Anti-dsDNA antibodies as a classification criterion and a diagnostic marker for systemic lupus erythematosus: critical remarks.

Antibodies to mammalian dsDNA have, for decades, been linked to systemic lupus erythematosus (SLE) and particularly to its most serious complication, lupus nephritis. This canonical view derives from studies on its strong association with disease. The dogma was particularly settled when the antibody was included in the classification criteria for SLE that developed during the 1970s, most prominently in the 1982 American College of Rheumatology (ACR), and recently in The Systemic Lupus International Collaborating Clinics (SLICC) classification criteria. There are several problems to be discussed before the anti-dsDNA antibody can be accepted without further distinction as a criterion to classify SLE. Old and contemporary knowledge make it clear that an anti-dsDNA antibody is not a unifying term. It embraces antibodies with a wide spectrum of fine molecular specificities, antibodies that are produced transiently in context of infections and persistently in the context of true autoimmunity, and also includes anti-dsDNA antibodies that have the potential to bind chromatin (accessible DNA structures) and not (specificity for DNA structures that are embedded in chromatin and therefore unaccessible for the antibodies). This critical review summarizes this knowledge and questions whether or not an anti-dsDNA antibody, as simply that, can be used to classify SLE.

Galectin-3 binding protein links circulating microparticles with electron dense glomerular deposits in lupus nephritis.

OBJECTIVE: A high level of galectin-3-binding protein (G3BP) appears to distinguish circulating cell-derived microparticles in systemic lupus erythematosus (SLE). The aim of this study is to characterize the population of G3BP-positive microparticles from SLE patients compared to healthy controls, explore putative clinical correlates, and examine if G3BP is present in immune complex deposits in kidney biopsies from patients with lupus nephritis. METHODS: Numbers of annexin V-binding and G3BP-exposing plasma microparticles from 56 SLE patients and 36 healthy controls were determined by flow cytometry. Quantitation of microparticle-associated G3BP, C1q and immunoglobulins was obtained by liquid chromatography tandem mass spectrometry (LC-MS/MS). Correlations between microparticle-G3BP data and clinical parameters were analyzed. Co-localization of G3BP with in vivo-bound IgG was examined in kidney biopsies from one non-SLE control and from patients with class IV (n = 2) and class V (n = 1) lupus nephritis using co-localization immune electron microscopy. RESULTS: Microparticle-G3BP, microparticle-C1q and microparticle-immunoglobulins were significantly (P < 0.01) increased in SLE patients by LC-MS/MS. Three G3BP-exposing microparticle populations could be discerned by flow cytometry, including two subpopulations that were significantly increased in SLE samples (P = 0.01 and P = 0.0002, respectively). No associations of G3BP-positive microparticles with clinical manifestations or disease activity were found. Immune electron microscopy showed co-localization of G3BP with in vivo-bound IgG in glomerular electron dense immune complex deposits in all lupus nephritis biopsies. CONCLUSIONS: Both circulating microparticle-G3BP numbers as well as G3BP expression are increased in SLE patients corroborating G3BP being a feature of SLE microparticles. By demonstrating G3BP co-localized with deposited immune complexes in lupus nephritis, the study supports cell-derived microparticles as a major autoantigen source and provides a new understanding of the origin of immune complexes occurring in lupus nephritis.

Low diagnostic and predictive value of anti-dsDNA antibodies in unselected patients with recent onset of rheumatic symptoms: results from a long-term follow-up Scandinavian multicentre study.

OBJECTIVES: To verify the diagnostic accuracy of anti-double-stranded DNA (anti-dsDNA) antibodies detected by the Crithidia luciliae immunofluorescence test (CLIFT) in a cohort of unselected patients, referred to a rheumatologist due to recent onset of rheumatic symptoms. METHOD: A total of 1073 consecutive patients were screened for anti-nuclear antibodies (ANAs). Serum samples from 292 ANA-positive and 292 matching ANA-negative patients were tested three times for anti-dsDNA antibodies, using two different CLIFT kits (ImmunoConcepts(®) and Euroimmun(®)). An initial clinical diagnosis was made by rheumatologists unaware of the results. The diagnoses were updated after a median follow-up of 4.8 years. RESULTS: CLIFT was positive at least once in 60 patients but only 23 patients were CLIFT positive in all of the assays. Diagnosis of systemic lupus erythematosus (SLE) was made initially in 65 patients, of whom 24 (37%) were CLIFT positive. Many other diagnoses were observed among the CLIFT-positive patients. Overall, 16 (5.5%) ANA-negative patients were CLIFT positive. After approximately 5 years, the diagnosis of SLE remained unchanged in 63 patients (23 CLIFT positive) and altered in only two (one CLIFT positive). Among the 36 CLIFT-positive patients who were not diagnosed with SLE at study entry, only one developed SLE during the follow-up period. CONCLUSIONS: CLIFT was not reliable as a diagnostic tool in unselected patients with rheumatic symptoms. ANAs were of little value as a screening test before the CLIFT analysis. CLIFT had surprisingly low positive predictive value (PPV) for the diagnosis of SLE despite its high specificity. For non-SLE patients, being CLIFT positive poses little risk of developing SLE within 5 years.

Human autoantibodies that react with both cell nuclei and plasma membranes display specificity for the octamer of histones H2A, H2B, H3, and H4 in high salt.

Sera of some patients with systemic lupus erythematosus and related diseases contain a polyclonal antibody population (cross-reactive antinuclear antibodies [X-ANA]) that react specifically with both core mononucleosomes and plasma membranes of viable nucleated cells. Native mononucleosomes and nucleosome cores assembled from long DNA and the inner histones were indistinguishable in terms of inhibition of binding of X-ANA to nuclei of tissue sections and to polynucleosomes on the walls of plastic tubes. In contrast, mononucleosomes selectively depleted of histones H2A and H2B did not inhibit these reactions. A method was developed for isolation of X-ANA from serum that took advantage of the dual specificity of these antibodies. Immunosedimentation in sucrose density gradients revealed that 125I-labeled Fab' fragments of highly pure X-ANA formed complexes with the inner histones H2A, H2B, H3, and H4 in 2 M NaCL, but not in 0.15 M salt. These results indicate that X-ANA recognize an epitope of the inner histone in 2 M salt, and that in 0.15 M NaCL this epitope is not formed unless the histones interact with DNA to generate a nucleosome structure. Furthermore, in light of the previous demonstration that the epitope is destroyed by trypsin, it may be localized in the N-terminal region of histone H2A or H2B.

European League Against Rheumatism recommendations for monitoring patients with systemic lupus erythematosus in clinical practice and in observational studies.

OBJECTIVES: To develop recommendations for monitoring patients with systemic lupus erythematosus (SLE) in clinical practice and observational studies and to develop a standardised core set of variables to monitor SLE. METHODS: We followed the European League Against Rheumatism (EULAR) standardised procedures for guideline development. The following techniques were applied: nominal groups, Delphi surveys for prioritisation, small group discussion, systematic literature review and two Delphi rounds to obtain agreement. The panel included rheumatologists, internists, dermatologists, a nephrologist and an expert related to national research agencies. The level of evidence and grading of recommendations were determined according to the Levels of Evidence and Grades of Recommendations of the Oxford Centre for Evidence-Based Medicine. RESULTS: A total of 10 recommendations have been developed, covering the following aspects: patient assessment, cardiovascular risk factors, other risk factors (osteoporosis, cancer), infection risk (screening, vaccination, monitoring), frequency of assessments, laboratory tests, mucocutaneous involvement, kidney monitoring, neuropsychological manifestations and ophthalmology assessment. A 'core set' of minimal variables for the assessment and monitoring of patients with SLE in clinical practice was developed that included some of the recommendations. In addition to the recommendations, indications for specific organ assessments that were viewed as part of good clinical practice were discussed and included in the flow chart. CONCLUSIONS: A set of recommendations for monitoring patients with SLE in routine clinical practice has been developed. The use of a standardised core set to monitor patients with SLE should facilitate clinical practice, as well as the quality control of care for patients with SLE, and the collection and comparison of data in observational studies.

Histones synthesized for use in early development of Xenopus laevis are stored as a complex with antigenic properties similar to those of the octamer core of nucleosomes.

Serum of patients with systemic lupus erythematosus (SLE) contains crossreacting autoantibodies which recognize histones in nucleosomes or when they are induced to form octamers in solution in the presence of 2 M NaCl, but not when they are dissociated free in solution at physiological ionic strength. We have found that histones stored in eggs of Xenopus laevis for use in rapid nuclear synthesis during early development react with this antibody. This reaction has been observed by radioimmunoassay, inhibition of chromatin assembly by the extracts in the presence of antibody, and, in a preliminary result, by identification of a histone-antibody complex bound to protein A-sepharose. Further evidence that the extract antigen corresponds to the stored histone pool comes from sedimentation and charge fractionation experiments where the chromatin assembly activity and antigen (measured by radioimmunoassay) were found to cofractionate. BEcause the extract histones are not bound to DNA, our results suggest that they are stored as a soluble complex in a conformation similar or identical to the octameric core of the nucleosome. Our data suggest that the histones in this complex are bound to an anionic factor or factors which presumably replaces the DNA in shielding the positive charges on the histones.

Experimental expression in mice and spontaneous expression in human SLE of polyomavirus T-antigen. A molecular basis for induction of antibodies to DNA and eukaryotic transcription factors.

We have previously demonstrated that experimental expression of the polyomavirus transcription factor T-antigen has the potential to induce anti-DNA antibodies in mice. Two sets of independent evidences are presented here that demonstrate a biological relevance for this model. First, we describe results demonstrating that mice inoculated with T-antigen-expressing plasmids produced antibodies, not only to T-antigen and DNA, but also to the DNA-binding eukaryotic transcription factors TATA-binding protein (TBP), and to the cAMP-response-element-binding protein (CREB). Secondly, we investigated whether polyomavirus reactivation occurs in SLE patients, and whether antibodies to T-antigen, DNA, and to TBP and CREB are linked to such events. Both within and among these SLE patients, frequent polyomavirus reactivations were observed that could not be explained by certain rearrangements of the noncoding control regions, nor by corticosteroid treatment. Linked to these events, antibodies to T-antigen, DNA, TBP, and CREB were detected, identical to what we observed in mice. Antibodies recognizing double-stranded DNA were confined to patients with frequent polyomavirus reactivations. The results described here indicate that cognate interaction of B cells recognizing DNA or DNA-associated proteins and T cells recognizing T antigen had taken place as a consequence of complex formation between T ag and DNA in vivo in the context of polyomavirus reactivations.

Nucleosomes possess a high affinity for glomerular laminin and collagen IV and bind nephritogenic antibodies in murine lupus-like nephritis.

AIM: Lupus nephritis is closely associated with in vivo autoantibody-binding to glomerular membrane-associated electron-dense structures (EDS). The biochemical nature and cellular origin of EDS are controversial, and definitive characterisation needs to be performed. METHODS: By using the terminal transferase biotin-dUTP nick end-labelling (TUNEL) assay at the electron microscopic level, we have traced extracellular chromatin within the glomerular basement membranes of nephritic (NZBxNZW)F1 mice. The TUNEL assay was subsequently used in combination with standard immune electron microscopy (IEM). To analyse why chromatin particles associate with membranes, we determined the affinity of nucleosomes and DNA for glomerular laminin, collagen IV and the mesangial matrix proteoglycan perlecan by surface plasmon resonance. RESULTS: This intra-assay colocalisation TUNEL IEM demonstrated that autoantibodies fully colocalised with extracellular TUNEL-positive chromatin observed as EDS in glomerular membranes, similar to results obtained by the same technique applied to human lupus nephritis. Most importantly, these data validate the murine variant of lupus nephritis as a model to study origin of extracellular chromatin as a key element in human lupus nephritis. Kinetic analyses demonstrated that nucleosomes had a high affinity for collagen IV and laminin, but not for perlecan. CONCLUSION: Collectively, these results provide firm evidence that dominant target structures for nephritogenic autoantibodies are constituted by TUNEL-positive chromatin associated with glomerular capillary and mesangial matrix membranes at high affinity.

Anti-DNA antibody subpopulations and lupus nephritis.

As a consequence of increased insight into the cellular and molecular mechanisms responsible for induction of B cell and T cell autoimmunity to DNA and nucleosomes, there is an obvious need to reconsider the dogma stating that anti-dsDNA antibodies serve as marker antibodies for SLE and also that anti-dsDNA antibodies per se are responsible for the initiation of lupus nephritis. Given that the potential to produce anti-dsDNA antibodies is an inherent property of the normal immune system and that few anti-DNA antibodies have nephritogenic potential, we must try to solve the problem whether it is avidity for DNA, specificity for unique DNA structures or cross-reactivity with non-DNA molecules, that make such antibodies pathogenic and thus potential markers for SLE and lupus nephritis. In this review, we will summarize contemporary problems related to these questions; (1) try to focus on phenotypic differences with respect to the ability to produce anti-dsDNA antibodies between individuals suffering from SLE and those not belonging to this diagnostic group, and (2) to describe differences between pathogenic and non-pathogenic anti-dsDNA antibodies.

Tumor necrosis factor and interleukin 1 appearance in experimental gram-negative septic shock. The effects of plasma exchange with albumin and plasma infusion.

To study the effect of plasma removal vs plasma administration on the appearance of tumor necrosis factor (TNF) and interleukin 1 in septic shock, 24 anesthetized piglets were inoculated with live Escherichia coli. Plasma exchange with albumin was performed in one group. Fresh-frozen plasma was administered to a second group. A third group served as nontreated controls. Following plasma exchange, a reduction in both TNF and interleukin 1 levels occurred, whereas plasma infusion was followed by a decrease in TNF levels only. No significant differences were observed between the two treated groups with respect to survival or cardiovascular performance, with both being significantly enhanced compared with the controls. High levels of TNF and interleukin 1 correlated with depressed cardiovascular performance in the early phase of the shock. Our results confirm the important role of TNF and interleukin 1 as early mediators of septic shock. However, the benefit of reducing cytokine activity in later stages of septicemia seems to be dubious.

Secretin exerts inhibition of the 8BrcAMP-stimulated 86Rb influx into avian red blood cells.

Avian erythrocytes possess a Na+ and K+ cotransport system which can be inhibited by loop diuretics. The newly discovered diuretic effect of secretin in man led us to study the effect of this hormone on the cotransport system. Secretin caused a 50% inhibition of the 8BrcAMP-stimulated 86Rb influx into red blood cells from goose at a concentration of 8.5 X 10(-6) M, while furosemide and bumetanide caused a 50% inhibition at concentrations of 7 X 10(-6) M and 9 X 10(-8) M, respectively. It is suggested that the diuretic effect of secretin is mediated through an inhibitory or blocking effect on the Na+ and K+ cotransport system.

Development of lupus nephritis is associated with qualitative changes in the glomerular collagen IV matrix composition.

Lupus nephritis is associated with thickening of the glomerular extracellular membranes. Distribution of collagen IV alpha-chains in the glomerular basement membrane in kidneys of lupus-prone B/W mice has been examined in this study. The results are indicative of a qualitative change in the collagen IV matrix occurring around the time of development of proteinuria, with an embryonic alpha1/alpha2 isoform replacing the normal glomerular basement membrane (GBM). These changes mimic alterations seen in Alport syndrome and coincide with an increase in collagenolytic activity within the glomerulus. It has been hypothesized that alterations in collagen matrix synthesis represent compensatory responses to an increase in GBM proteolysis and could represent an important step in the pathogenesis of nephritis through the formation of a dysfunctional glomerular filter. Also, aberrations in the collagen matrix composition could contribute to the deposition of autoantibodies within the glomerulus.

Circulating chromatin-anti-chromatin antibody complexes bind with high affinity to dermo-epidermal structures in murine and human lupus nephritis.

Murine and human lupus nephritis are characterized by glomerular deposits of electron-dense structures (EDS). Dominant components of EDS are chromatin fragments and IgG antibodies. Whether glomerular EDS predispose for similar deposits in skin is unknown. We analysed (i) whether dermo-epidermal immune complex deposits have similar molecular composition as glomerular deposits, (ii) whether chromatin fragments bind dermo-epidermal structures, and (iii) whether deposits in nephritic glomeruli predispose for accumulation of similar deposits in skin. Paired skin and kidney biopsies from nephritic (NZBxNZW)F1 and MRL-lpr/lpr mice and from five patients with lupus nephritis were analysed by immunofluorescence, immune electron microscopy (IEM) and co-localization TUNEL IEM. Affinity of chromatin fragments for membrane structures was determined by surface plasmon resonance. Results demonstrated (i) presence of EDS containing chromatin fragments and IgG in both organs in nephritic patients, (ii) chromatin fragments possessed high affinity for dermo-epidermal laminins and collagens, (iii) glomerular immune complex deposits did not predict similar interstitial deposits in skin, although such complexes were present in capillary lumina in glomeruli and skin of all nephritic individuals. Thus, chromatin-IgG complexes accounting for lupus nephritis seem to reach skin through circulation, but other undetermined factors are required for these complexes to deposit within skin membranes.

Diagnostic impact of contemporary biomarker assays for rheumatoid arthritis.

OBJECTIVE: The impetus towards early treatment for patients with rheumatoid arthritis (RA) requires more reliable disease markers than the non-specific immunoglobulin M (IgM) rheumatoid factor (RF). To determine the accuracy of newer biomarkers for RA testing for antibody against cyclic citrullinated peptides (anti-CCP Ab), IgA- and IgG-RF and cartilage oligomeric matrix protein (COMP) levels, measured by enzyme-linked immunosorbent assay (ELISA), were compared with IgM-RF isotyping. METHODS: Serum samples were investigated in patients with an established diagnosis of RA (n = 54), ankylosing spondylitis (AS) (n = 36), and non-inflammatory conditions (n = 18) (cohort A), and in 234 consecutive outpatients in a blinded fashion (cohort B). Non-parametric analysis of areas under the curve (AUC) of receiver operator characteristics were performed. RESULTS: The presence of anti-CCP Ab had the highest accuracy (96%) in distinguishing RA patients in cohort A and cohort B (accuracy 83%), and in both cohorts combined (accuracy 87%). This was related to the high specificity of anti-CCP Ab for RA (95-96%), even though IgM-RF was the most sensitive test (87-96%). Sensitivity (15-48%) and specificity (66-69%) of COMP as a marker for RA was low. Combining results of anti-CCP Ab and IgM-RF or any of the other assays did not increase the diagnostic accuracy for RA. CONCLUSION: The presence of anti-CCP Ab is the most accurate biomarker for RA in both selected and unselected cohorts, while the COMP assay is not very useful in RA diagnosis. Combining assays for anti-CCP Ab and IgM-RF or IgA-RF does not enhance RA diagnosis.

Glomerular apoptotic nucleosomes are central target structures for nephritogenic antibodies in human SLE nephritis.

Antibodies to double-stranded (dsDNA) are associated with systemic lupus erythematosus (SLE) and directly involved in human lupus nephritis. Information about their glomerular target antigens is inconsistent, and whether availability of target antigens, antibody specificity or avidity are nephritogenic parameters, is not determined. In this study, we analyzed renal tissue from anti-dsDNA antibody-positive lupus patients with nephritis by morphological and immunological assays, including immune electron microscopy (IEM) and colocalization IEM, an EM-based confocal microscopy assay. IEM demonstrated that antibody deposits were confined to electron dense structures (EDS) in glomerular membranes. These autoantibodies colocalized with nucleosome-binding anti-dsDNA/-histone/-transcription factor antibodies. To confirm the colocalization IEM-data, we developed a colocalization terminal deoxynucleotidyl-transferase (TdT) biotin-dUTP nicked end-labeled (TUNEL) IEM assay where extracellular DNA was traced by TdT-mediated introduction of biotinylated nucleotides and autoantibodies by IEM. Results consistently demonstrated that DNA colocalized with autoantibodies in glomerular membrane-associated EDS. The colocalization IEM and colocalization TUNEL IEM assays thus demonstrate that intra-glomerular membrane-associated nucleosomes are targeted by anti-dsDNA autoantibodies in human lupus nephritis. The data provide a new approach to understand basic molecular and immunological processes accounting for antibody-mediated nephritis in human SLE.

Rheumatoid factor by laser nephelometry and Waaler-Rose assay: prognostic value in patients with recent-onset rheumatoid arthritis.

OBJECTIVE: To evaluate the prognostic value of rheumatoid factor (RF), detected in the Waaler-Rose agglutination assay and by nephelometry, in patients with recent-onset rheumatoid arthritis (RA). METHODS: Consecutive patients with new-onset RA between 1993 and 1997 were followed for a median period of 4.7 years. Clinical data at baseline and drug use during the disease course were recorded. Outcome parameters studied were disease process, damage (erosions, joint surgery, extra-articular manifestations, and new co-morbidity), and death. Cut-off levels for RF were >40 IU/mL (nephelometry) and titres 1:160 (Waaler-Rose haemagglutination). RESULTS: RF tests were negative by both methods in 22% of RA patients (RF- group), while 33% were RF positive by nephelometry only (RF+ group) and 45% were positive by Waaler-Rose and nephelometry (RF++ group). Baseline clinical and laboratory findings as well as the number of subsequently used disease-modifying anti-rheumatic drugs (DMARDs), the number of patients starting and the time spent on steroid therapy were similar in the three RF groups. Odd ratios for death (n = 23), erosions (n = 62), and serious extra-articular disease manifestations (EAMs) (n = 13) as well as patient survival, erosion-free or surgery-free survival rates did not differ between the RF groups. Only rheumatoid nodules were more frequent in RF++ patients. CONCLUSION: The baseline presence of RF by either Waaler-Rose or nephelometry was not associated with differences in drug therapy, morbidity other than rheumatoid nodules, or mortality in RA patients in the first 5 years of disease. Being immunoglobulin M (IgM) RF positive thus had little impact on RA patient outcome.

Intrinsic cell membrane antigens recognized by antichromatin autoantibodies. The membrane antigens do not derive from the nucleus.

The main object of this study was to see whether or not chromatin constituents are present in cell membranes. The binding of antinuclear antibodies (ANA) to plasma membranes of leucocytes was studied by using autoantibodies and induced antibodies to different histones and histone peptides, and autoantibodies to double-stranded DNA (dsDNA) in adsorption-elution experiments. Eluates were subsequently tested for antinuclear antibodies by a solid-phase ELISA. No ANA activities in the eluates were observed, except when true cross-reacting ANA were employed in the study. Furthermore, no binding of these antibodies to plasma membranes could be visualized by indirect membrane fluorescence tests. The conclusion of these studies was that freshly isolated viable leucocytes did not carry detectable amounts of ectopic chromatin components at the level of plasma membranes. Thus, it seems fairly unlikely that chromatin autoantibodies in general exert some tissue damage by binding to homologous nuclear antigens associated with plasma membranes in vivo.

Antinuclear antibody screening in this new millennium: farewell to the microscope?

ANA testing by immunofluorescence technique (F-ANA) is nowadays still performed in much the same way as 45 years ago when the test was introduced. Due to its low specificity the F-ANA test has a poor predictive value for systemic autoimmune diseases and in addition has proven difficult to standardise. In the meantime, many of the nuclear and cytoplasmatic auto-antigens, related to specific types of autoimmune disease, have been characterised and can be tested for in specific ELISA assays (E-ANA). These assays are in large part automated and enable the large volume testing required, by the current attitude, to use ANA-testing for its high negative predictive value in the exclusion of systemic autoimmune disease. In addition, E-ANA assays give specific results for clinically relevant autoantibodies, while its test repertoire can be altered at any given time to reflect changes in current thinking on relevant auto-antigens. Thus, we suggest that the unspecific F-ANA test should no longer be considered the gold standard for the detection of clinically relevant autoantibodies.

Interleukin-2, but not interleukin-15, is required to terminate experimentally induced clonal T-cell anergy.

It has been demonstrated that T cells stimulated with nucleosome-polyomavirus T-antigen (self-nonself) complexes, but not nucleosomes, activate autoimmune nucleosome-specific T cells. As these cells may be naïve, such observations do not show that anergic T cells are reactivated. To understand the regulation of autoimmunity, this is important to assess, and this is the focus of this study. T-cell anergy was induced by antigen stimulation in the presence of antibodies to the costimulatory molecules CD80/CD86. Requirements for the reactivation of anergic T cells were analysed by the ability of antigen and interleukin-2 (IL-2) or IL-15 to increase T-cell proliferation and IL-2 transcription. Data demonstrate that stimulation of T cells with T-antigen and anti-CD80/86 antibodies promotes long-lasting clonal T-cell anergy. While T-antigen did not reactivate anergic T cells, proliferation and upregulation of IL-2 gene transcription was initiated by stimulation with antigen, costimulation and IL-2 added to the cultures. Proliferation per se was not sufficient to promote the reactivation of anergic T cells, as both IL-2 and IL-15 induced proliferation, while antigen and IL-2, but not IL-15, upregulated IL-2 mRNA levels. These data demonstrate that the innate immune system and IL-2 are central to the initiation and termination of T-cell anergy.

Mixed form of dermatitis herpetiformis and bullous pemphigoid.

Two cases of bullous disease had clinical and histological features suggestive of both dermatitis herpetiformis (DH) and bullous pemphigoid (BP). Immunofluorescence (IF) studies on both diseased and healthy skin tissue revealed in the one patient an intense linear deposit of IgG and C3 along the dermo-epidermal junction and in the other patient a linear deposit of IgA and C3 along the dermo-epidermal junction. No circulating anti-basement membrane antibodies were found. At the time of investigation in both cases normal jejunal mucosa was observed. Both patients carried the HLA-DRw3 antigen, which has been found associated with DH. One of them responded to treatment with diasone-sodium combined with a gluten-free diet (GFD). Combined treatment with a small dose of prednisone and diasone-sodium in addition to GFD controlled the disease of the other patient. Despite significant differences in typical cases of DH and BP, the diseases do overlap on rare occasions to form the so-called intermediate or mixed form.