In spite of a long and expensive CAEV eradication campaign, seropositive goats continue to be detected in Swiss flocks. We have previously shown that the serological tools used to implement this campaign are quite effective at detecting animals infected with classical CAEV strains (SRLV B subtypes), but perform poorly when applied to the detection of goats infected with SRLV A subtypes
. The combination of a long eradication campaign, based on the elimination of seropositive animals and their descendants, and the use of serological tools biased toward an efficient detection of SRLV B subtypes most likely favored the spread of SRLV A subtypes in the goat population. In sharp contrast to the situation preceding the CAEV eradication campaign, no clinical cases of SRLV induced pathology such as carpitis, encephalitis or mastitis have been recorded in Switzerland in the last 15 years. This is strong, albeit indirect, epidemiological evidence that for goats the SRLV A subtypes are attenuated. Based on this evidence the eradication campaign is now exclusively focused on goats infected with SRLV B subtypes and SRLV A infected goats are no longer eliminated from the flocks. As a direct consequence of this decision, we are expecting a constant increase in the number of SRLV A infected goats the coming years.
It is therefore crucial to study the SRLV A subtypes currently circulating in Switzerland and to actively search for potential subclinical manifestations of disease induced by these viruses in naturally infected goats. To this purpose we selected a flock of goats infected with the prevalent SRLV A4 subtype from a geographic region epidemiologically not linked to a previously described flock
. SRLV infections persist in the infected animals and, with the exception of encephalitis in young goats, the pathological manifestations appears months to years after infection
. We selected older animals in order to increase the chances of detecting subclinical manifestations of disease. The absence of clinical signs of arthritis in the entire flock and in these 5 chronically infected adult goats, as demonstrated by their inconspicuous carpal/metacarpal ratio, confirms previous reports on flocks infected with similar viruses.
The phylogenetic characterization of the viruses infecting these 5 goats was based on a region of the env gene encompassing the highly variable SU4 and SU5 domains
[16, 17]. This permitted us to confirm the SRLV A4 classification of these isolates, anticipated by their SU5 ELISA serological reaction (data not shown)
. The four isolates were closely related to previous Swiss A4 viruses but formed a separate cluster supported by strong bootstrap values (Figure
3). The fine phylogenetic resolution permitted by the variability of this env region revealed a coherent grouping of sequences obtained from different anatomical compartments of the same animals, such as PBMC and mammary gland tissue (Figure
3). Taken together, these observations convincingly exclude potential laboratory contaminations in this work.
Virus was readily isolated from 4 out of 5 goats, which points to a potentially high proviral load in the peripheral blood of these animals. The isolated viruses showed marked differences in their in vitro behavior. The isolates from goats #2 and #5 were manifestly cytopathogenic for macrophages, as already observed with a previous A4 isolate obtained from a goat. By contrast, the isolates from goat #3 and #4 were not cytopathogenic for macrophages behaving like SRLV A4 isolated from sheep in a previous work (Figure
. Cytopathogenicity appeared to be associated with the replication capacity of these viruses in macrophages, as shown by the rapid increase of RT activity in the supernatant of primary macrophage cultures of goats #2 and #5, measured at day 3 p.iso. (Figure
1). The behavior of these 4 primary isolates on GSM cells was similar to that observed with previous A4 isolates, inducing mild cytopathic effects on these cells. By contrast, these goat isolates did not induce cytopathic effects on LSM cells, as previously observed with an A4 goat isolate inducing large syncytia in LSM
. Similarly, SRLV B strains were shown to be severely restricted in their replication in ovine fibroblasts
. The lack of cytopathic effects correlated with poor replication on GSM and LSM as demonstrated by testing the supernatants of these cultures by PERT assay, often negative, and with a sensitive RT-PCR assay, generating positive results but with very weak bands (data not shown). Similar observations were reported by Singh and colleagues working with macrophage tropic SRLV isolated from sheep
. These authors demonstrated the capacity of infected macrophages to mediate the infection of non-permissive cells such as fibroblasts. In this respect, the preferential homing of infected monocytes to particular tissues, such as the mammary gland, may explain the peculiar pattern of proviral loads observed in our goats.
The final goal of this work was to examine the classical SRLV target organs for their proviral load and the potential presence of histopathological lesions. We have previously shown that there is an excellent correlation between the amounts of genomic RNA and proviral DNA in the peripheral blood and target organs of SRLV infected animals and that the proviral load is a more constant and reliable marker to assess the viral load of infected animals
Proviral load measurements were always performed in triplicates and, for the target organs, on three independent samples obtained from different locations.
With the exception of goat #2, the proviral load in the peripheral blood increased over time showing significant differences between the early time points and the latest sampling times in 3 out of 5 goats. This is a rather unexpected observation, suggesting a progressive loss of control by the immune system that, however, did not result in obvious pathological manifestations. The goats were kept under ideal conditions, excluding stress as a potential trigger for the increasing viral load. Moreover, these animals were adapted to their new environment for a period of 4 months prior to proviral load quantification in PBMC samples. In contrast to previous observations in an SRLV A4 infected, mixed flock of goats and sheep, the proviral loads in the PBMC of these goats were surprisingly high and comparable to those measured in SRLV A4 infected sheep, SRLV B infected goats or HIV infected patients
The concurrence of a high proviral loads in the absence of overt pathology in the infected animals is reminiscent of the situation observed in SIV infected natural hosts, such as sooty mongabeys and African green monkeys
. In both SIV and SRLV the absence of pathological manifestations is probably related to the long co-evolution between these lentiviruses and their respective hosts. In spite of their obvious similarities, such as the sharing of a marked tropism for macrophages, however, primate and small-ruminant lentiviruses are quite distinct. SRLV do not infect lymphocytes and do not induce overt immunosuppression. Therefore, we think that these two lentiviruses adopted different strategies to coexist with their hosts. The so called apathogenic SRLV restricted their expression to organs directly involved in their efficient transmission, such as the mammary gland, while keeping low viral loads in other tissues such as the carpal joints and the choroid plexus. By contrast, apathogenic primate lentiviruses are present at high titers in the entire body but their hosts adapted to the presence of these viruses by modulating the expression of receptors and co-receptors on the surface of target cells such as central memory T cells, which are indispensable to preserving immunocompetence.
The results obtained in target organs and, in particular, in the mammary gland showed small standard deviations between samples taken in the same location and quite large variations between samples taken in different locations of the same organs. This is not surprising and confirms previous observations indicative of a clear compartmentalization of the virus in the target organs
We did not observe an obvious association between the in vitro behavior of these viruses and their proviral load. Indeed, goats #2 and #5, from which the most cytopathogenic viruses were isolated, ranked at the opposite sides of the proviral load scale, with goat #5 showing the overall highest proviral load and #2 the second lowest. Independently of the lactation stage and with the exception of, again, goat #2 the highest proviral load was detected in the mammary gland confirming previous results obtained with goats experimentally infected with an SRLV B molecular clone
. These proviral loads were up to 3 orders of magnitude higher than those measured in other tissues, confirming the tropism of these viruses for the mammary gland. In view of the importance of colostrum and milk for the transmission of SRLV this is not surprising. Indeed, the mammary gland was shown to be the only target organ that efficiently permits transmission of highly attenuated SRLV E1 in goat familiar lineages
[7, 27]. Additionally, we had previously shown that SRLV A and B co-infected mothers tend to transmit the SRLV A subtype more efficiently to their offspring than the B subtype, suggesting that this particular subtypes are particularly efficient in lactogenic transmission
. This is in contrast to the situation observed in naturally infected SIV hosts that, in spite of the high viral load and in contrast to HIV infected mothers, do not efficiently transmit SIV to their offspring
The proviral load in the synovial membrane tissue was moderately low and comparable to that observed in PBMC. The lowest proviral load was detected in DNA extracted from choroid plexus cells, suggesting a poor neurotropism for these viruses
. Similar proviral loads, albeit at a slightly higher level, were found in the lung and lung macrophages suggesting that these viruses are unlikely to induce lung pathologies.
Gross pathological lesions were not observed and the histopathological findings were also unremarkable. With the exception of the mammary gland, no clear evidence of SRLV induced lesions was found.
In the light of the very high proviral loads detected in the mammary gland, the histopathological examination of this tissue was of particular interest. SRLV infections typically cause diffuse interstitial lympho-plasmacytic infiltrations and formation of lymphoid follicles in the mammary gland
. Such histopathological changes were present in all our goats and, interestingly, the different histopathological grades reflected proviral loads in the mammary tissue. The highest number of follicles was found in goat #5, which indeed showed a very high proviral load in the mammary gland tissue (Figure
6b). By contrast, in the mammary tissue of goat #2, which showed a low proviral load, only few lymphoid follicles were observed (Figure
6a). These results, however, should be interpreted with caution, indeed, mild histopathological lesions were observed in goats infected with the highly attenuated SRLV E subtype and also in control age matched animals (
[7, 27] and Sergio Rosati personal communication).