INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Human T-cell leukemia/lymphoma virus type 1 (HTLV-1), the causative agent of adult T-cell leukemia (36) and tropical spastic paraparesis/HTLV-1-associated myelopathy (TSP/HAM) (10), has also been associated with pediatric infectious dermatitis (22), uveitis (24), and some cases of arthropathy (13) and polymyositis (25).

Owing to the inherent difficulty of obtaining human specimens in early HTLV-1 infection, a relevant animal model is essential for better understanding of the seeding of HTLV-1 in various organs. We showed recently that the squirrel monkey Saimiri sciureus, a South American nonhuman primate which is free of simian T-cell leukemia virus, is susceptible to experimental infection with syngeneic or allogeneic HTLV-1-immortalized cells (19). As in humans, such experimental inoculation leads to chronic infection, HTLV-1 provirus being detectable in peripheral blood mononuclear cells (PBMCs) by PCR up to 36 months after inoculation. Chronically infected monkeys, like HTLV-1-infected humans, develop high titers of antibodies against the structural proteins of the virus (19).

In infections with human immunodeficiency virus type 1 (HIV-1) and simian immunodeficiency virus (SIV), the lymph nodes are reservoirs that contain a large viral load during the asymptomatic stage (8, 29). A significant percentage of latently infected lymphocytes are found in the lymph nodes and constitute a target for later viral reactivation (7). Indeed, during clinical latency in HIV or SIV infection, productively infected cells are detected at higher frequency in lymph nodes than in peripheral blood, indicating the progressive involvement of the lymphoid organs in HIV infection (2, 21, 27). These findings show that HIV is widely disseminated in lymphoid tissue relatively early in infection. In contrast, little is known about the role of the lymphoid tissues in the early phases of HTLV-1 infection.

The aim of the present study was to investigate the early events in HTLV-1 infection and specifically the distribution of the HTLV-1 provirus in various organs of serially sacrificed squirrel monkeys, to localize the reservoir of the latent virus, and to evaluate the humoral and cellular immune responses to infection.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and HTLV-1 infection. Three 6-year-old male, two 9-year-old male, and two 7-year-old female squirrel monkeys from the primate breeding center of the Pasteur Institute of French Guiana were used in this study. The experiments performed complied with French laws relative to animal experimentation. Five animals were inoculated intravenously with 5 × 10⁷ EVO/798 monkey HTLV-1-transformed cells, and one monkey (1491) had been infected previously with 10⁷ of these cells; the seventh animal (1657) had been infected previously by intravenous inoculation with 10⁷ monkey HTLV-1-transformed EVO/1540 cells (19). The monkeys were bled at various times after inoculation, and their PBMCs were separated by Ficoll-Paque. The sera were diluted 1/100, and antibodies against HTLV-1 were measured by enzyme-linked immunosorbent assay (ELISA; Diagnostic Biotechnology, Singapore) at various times after infection and then confirmed by Western blotting (HTLV blot 2.3; Diagnostic Biotechnology).

Detection of the HTLV-1 provirus by PCR and quantitative PCR. Genomic DNA was extracted from PBMCs and various organs by the method described by Ibrahim et al. (11). PCR (35 cycles) was performed, as previously described, with primers Rmtax1 and Rmtax2 (antisense) for the tax region (202 nucleotides) (23) and with primers gag1 and gag2 for the gag region (17). One microgram of genomic DNA from the PBMCs or organs was used for each PCR amplification. The amplified products were submitted to electrophoresis on a 1.4% agarose gel, transferred to nylon membranes, and hybridized with ³²P-labeled specific internal oligonucleotide probes for tax (5' GGGGCCCTAATAATTCTACCCGAAGACT 3') and gag (5' GCAAAGGTACTGCAGGAGGT 3'). The membranes were washed and exposed to Hyperfilm MP (Amersham, Little Chalfont, United Kingdom) at 80°C for 24 h and 1 week. The sensitivity of the PCR was evaluated with HTLV-1 plasmid p4.39, cloned from HTLV-1 cell line 2060 (26). Plasmid p4.39 DNA was diluted at various concentrations in 1 µg (150,000 cell equivalents) of DNA from uninfected PBMCs and analyzed by PCR.

Detection of tax/rex mRNA by RT-PCR. Total cellular RNA was extracted from PBMCs and tissues with Trizol reagent (Gibco BRL, Grand Island, N.Y.). The RNA was suspended in water and diluted to a final concentration of 0.1 µg/µl. It was then mixed with 15 pmol of the random hexamer primers in a final volume of 12.7 µl. Reverse transcription (RT) and cDNA synthesis were performed as described previously (30). Immediately after cDNA synthesis, a seminested PCR was carried out on the pX region, with RPX3 and RPX5 as the outer primers and RPX3 and RPX4 as the inner primers (20). The sequence of the RPX5 primer was 5' AGGCGGGCCGAACATAGTCCC 3' (antisense). To confirm the presence of amplifiable cDNA in the samples, primers 18SL (5' CCATGGAGAAGGCTGGGG 3'; sense) and 20SL (5' CAAAGTTGTCATGGATGACC 3'; antisense) were used to detect a portion of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA.

Fate of inoculated cells. To confirm that the expression of tax/rex mRNA of HTLV-1 and HTLV-1 provirus that was detected was not that of surviving cells from the inoculum, we evaluated the fate of the inoculum by inoculating two females with the HTLV-1-transformed cell line EVO/798, derived from a male, and sacrificing the two females 6 and 35 days after inoculation. By PCR with primers VIB and SRY Omega (35), a portion of the Saimiri Y chromosome gene SRY was amplified and identified by hybridization with the SRY probe (5' TATACCTTCTCATACAGAGATAACTGTACAAAATCC 3'). This probe was defined after cloning and sequencing of the amplified Saimiri SRY product.

In situ hybridization. In situ hybridization was performed on serial sections of various frozen tissues as described by Vazeux et al. (33) and Boche et al. (3). Briefly, ³³P- or ³⁵S-labeled antisense riboprobes corresponding to the complete tax mRNA sequence were used. These riboprobes were derived from the T3 promoter of a Bluescript vector. The sections were exposed for 10 to 20 days to 10⁶ cpm/slide. The monkey HTLV-1-transformed cell line EVO/798 was used as a positive control, and negative controls were performed with PBMCs from an HTLV-1-seronegative monkey and frozen monkey tissues in which tax expression was not detected by RT-PCR.

Detection of anti-HTLV-1 cell-mediated immunity. Autologous Epstein-Barr virus-transformed B-lymphoblastoid cell lines were obtained as described previously (4) and used as target cells in the assay for cytolytic activity after infection with vaccinia virus-HTLV-1 constructs or loading with exogenous HTLV-1 p40^Tax peptides. The recombinant vaccinia viruses were kindly provided by H. Shida, Kyoto University, Kyoto, Japan. The overlapping p40^Tax peptides described by Parker et al. (28) were kindly provided by C. Pique, Institut Cochin de Génétique Moléculaire, Paris, France.

Nucleotide sequence accession number. The GenBank accession number of the Saimiri SRY sequence reported (359 nucleotides) is AF151695.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Detection of the HTLV-1 provirus and its distribution in organs of squirrel monkeys. The distribution of the HTLV-1 provirus was evaluated by PCR with gag and tax primers in 20 organs from the six experimentally infected squirrel monkeys sacrificed at various times after infection (Table 1). This method has been shown to detect five copies of plasmid p4.39 diluted in 1 µg of carrier DNA.

Fate of the inoculum. The fate of the inoculum was evaluated in two females (90047 and 90083) sacrificed 6 and 35 days after inoculation. As seen in Fig. 1, PCR with primers that detect part of the SRY sequence of the Y chromosome showed the presence of the inoculum 24 h later but not thereafter. The presence of the inoculum correlates with the detection of HTLV-1 provirus (202 bp of the tax region) by PCR and tax/rex mRNA by RT-PCR at 24 h in the PBMCs of these inoculated monkeys (Fig. 1 and 2A). No positive signal for the SRY was detected in PBMCs or in any organs at necropsy. The absence of a positive signal for the SRY in PBMCs and organs shows that the presence of HTLV-1 provirus was due to infection of the PBMCs and not to persistence of the inoculum.

View larger version (34K): [in this window] [in a new window]   FIG. 1.   Detection of HTLV-1 tax sequence (202 bp) (first line) and SRY sequence of Y chromosome (second line) by PCR in PBMCs of monkeys 90047 (female), 92038 (male), and 90083 (female) sacrificed 6, 12, and 35 days after inoculation. C and C+ represent negative and positive controls, respectively.

View larger version (35K): [in this window] [in a new window]   FIG. 2.   Detection of HTLV-1 provirus expression (tax/rex mRNA) by RT-PCR. (A) Detection of tax/rex mRNA (first line) and GAPDH sequence (second line) in PBMCs of monkeys 90047, 92038, and 90083 sacrificed 6, 12, and 35 days after inoculation. (B) Detection of tax/rex mRNA (first line) and GAPDH sequence (second line) in spleens of different monkeys sacrificed at various times after inoculation. C and C+ represent negative and positive controls; respectively.

Dynamics of virus infection in PBMCs as detected by PCR. As seen in Fig. 1, no or very low positive signals were detected by PCR in the PBMCs of three monkeys 3 days after inoculation, but the signals increased after 8 days and became stable after 16 days in the monkey sacrificed at 35 days (Fig. 1). Furthermore, in three other animals (92106, 92039, and 1657) sacrificed at 3, 6, and 26 months after inoculation, the HTLV-1 provirus could be detected by PCR at 3 weeks after inoculation and thereafter. Quantitative PCR was performed to determine the proviral load in 10⁵ PBMCs from two monkeys, 90083 and 92039, sacrificed at 35 days and at 6 months. In the monkey sacrificed at 35 days, we detected 167 HTLV-1 proviral copies 8 days after infection and 557 copies at 35 days. In the animal sacrificed at 6 months, we detected 354 proviral copies 3 weeks after infection and 413 copies on the day of necropsy.

Detection of HTLV-1 provirus expression (tax/rex mRNA) and in situ hybridization in various organs of squirrel monkeys. RT-PCR was used to examine the expression of tax/rex mRNA in fresh PBMCs and in organs obtained from the animals at necropsy. tax/rex mRNA was detected by seminested PCR only in the PBMCs, spleen, and mesenteric and submaxillary lymph nodes of the animal sacrificed 12 days after inoculation and in PBMCs collected at 12 days from the animal sacrificed at 35 days (Fig. 2). In other animals, tax/rex mRNA could not be detected in PBMCs at 3 weeks or in organs at necropsy. In the animal sacrificed at 12 days, in situ hybridization on several frozen sections of the spleen revealed scattered HTLV-1-positive lymphocytes around the germinal centers and next to arterial capillaries (Fig. 3). Cells with signals for tax/rex mRNA were observed reproducibly in serial sections hybridized with ³³P- and ³⁵S-labeled riboprobes. In contrast, we observed no positive cells in frozen or paraffin-embedded organs, including spleen, lymph nodes, and PBMCs, from monkeys infected longer.

View larger version (119K): [in this window] [in a new window]   FIG. 3.   In situ hybridization with HTLV-1 tax riboprobe on frozen spleen sections from an infected squirrel monkey sacrificed 12 days after inoculation. (A and B) Positive signals in the spleen observed with antisense 33P-labeled riboprobe; arrows indicate positive spleen cells after 10 days of in situ hybridization. (C) Negative control (PBMCs from squirrel monkey not infected with HTLV-1). (D) Positive control (EVO/798 HTLV-1 monkey-transformed cell line).

Pathological and histological investigations. Some lesions and lymphocyte infiltrations were observed in the salivary glands of two monkeys sacrificed at 12 days and 3 months, and follicular hyperplasia was observed in the spleen and lymph nodes of four of the six monkeys.

Humoral responses of inoculated monkeys. When the antibody response to HTLV-1 was evaluated by ELISA, no anti-HTLV-1 antibodies were detected in the two animals sacrificed at 6 and 12 days; however, as seen in Fig. 4, the monkey sacrificed at 35 days had high antibody responses from 16 days after infection, and the three monkeys sacrificed at 3, 6, and 26 months after inoculation had antibodies 3 to 6 weeks after infection. The positive sera reacted on Western blotting against Gag proteins p19 and p24 and Env gp46 peptide (amino acids 162 to 209) and recombinant gp21. This antibody response remained stable for 3 to 26 months.

View larger version (17K): [in this window] [in a new window]   FIG. 4.   Antibody responses of monkeys against HTLV-1, tested by ELISA after inoculation with homologous HTLV-1 monkey-transformed cell lines. Monkey 90083 was sacrificed at 35 days, monkey 92106 was sacrificed at 3 months, monkey 92039 was sacrificed at 6 months, and monkey 1657 was sacrificed at 26 months. O.D., optical density.

Anti-HTLV-1 cell-mediated immunity. In the two monkeys sacrificed 6 and 12 days after inoculation, no significant cytolytic activity was detected in splenocytes against target cells infected with recombinant vaccinia virus expressing the whole p40^Tax protein (VacTax; Fig. 5A and B) or loaded with p40^Tax peptides. The monkey sacrificed 6 months after inoculation showed cytolytic activity in PBMCs against VacTax-infected target cells at 2 months (Fig. 5C) but not at 2 or 6 weeks (data not shown). This response against p40^Tax seemed to be directed against various Tax epitopes, as we detected cell-mediated immune responses against Tax peptides 151-165, 171-185, 261-275, 265-279, and 271-285 (Fig. 5D). Furthermore, at necropsy at 6 months, we observed strong cytolytic activity in splenocytes against p40^Tax but only a weak response against Env and no specific cytolytic activity against Gag (Fig. 5E).

View larger version (2332K): [in this window] [in a new window]   FIG. 5.   Cultured CD4+-depleted CD8+-enriched cells obtained from PBMCs or splenocytes, used as effectors in cytolytic chromium release assays. Target cells were autologous B-lymphoblastoid cell lines either uninfected (n.i.), vaccinia virus infected, or Tax peptide loaded. In some cases, a 25- or 50-fold excess of cold autologous vaccinia virus-infected B-lymphoblastoid target cells was added to reduce background. (A) Monkey 90047 sacrificed day 6 after inoculation; CD4-depleted splenocytes by day 12 of culture. (B) Monkey 92038 sacrificed day 12 after inoculation; CD4-depleted splenocytes by day 18 of culture. (C) Blood from monkey 92039 sampled at 2 months; CD4-depleted PBMCs by day 18 of culture. (D) Blood from monkey 92039 sampled at 2 months; CD4-depleted PBMCs by day 16 of culture. (E) Monkey 92039 sacrificed at 6 months; CD4-depleted splenocytes by day 24 of culture. (F) Blood from monkey 1491 sampled 20 and 30 months after inoculation; CD4-depleted PBMCs by days 23 and 15 of culture, respectively. (G) Blood from monkey 1491 sampled 20 and 30 months after inoculation; CD4-depleted PBMCs obtained from blood sample at 30 months after inoculation, by day 19 of culture. (H) Blood from monkey 1657 sampled 20 months after inoculation; CD4-depleted PBMCs by day 26 of culture. (I) Monkey 1657 sacrificed 26 months after inoculation; CD4-depleted splenocytes by day 21 of culture.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We show here that during early HTLV-1 infection of a nonhuman primate (S. sciureus), PBMCs, the spleen, and lymph nodes serve as major reservoirs for HTLV-1.

In our previous studies (11, 18), in which adult rats were used as the experimental model to examine the distribution of the HTLV-1 provirus in serially sacrificed animals, the most frequent locations of the provirus 12 weeks after infection were PBMCs and the spinal cord; however, 22 weeks after inoculation, when the provirus could no longer be detected in these organs, it was present in sympathetic nodes. In a study reported by Ishiguro et al. (14) in which rats were infected as neonates, the HTLV-1 provirus was widely disseminated in the lymphoid system, central nervous system, heart, liver, lung, and kidney. The difference from our results could be linked to the age at which the animals were infected, as neonates rather than as adults, and to the fact that the rats and monkeys in our studies were thoroughly perfused before necropsy. PBMCs and mesenteric lymph nodes were also found to be major reservoirs in HTLV-1-infected rabbits (6). Ibuki et al. (12) showed, however, the presence of HTLV-1 provirus in PBMCs, spleens, and lymph nodes of cynomolgus monkeys 51 weeks after inoculation of an HTLV-1-producing T-cell line derived from a macaque monkey.

No data are available on the reservoirs of HTLV-1 in humans during the asymptomatic period. In HIV or SIV infection, it is now accepted that the major viral reservoir during the asymptomatic period is lymphoid tissue (2, 8, 21, 27). In the case of lentiretroviruses, the lymphoid tissue harbors numerous virus-producing cells and virions trapped as immune complexes in the follicular dendritic cell network of the germinal centers. These trapped virions could act as a source of infectious virus (2). Infected cells can migrate into and through lymphoid tissue, and naive CD4⁺ cells, in particular, can be infected within the germinal centers of the lymphoid tissue.

In the present study, HTLV-1 provirus could be detected by PCR 6 days after infection, and we showed by competitive PCR that the proviral load increases in PBMCs of infected animals shortly after infection and remains stable thereafter. RT-PCR and in situ hybridization indicated, however, that HTLV-1 tax/rex mRNA is transcribed transiently during the early stage of infection (12 days). We cannot conclude that tax/rex expression is not maintained in PBMCs, but it is too weak to be detected by our method. Nevertheless, it is likely that tax/rex mRNA expression does not last long and that the latency occurs very early after infection of squirrel monkeys with HTLV-1. HTLV-1 expression is rarely detected in asymptomatic human chronic carriers of HTLV-1, despite a high proviral load, suggesting that HTLV-1 replication in vivo results predominantly from mitosis of latently infected cells rather than from RT (34). The genetic stability of HTLV-1, its weak expression, and the high proviral load cited above favor such a hypothesis. Wattel et al. (34) further suggested that primary infection with HTLV-1 must involve infected allogeneic cells, followed by RT and viral replication. Once a few rounds of RT have taken place, clonal expansion of the infected cells should predominate. Our results in the squirrel monkey model seem to confirm this observation, as we found expression of tax/rex mRNA in PBMCs and lymphoid organs only during the early stage of infection and found only latent HTLV-1 proviral sequences at these sites thereafter. This could reflect clonal expansion of infected cells lacking detectable viral RNA. Since only a small proportion of HTLV-1 carriers develop malignant disease after 30 to 50 years of infection, such clonal expansion must be actively controlled by specific humoral or cell-mediated immunity.

As a first step to addressing the role of the cell-mediated immune response in controlling HTLV-1 infection in the squirrel monkey model, we analyzed its presence at various times after infection in serially sacrificed animals. No clear phenotyping of the cytolytic effectors or of major histocompatibility complex (MHC) restriction could be done, as there are no relevant monoclonal antibodies and MHC class I molecules in the squirrel monkey have not been well characterized. Work is in progress to better characterize this response. Nevertheless, two observations support the hypothesis that the cytolytic effectors responsible for the responses we observed are CD8 T lymphocytes. (i) PBMCs were CD4 depleted before addition of T mitogen, which would favor expansion of CD8 T lymphocytes in the culture medium. (ii) NK activity is unlikely, as the B cells used as targets express large amounts of class I molecules, which are known to inhibit NK activity; furthermore, NK activity against unloaded viral peptide or wild vaccinia virus-infected B cells was not observed. We therefore conclude that the cell-mediated immune responses observed in our study are due to cytotoxic T-lymphocyte (CTL) activity.

While there was no detectable CTL response during the early stage of infection (6 and 12 days and 2 or 6 weeks after inoculation), activity was detected after 2 months and was present during chronic infection (up to 20 months). The CTL activity was directed mainly against HTLV-1 p40^Tax protein. In rats infected with HTLV-1, CTL responses can be detected against Gag, Env, and Tax, with a stronger reaction against Env protein (31). In asymptomatic human carriers of HTLV-1 and in patients with TSP/HAM, the CTL response is also directed mainly against p40^Tax protein (15). Bangham et al. (1) suggested that this response could play a major role in controlling HTLV-1 replication, but others consider that it could be detrimental in some genetically susceptible individuals, leading them to develop TSP/HAM (15). In our squirrel monkey model, the CTL response seem to be comparable to that observed in asymptomatic human HTLV-1 carriers but with different epitopic targets, probably because of differences in antigen presentation by MHC class I molecules.

It was of interest that virus expression in squirrel monkeys was detected only in the absence of humoral and CTL responses. In SIV infection, both neutralizing antibodies and the CTL response play important roles in limiting viral replication and in clearance of the agent (9, 16). The CTL response in our nonhuman primate model, which is probably mediated by CD8⁺-enriched lymphocytes, could be of interest for evaluating the role of cytotoxic CD8⁺ T-cell responses during HTLV-1 infection; however, immunological tools such as monoclonal antibodies against squirrel monkey CD4 and CD8 molecules will be necessary. Although one monoclonal antibody cross-reactive between human and Saimiri CD4 molecules has been identified (Leu-3a), none of the anti-human CD8 monoclonal antibodies tested has similar cross-reactivity. In an attempt to derive anti-squirrel monkey CD8 monoclonal antibodies, we cloned and sequenced the cDNAs encoding squirrel monkey CD8  and  chains (32). The expression of these chains in appropriate mouse cells should enable us to derive the monoclonal antibodies needed to evaluate the role of the antiviral CTL response in chronic HTLV-1 infection and in HTLV-1-induced neurological disease.