ID: Panel table number /Sample number) n ArrayTube test Confirmatory test 1 Patient K, bronchoalveolar lavage (1/1) 1 Cp. psittaci rtpcr positive 2 Patient K, urine (1/2) 1 Cp. psittaci rtpcr positive 3 Patient K, urine 11 days later (1/3) 1 No signal rtpcr negative 4 Patient R, bronchoalveolar lavage (1/4) 1 Cp. psittaci rtpcr positive 5 Patient F, pharyngeal swab (1/5) 1 Cp. psittaci rtpcr positive 6 Patient H, lymphatic tissue (1/6) 1 No signal rtpcr negative 7 Cloacal swabs from diseased poultry (1/7 12) 6 Cp. psittaci a rtpcr positive 8 Cloacal swab from non-diseased hen (1/13) 1 Chlamydophila sp. a,b rtpcr positive 9 Conjunctival swabs from diseased guinea pigs (1/14, 19 24) 7 Cp. caviae c rtpcr positive 10 Conjunctival swabs from diseased guinea pigs (1/15 16, 25 27, 30 31) 7 No signal c rtpcr negative 11 Conjunctival swabs from diseased guinea pig (1/18) 1 Cp. caviae c rtpcr negative 12 Conjunctival swabs from non-diseased guinea pigs (1/17, 28 29, 32 35) 7 No signal c rtpcr negative 13 Swab from contact lens of guinea pig owner (1/36 37) 2 No signal c rtpcr negative 14 Conjunctival swab from guinea pig owner (1/38 39) 2 No signal c rtpcr negative 15 Placental tissue from human case of abortion (2a/27) 1 Cp. abortus 16S PCR positive 16 Organ tissue from cases of abortion in sheep (2a/1 10, 12 15) 14 Cp. abortus d 16S PCR positive 17 Placental tissue from a case of abortion in sheep (2a/11) 1 Cp. abortus d 16S PCR negative 18 Placental tissue from cases of abortion in cattle (2a/17, 20, 22, 24) 4 Cp. abortus d 16S PCR positive 19 Placental tissue from cases of abortion in cattle (2a/16, 18, 19, 21, 23, 25) 6 No signal d 16S PCR positive 20 Placental tissue from a case of abortion in cattle (2a/26) 1 Cp. psittaci d 16S PCR positive 21 Organ tissue from frogs (2a/28 30) 3 Cp. pneumoniae d 16S PCR positive 22 Organ tissue from snakes (2a/30, 32) 2 Cp. pneumoniae d 16S PCR positive 23 Organ tissue from snake (2a/31) 1 Cp. pneumoniae d 16S PCR positive (C. suis) 24 Organ tissue from birds (2a/34 39, 41) 7 Cp. psittaci d 16S PCR positive 25 Organ tissue from bird (2a/40) 1 Cp. psittaci d 16S PCR negative 26 Organ tissue from bird (2a/42) 1 Cp. psittaci d 16S PCR positive (Cp. pneumoniae) 27 Organ tissue from bird (2a/43) 1 C. suis d 16S PCR positive 28 Organ tissue from bird (2a/44) 1 No signal d 16S PCR positive 29 Intestinal tissue from experimentally infected gnotobiotic piglets (2a/45 47, 49) 4 Cp. pecorum Exp. infection Cp. pecorum 30 Intestinal tissue from experimentally infected gnotobiotic piglet (2a/48) 1 No signal Exp. infection Cp. pecorum 31 Intestinal tissue from experimentally infected gnotobiotic piglet (2a/50) 1 No signal Exp. infection C. trachomatis 32 Intestinal tissue from experimentally infected gnotobiotic piglet (2a/51) 1 C. trachomatis Exp. infection C. trachomatis 33 Intestinal tissue from experimentally infected gnotobiotic piglets (2a/52 53) 2 Cp. psittaci Exp. infection Cp. psittaci 34 Fresh organ tissue from naturally infected, diseased piglets (2a/53 57) 4 C. suis rtpcr positive 35 Chlamydia-negative tissue samples from different animal species (2b/1 38) 38 No signal d 16S PCR negative 36 Milk from dairy herd with a history of elevated somatic cell counts in milk 9 Cp. abortus rtpcr positive and frequent cases of abortion (3/1 3, 5 8, 15 16) 37 Milk samples from the same herd (3/4, 18) 2 Cp. abortus+cp. rtpcr positive psittaci 38 Milk sample from the same herd (3/21) 1 Cp. psittaci rtpcr positive 39 Milk samples from the same herd (3/9, 13, 19) 3 No signal rtpcr positive 40 Milk samples from the same herd (3/10 12, 17, 20) 5 No signal rtpcr negative 41 Milk sample from the same herd (3/14) 1 Cp. abortus rtpcr negative 42 Conjunctival and nasal swabs from calves (4/1, 4, 7, 8, 11 16, 20 21, 27, 39, 17 Cp. pecorum rt PCR positive 41, 45, 46) 43 Conjunctival and nasal swabs from calves (4/25, 57) 2 Cp. pecorum rtpcr negative 44 Conjunctival and nasal swabs from calves (4/6, 23, 32, 37, 40, 55, 57) 7 Cp. abortus rtpcr positive 45 Conjunctival and nasal swabs from calves (4/2, 5, 19, 24, 26, 28, 29, 31, 34, 15 Cp. pecorum+cp. rtpcr positive 35, 42 44, 48, 56) abortus 46 Conjunctival swab from calf (4/30) 1 Cp. pecorum+cp. rtpcr negative abortus 47 Nasal swab from calf (4/10) 1 Cp. psittaci rtpcr positive 48 Nasal swabs from calves (4/3, 9) 2 Waddlia rtpcr positive chondrophila e 49 Conjunctival and nasal swabs from calves (4/17, 18, 22, 36, 38, 47, 49 54, 59) 13 No signal rtpcr positive 50 Nasal swab from calf (4/33) 1 No signal rtpcr negative Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

97 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] 7 Table 1 (continued ) Line no. Sample description (Sample ID: Panel table number /Sample number) n ArrayTube test Confirmatory test 51 Conjunctival swabs from sheep (4/60, 61, 65, 72 76) 8 No signal f rtpcr negative 52 Conjunctival swabs from sheep (4/62, 64, 66, 67, 79) 5 No signal f rtpcr positive 53 Conjunctival swabs from sheep (4/65, 77, 78, 80, 81) 5 Cp. abortus f rtpcr positive 54 Conjunctival swabs from sheep (4/63, 69, 70) 3 C. suis f rtpcr positive 55 Conjunctival swab from sheep (4/71) 1 C. suis f rtpcr negative n, number of samples tested. a Additional confirmation by partial ompa gene sequencing. b AT and sequencing data indicate that this strain has limited sequence homology to the genus Chlamydophila, but does not belong to any of the currently defined species of the family Chlamydiaceae. c Additional confirmation by 16S PCR and IGS-L PCR. d Additional confirmatory data from IGS-S PCR, immunohistochemistry, histology, antigen ELISA and bacteriology all tissue samples were formalin fixed and paraffin embedded, unless stated otherwise. e The family Waddliaceae is also a member of the order Chlamydiales. f 16S PCR as additional confirmatory test. Fig. 1. Examination of clinical samples using the ArrayTube test. Microarray images, barplots of hybridization signals and real-time PCR amplification curves are shown for four typical samples. (A) Cloacal swab from a laying hen suffering from psittacosis (sample 1/11); positive hybridization signals include all probes of genus Chlamydophila (A1) and species Cp. psittaci (A2). The asterisk denotes the signal bar of the internal staining control (biotinylated marker probe). (B) Bronchoalveolar lavage of a psittacosis patient (sample 1/4); positive hybridization signals include all probes of genus Chlamydophila (B1) and species Cp. psittaci (B2). (C) Conjunctival swab of diseased guinea pig (sample 1/19); positive hybridization signals include all probes of genus Chlamydophila (C1) and species Cp. caviae (C2). (D) Swab from contact lens of the owner of the guinea pig (sample 1/36); no positive hybridization signals. Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

98 Habilitationsschrift Nicole Simona Borel 8 ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] 3.4. Conjunctival and nasal swabs from cattle and sheep Results obtained from 81 samples are summarized in Table 1, lines In 37 nasal swabs from calves, the AT assay identified Cp. pecorum (n ¼ 9), Cp. abortus (n ¼ 5), both agents in mixed infection (n ¼ 9), and Cp. psittaci (n ¼ 1). In a panel of 22 conjunctival swabs from the same herds, 10 samples were found to contain Cp. pecorum, 2 had Cp. abortus, and 7 had both. Furthermore, examination of 22 conjunctival swabs from sheep revealed Cp. abortus in 5 samples and C. suis in 4 cases. As the latter was confirmed by 16S PCR, this unexpected finding is probably due to close contact between sheep and pigs. Notably, there was a considerable number of non-concordant findings between AT on the one hand and real-time PCR and 16S PCR on the other in this series of samples. Thus, 18/81 (22.2%) real-time PCR-positive samples failed to produce a measurable hybridization signal in the AT test, and in 4 cases (4.9%) were the AT-positive samples negative in real-time PCR Cell culture The DNA microarray assay was further validated on a series of 57 chlamydial cell culture samples. Comparison of the species identities found by PCR and AT test revealed complete agreement for all 26 cell cultures of chlamydial isolates from birds, cattle and swine. Furthermore, a comparative trial of 31 samples of experimentally infected cell cultures demonstrated the capability of the AT assay to detect and identify double and mixed infections involving up to five different chlamydial species in a single sample. Table 2 Number of samples subjected to each combination of tests (293 in total) Test combination AT test, real-time PCR, 16S PCR, and IGS-L PCR 48 AT test and real-time PCR 97 AT test, 16S PCR, IGS-S PCR, and 25 immunohistochemistry AT test and 16S PCR a 35 AT test, immunohistochemistry, and 16S PCR 39 AT test and immunohistochemistry 1 AT test and nested ompa PCR (from cell culture) 8 AT test only (experimentally infected or cell culture) 40 a 18 of these samples were from cell cultures. Table 3 Median and 95% credible intervals for the various tests Number of samples Test Sensitivity Specificity a (a) All samples included, n ¼ 293 AT test 0.81 ( ) 1 Real-time PCR 0.90 ( ) 1 16S PCR 0.84 ( ) 1 ompa PCR 0.90 (0.66 1) 1 IGS-L PCR 0.53 ( ) 1 IGS-S PCR 0.91 (0.70 1) 1 Immunohistochemistry 0.57 ( ) 0.96 (0.86 1) (b) n ¼ 212; excluding 81 swab samples of inferior quality AT test 0.87 ( ) 1 Real-time PCR 0.86 ( ) 1 16S PCR 0.84 ( ) 1 ompa PCR 0.90 (0.66 1) 1 IGS-L PCR 0.43 ( ) 1 IGS-S PCR 0.91 (0.69 1) 1 Immunohistochemistry 0.57 ( ) 0.95 (0.81 1) a All DNA-based tests were assumed to have a specificity of Specificity testing To verify the specificity of the AT assay, 42 different bacterial species and Candida sp. were examined by hybridizing 100-ng DNA aliquots to the microarray after biotinylation PCR. The complete list of the 46 strains involved is given in Supplement 1. While the vast majority of these samples showed no hybridization signals above the threshold value, minor cross-reactions were observed with salmonellae (serovar Enteritidis), mycobacteria (M. avium and M. intracellulare) and Pseudomonas aeroginosa, all of which reacted weakly with the genus-specific probe chlamydia_1. Additionally, Salmonella Enteritidis DNA partially hybridized to probe Cp_abortus_2. However, as the resulting hybridization patterns consisted only of one or two isolated signals there was no similarity to any of the species-specific patterns of chlamydiae, and none of these samples would have been identified as a false positive. Therefore, the assumption that the specificity of each of the DNA tests was 1 appears to be valid Statistical analysis Table 2 gives an overview of the various test combinations used for different groups of samples. To determine test parameters and their 95% credible intervals, two sets of samples were analyzed. Firstly, when all clinical samples (n ¼ 293) were included (Table 3a), the microarray test had a median sensitivity of 81% (75 88). This compares to the results of real-time PCR (90% [83 96]) and 16S PCR (84% [75 92%]). The ompa PCR had a median of 90% (65 100), which was not significantly different, but the IGS-L PCR had a significantly lower sensitivity of 53% (34 73). The sensitivity of 90% (70 100) of the IGS-S PCR was not significantly different from the array test. Immunohistochemistry also had a significantly lower sensitivity of 57% (41 71). There were significant correlations in the sensitivity of immunohistochemistry and the array test (gamma coefficient 0.17, lower 95 percentile 0.01, p ¼ 0.039). No other test correlations were detected. Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

99 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] 9 Secondly, when a subset of 81 swab samples was excluded from analysis because of inferior quality, the AT assay exhibited clearly improved sensitivity with 87% vs. 86% for real-time PCR and 84% for 16S PCR (Table 3b). 4. Discussion Rapid, reliable and highly informative laboratory diagnosis of chlamydioses is not only of importance for monitoring the health status of individual animals or entire flocks, but also for the prevention of zoonotic transmission to humans. Although a number of diagnostic methods for the detection of chlamydiae are available, i.e. cell culture, antigen and antibody ELISAs, conventional and real-time PCR, immunohistochemistry, immunofluorescence, etc., not all of them possess the necessary sensitivity and specificity. Veterinary diagnostic laboratories often have to deal with autolytic or formalin-fixed tissues which precludes the possibility to recover viable bacteria. Antibodies used for immunhistochemical detection of chlamydiae in formalin-fixed and paraffin-embedded tissues are often specific for the family Chlamydiaceae, so that speciesspecific diagnosis is impossible. PCR tests are known to be very sensitive, but there is no single test to allow the identification of all nine Chlamydia and Chlamydophila species. The present data show that the DNA microarray assay was capable of detecting and identifying the chlamydial species in many different specimens including autolytic material (placental tissue), formalin-fixed and paraffin-embedded tissues, as well as fresh material such as swabs, milk and feces. Although the total numbers of samples were relatively small, the data did allow for resolution of the model. The fact that up to four tests were conducted for a number of samples provided more than enough degrees of freedom to resolve the model, particularly since the experimentally infected group (n ¼ 40) had a different probability of infection than the naturally infected group. There is one correlation suggested by the model. This was a correlation in the sensitivity of the AT test and immunohistochemistry. The former detects the presence of chlamydial DNA and the latter detects antigens of the microorganism. The correlation may result from the fact that if there were relatively high levels of antigen detected by the immunohistochemistry test, this would increase the probability of detecting DNA in the sample. This may be due to the higher numbers of organisms present in immunohistochemistry-positive samples. The present results of AT testing of clinical samples from disease outbreaks (Table 1), tissue and milk samples demonstrate the suitability of this assay for the diagnostic laboratory. When the test results of all clinical samples (n ¼ 293) were included in the statistical analysis, the sensitivity of the AT test was a median of 0.81, which is comparable to the 16S PCR, a little lower than real-time PCR, but better than immunohistochemistry (Table 3a). However, when the somewhat critical subset (n ¼ 81) of nasal and conjunctival swabs from calves and sheep (Table 1, lines 42 55) was excluded, the microarray test s sensitivity was at 0.87 and completely equivalent to the other PCR tests (Table 3b). This subset of samples was included to explore the limits of the performance of the present DNA microarray test. Although cotton swabs represent a convenient and widely used sample collection tool, they usually carry only low amounts of bacterial cells from infected host tissue. According to our experience from routine diagnosis using real-time PCR, such samples typically contain 100 or less chlamydial inclusion-forming units. Consequently, only low amounts of DNA will be recovered for diagnostic examination. In the context of the present study, this meant that the AT assay was run at DNA concentrations around the detection limit [9]. Another relevant constraint was the fact that AT testing of these swab samples was conducted retrospectively, i.e. about 1 year after real-time PCR examination. Such a prolonged storage (at 20 1C) certainly led to partial DNA degradation. These complicating circumstances provide an explanation why 18 out of 81 real-time PCR-positive samples failed to react in the microarray assay. Nevertheless, we considered it worthwhile to include this panel of swab samples into the present study since the findings underline the importance of speedy sample delivery and processing. Altogether, the present data show that the DNA microarray assay for chlamydiae: (i) has the sensitivity required for routine testing of clinical samples and (ii) due to the parallel approach and high specificity, enables the diagnostician to save time by obtaining species identification in a single test within one working day. The assay s advantages will become most obvious in diagnosis of animal chlamydioses, where one has to deal with up to nine different species of the family Chlamydiaceae, some of which possess a broad host range. For instance, cattle can be infected by Cp. abortus, Cp. pecorum or Cp. psittaci, sheep can have Cp. abortus or Cp. psittaci, and swine are known to be host of C. suis, Cp. abortus, Cp. pecorum or Cp. psittaci. There are also rare species of narrow host range, such as C. muridarum and the potentially zoonotic agents Cp. felis and Cp. caviae, for which no specific test on clinical samples is available (16S PCR-RFLP works only with cell culture). On the other hand, the assay should also be of interest in diagnosis of human chlamydial infections. Although only three species are usually expected, i.e. C. trachomatis, Cp. pneumoniae and, less frequently, Cp. psittaci, the confirmation of species identity currently requires a specific PCR test to be conducted for each individual pathogen. For maximum accuracy, these tests should be run either in parallel or as a multiplex test. However, it is by no means certain that, with more detailed epidemiological research, the spectrum of chlamydiae in humans will remain confined to the three classical species. In this context, the microarray assay also represents an interesting and powerful tool that can be used to reassess the present perceptions on dissemination of chlamydial agents. Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

100 Habilitationsschrift Nicole Simona Borel 10 ARTICLE IN PRESS N. Borel et al. / Molecular and Cellular Probes ] (]]]]) ]]] ]]] Acknowledgments The authors thank Christine Grajetzki, Roseline Weilenmann, Elke Mu ller, Carmen Kaiser, Simone Bettermann, and Sabine Scharf for excellent technical assistance. The study is an integral part of the European COST Action 855 Animal Chlamydioses and the Zoonotic Implications. Appendix A. Supplementary material Supplement 1. Examination of microorganisms other than Chlamydiaceae (specificity controls). Supplementary data associated with this article can be found in the online version at doi: /j.mcp References [1] Hegde P, Qi R, Abernathy K, Rasooly A, Chizhikov V. A concise guide to cdna microarray analysis. Biotechniques 2000;29: [2] Perreten V, Vorlet-Fawer L, Slickers P, Ehricht R, Kuhnert P, Frey J. Microarray-based detection of 90 antibiotic resistance genes of grampositive bacteria. J Clin Microbiol 2005;43: [3] Monecke S, Ehricht R. Rapid genotyping of methicillin-resistant Staphylococcus aureus (MRSA) isolates using miniaturised oligonucleotide arrays. Clin Microbiol Infect 2005;11: [4] Al-Khaldi SF, Myers KM, Rasooly A, Chizhikov V. Genotyping of Clostridium perfringens toxins using multiple oligonucleotide microarray hybridisation. Mol Cell Probes 2004;18: [5] Kim BC, Park JH, Gu MB. Development of a DNA microarray chip for the identification of sludge bacteria using an unsequenced random genomic DNA hybridization method. Environ Sci Technol 2004; 38: [6] Sachse K, Hotzel H, Slickers P, Ellinger T, Ehricht R. DNA microarray-based detection and identification of Chlamydia and Chlamydophila spp. Mol Cell Probes 2005;19: [7] Roth SB, Jalava J, Ruuskanen O, Ruohola A, Nikkari S. Use of an oligonucleotide array for laboratory diagnosis of bacteria responsible for acute upper respiratory infections. J Clin Microbiol 2004;42: [8] Conejero-Goldberg C, Wang E, Yi C, Goldberg TE, Jones-Brando L, Marincola FM, et al. Infectious pathogen detection arrays: viral detection in cell lines and postmortem brain tissue. Biotechniques 2005;39: [9] Ehricht R, Slickers P, Goellner S, Hotzel H, Sachse K. Optimized DNA microarray assay allows detection and genotyping of single PCR-amplifiable target copies. Mol Cell Probes 2006;20:60 3. [10] Lutz-Wohlgroth L, Becker A, Brugnera E, Huat ZL, Zimmermann D, Grimm F, et al. Chlamydiales in guinea-pigs and their zoonotic potential. J Vet Med A Physiol Pathol Clin Med 2006;53: [11] Chanton-Greutmann H, Thoma R, Corboz L, Borel N, Pospischil A. Abortion in small ruminants in Switzerland: investigations during two lambing seasons ( ) with special regard to chlamydiae. Schweiz Arch Tierheilk 2002;144: [12] Borel N, Thoma R, Spaeni P, Weilenmann R, Teankum K, Brugnera E, et al. Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet Pathol 2006;43: [13] Pospischil A, Thoma R, Hilbe M, Grest P, Gebbers JO. Abortion in woman caused by caprine Chlamydophila abortus (Chlamydia psittaci serovar 1). Swiss Med Wkly 2002;132:64 6. [14] Blumer C, Zimmermann R, Weilenmann R, Vaughan L, Pospischil A. Chlamydia species as possible cause of mass deaths of free-ranging amphibians and frogs in Switzerland. In: Proceedings of the 3rd workshop of COST 855 on diagnosis and pathogenesis of animal chlamydioses, Siena, Italy, p. 43 [15] Soldati G, Lu ZH, Vaughan L, Polkinghorne A, Zimmermann DR, Huder JB, et al. Detection of mycobacteria and chlamydiae in granulomatous inflammation of reptiles: a retrospective study. Vet Pathol 2004;41: [16] Guscetti F, Schiller I, Sydler T, Corboz L, Pospischil A. Experimental Chlamydia psittaci serotype 1 enteric infection in gnotobiotic piglets: histopathological, immunohistochemical and microbiological findings. Vet Microbiol 1998;62: [17] Becker A, Wohlgroth L, Vaughan L, Brugnera E, Kaps S, Spiess B, et al. Chlamydial conjunctivitis in domesticated animals: preliminary results in guinea pigs, pigs and sheep. In: Proceedings of the fifth meeting of the European Society for Chlamydia Research, Budapest, Hungary, p. 317 [18] Sachse K, Grossmann E, Jager C, Diller R, Hotzel H. Detection of Chlamydia suis from clinical specimens: comparison of PCR, antigen ELISA, and culture. J Microbiol Methods 2003;54: [19] Sachse K, Hotzel H. Detection and differentiation of chlamydiae by nested PCR. Methods Mol Biol 2003;216: [20] Everett KDE, Andersen AA. Identification of nine species of the Chlamydiaceae using PCR-RFLP. Int J Syst Bacteriol 1999;49(Pt 2): [21] Everett KDE, Bush RM, Andersen AA. Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each contaning on monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int J Syst Bacteriol 1999;49: [22] Guscetti F, Hoop R, Schiller I, Corboz L, Sydler T, Pospischil A. Experimental enteric infection of gnotobiotic piglets with a Chlamydia psittaci strain of avian origin. J Vet Med B Infect Dis Vet Public Health 2000;47: [23] Brunscum AJ, Gardner IA, Johnson WO. Estimation of diagnostictest sensitivity and specificity through Bayesian modeling. Prev Vet Med 2005;68: [24] Dendukuri N, Joseph L. Bayesian approaches to modeling the conditional dependence between multiple diagnostic tests. Biometrics 2001;57: [25] Gardner IA, Stryhn H, Lind P, Collins MT. Conditional dependence between tests affects the diagnosis and surveillance of animal diseases. Prev Vet Med 2000;45: [26] Rapsch C, Schweizer G, Grimm F, Kohler L, Bauer C, Deplazes P, et al. Estimating the true prevalence of Fasciola hepatica in cattle slaughtered in Switzerland in the absence of an absolute diagnostic test. Int J Parasitol 2006;36: Please cite this article as: Borel N, et al. Direct identification of chlamydiae from clinical samples using a DNA microarray assay A validation study. Molecular Cellular Probes (2007), doi: /j.mcp

101 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Infectious Diseases BioMed Central Research article Tissue MicroArray (TMA) analysis of normal and persistent Chlamydophila pneumoniae infection Nicole Borel* 1, Sanghamitra Mukhopadhyay 2,5, Carmen Kaiser 1, Erin D Sullivan 2, Richard D Miller 3, Peter Timms 4, James T Summersgill 2,3, Julio A Ramirez 2 and Andreas Pospischil 1 Open Access Address: 1 Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 2 Division of Infectious Diseases, Department of Medicine, University of Louisville, Louisville, Kentucky, USA, 3 Department of Microbiology and Immunology, University of Louisville, Louisville, Kentucky, USA, 4 Infectious Diseases Program, School of Life Sciences, Queensland University of Technology, Brisbane, Australia and 5 Biological Defense Research Directorate, Naval Medical Research Center, Washington Avenue, Rockville, MD 20852, USA Nicole Borel* - n.borel@access.unizh.ch; Sanghamitra Mukhopadhyay - Mukhopadhyays.ctr.in@nmrc.navy.mil; Carmen Kaiser - c.kaiser@access.unizh.ch; Erin D Sullivan - edsull01@gwise.louisville.edu; Richard D Miller - rdmill01@louisville.edu; Peter Timms - p.timms@qut.edu.au; James T Summersgill - j.summersgill@louisville.edu; Julio A Ramirez - jarami01@gwise.louisville.edu; Andreas Pospischil - apos@vetpath.unizh.ch * Corresponding author Equal contributors Published: 19 October 2006 BMC Infectious Diseases 2006, 6:152 doi: / Received: 01 June 2006 Accepted: 19 October 2006 This article is available from: Borel et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: Chlamydophila pneumoniae infection has been implicated as a potential risk factor for atherosclerosis, however the mechanism leading to persistent infection and its role in the disease process remains to be elucidated. Methods: We validated the use of tissue microarray (TMA) technology, in combination with immunohistochemistry (IHC), to test antibodies (GroEL, GroES, GspD, Ndk and Pyk) raised against differentially expressed proteins under an interferon-gamma (IFN-γ) induced model of chlamydial persistence. Results: In the cell pellet array, we were able to identify differences in protein expression patterns between untreated and IFN-γ treated samples. Typical, large chlamydial inclusions could be observed in the untreated samples with all antibodies, whereas the number of inclusions were decreased and were smaller and atypical in shape in the IFN-γ treated samples. The staining results obtained with the TMA method were generally similar to the changes observed between normal and IFN-γ persistence using proteomic analysis. Subsequently, it was shown in a second TMA including archival atheromatous heart tissues from 12 patients undergoing heart transplantation, that GroEL, GroES, GspD and Pyk were expressed in atheromatous heart tissue specimens as well, and were detectable morphologically within lesions by IHC. Conclusion: TMA technology proved useful in documenting functional proteomics data with the morphologic distribution of GroEL, GroES, GspD, Ndk and Pyk within formalin-fixed, paraffinembedded cell pellets and tissues from patients with severe coronary atherosclerosis. The antibodies GroEL and GroES, which were upregulated under persistence in proteomic analysis, displayed positive reaction in atheromatous heart tissue from 10 out of 12 patients. These may be useful markers for the detection of persistent infection in vitro and in vivo. Page 1 of 8 (page number not for citation purposes) 97

102 Habilitationsschrift Nicole Simona Borel BMC Infectious Diseases 2006, 6:152 Background Chlamydophila (C.) pneumoniae is an obligate intracellular pathogen which causes both acute and chronic respiratory infections in humans [1-5]. Over the last decade, several reports in the literature have suggested that infection with C. pneumoniae may also contribute to the pathogenesis of atherosclerosis [6,7]. C. pneumoniae was detected in atheromatous lesions by isolation in pure culture, polymerase chain reaction (PCR), electron microscopy, in situ hybridization (ISH) and immunohistochemistry (IHC) [8-11]. In order to play a causative role in chronic diseases, C. pneumoniae would need to persist within infected tissue for extended periods of time, thereby stimulating a chronic inflammatory response. In vitro, an alteration of the normal developmental cycle of C. pneumoniae can be induced by interferon-γ-mediated induction of the host cell indoleamine 2,3-dioxygenase (IDO) activity, leading to a persistent form of the organism [12-15]. In addition, several other models of in vitro persistence have been described (i.e. iron-limitation and antibiotics) [16]. Nevertheless, it is unknown which genes and proteins of C. pneumoniae are involved in the development and maintenance of persistence. We have previously characterized an IFN-γ induced model of persistence by 2D gel electrophoresis [17-19], and identified several proteins that are differentially regulated during the induction of persistence. Tissue microarray (TMA) technology, developed by Kononen et al., 1998 [20] represents a promising approach in the field of proteomics for its potential usefulness in in situ analysis. Preparations for TMA are constructed by obtaining cylindrical tissue specimens from paraffin blocks, assembling several hundreds into a single block, and preparing sections containing multiple tissue specimens [20-22]. TMA sections can be analyzed using standard pathology methods, such as hematoxylin and eosin (HE) staining or special stains and in situ analyses, such as immunohistochemistry (IHC) [20,21,23-25]. The utility of TMA protocols for high-throughput expression profiling of tumors at the molecular and protein levels has been widely used in human cancer research on formalinfixed and paraffin-embedded biopsy specimens [20,21,26,27]. Since many markers of gene and protein expression are first established and studied in cell culture systems, it is useful to include cultured cells in TMAs for preliminary studies when translating experimental techniques from laboratory systems to studies of human tissue. Therefore, in the present study, we used 5 polyclonal antibodies directed against differentially regulated chlamydial proteins during in vitro persistence [18,19] to validate the use of TMA technology on non-persistently, persistently infected and uninfected HEp-2 cell pellets. In addition, archival atheromatous heart tissues [10] were tested by TMA, in combination with IHC, using the same antibodies, to determine their potential future use in detecting persistently infected tissue. Methods Cell line HEp-2 cells (ATCC CCL-23) were obtained from the American Type Culture Collection (Rockville, MD) and maintained in Iscove's Maintenance Medium (IMM) (Cellgro, Washington, DC), as described previously [28]. Bacterial isolate C. pneumoniae A-03 (ATCC VR-1452), previously isolated from an atheroma of a patient with coronary artery disease during heart transplantation at the Jewish Hospital Heart and Lung Institute, Louisville, KY [10], were propagated in HEp-2 cell monolayers in Iscove's Growth Medium (IGM), as described previously [28]. Elementary bodies (EBs) were harvested and purified by disruption of HEp-2 cell monolayers with a cell scraper, sonication and centrifugation over a renografin density gradient [28]. EB suspensions were stored in sucrose-phosphate-glutamic acid buffer at -80 C, after which viable titers were established using standard methods. Patients specimens Archival atheromatous tissue specimens from twelve patients undergoing heart transplantation were investigated. Results from culture, PCR, IHC, ISH, EM and serology testing have been described previously [10]. The study of Ramirez et al. was approved by the Institutional Review Boards (IRB) at both the University of Louisville and Jewish Hospital Healthcare Corporation. Preparation of antibodies Five proteins were identified as being differentially regulated in the IFN-γ-induced model of persistence [17,19]: (i) GroEL (60 kda chaperonin) and (ii) GroES (10 kda chaperonin) are both chaperons involved in protein folding, assembly and modification, (iii) GspD (general secretion protein D) involved in general protein secretion, (iv) Ndk (nucleoside-2-phosphate-kinase) involved in base and nucleotide metabolism of amino acid biosynthesis, and (v) Pyk (pyruvate kinase) involved in the energy metabolism (glycolysis and gluconeogenesis). GroEL was analysed and quantitated previously [17], and the remaining four proteins were selected and confirmed in an identical fashion [19]. GroEL and GroES were upregulated under persistence, whereas GspD and Pyk remained unchanged and Ndk was downregulated. Each protein was plugged from a silver-stained 2D gel, which was obtained from a purified EB preparation [18,19,28], and used for antibody production of polyclonal antibodies at Page 2 of 8 (page number not for citation purposes) 98

103 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Infectious Diseases 2006, 6:152 Harlan Bioproducts for Science, Inc. (Indianapolis, IN, USA). Pathogen-free, barrier-raised New Zealand white rabbits were immunized four times with the antigens, and sera from the final bleed were used in this study. The specificity of each antibody was confirmed in our laboratory by 2D gel electrophoresis of a purified C. pneumoniae EB preparation followed by western blot analysis, demonstrating specific reactivity on the blot which corresponded to the molecular weight and iso-electric point of each individual protein (data not shown). Other antibodies used The following antibodies were also used in these studies. (i) Chlamydiaceae family-specific mouse monoclonal antibody directed against the chlamydial lipopolysaccharide (mlps; Clone ACI-P, Progen, Heidelberg, Germany). (ii) Chlamydiaceae family-specific rabbit polyclonal antibody directed against both the chlamydial LPS and the chlamydial major outer membrane protein (MOMP) (plps/momp; Cygnus Technologies, Inc., Southport, NC). TMA analysis Infection protocol HEp-2 cells were grown in 75-cm 2 cell culture flasks (Costar, Cambridge, MA) to confluency and inoculated with purified C. pneumoniae EB ( IFUs/flask) in IGM with or without human recombinant IFN-γ (50 and 100 U per ml), followed by centrifugation at 675 g (Sorvall TR 6000D) for 30 min at 10 C and incubating at 37 C in 5% CO 2 for 24, 48 and 72 hpi. After the respective incubation period, the medium was aspirated and the monolayers were washed twice with 1 phosphate buffer saline (PBS). Monolayers were fixed with 4% formalin in 1 PBS for 10 min followed by two washes with 1 PBS. All monolayers were harvested from the flasks with a cell scraper and transferred into 15 ml of 1 PBS. After centrifugation at 675 g for 10 min at 20 C, the supernatant was discarded. The pellets were re-suspended in 5% BSA prepared in 1 PBS and transferred to an Eppendorf tube (Eppendorf-Netheler-Hinz GmbH, Hamburg, Germany) with one drop of hematoxylin for visualization. The cell suspensions were centrifuged for 5 min at 950 g and the supernatant was discarded. The resulting pellets were then embedded in paraffin using the Cytoblock cell block preparation system (Shandon, Pittsburg, USA). For each condition (timepoints 24, 48 and 72 h, concentrations of 50 U/ml and 100 U/ml of IFN-γ) four sets of infected monolayers were prepared. As controls, four sets of uninfected HEp-2 monolayers without and with 100 U/ml IFN-γ, were prepared as pellets. TMA setup Two equal cell culture array blocks including four equal prepared sets of cell pellets were created with the TMA Builder from Histopathology Ltd., Hungary according to the instructions of the manufacturer. Briefly, the recipient paraffin block with 24 holes of diameter 2 mm each arranged in four columns and six rows was moulded with the TMA Builder. The whole cell pellets from the donor blocks were punched out with the Paraffin-Punch-Extractor and were arrayed in the preformed recipient paraffin block according to protocol. Formalin-fixed and paraffin-embedded coronary artery specimens were available from the 12 patients (1 to 7 paraffin blocks per patient). Three TMA blocks including each tissue of the 12 patients, an uninfected and an infected control HEp-2 cell pellet (72 hpi, without IFN-γ) were created in an identical fashion. 4 µm slide were cut using a standard microtome. Immunohistochemistry Paraffin sections were stained with the following primary anti-chlamydial-antibodies: (1) mlps at a dilution of 1:50, (2) plps/momp at a dilution of 1:400, (3) GroEL, GroES, GspD, Ndk and Pyk at a dilution of 1:200. These optimal dilutions were previously determined. Detection was performed with the Detection Kit (Dako ChemMate Detection Kit, Glostrup, Denmark) according to the manufacturer's instructions. Antigen retrieval was performed by one min enzyme digestion (Pronase) (mlps and plps/ MOMP) and microwave heating (750 W for 10 minutes) two times in citrate buffer (ph 6,0; Target Retrieval Solution ( 10), Dako ChemMate, Glostrup, Denmark) (GroEL, GroES, GspD, Ndk and Pyk), respectively. For inhibition of the endogenous peroxidase activity, the slides were immersed in peroxidase-blocking solution (Dako ChemMate, Glostrup, Denmark) for 5 min at room temperature (RT). Two additional blocking solutions were added to the slides, which were incubated with the polyclonal antibodies GroEL, GroES, GspD, Ndk and Pyk: Dako Protein Block Serum-free for 5 min at room temperature (Dako ChemMate, Glostrup, Denmark) and 20 min Avidin D solution followed by 20 min Biotin solution at room temperature (Vector). The slides were incubated with the primary antibody for 60 min (mlps and plps/momp) or over night (GroEL, GroES, GspD, Ndk and Pyk) at room temperature in a moist chamber. In total, the IHC for each antibody was repeated at least four times on consecutive sections. Results Cell pellet array Results with the monoclonal antibody directed against LPS (mlps) and the polyclonal antibody directed against chlamydial LPS and MOMP (plps/momp) were similar. Page 3 of 8 (page number not for citation purposes) 99

104 Habilitationsschrift Nicole Simona Borel BMC Infectious Diseases 2006, 6:152 Typical large, uniformly-staining chlamydial inclusions, located near the host cell nucleus, could be seen at 48 and 72 hpi in the untreated samples with the mlps (Figure 1) and the plps/momp antibody. Results with the 5 polyclonal antisera tested, showed similar staining patterns in the untreated samples, however, not all inclusions were stained uniformly positive (Figure 1). Earlier in the chlamydial developmental cycle at 24 h, granular positive material was seen in the cytoplasm of untreated cell pellets with the mlps antibody and the plps/momp antibody, as well as the GroEL, GroES and Pyk antibodies, and to a lesser extent, with the GspD (data not shown). All pellets at 24 hpi were negative with the Ndk antibody. The IFN-γ-treated samples displayed an overall decrease in positive reactivity. At 48 and 72 hpi, the number of inclusions were decreased and were obviously smaller and atypical in morphology (Figure 1). This could be observed for all antibodies, except for Ndk, which was negative in all IFN-γ-treated samples (Figure 1). There was no difference in inclusion morphology and antibody staining patterns seen between the two concentrations of IFN-γ (50 U/ ml or 100 U/ml). Overall, the assay showed good reproducibility in all four replicates of monolayers. TMA with atheromatous heart tissues Granular, positive-staining material was seen in the cytoplasm of subintimal macrophages and smooth muscle cells in the medial part of the affected coronary arteries of all patients with at least one antibody, except the two patients with negative results for the presence of C. pneumoniae in the original study (patients # 7 and 11) [10]. Patients # 10 and 12 were positive with all antibodies used except of the Ndk, which was negative in all 12 heart tissue specimens. Patients # 1, 2, 3 and 9 were positive with the GroEL, GroES, GspD and Pyk, whereas patients # 4 and 6 were only positive with the GroEL, GspD and Pyk. Patient # 5 was positive with the GroES and Pyk, and patient # 8 revealed positive reaction with the GroEL and GspD. All patients were negative when tested with the mlps, whereas the plps/momp ab revealed positive reactions in patients # 10 and 12. Positive staining of the antibodies GroEL and GroES in patient # 10 and GspD and Pyk in patient # 12, respectively, is displayed in figure 2. Discussion In this study, TMA technology was useful in documenting functional proteomics data showing the morphologic distribution of GroEL, GroES, GspD, Ndk and Pyk within formalin-fixed, paraffin-embedded cell pellets and atheromatous heart tissues. GroEL, GroES, GspD, Ndk and Pyk, selected as differentially regulated proteins from proteomic analysis, were expressed in C. pneumoniaeinfected, untreated and IFN-γ treated HEp-2 cells and atheromatous heart tissues and were detectable morphologically by IHC. In general, IHC allows the detection of the presence of an antigen in tissue sections, however, the intensity of antigen labelling does not correlate with the amount of antigen present. The IHC labelling evaluated in this study, therefore, represents the presence or absence of proteins, but does not reflect quantitative expression of GroEL, GroES, GspD, Ndk, Pyk, mlps and/or plps/ MOMP. By using cell pellet array, in combination with IHC, it was possible to identify the number, size and morphology of the chlamydial inclusions. Numbers of inclusions at 48 hpi and 72 hpi were decreased under IFN-γ-persistence and the inclusions were smaller and atypical, as described previously [15], however, the staining intensity remained the same as that seen in untreated cells. The staining results obtained with the TMA method were generally similar to the changes observed between normal and IFN-γ persistence using proteomic analysis [17-19]. For example, GroEL, which was upregulated in proteomic analysis, displayed more positive reactions in the IFN-γ treated samples when compared to GspD, which remained unchanged in the proteomic analysis. Likewise, Ndk reactivity remained negative in all IFN-γ-treated samples by TMA, which corresponded to the downregulation observed in the proteomic analysis. In comparison to other methods, such as PCR, we were able to localize and visualize the positive reactions within the atheromatous heart tissue by TMA, in combination with IHC. The TMA method allowed a comparison between multiple cell pellets or tissue specimens on one microscope slide. The 2 mm punch, in combination with the Tissue Microarray Builder allowed easy manipulation during punching, facilitating a rapid preparation of the cell culture array block. The fact that the 2 mm punch can cause significant damage to the donor block was not of importance because each donor block contained only one single cell pellet that was entirely punched out and the whole area of each pellet could be examined. Other punch sizes such as 0.6 mm and 1.2 mm cause less damage to the original donor block and make it possible to array several hundreds of specimens on one single block, however the creation of this block is more laborious. The problem of reduction of the amount of tissue analyzed from the whole cell pellet to a disk, which may be not be representative of the protein expression of the entire tissue specimen, can be solved by performing the experiments in several-fold redundancy. For that reason, we created two equal cell pellet blocks including four sets of equally prepared monolayers containing the three timepoints (treated and untreated cell pellets and negative controls). Page 4 of 8 (page number not for citation purposes) 100

105 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Infectious Diseases 2006, 6:152 cell pellet array Figure 1 cell pellet array. Photomicrographs of TMA preparations of whole cell pellets showing differential expression pattern of C. pneumoniae proteins under untreated (A) and 50 U/ml IFN-γ treated (B) conditions at 48 hpi. Representative monolayers are shown to depict distinct differences in morphology and size of inclusions, as they were smaller and atypical under IFN-γinduced persistence in comparison to the untreated monolayers. Page 5 of 8 (page number not for citation purposes) 101

106 Habilitationsschrift Nicole Simona Borel BMC Infectious Diseases 2006, 6:152 TMA with atheromatous heart tissues Figure 2 TMA with atheromatous heart tissues. Photomicrographs of TMA preparations of atheromatous heart tissues of patient # 10 showing positive reaction within macrophages with GroEL (1) and with GroES (2) and in patient # 12 showing positive reaction within smooth muscle cells with GspD (3) and Pyk (4). Multiple sections of the two TMA blocks were cut and probed with each antibody. Multiple sections of 3 TMA blocks, including the atheromatous heart tissue, were investigated with each antibody. In our previous study, patients # 1, 2, 3, 6 and 12 were positive by IHC using Chlamydia-specific and C. pneumoniae-specific antibodies [10]. We were able to detect 5 more positive patients with experimentally produced anti- GroEL, anti-groes, anti-gspd and anti-pyk antibodies. Thus, differentially regulated proteins by proteomic analysis were expressed in C. pneumoniae-infected human atheromatous heart tissue specimens and were detectable morphologically within lesions by IHC. The staining results obtained with the TMA corresponded to the reactivity as determined by proteomic analysis, for example, GroEL and GroES, which were upregulated under persistence in proteomic analysis, were likewise positive in most heart tissue specimens. Patients # 7 and 11, which underwent heart transplantation primarily due to myocarditis, rather than severe atherosclerosis (unpublished data), displayed questionable positive results (GroEL, GroES, GspD and Pyk). As these 2 patients were in fifth decade of life and had elevated anti-c. pneumoniae titers in the microim- Page 6 of 8 (page number not for citation purposes) 102

107 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations BMC Infectious Diseases 2006, 6:152 munofluorescence assay [10], we tend to assume that they suffered from atherosclerosis and were Chlamydiainfected as patients # 1 6, 10 and 12, but perhaps to a lesser extent. Definitive demonstration of chlamydial particles in patients # 7 and 11 by more sensitive techniques (i.e. ultrastructural studies) is in progress [29]. Conclusion Overall, GroEL and GroES tend to be useful markers to detect persistent infection in vitro and in vivo. In addition, GspD and Pyk antibodies gave similar reactivity, indicating appropriate sensitivity and specificity, and may also allow the detection of C. pneumoniae in chronicallyinfected tissue. Antibody prepared against Ndk remained negative in tissue specimens from all 12 archived atheromatous tissue specimens, which corresponded nicely to the downregulation observed in proteomic analysis. These data represent the first thorough examination of atheromatous tissue by experimentally produced antibodies against C. pneumoniae and may provide a useful technique to further define the role of this organism in atherosclerosis and other chronic human diseases. Competing interests The author(s) declare that they have no competing interests. Authors' contributions NB carried out the TMA analysis and drafted the manuscript. SM performed the cell culture infections and the 2 D gel electrophoresis. CK performed the IHC. EDS participated in the cell culture infections and 2 D gel electrophoresis. RDM and PT participated in the coordination of the study and helped to draft the manuscript. JTS and AP participated in the design of the study. JAR provided the atheromatous heart tissue specimens. All authors read and approved the final manuscript. Acknowledgements We would like to thank Dr. Urs Ziegler and Claudia Dumrese, from the Institute of Anatomy, University of Zurich, Switzerland, for providing C. pneumoniae-infected HEp2 cells. This work was supported by grants from the NHMRC (PT) and the National Institutes of Health (HL68874 and A151255) (JTS). References 1. Grayston JT, Kuo CC, Wang SP, Altman J: A new Chlamydia psittaci strain, TWAR, isolated in acute respiratory tract infections. N Eng J Med 1986, 315: Grayston JT: Infections caused by Chlamydia pneumoniae strain TWAR. Clin Infect Dis 1992, 15: Hammerschlag MR, Chirgwin K, Roblin PM, Gelling M, Dumornay W, Mandel L, Smith P, Schachter J: Persistent infection with Chlamydia pneumoniae following acute respiratory illness. Clin Infect Dis 1992, 14: Kuo CC, Jackson LA, Campbell LA, Grayston JT: Chlamydia pneumoniae (TWAR). Clin Microbiol Rev 1995, 8: Peeling RW, Brunham RC: Chlamydiae as pathogens: new species and new issues. Emerg Infect Dis 1996, 2: Saikku P, Mattila K, Nieminen MS, Makela PH, Huttunen JK, Valtonen V: Serological evidence of an association of a novel Chlamydia, TWAR, with chronic coronary heart disease and acute myocardial infarction. Lancet 1988, ii: Saikku P, Leinonen M, Tenkanen L, Ekman MR, Linnanmaki E, Manninen V, Manttari M, Frick MH, Huttunen JK: Chronic Chlamydia pneumoniae infection as a risk factor for coronary heart disease in the Helsinki Heart Study. Ann Intern Med 1992, 116: Kuo CC, Shor A, Campbell LA, Fukushi H, Patton DL, Grayston JT: Demonstration of Chlamydia pneumoniae in atherosclerotic lesions of coronary arteries. J Infect Dis 1993, 167: Campbell LS, O'Brien ER, Cappuccio AL, Kuo CC, Wang SP, Stewart D, Patton DL, Cummings PK, Grayston JT: Detection of Chlamydia pneumoniae TWAR in human coronary atherectomy tissues. J Infect Dis 1995, 172: Ramirez JA, Ahkee S, Summersgill JT, Ganzel BL, Ogden LL, Quinn TC, Gaydos CA, Bobo LL, Hammerschlag MR, Roblin PM, LeBar W, Grayston JT, Kuo CC, Campbell LA, Patton DL, Dean D, Schachter J: Isolation of Chlamydia pneumoniae from the coronary artery of a patient with coronary atherosclerosis. Ann Intern Med 1996, 125: Jackson LA, Campbell LA, Kuo CC, Rodríguez KI, Lee A, Grayston JT: Isolation of Chlamydia pneumoniae from a carotid endarterectomy specimen. J Infect Dis 1997, 176: Summersgill JT, Sahney NN, Gaydos CA, Quinn TC, Ramirez JA: Inhibition of Chlamydia pneumoniae growth in HEp-2 cells pretreated with gamma interferon and tumor necrosis factor alpha. Infect Immun 1995, 36: Holmes EW: Expression and regulation of interferon-gammainduced tryptophan catabolism in cultured skin fibroblasts. J Interferon Cytokine Res 1998, 18: Mehta SJ, Miller RD, Ramirez JA, Summersgill JT: Inhibition of Chlamydia pneumoniae replication in HEp-2 cells by interferon-γ: role of tryptophan catabolism. J Infect Dis 1998, 177: Pantoja LG, Miller RD, Ramirez JA, Molestina RE, Summersgill JT: Characterization of Chlamydia pneumoniae persistence in HEp-2 cells treated with gamma interferon. Infect Immun 2001, 69: Hogan RJ, Mathews SA, Mukhopadhyay S, Summersgill JT, Timms P: Chlamydial persistence: beyond the biphasic paradigm. Infec Immun 2004, 72: Molestina RE, Klein JB, Miller RD, Pierce WH, Ramirez JA, Summersgill JT: Proteomic analysis of differentially expressed Chlamydia pneumoniae genes during persistent infection of HEp-2 cells. Infect Immun 2002, 70: Mukhopadhyay S, Miller RD, Summersgill JT: Analysis of altered protein expression patterns of Chlamydia pneumoniae by integrated Proteome-Works System. J Proteome Res 2004, 3: Mukhopadhyay S, Miller RD, Sullivan ED, Theodoropoulos C, Mathews SA, Timms P, Summersgill JT: Protein expression profiles of Chlamydia pneumoniae: persistent vs. heat shock stress response. Infect Immun in press. 20. Kononen J, Bubendorf L, Kallioniemi A, Barlund M, Schraml P, Leighton S, Torhorst J, Mihatsch MJ, Sauter G, Kallioniemi OP: Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 1998, 4: Bubendorf L: High-throughput microarray technologies: from genomics to clinics. Eur Urol 2001, 40: Bubendorf L, Nocito A, Moch H, Sauter G: Tissue microarray (TMA) technology: miniaturized pathology archives for highthroughput in situ studies. J Pathol 2001, 195: Moch H, Kononen T, Kallioniemi OP, Sauter G: Tissue microarrays: what will they bring to molecular and anatomic pathology? Adv Anat Pathol 2001, 8: Simon R, Mirlacher M, Sauter G: Tissue microarrays. Biotechniques 2004, 36: Van De Rijn M, Gilks CB: Applications of microarrays to histopathology. Histopathology 2004, 44: Schraml P, Kononen J, Bubendorf L, Moch H, Bissig H, Nocito A, Mihatsch MJ, Kallioniemi OP, Sauter G: Tissue microarrays for gene amplification surveys in many different tumor types. Clin Cancer Res 1999, 5: Page 7 of 8 (page number not for citation purposes) 103

108 Habilitationsschrift Nicole Simona Borel BMC Infectious Diseases 2006, 6: Camp RL, Charette LA, Rimm DL: Validation of tissue microarray technology in breast carcinoma. Lab Invest 2000, 80: Mukhopadhyay S, Clark AP, Sullivan ED, Miller RD, Summersgill JT: Detailed protocol for the purification of Chlamydia pneumoniae elementary bodies. J Clin Microbiol 2004, 42: Borel N, Summersgill JT, Mukhopadhyay S, Kaiser C, Nufer L, Miller RD, Ramirez JA, Pospischil A: Persistent Chlamydophila pneumoniae in human coronary atherosclerotic tissue: Tissue Microarray (TMA) analysis and ultrastructural study. In Proceedings of the Eleventh International Symposium on Human Chlamydial Infections: June 2006; Niagara-on-the-Lake, Ontario, Canada Edited by: Max Chernesky, Harlan Caldwell, Gunna Christiansen, Ian N Clarke, Bernhard Kaltenboeck, Charles Knirsch, Cho-Chou Kuo, James Mahony, Roger G Rank, Pekka Saikku, Julius Schachter, Walter E Stamm, Richard S Stephens, James T Summersgill, Peter Timms, Priscilla B Wyrick. San Francisco, CA 94110, USA; 2006: Pre-publication history The pre-publication history for this paper can be accessed here: pub Publish with BioMed Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours you keep the copyright Submit your manuscript here: BioMedcentral Page 8 of 8 (page number not for citation purposes) 104

109 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 5. Summary and Future Work The studies comprised in this work have focussed on serology, epidemiology and diagnostics of chlamydial abortion in ruminants. This has resulted in significant new findings that have been published during the last four years. In chapter 1, a brief summary on the Chlamydiales with special regard to chlamydial abortion in ruminants was given. In chapter 2, serological investigations in small ruminants in Switzerland revealed a high seroprevalence in the canton Graubünden. In this canton, economic losses due to chlamydial abortion in small ruminants are significantly higher than in the other cantons. Large flocks, extensive animal husbandry and mixing of different sheep flocks during summer time in the mountains promote the spread of disease. However, the seroprevalence in the canton Valais with similar management systems is much lower indicating other unknown factors favouring chlamydial abortion in small ruminants in the canton Graubünden. As wild ruminants are grazing on the same pastures as domestic sheep and goats in the mountains, the question raises if they could play a role in infectious spread and transmission of chlamydiosis. The aim of a future project is thus to define the prevalence of chlamydiae in wild ruminants (deer, chamois, alpine ibex) in the canton Graubünden. The range of chlamydial species and in particular the detection of Chlamydophila abortus in wild ruminants will elucidate their potential role as a reservoir for domestic ruminants and the potential zoonotic risk for humans having contact to wild ruminants (i.e. hunters, gamekeepers). Another aim of chapter 2 was to investigate the influence of vaccination against OEA on serological titers in sheep. The recent field study, the first of its kind, compared antibody titers in latently infected, diseased and vaccinated animals over a two year period. Several conclusions were drawn out of that study: a) anti-momp titers did not decline significantly in any group in the two year period, reconfirming previous observations after experimentally induced chlamydial abortion b) anti-momp titers elicited by vaccination with the live vaccine were individual and animal-specific and c) these titers were comparable to the titers observed in naturally infected animals. Moreover, the study highlighted the urgent need for assays capable of discriminating between infected and vaccinated animals. Analysis of the Cp. abortus proteome for detecting possible differences between the vaccinal strain and the abortion strain is planned in future projects. The first part of chapter 3 focussed on the significance of the venereal transmission of chlamydial infections in ruminants. PCR assays targeting two different chlamydial genes have been evaluated during the study. The most appropriate PCR method has been applied to investigate semen samples from bulls, rams and bucks. In Switzerland, the importance of venereal transmission seems to be low. Chlamydia-related abortions in cattle from Graubünden were investigated in the second part of chapter 3. In contrast to the situation in small ruminants in the canton Graubünden (chapter 2), bovine abortion due to Cp. abortus seems not to play an important role in Switzerland. The further investigation of abortion cases by 16S broadrange PCR resulted in the first description of Parachlamydia in bovine abortion. New methods to diagnose these recently described Chlamydia-like organisms have been developed during our studies. Sensitive and specific real-time PCR for different Chlamydia-like organisms as well as speciesspecific immunohistochemistry protocols for Parachlamydia and Waddlia have been validated on field samples. A future project investigating retrospectively sampled abortion material (placentae and fetal organs) from small 105

110 Habilitationsschrift Nicole Simona Borel ruminants for Chlamydia-like organisms will be soon accomplished. Prospective studies on 200 ruminant placentae and fetuses and on 200 human placentae suffering from miscarriage and preterm labour in collaboration with Gilbert Greub s group in Lausanne will further investigate the role of Chlamydia-like organisms in ruminant abortion and their zoonotic potential. New diagnostic tools developed during these projects may be used in the future for broad European epidemiological surveys. In chapter 4, the DNA microarray method has been validated for diagnostic purposes. The main advantage of using new molecular techniques such as the ArrayTube Microarray over cell culture for diagnosing chlamydial infections, other than the shorter time it takes to complete the test, is that they are not dependent on viability of Chlamydia, detecting both viable and non-viable EBs. The DNA microarray enables a fast and reliable diagnosis of chlamydial infections in different types of samples originating from a diversity of animal species. This assay has been thus accredited in the Institute of Veterinary Pathology as a new diagnostic tool for chlamydial infections. Furthermore it will be applied in further research projects to investigate: the prevalence of chlamydiae in wild ruminants as a reservoir for domestic ruminants and humans (see chapter 2) the prevalence of Chlamydophila psittaci in wild birds as a potential risk for domestic poultry, pet birds and public health in Switzerland. about the prevalence of the agent in the wild bird population. This lack of knowledge leads us to examine the occurence of Chlamydophila psittac in the Swiss wild bird population. Available samples from the surveillance for avian influenza in wild birds will be examined by microarray method for Chlamydophila psittaci and its genotypes to get insight into the epidemiological situation of ornithosis in wild birds in Switzerland. The results of the study will elucidate the potential risk for domestic poultry and public health due to ornithosis in wild birds in Switzerland and will show the potential need of preventive measurements. In the second part of chapter 4, the TMA method was established to study normal and persistent Cp. pneumoniae infection. The TMA technique will be further used in a prospective study on human transplanted hearts in collaboration with Prof. Summersgill, University of Louisville. Immunohistochemistry protocols for new antichlamydial antibodies will be established using TMA blocks. The TMA technology will be applied to test antisera raised against different Chlamydia-like organisms for their future use on animal tissue samples. There is an urgent need for more research on prevalence, significance, epidemiology and zoonotic potential of Chlamydiaceae and Chlamydia-like organisms in ruminants and other animals species studies to cover these research topics are planned. For that purpose, a new Microarray chip for genotyping of Chlamydophila psittaci, the agent of ornithosis/psittacosis in birds, has been developed recently. Ornithosis/psittacosis is a notifiable disease in Switzerland occuring world-wide in 375 different bird species. The different strains of Chlamydophila psittac in birds are named as genotypes A through F and have markedly different virulence traits and zoonotic potentials. To date, little is known 106

111 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 6. Literature Aitken ID. Ovine chlamydial abortion. In: Rickettsial and Chlamydial Diseases of Domestic Animals. Edited by Woldehiwet Z and Ristic M. Oxford: Pergamon Press. 1993: Anderson IE, Herring AJ, Jones GE, Low JC, Greig A. Development and evaluation of an indirect ELISA to detect antibodies to abortion strains of Chlamydia psittac in sheep sera. Vet Microbiol. 1995;43(1):1-12. Anderson IE, Baxter SI, Dunbar S, Rae AG, Philips HL, Clarkson MJ, Herring AJ. Analyses of the genomes of chlamydial isolates from ruminants and pigs support the adoption of the new species Chlamydia pecorum. Int J Syst Bacteriol. 1996;46(1): Anthony RM, Brown TJ, French GL. Rapid diagnosis of bacteremia by universal amplification of 23S ribosomal DNA followed by hybridization to an oligonucleotide array. J Clin Microbiol. 2000;38(2): Ball L, Griner LA, Carroll EJ. The bovine seminal vesiculitis syndrome. Am J Vet Res. 1964;25: Baud D, Thomas V, Arafa A, Regan l, Greub G. Waddlia chondrophila, a potential agent of human fetal death. Emerg Infect Dis. 2007;13(8): Borel N, Doherr MG, Vretou E, Psarrou E, Thoma R, Pospischil A. Ovine enzootic abortion: seroprevalence in Switzerland using a competitive enzyme-linked immunosorbent assay (celisa). Schweiz Arch Tierheilkd. 2002;144(9): German. Borucki MK, Krug MJ, Muraoka WT, Call DR. Discrimination among Listeria monocytogenes isolates using a mixed genome DNA microarray. Vet Microbiol. 2003;92(4): Brade H, Brade L, Nano FE. Chemical and serological investigations on the genus-specific lipopolysaccharide epitope of Chlamydia. Proc Natl Acad Sci USA. 1987;84(8): Bubendorf L. High-throughput microarray technologies: from genomics to clinics. Eur Urol. 2001;40(2): Bubendorf L, Nocito A, Moch H, Sauter G. Tissue microarray (TMA) technology: miniaturized pathology archives for high-throughput in situ studies. J Pathol 2001;195(1): Buendia AJ, Cuello F, Del Rio L, Gallego MC, Caro MR, Salinas J. Field evaluation of a new commercially available ELISA based on recombinant antigen on diagnosing Chlamydophila abortus (Chlamydia psittaci serotype 1) infection. Vet Microbiol. 2001;78(3): Buxton D. Potential danger to pregnant women of Chlamydia psittaci from sheep. Vet Rec. 1986;118(18): Buxton D, Barlow RM, Finlayson J, Anderson IE, Mackellar A. Observations on the pathogenesis of Chlamydia psittaci infection of pregnant sheep. J Comp Pathol. 1990;102(2): Call DR, Brockman FJ, Chandler DP. Detecting and genotyping Escherichia coli O157:H7 using multiplexed PCR and nucleic acid microarrays. Int J Food Microbiol. 2001;67(1-2): Cavirani S, Cabassi CS, Donofrio G, De Iaco B, Taddei S, Flammini CF. Association between Chlamydia psittaci seropositivity and abortion in Italian dairy cows. Prev Vet Med. 2001;50(1-2):

112 Habilitationsschrift Nicole Simona Borel Cevenini R, Moroni A, Sambri V, Perini S, La Placa M. Serological response to chlamydial infection in sheep, studied by enzyme-linked immunosorbent assay and immunoblotting. FEMS Microbiol Immunol. 1989;1(8-9): Chanton-Greutmann H, Thoma R, Corboz L, Borel N, Pospischil A. Abortion in small ruminants in Switzerland: investigations during two lambing seasons ( ) with special regard to chlamydial abortions. Schweiz Arch Tierheilkd. 2002;144(9): German. Chizhikov V, Rasooly A, Chumakov K, Levy DD. Microarray analysis of microbial virulence factors. Appl Environ Microbiol. 2001;67(7): Corsaro D, Greub G. Pathogenic potential of novel Chlamydiae and diagnostic approaches to infections due to these obligate intercellular bacteria. Clin Microbiol Rev. 2006;19(2): Dilbeck PM, Evermann JF, Crawford TB, Ward AC, Leathers CW, Holland CJ, Mebus CA, Logan LL, Rurangirwa FR, McGuire TC. Isolation of previously undescribed Rickettsia from an aborted bovine fetus. J Clin Microbiol. 1990;28(4): Donn A, Jones RE, Ruiu A, Ladu M, Machell J, Stancanelli A. Serological diagnosis of chlamydial abortion in sheep and goats: comparison of the complement fixation test and an enzyme-linked immunosorbent assay employing solubilised proteins as antigen. Vet Microbiol. 1997;59(1): Everett KD. Chlamydia and Chlamydiales: more than meets the eye. Vet Microbiol. 2000;75(2): Everett KD, Bush RM, Andersen AA. Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int J Syst Bacteriol. 1999;49(Pt 2): Fukushi H, Hirai K. Proposal of Chlamydia pecorum sp. nov. for Chlamydia strains derived from ruminants. Int J Syst Bacteriol. 1992;42(2): Greub G, Raoult D. Parachlamydiaceae: potential emerging pathogens. Emerg Infect Dis. 2002;8(6): Griffiths PC, Plater JM, Horigan MW, Rose MP, Venables C, Dawson M. Serological diagnosis of ovine enzootic abortion by comparative inclusion immunofluorescence assay, recombinant lipopolysaccharide enzyme-linked immunosorbent assay, and complement fixation test. J Clin Microbiol. 1996;34(6): Gut-Zangger P, Vretou E, Psarrou E, Pospischil A, Thoma R. Chlamydia abortion in sheep: possibilities of serological diagnosis using a competitive ELISA and insight into the epidemiologic situation in Switzerland. Schweiz Arch Tierheilkd. 1999;141(8): German. Hässig M, Lubsen J. Relationship between abortions and seroprevalences to selected infectious agents in dairy cows. Zentralbl Veterinarmed B. 1998;45(7): Hässig M, Eggenberger E, Künzle S, Rüsch P. Reassessment of the herd consultation in facilities with accumulated abortions in cattle. Schweiz Arch Tierheilkd. 2000;142(2): German. Henning K, Schares G, Granzow H, Polster U, Hartmann M, Hotzel H, Sachse K, Peters M, Rauser M. Neospora caninum and Waddlia chondrophila strain 2032/99 in a septic stillborn calf. Vet Microbiol. 2002;85(3): Hinton DG, Shipley A, Galvin JW, Harkin JT, Brunton RA. Chlamydiosis in workers at 108

113 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations a duck farm and processing plant. Aust Vet J. 1993;70(5): Hoelzle LE, Hoelzle K, Wittenbrink MM. Recombinant major outer membrane protein (MOMP) of Chlamydophila abortus, Chlamydophila pecorum, and Chlamydia suis as antigens to distinguish chlamydial species-specific antibodies in animal sera. Vet Microbiol. 2004;103(1-2): Hoorfar J, Cook N. Critical aspects of standardization of PCR. In: Methods in Molecular Biology: PCR Detection of Microbial Pathogens. Edited by Sachse K and Frey J. Totowa: Humana Press. 2002: Jee J, DeGraves FJ, Kim T, Kaltenboeck B. High prevalence of natural Chlamydophila species infection in calves. J Clin Microbiol. 2004;42(12): Jones GE, Low JC, Machell J, Armstrong K. Comparison of five tests for the detection of antibodies against chlamydial (enzootic) abortion of ewes. Vet Rec. 1997;141(7): Kaltenboeck B, Kousoulas KG, Storz J. Structures of and allelic diversity and relationships among the major outer membrane protein (ompa) genes of the four chlamydial species. J Bacteriol. 1993;175(2): Kaltenboeck B, Heard D, DeGraves FJ, Schmeer N. Use of synthetic antigens improves detection by enzyme-linked immunosorbent assay of antibodies against abortigenic Chlamydia psittac in ruminants. J Clin Microbiol. 1997;35(9): Kaltenboeck B, Hehnen HR, Vaglenov A. Bovine Chlamydophila spp. infection: do we underestimate the impact on fertility? Vet Res Commun. 2005;29 Suppl 1:1-15. Kampinga GA, Schröder FP, Visser IJ, Anderson JM, Buxton D, Möller AV. Lambing ewes as a source of severe psittacosis in a pregnant woman. Ned Tijdschr Geneeskd. 2000;144(52): Dutch. Kononen J, Bubendorf L, Kallioniemi A, Bärlund M, Schraml P, Leighton S, Torhorst J, Mihatsch MJ, Sauter G, Kallioniemi OP. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med. 1998;4(7): Livingstone M, Entrican G, Wattegedera S, Buxton D, McKendrick IJ, Longbottom D. Antibody responses to recombinant protein fragments of the major outer membrane protein and polymorphic outer membrane protein POMP90 in Chlamydophila abortusinfected pregnant sheep. Clin Diagn Lab Immunol. 2005;12(6): Longbottom D. Chlamydial infections of domestic ruminants and swine: new nomenclature and new knowledge. Vet J. 2004;168(1):9-11. Longbottom D, Psarrou E, Livingstone M, Vretou E. Diagnosis of ovine enzootic abortion using an indirect ELISA (romp91b ielisa) based on a recombinant protein fragment of the polymorphic outer membrane protein POMP91B of Chlamydophila abortus. FEMS Microbiol Lett. 2001;195(2): Longbottom D, Fairley S, Chapman S, Psarrou E, Vretou E, Livingstone M. Serological diagnosis of ovine enzootic abortion by enzyme-linked immunosorbent assay using a recombinant protein fragment of the polymorphic outer membrane protein POMP90 of Chlamydophila abortus. J Clin Microbiol. 2002;40(11): Longbottom D, Coulter LJ. Animal chlamydioses and zoonotic implications. J Comp Pathol. 2003;128(4): Longbottom D, Livingstone M. Vaccination against chlamydial infections of man and animals. Vet J. 2006;171(2): Markey BK, McNulty MS, Todd D. Comparison of serological tests for the diagnosis of Chlamydia psittac infection of sheep. Vet Microbiol. 1993;36(3-4):

114 Habilitationsschrift Nicole Simona Borel McCauley LM, Lancaster MJ, Young P, Butler KL, Ainsworth CG. Comparison of ELISA and CFT assays for Chlamydophila abortus antibodies in ovine sera. Aust Vet J. 2007;85(8): Moch H, Kononen T, Kallioniemi OP, Sauter G. Tissue microarrays: what will they bring to molecular and anatomic pathology? Adv Anat Pathol. 2001;8(1): Monecke S, Ehricht R. Rapid genotyping of methicillin-resistant Staphylococcus aureus (MRSA) isolates using miniaturised oligonucleotide arrays. Clin Microbiol Infect. 2005;11(10): Ossewaarde JM, Meijer A. Molecular evidence for the existence of additional members of the order Chlamydiales. Microbiology. 1999;145(Pt 2): Peeling RW, Brunham RC. Chlamydiae as pathogens: new species and new issues. Emerg Infect Dis. 1996;2(4): Philips HL, Clarkson MJ. Spontaneous change from overt to covert infection of Chlamydia pecorum in cycloheximidetreated mouse McCoy cells. Infect Immun. 1995;63(9): Piercy DW, Griffiths PC, Teale CJ. Encephalitis related to Chlamydia psittac infection in a 14-week-old calf. Vet Rec. 1999;144(5): Poppert S, Essig A, Marre R, Wagner M, Horn M. Detection and differentiation of chlamydiae by fluorescence in situ hybridization. Appl Environ Microbiol. 2002;68(8): Pospischil A, Thoma R, Hilbe M, Grest P, Gebbers JO. Abortion in woman caused by caprine Chlamydophila abortus (Chlamydia psittaci serovar 1). Swiss Med Wkly. 2002a;132(5-6): Pospischil A, Thoma R, von Bomhard W, Reitt K, Cantieni J, Zimmermann D, Polkinghorne A. Abortion in cattle caused by Chlamydia psittaci. Schweiz Arch Tierheilkd. 2002b;144(9): German. Reitt K, Hilbe M, Voegtlin A, Corboz L, Haessig M, Pospischil A. Aetiology of bovine abortion in Switzerland from 1986 to 1995 A retrospective study with emphasis on detection of Neospora caninum and Toxoplasma gondii by PCR. J Vet Med A Physiol Pathol Clin Med. 2007;54(1): Rodolakis A, Souriau A. Variations in the virulence of strains of Chlamydia psittaci for pregnant ewes. Vet Rec. 1989;125(4): Sachse K, Grossmann E. Chlamydial diseases of domestic animals--zoonotic potential of the agents and diagnostic issues. Dtsch Tierarztl Wochenschr. 2002;109(4): German. Sachse K, Hotzel H, Slickers P, Ellinger T, Ehricht R. DNA microarray-based detection and identification of Chlamydia and Chlamydophila spp. Mol Cell Probes. 2005;19(1): Salti-Montesanto V, Tsoli E, Papavassiliou P, Psarrou E, Markey BK, Jones GE, Vretou E. Diagnosis of ovine enzootic abortion, using a competitive ELISA based on monoclonal antibodies against variable segments 1 and 2 of the major outer membrane protein of Chlamydia psittaci serotype 1. Am J Vet Res. 1997;58(3): Sarma DK, Tamuli MK, Rahman T, Boro BR, Deka BC, Rajkonwar CK. Isolation of Chlamydia from a pig with lesions in the urethra and prostate gland. Vet Rec. 1983;112(22):525. Sherman JK, Jordan GW. Cryosurvival of Chlamydia trachomatis during cryopreservation of human spermatozoa. Fertil Steril. 1985;43(4): Simon R, Mirlacher M, Sauter G. Tissue microarrays. Biotechniques. 2004;36(1):

115 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations Stamp JT, Watt JA, Cockburn RB. Enzootic abortion in ewes; complement fixation test. J Comp Pathol. 1952;62(2): Sting R, Hafez HM. Purification of Chlamydia psittaci antigen by affinity chromatography on polymyxin B agarose for use in the enzymelinked immunosorbent assay (ELISA). Zentralbl Bakteriol. 1992;277(4): Storz J. In: Chlamydia and Chlamydia-induced diseases. Charles C. Thomas. Springfield, Illinois, USA, Storz J. Overview of animal diseases induced by chlamydial infections. In: Microbiology of Chlamydia. Edited by Barron AL. Boca Raton, Florida, USA: CRC Press. 1988: Storz J, Carroll EJ, Ball L, Faulkner LC. Isolation of a psittacosis agent (Chlamydia) from semen and epididymis of bulls with seminal vesiculitis syndrome. Am J Vet Res. 1968;29(3): Taton TA, Mirkin CA, Letsinger RL. Scanometric DNA array detection with nanoparticle probes. Science. 2000;289(5485): Teankum K, Pospischil A, Janett F, Bürgi E, Brugnera E, Hoelzle K, Polkinghorne A, Weilenmann R, Zimmermann DR, Borel N. Detection of chlamydiae in boar semen and genital tracts. Vet Microbiol. 2006;116(1-3): Thomson NR, Yeats C, Bell K, Holden MT, Bentley SD, Livingstone M, Cerdeño-Tárraga AM, Harris B, Doggett J, Ormond D, Mungall K, Clarke K, Feltwell T, Hance Z, Sanders M, Quail MA, Price C, Barrell BG, Parkhill J, Longbottom D. The Chlamydophila abortus genome sequence reveals an array of variable proteins that contribute to interspecies variation. Genome Res. 2005;15(5): and rifampin resistance testing with highdensity DNA probe arrays. J Clin Microbiol. 1999;37(1): Twomey DF, Griffiths PC, Horigan MW, Hignett BC, Martin TP. An investigation into the role of Chlamydophila spp. in bovine upper respiratory tract disease. Vet J. 2006;171(3): Vanrompay D, Ducatelle R, Haesebrouck F. Chlamydia psittac infections: a review with emphasis on avian chlamydiosis. Vet Microbiol. 1995;45(2-3): Van De Rijn M, Gilks CB: Applications of microarrays to histopathology. Histopathology 2004;44(2): Vezník Z, Svecová D, Pospísil L, Diblíková I. Detection of chlamydiae in animal and human semen using direct immunofluorescence. Vet Med (Praha). 1996;41(7): Czech. Vretou E, Psarrou E, Kaisar M, Vlisidou I, Salti- Montesanto V, Longbottom D. Identification of protective epitopes by sequencing of the major outer membrane protein gene of a variant strain of Chlamydia psittaci serotype 1 (Chlamydophila abortus). Infect Immun. 2001;69(1): Vretou E, Radouani F, Psarrou E, Kritikos I, Xylouri E, Mangana O. Evaluation of two commercial assays for the detection of Chlamydophila abortus antibodies. Vet Microbiol. 2007;123(1-3): Wilson WJ, Strout CL, DeSantis TZ, Stilwell JL, Carrano AV, Andersen GL. Sequencespecific identification of 18 pathogenic microorganisms using microarray technology. Mol Cell Probes. 2002;16(2): Troesch A, Nguyen H, Miyada CG, Desvarenne S, Gingeras TR, Kaplan PM, Cros P, Mabilat C. Mycobacterium species identification 111

116 Habilitationsschrift Nicole Simona Borel 112

117 Chlamydial Abortion in Ruminants: Serological, Epidemiological and Diagnostic Investigations 7. Acknowledgements This work would not have been possible without the help and support of my supervisor, laboratory technicians, collaborators, colleagues and family. First and foremost, I would like to thank my supervisor Professor Andreas Pospischil for his guidance and inspiration throughout the course of my scientific work. Without his continuous support, this work would not have been possible. He introduced me in the field of chlamydial research and made contacts to the international chlamydial community possible. He encouraged me to initiate new projects and independant research activities. Ziegler & Dr. Claudia Dumrese, University of Zurich. Finally, special thanks to my mother Elvira and my brother Alain for their support and patience throughout this adventure. I would like to thank my mother Elvira and my father Walter for enabling my studies and for supporting me during my studies and my doctoral thesis. I would like to thank Carmen Kaiser for her great support of the laboratory work. Beside her excellent technical support, I would like to mention the important friendship that developed between us during my research time at the Institute of Veterinary Pathology. Many thanks also to the whole laboratory technician group, especially to Roseline Weilenmann as the chief technician, Lisbeth Nufer and Andrea Schifferli for their technical support and their friendship. Many thanks to the doctoral fellows Patrick Späni, Komkrich Teankum, Evelyne Winter- Kempf, Andrea Gerber, Silke Ruhl and Daniela Zweifel, and to all members of the Institute of Veterinary Pathology, University of Zurich. Particular mention should be made to Dr. Enrico Brugnera and Dr. Adam Polkinghorne for their inspiration, scientific advice and their help in editing manuscripts. Special thanks to Rainer Egle for his computer support and for the layout of this manuscript. I would like to acknowledge the friendship, support and provision of material from our collaborators: Dr. Ruedi Thoma, Amt für Lebensmittelsicherheit und Tiergesundheit, Chur, Dr. Jim Summersgill, University of Louisville, USA, Dr. Gilbert Greub and his group, University of Lausanne and Dr. Urs 113

118 Habilitationsschrift Nicole Simona Borel 114

119

120 Nicole Simona Borel Dr. med. vet. von Couvet und Neuchâtel/NE