Antiguanosine antibodies: a new marker for procainamide-induced systemic lupus erythematosus

Serum antiguanosine antibodies as a marker for SLE disease activity and pathogen potential

BACKGROUND

This article reviews research conducted on the immunogenicity of the nucleosides of DNA, especially guanosine, the most immunologically active nucleoside. Discussed is the relationship between circulating antibodies to guanosine, their potential role in SLE disease activity, the binding properties of monoclonal antiguanosine antibody (4H2) compared to polyclonal antiguanosine antibodies in humans with SLE, cell membrane penetration by these antibodies and their interference with signal transduction possibly related to their binding to mitochondria and their apparent GTPase activity.

METHODS

Enzyme-linked immunosorbent assay methodology was used to show clinical relationships between antiguanosine antibody levels and disease activity in SLE. These results are discussed along with methods of detecting cell penetration by this antibody using special staining techniques, laser-scanning microscope detection of mitochondrial localization, and interference of cAMP and pKA production/activation. Additionally, there is some discussion regarding the assay used to detect enzymatic activity of antiguanosine antibodies.

RESULTS

Enhanced circulating levels of antiguanosine antibodies in patients with SLE correlate closely with SLE disease activity. Other factors are discussed that support the pathogenic potential of these antibodies, including their ability to penetrate lymphocytes, bind to mitochondria, inactivate mitochondrial function, interfere with signal transduction, and their potential enzymatically activity.

CONCLUSIONS

Antiguanosine antibodies correlate with SLE disease activity and may be pathogenically important in SLE by interfering with signal transduction, inactivating mitochondrial and cell function in patients with SLE.

A predictive equation for drug-induced lupus

Objective was to quantify the chemical mechanisms of drug-induced lupus (DIL), and to make the data of different authors comparable between them. I have used seven assumptions as a starting point, based on empirical evidence; then an equation is drawn, describing the relationship between the quantity of the drug taken by the patient(s) and the probability of DIL appearance. The DIL-causing potential of each chemical group is a measurable intrinsic characteristic of that group, and the probability of DIL appearance is directly proportional to the total dosage of the drug in question.

CONCLUSIONS

DIL is a syndrome that can be readily quantified and, as such, could be used as a model for idiopathic Systemic Lupus Erythematosus. On the other hand, the equation drawn here could be used to predict the lupus-inducing potential of many compounds that have not yet been studied. Other authors are encouraged to find similar equations in other drug-induced syndromes.

Circulating antibodies to guanosine in systemic lupus erythematosus: correlation with nephritis and polyserositis by acute and longitudinal analyses

Systemic lupus erythematosus (SLE) is characterized by autoantibodies, including antibodies to the nucleosides of DNA. Guanosine is the most immunogenic nucleoside. In this study serum antiguanosine antibody levels were compared with disease activity, determined by their SLEDI score, in 86 patients with SLE. Sera from these patients were tested, by ELISA, for autoantibodies to guanosine, single-stranded DNA (ssDNA), and double-stranded DNA (dsDNA). Anti-double-stranded DNA levels were also measured by RIA. Resultant values from these assays were correlated with SLE disease activity, and compared with specific features of SLE. The strongest correlation was higher levels of antiguanosine antibodies in patients with active lupus nephritis and polyserositis compared to patients with inactive disease (P < 0.0001). Antiguanosine levels also correlated with arthritis (P < 0.006), CNS lupus (P < 0.005), and hematologic manifestations of SLE (P < 0.002). To test the validity of this association in chronic SLE, serum antiguanosine antibodies were measured in patients with SLE at various phases of disease activity. Twelve patients with SLE had serum samples drawn at active, active-improved, and inactive phases over a 3-7 y period. Differences were significant for serum antiguanosine antibodies in the active group compared to the inactive group (P < 0.05) and the active vs the active-improved group (P < 0.02), unlike those for dsDNA and ssDNA by ELISA or RIA. Antiguanosine antibodies correlated more closely with disease activity in SLE patients in this longitudinal study than either anti-dsDNA or ssDNA antibodies. Thus, antibodies to guanosine correlated as well or better with disease activity than the other anti-DNA antibodies measured and should be considered to contribute to the pathology of SLE, especially lupus nephritis.

Effect of procainamide and hydralazine on poly (ADP-ribosylation) in cell lines

The prescription drugs procainamide (PA) and hydralazine (HYD) are associated with the induction of autoimmunity and a clinical syndrome called drug-induced lupus. Since PA- and HYD-induced autoantibodies are directed primarily against histones and histones are prime acceptors of poly (ADP-ribose) (PADPR), we have investigated the effects of PA and HYD on the activity of poly (ADP-ribose) polymerase (PADPRP). Control substances, with structures similar to PA and HYD but not known to induce lupus, included N-acetylprocainamide (NAPA) and the amino acids phenylalanine, tryptophan and proline, and their amide derivatives. Wil-2 cells were incubated in 0.5-50 microM PA, NAPA and HYD, which included therapeutic concentrations of these drugs. The mean enhancement of incorporation of [3H]-nicotinamide adenine dinucleotide (NAD) into PADPR was 1.84 (P = 0.005) with PA, with HYD 1.48 (P = 0.029), and with NAPA 1.38 (P = 0.036). This increase was suppressed by 3-aminobenzamide, an inhibitor of PADPRP activity. Little or no increase in [3H]-NAD incorporation was observed with equivalent concentrations of phenylalanine, phenylalaninamide or tryptophan. However, a 1.29-fold increase was noted with 0.5 microM tryptophanamide, a 1.26-fold increase with 0.5 microM prolinamide and a 1.4-fold increase with 50 microM proline. PA increased PADPRP activity in B- and T-cell lines but not in promyelocytic leukemia or epithelial cell lines. Since poly (ADP-ribosylation) is important in the cellular response to various agents, the increased ADP-ribosylation of intracellular molecules may be a key event in the induction of autoantibodies.

Autoantibodies associated with lupus induced by diverse drugs target a similar epitope in the (H2A-H2B)-DNA complex

IgG reactivity with the (H2A-H2B)-DNA complex, a subunit of the nucleosome, has been detected in many patients with lupus induced by procainamide and quinidine, but the similarity among the epitopes targeted by these antibodies in this heterogeneous patient group as well as the prevalence of this specificity in lupus induced by other drugs is unknown. Studies with histone-DNA complexes formed by sequential addition on a solid phase demonstrated that complexes containing single histones had negligible antigenicity, indicating that DNA stabilizes a protein epitope in the H2A-H2B dimer or that the complete epitope is generated by a surface feature involving H2A-H2B and DNA. F(ab')2 isolated from a patient with procainamide-induced lupus blocked greater than 90% of the anti-[(H2A-H2B)-DNA] reactivity in six of six sera from patients with lupus induced by procainamide, four of four quinidine-induced patients and in sera from patients with lupus induced by acebutolol, penicillamine, and isoniazid, but not methyldopa or auto-antibodies to the component macromolecules. Fab fragments purified from the IgG of two quinidine-induced lupus patients and patients with isoniazid- and procainamide-induced lupus retained 39% +/- 8% of their original IgG reactivity compared to 34 +/- 28% of the original anti-tetanus toxoid activity of Fab fragments in two of the same sera and two normal sera. These results indicate that anti-[(H2A-H2B)-DNA] does not require divalent antigen-antibody complexes for stability, and that the complete epitope is created by the monomeric, trimolecular histone-DNA complex. We conclude that despite their pharmacologic and chemical heterogeneity, many lupus-inducing drugs elicit near identical autoantibodies.

Free radical production and site-specific DNA damage induced by hydralazine in the presence of metal ions or peroxidase/hydrogen peroxide

Hydralazine caused site-specific DNA damage in the presence of Cu(II), Co(II), Fe(III), or peroxidase/H2O2. The order of inducing effect of metal ions on hydralazine-dependent DNA damage [Cu(II) greater than Co(II) greater than Fe(III)] was related to that of accelerating effect on the O2 consumption rate of hydralazine autoxidation. Catalase completely inhibited DNA damage by hydralazine plus Cu(II), but hydroxyl radical (.OH) scavengers and superoxide dismutase did not. On the other hand, DNA damage by hydralazine plus Fe(III) was inhibited by catalase and .OH scavengers. Hydralazine plus Cu(II) induced piperidine-labile sites predominantly at guanine and some adenine residues, whereas hydralazine plus Fe(III) caused cleavages at every nucleotide. Activation of hydralazine by peroxidase/H2O2 caused guanine-specific modification in DNA. ESR-spin trapping experiment showed that .OH and superoxide are generated during the Fe(III)- or Cu(II)-catalysed autoxidation of hydralazine, respectively, and that nitrogen-centered radical is generated during the Cu(II)- or peroxidase-catalysed oxidation. The generation of nitrogen-centered radical was also supported by HPLC-mass spectrometry. The results suggest that the guanine-specific modification by the enzymatic activation of hydralazine is due to the nitrogen-centered hydralazyl radical or derived active species, whereas .OH participates in DNA damage by hydralazine plus Fe(III). The mechanism of hydralazine plus Cu(II)-induced DNA damage is complex. The possible role of the DNA damage induced by hydralazine in the presence of Cu(II) or peroxidase/H2O2 is discussed in relation to hydralazine-induced lupus, mutation, and cancer.

Apheresis enhances the selective removal of antinuclear antibodies in systemic lupus erythematosus

Apheresis suppresses clinical manifestations of lupus and reduces levels of antinuclear antibodies implicated in the pathogenesis of systemic lupus erythematosus (SLE). It is not known, however, if reduced levels of antinuclear antibodies are due to nonspecific removal, or specific mechanisms associated with decreased production, or enhanced clearance from the circulation. In order to distinguish between specific and nonspecific effects of apheresis on antinuclear antibodies in SLE, we compared plasma levels of IgG antibodies to DNA and IgG antibodies to microbial antigens in 13 SLE patients before and after apheresis. Although apheresis lowered plasma levels of IgG (21% mean reduction), there was a disproportionate reduction in IgG antibodies to DNA (42% mean reduction, p less than 0.13). In marked contrast, reduction in antibodies to microbial antigens did not exceed those of plasma IgG. A rapid rebound of serum anti-DNA antibodies following apheresis in certain SLE patients suggests that the selective reduction in anti-DNA antibodies is due to enhanced clearance from the circulation rather than decreased production. These results indicate that apheresis enhances selective removal of antinuclear antibodies in some patients with SLE.

Poisoning due to class IA antiarrhythmic drugs. Quinidine, procainamide and disopyramide

Quinidine, procainamide and disopyramide are antiarrhythmic drugs in the class 1A category. These drugs have a low toxic to therapeutic ratio, and their use is associated with a number of serious adverse effects during long term therapy and life-threatening sequelae following acute overdose. Class 1A agents inhibit the fast inward sodium current and decrease the maximum rate of rise and amplitude of the cardiac action potential. Prolonged Q-T interval and, to a lesser extent, QRS duration may be observed at therapeutic concentrations of quinidine. With increasing plasma concentrations, progressive depression of automaticity and conduction velocity occur. 'Quinidine syncope' (a transient loss of consciousness due to paroxysmal ventricular tachycardia, frequently of the torsade de pointes type) occurs with therapeutic dosing, often in the first few days of therapy. Extracardiac adverse effects of quinidine include potentially intolerable gastrointestinal effects and hypersensitivity reactions such as fever, rash, blood dyscrasias and hepatitis. Procainamide produces electrophysiological changes that are similar to those of quinidine, although Q-T interval prolongation with the former is less pronounced at therapeutic concentrations. Hypersensitivity reactions including fever, rash and (more seriously) agranulocytosis are associated with procainamide, and a frequent adverse effect requiring cessation of therapy is the development of systemic lupus erythematosus. Of the 3 drugs, disopyramide has the most pronounced negative inotropic effects, which are especially significant in patients with pre-existing left ventricular dysfunction. As with quinidine, unexpected 'disopyramide syncope' at therapeutic concentrations has been described. Anticholinergic side effects are common with this drug and may require cessation of therapy. Disopyramide therapy may unpredictably induce severe hypoglycaemia. Severe intoxication with the class 1A agents may result from acute accidental or intentional overdose, or from accumulation of the drugs during long term therapy. Acute overdose can result in severe disturbances of cardiac conduction and hypotension, frequently accompanied by central nervous system toxicity. Decreased renal function can cause significant accumulation of procainamide and its active metabolite acecainide (N-acetyl-procainamide), resulting in severe intoxication. Mild to moderate renal dysfunction is less likely to lead to quinidine or disopyramide intoxication, unless renal failure is severe or concurrent hepatic dysfunction is present. Management of acute intoxication with class 1A drugs includes gut decontamination with provision of respiratory support and treatment of seizures as needed. Hypertonic sodium bicarbonate, by antagonising the inhibitory effect of quinidine on sodium conductance, may reverse many or all manifestations of cardiovascular toxicity.(ABSTRACT TRUNCATED AT 400 WORDS)

Drug-induced rheumatic syndromes. Diagnosis, clinical features and management

In order to avoid inappropriate therapy and prolonged morbidity, it is important to recognise when a patient's rheumatic complaints are due to drugs. However, this is often difficult because of the large number of drugs that have been implicated and the diversity of clinical presentations. Arthropathy may be seen with several different syndromes, including drug-induced lupus erythematosus (DILE), serum sickness and gout. The most widely reported of these is DILE, which usually develops after some months or even years of drug therapy. While many authors do not specifically require their presence for the diagnosis of DILE, antinuclear antibodies have been detected in the great majority of reported patients with DILE, whatever the causative drug. In contrast, patients who develop arthropathy soon after commencing a drug rarely have antinuclear antibodies and appear to be distinct from patients with DILE. Apart from arthropathy, a number of other syndromes that appear to have an immunological basis may be induced by drugs. Cutaneous vasculitis is not uncommon and drugs are frequently considered to be the aetiological factor. Whether drugs may cause larger vessel systemic vasculitis is less certain. Rarely, polymyositis and scleroderma-like syndromes have been associated with drug therapy. Corticosteroid-induced osteoporosis is a complication of all the corticosteroid preparations that are widely used at present. However, the development of deflazacort, a so-called 'bone-sparing' steroid, has raised the possibility that the effect of corticosteroids on bone may be separable, at least in part, from the other actions of these drugs. Data have been conflicting with regard to whether there is a 'safe' dose of corticosteroid. Similarly, it is unclear whether prophylactic therapy with agents such as calcium, fluoride and vitamin D is beneficial. Nonetheless, recent findings suggest that approaches will be developed to minimise the risk of osteoporosis in patients who require corticosteroids. There are a number of other ways in which drugs may affect bones. Osteomalacia is a well-known but uncommon complication of treatment with anticonvulsants and occasionally other drugs. The mechanism probably relates to the induction of hepatic enzymes and the consequent increased metabolism of vitamin D in patients with borderline levels initially. Osteosclerosis may also result from drug therapy; usually with fluoride or retinol (vitamin A) and its analogues. With continued research, the true spectrum of drug-induced rheumatic syndromes should become more clearly defined.(ABSTRACT TRUNCATED AT 400 WORDS)

Association of antibody to histone complex H2A-H2B with symptomatic procainamide-induced lupus

Antinuclear antibodies develop in most patients who are given prolonged procainamide therapy, but clinical symptoms resembling those of lupus appear in only 15 to 20 percent of such persons. No objective marker for symptomatic procainamide-induced lupus has been described. However, IgG antibodies to the histone complex H2A-H2B have previously been reported in this disorder, and it has been suggested that antiguanosine antibodies may be a marker for major manifestations of procainamide-induced lupus. We therefore tested for these antibodies in 20 symptomatic and 31 asymptomatic patients treated with procainamide. Most of the symptomatic patients had multiple manifestations of drug-induced lupus; resolution of symptoms after the discontinuation of procainamide was required for inclusion in the symptomatic group. All 20 symptomatic patients had elevated IgG antibodies to H2A-H2B, in contrast to only 2 asymptomatic patients (P less than 0.001). This activity was absent in patients not treated with procainamide and in patients with lupus induced by hydralazine or quinidine. IgG antiguanosine was elevated as compared with normal controls in 13 of 20 symptomatic and 19 of 31 asymptomatic patients--a finding that did not distinguish between symptomatic and asymptomatic patients. We conclude that IgG antibodies to H2A-H2B are a sensitive and specific marker for procainamide-induced lupus. The striking correlation between antibodies to H2A-H2B and symptomatic disease suggests a possible association between this antibody and the underlying pathogenic events.

Molecular mechanisms of autoimmunity

Autoimmune diseases result from a combination of genetic susceptibility factors and exogenous influences such as infection or chemical (including drug) exposure. Germline DNA variations in genetic type as well as defects in antigen recognition acquired during thymic education of developing T-lymphocytes both contribute to impaired self: nonself discrimination and set the stage for later development of such diseases as myasthenia gravis, polymyositis, or systemic lupus erythematosus. In addition, drugs such as D-penicillamine, hydralazine, procainamide, or quinidine induce T-cell or B-cell changes which precipitate auto-reactivity and cause drug-induced disease. Intervention in autoimmune diseases with prednisone, alkylating agents or the future use of more selective monoclonal antibody reagents may be life-saving in some of these disorders.

Antinuclear antibodies during psoralens plus ultraviolet A (PUVA) therapy--are they worthwhile?

From 1979 to 1985, 497 patients with psoriasis were started on psoralens plus ultraviolet A (PUVA) therapy at Northwestern University. Two hundred sixty-nine of these received therapy for greater than 3 months and had at least two antinuclear antibody (ANA) determinations. We have found that the difference between the number of significantly positive ANAs pre-PUVA therapy (4 of 269) compared to post-PUVA therapy (16 of 269) was not statistically significant. Furthermore, of the patients who did develop a significantly positive ANA, not one was found to have any symptoms, signs, or laboratory evidence of systemic lupus erythematosus. We therefore suggest obtaining ANAs prior to initiating PUVA therapy and obtaining follow-up ANAs only if the initial ANA is significantly positive. Patients with pre-PUVA--positive ANAs can be started on PUVA therapy if there is no evidence of lupus erythematosus.