TESTICULAR LESIONS AND ANTLER ABNORMALITIES IN COLORADO, USA MULE DEER (ODOCOILEUS HEMIONUS): A POSSIBLE ROLE FOR EPIZOOTIC HEMORRHAGIC DISEASE VIRUS

In deer and other Cervidae, antler abnormalities often result from a loss of or failure to form functional testicular tissue, resulting in low testosterone levels (Bubenik 1982). Testosterone plays a major role in driving normal antler growth cycles; mild increases in testosterone levels initiate antler growth (Bubenik 1982) along with other factors including insulin-like growth factor 1 (Suttie et al. 1985), but high testosterone is essential to stimulate antler mineralization/ossification, and subsequent velvet shedding (Wislocki et al. 1947). A marked decrease in testosterone is necessary to induce casting of antlers (Goss 1968). Low levels of testosterone therefore result in antler growth, but growth is incomplete or atypical, antlers do not completely harden and break easily, velvet is retained, and abnormal antlers are not shed annually (Wislocki et al. 1947; Goss 1968; Bubenik 1982). Aberrant growths can occur at broken points and continued growth occurs each year on the previous antler base. Short, aberrant points accumulate over several years, leading to the terming of “cactus bucks” (Clark 1953).

Antler abnormalities associated with testicular lesions have been described in white-tailed deer (Odocoileus virginianus) from Texas, with testicular atrophy (Taylor et al. 1964) and hypoplasia (Marburger et al. 1967); black-tailed deer (Odocoileus hemionus columbianus) from California, with testicular atrophy (DeMartini and Connolly 1975); and mule deer (Odocoileus hemionus) from Washington, with testicular atrophy (Tiller et al. 1997). In these cases, abnormal antler growth was attributed to noninflammatory testicular lesions of atrophy and hypoplasia, with proposed causes including toxins such as estrogenic substances and dietary deficiencies (Taylor et al. 1964; DeMartini and Connolly 1975; Tiller et al. 1997). However, inconsistent observations of chronic vascular lesions (DeMartini and Connolly 1975) and infarction (Tiller et al. 1997) in some populations also provided reason to consider a possible role for infectious agents including epizootic hemorrhagic disease and bluetongue viruses (Tiller et al. 1997). In 1971, Murphy and Clugston described a single mule deer buck from Colorado, US with bilateral testicular degeneration and polished antlers. The authors noted that this case was not consistent with lesions of testicular atrophy or hypoplasia, but rather the testes were affected by inflammation, necrosis, fibrosis, and mineralization. The authors suggested that these lesions were consistent with a “chronic infection in the testis of this animal” (Murphy and Clugston 1971).

Antler abnormalities attributed to testicular lesions have occasionally been observed and accepted at low levels in mule deer in Colorado for many years with minimal efforts to characterize associated testicular lesions. However, beginning in 1999, statewide management practices were initiated to increase deer populations and, specifically, to increase mature male mule deer numbers through limitation of hunting licenses (Bergman et al. 2011). In the area near Delta County, Colorado, deer hunting license limitations led to an approximate doubling of the estimated male mule deer population by 2006, producing a male∶female ratio of approximately 40∶100 (B.D. unpubl. data). Coinciding with this increase in male deer, we began to receive numerous reports of cactus bucks from hunters and outfitters. Affected males had retained velvet, brittle, or broken antlers, and in some cases short, knob-like proliferations along the antlers. Similar findings were also being reported in other mule deer herds across the state. We suspect this apparent increase in cactus buck prevalence is an artifact of management for mature male (trophy) deer. Antler abnormalities may have always been present in these herds, but only observed with enough frequency or severity to raise concern when the number of mature male deer was increased. Regardless, the apparent increase in cactus bucks led to an investigation into the cause of the abnormalities in these herds.

We describe testicular lesions associated with abnormal antler growth and assess whether lesions were more consistent with an inflammatory (chronic orchitis and epididymitis) or noninflammatory (atrophy or hypoplasia) process. We also screened cases to determine if lesions observed in the testes might be associated with infection by epizootic hemorrhagic disease virus (EHDV) or bluetongue virus (BTV). Exposure to these viruses was associated with antler abnormalities in a population of mule deer in Washington (Tiller et al. 1997), and we have seen unexpectedly high (100%) prevalence of EHDV antibody in mule deer cactus bucks (n = 25) in Colorado (L.L.W. unpubl. data). Prevalence of antibody to EHDV varies across the state, but averages approximately 13% (n = 717; seven herds) statewide. High EHDV antibody prevalence for any herd in the state is considered approximately 40% (L.L.W. unpubl. data). Of the 25 cactus bucks previously examined for EHDV antibody, 13 were also screened for antibodies to BTV, Brucella abortus, and bovine viral diarrhea virus (BVDV). Antibody prevalences were 7/13 (54%), for BTV, 6/13 (46%) for BVDV, and 0/13 (0%) for B. abortus. None of the 13 blood samples was positive for BVDV antigen by enzyme-linked immunosorbent assay. Of these results, BTV antibody prevalence was considered high, whereas BVDV and B. abortus antibody prevalences were similar to findings in mule deer across the state (L.L.W. unpubl. data).

Epizootic hemorrhagic disease is well described in white-tailed deer, in which the disease is often rapidly fatal (Fletch and Karstad 1971; Tsai and Karstad 1973; Howerth et al. 2001). Less is known about the pathogenesis of EHD in other North American wild ungulates such as mule deer, black-tailed deer, elk (Cervus elaphus nelsoni), bighorn sheep (Ovis canadensis), pronghorn (Antilocapra americana), moose (Alces alces), and bison (Bison bison). In many of these species, EHDV infection and seroconversion is common, but clinical disease is rare or absent (Hoff and Trainer 1973; Hoff and Hoff 1976; Stauber et al. 1977; Work et al. 1992; Dubay et al. 2006; Nol et al. 2010). Epizootic hemorrhagic disease is of concern to both wildlife and livestock managers in the US (Gibbs et al. 2008; Maclachlan and Guthrie 2010; Savini et al. 2011), and there is a need to better understand this disease in various wild species.

To investigate the occurrence of cactus bucks near Delta County, Colorado, we euthanized (by gunshot) and sampled six male mule deer with abnormal antlers (cases 1–6) and two male mule deer with normal antlers from this area in November 2010. Blood was collected, and serum was submitted for EHDV and BTV serology (Texas A&M Veterinary Medical Diagnostic Laboratory, College Station, Texas, USA). Tissues, including testis and epididymis, among others, were collected in 10% neutral buffered formalin for histopathology.

Given results from initial cases, we opportunistically sampled testis and epididymis from male mule deer in affected herds, but with minimal antler changes. In addition to the Delta County area, we identified four other regions of Colorado with recently increased observations of deer with retained velvet and abnormal antler growth. From these regions, 16 male mule deer from 11 counties were opportunistically sampled for testicular and splenic tissues. Deer with minimal antler lesions (retained velvet only) or essentially normal antlers were targeted for sampling. Velvet was considered to be retained for any case with velvet-covered antlers after 27 October on the basis of previous studies (Anderson and Medin 1971; Anderson and Wallmo 1984). We collected samples from late August through early December 2012, taking advantage of hunting seasons and partially targeting months when EHDV and BTV infections are likely to occur.

One entire testis from each carcass was collected in 10% neutral buffered formalin. Formalin-fixed testicular tissues were sectioned for histology at three levels: head of the epididymis, body of the epididymis, and tail of the epididymis. At each level, sections included both testicular and epididymal tissues. All fixed tissues were routinely processed for histologic examination via paraffin embedding, sectioning at 5 µm, and staining with hematoxylin and eosin. Each case was examined for testicular and epididymal lesions and a complete list of lesions was compiled. Each case was “scored” on the basis of the number of lesions. This scoring system was intended to objectively categorize cases such that higher scores indicated cases that were more representative of the disease process seen in the population. This system was also intended to help gain perspective regarding nonspecific lesions.

One testis from each carcass was sampled for fresh tissue at the level of the body of the epididymis. At this level, an approximately 3- × 3-mm section each of epididymis and testis was collected and placed in a single cryovial containing 1.5 mL of RNAlater (Qiagen Inc., Valencia, California, USA). A similar-sized sample of fresh spleen was also collected from each carcass and preserved in RNAlater. Each carcass was sampled with clean instruments, typically on separate days and at separate locations, with no potential for cross-contamination of tissues.

Splenic and testicular tissues were examined for for EHDV and BTV by RNA extraction and reverse-transcriptase real-time PCR (qRT-PCR). Tissues were homogenized with phosphate-buffered saline (1∶20 w/v) using a benchtop homogenizer and clarified by centrifugation (600 × G for 5 min). Eighty microliters of the clarified tissue suspension was transferred to a 96-well plate and RNA was extracted using a commercially available isolation kit (AM1836, Life Technologies, Grand Island, New York, USA) by a Kingfisher 96 magnetic particle processor (Fisher Scientific, Pittsburgh, Pennsylvania, USA). Extracted RNA was eluted with RNase-free water to a final volume of 80 µL.

Primers and probes used for BTV and EHDV detection are listed in Table 1. Primers and probes for EHDV were specific for detection of all eight EHDV serotypes (Clavijo et al. 2010). The PCR amplifications for both BTV and EHDV qRT-PCR assays were carried out using commercial one-step reagents (Path ID multiplex master mix, Life Technologies). We used the following PCR amplification profile: 48 C for 10 min, 95 C for 10 min, followed by 40 cycles of 95 C for 15 s, and 55 C for 45 s. Auto threshold cycle (Ct) analysis was used for determination of positive and negative results, with any Ct value <40 considered positive.

For samples that were EHDV positive by qRT-PCR, we used a RT-PCR assay to specify the EHDV serotype. Primers were selected from the VP2 gene for EHDV-1, 2, and 6, as described by Sun et al. (2014). Primer sequences are listed in Table 1. Briefly, a one-step multiplex RT-PCR assay was carried out using a OneStep RT-PCR kit (#21012; Qiagen). A no-template control and three positive-amplification controls containing RNA from EHDV-1, 2, and 6, respectively, were included. Reverse-transcriptase PCR amplification consisted of initiation at 50 C for 50 min and 95 C for 10 min, followed by 45 cycles at 95 C for 20 s, 57 C for 30 s, and 72 C for 45 s, with final extension of 72 C for 5 min. The PCR amplification products were visualized by electrophoresis on 1.5% agarose gels using a fluorescent nucleic acid gel staining method (GelRed™, Phenix Research Products, Candler, North Carolina, USA).

To assess a possible association between PCR results in the testes/epididymis and extent of testicular lesions, we performed a Mood's median test (Mood 1954) post hoc. Mood's median test compares the proportions of values in groups A (PCR-positive testes) and B (PCR-negative testes) that are above the common median (lesion score in the testis). Upton's chi-square test (Upton 1982; D'Agostino et al. 1990) is used for the comparison. Our null hypothesis was that deer with EHDV-positive and EHDV-negative testes were drawn from populations with the same median lesion score. The same test was used to assess BTV PCR results.

Of the six male mule deer sampled near Delta County, Colorado in November 2010, five were considered cactus bucks with retained velvet and short, knob-like points at the base or beams of the antlers (Fig. 1A and Table 2). Age was estimated by tooth wear as 2–4 yr in three of the deer and 7+ yr in the other two. The younger animals had antlers with retained velvet as the predominant feature, whereas the older deer had more knobby proliferations along the antlers. Serum was collected from four of the five cactus bucks. All four had detectable antibody to EHDV-2, and three of the four were antibody positive for BTV (Table 2). One of the six (case 3) had antlers that were abnormally shaped, but were well mineralized with appropriately shed velvet. For this case, a skull fracture with trauma to the pedicels was identified, explaining abnormal antler growth. This deer was negative for EHDV and BTV antibodies, and had no significant testicular lesions grossly or histologically (Table 2).

The five male mule deer with antler features of cactus bucks had end-stage lesions of testicular degeneration characterized by loss of seminiferous tubules, replacement fibrosis, interstitial lymphoplasmacytic inflammation, and mineralization (Fig. 2A). Macrophages containing intracytoplasmic hemosiderin were frequently observed, and testicular blood vessels were mildly cuffed by lymphocytes and plasma cells that minimally invaded the vessel wall. In three of five cases, minimal lymphocytic inflammation was present in the epididymis, predominantly centered on blood vessels. The ducts of the epididymis occasionally contained intraluminal sloughed epithelial cells, but sperm were absent (Table 2). The epithelial cells lining the ducts of the epididymis were within normal limits. The pituitary gland was examined for all six cases, and no histologic lesions were seen.

Of the 16 male mule deer opportunistically sampled for evidence of acute testicular lesions, 13 had normal antlers, two had retained velvet (cases 7 and 8) (Fig. 1B), and one had abnormal antler growth but normal antler mineralization and shed velvet (case 14, Table 3). For case 14, the left antler lacked normal branching and was curved downward toward the face of the deer (Fig. 1C). The testes of this animal were essentially normal, but the left forelimb had a chronic fracture with callus formation. Leg fractures have previously been associated with antler abnormalities, and forelimb fractures typically cause changes to the ipsilateral antler (Marburger et al. 1972).

Age was estimated by tooth wear for 10 of the 16 cases, including the two deer with retained velvet. Age estimates ranged from 2 to 5 yr, with no apparent association between age and minimal antler lesions of retained velvet only. Causes of death included: hit by car (n = 7), hunter killed (n = 3), euthanasia due to suspected trauma (n = 2), and found dead (n = 4). Complete necropsies were conducted on two of the four animals found dead. Cause of death was attributed to a brain abscess in one (case 11), and liver abscesses/bacteremia in the other (case 20). The postmortem interval before sampling for all carcasses was less than 24 h.

The two male mule deer with retained velvet had chronic testicular lesions including seminiferous tubular necrosis and loss, replacement fibrosis, inflammation, and mineralization, similar to cases 1–6 (Tables 2, 3). However, more acute lesions were noted in testes from males with normal antlers. Histologically, vascular changes within the testes and epididymis included: necrotizing vasculitis within the epididymis (n = 2), hemorrhage within the testis (n = 6) and epididymis (n = 1), and edema (n = 2). Tubular and interstitial changes included scattered seminiferous tubular necrosis with associated mild multifocal lymphoplasmacytic orchitis (n = 4), lymphocytic epididymitis (n = 10), and hypospermia/aspermia (n = 8) (Table 3 and Fig. 2B–F). Subtle lesions of seminiferous tubular degeneration were likely present in several cases, although the degree of autolysis varied between samples and therefore an interpretation of these more subtle lesions was not attempted.

Histologic scores of testicular/epididymal lesions ranged from one through six, with a median of two. Higher scores tended to correlate with more severe or chronic cases as expected. PCR results were positive for EHDV in seven of 16 cases for testis/epididymis and six of 16 cases for spleen. A Mood's median test examining PCR results for the testis/epididymis versus lesion score for the testis/epididymis demonstrated a significant difference between the EHDV-negative and EHDV-positive groups. None of the nine EHDV-negative testes scored above the median, whereas five of the seven (71%) EHDV-positive testes scored above the median (P = 0.003). All cases that were PCR positive in the spleen were also PCR positive in the testis. One case was PCR positive in the testis, but PCR negative in the spleen (Table 3). Serotyping of positive samples by PCR demonstrated the presence of EHDV-2 in all cases except case 9, for which results were consistent with EHDV-6 in testis and spleen.

PCR results for BTV were positive for three of 15 cases for testis/epididymis and five of 15 cases for spleen. A Mood's median test demonstrated no significant difference between the BTV-negative and BTV-positive groups on the basis of lesion score for testis/epididymis. Four of the 12 (33%) BTV-negative testes had scores above the median, whereas one of the three (33%) BTV-positive testes had a score above the median (P = 1.0). Three cases that were PCR positive in the spleen were also PCR positive in the testis. Two additional cases were PCR positive in the spleen but PCR negative in the testis (Table 3).

We described histologic lesions of the testes associated with antler abnormalities in male mule deer in Colorado including velvet retention and cactus buck-type proliferations. On the basis of histologic findings, these antler lesions are caused by severe testicular degeneration, likely initiated by vascular damage and inflammation. Breakdown of the blood-testis barrier may account for continued progression of these lesions after the initial insult. Although infarction was not specifically observed, severe end-stage lesions predominated by mineralization suggest that infarction may also be involved in the pathogenesis of these lesions. The inflammatory and vascular lesions were consistent with an infectious or inflammatory disease process versus noninflammatory processes that would be expected with toxin exposure or congenital hypoplasia.

On the basis of our histologic findings and serologic evidence of exposure to EHDV and potentially BTV in cactus bucks, we specifically examined testicular and epididymal tissues for these viruses by PCR. Comparison of PCR results with lesions in the testis/epididymis demonstrated a significant difference in lesion score on the basis of whether or not testes were PCR positive or PCR negative for EHDV. No significant difference was found on the basis of PCR results for BTV. Comparing results between the testis and spleen may suggest targeted infection of the testis by EHDV, as one deer (case 7) demonstrated EHDV RNA in the testis but not the spleen. This is an unexpected occurrence if positive testis results merely represent systemic infection; detection of systemic virus is expected to be more common in the highly vascular spleen (as was the observed pattern for BTV PCR results).

In general, the testicular and epididymal lesions that were associated with EHDV-positive results included orchitis, vasculitis, edema, and end-stage lesions of tubular loss, replacement fibrosis, and mineralization (Table 2). Our histologic finding of testicular hemorrhage (a lesion expected to be associated with EHDV infection) did not appear to be specifically associated with positive PCR results for EHDV, nor could testicular hemorrhage be attributed to BTV (Table 3). One explanation for the nonspecific testicular hemorrhage may be trauma. All of the EHDV-negative cases with testicular hemorrhage were hit by cars, whereas neither of the EHDV-positive cases with testicular hemorrhage was hit by cars. Of all cases with testicular hemorrhage, only one also had epididymal hemorrhage (case 9; Table 2), and this case was PCR positive for EHDV. Although we observed only this one occurrence of epididymal hemorrhage, that occurrence was specific to EHDV-positive cases in this study.

The results of this study suggest an association between EHDV infection, vascular and inflammatory lesions of the testis and epididymis, and cactus buck antler abnormalities seen in Colorado mule deer. We suspect that EHDV may have tropism for the testis and epididymis where infection causes acute lesions of vasculitis, edema, likely hemorrhage, and inflammation. Progression to end-stage lesions of seminiferous tubular loss, replacement fibrosis, and mineralization may be due to breakdown of the blood-testis barrier, infarction, or maintenance of viral infection in the testis. Antler abnormalities are not observed until the testes are nearly destroyed, which we suspect does not occur until at least the next season of antler growth after infection. After affected male deer are essentially castrated, low testosterone levels account for incomplete antler hardening, retained velvet, and failure to shed antlers annually. Subsequent breaks, proliferations, and annual growth on existing velvet antlers lead to a cactus buck appearance. Antler proliferations accumulate over time, and it may take several years to accumulate easily recognizable abnormalities.

It is likely that the apparent increase in cactus bucks in Colorado was caused by a recent change in management practices involving decreased hunting of male mule deer. This practice led not only to increased numbers of males in the population, but also to increased numbers of males living long enough to accumulate multiple years of abnormal antler growth on antlers that failed to shed annually. An increased occurrence of cactus bucks has also been observed in specific herds in other US western states associated with management for mature males, including New Mexico (Ryan Darr, pers. comm.), Arizona (Amber Munig, pers. comm.), and Oregon (Rod Klus, pers. comm.). In Wyoming, managers have noted increased numbers of deer with retained velvet and broken antlers after an outbreak of EHD (Daryl Lutz, pers. comm.).

Our results suggest a possible role for EHDV in the pathogenesis of cactus bucks in Colorado. However, limitations of the study are multiple. Although we targeted EHDV and BTV on the basis of serologic evidence of exposure to these viruses, we did not rule out other possible causes for inflammation in the testes, nor did we evaluate histologic lesions for the presence of EHDV using immunohistochemistry or in situ hybridization, due to the lack of commercially-available reagents. Our findings demonstrate the need for postmortem examinations of the testes in these cases, as the identification of inflammatory versus noninflammatory lesions is critical to propose differentials for targeted investigation.

If infection by EHDV is implicated as the cause of testicular degeneration in mule deer, there may be important management implications concerning what is effectively castration of male mule deer after infection by this virus. Diagnostic reagents that are currently unavailable for detection of EHDV, including antibodies for immunohistochemistry and in situ hybridization, would greatly assist in future studies. Additional studies of EHDV infection in other wild ungulates for which infection is nonfatal may also be important.

We thank Colorado Parks and Wildlife technicians Ivy LeVan and Tracy Dubovos for sample processing, and we thank the many Colorado Parks and Wildlife biologists and district wildlife managers who assisted with sample collection including Kyle Banks, Bob Holder, Melanie Kaknes, Brad Weinmeister, Brad Banulis, Joe Padia, Quentin Springer, Allen Vitt, and Jeromy Huntington. We thank Mike Miller for assistance with statistical analysis and comments on the manuscript. We thank the Western Association of Fish and Wildlife Agencies mule deer working group for insights on the manuscript. This work was funded, in part, by the Colorado Parks and Wildlife Auction and Raffle fund.