JOURNAL OF CLINICAL MICROBIOLOGY, May 1999, p. 1441–1446 0095-1137/99/$04.0010 Copyright © 1999, American Society for Microbiology. All Rights Reserved.
Vol. 37, No. 5
Latent Pneumocystis carinii Infection in Commercial Rat Colonies: Comparison of Inductive Immunosuppressants plus Histopathology, PCR, and Serology as Detection Methods STEVEN H. WEISBROTH,1 JAMES GEISTFELD,2 STEPHANIE P. WEISBROTH,1 BRUCE WILLIAMS,3 SANFORD H. FELDMAN,1,4 MICHAEL J. LINKE,5 SALLY ORR,5 AND MELANIE T. CUSHION5* Anmed/Biosafe, Inc., Rockville, Maryland1; Taconic Farms, Inc., Germantown, New York2; Armed Forces Institute of Pathology3; University of Cincinnati College of Medicine and the VA Medical Center, Cincinnati, Ohio5; and Health Services Center, University of Virginia, Charlottesville, Virginia4 Received 26 May 1998/Returned for modification 6 November 1998/Accepted 2 February 1999
Histopathologic evaluation combined with a period of immunosuppression has been the standard procedure for detection of Pneumocystis carinii in commercial rat colonies. Variation in induction regimens and in the sensitivity of detection methods may result in underreporting of the presence of P. carinii in breeding colonies or delay its detection. In the present study, methylprednisolone and cyclophosphamide were evaluated for the ability to induce P. carinii infection in rats from an enzootically infected commercial barrier colony. The presence of P. carinii was detected by histopathologic methods and by amplification of a targeted region of the P. carinii thymidylate synthase gene by PCR over the 8-week study period. Sera taken from rats prior to either induction regimen were evaluated for the presence of P. carinii-specific antibodies by the immunoblotting technique. Few significant differences in ability to induce organism burden or in histopathology were observed between the two immunosuppressive regimens. However, a dramatic loss of weight over the study period was observed in rats treated with methylprednisolone but not in rats treated with cyclophosphamide. Although histopathologic changes attributable to P. carinii did not appear before 2 weeks with either immunosuppressant, the presence of the organism in these animals was detected by immunoblotting and PCR. Cyst scores and the intensities of the histopathologic lesions increased during the study period, but the number of rats exhibiting evidence of P. carinii infection did not change after week 3. These results suggest that use of the PCR method on postmortem lung tissue of rats without prior induction regimens or identification of anti-P. carinii antibodies in antemortem serum samples is a sufficiently sensitive method for detection of the presence of a P. carinii carrier state in rodent breeding colonies. In the cost-benefit context of diagnosis, prolongation of the inductive period beyond a point enabling confident diagnosis becomes counterproductive. In the present study, specific parameters of the stress test were evaluated in an effort to establish optimal induction and detection methods for the presence of P. carinii. A large, commercial, gnotobiotically derived, barrier-maintained breeding colony of Fischer 344 (F344) inbred rats previously known to have an almost 100% incidence of clinically silent, latent P. carinii infection was selected as the study population (13). The high incidence of the organism in this colony helped to ensure that observed differences were due to experimental manipulations and not to sporadic infection levels. Only adolescent males in the 200- to 250-g weight range (corresponding to the 8- to 10-week age group) were used, permitting other parameters of the stress test equation to be manipulated while holding the age-weight-sex invariant. Two of the most widely used immunosuppressants, methylprednisolone and cyclophosphamide, were evaluated for progression of organism burden and pathological lesions associated with the infection in the rats’ lungs over an 8-week period. In addition, targeted amplification of a region of a putative single-copy gene of P. carinii, thymidylate synthase (TS), by PCR was compared to a standard histopathologic scoring system to determine the most effective method for detection of the presence of P. carinii. Serologic detection of antibodies specific for rat P. carinii by the immunoblotting technique was explored as a potential nonin-
Diagnosis of intercurrent but asymptomatic and histopathologically cryptic infection of laboratory rodents by Pneumocystis carinii has historically required chronic treatment with immunosuppressants to depress the hosts’ immune systems, permitting resident or environmentally acquired P. carinii to replicate to microscopically observable levels (1). The same technique, commonly referred to as the “stress test”, can also be used to provide evidence that rodents in commercial or laboratory colonies are free of P. carinii contamination (33). The stress test used by commercial vendors to certify the P. carinii-free status of breeding colonies has been based on animal models of experimental pneumocystosis and follows a protocol of chronic immunosuppression for several weeks. Few studies have focused on the early and most cost-efficient detection of P. carinii in large-scale-production situations. Critical parameters, such as minimum duration for induction of detectable infection, the qualities of alternative immunosuppressants, the dosing regimen, the scoring of infection intensity, the age-weight-sex status of the subject rodents, the utility of alternative detection methods, and the very definitions of success and failure in diagnosis, have all been directed towards provoking infections suitable for experimental manipulation.
* Corresponding author. Mailing address: VA Medical Center, Mail Location 151, 3200 Vine St., Cincinnati, OH 45220. Phone: (513) 8613100, ext. 4417. Fax: (513) 475-6415. E-mail:
[email protected]. 1441
1442
WEISBROTH ET AL.
vasive means for predicting progression to infection or a P. carinii-free state. MATERIALS AND METHODS Animals. The study was organized into two separately staged receipts of randomly selected groups of 24 and 27 rats from the enzootically infected F344 barrier rat production unit at Taconic Farms, Germantown, N.Y. The total census of rats during this period was approximately 3,500. Male rats in this study were taken from a population of ;1,000 nonbreeding animals. Based on periodic health surveillance tests, the colony was known to be free of common rodent viral, bacterial, and parasitic infections. On the day of receipt, each animal was bled to provide a preinduction serum sample, weighed, and treated with one of two immunosuppressants to initiate the inductive schedule. The first group of 24 adolescent (200- to 250-g) rats was randomly divided into eight groups of 3 rats per cage (four groups per immunosuppressant) and placed into sterilized rat microisolator-type cages (Lab Products, Seaford, Del.) provided with sterilized bedding, autoclaved bottled water supplemented with 1 mg of tetracycline (Sigma Chemical Co., St. Louis, Mo.)/ml, and sterilized pelleted diet (Purina Mills, St. Louis, Mo.). The second group of 27 rats was received, divided into groups of 3, and housed identically to the first, with the exception that one group of 3 rats was subjected to terminal sample collection procedures (described below) on the day of arrival to constitute a preinduction baseline. Immunosuppressive regimen. The first group of 24 rats was divided into eight cages with 3 rats per cage for the first stage of testing. The second shipment of 27 rats had 3 rats sacrificed upon receipt prior to induction of immunosuppression and then was likewise divided for the second stage of testing. At each stage, the rats in four cages were dosed with cyclophosphamide monohydrate (Sigma Chemical Co.) (33 mg/kg of body weight given once per week by intraperitoneal injection) and, in parallel, the rats in the other four cages were dosed with methylprednisolone acetate (Depomedrol; Upjohn, Kalamazoo, Mich.) (4 mg/kg given once per week by subcutaneous injection) (9). The animals were weighed weekly when removed for immunosuppressive injection (or for terminal procedures), and their clinical condition was evaluated by inspection. One cage of three rats dosed with each immunosuppressant was withdrawn from the first stage for terminal procedures at the end of 4, 5, 6, and 8 weeks of induction and, for the second stage, at the end of 1, 2, 3, and 4 weeks of induction. The 4-week group was replicated to confirm comparability and continuity of inductive effects between the two stages. Terminal procedures. Each animal was anesthetized with CO2 for blood sample collection by cardiac puncture and then euthanatized by overanesthesia. A gross necropsy dissection was performed on each rat in the study. The right anterior lung lobe was collected and frozen at 270°C for DNA extraction prior to PCR determinations. The large single left pulmonary lobe was excised and fixed in 10% neutral buffered formalin for histologic preparations. Histologic sections were made by standard methods for paraffin-embedded blocks, cut at a thickness of 5 mm, and stained alternatively with hematoxylin and eosin for assessment of overall inflammation and histopathologic lesions or with Gomori’s methenamine silver (GMS) for scoring of P. carinii cysts (Table 1, footnote a). Immunoblotting technique for detection of antibodies specific for P. carinii. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and immunoblotting were performed as previously described (19). Briefly, sera collected from the study rats were diluted 1:10 in phosphate-buffered saline (PBS) solution and used directly for reaction with the blotted, separated proteins of a standard P. carinii antigen preparation. The antigen preparation originated from organisms obtained from Sasco-Sprague-Dawley rats after induction of a naturally acquired infection of P. carinii f. sp. carinii form 1 (9) by chronic administration of methylprednisolone (4 mg/kg) for 8 to 12 weeks. Organisms representing all developmental stages were prepared as a lung homogenate with a Stomacher Lab Blender 80 (Tekmar Inc., Cincinnati, Ohio), with subsequent reduction of erythrocytes by exposure to aqueous 0.85% ammonium chloride. The preparations were stored at 220°C. Prior to use, the antigen preparation was solubilized in SDS lysis buffer (2% SDS–0.06 M Tris [pH 6.8]–1% glycerol–5% 2-mercaptoethanol) and subsequently electrophoresed through 0.1% SDS–discontinuous polyacrylamide minigels. Separated proteins were transferred to nylon membranes by use of a Trans-Blot apparatus (Bio-Rad) for 1 h at 100 V. The efficiency of transfer was verified by staining with Ponceau red. The blots were blocked with an aqueous solution of 1% powdered milk and probed overnight with the rat sera. Immunoreactive bands were detected with goat anti-rat immunoglobulin G conjugated with horseradish peroxidase followed by development in 4-chloro-1-naphthol substrate solution. Sera were scored as positive if an immunoreactive band appeared in the areas of migration between 100 and 120 kDa and/or 45 and 55 kDa (27). Tissue preparation and DNA extraction. Frozen lung tissue was received from Anmed/Biosafe, Rockville, Md., in the facilities of the VA Medical Center, Cincinnati, Ohio, and stored immediately at 270°C until use. After being rapidly thawed by immersion and shaking in a 37°C water bath, a 50- to 115-mg portion of the lung tissue was dissected from the entire sample with sterile forceps and a new razor blade for each sample. Two methods of preparation were used. For the first-stage samples, a volume of 13 Taq polymerase buffer (Promega, Madison, Wis.) was added to the tissue in an amount equal to the weight. Proteinase K (Boehringer Mannheim, Indianapolis, Ind.) was added to a final concentration
J. CLIN. MICROBIOL. of 1 mg/ml. Digestion at 55°C was allowed to proceed overnight and was stopped by heating the samples to 95°C for 10 min. Samples of the digestate were used directly or diluted 10-fold for PCR. The second-stage tissue samples within the same size range as those of the previous stage, were first homogenized in sterile plastic bags with a Stomacher 80 with a 1-ml volume of PBS for 5 min at the highest setting. After centrifugation at 2,400 3 g, the supernatant was removed, an equal volume of PBS was added to the pellet, and digestion proceeded as described for the first-stage tissue samples. This second method was used for retesting the tissue samples that yielded negative results. PCR (DNA amplification) conditions. The following primer sequences directed to a 403-bp portion of the rat P. carinii TS gene were used to evaluate the presence of P. carinii DNA in the rat lungs: PC1, 59-ATT TAT GGG TTT CAA TGG-39, and PC2, 59-GTT CCC TTT AAT ATT GCA-39, corresponding to 379 to 396 and 763 to 781 of the published TS gene of P. carinii (12). Primers for the rat globin gene, Glob1 and Glob2, were used to assess the integrity of DNA in the samples and to assay for inhibitory factors within the digests as previously described (21). The globin primers produce an amplicon of 400 bp. A 1.0-ml sample of each homogenate or digest, undiluted or a 10-fold dilution, was used as a template for each set of primers in a reaction mixture that contained 10 mM Tris-HCl (pH 8.0), 50 mM KCl, 3 mM MgCl2, 0.5 mM deoxynucleoside triphosphates, 0.5 mM (each) primer, and 1 U of Taq polymerase (Promega). Each sample was overlaid with ;25 ml of mineral oil, denatured for 2 min at 94°C, and subjected to 30 cycles of amplification that consisted of a 60-s 94°C denaturation step, a 60-s 50°C annealing step, and a 60-s 72°C elongation step. Each set of reactions with the TS primers included a negative (water) control and a positive P. carinii-infected rat lung homogenate. Amplifications with rat b-globin primers were conducted simultaneously in separate reaction tubes. Electrophoresis. A 15-ml aliquot from each reaction was electrophoresed at 100 V for 1 to 2 h through a 0.7% agarose gel containing ethidium bromide for the visualization of products. The amplicons were imaged on a Fotodyne Foto/ Eclipse system with NIH Image version 1.55 software. Reactions were scored as negative if no band of the appropriate size was detected after averaging 30 frames. Histopathologic evaluation. The pulmonary sections prepared as described above were evaluated at the end of the study by a veterinary pathologist blinded to the immunosuppressant group and the week of the study. The hematoxylinand-eosin-stained sections were used to score overall pulmonary inflammatory changes and were scored separately for the occurrence of lesions attributable by their character to the effects of P. carinii, as outlined in the grading scheme in footnote a to Table 1. The GMS-stained sections were used to score the presence and relative numbers of P. carinii cyst forms seen in pulmonary tissues and were likewise graded according to the scheme in footnote a to Table 1. Statistical evaluation. The Pearson correlation product moment was calculated for the development of histopathologic changes and changes in body weight as a function of time for both inductive regimens by the parameters in footnote a to Table 1. Student’s t test was performed to determine the level of significance for each product moment determined. A Wilcoxon rank sum test was used to compare the two immunosuppressive regimens for each histologic parameter in footnote a to Table 1. Cochran’s test was used to compare PCR and serology as methods for the detection of P. carinii.
RESULTS The animals in this study were evaluated weekly for weight loss, histopathologic changes in the lungs, organism burden, and the presence of P. carinii DNA in the lungs by use of PCR as induction progressed over the 8-week period. The weight loss experienced by the rats in the groups treated with methylprednisolone was found to be statistically different than the weight gain of the rats receiving the alternative immunosuppressant therapy, cyclophosphamide. The mean body weights for each group of three rats, taken at weekly intervals until the scheduled termination, are summarized in Fig. 1. A similarly sloped weight loss occurred with all methylprednisolone-treated groups, beginning after the first inductive injection (r 5 20.95; P , 0.001). In contrast, all groups treated with cyclophosphamide adhered essentially to the normative growth curve for F344 male rats (r 5 0.99; P , 0.001) (25). As discussed below, the observed weight loss was attributed to the corticosteroid treatment and not to the effects of progressive pneumocystosis. There were no significant findings based on the gross appearance of the lungs during necropsy dissection. Pulmonary changes within the rat lungs detected by histology are summarized in Tables 1 and 2. The criteria for assessment are found in footnote a to Table 1. In the rats sampled prior to induction, some background inflammation was ob-
DETECTION OF P. CARINII IN COMMERCIAL RAT COLONIES
VOL. 37, 1999
1443
TABLE 1. Histopathologic lesion and cyst scores and Western immunoblot and PCR results in rats during methylprednisolone inductiona
Animal no.
40–42 43–45 46–48 49–51 13–15 16–18 19–21 22 23–24 Total or avg
Weeks of induction
1 2 3 4 4 5 6 Died 8
Western immunoblotting (no. positive/no. tested)
Pulmonary histopathology GMS cyst scoreb (no. positive/no. tested)
Overall inflammation
Pneumocystosis lesionsb
0.66 1.00 1.00 1.00 1.00 1.33 0.66 NDc 1.50
0.33 1.00 0.33 0.66 1.00 1.66 1.00 ND 1.50
0.00 (0/3) 1.33 (3/3) 1.33 (3/3) 2.00 (3/3) 1.33 (3/3) 2.00 (3/3) 1.33 (3/3) ND 3.00 (2/2)
1.02
0.94
1.54 (20/23)
PCR (no. positive/ no. tested)
Preinduction
Terminal
3/3 3/3 1/3 0/3 2/3 3/3 3/3 ND 1/2
3/3 3/3 2/3 2/3 2/3 2/3 3/3 ND 2/2
2/3 3/3 3/3 2/3 2/3 2/3 1/3 ND 2/2
16/23
19/23
18/23
a
Histopathologic grading scales are as follows. Overall pulmonary inflammations (hematoxylin and eosin sections): 0, no discernible inflammation; 1, inflammation affecting less than 25% of examined tissue; 2, inflammation affecting 25 to 50% of examined tissue; 3, inflammation affecting 50 to 75% of examined tissue; 4, inflammation affecting over 75% of examined tissue. Pneumocystis lesions (hematoxylin and eosin sections): 0, no microscopic evidence of Pneumocystis infection; 1, multifocal poorly defined areas of inflammation consistent with Pneumocystis infection; 2, one or two well-defined areas of alveolar filling highly suggestive of Pneumocystis infection; 3, more than two well-defined areas of alveolar filling highly suggestive of Pneumocystis infection. Pneumocystis forms (Gomorri’s methenamine silver sections): 0, no cysts seen; 1, rare single cysts seen at a rate of less than 1 per 10 403 fields; 2, 1 to 10 cysts per 10 403 fields; 3, 10 to 100 cysts per 10 403 fields; 4, .100 cysts per 10 403 fields. b Means of groups of three rats. c ND, not done.
served, although there were no lesions or silver-staining forms suggestive of the presence of significant numbers of P. carinii. Since little is known about the first stages of P. carinii infection or colonization in the host, it is possible that this inflammation was due to the presence of small numbers of P. carinii undetected by the silver stain or to the presence of trophic forms of P. carinii that do not stain with the silver stain. Another possible explanation for the mild inflammatory reaction is the presence of an unidentified microbial population or idiopathic host process. Rats treated with methylprednisolone (Table 1) manifested lesions attributable to pneumocystosis, and silverstaining cyst forms typical of Pneumocystis were detected at the end of the second week of induction, as was also the case for the cyclophosphamide-treated series (Table 2). There was a significant trend for the intensity of lesions attributable to pneumocystosis (cyclophosphamide, r 5 0.56 and P , 0.01; methylprednisolone, r 5 0.65 and P , 0.01) and for the number of silver-staining cyst forms (cyclophosphamide, r 5 0.45 and P , 0.05; methylprednisolone, r 5 0.55 and P , 0.01) to increase between weeks 2 and 8 for both immunosuppressants. No significant differences in the efficiencies of the immunosuppressants in facilitating histopathologic diagnosis by inducing overt expression of disease were demonstrated in the groups of rats evaluated in this study. Additionally, the results indicate that there was no further increase in the efficiency of detection of histopathologically positive animals after the third week of induction with either immunosuppressant. In all, 20 of 23 (87%) rats induced with methylprednisolone were found histopathologically positive for pneumocystosis, whereas 16 of 23 (70%) rats were similarly detected when induced with cyclophosphamide. The seeming difference between the effects of the immunosuppressants was probably due to some variability in the incidence of natural infection between groups, although we chose this particular colony to minimize such differences. The dramatic body weight loss observed with all groups treated with methylprednisolone was interpreted as a direct effect of the corticosteroid rather than of pneumocystosis, as the incidences of P. carinii within all of the groups were quite similar and the weight loss was profound in only the methylprednisolone-treated animals. We acknowledge that some slight differences in the abilities of the immunosuppressants to provoke
histopathological lesions may be uncovered by the use of larger groups of animals, but it was the strategy of the present studies to use small numbers of animals for the purpose of cost efficiency in the commercial facility setting. The results of the PCR analysis showed this technique to be useful for diagnostic detection, especially during the first 2 weeks of induction. Two of the three untreated rats were detected as positive by PCR, and four of six rats were detected as
FIG. 1. Mean body weights of rats during immunosuppression induction with methylprednisolone and cyclophosphamide. The weeks of immunosuppression and immunosuppressant used are indicated in the legend within the figure.
1444
WEISBROTH ET AL.
J. CLIN. MICROBIOL.
TABLE 2. Histopathologic lesion and cyst scores and Western immunoblot and PCR results in rats during cyclophosphamide inductiona
Animal no.
28–30 31–33 34 35–36 37–39 1–3 4–6 7–9 10–12 Total or avg a b c
Weeks of induction
1 2 Died 3 4 4 5 6 8
Western immunoblotting (no. positive/no. tested)
Pulmonary histopathology GMS cyst scoreb (no. positive/no. tested)
Overall inflammation
Pneumocystosis lesionsb
0.66 2.00 NDc 1.00 1.33 2.33 1.33 2.33 1.33
0.33 1.33 ND 0.50 0.33 2.00 1.66 2.33 0.33
0.00 (0/3) 0.66 (2/3) ND 1.00 (2/2) 1.66 (3/3) 2.33 (3/3) 1.33 (2/3) 1.33 (3/3) 0.33 (1/3)
1.54
1.10
1.08 (16/23)
PCR (no. positive/ no. tested)
Preinduction
Terminal
2/3 1/2 ND 2/2 3/3 3/3 3/3 3/3 3/3
2/3 1/2 ND 2/2 3/3 3/3 3/3 3/3 2/3
2/3 2/3 ND 2/2 3/3 3/3 2/3 2/3 1/3
20/23
19/23
17/23
Histopathologic scoring is described in note a to Table 1. Means of groups of 3 rats. ND, not done.
positive after 1 week of induction (two of three in each immunosuppressant series) (Tables 1 and 2). Thus, six of nine rats were found positive by PCR under circumstances that would have resulted in false negatives if diagnosis were based on histopathologic results alone. This discrepancy was abolished by the end of the second through eighth weeks, since the results of histopathology and PCR were equivalent, 20 of 23 versus 19 of 23, respectively, with methylprednisolone and 16 of 23 versus 17 of 23, respectively, with cyclophosphamide. Thus, the chief advantage of PCR lay in the detection of P. carinii before microscopically observable recrudescence of the organisms. The serologic results were quite significant, in that 37 of 49 (75.5%) preinduction samples were positive for antibodies to P. carinii as detected by immunoblotting. The frequency of serologic positives in the terminal samples, 38 of 46, or 82%, was not diminished by the immunosuppressive regimen. Serologic results in general paralleled those of PCR, and for both immunosuppressants combined, there were no significant differences (Cochran’s test; Q 5 0.82) between the detection rates of terminal serology and PCR (about 83%). Results for the nine rats in the untreated group and those given a single week of immunosuppressant treatment were revealing: although all of the rats were histopathologically negative, six of nine were serologically positive and six of nine were positive by PCR. However, the correlation between the methods was not particularly high (three of nine were PCR positive but serologically negative, and likewise, three of nine were serologically positive but PCR negative). There were some discrepant results among PCR, immunoblot serology, and histopathology. In the cyclophosphamide series, PCR failed to detect 1 of 17 rats known to be positive by histopathology, and likewise, in the methylprednisolone series, 4 of 23 rats were positive by histopathology but undetected by PCR. Reevaluation of the lungs by resampling of the lung tissue followed by homogenization prior to digestion improved the sensitivity of the procedure, resulting in positive amplification of three of the four samples previously negative by PCR. These data suggest that the sensitivity of this PCR method depends on the method of tissue sampling, and they imply that the organisms may be focally distributed within the colonized or infected animal lung. DISCUSSION A primary goal of this study was to define the parameters of the immunosuppressive stress test sufficiently to establish
guidelines enabling confident diagnosis of rat populations harboring P. carinii organisms. Under the conditions of this study, i.e., given the rat strain (F344), age, and sex, these parameters were determined to be as follows. (i) Methylprednisolone (4 mg/kg/week) and cyclophosphamide (33 mg/kg/week) functioned equivalently in forcing expression of histopathologically confirmed pneumocystosis in ;80% of the rats in the study. (ii) A 2-week induction period was sufficient to enable histopathologic detection of at least two rats in a group of three, thereby establishing the presence of P. carinii in the rat colony. There was no further increase in the frequency of histopathologic confirmation after the third week of induction. (iii) Immunoblotting of sera from the 49 rats used in these studies showed reactivity to P. carinii antigens in 38 rats (77.5%) prior to any immunosuppressive therapy. There were no significant increases or decreases in serologic detection by virtue of induction. (iv) Amplification of a region of the P. carinii TS gene by PCR permitted detection of an apparent carrier state in all groups of rats, including those surveyed on the day of arrival, those in a noninduced group, and the terminal samples from rats receiving a single dose of either immunosuppressant. As with immunoblotting detection, a positive PCR did not require immunosuppressive induction to detect P. carinii carriers. The two drugs used as immunosuppressants in the study have very different mechanisms of action (5, 11). Methylprednisolone is a glucocorticoid with profound effects on the immune system, especially lymphocyte number and function, and on almost every organ system. Thus, chronic administration of glucocorticoids leads to skeletal muscle wasting, resulting in reduction of body mass, as observed for the rats in this study. The finding of weight loss in rats (and mice) undergoing corticosteroid immunosuppression for expression of pneumocystosis has been previously reported (3, 14, 32) and has been used to gauge the progression of the disease process in many laboratories. In contrast, cyclophosphamide is a DNA-alkylating agent, the effects of which result primarily in inhibition of DNA synthesis, particularly in proliferating cells. Cachexia has not been reported for this drug. The weight gain (or noninterference with normative growth) of the rats in this study during immunosuppression with cyclophosphamide appears to be a new finding, since previous studies in which cyclophosphamide was used to express latent agents have not followed animal weights during induction (6, 26, 29). An 8- to 10-week induction period has been a standard recommendation for detection of P. carinii in commercial rodent colonies (15), a time period used to maximize expression of
VOL. 37, 1999
DETECTION OF P. CARINII IN COMMERCIAL RAT COLONIES
pneumocystosis for purposes unrelated to detection of latent infection (2, 3, 22, 32). Data from the present study suggest that a 2-week induction period should be sufficient to histopathologically support or reliably refute the existence of the latent carrier state in sample groups drawn from breeding colonies. Serologic diagnosis by the immunoblotting technique was quite effective in detecting antibodies indicative of exposure to P. carinii in all groups of rats prior to induction. The presence of antibodies to P. carinii antigens in nonimmunosuppressed rats has been reported extensively in the literature (4, 7, 20, 31) but has not been routinely accepted as a criterion for the expected development of P. carinii pneumonia, since the presence of antibodies does not always correlate with development of pneumocystosis and their absence does not guarantee the lack of infection (7). Our study shows that .75% of a subpopulation of rats taken from an enzootically infected animal colony (representing about 5% of the male nonbreeding population) were seropositive to rat P. carinii. Thus, in such a setting, random sampling of animal sera for the presence of anti-P. carinii antibodies would be useful to qualitatively assess the presence of P. carinii within a large commercial rat colony. There are, nonetheless, several factors that hinder endorsement of serology as a generally useful diagnostic tool in the commercial setting at the present time. Chief among them is the practical problem of the lack of availability of validated testing reagents (antigens) in bulk and in sustained production to support high-volume testing programs. There are theoretical problems as well, which relate to documented demonstrations of genetic (hence, immunologic) diversity of putative P. carinii strains and species among host species (4, 17, 30), as well as evidence for genetic and antigenic diversity of P. carinii distributed within a given host species (8–10, 27), which argues against the designation of a universal antigen preparation. There are two important considerations favoring development of a diagnostic serologic tool. First, the test could be performed on individual experimental rats without requiring their sacrifice, prior to a large-scale animal study or purchase, using a relatively noninvasive blood collection procedure. Second, it is conceivable that an antigen preparation suitable for the survey of P. carinii within a particular mammalian species (e.g., the rat) will become available as the repertoire of the surface antigens is more fully defined and common epitopes are identified (17–19, 24). Although the animals were histopathologically negative, P. carinii amplicons were produced by PCR from the lungs of six of nine terminally collected samples from noninduced rats and from those that received a single immunosuppressant dose. While suggestions that PCR could be used to detect the latent carrier state in noninduced rodents have been made previously (16, 23), most studies have not focused on the assay as a means to detect the presence of the organism in largescale commercial rat colonies. Although our study used a single-copy gene as the target for PCR, it is likely that the sensitivity of detection would be increased by the use of a multicopy gene target, such as the mitochondrial large-subunit rRNA gene (13, 28). The case for diagnostic screening of rodent colonies for latent P. carinii carriers by PCR of lung samples taken from rats in noninduced sample groups is so compelling on the basis of time, availability of reagents, economy, and accuracy as to strongly endorse this technique as the method of choice with single or multiple genes as targets. The major drawback to a PCR-based assay is the requirement for postmortem tissue, as bronchoalveolar lavage and needle biopsies are difficult to perform on small laboratory animals with a high survival rate.
1445
Commercial vendors can ascertain the presence and even the type of P. carinii harbored by a given colony by using random surveys of animals sacrificed for this purpose, but the individual investigator would greatly benefit from a reliable, noninvasive assay, such as screening for anti-P. carinii antibodies. ACKNOWLEDGMENTS These studies were supported by grant RO1 AI29839-07 from the National Institutes of Health and the Medical Research Service of the Department of Veterans Affairs (M.T.C. and M.J.L.). REFERENCES 1. Armstrong, M. Y. K., and M. T. Cushion. 1994. Animal models, p. 181–222. In P. D. Walzer (ed.), Pneumocystis carinii pneumonia, 2nd ed. Marcel Dekker, Inc., New York, N.Y. 2. Bartlett, M. S., M. M. Durkin, M. A. Jay, S. F. Queener, and J. W. Smith. 1987. Sources of rats free of latent Pneumocystis carinii. J. Clin. Microbiol. 25:1794–1795. 3. Barton, E. G., and W. G. Campbell, Jr. 1969. Pneumocystis carinii in lungs of rats treated with cortisone acetate. Am. J. Pathol. 54:209–236. 4. Bauer, N. L., J. R. Paulsrud, M. S. Bartlett, J. W. Smith, and C. E. Wilde III. 1991. Immunologic comparisons of Pneumocystis carinii strains obtained from rats, ferrets and mice using convalescent sera from the same sources. J. Protozool. 38:1665–1685. 5. Chabner, B. A., C. J. Allegra, G. A. Curt, and P. Calabresi. 1996. Antineoplastic agents, p. 1233–1288. In J. G. Hardman, L. E. Limbird, P. B. Molinoff, R. W. Ruddon, and A. Gilman Goodman (ed.), Goodman and Gilman’s the pharmacological basis of therapeutics, 9th ed. McGraw-Hill, New York, N.Y. 6. Chandler, F. W., Jr., J. K. Frenkel, and W. G. Campbell. 1979. Animal model: Pneumocystis carinii pneumonia in the immunosuppressed rat. Am. J. Pathol. 95:571–574. 7. Cushion, M. T., and M. J. Linke. 1991. Factors influencing Pneumocystis infection in the immunocompromised rat. J. Protozool. 38:133S–135S. 8. Cushion, M. T., J. Zhang, M. Kaselis, D. Giuntoli, S. L. Stringer, and J. R. Stringer. 1993. Evidence for two genetic variants of Pneumocystis carinii coinfecting laboratory rats. J. Clin. Microbiol. 31:1217–1223. 9. Cushion, M. T., M. Kaselis, S. L. Stringer, and J. R. Stringer. 1993. Genetic stability and diversity of Pneumocystis carinii infecting rat colonies. Infect. Immun. 61:4801–4813. 10. Cushion, M. T. 1998. Genetic heterogeneity of rat-derived Pneumocystis. FEMS Immunol. Med. Microbiol. 22:51–58. 11. Diasio, R. B., and A. F. LoBuglio. 1996. Immunomodulators: immunosuppressive agents and immunostimulants, p. 1291–1308. In J. G. Hardman, L. E. Limbird, P. B. Molinoff, R. W. Ruddon, and A. Gilman Goodman (ed.), Goodman and Gilman’s the pharmocological basis of therapeutics, 9th ed. McGraw-Hill, New York, N.Y. 12. Edman, U., J. C. Edman, B. Lundgren, and D. V. Santi. 1989. Isolation and expression of the Pneumocystis carinii thymidylate synthetase gene. Proc. Natl. Acad. Sci. USA 86:6503–6507. 13. Feldman, S. H., S. P. Weisbroth, and S. H. Weisbroth. 1996. Detection of Pneumocystis carinii in rats by polymerase chain reaction: comparison of lung tissue and bronchiolar lavage specimens. Lab. Anim. Sci. 46:628–634. 14. Frenkel, J. K., J. T. Good, and J. A. Shultz. 1996. Latent Pneumocystis infection of rats, relapse and chemotherapy. Lab. Investig. 15:1559–1577. 15. Institute of Laboratory Animal Resources. 1991. Infectious disease of mice and rats, p. 64–69. NRC. National Academy Press. 16. Kitada, K., S. Oka, S. Kimura, K. Shimada, T. Serikawa, J. Yamada, H. Tsunoo, K. Egawa, and Y. Nakamura. 1991. Detection of Pneumocystis carinii sequences by polymerase chain reaction: animal models and clinical application to noninvasive specimens. J. Clin. Microbiol. 29:1985–1990. 17. Kovacs, J. A., J. L. Halpern, B. Lundgren, J. C. Swan, J. E. Parillo, and H. Masur. 1989. Monoclonal antibodies to Pneumocystis carinii: identification of specific antigens and characterization of antigenic differences between rat and human isolates. J. Infect. Dis. 159:60–70. 18. Linke, M. J., A. G. Smulian, J. R. Stringer, and P. D. Walzer. 1994. Characterization of multiple unique cDNAs encoding the major surface glycoprotein of rat-derived Pneumocystis carinii. Parasitol. Res. 80:478–486. 19. Linke, M. J., P. D. Walzer, and M. T. Cushion. 1989. Properties of the major rat and human Pneumocystis carinii antigens. Infect. Immun. 57:1547–1555. 20. Milder, J. E., P. D. Walzer, J. D. Coonrod, and M. E. Rutledge. 1980. Comparison of histological and immunological techniques for detection of Pneumocystis carinii in rat bronchial lavage fluid. J. Clin. Microbiol. 11: 409–417. 21. O’Leary, T. J., M. M. Tsai, C. F. Wright, and M. T. Cushion. 1995. Use of semiquantitative PCR to assess onset and treatment of Pneumocystis carinii infection in rat model. J. Clin. Microbiol. 33:718–724. 22. Ogino, K. 1978. Pneumocystis carinii. Experimental pulmonary infection in rats. Jpn. J. Parasitol. 27:77–89. 23. Sepkowitz, K., N. Schluger, T. Godwin, D. Armstrong, A. Cerami, and R.
1446
24. 25. 26. 27. 28.
WEISBROTH ET AL.
Bucala. 1993. DNA amplification in experimental pneumocystosis: characterization of serum Pneumocystis carinii DNA and potential P. carinii carrier states. J. Infect. Dis. 168:421–426. Stringer, J. R., and M. T. Cushion. 1998. The genome of Pneumocystis carinii. FEMS Immunol. Med. Microbiol. 22:15–26. Taconic Farms, Inc. 1998. Catalogue. F344 Fischer rats. Taconic Farms, Inc., Germantown, N.Y. Ueda, K., Y. Goto, S. Yamazaki, and K. Fugiwara. 1977. Chronic fatal pneumocystosis in nude mice. Jpn. J. Exp. Med. 47:475–482. Vasquez, J., A. G. Smulian, M. J. Linke, and M. T. Cushion. 1996. Antigenic differences associated with genetically distinct Pneumocystis carinii. Infect. Immun. 64:691–701. Wakefield, A. E. 1996. DNA sequences identical to Pneumocystis carinii f. sp. carinii and Pneumocystis carinii f. sp. hominis in samples of air spora. J. Clin. Microbiol. 34:1754–1759.
J. CLIN. MICROBIOL. 29. Wagner, M. 1985. Absence of Pneumocystis carinii in Lobund germfree and conventional rat colonies, p. 51–54. In W. B. Wostman (ed.), Germfree research: microflora control and its application to the biomedical sciences. Alan R. Liss, New York, N.Y. 30. Walzer, P. D., and M. E. Rutledge. 1980. Comparison of rat, mouse and human Pneumocystis carinii by immunofluorescence. J. Infect. Dis. 142:449. 31. Walzer, P. D., and M. E. Rutledge. 1981. Humoral immunity in experimental Pneumocystis carinii infection. 1. Serum and bronchial lavage fluid antibody responses in rats. J. Lab. Clin. Med. 97:820–833. 32. Walzer, P. D., R. Powell, K. Yoneda, M. E. Rutledge, and J. E. Milder. 1980. Growth characteristics and pathogenesis of experimental Pneumocystis carinii pneumonia. Infect. Immun. 27:928–937. 33. Weisbroth, S. H. 1995. Pneumocystis carinii: review of diagnostic issues in laboratory rodents. Lab. Anim. 24:36–40.
