OBJECTIVE A high seroprevalence of HTLV-I in female Sjögren’s syndrome (SS) patients has been reported in Nagasaki, Japan, an area that is heavily endemic for HTLV-I infection. Salivary IgA class antibodies to HTLV-I were common among HTLV-I seropositive patients with SS. This study was undertaken to elucidate the pathogenesis of SS caused by HTLV-I infection.
METHODS The clinical features and histological findings of SS and the prevalence of serum autoantibodies in 10 patients with HTLV-I associated myelopathy (HAM) who were consecutively admitted into Nagasaki University School of Medicine, were compared with those of 20 HTLV-I seropositive and 20 HTLV-I seronegative patients with SS.
RESULTS Ocular and oral manifestations of SS were commonly detected in HAM patients. These patients also had extraglandular manifestations including recurrent uveitis, arthropathy, interstitial pneumonitis, Raynaud’s phenomenon, and inflammatory bowel disease. All patients with HAM histologically showed a mononuclear cell infiltration in the labial salivary grands. Six of 10 patients had a mononuclear cell infiltration with a focus score of 1 or greater. According to the preliminary criteria for SS proposed by the European Community, definitive SS was diagnosed in six patients and probable SS in two patients. Serum γ globulin and IgG values were increased in HAM patients. Patients with HAM had lower prevalence of rheumatoid factor, anti-nuclear antibody, and anti-SS-A (Ro) antibody than those of HTLV-I seropositive and HTLV-I seronegative SS patients. However, there was no significant difference in the prevalence of these antibodies among HAM patients with definitive SS, HTLV-I seropositive and HTLV-I seronegative SS patients. The CD3+CD4+ T cells preferentially infiltrated into the salivary glands in HAM patients as well as the salivary glands of patients with HTLV-I seropositive and seronegative patients. It seems probable that peripheral blood mononuclear cells from HAM patients preferentially infiltrated into the salivary glands, and that these cells produced the autoantibodies as well as anti-HTLV-I antibody.
CONCLUSION The results strongly support the idea that HTLV-I is involved in the pathogenesis of the disease in a subset of patients with SS in endemic areas.
- Sjögren’s syndrome
- HTLV-I associated myelopathy.
Statistics from Altmetric.com
The T lymphotropic virus type I (HTLV-I) is a human retrovirus originally identified as the aetiological agent of adult T cell leukemia/lymphoma (ATL).1 2 It is closely associated with a slowly progressive myelopathy known as HTLV-I associated myelopathy/tropical spastic paraparesis (HAM/TSP).3 4While the mechanisms by which HTLV-I causes HAM/TSP are not well understood, a high viral load and spontaneous proliferation of infected T cells are observed in patients with HAM/TSP compared with patients with ATL.5-7 Sjögren’s syndrome (SS), which is considered an autoimmune disease, is characterised by chronic lymphocytic infiltration of salivary and lacrimal glands and occasional presence of serum autoantibodies.8 It may occur alone (primary SS), or in association with a variety of connective tissue diseases and autoimmune disorders (secondary SS). Although several aetiological mechanisms have been proposed, the pathogenesis of SS remains unknown. Viruses have long been considered potential aetiopathological agents9 and retroviruses have been recently considered as possible agents in inducing SS.10 We have recently reported a high seroprevalence of HTLV-I in female SS patients, compared with female blood donors, residing in Nagasaki, a city located in the most western part of Japan, which is heavily endemic for HTLV-I infection.11 12 Furthermore, salivary IgA class antibodies to HTLV-I were common among HTLV-I seropositive patients with SS.12 These findings suggest a close relation between HTLV-I infection and SS in endemic areas, and that HTLV-I might induce SS. In this study, we examined the SS-like clinical features in HAM/TSP patients. We also compared the prevalence of serum autoantibodies and phenotypic markers of mononuclear cells infiltrating the labial salivary glands among HAM/TSP patients, HTLV-I seropositive and seronegative patients.
We examined 10 patients with HAM (three men and seven women) aged 60.7 (15.3) years (mean (SD), range: 32-73). They represented all HAM patients that were consecutively admitted into Nagasaki University Hospital between January 1995 to January 1996. The diagnosis of HAM was based on the criteria described by Osame et al.13 Two groups served as control subjects. The first group included 20 patients with SS (one male and 19 females) who were positive for anti-HTLV-I antibody (HTLV-I seropositive SS patients) of a mean age of 56.4 (9.5) years (range: 38-70). The second group consisted of 20 age and sex matched patients with SS (one male and 19 females) who were negative for anti-HTLV-I antibody (HTLV-I seronegative SS patients) of a mean age of 50.4 (13.5) years (range: 23-75). All patients fulfilled the preliminary criteria for diagnosis of SS as defined by the European Community.14 Patients with secondary SS complicating other autoimmune diseases, such as systemic lupus erythematosus, rheumatoid arthritis, and progressive systemic sclerosis, were not included in the study. Informed consent was obtained from all participating patients and the study was conducted in accordance with the human experimentation guidelines of our institution.
Serum samples were screened for antibodies to HTLV-I using the enzyme linked immunosorbent assay (ELISA; Eitest-ATL kit, Eisai, Tokyo, Japan) and the particle agglutination assay (Serodia-ATL kit, Fujirebio, Tokyo). Anti-HTLV-I antibody was confirmed using a commercially available immunoblotting kit (Problot-HTLV-I, Fuji Rebio). In this test kit blotted HTLV-I antigens include three gag proteins (p19, p24, and p53) and 1 env protein (gp46). According to the criteria proposed by the World Health Organisation,15 reactions with both HTLV-I gp46 and more than one of the gag proteins were scored positive. Antinuclear antibodies were detected by an indirect immunofluorescence procedure using HEp-2 cells (Fluoro Hep Ana test, Medical & Biological Laboratories (MBL), Nagoya, Japan). Antibodies to SS-A (Ro) and SS-B (La) antigens were determined by ELISA (Mesacup SS-A/Ro test and Mesacup SS-B/La test, MBL). The presence of the rheumatoid factor in serum was determined by agglutination of red blood cells coated with rabbit IgG (RAHA) and ELISA of cells precoated with human IgG.
PREPARATION OF MINOR LABIAL SALIVARY GLANDS
Minor labial salivary glands were obtained from the mucosa of the lower lip, 1.0 cm lateral from midline. Tissue sections were transferred in 4% paraformaldehyde immediately after biopsy and dehydrated by 10, 15, and 20% sucrose. The tissues were then snap frozen in liquid nitrogen and stored at −80°C.
The phenotypes of mononuclear cells infiltrating the minor salivary glands were determined using several monoclonal antibodies, including anti-CD3 (Leu-4), anti-CD4 (Leu 3a+3b), anti-CD8 (Leu-2b), and anti-CD20 monoclonal antibodies. These were purchased from Becton Dickinson Monoclonal Center (Mountain View, CA). Their production and characterisation have been described elsewhere.18 Gin 14 (monoclonal mouse antibody reactive with the HTLV-I core proteins p19 and p28) was purchased from Fujirebio Inc, (Tokyo).
Tissue sections (5-7 μm thick) were cut and mounted on glass slides precoated with aminopropyltriethoxysilane. The sections were stained by the labelled-streptavidin-biotin method (HISTOFINE staining kit, Nichirei Co, Tokyo), as described previously.19Briefly, endogenous peroxidase was inactivated by immersing the section in 3% H2O2 solution. The sections were then incubated with non-immune rabbit IgG, followed by incubation with mouse monoclonal antibodies (CD3, CD4, CD8, CD20, GIN 14) in a humid chamber for 60 minutes at room temperature. The sections were treated with biotinylated rabbit antimouse IgG for 12 minutes. After washing, the sections were incubated with peroxidase conjugated streptavidin. The colour was developed using 3.3’-diaminobenzidin and H2O2. The slides were counterstained with haematoxylin. Negative control sections were treated with normal rabbit serum. The percentage of positive cells was subjectively divided into four grades by an observer blind to the type of tissue. The different grades included (−): 0%; (±): 0-10%; (+): 10-50%; and (++): >50%.
Data were expressed as mean (SD). Statistical analysis was performed using the χ2 test. p Values less than 0.05 were considered statistically significant.
CLINICAL CHARACTERISTICS OF HAM PATIENTS
Table 1 summarises the clinical features of patients with HAM. Eight had both spastic paraparesis and involvement of the urinary bladder. Two other patients had either spastic paraparesis or involvement of the bladder. Western blot analysis demonstrated that all serum samples were reactive with both HTLV-I gp46 and more than one of the gag proteins. The mean age at onset of clinical symptoms was 44.3 (20.0) years. The mean duration of the disease was 14.4 (14.3) years. Two female patients had a history of blood transfusion, while three patients had a family history of HAM.
COMPLICATIONS AND SEROLOGICAL EXAMINATIONS OF HAM PATIENTS
As shown in table 2, a positive history of recurrent uveitis was identified in two of 10 HAM patients. Three patients had arthropathy involving the knee, foot, finger, and wrist joints. One female patient had interstitial pneumonitis and Raynaud’s phenomenon, while HAM was complicated in one male patient by inflammatory bowel disease and uveitis. Leucopenia was detected in two patients (cases 3 and 8, table2). A pronounced increase in IgG and IgA concentrations was present in six and three patients, respectively.
COMPLICATIONS OF SJÖGREN’S SYNDROME IN HAM PATIENTS
We studied the clinical features according to the preliminary criteria for the classification of SS defined by the European Community14 (table 3). Ocular symptoms (dry eyes) and oral symptoms (dry mouth) were present in seven and seven patients, respectively. Nine patients had objective evidence of ocular involvement (more than one of the two tests including Schimer’s and Rose bengal tests). The presence of at least one focus (50) of mononuclear cells in 4 mm2 of glandular tissue was considered to represent a positive sign. These histopathological features were identified in six patients with HAM. The other four patients who were histopathologically negative had only a mild form of mononuclear cell infiltration. Two of six patients who were examined by sialography were positive. Antibodies to SS-A (Ro), SS-B (La), and nuclear antigens were positive in two, one, and four patients, respectively. Two patients had increased serum titres of rheumatoid factor. On the basis of the preliminary criteria for SS, six patients were classified as definitive SS (cases 1, 3, 6, 7, 8, 9, table 3) while two patients were probable SS (cases 5 and 10). Two patients (cases 2 and 4), who did not fulfil the criteria were younger patients (35 and 32 years), and case 2 had a history of recurrent uveitis.
CLINICAL LABORATORY DATA
We compared the clinical laboratory data among 10 patients with HAM, 20 HTLV-I seropositive, and 20 HTLV-I seronegative SS patients. High concentrations of IgG were found in the serum samples of six (60%) HAM patients, 13 of 19 (68%) seropositive SS patients, and 14 of 20 (70%) seronegative SS patients. The serum concentrations of IgG and total γ globulin were not significant differently among these patients (table 4). Rheumatoid factor was present in two (20%) HAM patients, 11 (57.9%) HTLV-I seropositive patients, and 13 (65%) HTLV-I seronegative patients. Antibodies to nuclear antigen were detected in four (40%) HAM patients, 16 (80%) HTLV-I seropositive patients, and 18 (90%) HTLV-I seronegative patients. Antibodies to SS-A (Ro) antigen were also detected in two (20%) HAM patients, in 12 (63%) seropositive patients, and in 15 (75%) seronegative patients. The prevalences of rheumatoid factor, antinuclear antibodies, and anti-SS-A (Ro) antibodies in HAM patients were significantly lower than those in HTLV-I seropositive and seronegative SS patients. The prevalence of these antibodies in HTLV-I seropositive patients was similar to that in HTLV-I seronegative SS patients. Anti-SS-B (La) antibodies were also found in one (10%) HAM patient, two (11%) seropositive, and two (10%) seronegative patients. However, rheumatoid factor, antinuclear antibody, anti-SS-A (Ro) antibody, and anti-SS-B (La) antibody were found in two (33.3%), four (67%), two (33.3%), and one (17.7%) of six HAM patients with definitive SS, respectively. These data indicated that the prevalences of rheumatoid factor, antinuclear antibody, and anti-SS-A (Ro) antibody in HAM patients with definitive SS were similar to those of HTLV-I seropositive and seronegative patients.
IMMUNOHISTOPATHOLOGICAL FINDINGS IN MINOR SALIVARY GLANDS
Phenotypic markers of infiltrating mononuclear cells in minor salivary glands from six HAM patients were studied immunohistopathologically. Figure 1 shows phenotypic markers of infiltrating mononuclear cells from a representative patient (case 9). The CD3+CD4+ cells were the predominant cells infiltrating the minor salivary glands (fig 1A, B). A small number of infiltrated CD8+ and CD20+ cells were also present (fig 1C, D). We also examined the minor salivary glands of three HTLV-I seropositive and 10 HTLV-I seronegative SS patients. As shown in table 5, there was no significant difference in phenotypic markers of infiltrating mononuclear cells among these groups. Furthermore, MT-2 cells were stained with GIN14 monoclonal antibody, but staining was not detected in all tissue samples of HAM patients, HTLV-I seropositive, and seronegative SS patients.
The aetiology of SS has not yet been established, but it is probably multifactorial, representing a complex interaction between genetic, hormonal, and environmental factors.16Retroviruses have been implicated as causative agents in SS. Of particular interest in humans are HTLV-I and HIV-I.10 To investigate whether HTLV-I is a causative agent in SS, we analysed the incidence, clinical, and histological features of SS in a group of HAM/tropical spastic paraparesis patients.
Patients with HAM exhibited clinical features of extraglandular manifestations, including recurrent uveitis, arthropathy, and interstitial pneumonitis. In this regard, Mochizuki et al 17 reported that patients with uveitis had a high seroprevalence for HTLV-I throughout all age groups. Chronic inflammatory arthropathy18 and T lymphocyte alveolitis19 are known to be present in patients with HAM/tropical spastic paraparesis. Furthermore, Vernantet al 20 reported an association between SS and HTLV-I associated tropical spastic paraparesis in five natives of the West Indies island of Martinique. Although only one of the five patients complained of xerophthalmia and xerostomia, all had mononuclear cell infiltration in minor salivary glands.
Histological examination of the minor salivary glands in the present study showed mononuclear cell infiltration in all HAM patients. Using a system of T cell transendothelial migration on collagen gels, we have previously shown that peripheral blood T cells from HAM patients have a greater capacity to migrate through endothelial cells than those from HTLV-I negative healthy subjects.21 From the above findings, it seems probable that peripheral blood mononuclear cells from HAM patients preferentially migrate into the salivary glands as well as the central nervous system.
We also investigated whether SS found in HAM patients was distinct from HTLV-I seropositive and HTLV-I seronegative SS. The prevalences of rheumatoid factor, antinuclear antibody, and anti-SS-A(Ro) antibody were not significantly different in these autoantibodies among HAM patients with definitive SS, HTLV-I seropositive, and seronegative SS patients. We also previously reported that IgA class anti-HTLV-I antibodies are commonly detectable in the saliva of HTLV-I seropositive patients with SS.12 Local synthesis of IgA and IgG antibodies to HTLV-I in HTLV-I seropositive SS was also confirmed by Bélec et al.22 These results suggest that mononuclear cells infiltrated into salivary glands may secrete the autoantibodies such as antinuclear antibodies and anti-SS-A(Ro) antibody as well as antibodies to HTLV-I. The precise mechanism, however, by which HTLV-I infected T cells induce the production of autoantibodies has yet to be clarified.
Recently, Umehara et al 23 reported that CD4+ cells, CD8+ cells, and macrophages infiltrated active chronic inflammatory lesions in spinal cords of HAM patients with a short duration of illness. In contrast, CD8+ cells, rather than CD4+, were the predominant cell type in inactive lesions of HAM patients with long duration of illness. The present immunohistochemical findings showed that CD4+ cells preferentially infiltrated minor salivary glands of HAM and HTLV-I seropositive SS patients. Furthermore, in HTLV-I seronegative SS, CD4+ cells were the predominant cells among infiltrated mononuclear cells, as has been described previously by Coll et al. 24
It will be interesting to identify the exact cell type infected with HTLV-I in the salivary gland. We have detected HTLV-I proviral DNA in all biopsy specimens of minor salivary glands of HTLV-I seropositive and HAM patients by the polymerase chain reaction method (data not shown), but could not detect the HTLV-I gag proteins by GIN14 monoclonal antibody reactive to p19 and p24. T cells are thought to be the natural target cells for HTLV-I infection25 and HTLV-I proviral DNA can be detected in lymphocytes from saliva of HTLV-I infected patients.26 Infiltration of HTLV-I infected T cells into tissues and an immunological reaction to HTLV-I may contribute to the development of exocrinopathy.
In this study, HAM patients had a high prevalence of SS. There are many causes of infiltrative disease of the salivary and lacrimal glands. Some of the patients with HAM developed sicca complaints, but had no clinical manifestations of seronegative spondyloarthritis or reactive arthritis such as sacroiliitis, spondylitis, and enthesopathy. More recently, HIV, and hepatitis C have been identified to have latency in these glands associated with sicca symptoms.27 The sicca syndrome with these infection may be thought to be distinct from the clinical entity of primary SS. However, both HAM patients with definitive SS and HTLV-I seropositive SS patients had a high frequency of autoantibodies, the same as HTLV-I seronegative SS (primary SS) patients. These HTLV-I seropositive SS patients did not serologically discriminate from HTLV-I seronegative SS patients. These results strongly supported the idea that HTLV-I is involved in the pathogenesis of the disease in a subset of patients with SS in endemic areas.
Recently, Renjifo et al 28 reported that a mutation in Tax nucleotide is associated with tropical spastic paraparesis/HAM disease outcome independent of the geographical origin of the patients. Furthermore, it was also recently reported that the predicted protein sequence of Tax is significantly more variable in proviruses isolated from healthy HTLV-I carriers than in those from patients with tropical spastic paraparesis/HAM.29 Whether the Tax protein sequences are different between tropical spastic paraparesis/HAM and HTLV-I associated SS remain to be determined. These studies are currently in progress in our laboratory.
Supported in part by a grant in aid (05670426) from the Ministry of Education, Science, Sport and Culture, Japan.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.