Alloantiserum-mediated suppression of histocompatibility-linked Ir-gene-controlled immune responses. Suppressive effects of IgG fragments derived from alloantisera.

Fab, Fc, and F(ab)'(2) fragments were prepared by enzymatic hydrolysis of the IgG fraction of strain 13 antistrain 2 alloantisera. These fragments were not cytotoxic to lymphocytes bearing strain 2 histocompatibility antigens, but the Fab and F(ab)'(2) fragments retained functional combining sites as indicated by their ability to suppress the cytotoxicity mediated by the intact antistrain 2 antibodies. The F(ab)'(2) fragments were much more efficient as inhibitors in this system than the Fab fragments. F(ab)'(2) at 0.06 mg/ml and 0.45 mg/ml Fab produced comparable degrees of suppression. The F(ab)'(2) at 0.06 mg/ml completely suppressed DNP copolymer of L-glutamic acid and L-lysine (GL)-stimulated tritiated thymidine incorporation. The monovalent Fab at 0.45 mg/ml, however, had no significant effect on the in vitro responses to DNP-GL. Addition of the intact alloantisera can be delayed 3 h after initiation of the antigen-stimulated cultures with no loss of suppression. After a delay of 6 h 45% suppression was observed. The requirement for the divalent molecule and the observation that effective suppression of the in vitro responses is still obtained when the alloantiserum is added several hours after initiation of the cultures both suggest that the immunosuppression results from an active process affecting the lymphocyte membrane that renders the cell refractory to the antigenic stimulus.

Histocompatibility specificity.

The lymph node cells from all L-glutamic acid and L-tyrosine (GT) responder random-bred guinea pigs were susceptible to lysis by strain 2 anti-strain 13 isoantisera in the presence of complement. These same antisera were cytolytic for lymph node cells of only some of the GT nonresponder animals. However, after absorption with cells, from a nonresponder guinea pig, susceptible to lysis, the anti-strain 13 antisera were no longer able to lyse cells from any GT nonresponder guinea pigs while retaining a large measure of their cytolytic activity for cells of all GT responder guinea pigs. Thus, at least two major strain 13 histocompatibility specificities are expressed on the cells of random-bred guinea pigs. The genetic locus controlling the expression of only one of those strain 13 histocompatibility specificities is linked to the GT immune response gene.

Specific immune response genes of the guinea pig. II. Relationship between the poly-L-lysine gene and the genes controlling immune responsiveness to copolymers of L-glutamic acid and L-alanine and L-glutamic acid and L-tyrosine in random-bred Hartley guinea pigs.

The ability of guinea pigs to make immune responses to GA, a linear random copolymer of L-glutamic acid and L-alanine, GT, a random linear copolymer of L-glutamic acid and L-tyrosine, and PLL, a linear homopolymer of L-lysine, is controlled by different autosomal dominant genes specific for each of those polymers. We have investigated the relationship between the PLL gene and the GA and GT immune response genes by simultaneously immunizing random-bred Hartley strain guinea pigs with GA and PLL, GT and PLL, or GA and GT. In most Hartley guinea pigs the ability to respond immunologically to GA and to PLL is inherited together; that is, most animals responding to GA respond to PLL and vice versa. However, a few animals respond to either GA or to PLL but not both, demonstrating that the GA and PLL immune response genes are not identical but linked in most Hartley animals. Conversely, when simultaneously immunized with GT and PLL, most Hartley guinea pigs respond to either PLL or GT but not both, indicating that GT and PLL responsiveness tends to segregate away from each other. Thus, the GT and PLL immune response genes also are not inherited independently but, rather, behave as alleles or pseudoalleles. Similar results are observed when Hartley guinea pigs are simultaneously immunized with GA and GT. The ability to respond to GA segregates away from the ability to respond to GT. Our studies demonstrated that the specific immune response genes thus far identified in guinea pigs controlling the ability to respond to GA, GT, and PLL, respectively, are found on the same chromosome. In most Hartley animals, the GA and PLL immune response genes are often linked, i.e. occur on the same chromosome strand, and tend to behave as alleles or pseudoalleles to the GT immune response gene.

Specific immune response genes of the guinea pig. I. Dominant genetic control of immune responsiveness to copolymers of L-glutamic acid and L-alanine and L-glutamic acid and L-tyrosine.

The immunogenicity of three random copolymers of amino acids with L-glutamic acid and L-alanine (GA), L-glutamic acid and L-tyrosine (GT), or L-glutamic acid, L-alanine, and L-tyrosine (GAT), administered in complete Freund's adjuvant, was studied in several inbred and random-bred guinea pig strains. The animals were tested for delayed sensitivity and their sera were assayed for the presence of antibody directed against the immunizing polymer. All of the guinea pigs developing delayed hypersensitivity also had significant antibody levels in their sera. Inbred strain 2 guinea pigs responded to immunization with GA, but failed to form detectable responses to GT. Inbred strain 13 animals, on the other hand, responded to GT, but not to GA. The (2 x 13)F(1) hybrids responded to both GA and GT with both delayed hypersensitivity and circulating antibody. Thus, the ability of these inbred guinea pigs to respond immunologically to GA or GT is controlled by distinct autosomal dominant genes. A variable percentage of random-bred guinea pigs, depending on their source as well as their strain, responded to immunization with GA and with GT. All guinea pigs, both inbred and random bred, responded to immunization with GAT. The ability to respond immunologically to GAT, therefore, does not seem to be under simple genetic control. However, the levels of anti-GAT antibody found in the sera of animals lacking the ability to respond to GA were much lower than those detected in GA responder animals.

Analysis of the defects responsible for the impaired regulation of Epstein-Barr virus-induced B cell proliferation by rheumatoid arthritis lymphocytes. I. Diminished gamma interferon production in response to autologous stimulation.

T cells of patients with rheumatoid arthritis (RA) do not control the rate of B lymphoblast transformation induced by Epstein-Barr virus (EBV) as efficiently as T cells from healthy individuals; thus, lymphoblast cell lines are established more readily in RA lymphocytes in vitro after EBV infection. In the present experiments, we have asked whether this T cell regulation can be reproduced by lymphocytes. We found that normal T cells, activated in allogeneic or autologous mixed leukocyte reactions (MLR), produce lymphokines that inhibit in vitro EBV-induced B cell proliferation. Allogeneic MLR supernatants inhibited EBV-induced DNA synthesis 62 +/- 4% (mean +/- SE) at 10 d post-infection, whereas autologous MLR supernatants suppressed it 50 +/- 3%. RA T cell supernatants produced in an allogeneic MLR suppressed as well as normal T cell supernatants (64 +/- 5% inhibition). In contrast, supernatants from RA autologous MLR had little inhibitory activity. EBV-induced DNA synthesis at 10 d was reduced only 8 +/- 3%, compared with the 50 +/- 3% suppressive activity of normal autologous MLR supernatants. The magnitude of the proliferative responses in the autologous MLR regenerating the lymphokines was similar in the normal and RA populations. After depletion of adherent cells from the RA auto-MLR stimulators, supernatant inhibitory activities increased to normal levels (from 11 +/- 6 [SE] to 52 +/- 6% [SE]). The inhibitory factor involved in the regulation of in vitro EBV infection is a protein with a molecular weight of approximately 50,000. Its activity is eliminated by hearing at 56 degrees C and by exposure to acid at pH 2. The inhibitory activity is blocked by mixing the MLR supernatants with a polyvalent antisera or monoclonal antibodies specific for human gamma interferon. Gamma interferon produced by activating T cells in allo- or auto-MLR can reproduce T cell-mediated regulation of EBV-induced B cell proliferation, and the failure of RA auto-MLR to generate that lymphokine parallels the defective T cell regulation of EBV-induced B cell proliferation characteristic of RA lymphoid cells.

Specific immune response genes of the guinea pig. 3. Linkage of the GA and GT immune response genes to histocompatibility genotypes in inbred guinea pigs.

The ability of guinea pigs to form immune responses specific for each of the random copolymers, L-glutamic acid and L-alanine (GA) and L-glutamic acid and L-tyrosine (GT), is under the control of distinct autosomal dominant genes. By testing for the ability to respond to these copolymers among the progeny from the reciprocal backcross mating of responder (2 x 13)F(1) animals with the appropriate nonresponder parental strain, we have demonstrated that different unigenic autosomal dominant traits control the ability to respond to GA and GT respectively. The data further shows that the GA gene is linked to the poly-L-lysine (PLL) gene and to the locus determining the major strain 2 histocompatibility specificities and that the GT gene is linked to the locus controlling the expression of major strain 13 histocompatibility specificities. Analysis of the inheritance of the GT and PLL genes among the offspring from a mating of responder (2 x 13)F(1) guinea pigs with random-bred guinea pigs unable to respond to GT or PLL demonstrate that these genes segregate away from each other. Thus, the PLL gene and the genes to which it is linked, the GA gene and the major strain 2 histocompatibility locus, behave as alleles or pseudoalleles to the GT gene and the major strain 13 histocompatibility locus.

Specific immune response genes of the guinea pig. V. Influence of the GA and GT immune response genes on the specificity of cellular and humoral immune responses to a terpolymer of L-glutamic acid, L-alanine, and L-tyrosine.

The ability of guinea pigs to make immune responses to the random linear copolymer of L-glutamic acid and L-alanine, GA, and to L-glutamic acid and L-tyrosine, GT, is each controlled by a different immune response gene. On the other hand, the random linear terpolymer of L-glutamic acid, L-alanine, and L-tyrosine, GAT, which contains both GA and GT antigenic determinants, is immunogenic in all guinea pigs. After GAT immunization, all animals develop delayed hvpersensitivity and serum antibody specific for GAT. However, only those guinea pigs possessing the GA immune response gene demonstrate cross-reactive delayed hypersensitivity when challenged with GA. In addition, the anti-GAT antisera produced by those animals having the GA gene contain cross-reacting anti-GA antibodies. The sera from guinea pigs lacking the GA gene have no anti-GA antibody activity. Thus, we have demonstrated that a specific immune response gene controlling responsiveness to a "simple" antigen can determine the specificity of both cellular and humoral immune responses to a more complex antigen.

Very late activation antigens (VLA) are human leukocyte-neuronal crossreactive cell surface antigens.

The antigenic relationship between human neuronal and lymphocyte cell surface antigens has been analyzed using heteroantisera raised against human peripheral blood mononuclear cells (PBMC). The specificities of the crossreactive antigens were examined by immunoprecipitation of 125I-labeled SK-N-SH cultured neuronal cells using rabbit anti-PBMC (RAPBMC) sera and compared to known specificities using mAb. The predominant reactivity of each rabbit antiserum tested against SK-N-SH cells was with three molecules of 130,000, 160,000, and 180,000 Mr. These three chains comigrated with three molecules precipitated with the very late activation antigen (VLA)-specific mAb A-1A5. Sequential precipitations with mAb A-1A5 established that the three RAPBMC-precipitated bands were members of the VLA complex. This was confirmed by two-dimensional PAGE of the RAPBMC and A-1A5 immunoprecipitates, which were indistinguishable from one another. The two-dimensional pattern was more complex than was anticipated from the heterodimeric model of VLA chain association, and suggests an additional 130,000 Mr component of VLA. The three chains of the VLA complex precipitated by RAPBMC or mAb A-1A5 from SK-N-SH neurons closely resembled the VLA pattern present on activated T cells, including the 180,000 Mr activation-specific alpha 1 chain recognized by mAb TS2/7. Normal brain cell membranes also contain VLA molecules that are precipitated by RAPBMC and mAb A-1A5. Thus the VLA complex provides potentially important shared immunogens on human neurons and T cells.

Brain-reactive lymphocytotoxic antibodies in the serum of patients with systemic lupus erythematosus.

Homogenized tissue from the frontal cortex of normal human brains obtained at postmortem examination was used to absorb lymphocytotoxic antibody from the serum of six patients with systemic lupus erythematosus (SLE). Four absorptions of all of the SLE sera with equal volumes of homogenized brain tissue at 4 degrees C depleted their cytotoxic capacity more than 90%. Three of the six sera, however, retained some lymphocytotoxicity despite extensive brain absorption. Absorbed lymphocytotoxic antibodies were eluted from brain tissue absorbents at 37 degrees C. Cytotoxicity of the brain eluates was blocked by antibodies to human IgM (mu-chain specific) but not anti-IgG. The unabsorbed SLE sera, brain-absorbed sera, and brain eluates were equally cytotoxic to T (thymus-derived) and B (bone marrow-derived) cells fractionated from normal human peripheral blood lymphocytes. Thus, the lymphocytotoxic antibodies in SLE serum exhibit no preference for circulating human T cells. An analysis of the clinical records of 40 patients with SLE whose serum cytotoxic capacity had been determined revealed that circulating lymphocytotoxicity is greater in sera of patients with central nervous system (CNS) manifestations than in other SLE patients. This observation suggests a possible role for brain-reactive lymphocytotoxic antibodies in the development of CNS disease in SLE.

Platelet glycoproteins Ia, Ic, and IIa are physicochemically indistinguishable from the very late activation antigens adhesion-related proteins of lymphocytes and other cell types.

The very late activation antigens (VLA) are a subset of the superfamily of cell surface glycoproteins that serve as receptors from extracellular matrix proteins. One or more of the VLA heterodimers are present on T lymphocytes and most other cell types, including platelets. We have used VLA-specific monoclonal antibodies to isolate the reactive platelet membrane molecules. We have identified them as previously characterized platelet surface glycoproteins and have compared them with VLA molecules isolated from lymphocytes and other cells. Utilizing one-dimensional SDS-PAGE, two-dimensional O'Farrell gel electrophoresis, and nonreduced-reduced two-dimensional gel electrophoresis, we show that reduced VLA molecules of platelets are composed of three chains of molecular weights 165,000, 145,000, and 140,000 that possess the physicochemical properties of platelet glycoproteins GPIa, GPIc alpha, and GPIIa. GPIa corresponds to the VLA 165,000 alpha 2-chain, GPIIa corresponds to a 145,000 Mr VLA beta-chain, and GPIc alpha corresponds to a 140,000 Mr VLA alpha-chain. The polypeptide structure of VLA molecules on platelets and lymphocytes are very similar or identical. Platelet proteins GPIa and GPIIa exist as a mixed heterodimer in detergent lysates and correspond with the VLA-2 heterodimer found on activated T lymphocytes and other cell types. The platelet glycoproteins GPIIa and GPIc form a second mixed heterodimer. The mAb A-1A5, which binds to the VLA beta chain, binds to platelet GPIIa and precipitates both the GPIIa-GPIa and GPIIa-GPIc heterodimers, and binds to 4,926 +/- 740 sites per platelet. A VLA-2-specific mAb, 12F1, which binds to the VLA alpha 2-chain reacts with GPIa and immunoprecipitates only the GPIIa-GPIa heterodimer, and binds to 1,842 +/- 449 sites per platelet. The similarity of VLA chains and platelet GPIIa, GPIa, and GPIc molecules suggests that these molecules may have similar functions on various cell types.

Antigenic polymorphism of human very late activation protein-2 (platelet glycoprotein Ia-IIa). Platelet alloantigen Hca.

We have found evidence for a human alloantigenic system on the very late activation protein -2 (VLA-2) heterodimer (platelet GPIa/IIa). Sera from two patients with systemic lupus erythematosus (SLE) contained antibodies that immunoprecipitated surface molecules from platelets and fibroblasts that comigrated on SDS-PAGE and two-dimensional O'Farrell gels with platelet GPIa (VLA-alpha2 chain) and platelet GPIIa (VLA-beta chain). These SLE antibodies were alloreactive as they precipitated VLA molecules from only 5 of 22 normal donors' platelets and did not react with the lupus patients' own platelets, despite the expression of apparently normal amounts of VLA on the donors' cells. Two-dimensional O'Farrell analysis demonstrated no differences in the molecular weight or isoelectric point of GPIa and GPIIa obtained from platelets of alloantibody reactive or unreactive donors. Sequential immunoprecipitation experiments with VLA chain-specific monoclonal antibodies, and the pattern of immunoprecipitation of several different VLA heterodimers demonstrated that the alloantibody-reactive determinant was present on the VLA-2 heterodimer, and not other VLA molecules. Thus, these SLE sera demonstrate a previously unrecognized antigenic polymorphism of the VLA-2 (platelet GPIa/IIa) heterodimer, platelet alloantigen Hca.

Polymorphism of lymphocyte function-associated antigen-1 demonstrated by a lupus patient's alloantiserum.

We have found a human serum, E27, obtained from a multiply transfused patient with systemic lupus erythematosus, which immunoprecipitates the lymphocyte function associated antigen-1 (LFA-1). The immunoprecipitated molecules were identified as the LFA-1 alpha and beta chains by their comigration on SDS-PAGE, two-dimensional SDS-PAGE, and by sequential clearance experiments. Serum E27 did not immunoprecipitate LFA-1 from autologous cells, though LFA-1 molecules were present. In contrast, serum E27 immunoprecipitated LFA-1 from most but not all normal donor lymphocytes. Thus, serum E27 defines two serological phenotypes of LFA-1. 95% of normal individuals tested exhibited the LFA-1 phenotype precipitated by serum E27. Serum E27 appears to be directed at determinants of the LFA-1 alpha-chain and not the beta-chain since it immunoprecipitated LFA-1 molecules but not the Mac-1 molecules. Additional evidence for the alpha chain specificity was provided by immunoprecipitation of mouse-human heterohybridoma cells. LFA-1 was immunoprecipitated by serum E27 from mouse-human heterohybridoma cells expressing the human alpha-chain, not from a hybrid cell line expressing the human beta-chain. Together these findings demonstrate an antigenic polymorphism of the human LFA-1 alpha-chain molecule.

Generation of human cytomegalovirus-specific cytotoxic T-lymphocytes in a short-term culture.

A method to generate human cytomegalovirus (HCMV)-specific CTL (cytotoxic T lymphocytes) from human peripheral blood mononuclear cells is described. This assay is unique in comparison with other methods reported to date, because it only requires a short-term (6 days) coculture of PBM and autologous infected fibroblasts without the addition of exogenous IL-2 (interleukin-2) and nevertheless is sensitive enough to determine HCMV-specific killing in a short (6 h) 51Cr-release assay using autologous HCMV-infected fibroblasts as targets. The virus-specific killing is mediated by CTL of the CD8 phenotype and it can be inhibited by a HLA class I monoclonal antibody. The sensitivity of the assay can be significantly enhanced by pretreating the targets with interferon-gamma (IFN-gamma) prior to infection with HCMV. HCMV-specific 51Cr-release is more than doubled when the IFN-gamma pretreated targets are used. This increase is mostly due to enhanced sensitivity of the fibroblasts to killing mediated by CD8-positive CTL, but some killing can be attributed to CTL of the CD4 phenotype.

Neuropsychiatric disease in Sjögren's syndrome: anti-ribosomal P and anti-neuronal antibodies.

PURPOSE: Patients with Sjögren's syndrome (SS) may develop nonfocal (i.e., psychiatric and/or cognitive dysfunction) as well as focal, neuropsychiatric disease (CNS-SS). Anti-ribosomal P and anti-neuronal antibodies have been associated with nonfocal neuropsychiatric disease in systemic lupus erythematosus (SLE), particularly psychosis and depression. This study examines the spectrum of psychiatric and cognitive dysfunction observed in SS patients with focal, as well as nonfocal, central nervous system (CNS) disease and relates these observations to the presence of serum and cerebrospinal fluid (CSF) anti-ribosomal and anti-neuronal antibodies. PATIENTS AND METHODS: One hundred thirty-one patients--patients with primary SS (n = 91), patients with secondary SS (n = 34), and mothers of infants with neonatal lupus erythematosus (NLE) (n = 6)--were studied. Patients were referred to a large tertiary referral center and the population was highly selected for CNS disease. Patients were evaluated clinically for focal and nonfocal CNS disease. Sera from 131 patients and 34 paired sera/CSF samples were examined by enzyme-linked immunosorbent assay and radioimmunoassay for the presence of anti-ribosomal P and anti-neuronal autoantibodies, respectively. Clinical features were categorized and autoantibody profiles obtained and correlated independently for statistical significance. Data were analyzed using the two-tailed Fisher exact test. RESULTS: Psychiatric or cognitive impairment, usually mild or moderate, occurred in over 80% (63 of 77) of this highly selected population of SS patients, and more than 60% of patients (48 of 77) had both. Anti-ribosomal P antibodies occurred in six (4.6%) patients with SS and related disorders. None of the patients with primary SS had anti-ribosomal P antibodies, whereas they were present in a small number of patients with secondary SS (i.e., 4 of 34 [12%]) and in 2 of 6 mothers of infants with NLE. There was no correlation between nonfocal CNS disease, including psychosis or severe depression, and the presence of anti-ribosomal P antibodies. Paired serum CSF samples from 34 SS patients with active CNS disease, including 6 with psychosis and 5 with severe depression, did not contain either anti-ribosomal P or anti-neuronal antibodies. Anti-ribosomal P and anti-neuronal antibodies were present in a control subset of SLE patients defined serologically by the presence of anti-nDNA antibodies. CONCLUSION: Patients with primary SS associated with CNS disease, including psychosis and depression, do not have serum or CSF autoantibodies to ribosomal P peptide or neuronal antigens, detected by binding to neuroblastoma cells. Thus, autoantibodies associated with nonfocal or diffuse CNS disease in classical SLE (particularly psychosis and depression) are not present in CNS-SS. The observations suggest that nonfocal CNS disease in CNS-SS and CNS-SLE may be mediated by different immunopathologic mechanisms. Potentially, these observations may have diagnostic and therapeutic implications in the management of patients with CNS-SS and patients with CNS-SLE.

Cerebrospinal fluid antibodies to neuronal cells: association with neuropsychiatric manifestations of systemic lupus erythematosus.

The validity of the hypothesis that some of the neuropsychiatric manifestations of systemic lupus erythematosus (SLE) are mediated by the direct effects of antibody binding to neuronal cell membranes is dependent on the demonstration of antineuronal activity within the central nervous system of patients with active central nervous system disease. Using a radiolabelled staphylococcal protein A assay, we tested cerebrospinal fluid from 27 patients with SLE and central nervous system manifestations, and cerebrospinal fluid from 18 additional patients with SLE but free of central nervous system disease for antibody reactive with the cultured human neuronal cell line SK-N-SH. Cerebrospinal fluid from 20 of 27 patients with active lupus central nervous system disease had increased immunoglobulin G (IgG) antineuronal activity compared with cerebrospinal fluid from two of 18 patients with SLE without central nervous system disease. Ninety percent of the patients with psychosis, organic brain syndrome or generalized seizures had increased IgG antineuronal activity as compared with only 25 percent of the patients who presented with hemiparesis or with chorea/hemiballismus. Antineuronal activity per microgram of IgG was concentrated eightfold in the cerebrospinal fluid of patients with active central nervous system disease as compared with the serum activity. Patients with or without active central nervous system disease did not differ significantly in the amount of serum antineuronal binding activity. The results are consistent with the hypothesis that the more diffuse central nervous system manifestations of SLE are a direct result of the interaction of antibody with neuronal cell membranes.

Antibodies reactive with central nervous system antigens.

The pathogenesis of the neuropsychiatric manifestations of systemic lupus erythematosus (SLE) remains an enigma. The observation that many of the lymphocytotoxic antibodies in SLE are also brain-reactive has led to the hypothesis that central nervous system (CNS) lupus, like the autoimmune hematologic manifestations of SLE, is due to the direct effects of autoantibodies to cell membrane antigens. Studies of neuron-reactive antibodies in SLE sera and cerebrospinal fluid support that hypothesis and suggest that the diffuse neuropsychiatric manifestations require the co-existence of serum antibodies to nerve cells and an alteration in the blood-brain barrier that allows those antibodies to enter the CNS.

Neuropsychologic deficits and antineuronal antibodies in pediatric systemic lupus erythematosus.

Pediatric research has been limited regarding the neuropsychologic status in systemic lupus erythematosus (SLE) despite frequent involvement of the central nervous system early in the disease process. SLE is a multisystem autoimmune disorder which often presents with significant neuropsychiatric manifestations including objective neurologic findings and severe psychiatric symptoms. Neuropsychological evaluation provides an objective method for delineating changes in higher cortical functions. We studied 21 pediatric patients who met SLE criteria (12 moderate, 9 mild disease activity) and had no history of CNS damage unrelated to lupus. Mean age was 15.8 years; mean SLE duration at the time of the neuropsychological examination was 2.4 years. Comparison of these SLE patients to a contrast group of 11 patients with juvenile rheumatoid arthritis (JRA) revealed decreased complex problem solving ability for the SLE group. Individual, IQ-adjusted neuropsychological profile analysis yielded a significant difference in the number of specific neuropsychologic deficits for the 2 groups, with impairment rates of 43% for SLE and 18% for JRA. Longer duration of lupus was associated with lower cognitive status. Neuron-reactive antibody studies for IgG and IgM were negative. Results suggest that the prevalence of higher cortical impairment may be as great for younger individuals with lupus as has been documented for older populations.

Anatomic and genetic considerations in the pathogenesis of ankylosing spondylitis.

Investigations have focused on the association between ankylosing spondylitis and HLA-B27; several theories have been offered, including linkage disequilibrium, molecular mimicry involving either autoimmunity or tolerance, and the functional role of the HLA molecule, specifically its role in immune recognition or as a membrane receptor for an infectious agent. The widespread distribution of B27 in the cells of the body versus the restricted anatomic distribution of the pathology of spondyloarthritis remains a major problem in explaining the association. Anterior uveitis, cardiac complications, and inflammation of the gastrointestinal and genitourinary tracts and other characteristic extraskeletal sites of involvement may reflect the basic disease process. Investigations need to focus on potential etiologic factors and their tropism for connective tissue at fibromuscular and fibroskeletal junctions, the cell mediated inflammatory response they precipitate, and their vascular routes to target sites.