Initial lentivirus-host interactions within lymph nodes: a study of maedi-visna virus infection in sheep

Evaluation of Serological Methods and a New Real-Time Nested PCR for Small Ruminant Lentiviruses

Small ruminant lentiviruses (SRLVs), i.e., CAEV and MVV, cause insidious infections with life-long persistence and a slowly progressive disease, impairing both animal welfare and productivity in affected herds. The complex diagnosis of SRLVs currently combines serological methods including whole-virus and peptide-based ELISAs and Immunoblot. To improve the current diagnostic protocol, we analyzed 290 sera of animals originating from different European countries in parallel with three commercial screening ELISAs, Immunoblot as a confirmatory assay and five SU5 peptide ELISAs for genotype differentiation. A newly developed nested real-time PCR was carried out for the detection and genotype differentiation of the virus. Using a heat-map display of the combined results, the drawbacks of the current techniques were graphically visualized and quantified. The immunoblot and the SU5-ELISAs exhibited either unsatisfactory sensitivity or insufficient reliability in the differentiation of the causative viral genotype, respectively. The new truth standard was the concordance of the results of two out of three screening ELISAs and the PCR results for serologically false negative samples along with genotype differentiation. Whole-virus antigen-based ELISA showed the highest sensitivity (92.2%) and specificity (98.9%) among the screening tests, whereas PCR exhibited a sensitivity of 75%.

Characterization of minimal lesions related to the presence of visna/maedi virus in the mammary gland and milk of dairy sheep

Background

In order to characterize the complete range of lesions, especially minimal, affecting mammary gland and viral antigen distribution and target cells using immunohistochemistry in naturally Visna/maedi (VM) 84 infected sheep were studied, forty-four from flocks with clinical cases (A) and 35 randomly sampled from two abattoirs (B) together with five negative controls (C). An immunocytochemistry technique was developed and further milk samples (n = 39) were used to study viral excretion, carrier cells and the role of milk and colostrum in the transmission of the disease.

Results

All sheep from group C and three sheep from group B were negative to VM in tissue sections by histopathology, immunohistochemistry and PCR, and also in serum using ELISA. Several degrees of CD3 + lymphocytic interstitial mastitis were observed in groups A and B: minimal (+) n = 26 sheep; moderate (++), n = 32 and severe (+++), n = 12. No differences in lesion distribution were observed between groups A and B. Viral presence was confirmed by immunohistochemistry using two different antibodies and/or PCR in every tissue with lesions while serology was negative in six sheep with lesions. Two milk samples taken from milk tanks from two flocks from group A and fourteen milk samples from 29 infected sheep from group B were positive to VM (most of them from animals with moderate and severe lesions). Positivity was only found in macrophages, even in focal and minimal lesions, while no positivity was observed in epithelial or any other cells in either tissue and milk samples.

Conclusions

This new observation of the minimal lesions described in this work increased the prevalence of VM lesions in mammary gland up to 90.9% and VM should be considered as a differential diagnosis when minimal interstitial lesions are detected. A high prevalence of VM was observed in intensive milk-producing sheep, ELISA serology did not detect as positivity all infected animals, while histology, IHC or PCR showed higher sensitivity. The cytological technique developed was very useful in milk-cell studies using hematoxylin and eosin and immunocytochemistry. Viral detection in milk samples (16/39) confirms a potential but limited role of milk/colostrum in viral transmission.

Electronic supplementary material

The online version of this article (10.1186/s12917-019-1855-3) contains supplementary material, which is available to authorized users.

Small ruminant lentivirus infection influences expression of acute phase proteins and cathelicidin genes in milk somatic cells and peripheral blood leukocytes of dairy goats

The aim of the study was to analyze acute phase protein and cathelicidin gene responses to small ruminant lentivirus (SRLV) infection in goats. In uninfected goats, we found higher Cp and lower Fbγ mRNA levels in blood leucocytes (BL) than in milk somatic cells (MSC), as well as lower SAA, Hp, and CRP and higher Cp and AGP concentrations in blood serum than in milk. In SRLV-infected goats, we found higher Fbγ and MAP28 and lower Cp expression in MSC than in BL, and higher SAA, Hp, Fb, and MAP28 and lower AGP concentrations in milk than in blood serum. Higher SAA and Hp expressions in BL and Hp expression in MSC were found in SRLV-infected goats. In SRLV-infected goats, we observed a higher concentration of SAA in blood serum, while in milk, lower SAA, Cp, and MAP28 and higher MAP34 concentrations were observed. The expression profiles of the studied genes differed between BL/serum and MSC/milk. The elevated SAA concentration in blood serum was accompanied by a decreased concentration of SAA and Cp in the milk of infected goats. No differences in the expression of the other studied genes may mean that the SRLV has the ability to evade the immune system, continuing to replicate. The elevated concentration of SAA in blood serum may promote viral multiplication. This higher concentration of SAA in blood serum and simultaneous reduced concentration of SAA and Cp in milk may be additive indicators of this infection.

Small ruminant lentiviruses: immunopathogenesis of visna-maedi and caprine arthritis and encephalitis virus

The small ruminant lentiviruses include the prototype for the genus, visna-maedi virus (VMV) as well as caprine arthritis encephalitis virus (CAEV). Infection of sheep or goats with these viruses causes slow, progressive, inflammatory pathology in many tissues, but the most common clinical signs result from pathology in the lung, mammary gland, central nervous system and joints. This review examines replication, immunity to and pathogenesis of these viruses and highlights major differences from and similarities to some of the other lentiviruses.

Development and comparison of strain specific gag and pol real-time PCR assays for the detection of Visna/maedi virus

The aim of this study was the development of gag and pol dual labelled probe real-time PCR and RT PCR assays to quantify the proviral load and the transcripts of the British Visna/maedi virus EV1 strain. Primers and probes were chosen based on the consensus sequences of gag and pol clones representative of EV1 genetic variants. Both PCRs had a detection limit of 3 copies of target gene, with a linearity over 6 orders of magnitude. The performances of the two PCRs in vivo were evaluated and compared on a panel of DNAs extracted from blood of sheep infected experimentally with EV1. The pol assay detected in most cases lower numbers of viral molecules than gag assay, yielding some false negative results. The gag real-time RT PCR had a detection limit of 100 RNA molecules with a linearity over 5 orders of magnitude. This did not result in a lower performance of the RT PCR compared to the PCR in cells permissive for virus replication, which contain higher numbers of viral transcripts than proviral genomes. The real-time assays developed in this study, particularly the gag assay, provide a sensitive tool which can be used to quantify the viral load in experimental infections.

Mapping and characterization of visna/maedi virus cytotoxic T-lymphocyte epitopes

CD8(+) cytotoxic T-lymphocyte (CTL) responses have been shown to be important in the control of human and simian immunodeficiency virus infections. Infection of sheep with visna/maedi virus (VISNA), a related lentivirus, induces specific CD8(+) CTL in vivo, but the specific viral proteins recognized are not known. To determine which VISNA antigens were recognized by sheep CTL, we used recombinant vaccinia viruses expressing the different genes of VISNA: in six sheep (Finnish LandracexDorset crosses, Friesland and Lleyn breeds) all VISNA proteins were recognized except TAT. Two sheep, shown to share major histocompatibility complex (MHC) class I alleles, recognized POL and were used to map the epitope. The pol gene is 3267 bp long encoding 1088 aa. By using recombinant vaccinia viruses a central portion (nt 1609-2176, aa 537-725) was found to contain the CTL epitope and this was mapped with synthetic peptides to a 25 aa region (aa 612-636). When smaller peptides were used, a cluster of epitopes was detected: at least three epitopes were present, at positions 612-623: DSRYAFEFMIRN; 620-631: MIRNWDEEVIKN; and 625-635: EEVIKNPIQAR. A DNA-prime-modified vaccinia virus Ankara (MVA)-boost strategy was employed to immunize four sheep shown to share MHC class I allele(s) with the sheep above. Specific CTL activity developed in all the immunized sheep within 3 weeks of the final MVA boost although half the sheep showed evidence of specific reactivity after the DNA-prime immunizations. This is the first report, to our knowledge, of induction of CTL by a DNA-prime-boost method in VISNA infection.

Serum containing ovine IgG2 antibody specific for maedi visna virus envelope glycoprotein mediates antibody dependent cellular cytotoxicity

Antibody-dependent cell-mediated cytotoxicity (ADCC) specific for maedi visna virus (MVV) has never been described. The IgG antibody response to MVV is restricted to an IgG1 response whilst MVV specific IgG2 is never seen in persistently infected sheep. To determine whether the isotypic restriction of the antibody response is responsible for the lack of ADCC, an ADCC assay was developed using polyclonal serum raised to recombinant MVV ENV protein. Sheep immunised with a recombinant GST:SUenv fusion protein in complete Freund's adjuvant produced an antibody response which contained IgG1 and IgG2 antibodies. The activity of this serum in an ADCC assay was compared to serum from persistently infected sheep. Serum from immunised sheep mediated ADCC reactions whilst no activity was ever seen in persistently infected sheep serum. IgG2 may therefore be the possible effector isotype for ADCC reactions against MVV. Failure of the IgG2 dependent ADCC system in vivo may contribute to the persistence of MVV-infected macrophages in vivo.

Viral load, organ distribution, histopathological lesions, and cytokine mRNA expression in goats infected with a molecular clone of the caprine arthritis encephalitis virus

Caprine arthritis encephalitis virus (CAEV) is a lentivirus of goats that causes persistent infection characterized by the appearance of inflammatory lesions in various organs. To define the sites of persistence, 5 goats were infected with a molecular clone of CAEV, and the viral load was monitored by real-time-PCR and RT-PCR in different sites 8 years after infection. The lymph nodes proved to be an important virus reservoir, with moderate virus replication relative to what is reported for lentiviruses of primates. Mammary gland and milk cells were preferred sites of viral replication. The viral load varied significantly between animals, which points to an important role of the genetic background. We found a clear association between occurrence of histopathological lesions and viral load in specific sites. The mRNA expression analysis of several cytokines did not reveal differences between animals that could explain the considerable individual variations in viral load observed.

The phosphorylation status of the serine-rich region of the human cytomegalovirus 86-kilodalton major immediate-early protein IE2/IEP86 affects temporal viral gene expression

The 86-kDa major immediate-early protein (IE2/IEP86) of human cytomegalovirus (HCMV) contains a serine-rich region (amino acids 258 to 275) with several consensus casein kinase II (CKII) sites. We performed extensive mutational analysis of this region, changing serines to alternating alanines and glycines. Mutation of the serines between amino acids 266 and 275 eliminated in vitro phosphorylation by CKII. In vitro CKII phosphorylation of the serines between amino acids 266 and 269 or between amino acids 271 and 275 inhibited the ability of IE2/IEP86 to bind to TATA-binding protein. Correspondingly, nonphosphorylatable mutants in these regions showed increased activation of specific HCMV gene promoters in transfection studies. Viruses containing mutations of the serines throughout the entire region (amino acids 258 to 275) or the second half (amino acids 266 to 275) of the region showed delayed expression of all viral proteins tested and, correspondingly, delayed growth compared to wild-type HCMV. Mutation of the serines in the first half of the serine-rich region (amino acids 258 to 264) or between amino acids 266 and 269 propagated very slowly and has not been further studied. In contrast, mutation of the serines between amino acids 271 and 275 resulted in accelerated virus growth and accelerated temporal expression of viral proteins. These results suggest that the serine-rich region is structurally complex, possibly affecting multiple functions of IE2/IEP86. The data show that the phosphorylation state of the serine-rich region, particularly between amino acids 271 and 275, modulates the temporal expression of viral genes.

Role of murine cytomegalovirus US22 gene family members in replication in macrophages

The large cytomegalovirus (CMV) US22 gene family, found in all betaherpesviruses, comprises 12 members in both human cytomegalovirus (HCMV) and murine cytomegalovirus (MCMV). Conserved sequence motifs suggested a common ancestry and related functions for these gene products. Two members of this family, m140 and m141, were recently shown to affect MCMV replication on macrophages. To test the role of all US22 members in cell tropism, we analyzed the growth properties in different cell types of MCMV mutants carrying transposon insertions in all 12 US22 gene family members. When necessary, additional targeted mutants with gene deletions, ATG deletions, and ectopic gene revertants were constructed. Mutants with disruption of genes M23, M24, m25.1, m25.2, and m128 (ie2) showed no obvious growth phenotype, whereas growth of M43 mutants was reduced in a number of cell lines. Genes m142 and m143 were shown to be essential for virus replication. Growth of mutants with insertions into genes M36, m139, m140, and m141 in macrophages was severely affected. The common phenotype of the m139, m140, and m141 mutants was explained by an interaction at the protein level. The M36-dependent macrophage growth phenotype could be explained by the antiapoptotic function of the gene that was required for growth on macrophages but not for growth on other cell types. Together, the comprehensive set of mutants of the US22 gene family suggests that individual family members have diverged through evolution to serve a variety of functions for the virus.

Inhibition of nitric oxide enhances ovine lentivirus replication in monocyte-derived macrophages

Ovine lentivirus (OvLV) also known as maedi-visna virus, infects and replicates primarily in macrophages. This investigation examined the role of nitric oxide in the replication of OvLV in cultured macrophages. Peripheral blood mononuclear cells were collected from OvLV-free sheep and cultured in Teflon coated flasks at a high concentration of lamb serum. The cells were subsequently infected with OvLV strain 85/34. OvLV replication was assessed under different experimental treatments by comparison of reverse transcriptase (RT) activity in culture supernatant. Cultures that were treated with exogenous nitric oxide via S-nitroso-acetylpenicillamine did not have altered levels of RT activity compared to cultures treated with the inactive control compound, acetylpenicillamine. However, blockage of nitric oxide production by treatment with aminoguanidine, a competitive inhibitor of inducible nitric oxide synthase (iNOS), led to a significant rise in RT activity. This rise in RT activity was partially reversed in aminoguanidine treated cultures by L-arginine, the normal substrate for iNOS. Finally, the number of viral antigen producing cells was also quantified after aminoguanidine treatment and found to be significantly higher than untreated cultures. Collectively, these results indicate that nitric oxide is a negative regulator of OvLV replication in macrophages.

Phenotypic and ultrastructural characteristics of bronchoalveolar lavage cells of lentivirus-infected lambs treated with recombinant ovine IFN-tau

Ovine lentivirus (OvLV) belongs to the family Retroviridae and closely resembles the human immunodeficiency virus (HIV). Pulmonary lesions in OvLV-infected sheep consist of lymphoid interstitial pneumonia (LIP) and lymphocytic alveolitis. Similar pulmonary lesions occur in up to 40% of HIV-infected children and in some adults with AIDS. Interferon-tau (IFN-tau), a type I IFN, is produced by trophectoderm of ruminant conceptuses and is the pregnancy recognition signal in these species. To evaluate changes in phenotypes of bronchoalveolar lavage (BAL) cells of OvLV-infected lambs treated with recombinant ovine IFN-tau (rOvIFN-tau), 24 lambs were randomly allocated to one of four groups (n = 6 per group): 1, no virus + placebo (NVP); 2, no virus + rOvIFN-tau (NVI); 3, virus + placebo (VP); 4, virus + rOvIFN-tau (VI). The BAL cells from 3 lambs in each group were labeled with monoclonal antibodies (mAb) to cell surface markers at 16 weeks of treatment, and cells from the remaining 3 lambs in each group were labeled with mAb at 34 weeks of treatment. After labeling, BAL cells were analyzed by flow cytometry. The morphology of BAL cells from all experimental lambs was examined by transmission electron microscopy (TEM). At week 16, no differences in the relative proportions of BAL cell phenotypes were detected among the experimental groups. At week 34, VI lambs had higher proportions of CD8(+), gammadelta(+), MHC class II(+), and L-selectin (LS(+)) BAL cells compared with VP lambs. Higher proportions of CD14(+) and CD44(+) cells were found in VP lambs compared with NVP lambs at 34 weeks. OvLV-like particles were detected only in bronchoalveolar macrophages of VP lambs. In this study, rOvIFN-tau increased the proportions of primary antiviral gammadelta(+) and CD8(+) immune cells in OvLV-infected lambs. This may represent a cellular mechanism to explain the antiviral and therapeutic efficacy of this cytokine, in addition to its direct antiviral effect. However, because the actual number of cells labeled with mAb CD8 was low and some subsets of gammadelta cells may coexpress the CD8 marker, further studies are necessary to better define the role of rOvIFN-tau in the modulation of these cells in vivo.

Infection of dendritic cells by the Maedi-Visna lentivirus

The early stages of lentivirus infection of dendritic cells have been studied in an in vivo model. Maedi-visna virus (MVV) is a natural pathogen of sheep with a tropism for macrophages, but the infection of dendritic cells has not been proven, largely because of the difficulties of definitively distinguishing the two cell types. Afferent lymphatic dendritic cells from sheep have been phenotypically characterized and separated from macrophages. Dendritic cells purified from experimentally infected sheep have been demonstrated not only to carry infectious MVV but also to be hosts of the virus themselves. The results of the in vivo infection experiments are supported by infections of purified afferent lymph dendritic cells in vitro, in which late reverse transcriptase products are demonstrated by PCR. The significance of the infection of afferent lymph dendritic cells is discussed in relation to the initial spread of lentivirus infection and the requirement for CD4 T cells.

Langerhans cells migrate to local lymph nodes following cutaneous infection with an arbovirus

Whereas there has been recent interest in interactions between dendritic cells and pathogenic viruses, the role of dendritic cells in the initiation of protective immunity to such organisms has not been elucidated. The aim of this study was to examine whether a resident dendritic cell population in the skin, Langerhans cells, respond to cutaneous viral infections which are effectively cleared by the immune system. We therefore characterized the ability of Langerhans cells to migrate to local draining lymph nodes following infection with the arthropod-borne viruses, West Nile virus or Semliki Forest virus. The data show that major histocompatibility complex class II+/NLDC145+/E-cadherin+ Langerhans cell numbers are increased in the draining lymph nodes of infected mice and this increase is accompanied by a concomitant decrease in the Langerhans cell density in the epidermis. Langerhans cell migration is associated with an accumulation of leukocytes in the lymph node, which is one of the earliest events in the initiation of an immune response. Both the migratory response and the draining lymph node leukocyte accumulation were abrogated if ultraviolet-inactivated instead of live viruses were used, suggesting the activation and subsequent migration of Langerhans cells requires a live, replicating antigen. Our findings are likely to have wider implications for the development of epidermally delivered vaccines and suggest that mobilization of dendritic cells may be involved in the development of immune responses to arthropod-borne viruses.

In vivo depletion of CD8+ cells does not affect primary maedi visna virus infection in sheep

T-cells have been implicated both, in promoting and reducing viral replication during lentivirus infection. CD8+ lymphocytes are believed to be important in controlling viral load through direct killing of virus-infected cells and by secretion of inhibitory chemokines and cytokines. To evaluate the role of CD8+ T-cells in the induction and control of the primary phase of a lentivirus infection, we have used a non-T-cell tropic lentivirus, maedi-visna virus (MVV), to study the initial pathogenesis and subsequent immune responses in sheep depleted in vivo of CD8+ cells. Sheep were depleted of CD8+ cells in both blood and efferent lymph for up to 14 days. No difference in MVV replication was observed in either the draining efferent lymph or lymph node of these sheep. Surprisingly, these animals displayed a normal induction of pCTL whereas the virus-specific proliferative responses were reduced. This could reflect either that a proportion of functional CD8+ lymphocytes remained in these animals, as suggested by the appearance of pCTLs, or that CD8+ cells are not required for control of primary MVV infection.

CD4(+) T-cells are required for the establishment of maedi-visna virus infection in macrophages but not dendritic cells in vivo

The role of CD4(+) lymphocytes in the establishment of lentivirus infection in macrophages has been studied in an in vivo system of lentivirus infection where CD4(+) lymphocytes are not the targets for infection. Using the non-T-cell-tropic lentivirus, maedi-visna virus (MVV), in CD4-depleted sheep, we have found that CD4(+) T cells were required for MVV infection in macrophages but not dendritic cells. CD4-depleted sheep had significantly lower levels of MVV-infected cells in lymph nodes and efferent lymph after MVV challenge in the drainage area of the lymph node. Due to the absence of virus in combination with the lack of CD4(+) T helper cells, virus-specific immune responses were reduced. There was delayed induction of cytotoxic T cell precursors, a marked reduction in virus-specific in vitro proliferative responses, and a delay in the appearance of MVV-specific antibodies. By contrast, CD4 depletion had no effect on the establishment of MVV infection in afferent lymph dendritic cells migrating from the skin infection site to the lymph node.

T-lymphocyte populations in the blood of caprine arthritis encephalitis virus-infected goats

The effect of natural infection by Caprine Arthritis Encephalitis Virus on the phenotypic pattern of T lymphocytes in peripheral blood was studied in a herd of 127 milking goats by flow cytometry. Total leukocyte and T-lymphocyte numbers tend to decrease with age, with only small changes in the CD4/CD8 ratio. The lymphocyte phenotypes show no strong correlation with seropositivity to CAEV or the presence of clinical symptoms, suggesting that this macrophagetropic lentivirus does not greatly effect the lymphocyte population.

Immunohistochemical identification of caprine arthritis-encephalitis virus in paraffin-embedded specimens from naturally infected goats

The expression of caprine arthritis-encephalitis virus capsid protein was studied in seropositive naturally infected asymptomatic goals (10< seropositive naturally infected encephalitic kids (12) and goats (4), and noninfected control goats (3). Rabbit antiserum to recombinant viral capsid and matrix proteins were used in a biotin-streptavidin-alkaline phosphatase complex immunohistochemical method on sections of formalin- and ethanol-fixed tissue specimens. Macrophages in inflamed areas of the lung (8/12), in the brain (5/16), and in the spinal cord (4/16) from encephalitic animals harbored viral antigens, as revealed by immunohistochemistry and use of a capsid protein-specific antiserum. Altogether 12/16 encephalitic animals tested positive for viral antigen. Viral antigens were found in 5/10 seropositive asymptomatic goals in macrophages located in the lung (3), the udder (1), and the medulla of lymph nodes (4). None of the control animals tested positive for viral antigen. Ethanol fixation showed highest sensitivity, and the lowest antigen concentration that revealed a positive signal discernible from background was twofold higher in ethanol-fixed specimens than in formalin-fixed specimens. The evaluation was performed on artificial antigen substrates embedded with defined concentrations of recombinant viral capsid protein. Immunohistochemistry is a valuable supplement to the methods presently available for diagnosis in cases suspicious of caprine arthritis-encephalitis.

The changing preference of T and B cells for partners as T-dependent antibody responses develop

Recirculating virgin CD4+ T cells spend their life migrating between the T zones of secondary lymphoid tissues where they screen the surface of interdigitating dendritic cells. T-cell priming starts when processed peptides or superantigen associated with class II MHC molecules are recognised. Those primed T cells that remain within the lymphoid tissue move to the outer T zone, where they interact with B cells that have taken up and processed antigen. Cognate interaction between these cells initiates immunoglobulin (Ig) class switch-recombination and proliferation of both B and T cells; much of this growth occurs outside the T zones B cells migrate to follicles, where they form germinal centres, and to extrafollicular sites of B-cell growth, where they differentiate into mainly short-lived plasma cells. T cells do not move to the extrafollicular foci, but to the follicles; there they proliferate and are subsequently involved in the selection of B cells that have mutated their Ig variable-region genes. During primary antibody responses T-cell proliferation in follicles produces many times the peak number of T cells found in that site: a substantial proportion of the CD4+ memory T-cell pool may originate from growth in follicles.

Maedi-Visna and ovine progressive pneumonia

Maedi-Visna and ovine progressive pneumonia are disease of sheep that are caused by ovine lentivirus and characterized by chronic inflammation of the lungs, mammary glands, joints, and central nervous system. Although tremendous progress in research has led to a better understanding of the pathogenesis of these diseases, many questions still remain. Much of the mystery is the result of the complexity of the ovine lentivirus genome and the intricate interactions of the virus with the host during replication. Discoveries in molecular virology are shedding light on these interactions and novel approaches to prevent and control lentivirus infections are being explored. There is hope that some of these approaches will eventually be used to eradicate these diseases.

The phenotype and fate of the antibody-forming cells of the splenic foci

The first product of the humoral response to antigen is low-affinity antibody, produced by extrafollicular foci of antibody-forming cells (AFC) in organs such as spleen and lymph node. These cells proliferate rapidly but then undergo an equally rapid decline, so that they are present in only small numbers 14 days after immunization. We have used 6-parameter flow cytometry to isolate and examine the characteristics of (4-hydroxy-5-nitrophenyl)acetyl-specific AFC, looking in particular for those markers that might differentiate them from cells of the intrafollicular (germinal center) arm of the T-dependent immune response. At day 7 of the primary response, most AFC were found to express surprisingly low levels of B220, high levels of syndecan, and retain significant levels of surface IgG1. We then used enzyme-linked immunospot assays to demonstrate that the rapid decline of these cells was not likely to be due to migration to organs such as the bone marrow. Their decline could, however, be explained by apoptosis in situ, which was demonstrated immunohistologically by nick-end labeling.