Prevalence of small ruminant lentivirus and Mycobacterium avium subsp. paratuberculosis co-infection in Ontario dairy sheep and dairy goats

Abstract

Infection with small ruminant lentiviruses (SRLV) causes a variety of chronic inflammatory conditions that limit production. Mycobacterium avium subsp. paratuberculosis (MAP) is also a major production-limiting disease of sheep and goats, which causes severe inflammation of the small intestine. Previous studies have indicated that both SRLV and MAP are widespread in small ruminants in Ontario. This study estimated the prevalence of SRLV and MAP co-infection. Serum samples that were previously tested for MAP infection were re-tested for SRLV. The apparent prevalence of co-infection was low, with 3.4% [95% confidence interval (CI): 1.9 to 5.9] and 14.3% (95% CI: 11.6 to 17.5) of sheep and goats respectively, positive for both infections. However, co-infection is widespread with 36.8% (95% CI: 19.1 to 59.1) and 71.4% (95% CI: 52.8 to 84.9) of sheep and goat farms with 1 or more co-infected animals. A significant association was found between SRLV seropositivity and MAP fecal culture (P = 0.021), suggesting that co-infected goats may be more likely to shed MAP in their feces.

The maedi-visna virus (MVV) and caprine arthritis encephalitis virus (CAEV) are lentiviruses affecting sheep and goats, respectively (1). The MVV and CAEV can infect either species and an individual animal can be infected with 1 or more strains; therefore, these viruses are often referred to as the small ruminant lentivirus (SRLV) (1). The SRLV infects monocytes and macrophages, and causes chronic inflammation in tissues, such as the lungs, mammary glands, central nervous system, and joint synovia (2). Common clinical signs include progressive weight loss, hard udders with reduced milk production, chronic respiratory disease in sheep, and enlarged joints with lameness in goats (2). However, due to the long incubation period of SRLV, it can take months or years after infection for clinical signs of disease to emerge.

Transmission of SRLV primarily occurs when the newborn lamb or kid is exposed to shed viron particles and infected cells in colostrum and milk, or through infected respiratory secretions from the dam or other near-by adults (2). However, SRLV can also be transmitted directly to the fetus in utero or during the birth process (2). Horizontal transmission can also occur among adult sheep through direct contact with infected respiratory secretions, and sexual transmission may also be possible (2).

Johne’s disease is a production-limiting disease of ruminants caused by infection with Mycobacterium avium subsp. paratuberculosis (MAP) (3). In small ruminants, the infection causes chronic inflammation of the distal ileum leading to progressive weight loss with or without diarrhea (3). As with SRLV, Johne’s disease has a long incubation period from 1 to several years (3). However, MAP bacteria can be shed in the feces for many months prior to the onset of clinical signs making control difficult (3). The MAP bacteria infect macrophages located in Peyer’s patches beneath the intestinal epithelium of the distal ileum. The bacteria replicate inside these cells forming granulomatous lesions (4). This localized infection allows MAP to evade host immune responses while shedding bacteria into the feces and thus into the environment (4). As the infection progresses, monocytes and lymphocytes traffic to the infection site, lesions continue to grow, and, as the epithelial cell layer becomes increasingly damaged, clinical signs emerge (4). The primary mode of MAP transmission is fecal-oral from feces and the contaminated environment and skin, such as teats of the dam (5). The MAP can also be transmitted to newborn animals through bacterial shedding in the colostrum and milk, and transplacental infection may occur (5).

Previously, the apparent prevalence of SRLV infection in Ontario goats based on serology was 17% and 80.4% in meat and dairy goats, respectively (6), and 20.9% in Ontario sheep (7). Using fecal culture, the apparent prevalence of MAP infection and the individual animal level was estimated at 18.3% (95% CI: 12.0 to 24.6) and 7.6% (95% CI: 4.7 to 9.7) in dairy goats and dairy sheep, respectively, while herd-level apparent prevalence was 79.3% (95% CI: 60.3 to 90.2) and 57.1% (95% CI: 34.0 to 78.0) in dairy goat herds and dairy sheep flocks, respectively (8). Given the high prevalence of SRLV and the widespread nature of MAP infection, it is probable that herds and, likely, individual animals may be co-infected. Since both SRLV and MAP preferentially infect macrophages and monocytes, it is unclear how animals respond when faced with both pathogens. Additionally, both SRLV and MAP cause severe immune dysregulation as the disease progresses (9,10); therefore, understanding how these pathogens interact and affect host immunity is important for overall herd and flock health. The prevalence of SRLV and MAP co-infection, however, is currently unknown. A follow-up investigation using samples from a previous MAP study (8) was conducted to validate the prevalence of SRLV in Ontario and estimate the prevalence of co-infection in Ontario dairy sheep flocks and dairy goat herds.

Serum samples from 355 sheep and 558 goats that were previously collected to evaluate diagnostic test accuracy for MAP infection (8) were then evaluated for evidence of SRLV infection, thus allowing for the determination of the prevalence of co-infection with MAP and SRLV. These samples were collected between October 2010 and August 2011 from 19 dairy sheep farms and 28 dairy goat farms in Ontario. Dairy goat herds were selected randomly after stratification based on herd size, while the sheep flocks were selected conveniently since a list of Ontario dairy sheep producers was unavailable (8). From each herd/flock, blood and feces were collected from 20 randomly selected, healthy-appearing lactating females > 2 y of age. The MAP and SRLV disease status of all herds and flocks was unknown at the time of sampling. Fecal samples were evaluated for presence of MAP bacteria by culture using a detection system (BD BACTEC MGIT 960 Mycobacterial detection system; Becton Dickinson and Company, Franklin Lakes, New Jersey, USA) and BACTEC MGIT Para TB medium confirmed by acid-fast staining and polymerase chain reaction (PCR) (Tetracore, Rockville, Maryland, USA) (8). A sample was confirmed fecal culture positive if it fluoresced positive in the BACTEC system, had a positive acid-fast stain and positive PCR confirmatory test.

The serum samples were subsequently submitted to Animal Health Laboratory, University of Guelph, Guelph, Ontario to test for the presence of antibodies to SRLV. Sheep sera were tested using the Hyphen ELISA test (ELITEST MVV/CAEV; Hyphen Biomed, Neuville sur Oise, France), and goat sera was tested using an enzyme-linked immunosorbent assay (ELISA) kit (CAEV ELISA kit, IDEXX CHEKIT/MVV, IDEXX Laboratories, Westbrook, Maine, USA) according to the manufacturer’s instructions. The standard protocols for both tests were previously described by DeAndres et al (11) and Stonos et al (6), respectively.

For the purposes of this study, an animal was considered positive for MAP infection if positive by fecal culture. This test was chosen because it is more sensitive than serology (8). An animal was considered positive for SLRV infection if the ELISA result was positive. A farm was considered MAP or SRLV positive if ≥ 1 animal in the herd/flock tested positive for either. An animal is defined as co-infected if positive for both SLRV serology and MAP fecal culture.

The data were organized and stored using computer software (Microsoft Excel version 14.4.8 Redmond, Washington, USA). Sheep and goat data were analyzed separately due to species-specific differences in test performance (12). The SRLV prevalence data were analyzed using the computer software (GLIMMIX, SAS Version 9.2; SAS Institute, Cary, North Carolina, USA). This procedure used a generalized linear model with a logit-link function [log (p(infected)/ (1 – p(infected)] and included farm as a random effect with MAP serum results, MAP fecal culture results, and MAP polymerase chain reaction (PCR) results as fixed effects. This model allowed for the inclusion of SRLV infection as a binary trait. Further analysis with the mixed model involved the logistical regression analysis using computer software (SAS). This model used log transformed SRLV optical densities as indicators of antibody levels. Farm was included as a random effect, MAP serum data was included as a covariate, and MAP fecal culture, MAP PCR data and farm size data were included as classification variables. These models were used to assess associations in all the animals as a whole, as well as only those animals that tested positive for SRLV.

The overall apparent prevalence of SRLV, based on serology, in the 355 Ontario dairy sheep was 41.3% (95% CI: 36.3 to 46.5). Of the 19 farms sampled, 3 farms had a within flock prevalence of 100% and 3 farms had a prevalence of 0% (Figure 1). Flock-level prevalence was 78.9% (95% CI: 56.1 to 92.1) with 15/19 farms having at least 1 infected sheep. The prevalence of co-infection of SRLV and MAP in Ontario sheep was estimated at 3.4% (95% CI: 1.9 to 5.9) and 36.8% (95% CI: 19.1 to 59.1) of sampled farms had at least 1 animal test positive for both SRLV and MAP.

Figure 1.

Frequency distribution of small ruminant lentivirus (SRLV) and Mycobacterium avium subsp. paratuberculosis (MAP) fecal culture infection in Ontario dairy sheep flocks based on testing 20 adult females ≥ 2 y of age per flock (n = 19).

The seroprevalence of SRLV in the 558 sampled Ontario dairy goats was 80.1% (95% CI: 77.7 to 84.2). Of the 28 farms sampled, 3 farms had a prevalence of 0% and 11 farms had a within-herd prevalence of 100% (Figure 2). Additionally, 89.3% (95% CI: 71.9 to 97.1) or 25/28 of the farms sampled had at least 1 SRLV-infected goat. With respect to co-infection, 14.3% (95% CI: 11.6 to 17.5) of goats tested positive for SRLV and MAP, and 71.4% (95% CI: 52.8 to 84.9) or 20 of the 28 farms sampled had at least 1 co-infected animal. Additionally, 91.2% of MAP infected goats also tested positive for SRLV.

Figure 2.

Frequency distribution of small ruminant lentivirus (SRLV) and Mycobacterium avium subsp. paratuberculosis (MAP) fecal culture infection in Ontario dairy goat herds based on testing 20 adult females ≥ 2 y of age per flock (n = 28).

Analysis of the sheep SRLV data revealed no significant associations (P = 0.592) (Table I); however, analysis of the goat SRLV data found a significant association between SRLV infection and MAP fecal culture (P = 0.021; Table II).

Table I.

A 2 × 2 representation of SRLV and MAP infection in lactating dairy sheep > 2 years of age (n = 355) from 19 Ontario sheep herds (P = 0.592)

	SRLV positive	SRLV negative	Total
MAP negative	134	197	331
MAP positive	12	12	24
Total	146	209	355

Table II.

A 2 × 2 representation of SRLV and MAP infection in lactating dairy goats > 2 years of age (n = 558) from 28 Ontario goat herds (P = 0.0213)

	SRLV positive	SRLV negative	Total
MAP negative	354	102	456
MAP positive	93	9	102
Total	447	111	558

A cross-sectional study was done to assess the prevalence of SRLV infection in Ontario dairy sheep flocks and dairy goat herds, and to estimate the prevalence of SRLV and MAP co-infection in those same farms and animals. The seroprevalence of SRLV in Ontario dairy sheep was found to be 41.3% (95% CI: 36.3 to 46.5) and 78.9% (95% CI: 56.1 to 92.1) of the farms sampled had ≥ 1 infected animal suggesting that SRLV infection is common in Ontario dairy sheep flocks.

In Ontario dairy goats, the seroprevalence was 80.1% (95% CI: 77.7 to 84.2), and 89.3% (95% CI: 71.9 to 97.1) of study farms had ≥ 1 infected animal, which is in agreement with previous work that estimated a seroprevalence of SRLV in Ontario dairy goats to be 80.4% (6). This high prevalence of SRLV infection in both dairy sheep and dairy goats at the individual and farm level, suggests that SRLV may represent a serious problem to those industries, and emphasizes the need to implement preventative strategies into current farm management practices to prevent SRLV transmission.

This study also investigated SRLV and MAP co-infection in Ontario small ruminants. It is possible that the degree of immune dysregulation that occurs both during SRLV and MAP infections is exacerbated by co-infection. For example, HIV-1 and M. tuberculosis co-infected patients respond inappropriately to both pathogens, which can accelerate disease progression (13), and these altered immune responses may compromise the ability to detect both pathogens. Therefore, understanding how SLRV and MAP interact and how many animals are infected with both pathogens is important for assessing overall animal health. The apparent co-prevalence of SRLV and MAP was estimated at 3.4% (95% CI: 1.9 to 5.9) and 14.3% (95% CI: 11.6 to 17.5) in sheep and goats, respectively. Despite the high prevalence of both SRLV and widespread nature of MAP, the animal-level prevalence of co-infection was surprisingly low. However, the farm level prevalence was much higher with 36.8% (95% CI: 19.1 to 59.1) and 71.4% (95% CI: 52.8 to 84.9) of sheep and goat farms, respectively, having ≥ 1 co-infected animal.

Although the animal level co-prevalence is low, it may be that the testing protocols used have underestimated the level of infection particularly in sheep. Detecting SRLV infection in these animals likely wasn’t an issue. Test sensitivity for the SRLV ELISAs are 99.4% and 100% in sheep and goats according to the manufacturers and since all animals sampled were over 2 y of age, animals infected at birth would have seroconverted (14). However, test sensitivity for MAP infection regardless of the diagnostic test used, has often been reported as moderate to low (15). The MAP test evaluations used in this study were previously conducted (8) and identified MAP fecal culture to be the most sensitive test in both sheep and goats. In goats, fecal culture is 81.1% (95% CI: 65.8 to 93.0) sensitive, while in sheep it is only 49.5% (95% CI: 27.4 to 72.5). While culture still outperformed serological testing and fecal PCR (8), MAP infection may still be underestimated. The decontamination procedure used prior to culturing may have killed some bacteria and the fecal culture medium used does not grow the sheep strain of MAP very well (8,16). Additionally, a number of factors affect the sensitivity of the tests, including the stage of infection and the degree of fecal shedding (15). Therefore, if MAP infection is underestimated, co-infection is likely to also be underestimated. Due to the difficulty identifying MAP infected and co-infected animals this study was only able to report the apparent prevalence and was unable to estimate true prevalence in these populations.

This study failed to reveal any significant associations between the different MAP test results and SRLV infection in sheep. One possible reason for the lack of an association in sheep may be attributed to the ovine MAP test sensitivity as previously discussed; however, the small sample size in this study was also a limitation. Additionally, this study was a cross-sectional point prevalence study, so several animals may have been missed at the time of sampling due to intermittent shedding of MAP (17) or fluctuating SLRV antibody levels (14). Thus a cohort study may be more effective to track disease progression in co-infected animals over time.

In goats, a significant positive association between animals that tested positive for SRLV and animals positive for MAP by fecal culture was identified, which suggests that those animals that are SRLV positive were more likely to be MAP infected or shed more bacteria. This supports the hypothesis that those animals that are infected with one pathogen may not respond to further infections with an appropriate immune response. The animal would then shed more bacteria or fail to absorb the bacteria and, therefore, it passes through into the feces. However, this association may also indicate that co-infected animals experience enhanced MAP growth within the small intestine and therefore shed more. Similar results have been found with HIV-1 and M. tuberculosis co-infection in human macrophages, where an increase in M. tuberculosis replication was noted in co-infected cells (18). This may also suggest that infection with SRLV may increase susceptibility to other infections, as is the case with HIV-1 patients (13). However, it is currently unknown how SRLV infection affects MAP replication and disease progression.

Further research is required to accurately identify animals that are co-infected with MAP and SRLV. Future studies should increase the number of farms sampled and collect more than one fecal sample from each animal to help identify those that may not be shedding the bacteria during the initial sampling period. The possibility of accelerated disease progression should also be investigated by monitoring the time to SRLV seroconversion in herds and flocks that are positive for both pathogens. The high prevalence of SRLV, further exemplifies the need for more producer education and implementation of on-farm SRLV control measures. These control measures include, removing newborn kids and lambs at birth and providing replacement colostrum and milk, limiting animal movements between farms, providing separate housing for animals of different ages, and routine testing and culling of infected animals (19). These prevention measures are also effective for preventing the spread of MAP and other pathogens, and will thus help ensure a healthy flock or herd. Additionally, further studies should focus on understanding the immune responses to SRLV and MAP infection, and how the pathogens interact as a means of improving diagnostic testing and vaccine development.

This is the first study to estimate the prevalence of SRLV and MAP co-infection in Ontario dairy sheep flocks and goat herds. The apparent prevalence of 3.4% and 14.3% in sheep and goats, respectively, suggests that few animals are in fact co-infected, however, this is likely underestimated. The high flock and herd level prevalence for co-infection suggests that both MAP and SLRV are widespread in Ontario dairy sheep flocks and goat herds. This study further validates a high prevalence of SRLV in Ontario sheep and goats indicating the widespread nature of SRLV in Ontario. Therefore, the Ontario sheep and goat industries should introduce measures to educate producers and promote SRLV and MAP control programs.