Assessment of the conventional detection of fecal Cryptosporidium serpentis oocysts in subclinically infected captive snakes

Transcription

1 Assessment of the conventional detection of fecal Cryptosporidium serpentis oocysts in subclinically infected captive snakes Tk Graczyk, Mr Cranfield To cite this version: Tk Graczyk, Mr Cranfield. Assessment of the conventional detection of fecal Cryptosporidium serpentis oocysts in subclinically infected captive snakes. Veterinary Research, BioMed Central, 1996, 27 (2), pp <hal > HAL Id: hal Submitted on 1 Jan 1996 HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

2 Original article Assessment of the conventional detection of fecal Cryptosporidium serpentis oocysts in subclinically infected captive snakes TK Graczyk MR Cranfield 1 Department of Molecular Microbiology and Immunology, School of Hygiene and Public Health, Johns Hopkins University, 615 North Wolfe Street, Baltimore, MD 21205; 2 Medical Department, The Baltimore Zoo, Druid Hill Park, Baltimore, MD 21217; 3 Division of Comparative Medicine, Department of Pathology, School of Medicine, Johns Hopkins University, 720 Rutland Avenue, Baltimore, MD 21205, USA (Received 28 August 1995; accepted 19 December 1995) Summary &horbar; Fecal specimens of seven captive snakes naturally subclinically infected with Cryptosporidium serpentis were monitored for 131 days for the presence and concentration of oocysts. Thirty-three of 81 (41 %) of the monoclonal antibody positive stools were read as negative (sensitivity 59%) by the acid-fast-stained (AFS) fecal smears. Oocyst concentrations in the false-negative stools ranged between 6.0 x 103/g to 2.5 x 10 4/g. The experimentally determined concentration-threshold of oocyst detection by AFS fecal smear was 3.0 x 10 4/g. The stool weights did not conform to a normal distribution; the oocyst concentration was significantly correlated with the stool weight (P < 0.01 ). Due to this correlation, stools which constituted less than 0.41 % of the snake weight were classified as negative by the AFS fecal smears. The AFS fecal smear technique should be used exclusively for the determination of Cryptosporidium-positive snakes, but not for diagnosis of snake negativity for Cryptosporidium; it should be applied only for larger fecal specimens, such as those which constitute more than 0.41 % of snake weight. At least five to seven stool samples should be examined by fecal smear in order to determine snake positivity for Cryptosporidium in subclinical infection. Cryptosporidium serpentis / snake cryptosporidiosis / oocyst / captive snake / fecal smear Résumé &horbar; Évaluation de la méthode de détection classique des oocystes fécaux de Cryptosporidium serpentis chez des serpents captifs infectés de manière subclinique. Des échantillons fécaux provenant de sept serpents captifs infectés naturellement de manière subclinique par Cryp- * Correspondence and reprints

3 tosporidium serpentis ont été contrôlés pendant 131 jours pour la présence et la concentration d oocystes. Trente-trois des 81 échantillons fécaux (41 b) réagissant positivement avec un anticorps monoclonal étaient négatifs (sensibilité 59 %) par la méthode de coloration rapide de frottis fécaux à l acide. La concentration en oocystes dans les selles faussement négatives s échelonnait de 6,0 x 10 3/g à 2,5 x 10 4/g. La concentration limite pour la détection d oocystes par la méthode de coloration rapide à l acide, déterminée expérimentalement, était 3,0 x 10 lg. La répartition du poids des selles n était pas conforme à une répartition normale ; la concentration en oocystes était corrélée de manière significative avec le poids des selles (p < 0.01). Du fait de cette corrélation, les selles qui constituaient moins de 0,41 % du poids du serpent étaient classées comme négatives par la méthode de coloration rapide de frottis fécaux à l acide. Cette technique devrait être utilisée exclusivement pour la détermination de la positivité des serpents pourcryptosporidium, mais non pour le diagnostic de la négativité des serpents pour Cryptosporidium ; elle devrait être appliquée seulement pour des échantillons fécaux constituant plus de 0,41 % du poids du serpent. Au moins cinq échantillons de selles devraient être examinés afin de déterminer la positivité d un serpent pourcryptosporidium en cas d infection subclinique. Cryptosporidium serpentis l cryptosporidiose du serpent l oocyste / serpent captif / frottis fécal INTRODUCTION Cryptosporidium serpentis-associated cryptosoridiosis is the cause of considerable morbidity and mortality in snakes and eliminates valuable specimens via natural death or obligatory euthanization (Upton, 1990; Cranfield and Graczyk, 1995). The early diagnosis of subclinical infections, which allows for the isolation of snakes from collections, and management decisions, which reduce pathogen spread, is beneficial to ophidian centers, enabling them to avoid economic losses. Snake cryptosporidiosis is insidious and difficult to diagnose in a subclinical phase, the most dangerous form of the disease for a colony of captive snakes due to its transmission potential. Detection of Cryptosporidium oocysts by acid-fast stain (AFS) is known to be of low sensitivity and specificity (Graczyk et al, 1996); however, it is still commonly used because its cost is low and no sophisticated equipment is required. Reliable classification of an AFS direct wet smear (DWS) is influenced by the threshold of oocyst detection and human-derived error. Graczyk et al (1996), using fluorescein labeled monoclonal antibody (mab), estimated the concentration of Cryptosporidium oocysts in snake stools within the range of 4.0 x 12/mL 0 to 1.4 x 10 5/mL, and noted 100% false-negativity of AFS DWS when the oocyst concentration was < 7.5 x 10 4/mL. In this previous study, the threshold of detection was estimated, based on the number of oocysts extracted from feces, and semi-quantified by mab. However, the detection threshold associated with the oocyst concentration in the stool specimens remained unknown. Fatigue derived from the time-consuming examination of the fecal smear, and the exposure to fecal fumes during the stool processing may generate discomfort and may contribute to human error. It is therefore important to recognize which features of snake cryptosporidiosis and stool-processing parameters facilitate an efficient conventional diagnosis, and which represent unnecessary tasks being carried out by personnel. The purpose of the present study was to provide useful guidelines for diagnosis of subclinical cryptosporidiosis in snakes for the ophidian centers that routinely use fecal smear technique. These include the determination of the oocyst concentration in the stools of subclinically infected snakes and the threshold of oocyst detection in the fecal smear.

4 MATERIALS AND METHODS Seven snakes (three species, three genera) from the Philadelphia Zoo (Philadelphia, PA, USA) naturally subclinically infected with C serpentis (table I), and one uninfected Elaphe obsoleta obsoleta were used. The infected snakes shed Cryptosporidium oocysts, as determined by Merifluor TM Cryptosporidium/Giardia test (Meridian Diagnostic, Cincinnati, OH, USA), but did not display any clinical signs of infection such as midbody swelling or food regurgitation. All animals were kept individually in separate cages in the same room, isolated from the rest of the ophidian collection though under similar ambient conditions, management protocol (Cranfield and Graczyk, 1995), and diet (Cranfield and Graczyk, 1994). The stool specimens collected for 131 days at defecation were weighed, a DWS prepared from the specimen was stained with AFS, and the material was stored in g-ml-equal volume of 0.37% HCI/0.25% Pepsin A (91 units/mg solid, 620 units/mg protein) medium (Sigma Chemical Co, Saint Louis, MO, USA). The preparation of AFS solution and DWS followed the protocol of Ash and Orihel (1987). DWS were air-dried, fixed with absolute methanol, and stained for 5 min with AFS. All AFS DWS were processed by the same person. Ten negative and 20 positive stool specimens, as determined by AFS DWS, were selected randomly and used for oocyst extraction and purification by cesium chloride (C SCI) gradient centrifugation (Kilani and Sekla, 1987). The extraction and purification procedures were performed at 4 C. The volume of fecal specimens was adjusted to 350 ml with HCI-Pepsin A medium. The fecal material was homogenized by stirring in 500 ml beaker for 1 h and passed through two sieves (WS Tyler Company, Cleveland, OH, USA), 100 and 45 pm, to remove rodent bristles and phosphate/urate crystals, respectively. The specimen was placed in 500 ml separator funnel (Thomas Scientific, Swedesboro, NJ, USA) mounted on a tripod base stand (Thomas Scientific, Swedesboro, NJ) with two GE 2-h-interval timers (Thomas Scientific, Swedesboro, NJ) mounted on the upper distal tripod stand arm. Each tripod stand had three separator funnels attached with clamps. The timers were set up for 1 and 2 h, respectively, and timer-off-derived vibrations caused a smooth descent of the fecal oocyst-enriched sediment from the oblique funnel walls. After 3 h, the bottom sediment was collected by the valve and used for CsCl gradient centrifugation. Extending the sedimentation period for more than 3 h caused packing of the sediment and consequent difficulties in removing the substrate. The CsCl cen-

5 trifugation was performed according to the protocol of Kilani and Sekla (1987) with the modification that all the upper layer (1.05 g/ml CSCI) was collected, and centrifuged ( g, 10 min). The pelleted oocysts were washed five times with water by centrifugation ( g, 10 min) to remove the CsCI. The number of oocysts counted with a hematocytometer (Baxter Healthcare, Columbia, MD, USA) (Fayer and Ellis, 1993) was divided by the fecal weight to determine the concentration of oocysts. To determine the detection threshold of AFS DWS, the extracted and CsCl-purified oocysts were resuspended in water. One milliliter of suspension (10 5 oocysts/ml) was added to 3.0, 2.3 and 1.8 g of the negative stools of E o obsoleta, as determined by Merifluor TM CryptosporidiumlGiardia test, to produce oocyst concentrations of 2.5 x 10 4/g, 3.0 x 10 4/g, and 3.5 4/g, x 10 respectively. Ten fecal smears prepared from each of the three concentration specimens (total 30 smears) were processed and examined blindly, that is without knowledge of the smear identity, by the same person who routinely screened AFS DWS for Cryptosporidium oocysts. All fecal smear-negative stools were processed by Merifluor TM CryptosporidiumlGiardia test as described previously (Graczyk et al, 1996). Ten-well (two rows of five wells) microslides (Carlson Scientific Inc, Peotone, IL, USA) were coated with lysine (Sigma Co, Saint Louis, MO, USA) to enhance the adherence of the fecal substrate. were used to cover three Three 5 pl aliquots 5 mm diameter wells in one row. The wells were air-dried. The Merifluor Tll test procedures were applied according to the manufacturer s instruction. Because of the higher sensitivity of the Merifluor TM test over the AFS DWS method (Graczyk et al, 1995), the fecal specimens positive for Cryptosporidium by the AFS DWS were not processed by Merifluor Tm. In addition, the oocysts isolated from the snakes were processed by the Cryptosporidium Rapid Assay (Alexon, Inc, Sunnyvale, CA, USA) test kit according to the manufacturer s instructions to determine if the ProSpect-11 test can be applied to snake cryptosporidiosis. ProSpect! Cryptosporidium Rapid Assay is a qualitative in vitro diagnostic test for the detection of Cryptosporidium specific antigen (CSA) in stool samples. Stool specimens are applied to membranes on which anti-csa mab are immobilized. If CSA is present in a specimen, it is captured by the mab on the membrane, and a blue color is developed at the reaction site. Statistical analysis was performed with Statistix 4.1 (Analytical Software, Saint Paul, MN, USA). The degree of linear association between variables was compared using Pearson s correlation coefficient (r). The mean values were associated with SD. The variables were examined by the Runs test to determine if their distribution conformed to a normal distribution, and if not, the nonparametric tests were used to assess the significance of differences among variables. The within-snake-species differences in the percentage of positive feces was determined by G-heterogeneity test (Sokal and Rohlf, 1981 The smallest theoretical number of samples (N) (containing at least one positive AFS DWS determination) sufficient to diagnose subclinical Cryptosporidium infection in a snake was computed according to the sampling program equation: N = t2x p x q / D2 (Southwood, 1991 ), where t = Students t of standard statistical tables with n ranging from 4 to 19 (table I), P < 0.05, p probability of positive result (table 1), q = -p, and D = the = predetermined half-width of the confidence limits (D 0.48). The lower 95% confidence interval = limits of the mean weight of Cryptosporidiumpositive stools as determined by fecal smears were compared with the snake weight to determine the percentage of snake weight contributed by the stools. Statistical significance was considered to be P < RESULTS Forty-eight of 81 (59%) stool specimens tested positive by fecal smear (fig 1 and the percentage of positivity varied from 50.0% to 75% (table I). Although the three Pituophis melanoleucus melanoleucus produced significantly higher mean number of stools than the four other species of snakes (Rank sum test: t 1.95, P < 0.04) (table I), the differences in the percentages of positive specimens were not significant among all animals (G-heterogeneity test: G 6.7, = P > 0.05). The highest frequency of defecation, on average every 6-7 days, was observed for the three P m melanoleucus with the lowest for Epicrates cenchria cenchria (on average every 33 days). For the four remaining snakes the frequency of

6 defecation varied between 16 and 22 days (table I). When ten of 33 fecal smear-negative specimens were used for oocyst extraction and CsCl processing, it appeared that all of them contained Cryptosporidium oocysts with a concentration ranging between 6.0 x 10 3/g to 2.5 x 10 4/g (mean: 1.5 x 10 4/g ± 6.7 x 10 3/g). Oocyst concentration in the 20 selected fecal smears which were AFS-positive ranged from 2.9 x 10 4/g to 1.5 x 10 5/g (mean: 6.6 x 10 4/g ± 8.0 x 10 3/g). The concentration of oocysts in the AFS DWS-false negative stools was significantly correlated with the weight of fecal specimens (r= 0.76, P < 0.01 Significant correlation was also observed when the oocyst concentrations of fecal smear-positive stools were plotted against the weight of stools (r= 0.77, P < 0.01) (fig 2). Overall, the oocyst concentration in all 30 fecal specimens was significantly correlated with the weight of the stools (r= 0.86, P < 0.01 ) (fig 2). All ten fecal smears prepared from the specimens with 2.5 x 104 oocysts per gram were classified as negative, whereas the

7 remaining 20 smears prepared from the fecal samples with oocyst concentrations of 3.0 x 10 4/g and 3.5 x 10 4/g were determined as positive by routine examination. The Runs test revealed that the distribution of the stool weights did not conform to normal distribution for the three P m melanoleucus or for E c cenchria (P < 0.04). The difference in weight among the three P m melanoleucus was not significant (Kruskal-Wallis ANOVA: F= 0.12, P > 0.05). The total weight range of the 33 AFS DWS-false negative stools was 0.3 to 1.3 g with a mean of 0.6 ± 0.2 g. The mean weight of those samples was significantly lower (Rank sum test: t = 7.72, P < 0.01) than the mean weight (9.6 ± 7.9 g, range: g) of the 48 samples tested positive by the fecal smear. Interestingly, the lower the snake weight, the lower the mean weight of positive stools, and the smaller the difference between the means of positive and negative stools. The last two variables were significantly correlated with snake weight (r = 0.62, P < 0.03; r = 0.63, P < 0.03, respectively). The differences in the weight of the smear-positive stools were significant among the snakes (Kruskal-Wallis ANOVA: F = 6.33, P < 0.05), whereas these differences were non-significant for the smearfalse negative stools. Overall, all fecal specimens that constituted more than 0.73% of the snake weight were diagnosed as positive by the fecal smears. The smear-positive stools exceeded 0.68, 0.99 and 0.94% of the weight of the three P m melanoleucus, respectively (from the data expressed in table I). They were 2.00%, and 0.70% of the weight of the two Lampropeltis triangulum conati, and 0.41 % for E c cenchria, and 0.48% for L t annuata (table I). The smallest theoretical number of stools to be examined by fecal smear to result in at least one positive determination varied within the range of five to seven for the seven snakes. With 95% confidence limits, one positive fecal smear determination would be identified from at least five fecal smears from each of the three P m melanoleucus, from six stools of each of the two L t conati, and from 7 stools of each E c cenchria and L t annuata. Overall, all 81 fecal specimens were positive for Cryptosporidium oocysts; 48 were diagnosed by the fecal smear, and the other 33 by the MerifIuorT&dquo;^ CryptosporidiumlGiardia test. In two snakes, oocyst-positive stools were misdiagnosed by AFS DWS three subsequent times, and a double-subsequent misdiagnosis occurred three times in two snakes (fig 1 ). All seven oocyst isolates originating from the seven snakes produced a negative reaction with the ProSpect test. The false-negativity of the fecal smear was 41 % and its sensitivity 59%. DISCUSSION In the present study, 41 % of the oocyst positive stools were missed by the AFS fecal smear. In two snakes, positive samples were misdiagnosed three times subsequently, and a double-subsequent misdiagnosis occurred three times in two snakes (fig 1 ). In a possible reptile center scenario, these snakes would very likely pass the quarantine and join a collection of snakes thus serving as a reservoir for the pathogen. A previous study (Cranfield and Graczyk, 1994) showed that spreading the pathogen via mechanical water-associated contamination is effi- occur. Outbreaks cient and would most likely of cryptosporidiosis have been frequently reported in colonies of captive snakes (Carmel and Groves, 1993). The threshold of detection for Cryptosporidium oocysts by the fecal smear was 3.0 x 10 4/g; and all the stools having lower oocyst concentration than the threshold were false-negatively misdiagnosed. The present study showed that AFS fecal

8 smears should be used exclusively for the determination of Cryptosporidium-positive snakes, but not for diagnosis of snake negativity for Cryptosporidium. Even multiple subsequent oocyst-negative fecal smears should not establish a base for any conclusions or management decisions regarding snake Cryptosporidium infections. If this technique is going to be used for the determination of Cryptosporidium-negativity of snakes, subclinically infected animals will most likely pass the quarantine. The improper and poorly understood use of this technique may result in the spread of the pathogen within a snake collection. The negativity of a snake for Crypfosporidium should be established based on another technique, eventually the MerifIuor T&dquo;&dquo; Cryptosporidium/Giardia test (Graczyk et al, 1996). Unfortunately, as shown in the present study, the Cryptosporidium Rapid Assay test produced a negative reaction with seven C serpentis oocyst isolates, indicating its low applicability for the diagnosis of snake cryptosporidiosis. Despite the high number of misdiagnosed fecal specimens by AFS DWS, we concluded that this technique should not be rejected from laboratory procedures but that its approach simply needs to be modified. The technique should selectively target bigger, consequently heavier, fecal specimens. As shown herein, the stools that did not constitute at least 0.41 % of snake weight contained a low number of oocysts that resulted in a concentration below the threshold of the fecal smear detection. In practice, such fecal specimens are very likely to be falsenegatively misdiagnosed. Because of the availability of the weight records of quarantined snakes, the limits for weight of fecal specimens considered for screening by fecal smears would be easy to establish and could be incorporated in snake diet-datarecord sheet. The assertion of the present study that subclinical cryptosporidiosis in a snake can be diagnosed based on one oocyst-positive fecal smear required that at least seven fecal specimens be examined. Considering the frequency of snake defecation (table I), the screening procedure can be accomplished within a month to a half year depending on the snake species. Thus, the quarantine period in an ophidian center routinely using this conventional technique should be adjusted accordingly. The overall distribution of snake stool weights did not conform to normal distribu- true for the three tion. This was particularly frequently defecating P m melanoleucus. The disproportion in stool mass is a characteristic feature of the physiology of the snake digestive tract (Skoczylas, 1978). The gut contents may remain in the intestine for various time-periods, and the number of defecations is unequal to the number of meals consumed (Skoczylas, 1978). The snake cloaca serves as a passway or as a storage space for feces and urine (McDonald, 1976); thus, in the low-weight fecal specimens, the urine constituted a greater weight-fraction than in large stools. This may explain the low concentration of the oocysts in the light-weight fecal specimens. Studies by Secor et al (1994) showed an upregulation of snake intestine activity in response to feeding. They found that metabolic rates and brush-border nutrient transport (glucose, leucine, and proline) exceeded at least 22 times the fasting values, and the intestinal mass increased more than twofold. The increased gastric epithelium activity and availability of intra- and extracellular nutrients might affect the reproduction of the intracellular ectoplasmatic trophozoites of Cryptosporidium in the gastric epithelium. Thus, the intensity of the pathogen reproduction processes could be related to the magnitude of the gastric epithelium responses induced by the different-sized meal (Secor et al, 1994). This would explain the higher concentration of Cryptosporidium oocysts in larger snake

9 stools. The concentration of Cryptosporidium oocysts was positively correlated with the stool mass; however, because all of the snakes were subclinically infected, the question of how this correlation is applicable to snake clinical cryptosporidiosis needs to be answered. The described modification of the oocyst extraction considerably accelerated this process and reduced the laboratory space necessary for the simultaneous processing of a high number of stools. The use of the cold room (4 C) reduces the discomfort associated with exposure to fecal fumes, and slows the process of Crypfosporidium-antigen degradation, which is important if the oocysts are designed for molecular studies. In addition, the use of individual snake separator funnels prevents oocyst isolate-toisolate contamination. The results of the present study may assist ophidian centers in improving their diagnosis of subclinical cryptosporidiosis in snakes by the proper interpretation of the fecal smear results. In particular, the discomfort of the task related to the stool-processing can be reduced to a minimum without decreasing efficiency of the conventional detection of Crypfosporidium oocysts. ACKNOWLEDGMENTS We thank SJ Maltese, D Heyl and T Horach for their technical assistance, and K Wolff for her editorial comments. We acknowledge the Philadelphia Zoo (Philadelphia, PA, USA) for providing the snakes. This study was supported by the Maryland Zoological Society, and the AKC Fund of New York. REFERENCES Ash LR, Orihel TC (1987) Parasites: A Guide to Laboratory Procedures and Identification. ASCP Press, American Sociey of Clinical Pathologists Inc, Chicago, 328 Carmel BP, Groves V (1993) Chronic cryptosporidiosis in Australian elaphid snakes: control of an outbreak in a captive colony. Aust Vet J 70, Cranfield MR, Graczyk TK (1994) Experimental infection of elaphid snakes with Cryptosporidium serpentis (Apicomplexa: Cryptosporidiidae). J Parasitol 80, Cranfield MR, Graczyk TK (1995) Cryptosporidiosis. In: Manual of Reptile Medicine and Surgery (DR Mader, ed), WB Saunders Company, Philadelphia, Fayer R, Ellis W (1993) Paromomycin is effective as prophylaxis for cryptosporidiosis in dairy calves. J Parasitol79, Graczyk TK, Cranfield MR, Fayer R (1996) A comparative assessment of direct fluorescence antibody, modified acid fast stain, and sucrose flotation techniques for detection of Cryptosporidium serpentis oocysts in snake faecal specimens. J Zoo lnildl Med 26 (in press) Kilani RT, Sekla L (1987) Purification of Cryptosporidium oocysts and sporozoites by cesium chloride and percoll gradients. Am J Trop Med Hyg 36, McDonald HS (1976) Methods for the physiological study of reptiles. In: Biology of the Reptilia (C Gans, WR Dawson, eds), vol 5, Academic Press, Hartcourt Brace Jovanovich, New York, Secor SM, Stein ED, Diamond J (1994) Rapid upregulation of snake intestine in response to feeding: a new model of intestinal adaptation. Am J Physiol 266, G695-G705 Skoczylas R (1978) Physiology of the digestive tract. In: Biology of the Reptilia (C Gans, WR Dawson, eds), vol 8, Academic Press, Hartcourt Brace Jovanovich, New York, Sokal RR, Rohlf FJ (1981) Biometry. WH Freeman and Company, New York, 859 Southwood TRE (1991) The sampling programme. In: Ecological Methods (TRE Southwood, ed), Chapman and Hall, New York, Upton SJ (1990) Cryptosporidium spp in lower vertebrates. In: Cryptosporidiosis in Man and Animals (JP Dubey, CA Speer, R Fayer, eds), CRS Press, Boca Raton, FL,