Neuropathology of HTLV‐1‐associated myelopathy (HAM/TSP)

A series of our neuropathological studies was reviewed in order to clarify pathogenesis of human T lymphotropic virus type 1(HTLV‐1)‐associated myelopathy (HAM)/tropical spastic paraparesis (TSP). The essential histopathologic finding was chronic inflammation in which inflammatory infiltrates of mononuclear cells and degeneration of myelin and axons were noted in the entire spinal cord. Immunohistochemical analysis demonstrated T‐cell dominance, and the numbers of CD4+ cells and CD8+ cells were equally present in patients with shorter clinical courses. Apoptosis of helper/inducer T‐cells were observed in these active inflammatory lesions. Horizontal distribution of inflammatory lesions was symmetric at all spinal levels and was accentuated at sites with slow blood flow in the middle to lower thoracic levels. HTLV‐1 proviral DNA amounts were well correlated with the numbers of infiltrated CD4+ cells. In situ PCR of HTLV‐1 proviral DNA and in situ hybridization of HTLV‐1 Tax gene demonstrated the presence of HTLV‐1‐infected cells exclusively in the mononuclear infiltrates of perivascular areas. From these findings, it is suggested that T‐cell mediated chronic inflammatory processes targeting the HTLV‐1 infected T‐cells is the primary pathogenic mechanism of HAM/TSP. Anatomically determined hemodynamic conditions may contribute to the localization of infected T‐cells and the formation of main lesions in the middle to lower thoracic spinal cord.     

Pathological mechanisms of human T-cell lymphotropic virus type I-associated myelopathy (HAM/TSP)

The recent studies have greatly improved our understanding of the pathological mechanisms of human T cell lymphotropic virus type I (HTLV-I)-associated myelopathy/tropical spastic paraparesis (HAM/TSP). The pathological mechanisms of HAM/TSP based on the histopathological, immunological, and molecular analysis with emphasis on the longitudinal alterations of the disease will be discussed. Immunohistological examination revealed the existence and the activation both of HTLV-I-infected CD4+ cells and HTLV-I-specific CD8+ cytotoxic T lymphocytes in the spinal cord lesions, which suggest that they play an important role in the pathogenesis. Increased expression of several cytokines, Fas/Fas ligand, adhesion molecules, and molecules influencing T cell migration in the lesions have been reported. These cell infiltrates and cytokines they secrete in the lesions may damage bystander neural tissue. Furthermore, longitudinal alterations in the affected spinal cords suggest that the inflammatory process is gradually decreased. Epidemiological studies show that less than 5% of infected individuals develop HAM/TSP and indicate that increased proviral load of HTLV-I is a strong predictor for the development of HAM/TSP. A recent study has shown that the autoantibody for the ribonuclear protein-A1 can cross-react with HTLV-I Tax protein and inhibit neuronal firing ex vivo, indicating that a molecular mimicry of the humoral immune response may be involved in the pathogenesis of HAM/TSP. Based on these studies, two hypotheses can be proposed for the pathogenesis of HAM/TSP, where cellular and humoral immune responses both play important roles.     

Perivascular T cells are infected with HTLV-I in the spinal cord lesions with HTLV-I-associated myelopathy/tropical spastic paraparesis: double staining of immunohistochemistry and polymerase chain reaction in situ hybridization

HTLV-I-infected cells play an important role in pathogenesis HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP). Our previous studies of quantitative polymerase chain reaction (PCR) and in situ PCR suggested that T cells infiltrating in the spinal cord lesion were infected with HTLV-I. To elucidate the localization of HTLV-I proviral DNA directly, we performed double staining using immunohistochemistry and PCR in situ hybridization (PCR-ISH). Fresh frozen sections of the spinal cord from four HAM patients taken at autopsy were first immunostained with antibodies to pan T cells (UCHL-1), macrophages (KP-1) and helper/inducer T cells (OPD4). Then PCR-ISH was carried out with specific primers and probe for the HTLV-I pX region. UCHL-1-positive cells were noted around perivascular areas and, to some extent, in the parenchyma. Of the UCHL-1-positive cells, 9.4% (case 1), 9.6% (case 2), 1.1% (case 3) and 6.7% (case 4) became positive in HTLV-I PCR-ISH. UCHL-1-negative cells were HTLV-I PCR-ISH negative and almost all KP-1-positive cells were HTLV-I negative. HTLV-I was localized to OPD4-positive cells in examined lesions of cases 2 and 4. These data are a direct demonstration of HTLV-I proviral DNA localizing to infiltrated T cells in HAM/ TSP spinal cord lesions. Received: 1 December 1997 / Revised, accepted: 20 March 1998     

T cell transformation with Herpesvirus saimiri: a tool for neuroimmunological research

Splice variants of CD44 molecule-harboring exon 10 (v6), often called v6 variants (v6v), are shown to confer tumor progressive, metastatic or invasive capacities. Furthermore, CD44 molecule on activated T-cells are shown to be required for infiltration of these cells into the inflammatory site and for accelerated immune response. Human T-cell lymphotropic virus type I (HTLV-I)-associated myelopathy/tropical spastic paraparesis (HAM/TSP) is caused by HTLV-I infection and characterized by spastic paraparesis and urinary disturbance with perivascular HTLV-I-infected and activated CD4+ T-cell infiltration. In order to explore the underlying mechanism causing the disease after HTLV-I infection, we analyzed CD44 variant expression on peripheral blood mononuclear cells (PBMC) and in the spinal cord specimens from patients with HAM/TSP, and compared them with those from other HTLV-I-infected individuals and controls. We found that v6v expression with special direct link of exons 10 (v6) and 14(v10) was highly expressed in PBMC from patients with HAM/TSP and that v6v and CD4 double positive T-cell infiltration into the spinal cord lesion of HAM/TSP. This combination of CD44 splice variant has not been previously reported in the study of chronic inflammatory disorders and may be a marker molecule for T-cells infiltrating into the central nervous system (CNS), especially the spinal cord.     

Cell surface phenotype of in vitro proliferating lymphocytes in HTLV-I-associated myelopathy (HAM/TSP)

The in vitro proliferation of peripheral blood lymphocytes (PBLs) without any mitogenic stimulation is one of the hallmarks of human T lymphotropic virus type I (HTLV-I) infection. Recent evidence suggests a difference in the degree of the phenomenon between HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP) and asymptomatic HTLV-I carriers (AC). In this article, we demonstrated several alterations in the features of the in vitro transformed lymphocytes between patients with HAM/TSP (n = 16) and AC (n = 8). The percentages of total CD8+ and CD8+CD28+ cells were significantly increased in the in vitro proliferating T lymphocytes derived from the patients with HAM/TSP when compared to those from AC. HAM/TSP was segregated from AC by the high degree of the proliferation of CD8+CD28+ cells. The expression of HTLV-I-specific antigens on the cultured PBLs was detected only in the subjects which showed low CD8+CD28+/CD4+ ratio of the in vitro proliferating lymphocytes. These findings suggest that this phenomenon distinguishes HAM/TSP from AC, not only in quantity but also in quality.     

Detection of human T-lymphotropic virus type I (HTLV-I) tax RNA in the central nervous system of HTLV-I–associated myelopathy/tropical spastic paraparesis patients by in situ hybridization

Autopsy specimens from 3 patients with human T-lymphotropic virus (HTLV-I)–associated myelopathy/tropical spastic paraparesis (HAM/TSP) were examined for the presence of HTLV-I in the central nervous system (CNS). In situ hybridization using an HTLV-I tax RNA probe detected cells containing HTLV-I RNA in spinal cord and cerebellar sections. HTLV-I infected cells were located within the white matter and, in particular, within the anterior and lateral funiculi of the spinal cord. Consistent with previously described HAM/TSP pathology, there were perivascular infiltrates in these CNS specimens. Significantly, HTLV-I RNA was not localized to these infiltrates but was detected deeper within the neural tissue. Furthermore, phenotypic analysis demonstrated that at least some of the infected cells were astrocytes. While previous polymerase chain reaction studies have demonstrated the presence of proviral HTLV-I in CNS specimens, here we provide evidence for the in situ expression of HTLV-I RNA in the CNS of HAM/TSP patients.     

Human T-cell lymphotropic virus type I and neurological diseases

Human T-cell lymphotropic virus type I (HTLV-I) infection is associated with a variety of human diseases. In particular, there are two major diseases caused by HTLV-I infection. One is an aggressive neoplastic disease called adult T-cell leukemia (ATL), and another is a chronic progressive inflammatory neurological disease called HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP). It is still unknown why one virus causes these different diseases. With regard to HAM/TSP, virus-host immunological interactions are an considered to be important cause of this disease. Coexisting high HTLV-I proviral load and HTLV-I-specific T cells (CD4+ T cells and CD8+ T cells) is an important feature of HAM/TSP. Histopathological studies indicate the existence of an inflammatory reaction and HTLV-I-infected cells in the affected lesions of HAM/TSP. Therefore, the immune response to HTLV-I probably contributes to the inflammatory process of the central nervous system lesions in HAM/TSP patients.     

Flow cytometry evaluation of the T-cell receptor Vbeta repertoire among human T-cell lymphotropic virus type-1 (HTLV-1) infected individuals: effect of interferon alpha therapy in HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP)

Human T-cell lymphotropic virus type-1 (HTLV-1)-associated myelopathy/tropical spastic paraparesis (HAM/TSP) is chronic inflammatory disease of the spinal cord characterized by perivascular lymphocytic cuffing and parenchymal lymphocytic infiltration. In this study using flow cytometry, we have investigated the T-cell receptor (TCR) Vbeta repertoire of peripheral blood T lymphocytes in 8 HAM/TSP patients, 10 HTLV-1 infected healthy carriers, and 11 uninfected healthy controls to determine if there is a biased usage of TCR Vbeta. We found that TCR Vbeta7.2 was under-utilized and Vbeta12 was over-utilized in CD4+ T cells of HTLV-1 infected individuals compared with healthy uninfected controls, whereas there were no such differences in CD8+ T cells. Comparison of Vbeta repertoire changes before and after interferon-alpha (IFN-alpha) treatment for HAM/TSP revealed that one out of five patients showed dramatic decrease of specific Vbeta in CD8+ T cells. Our results suggest that dominant Vbeta subpopulations in CD4+ T cells evolved associated with chronic HTLV-1 infection, and IFN-alpha treatment for HAM/TSP does not induce a specific pattern of TCR Vbeta changes.     

Pathogenesis of chronic progressive myelopathy associated with human T-cell lymphotropic virus type I

Human T-cell lymphotropic virus type I (HTLV-I) induces a chronic demyelinating disease known as HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP). While only 0.25% of HTLV-I-infected individuals develop HAM/TSP, the mechanisms responsible for the progression of an HTLV-I carrier state to clinical disease are not clear. In particular, no specific sequence differences have been found between HTLV-I recovered from HAM patients and HTLV-I-infected carriers. Since CD4 T cells are the major reservoir of the virus, at least three hypotheses implicating CD4 T cells directly or indirectly have been proposed: 1) The cytotoxic hypothesis predicts that activated and HTLV-I-infected CD4 T cells migrate to the CNS and infect resident cells. Cytotoxic CD8 T cells may then recognize viral antigens on HTLV-I-infected CNS cells causing a cellularly mediated cytotoxic demyelination. 2) The autoimmune hypothesis predicts that either (a) virally reactive T cells cross-react with a CNS antigen, or (b) random infection of CD4 T cells eventually results in the infection of CNS-autoreactive CD4 T cells that, by virtue of the productive HTLV-I infection, become activated, expand and migrate to the CNS, where they encounter their antigen. This results in a specific immune response and demyelination, as is known to occur in experimental autoimmune encephalomyelitis. 3) The bystander damage hypothesis does not implicate a specific response against CNS cells. Instead this hypothesis suggests that the presence of IFN-γ-secreting HTLV-I-infected CD4 T cells and their recognition by virally specific CD8 T cells in the CNS induce microglia to secrete cytokines, such as TNF-α, which may be toxic for the myelin.     

Widespread active inflammatory lesions in a case of HTLV-I-associated myelopathy lasting 29 years

An autopsy case of human T-lymphotropic virus type I (HTLV-I)-associated myelopathy (HAM) of 29 years duration is reported. The patient had no history of surgery or blood transfusion and likely contracted HTLV-I sexually while traveling in an endemic area. At age 45, the patient began to experience gait disturbance; he later developed spastic tetraparesis. Autopsy revealed marked gross spinal cord atrophy, particularly in the middle to lower thoracic levels. Myelin and axonal degeneration were identified predominantly in the middle to lower thoracic spinal cord, extending into the medulla oblongata and lumbar cord. Inflammatory infiltrates of mononuclear cells were diffuse in the white and gray matter of the spinal cord and medulla oblongata, particularly in perivascular areas. These infiltrates were also observed in perivascular areas of the pons, midbrain, cerebellum, and cerebrum. More than half of the infiltrating cells were positive for the pan-T cell marker UCHL-1, but some were positive for the B cell marker SL-26. There were far more CD8-positive cells than CD4-positive cells in the spinal parenchyma and perivascular areas. Neurons in the anterior horn, Clarkes column, and intermediolateral column were relatively well preserved. Active chronic inflammation was indicated. Despite the 29-year history of HAM, the presence of an active inflammatory reaction is surprising. We discuss possible modulation of the histopathological manifestations of HAM by corticosteroid therapy.     

Pathogenetic significance of HTLV-I infection and immune surveillance in HAM]

HTLV-I proviral DNA load is significantly increased in HTLV-I associated myelopathy (HAM) compared with asymptomatic HTLV-I seropositive carriers (SPC), and this spread of HTLV-I infection seems to be critically important in the pathogenesis of HAM. Thus, in this report, cellular immune surveillance against HTLV-I was reviewed. (1) MHC class I-restricted cytotoxic T lymphocytes (CTL) activities are detected in peripheral blood mononuclear cells (PBMC) of HAM. CTL release various proinflammatory and cytotoxic cytokines, chemokines, and proteases. Since CTL are also found in the spinal lesions of HAM, CTL may contribute to the tissue damage. In spontaneous proliferation of PBMC in HAM, CD 4/CD 8 is decreased due to the proliferation of CD 8 + CTL. CD 4/CD 8 is inversely correlated with the clinical severity of HAM. Collectively, CTL may be involved in the pathogenesis of HAM. (2) Activity and subsets of natural killer (NK) cells are lower in HTLV-I-seropositive individuals. Moreover, NK have only a weak cytotoxicity against HTLV-I infected cells. (3) Antibody-dependent cell-mediated cytotoxicity (ADCC) are impaired in HAM compared with SPC due to the suppressed effector cell activity. Since ADCC effectively lyse HTLV-I infected cells in vitro, the impaired ADCC may in part allow the spread of HTLV-I infection in HAM, and potentiation of ADCC may have an anti-HTLV-I therapeutic effect.     

HTLV-I specific IFN-gamma+ CD8+ lymphocytes correlate with the proviral load in peripheral blood of infected individuals

Human T lymphotropic virus type I (HTLV-I)-associated myelopathy/tropical spastic paraparesis (HAM/TSP) is an inflammatory neurological disease caused by HTLV-I infection. It has been shown that HAM/TSP patients have high proviral loads and an extraordinarily high frequency of circulating CD8 + cytotoxic T lymphocytes specific for HTLV-I in their peripheral blood when compared to asymptomatic HTLV-I carriers (AC). We have previously described an intracellular cytokine detection assay, in which interferon-gamma (IFN-gamma) + CD8 + lymphocytes are specific for HTLV-I in infected individuals. Here, we have established a competitive polymerase chain reaction assay to measure the proviral load of patients and investigate a potential relationship between proviral load and virus-specific CD8 + lymphocytes. Genomic DNA was extracted from peripheral blood lymphocytes (PBL) from eight HAM/TSP patients and seven AC for the measurement of HTLV-I measuring proviral loads. The same PBL were analyzed for intracellular IFN-gamma expression by flow cytometry. In the HAM/TSP patients and AC, the average proviral loads were 34,482 and 9784 copy/microg DNA (P = 0.021), and the average of IFN-gamma + CD8 + lymphocytes in total PBL were 1.47 and 0.08% (P = 0.001), respectively. It was confirmed that HAM/TSP patients have both high proviral loads and increased HTLV-I-specific CD8 + lymphocytes. Furthermore, we found a positive correlation between both factors in the patients with HAM/TSP (P = 0.044) but not in the AC (P = 0.508). These findings suggest that the high number of HTLV-I-specific lymphocytes may result from the increased proviral load in HAM/TSP patients.     

Analysis of HTLV-I proviral load in 202 HAM/TSP patients and 243 asymptomatic HTLV-I carriers: high proviral load strongly predisposes to HAM/TSP

In order to examine the effect of HTLV-I proviral load on the pathogenesis of HAM/TSP, we measured the HTLV-I proviral load in peripheral blood mononuclear cells (PBMC) from a large number of HAM/TSP patients and asymptomatic HTLV-I carriers. To measure the proviral load, we used an accurate and reproducible quantitative PCR method using a dual-labeled fluorogenic probe (ABI PRISM 7700 Sequence Detection System). The mean +/- standard error of mean (s.e.m.) HTLV-I proviral copy number per 1 x 10(4) PBMC was 798 +/- 51 (median 544) in 202 HAM/TSP patients; 120 +/- 17 (median 34) in 200 non HAM-related (general) asymptomatic HTLV-I carriers (RC); and 496 +/- 82 (median 321) in 43 asymptomatic HTLV-I carriers genetically related to HAM/TSP patients (FA). The prevalence of HAM/TSP rises exponentially with log (proviral load) once the proviral load exceeds 1% PBMC. The HTLV-I proviral load of female patients with HAM/TSP was significantly higher than that of male patients, however there was no significant difference in proviral load between sexes in RC. There was a significant correlation between the proviral load and the concentration of neopterin in CSF of HAM/TSP patients. These results indicate that the HTLV-I proviral load in PBMC may be related to the inflammatory process in the spinal cord lesion. The increased proviral load in FA suggests the existence of genetic factors contributing to the replication of HTLV-I in vivo.     

Expression of adhesion molecules and monocyte chemoattractant protein-1 (MCP-1) in the spinal cord lesions in HTLV-I-associated myelopathy

Leukocyte adhesion molecules to endothelium plays an important role in the pathogenesis of inflammatory diseases, including HTLV-I-associated myelopathy (HAM)/tropical spastic paraparesis (TSP). To help define the role of adhesion molecules in HAM/TSP, we studied the expression of lymphocyte function-associated antigen-1 (LFA-1), Mac-1, very late antigen-4 (VLA-4), Sialyl Lewis^x (SLex), intercelluar adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), endothelial leukocyte adhesion molecule-1 (ELAM-1) and monocyte chemoattractant protein-1 (MCP-1) in the spinal cord lesions of HAM/TSP. The results indicate that spinal cord lesions of HAM/TSP have greater VCAM-1 expression on endothelium compared with those of controls. Infiltrating mononuclear cells, especially perivascular lesions, expressed VLA-4. Although the expression of ICAM-1 in the spinal cords was not distinctive between HAM/TSP and controls, infiltrating mononulcear cells in the spinal cords of HAM/TSP strongly expressed LFA-1 and Mac-1. ELAM-1 was expressed on endothelium in the inactive-chronic lesions from three of five HAM/TSP, but was not detectable in the spinal cords of controls. SLex reaction was detectable on occasional perivascular cells in the spinal cord of HAM/TSP, but not in those of controls. MCP-1 was detectable on perivascular infiltrating cells and vascular endothelium in active-chronic lesions. This study suggests that VLA-4/VCAM-1 interaction may play an important role for lymphocyte migration into the central nervous system (CNS), and MCP-1 may also be involved in inflammatory cell recruitment to the CNS in HAM/TSP. Received: 4 September 1995 / Revised, accepted: 23 October 1995     

Enhanced adherence of endothelial cells to blood mononuclear cells in HAM/TSP]

We investigated the adhesion of blood mononuclear cells (MNC) isolated from patients with HTLV-1-associated myelopathy (HAM/TSP). MNC from HAM/TSP patients were significantly more adherent to activated endothelial monolayers than MNC from non-HAM/TSP (controls and HTLV-1 carriers) subjects. Blocking studies demonstrated that the adhesion molecules VLA-4 (CD49d), ICAM-1 (CD54), and L-selectin (CD62L) all contributed to increased binding. However, anti-ICAM-1 antibody was the most efficient in inhibiting binding HAM/TSP patients MNC to activated endothelial cells. Expression on MNC of molecules involved in adhesion was also studied by flow cytometry in HAM/TSP patients, HTLV-1 carriers, and healthy control subjects after two days culture without any mitogen. In HAM/TSP patients, L-selectin expression on CD4+ and CD8+ subsets was lower than in controls; interestingly, HAM/TSP patients had lower percentage of CD4+ subset expressing L-selectin than HTLV-1 carriers. The percentage of CD4+ and CD8+ cells expressing VLA-4 was found to be similar to controls in both HAM/TSP patients and HTLV-1 carriers. Following two days in culture without mitogen, the percentage of T cells expressing ICAM-1 increased in HAM/TSP and carriers, but not in controls. This study provides information regarding trans-endothelial migration of MNC across the blood brain barrier in HAM/TSP and suggests ICAM-1 and its counterpart molecule LAF-1 are involved in massive infiltration of lymphocytes observed in the spinal cord.     

Characterization of a unique T-cell clone established from a patient with HAM/TSP which recognized HTLV-I-infected T-cell antigens as well as spinal cord tissue antigens

Five T-cell clones reactive to autologous HTLV-I-infected T-cells (KODA-TV) were established from peripheral blood lymphocytes of a HAM/TSP patient (KODA) by the limiting dilution method. All the clones showed CD3⁺, CD4⁺ and CD25⁺ surface markers and expressed αβ⁺ T-cell receptors to recognize KODA-TV antigens. One of the five T-cell clones (KODA-408) was infected with HTLV-I but the remaining four clones (KODA-400, 404, 405 and 409) were free of HTLV-I infection. KODA-408 recognized both KODA-TV and spinal cord antigens, the latter being extracted from autopsy tissues of a HTLV-I seronegative donor. KODA-408 did not recognize either alloantigens of peripheral blood mononuclear cells extracted from unrelated HTLV-I seronegative donors or purified human myelin basic protein. KODA-408 T-cell clone produced a considerable amount of TNF-aβ, IFN-γ, and IL-6. The CDR3 motif of KODA-408 T-cell receptor showed a unique sequence CASSAGQS of Vβ8-Dβ-Jβ1.5. These results indicated that HAM/TSP CD4⁺ T-cells were polyclonally activated by HTLV-I infection and antigenic stimulation. The T-cell repertoire shaped by HTLV-I infection included T-cells which recognized HTLV-I-infected T-cell antigens as well as spinal cord antigen in particular.     

MRI-pathological correlate of brain lesions in a necropsy case of HTLV-I associated myelopathy.

A postmortem case of HTLV-I associated myelopathy (HAM)/tropical spastic paraparesis (TSP) with a history of remission and exacerbation of neurological signs and symptoms, resembling those of multiple sclerosis is reported. MRI analysis revealed lesions in the periventricular white matter in addition to atrophy of the thoracic spinal cord, characteristic of HAM/TSP. The cerebral periventricular areas consisted of ill-defined paucity of myelin sheaths with astrocytic gliosis and hyaline thickening of blood vessels. The poorly demarcated white matter lesions found in both brain and spinal cord were different from plaques found in multiple sclerosis. It is suggested that, in some cases of HAM/TSP, inflammatory lesions that destroy myelin can involve not only the spinal cord but also the cerebral periventricular white matter.     

Presence of HTLV-I proviral DNA in central nervous system of patients with HTLV-I-associated myelopathy.

The polymerase chain reaction (PCR) method was used to determine the presence and amount of human T lymphotropic virus type I (HTLV-I) proviral DNA in central nervous system (CNS) tissue obtained at autopsy from 6 patients with HTLV-I-associated myelopathy (HAM)/tropical spastic paraparesis (TSP), 1 patient with adult T-cell leukemia (ATL) and CNS infiltration of leukemic cells, and 9 control subjects with other neurological disorders. HTLV-I pX and env but not pol DNA were detected in CNS tissue from 5 of 6 patients with HAM/TSP. The ATL samples were positive for pX, env, and pol DNA by PCR. None of the control samples was consistently positive for HTLV-I by PCR, but all showed positive bands on beta-globin PCR. Quantitative PCR combined with histological studies showed no correlation between HTLV-I proviral DNA amounts and extent of perivascular mononuclear cell infiltration in the HAM/TSP CNS. Also, the amounts of pX and probably env DNA were greater in the HAM/TSP samples than in the ATL sample, although the extent of mononuclear cell infiltration was far less in the HAM/TSP samples than in the ATL sample. Therefore, in addition to infiltrating mononuclear cells, constituent cells of the CNS may harbor the HTLV-I genome, at least in the pX and env regions, in patients with HAM/TSP.