Longitudinal study on seroreactivity of goats exposed to colostrum and milk of small ruminant lentivirus–infected dams

Small ruminant lentivirus (SRLV), classified in the Lentivirus genus and Retroviridae family, causes two diseases of small ruminants – caprine arthritisencephalitis (CAE) in goats and maedi-visna disease (MV) in sheep. Both are chronic and progressive, and have a considerable impact on animal productivity and welfare (34). Pathological processes comprise predominantly arthritis in CAE and interstitial pneumonia in MV, occasionally indurative mastitis and very rarely leukoencephalomyelitis (43). The clinical form of CAE develops in only approximately one third of infected goats and usually several months to years after infection (64). Historically, the aetiological agents of CAE and MV were considered related but distinct viral species referred to as caprine arthritis-encephalitis virus (CAEV) and maedi-visna virus (MVV). However, in the late 1990s, marked genetic overlapping between SRLV isolates from goats and sheep was observed (65), and over the first two decades of the 21^st century the taxonomic classification was modified. At present, SRLV is classified into at least four genotypes (A, B, C, and E, with D being in question) and the genotypes are further divided into sub-genotypes (12). Most genotypes and sub-genotypes have been shown to infect both goats and sheep (22, 40, 61). Genotypes A and B are widespread all over the world, while genotypes C and E are geographically restricted to Norway and Italy, respectively (22, 53). As genotypes A and B comprise most strains formerly classified as MVV and CAEV, they are customarily referred to as MVV-like and CAEV-like genotypes, respectively.

The major routes of SRLV transmission are the ingestion of colostrum and milk from infected dams (vertical postnatal transmission) and long-term direct contact between animals (horizontal transmission) (9, 50). The infection may also spread horizontally via contaminated milking cups (3). Intrauterine transmission is generally regarded as possible but sporadic (55). However, recent studies have yielded positive PCR results in a considerable proportion of lambs and kids taken from their mothers immediately after birth, strongly indicating the occurrence of intrauterine transmission (6, 21, 25). Traditionally, ingestion of colostrum and milk has been considered the most efficient mode of SRLV transmission in both goats and sheep, with most kids and lambs seroconverting after 4–12 weeks, which results in an overlap between periods of passive and active humoral immunity (1, 55, 56). On the other hand, more recent studies carried out in sheep infected with genotype A imply that the non-maternal route may play the principal role in SRLV transmission (10) and lactogenic transmission may be much less effective in natural conditions (5, 6), perhaps due to an ability of lambs that have acquired passive colostral immunity to clear the lactogenic infection (27). Whether a similar phenomenon may be observed in goats remains unknown. Therefore, we carried out a longitudinal study investigating the serological status of goats from the moment of natural exposure to colostrum and milk of dams contaminated with SRLV genotype A (MVV-like) through the age of 24 months.

Seroconversion was evident in 13 out of the 31 goats (42%; CI 95%: 26%–59%) during the study; in age breakdown they were 6 out of 20 goats (30%) born in 2014 and 7 out of 11 goats (64%) born in 2015 (Fig. 1). This difference was not significant (P = 0.128). Four of them tested seropositive only in one ELISA – three in the sp-iELISA and one in the SU-cELISA. The age at which the goats seroconverted ranged from 3 to 22 months with a median of 5 months (IQR 4–7 months). In 11 goats seroconversion was early and in the other 2 goats (no. 13 and no. 20) it was late (occurring in the 22^nd and 19^th months, respectively). Two of the goats that seroconverted early became seronegative again after 2 (goat no. 9) or 5 (goat no. 18) months. The former goat reverted to seropositive status after a further 13 months, while the latter remained seronegative until the end of the study (Fig. 1). These two goats with intermittent seroreactivity and one goat with late seroconversion (no. 20) tested positive only in the sp-iELISA (Fig. 2).

Fig. 1

Fig. 2

Only 9 of 31 goats (29%; CI 95%: 16%–47%) seroconverted in the first year of life and remained seropositive until the end of the study (i.e. were early and stable seroreactors). Eight of them seroconverted demonstrably in both ELISAs and one (goat no. 30) only in the SU-cELISA (Fig. 3). Early and stable seroconversion occurred in only 2 out of 20 kids (10%) born in 2014 but in all kids born in 2015 that seroconverted (7 out of 11; 64%) (P = 0.003). The age at which they seroconverted ranged from 3 to 10 months with a median of 5 months (IQR 4–6 months) and did not differ significantly between the sp-iELISA and SU-cELISA (P = 0.263).

Fig. 3

Eighteen goats remained seronegative for the entire study. In eight of them a single isolated seropositive result was obtained: in the sp-iELISA in two goats (nos. 12 and 26) and in the SU-cELISA in six goats (nos. 10, 16, 17, 19, 21, 23). In the sp-iELISA, the S/P% in seronegative goats ranged from 0% to 96% with a median of 3.6% (IQR 2.0%–7.4%). Compared to the S/P% at the age of 4 months, it was significantly higher at the ages of 6 months (P = 0.004) and 7 months (P = 0.029), and significantly lower at the age of 20 months (P = 0.017). In the SU-cELISA, the PI in seronegative goats ranged from 0% to 40.3% with a median of 13.7% (IQR 9.8%–20.3%) (Fig. 4). Compared to the PI at the age of 4 months, it was significantly higher at the ages of 11 months (P = 0.016), 12 months (P = 0.003), and 13 months (P <0.001), and then from the age of 20 months until the end of the study (P = 0.001) (Fig. 5).

Fig. 4

Fig. 5

None of the 31 goats showed any clinical signs of arthritis (joint swelling or lameness) during the observation period. The C/MC ranged from 1.5 to 1.7 with the median of 1.6.

Neither the S/P% of the sp-iELISA nor the PI of the SU-cELISA when the goats were one week old differed significantly between the stable seroreactors and the rest of the goats (P = 0.966 and P = 0.481, respectively), which disqualifies these variables as potential predictors of stable seroconversion.

Our study showed that seroconversion was an uncommon and delayed event in goats exposed to oral infection with heterologous SRLV genotype A (sub-genotype A17) via ingestion of colostrum and milk from infected dams. Only approximately one third of goats seroconverted and it took them 3 to 10 months from the moment of first colostrum ingestion to mount an antibody response with levels detectable in commercial ELISAs. These observations are at odds with many older studies on SRLV-infected goats which showed that ingestion of SRLV-contaminated colostrum and milk is an effective mode of SRLV transmission in goats and that most of them seroconvert three weeks to five months after oral infection with CAEV-like genotype B SRLV (1, 9, 17, 18, 56).

There are several possible explanations for this phenomenon. First, the colostrum and milk of dams enrolled in this study could have been a poor source of infective SRLV. It cannot be directly excluded, as we neither confirmed the presence of infective SRLV in colostrum or milk ingested by the kids nor made any attempt to detect SRLV in the kids. This is undoubtedly an important limitation of our study. On the other hand, such a scenario is very unlikely as all the dams enrolled in the study had been shown to be infected with SRLV genotype A (47, 48, 49) and seropositive for at least a year preceding the start of the study. Both the colostrum and milk of does infected with SRLV were shown to be potent and equally effective sources of infection (2, 18, 19, 36). The kids in our study certainly ingested colostrum from their infected dams and were freely consuming milk from infected does for three weeks in total. Such a time period has been shown to be more than enough to ensure lactogenic infection with SRLV. Therefore, we think that infection of the kids did occur.

The second possible explanation is that seroconversion may have in fact occurred in all kids, and the diagnostic sensitivities of the two ELISAs used were too low to detect it. This scenario is virtually impossible, as both ELISAs have been shown to be highly sensitive (42, 46, 60). Moreover, goats which did not seroconvert in our study had consistently low quantitative results in the ELISAs. Even lowering the cut-off value of the sp-iELISA to for example 20%, which, according to Nowicka et al. (46) ensures >95% diagnostic sensitivity, would not considerably change the interpretation of the results – goat no. 30 positive in the SU-cELISA would become positive also in the sp-iELISA for most of the time and goats no. 7 and 16 would be classified as intermittent seroreactors; however, no additional goats would be classified as stable seroreactors.

The third reason to consider for the low seroconversion rate and extended time in the present study is that seroconversion or lactogenic infection may have not occurred in some goats because they were exposed to the heterologous (MVV-like, A) SRLV genotype. Genotypes of SRLV have been shown to vary in terms of their potential for lactogenic transmission (51) as well as antigenic properties (11, 22, 45). As a result, the diagnostic performance of serological assays varies depending on the degree of sameness of the genotype responsible for the infection and the genotype from which the antigens used in the assay were derived (15, 33, 42, 52, 58, 60). Although very few studies have investigated the differences in antibody response between goats infected with genotype A and those infected with genotype B, the humoral response after infection with the homologous SRLV genotype has been shown to be stronger than that which followed the heterologous genotype infection in both goats and sheep (33, 41). Moreover, genotype A generally appears to stimulate slower development of the humoral response in both species (41). Notwithstanding the possible weaker lactogenic transmissibility and antigenicity of genotype A, the less pronounced humoral response associated with genotypic heterology, and the less rapid initiation of the same by genotype A, the goats were followed for two years and during this time seroconversion should have occurred, even if it was markedly delayed (54). Although the three goats that seroconverted at the age of >12 months (one of them, goat no. 9, after a temporary seroconversion) might have represented animals with a delayed humoral immune response following colostral or lactogenic infection, they were more likely to have been horizontally infected through sustained direct contact with their seropositive companions.

Interestingly, significantly more goats born in 2015 seroconverted than goats born a year before. It is possible that transferring these goats to the pen already inhabited by adult males resulted in a stress reaction which promoted faster infection and seroconversion. Obviously, it is impossible to definitively state whether seroconversion in those 7 of the 11 goats placed in the pen already inhabited by goats born in the previous year was the consequence of colostrum and milk ingestion or horizontal transmission. The latter cannot be excluded, as several studies have shown that horizontal transmission plays a very important role, possibly even more important than vertical transmission, and little benefit results from preventing lactogenic transmission unless strict segregation from infected individuals can be ensured (5, 8, 57). Nevertheless, the relatively short time in which most of the seroconversions of goats born in 2015 took place (between the ages of 3 and 6 months in six out of the seven goats that seroconverted) implies that they resulted from the point-source exposure corresponding to colostrum and milk ingestion in the first days of life (59).

The fourth possible explanation of the low seroconversion rate in our study is that the infection did occur, but was cleared by the immune system. Such a phenomenon has been described in sheep (27): 21 lambs born to SRLV-infected ewes were positive for proviral DNA in PCR at 8 weeks of age, but by week 24 proviral DNA became undetectable. Moreover, their serum antibody titres, high a week after birth, had fallen drastically by nine months of age. It was proposed that colostral antibodies could have activated natural killer cells through antibody-dependent cellular cytotoxicity (ADCC). This cytotoxicity and the neutralising activity of anti-lentiviral antibodies have been described in some studies in cattle and humans. Bovine colostral immunoglobulins of class G have been shown to have strong ADCC and neutralising activity against human immunodeficiency virus (HIV) type 1 (31, 32). High serum titres of ADCC-mediating anti-HIV antibodies are also negatively correlated with plasma viremia (3). However, no evidence of such activity of colostral antibodies exists in goats. On the contrary, Ellis et al. (18) showed that four out of six kids exposed to SRLV-contaminated colostrum became infected and this proportion was the same as in the group of kids deprived of colostral antibodies and exposed to SRLV-contaminated milk. Therefore, our results are the first such observation in goats, which should be treated with the highest caution and verified in further studies involving molecular and virological diagnostic methods.

We must firmly stress that the lack of molecular testing precludes drawing the conclusion that most of the goats enrolled in our study remained uninfected despite extensive exposure to the natural lactogenic source of SRLV. We may only state that most goats remained seronegative in two different serological assays of high diagnostic accuracy for two years after exposure, which is strong evidence that they did not mount a humoral immune response at all or that it was markedly delayed. Whether it was because they remained uninfected, cleared the infection, or were infected but remained seronegative, and in the latter scenario, whether they were about to seroconvert later or not, remains unknown. Several studies based on molecular testing have shown that both goats and sheep may be infected with SRLV without detectable seroreactivity (7, 16, 24, 25, 37, 54, 62). In one study, as many as 6 out of 12 goats with presumed intrauterine SRLV infection were positive in a nested-PCR and negative in an SU-cELISA and agar gel immunodiffusion test (25). Usually, the proportion of molecular test–positive and seronegative individuals does not exceed 10–20%. Therefore, we think that it is very unlikely that all goats that remained seronegative in our study were in fact infected at any moment of the study. However, even if one quarter of them were truly infected, it still means that the rate of infection did not exceed 50%, which is lower than expected and indicates that the natural lactogenic route of transmission of SRLV genotype A in goats may be less effective than the same route transmission of genotype B reported in early studies. Several studies in sheep have shown that the respiratory route of transmission appears to be much more efficient than the lactogenic route in this species (8, 10, 63). Recent studies have also shown that goats seroconvert 2–8 weeks after intratracheal inoculation with SRLV (41, 52, 53), which is a considerably shorter period than observed in older studies which arranged natural lactogenic transmission (1, 9, 17, 18, 56). Obviously, an experimental intratracheal inoculation only partially reflects the mode of SRLV transmission via the respiratory route in field conditions. The pathogenesis of SRLV infection in sheep, in which the lungs are the main target organ, naturally promotes the respiratory route of transmission in this species (38, 39). Nevertheless, some studies indicate that subclinical interstitial pneumonia also develops in SRLV-infected goats (20, 44). These observations give rise to the suspicion that the respiratory route may play a more important role than the lactogenic route in SRLV transmission in goats as it does in sheep.

Concluding, seroconversion appears to occur in less than 50% of goats exposed to oral infection with heterologous SRLV genotype A via ingestion of colostrum and milk from infected dams and may be delayed by 3 to 10 months. The natural lactogenic route of transmission of SRLV genotype A in goats appears to be less effective than the transmission by this route of genotype B observed in investigations carried out previously.