Ablation of Stat3 by siRNA alters gene expression profiles in JEG-3 cells: a systems biology approach.

Control of inflammation at the maternal-fetal interface is a critical element in mammalian pregnancy. Previous work from our laboratory has shown that Stat3 may be a placental mediator involved in maintaining immunologic homeostasis at the maternal-fetal interface. The aim of the current study is to further elucidate the role of Stat3 in response to inflammation. As ablation of Stat3 in mice results in embryonic lethality, we evaluated the role of Stat3 in vitro using an siRNA approach. Trophoblast-like JEG-3 cells were transfected with an siRNA construct specific to Stat3. Experimental and control cells were exposed to conditioned medium from PHA-activated peripheral blood mononuclear cells and incubated for 45 min. Cells were then collected and RNA isolated for transcriptional profiling using human Affymetrix U133 plus 2.0 GeneChips. Differences in gene expression between control and Stat3-ablated cells were evaluated using conventional statistical methods. Fifty-two genes were detected as up-regulated in conditioned medium in both mock transfected and in Stat3 siRNA transfected JEG-3 cells. Two genes (EPAS1 and RASGEF1B) were up-regulated only in cells transfected with negative control siRNA, while 36 genes were up-regulated only in cells transfected with Stat3 siRNA. Sixty genes were differentially expressed between Stat3 siRNA transfected cells relative to mock transfected cells both in basal and conditioned medium. These included 31 genes up-regulated with Stat3 siRNA transfected cells and 29 genes down-regulated with Stat3 siRNA. Eleven genes were differentially expressed only in basal medium. Seven of these were up-regulated in the presence of Stat3 siRNA and four were down-regulated. Nine genes were differentially expressed only in conditioned medium. Six of these were up-regulated and three down-regulated in the presence of Stat3 siRNA. Off-target effects were excluded in a second set of experiments in which Stat3 mRNA was targeted at a different site and quantitative real-time PCR performed on selected genes derived from the microarray analysis. While some of the genes that showed differential expression between Stat3-ablated cells and mock transfected controls were genes typically associated with immune response (e.g., CCR7 and IRAK1), in silico modeling of the microarray data also revealed complex networks of signaling molecules and molecules associated with cellular metabolism previously seen in transcription factor ablation in model organisms. We conclude thus: Stat3 controls a specific gene set in trophoblast-like JEG-3 cells. While some differentially expressed genes and in silico models of their functions are consistent with the hypothesis that Stat3 plays a role in regulating inflammation, Stat3-mediated response to inflammation appears to also involve complex homeostatic adaptations of a non-immunologic nature.

Multiplex serum cytokine monitoring as a prognostic tool in rheumatoid arthritis.

OBJECTIVE: Early optimized therapy of rheumatoid arthritis (RA) results in improved outcomes. The initiation of optimized therapy is hindered by the difficulty of early diagnosis and the limitations of current disease activity and therapeutic response assessment tools. Identifying patients requiring early combination DMARD/biologic therapy is currently a significant clinical challenge given the lack of definitive prognostic criteria. Since cytokines are soluble intracellular signaling molecules that modulate disease pathology in RA, we tested the recent conjecture that en mass serum cyto-kine measurement and monitoring will provide a useful tool for effective therapeutic management in RA. METHODS: We assayed the levels of 16 serum cytokines in 18 RA patients treated prospectively with methotrexate and from 18 unaffected controls. Specific mechanistic aspects of inflammatory pathology in the periphery could be discerned on a patient-specific basis from patients' serum cytokine profiles, information that may aid in the design of anti-cytokine biologic therapy. A serum Cytokine Activity Index (CAI) was also created using multi-variant analysis methods. RESULTS: Distinct cytokines were significantly elevated in RA patients relative to controls, and three distinct clusters with correlations to disease activity were identified. The Cytokine Activity Index correlated well with the therapeutic res-ponse; responders and non-responders in this cohort were distinguishable as early as one month post initiation of methotrexate therapy, well before clinical assessments of response are commonly completed. CONCLUSION: Clinical assessment tools could be derived from this approach that may provide a means to continually track patients, allowing intervention strategies to be better evaluated on a patient-specific basis and to identify residual cytokine activity that could be used to guide combination therapy.

Genome-scale assessment of molecular pathology in systemic autoimmune diseases using microarray technology: a potential breakthrough diagnostic and individualized therapy-design tool.

Systemic autoimmune rheumatic diseases are of complex aetiology, characterized by an intricate interplay of various factors. A myriad of genes lies behind the heterogeneous manifestations of these diseases, and the overexpression and repression of particular genes form a specific gene-expression profile (genetic fingerprints) that is characteristic to the given disease phenotype. Besides the description of various cell types by using gene-expression profiling, the data should be directly applicable to the design of individual therapeutic protocols for patients suffering from various autoimmune diseases. In this review, we summarize the gene-expression profile, various genetic signatures of different autoimmune diseases and give an overview on the possible interpretations of the data. The application of recent breakthroughs in high-throughput molecular profiling technologies, such as microarray technology has been the basis for a revolution in biomedical research, as well as diagnostics and pharmaceutical development. It is easy to envision a day when personalized medicine, which is the diagnosis and treatment of a given patient with agents and procedures tailored to that patient's genetics, physiology and pathology, will become the standard of care.

A genome-scale assessment of peripheral blood B-cell molecular homeostasis in patients with rheumatoid arthritis.

OBJECTIVE: While rheumatoid arthritis (RA) is considered a prototypical autoimmune disease, the specific roles of B-cells in RA pathogenesis is not fully delineated. METHODS: We performed microarray expression profiling of peripheral blood B-cells from RA patients and controls. Data were analysed using differential gene expression analysis and 'gene networking' analysis (characterizing clusters of functionally inter-relelated genes) to identify both regulatory genes and the pathways in which they participate. Results were confirmed by quantitative real-time polymerase chain reaction and by measuring the levels of 10 serum cytokines involved in the pathways identified. RESULTS: Genes regulating and effecting the cell-cycle, proliferation, apoptosis, autoimmunity, cytokine networks, angiogenesis and neuro-immune regulation were differentially expressed in RA B-cells. Moreover, the serum levels of several soluble factors that modulate these pathways, including IL-1beta, IL-5, IL-6, IL-10, IL-12p40, IL-17 and VEGF were significantly increased in this cohort of RA patients. CONCLUSIONS: These results outline aspects of the multifaceted role B-cells play in RA pathogenesis in which immune dysregulation in RA modulates B-cell biology and thereby contributes to the induction and perpetuation of a pathogenic humoral immune response.

Transcriptional modulation of TCR, Notch and Wnt signaling pathways in SEB-anergized CD4+ T cells.

Gene expression changes in CD4 + Vbeta8+ T cells energized by in vivo exposure to staphylococcal enterotoxin B (SEB) bacterial superantigen compared to CD4 + Vbeta8+ non-energic T cells were assessed using DNA microarrays containing 5184 murine complementary DNAs. Anergy in splenic T cells of SEB-immunized BALB/c mice was verified by dramatically reduced proliferative capacity and an 8 x overexpression of GRAIL mRNA in CD4 + Vbeta8+ T cells taken from mice 7 days after injection. At an Associative t-test threshold of P<0.0005, 96 genes were overexpressed or detected only in anergic T cells, while 256 genes were suppressed or not detected in anergic T cells. Six of eight differential expressions tested using real-time quantitative PCR were validated. Message for B-Raf was detected only in non-anergic cells, while expression of the TCR signaling modulator Slap (Src-like adapter protein) and the TCR zeta-chain specific phosphatase Ptpn3 was enhanced. Modulation of multiple genes suggests downregulation of Wnt/beta-catenin signaling and enhanced Notch signaling in the anergic cells. Consistent with previous reports in a non-superantigen in vivo anergy model, mRNA for CD18 and the transcription factor Satb1 (special AT-rich-binding protein 1) was increased in SEB-energized T cells. This is the first report of global transcriptional changes in CD4+ T cells made anergic by superantigen exposure.

5alpha-Androstane-3alpha,17beta-diol activates pathway that resembles the epidermal growth factor responsive pathways in stimulating human prostate cancer LNCaP cell proliferation.

5alpha-Androstane-3alpha,17beta-diol (3alpha-diol) is considered to have no androgenic effects in androgen target organs unless it is oxidized to 5alpha-dihydrotestosterone (5alpha-DHT). We used microarray and bioinformatics to identify and compare 3alpha-diol and 5alpha-DHT responsive gene in human prostate LNCaP cells. Through a procedure called 'hypervariable determination', a similar set of 30 responsive genes involving signal transduction, transcription regulation, and cell proliferation were selected in 5alpha-DHT-, 3alpha-diol-, and epidermal growth factor (EGF)-treated samples. F-means cluster and networking procedures showed that the responsive pattern of these genes was more closely related between 3alpha-diol and EGF than between 5alpha-DHT and 3alpha-diol treatments. We conclude that 3alpha-diol is capable of stimulating prostate cell proliferation by eliciting EGF-like pathway in conjunction with androgen receptor pathway.

Production and characterization of human monoclonal anti-idiotype antibodies to anti-dsDNA antibodies.

Anti-dsDNA autoantibodies are the hallmark of systemic lupus erythematosus (SLE) and frequently correlate with disease activity. In this study we report the isolation and characterization of human anti-Id monoclonal antibody fragments as single-chain Fv fragments (scFv) against anti-dsDNA antibody. The anti-Id monoclonal antibodies, specific for anti-dsDNA antibodies, have been cloned from phage display antibody scFv libraries derived from a patient with SLE. The V gene repertoires were derived from the RNA obtained from the B cells of an SLE patient with anti-Ro/SSA and anti-La/SSB antibodies. Affinity-purified anti-dsDNA antibodies were used for selection of bacterial clones producing specific scFv antibody fragments against anti-dsDNA antibodies and little reactivity with normal IgG and other IgG antibodies by ELISA. The anti-Id antibody recognizes a public idiotope that is broadly cross-reactive with polyclonal and monoclonal anti-dsDNA antibodies. This binding was largely inhibited by dsDNA antigen. The anti-Id antibody inhibited anti-dsDNA binding to dsDNA antigen in immunoassays and in the Crithidia luciliae assay. The anti-Id scFv antibody fragments derived from human genes could modulate the pathogenicity of anti-dsDNA autoantibodies and may have therapeutic implications in SLE. They may also be used as probes in studies of the structure of the idiotype.

Polymorphisms of coagulation factor XIII subunit A and risk of nonfatal hemorrhagic stroke in young white women.

BACKGROUND AND PURPOSE: Although family studies have suggested a genetic influence on hemorrhagic stroke, the underlying genetic risk factors remain poorly defined. Coagulation factor XIII, which is involved in hemostasis, fibrinolysis, vascular remodeling, and tissue repair, represents a candidate gene for hemorrhagic stroke. We assessed the potential role of 3 factor XIII subunit A coding-sequence polymorphisms, along with a promoter polymorphism of plasminogen activator inhibitor-1 (PAI-1, which is also involved in fibrin stabilization and vascular remodeling), in young white women with hemorrhagic stroke. METHODS: Genotype analysis for factor XIII subunit A Val34Leu, Tyr204Phe, and Pro564Leu and for PAI-1 -675 4G/5G was performed in a population-based case-control study of 42 white women aged <45 years with nonfatal hemorrhagic stroke and 345 demographically similar control subjects. RESULTS: Compared with the respective homozygous wild-type genotypes, the Tyr204/Phe204 genotypes (age-adjusted odds ratio [OR] 2.9, 95% 95% CI 1.1 to 7.5) and the Leu564/Leu564 genotype (OR 4.3, 95% CI 1.4 to 13.7) were each associated with an increased risk of nonfatal hemorrhagic stroke. The risk estimate associated with the Phe204 variant was highest in women with subarachnoid hemorrhage and in nonsmokers, whereas the risk estimate of the Leu564/Leu564 genotype was highest in women with intracerebral hemorrhage and in smokers. Women who carried either the Phe204 allele or the Leu564/Leu564 genotype in combination with the PAI-1 5G/5G genotype had a nearly 20-fold increased risk of hemorrhagic stroke (OR 18.9, 95% CI 3.8 to 95.1). CONCLUSIONS: Our findings suggest that the Phe204 and Leu564 variants of coagulation factor XIII may be markers for genetic susceptibility to hemorrhagic stroke in women aged <45 years.

Autoantibody to the leucine zipper region of 52 kDa Ro/SSA binds native 60 kDa Ro/SSA: identification of a tertiary epitope with components from 60 kDa Ro/SSA and 52 kDa Ro/SSA.

Anti-Ro (or SSA) is found in the sera of patients with autoimmune rheumatic illnesses. All patients with anti-Ro defined by precipitation bind a 60 000 Da antigen (60 kDa Ro), whereas some patients also bind a 52 000 Da molecule (52 kDa Ro). In general, antibody binding is directed against native 60 kDa Ro and denatured 52 kDa Ro. The mechanism by which anti-52 kDa Ro arises in the setting of anti-60 kDa Ro is unknown. Conflicting data exist as to the existence of a physical interaction between the two proteins in cells and as to cross-reacting antibodies. Antibodies were affinity purified from a peptide within the leucine zipper region of 52 kDa Ro. These purified antibodies binding the 197-207 peptide from 52 kDa Ro (anti-52LZ) bound native 60 kDa Ro as well as denatured 52 kDa Ro. In addition, anti-52LZ also bound up to four regions from the sequence of 60 kDa Ro and a single conformational epitope of 60 kDa Ro. Thus, these primary sites represent components of the tertiary epitope. We hypothesized that if this was the case, these peptides making up a tertiary epitope would show molecular interaction. In fact, peptides from 60 kDa Ro have a molecular interaction with the 52 kDa Ro peptide as well as full-length 52 kDa Ro when assessed by surface plasmon resonance. The leucine-zipper region peptide from 52 kDa Ro bound three of the four peptides from 60 kDa Ro. These data suggest that these two molecular species, 60 and 52 kDa Ro, form a conformational epitope. This relationship may explain why anti-52 kDa Ro is found in association with anti-60 kDa Ro.

The Kozak sequence polymorphism of platelet glycoprotein Ibalpha and risk of nonfatal myocardial infarction and nonfatal stroke in young women.

Several platelet membrane glycoprotein polymorphisms have been identified as potential risk factors for cardiovascular disease. Recently a nucleotide -5T/C dimorphism in the translation initiation site (Kozak sequence) of the platelet glycoprotein Ibalpha (GPIbalpha) gene was associated with increased platelet surface levels of the GPIb-IX-V receptor complex. The role of this GPIbalpha Kozak sequence polymorphism in the occurrence of arterial thrombotic disease is unknown. We performed genotype analysis of the Kozak sequence polymorphism of GPIbalpha in a population-based study of 18- to 44-year-old women with nonfatal myocardial infarction (MI) (n = 78), nonfatal stroke (n = 106), and 384 demographically similar female control subjects. Analysis of -5T/C genotypes revealed that at least one copy of the C allele was present in 14.1% of MI cases, 23.6% of stroke cases, and 23.7% of controls. The age-adjusted odds ratio for MI in women carrying at least one copy of the C allele was 0.53 (95% confidence interval [CI] 0.27-1.05). The age-adjusted odds ratio for stroke in women carrying at least one copy of the C allele was 0.99 (95% CI 0.59-1.65). Analyses stratified by stroke type (ischemic, hemorrhagic) yielded similar results. In conclusion, young women carrying the C allele of the Kozak sequence polymorphism of GPIbalpha are not at increased risk of MI or stroke. Paradoxically, the C allele may even be associated with a reduced risk of MI in this population. This finding requires further study.

Chimeric del20q in a case of chronic neutrophilic leukemia.

Chronic neutrophilic leukemia (CNL) is a rare myeloproliferative disorder in which recurrent abnormalities of chromosome 20 have been reported. We report the case of a 76-year-old woman with CNL with partial deletion of the long arm of chromosome 20 in a subset of bone marrow metaphases, suggesting coexistence of a clonal stem cell disorder and normal hematopoiesis. Review of the literature suggests that such mosaicism is common in CNL, possibly accounting for the favorable prognosis observed in many patients with this disorder.

Characterization of DNA binding properties and sequence specificity of the human 52 kDa Ro/SS-A (Ro52) zinc finger protein.

The Ro52 protein is an autoantigen in Sjögren's Syndrome and systemic lupus erythematosus. Although its function is not yet known, sequence similarities to other proteins suggest that it binds to DNA. In this study, the hypothesis that Ro52 recognizes particular nucleic acid sequences was tested. Ro52 bound to double stranded but not single stranded DNA. 1,10-Phenanthroline, a chelater of zinc, was found to inhibit this interaction, suggesting that the zinc fingers of Ro52 are functional. DNA sequences were selected from an oligonucleotide library by binding to Ro52 followed by amplification by Taq DNA polymerase in order to characterize the DNA sequence-binding motif for this protein. These studies support the hypothesis that Ro52 is functionally a member of a family of zinc finger proteins, many of which are known to bind to DNA or regulate gene expression. We speculate that Ro52 functions as a transcription factor, and that its disregulation may have important consequences in the expression or susceptibility of certain autoimmune diseases.

The association of anti-Ro52 autoantibodies with myositis and scleroderma autoantibodies.

The presence of autoantibodies to the Ro52 protein in sera from patients with idiopathic inflammatory myopathies has recently been reported. These antibodies were found predominately in sera with the myositis-specific autoantibody anti-histidyl-tRNA synthetase (anti-Jo-1). In this report, we analysed sera from 216 patients to determine whether anti-Ro52 antibodies are associated with myositis autoantibodies other than anti-Jo-1. These included sera containing antibodies that recognize threonyl- or alanyl-tRNA synthetases, Mi-2, PM-Scl, signal recognition particle (SRP), as well as the systemic sclerosis-related antibodies anti-topoisomerase I (Scl-70) and anti-centromere. A high proportion of sera that contain anti-aminoacyl-tRNA synthetase antibodies, anti-SRP, or anti-PM-Scl antibodies were found to contain antibodies to the Ro52 protein. In contrast, in sera containing anti-Mi-2, anti-Scl-70 or anti-centromere antibodies, anti-Ro52 antibodies were absent or occurred infrequently. In addition, only one serum from 41 rheumatoid arthritis patients was positive for anti-Ro52 autoantibodies. These data indicate that anti-Ro52 antibodies are produced in particular subsets of myositis patients, and are not limited to sera with anti-Jo-1 antibodies.

Effect of immunoglobulin variable region structure on C3b and C4b deposition.

Many of the biological activities of immunoglobulins, including interaction with the complement system, are attributed to the structure of the heavy chain constant domains. However, previous studies indicated that immune complexes formed with independently derived isotype-matched pairs of monoclonal antibodies vary with respect to their capacity to activate complement and to serve as targets for C3b and C4b deposition. The goal of the present study was to provide a structural basis for explaining how variable domains influence C3b and C4b deposition on immunoglobulins. Heavy and light chain variable domains from a pair of IgG2a antibodies previously shown to differ in terms of complement activation and C3b and C4b deposition were cloned and sequenced. The two clones utilize distinct heavy and light variable region genes and the translated amino acid sequence reveals several residues that could serve as potential targets for complement deposition which differs between the two antibodies. Molecular modeling suggests that many of the relevant differences between the two antibodies are located in solvent exposed portions of the heavy and light chain variable domains and that some of the relevant sites are located within the complementarity determining regions. Differences in antibody affinity do not provide an explanation for the previously observed role of variable domains on interactions with the complement system. These data suggest that sequence variations within solvent-exposed variable domain residues may play a key role in C3b and C4b deposition on immunoglobulins.

Lupus autoantibodies to double-stranded DNA cross-react with ribosomal protein S1.

One cDNA clone (G7) was isolated from a lambda gt11 human liver cDNA library by the reaction with a serum containing anti-dsDNA Abs and was ligated into pGEX-1 lambda T vector. All the 10 SLE sera with anti-dsDNA, 2 samples of human monoclonal anti-dsDNA (33.C9 and 33.H11), and 2 affinity-purified anti-dsDNA Abs recognized the glutathione S-transferase fusion protein expressed by G7 (G7-FP). Ab binding to the recombinant protein expressed by G7 (G7-RP) and to G7-FP was inhibited completely by calf thymus dsDNA. The cDNA was 1314 nucleotides in length and contained an open reading frame encoding 352 amino acids. However, it seems to be a partial length cDNA because the affinity-purified Ab from G7-FP recognized only a 104-kDa protein on Western blot using MOLT4 cell extract. The nucleotide sequence of G7 was homologous (99% identity) to a cDNA encoding human ribosomal protein (r-protein) S1 homologue mRNA. The encoded protein contains repeating residues as a feature of r-proteins S1. Cytoplasmic and nucleolar staining of 33.H11 on indirect immunofluorescence (IF) using HEP 2 cells was inhibited by both G7-RP and calf thymus dsDNA. On ELISA, 33.H11 had a higher affinity for G7-RP than for DNA while 33.C9 had a higher affinity for DNA than for G7-RP and binds nuclei on IF. We conclude that G7 encodes a portion of human r-protein S1 and anti-dsDNA Abs cross-react with this protein.

Molecular analysis of a major antigenic region of the 240-kD protein of Mi-2 autoantigen.

Anti-Mi-2 autoantibody is strongly associated with dermatomyositis and found in sera of 20% of patients. Mi-2 antigen contains at least eight components and previous evidence suggested that the 240-kD protein was the antigenic component for at least some sera. In this study, anti-M-2 patient sera were used to screen human thymocyte and HeLa cell lambda gt11 expression libraries, and two clones from each had plaques specifically reactive with anti-Mi-2 sera. Studies with affinity-purified antibody supported the identification of the clones. All of 44 anti-Mi-2 sera reacted with the plaques, but none of 44 control sera reacted significantly. The cDNAs were identical, and full sequencing of one revealed an open reading frame spanning a 1,054-bp insert. Rescreening the library with the cDNA yielded a 1,589-bp cDNA that continued the open reading frame. The Mi-2 cDNA hybridized to a single 7.5-8.0 kb mRNA of HeLa cells, by Northern blot. Rabbit antiserum directed at a portion of the cDNA product reacted with HeLa 240-kD Mi-2 protein. The sequence was notable for four potential zinc-fingers and several charged regions. The protein encoded by the cDNA produced in vitro reacted with only one of five of the Mi-2 sera. These findings indicate that the Mi-2 240 kD is a novel protein that is antigenic for all Mi-2 sera, and strongly suggests that a major common epitope is conformational in nature.

Immunogenetics of epitopes of the carboxyl terminus of the human 60-kD Ro autoantigen.

Systemic lupus erythematosus is associated with the presence of autoantibodies which bind several ribonucleoproteins, including Ro (or SS-A). We have explored the relationship of the HLA-DQ and T cell receptor alleles in patients producing autoantibodies binding the 13-kD carboxyl terminus fragment of the 60-kD Ro and with autoantibodies binding a peptide epitope within this fragment (amino acid residues 480-494). Antibodies binding the 13-kD fragment are more likely to be found in the sera of patients with particular DQA1 and DQB1 alleles, while antibodies binding the epitope at 480-494 are found almost exclusively in the sera of patients with a Bg/II 9.8-kb polymorphism of the T cell receptor beta gene. Meanwhile, in these same patient sera the level of autoantibodies binding the complete 60-kD Ro particle is associated with a distinct pattern of alleles at these same immunoregulatory loci. These data demonstrate that component parts of autoantibody responses may be under genetic control which can be distinguished from the HLA associations characteristic of the response to the intact, complete autoantigen.

Expression and DNA binding of the human 52 kDa Ro/SSA autoantigen.

The 52 kDa Ro/SSA protein is an intracellular autoantigen that is frequently recognized by antibodies in sera of patients with systemic lupus erythematosus or Sjogren's syndrome. While the function of this molecule is not known, zinc finger and leucine zipper motifs have been identified in its predicted amino acid sequence which suggest that it may interact with nucleic acids. To test this hypothesis, the human gene which encodes this protein was cloned in a baculovirus and expressed in Spodoptera frugipoda cells. Extracts from these infected insect cells were used as a source of protein for this study. The protein is similar in size and antigenicity to that expressed in human cells. This protein binds to DNA at physiological temperature and is eluted with high concentrations of sodium chloride. Striking similarities were found between the sequence in, and adjacent to, the nucleic acid-binding motifs of 52 kDa Ro/SSA and a growing family of zinc finger proteins which have been shown to bind to DNA or regulate gene expression. The findings presented here place this protein structurally and functionally in this family and demonstrate a biochemical assay which can be used to study its function.

The location of a disease-associated polymorphism and genomic structure of the human 52-kDa Ro/SSA locus (SSA1).

Sera from approximately 30% of patients with systemic lupus erythematosus (SLE) contain high titers of autoantibodies that bind to the 52-kDa Ro/SSA protein. We previously detected polymorphisms in the 52-kDa Ro/SSA gene (SSA1) with restriction enzymes, one of which is strongly associated with the presence of SLE (P < 0.0005) in African Americans. A higher disease frequency and more severe forms of the disease are commonly noted among these female patients. To determine the location and nature of this polymorphism, we obtained two clones that span 8.5 kb of the 52-kDa Ro/SSA locus including its upstream regulatory region. Six exons were identified, and their nucleotide sequences plus adjacent noncoding regions were determined. No differences were found between these exons and the coding region of one of the reported cDNAs. The disease-associated polymorphic site suggested by a restriction enzyme map and confirmed by DNA amplification and nucleotide sequencing was present upstream of exon 1. This polymorphism may be a genetic marker for a disease-related variation in the coding region for the protein or in the upstream regulatory region of this gene. Although this RFLP is present in Japanese, it is not associated with lupus in this race.