Analysis of the Humoral Immune Response against Surface and Virulence- Associated Chlamydia abortus Proteins in Ovine and Human Abortion Using a

JCM Accepted Manuscript Posted Online 18 May 2016 J. Clin. Microbiol. doi:10.1128/jcm.00351-16 Copyright 2016, American Society for Microbiology. All Rights Reserved. 1 2 3 Analysis of the Humoral Immune Response against Surface and Virulence- Associated Chlamydia abortus Proteins in Ovine and Human Abortion Using a Newly Developed Line Immunoassay 4 Running title: Serodiagnosis of Chlamydia abortus infections 5 6 7 Jürgen Benjamin Hagemann 1#, Ulrike Simnacher 1, David Longbottom 2, Morag Livingstone 2, Julia Maile 3, Erwin Soutschek 3, Gernot Walder 4, Katharina Boden 5, Konrad Sachse 6, Andreas Essig 1 8 9 1 Institute of Medical Microbiology and Hygiene, University Hospital of Ulm, Albert- Einstein-Allee 23, D-89081 Ulm, Germany 10 11 2 Moredun Research Institute, Pentlands Science Park Bush Loan, Edinburgh, Midlothian EH26 0PZ, UK 12 3 MIKROGEN GmbH, Floriansbogen 2-4, D-82061 Neuried, Germany 13 14 4 Institute of Hygiene and Medical Microbiology, University of Innsbruck, Schöpfstraße 41, A-6020 Innsbruck, Austria 15 16 5 Institute of Clinical Chemistry and Laboratory Medicine, University Hospital Jena, Erlanger Allee 101, D-07747 Jena, Germany 17 18 6 Institute of Molecular Pathogenesis, Friedrich-Loeffler-Institut (Federal Research Institute for Animal Health), Naumburger Straße 96a, D-07743 Jena, Germany 19 20 1

21 # Corresponding author: 22 Dr. med. Jürgen Benjamin Hagemann 23 Institute of Medical Microbiology and Hygiene, University Hospital of Ulm 24 Albert-Einstein-Allee 23 25 D-89081 Ulm, Germany 26 Tel.: 0049 731 500 65313 27 Fax: 0049 731 500 65304 28 Mail: benjamin.hagemann@uniklinik-ulm.de 29 30 31 32 33 34 35 36 37 38 2

39 Abstract 40 41 42 43 44 The obligate intracellular bacterium Chlamydia abortus is the causative agent of enzootic abortion of ewes (EAE) and poses a significant zoonotic risk for pregnant women. Using proteomic analysis and gene expression library screening in a previous project, we identified potential virulence factors and candidates for serodiagnosis, of which nine have been scrutinised here by using a strip immunoassay. 45 46 47 48 49 50 We have shown that aborting sheep exhibited a strong antibody response against surface (MOMP, MIP, Pmp13G) and virulence-associated (CPAF, TARP, SINC) antigens. While the latter disappeared within 18 weeks following abortion in the majority of animals, antibodies against surface proteins persisted beyond the duration of the study. In contrast, non-aborting experimentally infected sheep developed antibodies mainly against surface antigens (MOMP, MIP, Pmp13G), all of which did not persist. 51 52 53 54 55 56 We were also able to detect antibodies against these surface antigens in C. abortusinfected women who had undergone septic abortion, whereas a group of shepherds and veterinarians with occupational exposure to C. abortus-infected sheep revealed only sporadic immune responses to the antigens selected. The most specific antigen for serodiagnosis of human C. abortus infections was Pmp13G, which showed no crossreactivity with other chlamydiae infecting humans. 57 58 59 We suggest that Pmp13G-based serodiagnosis accomplished by the detection of antibodies against virulence-associated antigens such as CPAF, TARP, and SINC, may improve laboratory diagnosis of human and animal C. abortus infections. 60 3

61 Introduction 62 63 64 65 66 67 68 69 70 Chlamydia (C.) abortus is an obligate intracellularly replicating zoonotic bacterium that shares a characteristic biphasic developmental cycle with all other members of the family Chlamydiaceae [1]. Among chlamydiae affecting humans, C. trachomatis is of most clinical and epidemiological relevance as a cause of oculogenital infections, including non-gonococcal urethritis and cervicitis, lymphogranuloma venereum, and trachoma. C. pneumoniae is known to be involved in community-acquired pneumonia, pharyngitis, bronchitis, and sinusitis. In addition, the avian pathogen C. psittaci has a well-documented zoonotic potential causing human psittacosis (ornithosis), which may present as generalized and life-threatening pneumonia [2]. 71 72 73 74 75 76 77 78 79 80 81 C. abortus is typically occurring in ruminants, such as sheep and goats, and is the leading cause of enzootic abortion of ewes (EAE) worldwide [3]. Even in the absence of elevated abortion rates, the agent was shown to be widespread in German sheep flocks, with 50% of flocks testing PCR-positive and 94% harbouring seropositive animals [4]. Afterbirths and foetuses of abortion cases can contain high loads of the pathogen and represent the major source of transmission to susceptible humans and naïve ewes, as well as environmental contamination [3]. Inhalation of infective aerosols by pregnant women poses the risk of severe infection including spontaneous abortion, stillbirth, and septicaemia [3, 5, 6]. However, only very little is known about clinical relevance, epidemiology, and transmission of human C. abortus infection, since specific diagnostic tools are currently not available [3]. 82 83 Recently, it was shown that relatively low doses of intranasally inoculated C. abortus organisms induced latent infection in non-pregnant ewes [7]. When latently infected 4

84 85 86 87 88 89 90 91 92 ewes became pregnant, this resulted in placental infection and consecutive abortion, whereas animals infected with high doses were better protected and showed a much lower abortion rate. While laboratory diagnosis of EAE can be conducted using DNA- or protein-based tests, serology remains the preferred option in many laboratories [3]. Despite limited sensitivity and specificity, the complement fixation test (CFT) is still the most widely used procedure for detecting infection and vaccination titres [3, 8]. A test based on polymorphic membrane protein Pmp12G in an ELISA format [3, 9] has been commercialised in 2015. However, serological assays based on a panel of both surface and virulence-associated C. abortus antigens have not been established until now. 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 To extend the spectrum of potential diagnostic marker proteins, we identified 48 immunoreactive proteins using 2D immunoblot analysis and screening of a C. abortus gene expression library in a previous project [10]. Out of these, we have selected nine proteins for recombinant synthesis and further evaluation. These comprised i) three surface proteins, specifically the major outer membrane protein (MOMP), macrophage infectivity potentiator (MIP), and polymorphic membrane protein 13G (Pmp13G); ii) three virulence-associated proteins, namely the homologs of Chlamydia protease-like activity factor (CPAF), translocated actin-recruiting phosphoprotein (TARP), and secreted inner nuclear membrane-associated Chlamydia protein (SINC) [11], and iii) three hypothetical proteins, specifically CAB031, CAB821, and CAB408, two of which (CAB821 and CAB408) are predicted to be secreted by the type III secretion system [12]. For a comprehensive analysis of the ovine and human antibody response against these antigens, we have used the so-called line immunoassay since this format allows the simultaneous detection of antibodies to multiple antigens in a single run [13]. Characterized sera from i) experimentally infected sheep, ii) naturally infected sheep, iii) 5

108 109 110 infected humans, iv) healthy blood donors, and individuals with chlamydial infections other than C. abortus, as well as v) humans at risk of exposure to C. abortus have been analyzed. 111 112 Material and Methods 113 Sera 114 115 116 117 118 119 The human and animal sera were taken from already existing serum collections of previous studies [4, 5, 7, 10, 13, 14]. Human serum samples were anonymized and their use was approved by the local ethics committee of the University of Ulm (#96/09) and the local ethics committee of the University Hospital Jena (2525-04/09). Table 1 gives an overview of the sera used for this study and lists their main characteristics (Tab.1). Further details are described below. 120 121 Animal sera 122 123 124 125 126 127 128 129 Experimentally infected sheep. Sera were obtained from three different groups of experimentally infected sheep, as described in a previously published experimental model of latency [7]. In that study, animals received different intranasally administered doses of infectious C. abortus elementary bodies (EBs) eight weeks before mating, and were monitored and bled over several months during pregnancy. Here, we investigated the sera of 10 experimentally challenged sheep undergone abortion as well as 10 experimentally inoculated sheep that lambed normally with clinically healthy offspring. Three uninfected sheep served as a negative control group. Sera were obtained at 0, 2, 6

130 131 5, 12, 21, 29 and 47 weeks after inoculation with abortion and lambing occurring around 26 30 weeks post-inoculation (pi) (Tab.1). 132 133 134 135 136 137 138 Naturally infected sheep. Sera were taken from 32 German sheep flocks with endemic C. abortus infections [4]. 11 sera originated from sheep with abortion and rectal or vaginal swabs positive for C. abortus by PCR. 29 sera originated from clinically asymptomatic sheep with vaginal and rectal swabs positive for C. abortus by PCR. Sera of asymptomatic, C. abortus PCR-negative sheep (n = 25), as well as sera of sheep positive for C. pecorum (n = 11) and / or C. psittaci (n = 5) served as a negative- and specificity control (Tab.1). 139 140 Human sera 141 142 Pregnant women with septic abortion. Sera were collected from women who had undergone septic abortion due to C. abortus as proven by PCR [5]. 143 144 145 146 147 Shepherds and veterinarians. We investigated 88 sera obtained from shepherds and veterinarians with close contact to infected sheep from flocks with endemic C. abortus infections and cases of abortion. The sera were collected during meetings of shepherds and veterinarians in parallel with the study on sheep by Lenzko et al. between 2009 and 2011. 148 149 150 151 Negative-control and specificity control group. Sera were obtained from healthy blood donors (n = 20, 50:50 male:female), as well as from clinically symptomatic patients with infections due to C. pneumoniae (n = 20), C. trachomatis (n = 20) or C. psittaci (n = 3), as proven by PCR [14]. 7

152 153 Gene cloning and expression of recombinant antigens 154 155 156 157 158 159 160 161 162 Nine immunoreactive C. abortus antigens were selected based on previous work by Forsbach-Birk et al. [10] to be recombinantly expressed for the production of a line immunoassay [13]. These were CAB048 (MOMP, Gene ID: 3337752), CAB080 (MIP, Gene ID: 3337460), CAB281 (Pmp13G, Gene ID: 3337669), CAB712 (CPAF homolog, Gene ID: 3337730), CAB167 (TARP homolog, Gene ID: 3337791), CAB063 (SINC homolog, Gene ID: 3337689), as well as the hypothetical proteins CAB031 (Gene ID: 3337919), CAB821 (Gene ID: 3338162), and CAB408 (Gene ID: 3337407). All antigens except CAB031, which was truncated at its C-terminus, were expressed as full-length proteins and highly purified by chromatographic methods [15]. 163 164 Production of the Chlamydia abortus line assay 165 166 167 168 169 170 171 172 The line assay enables the simultaneous detection of antibodies against selected antigens. All of the nine recombinant antigens MOMP, MIP, Pmp13G, CPAF, TARP, SINC, CAB031, CAB821, CAB408 were deposited in a line format on nitrocellulose membranes in separate lanes using a dispense platform. A reaction control, three conjugate controls for the detection of human immunoglobulin IgG, IgA, and IgM as well as a cut off line were also applied. For each C. abortus protein, individual dilutions and buffers were used. The membranes were saturated with milk solution and cut into test strips. 173 8

174 Incubation of line immunoassays and read-out of human and animal sera 175 176 177 178 179 180 181 182 183 184 185 186 187 188 Strips were incubated with either human or animal sera (both diluted 1:100) overnight (human sera) or for one hour (animal sera) at room temperature, allowing specific antibodies to bind to the C. abortus antigens. After a repetitive wash step with a phosphate buffer for 3 x 5 minutes, peroxidase-labeled rabbit-anti-human IgG (diluted 1:100) or polyvalent rabbit-anti-sheep antibodies (DakoCytomation, Denmark A/S, Glostrup) were added (diluted 1:1000) and incubated for 45 minutes at room temperature. After a second wash step, binding of specific antibodies was detected by the use of tetramethylbenzidine, incubated for 8 minutes at room temperature. Strips were first examined for positive bands visible to the naked eye before being analyzed by intensity measurements. Intensities of reactive bands were measured by use of an OpticPro S28 scanner (Plustek, Norderstedt, Germany) and recomscan software (BioSciTec GmbH, Frankfurt, Germany) according to the manufacturer s instruction. Bands were considered positive if (intensity of antigen band) : (intensity cut off 15% tolerance) 1.0, further referred to as cut off-adjusted optical density (OD). 189 190 Statistical analysis 191 192 193 Statistical analysis was performed using Microsoft Excel 2013. Mann-Whitney U-test was used for the calculation of the level of significance. Statistical significance was accepted at p 0.05. 194 195 Results 9

196 Animal sera 197 198 199 Aborting experimentally infected sheep show strong and long-lasting antibody response against surface and virulence-associated as well as hypothetical C. abortus proteins 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 Experimental intranasal inoculation of low or moderate doses of infectious C. abortus EBs (5x10 3 and 5x10 5 IFU, respectively) resulted in abortion in the majority of animals. Their placental tissue was investigated by histopathology, immunohistochemistry, and quantitative real-time PCR and revealed pathological tissue findings typical for EAE and evidence of C. abortus in placental tissue samples as described previously [7]. Here, we analyzed serum samples of sheep with abortion following inoculation of low (n = 5) and medium (n = 5) doses of EBs. A late antibody response not detectable until week 21 pi and peaking at the time of abortion at 29 weeks pi proved characteristic for these animals (Fig.1A). An increase of antibodies against Pmp13G was detected in 50% of sheep at 21 weeks pi (Tab.S1A), with bands being markedly more intense compared to the non-abortion group of experimentally infected sheep (p 0.02) (Fig.2A). At the time of abortion at 29 weeks pi, sera of all animals contained antibodies against Pmp13G, with OD values higher than for the non-abortion group (p 0.02). In addition, strong and intense bands with distinctly higher OD values than those of the non-abortion group were found for the other surface antigens MOMP (p 0.02) and MIP (p 0.02). Apart from those, we were also able to measure higher OD values for the virulence-associated antigens CPAF (p 0.02), TARP (p 0.02), and SINC (p 0.02) (Fig.1B), as well as for the hypothetical protein CAB031 (p 0.02) (Fig.1C). Strong bands were still present after 47 10

218 219 weeks pi, especially for the surface antigens MOMP (p 0.02) and Pmp13G (p 0.02) in the majority of animals (Fig.2A, Tab.S1A). 220 221 222 Normally lambing experimentally infected sheep show a rapid, but short-term immune response against surface antigens 223 224 225 226 227 228 229 230 231 232 233 Sheep that were experimentally inoculated with a high dose (5x10 7 IFU) of C. abortus EBs lambed normally with healthy offspring and no evidence of placental pathology [7]. These sheep (n = 10), which cleared infection, showed a clearly different pattern of humoral immune response than animals that underwent abortion. Two weeks pi (six weeks before mating), animals showed band patterns with increased OD values for the surface antigens MOMP (n.s.), MIP (n.s.), and Pmp13G (p 0.02), the latter being positive in 90% of sheep (Fig.2B, Tab.S1B). Band intensity rapidly decreased below the level of detection in subsequent sampled time points, with little or no detection of any response at the time of lambing. Uninfected animals kept under controlled laboratory conditions showed no measurable antibodies against the selected antigens (Fig.2C) and are therefore not listed in Tab.S1. 234 235 236 Aborting naturally infected sheep show antibodies against surface and virulence- associated as well as hypothetical antigens 237 238 239 Sera obtained from naturally infected ewes at the time of abortion were analyzed (n = 11). Compared to asymptomatic sheep that were C. abortus PCR-negative or PCRpositive for C. pecorum or C. psittaci (n = 41 in total), we observed greater antibody 11

240 241 242 243 244 245 reactivity for the surface antigens MOMP (p 0.02), MIP (p 0.02), and Pmp13G (p 0.02), for the virulence-associated antigens CPAF (p 0.02) and SINC (p 0.02), as well as for the hypothetical protein CAB031 (p 0.02) (Fig.2D-E, Fig.S1). In total, 54.5% of naturally infected sheep exhibited antibodies against MOMP, 72.3% against MIP and Pmp13G, 54.5% against CPAF, 72.3% against SINC, and 63.6% against CAB031 (data not shown). 246 247 248 Asymptomatic C. abortus PCR-negative and C. pecorum- or C. psittaci-positive sheep show little non-specific antibody reactivity 249 250 251 252 253 254 Sera of 41 asymptomatic sheep that were C. abortus PCR swab-negative (n = 25) or C. pecorum (n = 11) or C. psittaci swab-positive (n = 5) were investigated for the presence of antibodies against the antigen panel. We observed only single non-specific antibody binding in individual sheep, but none of them showed more than one positive band. In total, 2.4% of the animals showed detectable antibody reactivity against MIP, Pmp13G, CPAF, or TARP, and none of the sera reacted with MOMP, SINC, or CAB031 (Fig.2F). 255 256 257 Asymptomatic carriers of C. abortus may show a weak to moderate antibody response 258 259 260 261 Sera of 29 ewes that did not abort during the current lambing season, but with a positive C. abortus PCR result from vaginal or rectal swabs were tested. We observed that 10.3% of animals showed antibody reactivity to MOMP, 27.6 % to MIP, and 17.2% to Pmp13G. A humoral immune response to CPAF, TARP, and CAB063 could be detected 12

262 263 in 31.0%, 27.6%, and 13.8% of sheep, respectively. Bands for CAB031 were present in 13.8% (data not shown). 264 265 Human sera 266 267 Women that have aborted as a result of C. abortus infection reveal an antibody response against surface antigens of the bacterium 268 269 270 271 272 Due to the rarity of disease in humans, only two serum samples obtained from women that aborted as a result of C. abortus infection were available for analysis. Both samples revealed an immune response to the surface antigens MOMP, MIP, and Pmp13G. One of them showed antibody reactivity to the hypothetical protein CAB821, which is predicted to be type III secreted (Fig.3A, Fig.S2). 273 C. abortus Pmp13G is highly specific 274 275 276 277 278 279 Blood donors failed to exhibit any measurable antibody response to any of the antigens included, whereas we observed sporadic cross-reactivity covering almost all antigens when testing sera of patients infected with chlamydiae other than C. abortus (Fig.3B, Fig.S2). Notably, surface antigen Pmp13G was the only non-cross-reacting antigen, which indicates its potential to serve as a specific marker for C. abortus infection (Tab.S2). 280 One shepherd showed specific antibody response to C. abortus 281 282 We analyzed serum samples obtained from shepherds and veterinarians (n = 88) who had close contact with sheep flocks with high prevalence of C. abortus infections [4]. 13

283 284 285 286 287 288 289 Only a minority showed a detectable antibody response against any of the C. abortus antigens. In total, a signal to MOMP was detected in 6.9%, MIP in 14.9%, CPAF in 1.2%, TARP in 5.8%, and SINC in 4.6%. In addition, the response to the hypothetical proteins CAB031 and CAB821 was found in only 3.5% and 1.2%, respectively (Fig.3C, Fig.S2). A single shepherd showed measurable amounts of IgG antibodies against multiple surface and virulence-associated proteins including MOMP, MIP, Pmp13G, and SINC (Fig.3C), which suggests that transmission had occurred. 290 291 Discussion 292 293 294 295 296 297 298 299 300 301 302 303 304 305 To analyze the kinetics of antibody development that occurs during ovine abortion, we compared consecutive sera of latently infected aborting sheep with sera of sheep that had been inoculated with a high dose of C. abortus and lambed normally. We were able to show that, around the time of abortion, aborting sheep developed a strong antibody response to the surface proteins MOMP, MIP, and Pmp13G, as well as to the virulenceassociated proteins CPAF, TARP, and the SINC homolog CAB063. However, in the majority of experimentally infected aborting animals, antibody generation was not detectable until week 13 of gestation (week 21 pi). This is in agreement with the observation that chlamydial growth and pathology in the placenta are not evident any earlier than day 90 of gestation [7, 16], and that severe placentitis with massive chlamydial replication occurs in late pregnancy, when infected ewes are undergoing hormonal and immunological changes. As a consequence, detection of virulenceassociated antibodies around the time of abortion may reflect the increased expression of virulence-associated proteins that are required to regulate and sustain intra-placental 14

306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 infection [17]. Among these, TARP is a type III-secreted protein that was shown to modulate host cell cytoskeleton function in C. trachomatis infection [18, 19]. CPAF has initially been described as a chlamydial protease degrading host cell transcription factors, however its impact on virulence and pathogenicity is currently controversially discussed [20]. SINC is a type III secreted protein targeting the nuclear membrane of infected cells, which may modulate the nuclear envelope function [10, 11]. Here, we have demonstrated that virulence-associated antibodies are predominantly generated in animals that have aborted. This report therefore confirms the observation that virulenceassociated proteins, including the newly described SINC homolog of C. abortus (CAB063), are immunogenic in ovine abortion [10]. Although ewes that have aborted are considered protected from further abortion due to C. abortus, they represent a major reservoir of the pathogen for animals and humans and therefore need to be rapidly identified. As a major conclusion, detection of antibodies against both surface and virulence-associated proteins in a ewe may indicate imminent or recent abortion, since we have shown that antibody levels against virulence-associated proteins rapidly decreased below the level of detection within the subsequent observation period, while antibodies to surface antigens persisted. Further studies are needed to clarify, whether antibodies against virulence-associated antigens in EAE may become a useful marker suggesting the implementation of control measures at the flock level [3] such as segregation of seropositive animals to limit dissemination to susceptible animals, or treatment to prevent abortion. 327 15

328 329 330 331 332 333 334 335 336 337 338 339 The surface antigens investigated in this study had been shown to be immunoreactive in C. abortus infections [10] as well as in other chlamydial infections [13, 21, 22]. MOMP makes up to 50-60% of the total protein mass of the outer membrane in chlamydiae [23]. The type V autotransporter Pmp13G [24] is a member of the polymorphic membrane protein (Pmp) family, representing proteins that play an important role in the pathogenesis of chlamydial infections. Several Pmps of C. trachomatis, C. pneumoniae, C. psittaci, and C. abortus have been shown to elicit a humoral immune response [13, 25], with PmpD of C. trachomatis being described as a pan-neutralizing antigen [26]. Pmps are differentially expressed during the chlamydial developmental cycle [27, 28], therefore suspected to play a role in antigenic diversity and evasion from the host immune response. Pmp21 of C. pneumoniae has been described as an invasion protein that recruits the EGF receptor for host cell entry [29]. 340 341 342 343 344 345 346 347 348 349 350 351 In agreement with a study by Longbottom et al. [7], an early but transient antibody response to surface proteins was observed in sheep that were experimentally inoculated with high doses of C. abortus and lambed normally. The authors suggested that animals inoculated with high doses had in principle been vaccinated, since the dose was equivalent to that used with commercial live attenuated vaccines [3]. Our data support this equivalence as the absence of virulence factor-associated antibodies and the rapid decrease of antibodies below detection levels suggests loss of the immunogenic stimulus and indicates elimination of the pathogen. The finding that 90% of normally lambing experimentally infected animals showed a rapid antibody response to Pmp13G (CAB281) could indicate a protective role of neutralizing antibodies to this antigen during clearance of infection. In aborting experimentally infected animals, we have demonstrated that all of them developed antibodies against Pmp13G around the time of 16

352 353 354 abortion and that these antibodies were still present in 90% of animals 18 weeks after abortion. As sheep that have had abortion will normally not abort again, we speculate that Pmp13G antibodies may contribute to prevention of reinfection and abortion. 355 356 357 358 359 360 361 362 363 364 365 366 367 The antibody responses observed in experimentally infected animals correspond well to the present results on sera from German sheep flocks with a high prevalence of C. abortus infections. More than 70% of aborting ewes presented antibodies to Pmp13G and the virulence-associated SINC homolog at the time of abortion. As stated above, simultaneous detection of antibodies to surface and virulence-associated proteins may serve as a criterion to identify aborting animals and those on the verge of abortion. In contrast, non-aborting asymptomatic carriers of C. abortus either remained serologically negative or developed only a weak to moderate antibody response with much less prominent band patterns than did aborting animals. Therefore, we can conclude that mere colonization with C. abortus cannot be reliably detected by serological tools. On the flock level, it is therefore not possible to identify asymptomatic C. abortus shedders serologically, even though recent abortion leads to significantly higher ODs of reactive antigen bands compared to mere colonization (Fig.S1). 368 369 370 371 372 373 The complement fixation test (CFT) is the only test currently recognized by the World Organization For Animal Health (OIE) for diagnosing ovine chlamydiosis [8]. However, concerns on cross-reactivity preclude its use in species-specific diagnosis, as the antigen used includes a heat-resistant lipopolysaccharide (LPS), which is present in all Chlamydiaceae [30]. Further systematic studies on the sensitivity and specificity of the most promising antigens such as Pmp13G, CPAF, and SINC, or a combination of them 17

374 375 including quantification of band ODs are needed to compare their diagnostic potential with CFT and recently developed serodiagnostic assays [3]. 376 377 378 379 380 381 382 383 384 385 386 387 388 389 Concerning human C. abortus infection, pregnant women exposed to the pathogen run a substantial risk of developing severe infection and abortion, even though knowledge about the epidemiology of C. abortus infections in humans is poor. In most cases reported, diagnosis was established through culture, PCR, or immunohistochemistry of placental tissue following abortion [5, 6]. It is hard to deny that earlier microbiological diagnosis and earlier adequate antibiotic treatment would have substantially improved the clinical course of infection. Both patients examined here presented antibodies against MOMP, MIP, and Pmp13G. Even though two sera of confirmed infection would not justify definitive conclusions in terms of sensitivity, we suggest that the Pmp13G response is highly specific for C. abortus infection, since neither healthy blood donors nor patients suffering from other chlamydial infections revealed (cross-reactive) antibodies against Pmp13G in their sera. In contrast, the use of MOMP and MIP seems to be of limited value for species-specific serodiagnosis, as antibodies against these proteins were also detected in patients with chlamydial infections other than C. abortus. 390 391 392 393 394 395 Clinically relevant human C. abortus infections not associated with pregnancy have only sporadically been reported to date. We provide serological evidence that antigenic exposure to C. abortus may lead to a specific humoral immune response outside of pregnancy. Nevertheless, the case of a male shepherd who presented antibodies to Pmp13G and the SINC homolog, appears to be exceptional, since he was the only one among a group of 88 individuals. 18

396 397 398 399 400 401 402 403 404 405 406 407 In summary, we have analyzed the humoral response during both animal and human C. abortus infection. In animals that have undergone abortion, we observed a strong antibody response against surface and virulence-associated proteins. Comparing experimentally infected animals with either asymptomatic infection or abortion, we were able to show that antibodies against virulence-associated proteins are predominantly raised in animals that aborted. The surface protein Pmp13G of C. abortus seems to be a sensitive and highly specific immunogen in animal infection. Even though the human portion investigated here is too limited to make definitive conclusions on antigen sensitivities in humans, Pmp13G has proven a highly specific antigen that warrants further investigation. Further studies are needed to clarify, whether antibodies against Pmp13G are protective and may contribute to prevent reinfection in sheep and whether they are suitable as a diagnostic marker, both in sheep and humans. 408 409 Acknowledgements 410 411 412 413 414 415 416 417 This study was supported by the German BMBF ( Bundesministerium für Bildung und Forschung, Federal Ministry of Education and Research), project funding reference numbers 01KI1011C to AE and 01KI1001C to KB, as well as The Scottish Government's Rural and Environment Science and Analytical Services Division (RESAS) to DL. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. E. Soutschek and J. Maile are employees of MIKROGEN Molekularbiologische Entwicklungs-GmbH, Neuried, Germany. The authors declare that they have no other conflicts of interest in relation to this work. 418 19

419 References 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 1. Wilkat M, Herdoiza E, Forsbach-Birk V, Walther P, Essig A. 2014. Electron tomography and cryo-sem characterization reveals novel ultrastructural features of host-parasite interaction during Chlamydia abortus infection. Histochem Cell Biol 142(2):171-184. 2. Rohde G, Straube E, Essig A, Reinhold P, Sachse K. 2010. Chlamydial zoonoses. Dtsch Arztebl Int 107(10):174-180. 3. Essig A, Longbottom D. 2015. Chlamydia abortus: New Aspects of Infectious Abortion in Sheep and Potential Risk for Pregnant Women. Curr Clin Micro Rpt 2:22-34. 4. Lenzko H, Moog U, Henning K, Lederbach R, Diller R, Menge C, Sachse K, Sprague LD. 2011. High frequency of chlamydial co-infections in clinically healthy sheep flocks. BMC Vet Res 7:29. 5. Walder G, Hotzel H, Brezinka C, Gritsch W, Tauber R, Würzner R, Ploner F. 2005. An unusual cause of sepsis during pregnancy: recognizing infection with chlamydophila abortus. Obstet Gynecol 106(5 Pt 2):1215-1217. 6. Meijer A, Brandenburg A, de Vries J, Beentjes J, Roholl P, Dercksen D. 2004. Chlamydophila abortus infection in a pregnant woman associated with indirect contact with infected goats. Eur J Clin Microbiol Infect Dis 23(6):487-490. 7. Longbottom D, Livingstone M, Maley S, van der Zon A, Rochhi M, Wilson K, Wheelhouse N, Dagleish M, Aitchison K, Wattegedera S, Nath M, Entrican G, Buxton D. 2013. Intranasal infection with Chlamydia abortus induces dosedependent latency and abortion in sheep. PLoS One 8(2):e57950. 20

442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 8. Sachse K, Vretou E, Livingstone M, Borel N, Pospischil A, Longbottom D. 2009. Recent developments in the laboratory diagnosis of chlamydial infections. Vet Microbiol 135(1-2):2-21. 9. Wilson K, Livingstone M, Longbottom D. 2009. Comparative evaluation of eight serological assays for diagnosing Chlamydophila abortus infection in sheep. Vet Microbiol 135(1-2):38-45. 10. Forsbach-Birk V, Foddis C, Simnacher U, Wilkat M, Longbottom D, Walder G, Benesch C, Ganter M, Sachse K, Essig A. 2013. Profiling antibody responses to infections by Chlamydia abortus enables identification of potential virulence factors and candidates for serodiagnosis. PLoS One 8(11):e80310. 11. Mojica SA, Hovis KM, Frieman MB, Tran B, Hsia RC, Ravel J, Jenkins-Houk C, Wilson KL, Bavoil PM. 2015. SINC, a type III secreted protein of Chlamydia psittaci, targets the inner nuclear membrane of infected cells and uninfected neighbors. Mol Biol Cell 26(10):1918-1934. 12. Arnold R, Brandmaier S, Kleine F, Tischler P, Heinz E, Behrens S, Niinikoski A, Mewes HW, Horn M, Rattei T. 2009. Sequence-based prediction of type III secreted proteins. PLoS Pathog 5(4):e1000376. 459 460 461 462 13. Forsbach-Birk V, Simnacher U, Pfrepper KI, Soutschek E, Kiselev AO, Lampe MF, Meyer T, Straube E, Essig A. 2010. Identification and evaluation of a combination of chlamydial antigens to support the diagnosis of severe and invasive Chlamydia trachomatis infections. Clin Microbiol Infect 16(8):1237-1244. 21

463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 14. Walder G, Schönherr H, Hotzel H, Speth C, Oehme A, Dierich MP, Würzner R. 2003. Presence of Chlamydophila psittaci DNA in the central nervous system of a patient with status epilepticus. Scand J Infect Dis 35(1):71-73. 15. Soutschek E, Hoflacher B, Motz M. 1990. Purification of a recombinantly produced transmembrane protein (gp41) of HIV I. J Chromatogr 521(2):267-277. 16. Buxton D, Barlow RM, Finlayson J, Anderson IE, Mackellar A. 1990. Observations on the pathogenesis of Chlamydia psittaci infection of pregnant sheep. J Comp Pathol 102(2):221-237. 17. Jorgensen I, Bednar MM, Amin V, Davis BK, Ting JP, McCafferty DG, Valdivia RH. 2011. The Chlamydia protease CPAF regulates host and bacterial proteins to maintain pathogen vacuole integrity and promote virulence. Cell Host Microbe 10(1):21-32. 18. Jewett TJ, Miller NJ, Dooley CA, Hackstadt T. 2010. The conserved Tarp actin binding domain is important for chlamydial invasion. PLoS Pathog 6(7):e1000997. 19. Jiwani S, Ohr RJ, Fischer ER, Hackstadt T, Alvarado S, Romero A, Jewett TJ. 2012. Chlamydia trachomatis Tarp cooperates with the Arp2/3 complex to increase the rate of actin polymerization. Biochem Biophys Res Commun 420(4):816-821. 20. Hacker G, Heuer D, Ojcius DM. 2014. Is the hoopla over CPAF justified? Pathog Dis 72(1):1-2. 21. Tanzer RJ, Hatch TP. 2001. Characterization of outer membrane proteins in Chlamydia trachomatis LGV serovar L2. J Bacteriol 183(8):2686-2690. 22. Tanzer RJ, Longbottom D, Hatch TP. 2001. Identification of polymorphic outer membrane proteins of Chlamydia psittaci 6BC. Infect Immun 69(4):2428-2434. 22

487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 23. Kuo C, Takahashi N, Swanson AF, Ozeki Y, Hakomori S. 1996. An N-linked high-mannose type oligosaccharide, expressed at the major outer membrane protein of Chlamydia trachomatis, mediates attachment and infectivity of the microorganism to HeLa cells. J Clin Invest 98(12):2813-2818. 24. Henderson IR, Lam AC. 2001. Polymorphic proteins of Chlamydia spp.-- autotransporters beyond the Proteobacteria. Trends Microbiol 9(12):573-578. 25. Livingstone M, Entrican G, Wattegedera S, Buxton D, McKindrick IJ, Longbottom D. 2005. Antibody responses to recombinant protein fragments of the major outer membrane protein and polymorphic outer membrane protein POMP90 in Chlamydophila abortus-infected pregnant sheep. Clin Diagn Lab Immunol 12(6):770-777. 26. Crane DD, Carlson JH, Fischer ER, Bavoil P, Hsia RC, Tan C, Kuo CC, Caldwell HD. 2006. Chlamydia trachomatis polymorphic membrane protein D is a species-common pan-neutralizing antigen. Proc Natl Acad Sci USA 103(6):1894-1899. 27. Wheelhouse N, Sait M, Wilson K, Aitchison K, McLean K, Smith DG, Longbottom D. 2012. Expression patterns of five polymorphic membrane proteins during the Chlamydia abortus developmental cycle. Vet Microbiol 160(3-4):525-529. 28. Tan C, Hsia RC, Shou H, Carrasco JA, Rank RG, Bavoil PM. 2010. Variable expression of surface-exposed polymorphic membrane proteins in in vitro-grown Chlamydia trachomatis. Cell Microbiol 12(2):174-187. 23

509 510 511 512 513 514 29. Mölleken K, Becker E, Hegemann JH. 2013. The Chlamydia pneumoniae invasin protein Pmp21 recruits the EGF receptor for host cell entry. PLoS Pathog 9(4):e1003325. 30. Brade H, Brade L, Nano FE. 1987. Chemical and serological investigations on the genus-specific lipopolysaccharide epitope of Chlamydia. Proc Natl Acad Sci U S A 84(8):2508-2512. 515 516 517 518 31. Essig A, Zucs P, Susa M, Wasenauer G, Mamat U, Hetzel M, Vogel U, Wieshammer S, Brade H, Marre R. 1995. Diagnosis of ornithosis by cell culture and polymerase chain reaction in a patient with chronic pneumonia. Clin Infect Dis 21(6):1495-1497. 519 520 521 522 523 524 525 526 527 528 24

Figure 1 Kinetics of the humoral response against Chlamydia abortus antigens in experimentally infected sheep with and without abortion. Antibody response against surface (A), virulence-associated (B), and hypothetical protein antigens (C) at different periods following inoculation (0, 2, 5, 12, 21, 29, and 47 weeks). Different lines and symbols indicate different infectious doses (abortion group included n = 5 animals inoculated with 5x10 3 inclusion forming units (IFU) and n = 5 animals inoculated with 5x10 5 IFU, lambing group included n = 10 animals inoculated with 5x10 7 IFU of C. abortus elementary bodies). Mating took place at 8 weeks after inoculation and abortion or lambing at 29 weeks after inoculation, respectively. Ordinate shows the median cut off-adjusted optical density (OD) with the first (lower margin) and third quartile (upper margin) of the data, calculated from OD measurements of the reactivity of sera against the different antigens and the cut off OD of the corresponding line assay. p 0.02 (*) Figure 2 Representative line immunoassays of animal sera. Sera of experimentally infected sheep that have (A) and that have not aborted (B) as well as the corresponding negative control group consisting of healthy, uninfected sheep (C) were investigated. Numbers on the left-hand side indicate the number of weeks after inoculation. Abortion (A) or lambing (B and C) took place at 29 weeks after inoculation. Lower case characters on the right of the stripes denote individual sheep inoculated with 5x10 3 inclusion forming units (IFU) of infectious C. abortus elementary bodies (a), 5x10 5 IFU (b) and 5x10 7 IFU (c and c'), respectively. Lower case characters d and d' denote healthy, uninfected sheep. (D) Sera of sheep having undergone abortion due to natural C. abortus infection, (E) sera of asymptomatic C. abortus PCR-negative sheep with healthy offspring, (F) specificity control sera of sheep infected with C. pecorum or C. 529 psittaci. Each line assay represents one individual animal. 25

Figure 3 Representative line immunoassays of human sera. (A) Samples of two women who had aborted due to severe septic C. abortus infection were examined. (B) Specificity control on sera of males and females infected with C. trachomatis (C. tr.), C. pneumoniae (C. pn.), or C. psittaci (C. ps.). (C) Sera of shepherds and veterinarians exposed to C. abortus in high-prevalence flocks, one of them (1 out of 88) showing reactivity against the surface antigens MOMP, MIP, and pmp13g as well as against the virulence-associated protein SINC. Each line assay represents one individual. 530 26

Figure 1 Kinetics of the humoral response against Chlamydia abortus antigens in experimentally infected sheep with and without abortion. Antibody response against surface (A), virulence-associated (B), and hypothetical protein antigens (C) at different periods following inoculation (0, 2, 5, 12, 21, 29, and 47 weeks). Different lines and symbols indicate different infectious doses (abortion group included n = 5 animals inoculated with 5x10 3 inclusion forming units (IFU) and n = 5 animals inoculated with 5x10 5 IFU, lambing group included n = 10 animals inoculated with 5x10 7 IFU of C. abortus elementary bodies). Mating took place at 8 weeks after inoculation and abortion or lambing at 29 weeks after inoculation, respectively. Ordinate shows the median cut off-adjusted optical density (OD) with the first (lower margin) and third quartile (upper margin) of the data, calculated from OD measurements of the reactivity of sera against the different antigens and the cut off OD of the corresponding line assay. p 0.02 (*)

Pmp13G MIP SINC CAB031 CAB821 CPAF TARP MOMP CAB408 RK Cut off IgG IgM IgGM Pmp13G MIP SINC CAB031 CAB821 CPAF TARP MOMP CAB408 RK Cut off IgG IgM IgGM Pmp13G MIP SINC CAB031 CAB821 CPAF TARP MOMP CAB408 RK Cut off IgG IgM IgGM Pmp13G MIP SINC CAB031 CAB821 CPAF TARP MOMP CAB408 RK Cut off IgG IgM IgGM Pmp13G MIP SINC CAB031 CAB821 CPAF TARP MOMP CAB408 RK Cut off IgG IgM IgGM Pmp13G MIP SINC CAB031 CAB821 CPAF TARP MOMP CAB408 RK Cut off IgG IgM IgGM

Figure 2 Representative line immunoassays of animal sera. Sera of experimentally infected sheep that have (A) and that have not aborted (B) as well as the corresponding negative control group consisting of healthy, uninfected sheep (C) were investigated. Numbers on the left-hand side indicate the number of weeks after inoculation. Abortion (A) or lambing (B and C) took place at 29 weeks after inoculation. Lower case characters on the right of the stripes denote individual sheep inoculated with 5x10 3 inclusion forming units (IFU) of infectious C. abortus elementary bodies (a), 5x10 5 IFU (b) and 5x10 7 IFU (c and c'), respectively. Lower case characters d and d' denote healthy, uninfected sheep. (D) Sera of sheep having undergone abortion due to natural C. abortus infection, (E) sera of asymptomatic C. abortus PCR-negative sheep with healthy offspring, (F) specificity control sera of sheep infected with C. pecorum or C. psittaci. Each line assay represents one individual animal.

Figure 3 Representative line immunoassays of human sera. (A) Samples of two women who had aborted due to severe septic C. abortus infection were examined. (B) Specificity control on sera of males and females infected with C. trachomatis (C. tr.), C. pneumoniae (C. pn.), or C. psittaci (C. ps.). (C) Sera of shepherds and veterinarians exposed to C. abortus in high-prevalence flocks, one of them (1 out of 88) showing reactivity against the surface antigens MOMP, MIP and pmp13g as well as against the virulence-associated protein SINC. Each line assay represents one individual.

Animal sera Group Reference n i n s Characteristics 5 35 Inoculation with 5x10 3 IFU, abortion / stillbirth; histopathological findings in placenta specimen, significant amounts of C. abortus DNA in PCR. Experimentally infected sheep Longbottom et al., 2013 5 35 10 70 Inoculation with 5x10 5 IFU, abortion / stillbirth; histopathological findings in placenta specimen, significant amounts of C. abortus DNA in PCR. Inoculation with 5x10 7 IFU, normal lambing; no histopathological findings in placenta specimen, no significant amounts of C. abortus DNA in PCR. Naturally infected sheep Negative- and specificity control group Forsbach-Birk et al., 2013 Lenzko et al., 2011 Forsbach-Birk et al., 2013 Forsbach-Birk et al., 2013 Lenzko et al., 2011 Lenzko et al., 2011 3 21 Uninfected, asymptomatic sheep with normal lambing. 11 29 25 11 5 Sheep from German flocks with endemic C. abortus infections; abortion, vaginal or rectal swabs positive for C. abortus by PCR. Sheep from German flocks with endemic C. abortus infections; normal lambing, vaginal or rectal swabs positive for C. abortus by PCR. Sheep from German flocks with endemic C. abortus infections; normal lambing, vaginal or rectal swabs negative for C. abortus by PCR. Sheep from German flocks with endemic C. abortus infections; normal lambing, vaginal or rectal swabs negative for C. abortus and positive for C. pecorum by PCR. Sheep from German flocks with endemic C. abortus infections; normal lambing, vaginal or rectal swabs negative for C. abortus and positive for C. psittaci by PCR. Human sera Group Reference n i n s Characteristics Infected humans Walder et al., 2005 2 Women with septic abortion, PCR-positive for C. abortus. Humans at risk of exposure Negative-control group Specificity control group Lenzko et al., 2011 this study 88 Shepherds and veterinarians, unknown clinical history, close contact to sheep flocks with endemic C. abortus infections as proven by PCR methods. this study 20 Healthy blood donors, 10 female and 10 male individuals. Forsbach-Birk et al., 2010 Essig et al., 1995 Walder et al., 2003 20 20 3 Clinically symptomatic patients with infection due to C. pneumoniae as proven by PCR. Clinically symptomatic patients with infection due to C. trachomatis as proven by PCR. Clinically symptomatic patients with infection due to C. psittaci as proven by PCR. Σ 257 395 Table 1 Sera used in this study. Animal and human sera are listed with a declaration of origin, number and their specific characteristics. n i indicates the number of individuals, n s the number of sera investigated, which is equal unless for the experimentally infected group of sheep as sera at seven different periods following inoculation were collected per animal.