Systemic mycoplasmosis with dystocia and abortion in a North American bison ( Bison bison ) herd

Abstract

The current study describes a fatal Mycoplasma bovis infection in a North American bison (Bison bison) cow and her aborted fetus in a herd suffering unusual mortality associated with dystocia and abortion. Postmortem evaluation of the subject case found severe caseonecrotic bronchopneumonia, chronic fibrinous pleuritis and pulmonary sequestra, foci of caseous necrosis in the kidneys, and necrotizing endometritis and placentitis. Histologic findings in the maternal tissues include endometrial and placental necrotizing vasculitis and changes in the lung similar to those previously described for M. bovis–associated pneumonia in feedlot bison. Gross and microscopic lesions were not observed in the fetus. Maternal lung, uterus, kidney, and placenta as well as fetal lung and kidney were positive for M. bovis by polymerase chain reaction (PCR) as were the Mycoplasma-like colonies cultured from these tissues. The presence of M. bovis in maternal and fetal tissues was further demonstrated using nucleic acid extracts in a pan-Mycoplasma SYBR Green PCR assay targeting the 16S-23S ribosomal RNA spacer region with post-PCR dissociation curve analysis and sequencing of the resulting amplicons. Immunohistochemistry (IHC) testing on maternal lung and uterine caruncle was strongly positive for M. bovis antigen. A variety of methods, including culture, PCR, and IHC, failed to identify other bacterial or viral pathogens in any of the tissues evaluated. These data are the first to implicate M. bovis as a cause of placentitis and abortion in bison.

Introduction

In cattle, pneumonia, mastitis, and arthritis are the most frequently occurring manifestations of Mycoplasma bovis infection, but the bacterium has also been implicated in bovine otitis media, conjunctivitis, and reproductive disease.^8,13 Although anecdotal evidence for the emergence of M. bovis as a primary pathogen in North American bison (Bison bison) is rapidly accumulating, only 3 reports describing clinical manifestations, which include pneumonia, respiratory distress, polyarthritis, and necrotic pharyngitis, have been published to date.^4,5,11 While M. bovis is known to cause chronic endometritis, infertility, and abortion in cattle,^2,8,10 the authors found no published accounts of similar conditions in bison. The current study describes the isolation and identification of M. bovis and associated pathology in a bison cow and her aborted fetus from a herd with reproductive failure and pneumonic mycoplasmosis.

During the winter feeding period, the affected herd of 80 cows (68 pregnant) was maintained on an 32-hectare paddock close to the home buildings, with an additional 120 yearlings in an adjacent feeding pen. In March 2012, 2 breeding bulls were purchased from a bison sale and placed in a feeding pen that allowed fence-line contact with both cows and yearling feeders. The major presenting complaint in the cow herd was obstetrical in nature and, despite the postmortem finding of pneumonia, there was little mention by the owner of coughing or respiratory distress in the affected animals. In bison, dystocia from any cause is relatively uncommon, but when it does occur, there is little opportunity for intervention because of the nature of the species and problems with close handling. As a result, neither calf nor dam is likely to survive.

During a 3.5-month interval, from April 1 to July 15, 2012, 15 out of 68 pregnant cows in the herd (22%) died in the peri- or postparturient period. All were suspected of having M. bovis–related disease, and most had a body condition score less than 2.0 at the time of death. All 15 calves from these cows presented as stillbirth deliveries or were never found. Four of the affected cows were necropsied. Only 1 (the subject case) had an intrauterine calf, and all affected cows had caseonecrotic pleuropneumonia and endometrial necrosis. Of the 53 calves born to surviving pregnant cows, 4 died of unknown causes in the summer months and 49 remained, as of September 2012. None of the yearlings or recently purchased bulls displayed clinical signs of disease.

Among the 4 field necropsies performed by the referring veterinarian, the most frequently reported gross lesions included a remarkable loss of body condition and severe caseonecrotic bronchopneumonia with chronic fibrinous pleuritis and extensive zones of coagulation necrosis (pulmonary sequestra). In addition to these common findings, the subject case had multifocal, randomly scattered foci of caseous necrosis in the kidneys as well as multiple renal infarcts. There were multifocal to coalescing areas of caseous necrosis involving both the endometrium and the chorioallantoic surface of the placenta, primarily at the level of the cotyledon (Fig. 1). These lesions were interpreted as necrotizing endometritis and placentitis, respectively. Fetal lesions were not detected grossly.

Figure 1.

Uterus and fetal membranes; bison (Bison bison) cow. Multifocal to coalescing foci of caseous necrosis (arrow) involving the endometrium and chorioallantois are indicative of necrotizing endometritis and placentitis caused by Mycoplasma bovis infection.

Samples of maternal lung, kidney, and uterine caruncle as well as fetal lung, liver, and kidney were submitted for histology to the Diagnostic Services Unit of the University of Calgary Faculty of Veterinary Medicine (Calgary, Alberta, Canada). Tissue samples were fixed in 10% neutral buffered formalin for 24 hr, processed routinely for histological examination, and stained with hematoxylin and eosin. Sections of caruncle were also evaluated with Gram stain, Warthin–Faulkner stain, and Grocott modification of Gomori methenamine silver (GMS) method.

Microscopically, the maternal lung lesions include chronic caseonecrotic bronchopneumonia with bronchiectasis, chronic fibrinous pleuritis, necrotizing vasculitis, and large areas of coagulation necrosis, characteristic of pulmonary sequestra. These changes are similar to those previously described for M. bovis–associated pneumonia in feedlot cattle and bison.^4,7,11 Foci of caseous necrosis were noted in the maternal lung and renal cortex, characterized by a brightly eosinophilic coagulum encircled by a rim of degenerate neutrophils, macrophages, lymphocytes, plasma cells, and fibroblasts. Dystrophic mineralization was noted within the caseous foci. A single uterine caruncle was available for histology; intercotyledonary placenta was not sampled. The architecture of the placentome was almost completely effaced by a zone of brightly eosinophilic necrosis, bordered by an intense band of degenerate neutrophils with fewer numbers of macrophages, lymphocytes, and plasma cells. The underlying placental stroma was edematous with increased numbers of leukocytes. Affected vessels occasionally contained thrombi. Severe necrotizing vasculitis, characterized by fibrinoid change within the tunica media, degenerate neutrophils, and nuclear debris were also noted (Fig. 2A). Gram, Warthin–Faulkner, and GMS stains failed to identify any infectious organisms in the sections of placentome examined.

Figure 2.

Section of a uterine caruncle; bison (Bison bison) cow: hematoxylin and eosin (A) or immunohistochemical staining for Mycoplasma bovis (B). The normal tissue architecture is largely replaced by a zone of caseous necrosis bordered by a rim of degenerate neutrophils. Vasculitis is a feature and is characterized by fibrinoid change and degenerate neutrophils within the tunica media (arrow). Immunohistochemical staining demonstrates M. bovis antigen free within the areas of caseous necrosis and the band of degenerate neutrophils. The arrow highlights the same vessel indicated in panel A. Note the absence of positive staining for M. bovis in vessels demonstrating vasculitis.

Immunohistochemistry for M. bovis (caruncle and maternal lung) and Bovine herpesvirus 1 (BHV-1; caruncle and fetal lung, liver, and kidney) was carried out with paraffin-embedded tissue samples by Prairie Diagnostic Services (Saskatoon, Saskatchewan, Canada), using previously described methods ⁹ adapted for an automated slide stainer. ^a After pretreatment and enzymatic epitope retrieval, ^b sections were incubated with either monoclonal antibody clone 5A10 ^c to detect M. bovis antigen ¹ or a cocktail of monoclonal antibodies, clones 3F11 and 1H6, ^d to detect BHV-1. Binding of the primary antibody was detected using a streptavidin–biotin detection system. ^e Sections of pneumonic bovine lung known to be infected with M. bovis and fetal lung and liver known to be infected with BHV-1 were used as positive controls, respectively. Abundant M. bovis antigen was detected in the maternal lung with a staining pattern typical to that previously described for M. bovis–associated pneumonia in cattle. ⁷ M. bovis antigen was most plentiful at the periphery of foci of caseous necrosis and within the bronchiolar exudates, but was also observed in moderate amounts in alveolar macrophages and neutrophils. In addition, mycoplasmal antigen was plentiful within the alveolar exudate and alveolar leukocytes in areas of sequestration. Large amounts of M. bovis antigen were noted in the uterine caruncle (Fig. 2B). Positive staining was most abundant within the area of caseous necrosis and within the zone of necrotic neutrophils that bordered it. Mycoplasmal antigen was not detected within any vessels of the uterine caruncle. Immunohistochemistry for BHV-1 was negative in the caruncle and fetal tissues. The histologic and immunohistochemical findings are consistent with M. bovis–associated pneumonia in the cow with hematogenous dissemination to the kidneys and placentomes, resulting in caseonecrotic nephritis, necrotizing placentitis, and abortion.

To confirm the suspected etiology of infection, tissues from the cow (lung, uterus, and kidney) and fetus (lung, kidney, and placenta) were sent to the U.S. Department of Agriculture/National Animal Disease Center (USDA/NADC; Ames, Iowa) and Vaccine and Infectious Disease Organization–International Vaccine Centre (VIDO-InterVac; University of Saskatchewan, Saskatoon, Saskatchewan, Canada) for culture and analysis by PCR. At both institutions, all cultures were incubated at 37°C in an atmosphere of 5% CO₂. Samples provided to the USDA/NADC were homogenized at a 1:10 dilution (weight/volume) in selective PPLO broth ^f supplemented with 10 g/l of yeast extract, ^g 20% horse serum ^h and, as a selective inhibitor, 0.05% thallium acetate. ⁱ Aliquots were used to inoculate selective PPLO broth. After incubation for 24–48 hr, 2 µl of each culture was removed for testing with an M. bovis–specific PCR. ³ Aliquots of each culture were also propagated on selective PPLO agar plates and examined daily for growth of Mycoplasma-like colonies, which were similarly tested by PCR. ³ Broth cultures were incubated for up to 2 days and plates for up to 4 days before being discarded as negative. At VIDO-InterVac, tissue samples were homogenized in modified Hayflick broth ⁶ and cultured in the same for 2 days. Aliquots of the broth cultures were used to inoculate solid Hayflick medium and plates were examined daily for 5 days to detect the growth of Mycoplasma-like colonies. Suspect colonies were tested by PCR using a method different from that employed at the USDA/NADC.^15,16 Tissue homogenates were also used to inoculate 5% sheep blood agar plates, which were examined daily for 2 days. All tissues were culture and PCR positive for M. bovis and, in most cases, identical results were obtained at both laboratories (Table 1). Discrepant results for cow and fetal kidney may be related to the different samples provided to each laboratory and/or different limits of detection for each of the culture methods used. A role for bacterial pathogens other than M. bovis seems unlikely because additional pathogenic bacteria were not identified from blood agar plates inoculated with tissue homogenates.

Table 1.

Mycoplasma culture and polymerase chain reaction (PCR) results from maternal and fetal tissues.

Tissue	Culture*	Mycoplasma bovis–specific PCR†	Pan-Mycoplasma SYBR PCR‡
Fetal lung	+	+	+
Fetal kidney	+ 4	+	–
Fetal abomasal fluid	+	+	+
Lung I	+	+	+
Lung II	+	+	+
Lung III	+	+	+
Placenta I	+	+	+
Placenta II	+	+	+
Uterus, caruncle	+	+	+
Kidney	+§	+§	+

*

Presence of Mycoplasma-like colonies on solid medium inoculated with broth culture of the tissue homogenate indicated.

†

Carried out on cultures inoculated with a homogenate of the tissue indicated.

‡

Carried out on nucleic acid extracted from a homogenate of the tissue indicated; melting curve analysis indicated M. bovis as the sole source of amplicons from all positive samples.

§

Result obtained at Vaccine and Infectious Disease Organization–International Vaccine Centre (University of Saskatchewan, Saskatoon, Saskatchewan, Canada) was positive; result obtained at the U.S. Department of Agriculture/National Animal Disease Center (Ames, Iowa) was negative.

To more thoroughly investigate the etiology of the current case, nucleic acid extracts were prepared from tissues and abomasal fluid sent to the USDA/NADC and used for additional testing by PCR. One hundred milligrams of each solid tissue, minced with sterile scissors, or 250 μl of abomasal fluid was homogenized ^j and further processed using a kit, ^k according to the manufacturer’s protocols. A fresh broth culture of M. bovis and extraction buffer with no added material were also processed separately as positive and negative controls, respectively. Extracts were forwarded to the Center for Advanced Host Defenses Immunobiotics and Translational Comparative Medicine at Iowa State University (Ames, Iowa) for pan-Mycoplasma real-time SYBR PCR and precision melt analysis (Trujillo JD, Justice-Allen A, Morley T: 2009, SYBR Green real-time PCR detection and differentiation assay for Mycoplasma species in biological samples. In: Proceedings of the 52nd Annual AAVLD/USAHA Meeting, p. 97, October 7–14, San Diego, CA), which detects and differentiates among Mycoplasma species on the basis of dissociation curve analysis of SYBR Green–containing amplicons. The assay employs proprietary primers designed to amplify a genus-specific, 150–230-bp segment of the spacer region between the 16S and 23S genes of Mycoplasma ribosomal RNA (rRNA) loci. Species-specific sequence heterogeneity in the resulting amplicons allows identification of the particular species present based on amplicon melting temperature (T_m). Polymerase chain reaction was performed on a real-time system ^l using 12-μl reactions prepared with a commercially available master mix. ^m The thermocycling conditions used were 95°C for 2 min, 40 cycles of 55°C for 2 min, and 72°C for 1 min, followed by a continuous melt of 95°C for 15 sec, 60°C for 1 min, and 95°C for 15 sec. Positive controls on each PCR plate included purified DNA from a reference strain of M. bovis (ATCC no. 25025) ⁿ and from a field strain of Mycoplasma bovoculi as well as the nucleic acid extraction positive control described above. The PCR master mix with no added template and the nucleic acid extraction control prepared using only extraction buffer were included as negative controls. Mycoplasma species present were identified by analysis of melt curve data and determination of amplicon T_ms.

With the exception of fetal kidney, all samples were strongly positive for Mycoplasma on the basis of SYBR Green fluorescence, speciated as M. bovis by melting curve analysis (Tm = 72.9–73.0; Table 1). No additional Mycoplasma species were detected. To confirm the identification of M. bovis, amplicons generated by pan-Mycoplasma SYBR PCR were sequenced at the USDA/NADC on a genetic analyzer ^o using standard methods. Final consensus sequences derived from 1 forward and 1 reverse read were used as queries in nucleotide BLAST searches of the National Center for Biotechnology Information nonredundant nucleotide database (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Highest scoring matches were 100% identical to the 16S-23S rRNA spacer region of M. bovis. Pan-Mycoplasma SYBR PCR results for M. bovis are concordant with the prior PCR data obtained at the USDA/NADC, confirming the negative status of the fetal kidney sample assessed there. These results, as noted above, are not necessarily inconsistent with the positive result from a different sample of the same tissue analyzed at VIDO-InterVac.

Nucleic acid extracts from fetal lung and cow lung were also tested with a TaqMan reverse transcription PCR for Bovine viral diarrhea virus and Border disease virus ¹² and by conventional PCR for Ureaplasma species. ¹⁷ Extracts from fetal lung, abomasal fluid, placenta, and uterus were tested by a TaqMan real-time PCR for Tritrichomonas foetus. ¹⁴ Negative results were obtained with all the assays for all samples evaluated.

These results document the first report of M. bovis–associated abortion in bison. No evidence was found to support any alternative, nor any contributing, infectious etiology. Others have noted the potential for M. bovis infecting bison to disseminate from the lung to a variety of anatomic locations, including the uterus. ¹¹ In the case reported herein, hematogenous spread of the bacterium from the lung similarly appears to underlie the progression of disease, although an ascending infection through the lower urogenital tract resulting from an environment heavily contaminated with M. bovis is possible. This novel systemic syndrome highlights the importance of prevention and control of M. bovis infection in bison, particularly in breeding herds.