WO1985005689A1

WO1985005689A1 - A diagnostic assay for the presence of antibodies associated with drug-induced lupus erythematosus

Info

Publication number: WO1985005689A1
Application number: PCT/US1985/001051
Authority: WO
Inventors: Robert L. Rubin; Eng M. Tan
Original assignee: Scripps Clinic And Research Foundation
Priority date: 1984-06-06
Filing date: 1985-06-06
Publication date: 1985-12-19
Also published as: AU4493585A; EP0182896A1

Abstract

A method for assaying for the presence of antibodies associated with drug-induced lupus erythematosus in which the activities of anti-histone H2A-H2B, H2A and H2B antibodies are determined. An activity of anti-histone H2A-H2B antibodies that is in excess of the sum of anti-H2A and anti-H2B antibody activities indicates the presence of antibodies associated with the desease, and warns of impending lupus. An assay system for the detection of and prediction of impending drug-induced lupus erythematosus is also disclosed.

Description

A DIAGNOSTIC ASSAY FOR THE

PRESENCE OF ANTIBODIES ASSOCIATED

WITH DRUG-INDUCED LUPUS ERYTHEMATOSUS

DESCRIPTION TECHNICAL FIELD OF THE INVENTION

The present invention relates to an assay method and diagnostic system for determining the activity of antihistone antibodies present in a liquid body sample taken from an individual having drug-induced lupus erythematosus and for predicting the incidence of drug-induced lupus. BACKGROUND OF THE INVENTION

Antinuclear antibodies and symptoms of systemic lupus erythematosus may arise in patients undergoing prolonged therapy with certain drugs. Of the many drugs associated with the chemically-induced or drug-induced lupus syndrome, hydralazine and procainamide pose the greatest risk. Approximately 5 percent of hydralazine-treated and 15-20 percent of procainamide-treated individuals develop the symptomatic disease state of lupus as a side effect of drugs received. These side effects in the past could not be predicted. Patients receiving these drugs therefore had to bear the risk of incurring deleterious physical side effects without the advantage of withdrawal of the drugs prior to the onset of severe physical symptoms.

The word "lupus" (Latin for wolf) has been used since the Middle Ages to describe a disease state wherein the face of the patient acquires a distinctive coloration or rash characteristic of the malar erythema of a wolf, a rash also characterized by its "butterfly-like" appearance. Lupus is now known to be a multi-faceted disease state. In the 1800's the systemic manifestations of the disease were described,, and those effects were utilized to diagnose lupus syndrome in the absence of the butterfly rash. In the 1940's the lupus erythematosus (LE) test was developed and more recently, the autoimmune aspects of the disease have been used to differentiate lupus from other similar disease states.

Lupus erythematosus is characterized by autoimmune phenomena. Patients develop antibodies to many of their own cells and cell constituents. Idiopathic systemic lupus erythematosus is treated symptomatically whereas drug-induced lupus erythematosus can be prevented altogether by timely withdrawal of drug therapies inducing this disease state. The cause of idiopathic systemic lupus erythematosus remains essentially unknown. Many theories as to the cause of this disease state have been propounded, including exposure to sunlight or ultraviolet radiation, endrocrine factors relating to the remission tendency within the last two trimesters of pregnancy and relapse post partum, and genetic factors related to the high incidence of lupus in relatives of affected patients. In addition, the lupus syndrome has been found secondarily to administration of procaina ide, hydralazine and other drugs where treatment is prolonged.

Lupus as a result of drug or chemical therapy presents with antinuclear antibodies (ANA) including antihistone antibodies (AHA) and anti-denatured DNA (dDNA) antibodies, similar to those occurring from an idiopathic disease state. The drug-induced autoimmune phenomena regress after drug therapies are discontinued.

The lupus syndrome presents a variety of symptoms, none of which alone is adequate for differential diagnosis of the disease state. In addition, lupus is characterized by a variety of unobvious and varied symptoms. Often the symptoms of lupus may be mistaken for other diseases, such as pleurisy, pericarditis, edema, dyspnea, Raynaud's phenomenon and seizures. The disease is considered a chronic illness with periods of remission and activity in a disease state of idiopathic origin. A common complaint of lupus patients is a pain in joints (arthralgia) and muscles (myalgia) .

Arthritis, or inflammation of the joints, is observed less frequently. Slight deformities at some joints may occur. However, major joint swelling, with synovial thickening, flexion deformities, and deviation are uncommon.

The classic butterfly rash is seen in- less than half of the patients with systemic lupus erythematosus. Kidney involvement is found in approximately half of lupus patients, usually within the first two years of the onset of symptoms. The endocardium, myocardium and pericardium are involved in approximately half of the patients with lupus. Clinical myocarditis is seen infrequently, although non-specific wave changes and other electroσardiographic changes are frequently observed. Anorexia, nausea, vomiting, and abdominal pain are commonly observed.

Laboratory findings of lupus patients generally include mild anemia, and auto-agglutination of red blood cells. The differential cell count is usually normal, although mononuclear cells may be more suppressed than neutraphils. Complicating infections generally result in a rise of the white blood count into either the normal or elevated range. A circulating anticoagulant has been noted by some investigators in as. many as a quarter of patients with SLE. This anticoagulant, either as an antibody to Factor VIII or more commonly as an inhibitor to the formation of prothombinase, results in prolonged clotting and prothombin times and may be associated with mild or_r rarely, severe hemorrhaging. The erythrocyte sedimentation rate is often elevated. Serum albumin levels are low, especially when nephritis is present. Gamma globulin levels, that are elevated in many patients, may be low in nephrotics.

Abnormal immunologic reactions often are the benchmark of a finding of lupus. Most characteristic of lupus are the large number of auto-antibodies that react with cells and their nuclear constituents. Historically of greatest significance is the so-called lupus erythematosus (LE) cell phenomenon. The LE cell consists of a leukocyte with a large homogenous inclusion body that compresses the nucleus against the cell membrane, leaving only a thin rim of cytoplasm. Because LE cells are not always present in patients with lupus, more sensitive tests have been developed for the detection of various antinuclear antibodies (ANA) . Although the LE test is often negative, the antinuclear antibody test results are positive in virtually all patients with lupus, whether the disease is spontaneous or secondary to drug therapies.

An im uno-flourescent (IF) technique employing rodent liver or kidney as a source of nuclei detects ANA in over 99% of lupus patients. The detecting reagents used in these tests are usually anti-IgG or anti-IgM antibodies.

It has been well established that one of the long term side effects of procainamide, hydralazine, and isoniazid treatments includes the development of antinuclear antibodies (ANA) and symptoms of systemic lupus erythematosus. ANA's have been detected in 30% of hydralazine treated patients. Perry, Am. J. Med., 54:58-72(1973). It appears that most patients treated with procaina ide for one year develop positive ANA's; Woolsley, et. al., N. Enql. J. Med., 298:1157-1159 (1978); approximately one-quarter of patients with ANA's progress to a systemic disease state. Perry, supra, and Woolsley, supra. Physicians are typically hesitant to discontinue procainamide or other drug therapies in light of the fact that a low number of patients with ANA develop the lupus syndrome with symptoms of SLE. In these cases, the benefits of the drug therapies outweigh the risk of discontinuation of the therapies involved. The difficulty, then, for the physician, is deciding which patients will continue to receive the phar acologic regime that may induce SLE symptoms such as myalgia, arthralgia, arthritis, pleuritis, and pericarditis, and which, although induced by the drug therapy, will generally resolve after the termination of the drug therapy, while the ANA may persist for several years. The pharmacologic therapies are valuable as they are effective therapeutic regimes for the treatment of very severe disease states such as cardiac arrythmias, hypertension and tuberculosis.

The clinical significance and possible pathologic role of auto-antibodies elicited by drugs is not well understood. The ANA specificity in procainamide-induced lupus patients was first shown to be predominantly antihistone activity by modified immuno-flourescence assay. Fritzler et. al., J. Clin. Invest., 62:560-567 (1978). This immuno-flourescence assay, when utilized to evaluate asymptomatic drug-treated patients whose liquid body samples such as sera present positive ANA, results in only 40% of the sera displaying antihistone antibodies (AHA's); C. Grossman, et al, Arthritis Rheum., 24:927-931 (1981). Prior assays have shown an artifact effect, and further have not definitively predicted the likelihood of lupus-like symptoms in patients receiving procainamide or hydralazine therapies. Antihistone antibodies (AHA's) have been detected by solid phase assay in many sera that were negative by immuno-flourescent assay, including sera from patients with hydralazine-induced lupus. Portanova, et al., Clin. Immunol. Immunopathol. , 25:67-79 (1982). These results have suggested that AHA may be the predominant ANA in asymptomatic patients, but their immune specificity and/or isotype (antibody class) may not be the same as that of the AHA in patients with symptoms of lupus induced by pharmacologic therapy. In fact, in utilization of the activated polystyrene-based assay of Portanova et al., supra, the activity of the histone complex H2A-H2B, may be affected by artificial denaturation of the histone complex, resulting in a more active binding of antibodies not specific to the complex. Therefore, the report of enhanced H2A-H2B may have been a result of denaturation, rather than an indication of ANA specificity that is predictive of the future symptomatic state in patients undergoing pharmacologic therapy. Reliance on AHA assays affected by artifacts may result in physicians unnecessarily withdrawing drug therapy from patients where a high AHA is present, but where the individual will not incur lupus syndrome. Often, other assays have yielded methods that provide false positive indications of the future disease state of lupus syndrome. Thus, the art has recognized the presence of AHA's in patients displaying drug-induced lupus, and has devised several techniques for assesing the activity of such antibodies. However, the significance of the presence or absence of these antibodies in various patients has not heretofore been properly appreciated. SUMMARY OF THE INVENTION

The present invention contemplates the measurement of activities of each of three members of the class of anti-histone antibodies (AHA) , and particularly the relationship of the activities of one class of AHA relative to the remaining two AHA's. In accordance with this invention, it has been found that AHA's are readily detected by a solid phase assay in sera that were negative by IF assay, Rubin, et al., Scand. J. Immunol., 15:63-70 (1982), including sera from patients with hydralazine-induced lupus, Portanova, et al. , supra. . This invention concerns the determination of the activity of anti-histones H2A and H2B and anti-histone H2A-H2B complex present in an aliquot of a liquid antibody-containing body sample taken from a human receiving chemical therapy. The activity of anti-H2A-H2B complex within the body sample is determined. The activity of each of anti-H2A and anti-H2B within a body sample from the same patient is also determined. The difference between the activity of anti-histone H2A-H2B antibody and the activities of anti-histone H2A and anti-histone H2B antibodies are determined. Activity of anti-histone H2A-H2B greater than the sum of the activities of anti-histone H2A and H2B indicates the presence of antibodies associated with drug-induced lupus erythematosus, and also warns of impending lupus erythematosus. This finding alerts the physician to avoid future occurrence of lupus symptoms.

It is particularly preferred that the determination of the anti-histone antibody activity be performed by enzyme linked immunosorbent assay (ELISA) . In that particularly preferred assay, individual, purified histones are preferably coated on a solid substrate as the antigen. Anti-histone antibodies present in an aliquot of a liquid body sample to be assayed immunoreact with the coated solid support following admixture of the liquid aliquot, and maintainence of the resulting solid-liquid phase composition for a predetermined period of time sufficient for the antibodies to immunoreact with the coated antigen. After separation of the solid and liquid phases, an aqueous liquid composition containing second antibodies such as anti-human Ig antibodies, having a linked indicating means, preferably an enzyme, is admixed with the solid support. The second solid-liquid phase composition that results is maintained for a period of time sufficient for the second antibodies to immunoreact with the immunoreacted anti-histone antibodies bound to the solid phase. The activity of the second antibodies is determined by means of the linked indicating means, e.g., enzyme, thereby also determining the activity of the immunoreacted anti-histone antibodies. Suitable enzymes useful as an indicating means include horseradish peroxidase and alkaline phosphatase.

An advantage of the method, and particularly the use of the disclosed ELISA is that artificial deactivation of the H2A-H2B complex does not occur, and therefore the apparent loss of activity of individual histones as can result from artificial denaturation is avoided. This in turn results in a more accurate assay of the activity of antibodies to the histone H2A-H2B complex, and to anti-H2A and anti-H2B antibodies. A further advantage to this diagnostic method is that it can predict the occurrence of future lupus symptoms in presently asymptomatic patients.

Still a further advantage of the diagnostic method of the present invention is that it differentiates between benign AHA that lead to symptoms of lupus and AHA associated with development of secondary, drug-induced lupus syndrome, because of the specificity of the antibody activities to individual histones.

DESCRIPTION OF FIGURES

In the drawings forming a portion of this disclosure:

Figure 1 contains two graphs that show the anti-histone profile of a procainamide-treated patient with symptoms of lupus (A) , and of a patient treated with procainamide for 7 years without developing symptoms (B) . Antibody binding to individual histones and to the histone H2A-H2B complex was detected by ELISA with either anti-human IgM or anti-human IgG linked to horseradish peroxidase as a means for detecting AHA bound to the solid support. The ordinate is in units of optical density (O.D.) determined at 410 nanometers. Details of the ELISA are provided in the Materials and Methods section.

Figure 2 contains four graphs that show the IgG anti-histone activity in four asymptomatic patients (A.H., R.P., H.K., and L.C.) treated with procainamide. These patients also displayed IgM -10- activity to most or all histone elements. Details of this ELISA are as in Figure 1.

Figure 3^'contains two graphs showing the reactivity of sera from eleven patients having procainamide-induced lupus with individual histones and histone complexes, as well as the classes of the reactive antibodies. Peroxidase-conjugated anti-IgM was the detecting reagent for the data on the left and peroxidase-conjugated an i-IgG was used for the 0 data on the right. An optical density two standard deviations above the mean of the binding of 20 normal sera to each antigen is indicated by the T-shaped symbols. Details for these graphs are as described in Figure 1.

15 Figure 4 contains two graphs showing the reactivity of sera from eleven SLE patients with individual histones and histone complexes. Details for these graphs are as described in Figure 3. DETAILED DESCRIPTION OF THE INVENTION 0 I. DISCUSSION

A. Introduction

The present invention is directed to an assay method for detecting and measuring the incidence, immunoglobulin class, and antigen binding 5 specificity of anti-histone antibodies in an aliquot of a liquid antibody-containing body sample. The preferred body sample for the assay is serum, but the other body fluids, such as plasma, blood, synovial fluid, cerebrospinal fluid, fluid within bullae, 0 pleural and pericardial effusions are useful sources of the anti-histone antibody-containing body sample. For ease of description, serum will be used as illustrative of such body fluids that are evaluated for the presence of the assayed anti-histone antibody. 5 The presence of the anti-histone antibodies in a body fluid is indicative of the lupus syndrome. The evaluation of the interrelationship of the levels of activity provides to the diagnostician the indication of the type of anti-histone antibody activation. When the relationship of the activities of the antibodies to histones H2A and H2B, and to the H2A-H2B histone complex is such that the activity of antibodies to the complex is greater than the sum of the activities to the complex components, antibodies associated with drug-induced lupus are present, and the presence of a drug-induced lupus disease state is indicated, with lupus symptoms being present or being manifest in the future.

The measurement of auto-antibodies elicited by chemical therapy such as procainamide, hydralazine or isoniazid with a characteristic immunoglobulin class and unique specificity is of pathological significance in evaluating the treated individual for presence of chemically-induced or secondary lupus, as well as for predicting the incidence of drug- or chemically-induced lupus syndrome where the individual evaluated has no overt clinical symptoms of the lupus syndrome.

The anti-histone antibodies, as measured, are present in sera only when the anti-histone antibodies or anti-nuclear antibodies have been activated, either from drug therapy or from other chemical treatment. The anti-histone antibodies include a series of individual antibodies whose activities are separately evaluated and compared.

That comparison determines the unique fingerprint of anti-histone antibody activity as shown in Figure 1, and indicates the likelihood of an individual manifesting lupus syndrome without current overt symptoms. B. Results

Auto-antibodies elicited by procainamide may have remarkably discrete specificities for individual histones or histone-histone complexes and a wide diversity of specificities occurs within the patient populations. Serum from one patient may exhibit IgG antibodies to only one histone, while another patient may have antibodies to a different histone. IgM reactivity to all histones was commonly found in the studies discussed herein, but whether this activity was due to separate populations of AHA, reactive with individual histones, or to pure antibody species with cross-reactivity to numerous histones was not definitively established. The results described herein indicate that different AHA specificities are present in symptomatic and asymptomatic patients. This new finding of AHA with different specificities in symptomatic compared to asymptomatic patients may explain the variation of incidents of drug-induced AHA and of other diseases such as SLE.

Complement fixation methods, immunoflourescent assays and solid phase assays (data not shown) were found to be less accurate in detecting AHA of the wide diversity of specificities apparently present in auto-immune sera that were detected by the present ELISA, especially if the antigens used were poorly characterized or subjected to irreversible denaturing conditions within the assay protocol.

Most patients with symptoms of procainamide-induced lupus had an antibody activity with a unique specificity for a complex of histones H2A and H2B, i.e., H2A-H2B. Anti-H2A-H2B activity is defined by the presence of antibody activity to a complexed mixture of these histones (or to native H2A-H2B complexes) that is in excess of the sum of the anti-H2A plus anti-H2B activities. Thus, the complex H2A-H2B apparently exhibits a new epitope that is not present in either single histone.

H2A-H2B has been previously reported to have antigenic activity in sera from patients with procainamide induced lupus, Fritzler, J. Clin. Invest. , 62:560-567 (1978). This complex was shown to possess an antigenicity greater than the sum of the antigenic activities of the individual separate histones, Portanova, supra.

However, it was not known whether the presence of that anti-H2A-H2B activity had any pathological meaning, or could be used to predict that patients would exhibit lupus symptoms. The present findings that show asymptomatic patients who are on chemical therapies that can lead to drug-induced lupus do not exhibit antibodies to the H2A-H2B complex. Those findings now permit use of the known elevated level of anti-H2A-H2B antibody activity to indicate the presence of antibodies associated with or related to drug-induced lupus, and also to indicate the presence or future occurence of drug-induced lupus.

It is well known that dimers of H2A and H2B readily form in. solution, and use of this fact was incorporated into the ELISA assay described herein. The H2A-H2B complex is an integral component of the nucleosome, the structural subunit of chromatin. However, H3 and H4 complexes, which are also important nucleosome constituents, did not display unique antigenicity with any serum according to results obtained from this study. The presence of anti-H2A-H2B activity is thus believed to be of unparalleled diagnostic value in warning of an impending clinical deterioration and may also have pathologic significance.

Both IgG and IgM anti-dDNA antibodies were elicited by procainamide, and serial studies - suggested that these antibody classes arose concordantly de novo. The concordant appearance of both IgG and IgM AHA's is not expected of a classical immunization phenomenon for the underlying mechanism of the genesis of the antibodies. However, without regard to the mechanism underlying its elicitation, the presence of increased anti-H2A-H2B complex activity, when compared to the anti-H2A and anti-H2B activities, can indicate the presence of impending clinical deterioration, and provides an accurate method of prognosticating the induction of pathologic symptoms.

The utilization of the ELISA in the manner described herein points to a d_e novo induction of AHA having both unique specificities as well as diversity within the patient population. The mechanism for induction of AHA's by drugs appears to be different in patients who remain asymptomatic compared to those who have developed, or will develop in the near future, the adverse pathologic symptoms of lupus syndrome.

The overall prevalence of anti-histone in the population tested is shown in Figure 1. Elevated anti-histone and anti-dDNA (data not shown) were commonly observed in both symptomatic and asymptomatic patients (Figure 4) . All symptomatic patients had elevated anti-histone as is seen in Figures 1, 2 and 3. Seven of these patients also had elevated anti-dDNA, at varying activities. Of asymptomatic patients tested, 66% had elevated anti-histone and anti-dDNA. Of this group. two patients had been symptomatic but were in remission at the time that the serum sample was drawn. Others had converted to elevated auto-antibodies during a one-year observation period or had presented consistently elevated auto-antibodies on numerous occasions, and tended to have been treated with procainamide for long periods of time (approximately four years) . Most of the patients without elevated auto-antibodies were exposed to procainamide for a substantially shorter duration of time (less than one year) .

Examination of antibody reactivity to the five histone classes revealed patterns which tended to distinguish between asymptomatic and presently or predictively symptomatic patients. Of the six procainamide treated patients who were symptomatic at the time blood was drawn, four showed a unique anti-histone profile similar to that shown in Figure 1A. This profile is unique in two respects, it shows (a) a relative absence of both IgM and IgG antibody activity to individual histones, and (b) a substantial binding activity to an antigen consisting of a complex of histones H2A and H2B.

Patients with symptomatic drug-induced lupus presenting elevated anti-H2A-H2B activities, had an elevated level of antibody activity more than two standard deviations above the normal anti-H2A or anti-H2B activities, providing a specific reaction indicative of symptomology (See Figure 3) . It is the relationship of the anti-H2A-H2B complex compared to the anti-H2A and anti-H2B that is considered as a unique fingerprint when graphing the anti-histone responses. That comparison yields the conclusion that the auto-antibody anti-H2A-H2B is crucial in determining the onset or presence of symptoms of lupus in drug treated patients. (See Figures 1 through 4) .

Asymptomatic patients taking procainamide for a number of years commonly displayed a pattern of high IgM anti-histone antibody activity and a relative uniform reaction with all histones similar to the activities shown in Figure IB. The IgG antibody activity in asymptomatic patients was largely restricted to only one or two histones, and the histone class displaying predominant antigenicity varied among the patients.

Figure 2 shows the IgG anti-histone activities of four asymptomatic patients receiving procainamide against five individual histones. Activities of such anti-H2A-H2B complex are not shown. The anti-H2A-H2B activity was generally normal for such asymptomatic patients such as A.H. , H.K. and L.C.

Serum from patient R.P. had an increased anti-H2A-H2B activity, but that increased activity could be accounted for by the activities of the anti-H2A and anti-H2B antibodies. Thus, the sum of the anti-H2A and anti-H2B antibody activities was greater than the activity shown for the anti-H2A-H23. antibodies.

An elevation of anti-H2A-H2B complex activity, as shown in Figures 1 and 3, more than two standard deviations above the normal value for the anti-histone H2A and H2B antibody activities, measuring individual classes, can also indicate the existence of present or future pathological lupus syndrome. While a determination that the anti-histone H2A-H2B complex activity is greater than two standard deviations above normal levels for anti-histone H2A and H2B antibody activity can be used to indicate drug-induced lupus, the described assay wherein the difference in the three antibody activities is utilized is believed to be a more accurate determination for the existence of present or future pathological lupus syndrome.

The utilization of the ELISA as described herein yields an accurate anti-histone profile. With the diagnostic test employed, ELISA avoids the alteration of response due to an artifact within the test itself. This particularly preferred assay shows the unique relationship amongst the histone classes. It also shows that the anti-H2A-H2B complex activity elevated by more than two standard deviations above the mean provides an excellent fingerprint indicator of lupus symptomology, either present or predictive. C. General ELISA Assay Method In the method of this invention, a liquid body sample from a human patient believed to contain anti-histone antibodies is assayed to determine the presence and the activity of the anti-histone antibodies. The procedure utilized requires that the anti-histone antibodies be evaluated through individual assays. Individual histones are purified and isolated as by gel filtration chromotagraphy, and the purity of each histone; i.e., the freedom from interfering histones, is confirmed.

ELISA (Enzyme Linked Immunosorbent Assay) was utilized for the results discussed hereinbefore in section I B. Indeed, ELISA is the particularly preferred method of carrying out the assay of this 'invention. Other assay methods, described herinafter, may also be used. In an ELISA, enzymes linked to either an antigen or antibody can be used as a label or indicating means that is easily detected by measurement of enzyme activity. The enzyme horseradish peroxidase was linked to an antibody in the particularly preferred ELISA whose results are described hereinafter. That ELISA is carried out as follows. A histone such as histone H2A, H2B, or the

H2A-H2B complex was coated on a solid phase support, such as a well of a polystyrene microtiter plate by incubating for about six to eighteen hours, to form a coated solid support. An aliquot of a liquid antibody-containing body sample from a human to be assayed was admixed with each coated solid support to form a solid-liquid phase admixture. The admixture was maintained for about one and one-half hours, sufficient for anti-histone antibodies to immunoreact with each, separate, histone antigen of the coated solid support as by incubation of the serum with the coated solid support. The solid and liquid phases were separated as by rinsing.

In preferred practice, the coated support is contacted with a solution containing a protein that binds to the solid support to block non-specific protein binding but does not interfere with the remaining steps of the method. That second coating step is carried out prior to admixture of the aliquot of body sample. Exemplary useful proteins include serum albumins such as bovine and human serum albumins and gelatin.

The amount of anti-histone activity for each histone (H2A, H2B and the H2A-H2B complex) , was then determined by admixture of the solid phase with an aqueous liquid composition containing second antibodies such as anti-immunoglobulins that are linked to an indicating means such as the above enzyme. Each second solid-liquid phase composition so prepared was maintained for about one and one-half hours, sufficient time for the second antibodies to immunoreact with the bound, immunoreacted anti-histone antibodies. The solid and liquid phases were again separated and the activity of the enzyme remaining with the solid phase was measured.

That measurement provided the activity of the second antibodies, and thereby a measurement of the activity of the anti-histone antibodies in the aliquot of body sample. Comparison of the anti-H2A-H2B complex activity to the individual activities for anti-H2A and anti-H2B provided a determination for the presence of the anti-H2A-H2B comlex antibodies associated with drug-induced lupus, and thus a determination fo the presence of that disease state, as discussed before.

It is noted that aliquots from the same body sample of the same human are utilized in the above method. It is preferred that each sample analyzed for activity of a particular anti-histone antibody be diluted to the same extent as are the other body samples that are used so that the relative antibody activities can be readily compared. However, it is not necessary that each sample be equally diluted so long as the activities obtained are normalized to provide the relative antibody activities of a single body sample. Of course, when each aliquot used is substanially identical to the other aliquots used, normalization is not required.

It is also reiterated that the antibody activities against histone H2A, histone H2B and the histone H2A-H2B complex are separately determined. Preferably in the presently preferred ELISA assay, each histone is coated on a separate solid support so that the anti-histone activities are not only determined separately, but each is preferably determined in the absence of the other histones. Thus, three substantially identical coated solid supports are used, three first solid-liquid phase compositions are prepared, three second solid-liquid phase compositions are prepared, and the like. While each anti-histone activity is determined separately, those activities are preferably determined by substantially identical techniques. Those techniques are discussed hereinbefore and differed herein only in the specific histone, or histone complex, used as the antigen coating on the solid support.

The second, indicating means-containing antibodies are preferably anti-human Ig antibodies and are preferably of the IgG and IgM classes. Such antibodies may be conveniently raised in mammals such as goats, mice, rats, rabbits, and horses and immunoreact with human IgG or IgM antibodies, respectively. The data shown in the Figures illustrates the usefulness of knowing the class of antibodies whose activities are being assayed. Of the two antibody types used, human anti-IgG antibodies appear to give superior results to anti-human IgM antibodies, and thus anti-human IgG antibodies are preferred as the second antibodies. Alternative solid supports for use in the ELISA method or for other varieties of assay, include polystyrene beads, about 1 micron to 5 millimeters (mm) in diameter, available from Abbott Laboratories, North Chicago, polystyrene tubes, sticks or paddles of any convenient size, nitrocellulose sheets, sticks or paddles, and polystyrene latex whose polystyrene particles are of a size of about one micron and can be centrifugally separated from the latex. Although the above ELISA technique is preferred, anti-histone activity can also be determined in an ELISA by binding the antibodies present in the aliquot of body sample to the solid support to form a coated solid support. A liquid composition containing a known amount of an individual histones, each linked to an enzyme as an indicating means, is then added to form the solid-liquid phase composition. The remaining steps in this assay are substantially identical to those discussed hereinabove. Enzyme immunoassay has advantages over radioimmunoassay in that radioactive materials with relatively short half lives need not be handled and stored. Furthermore, artifacts that may be found in a solid phase enzyme immunoassay or radioimmunoassay or immunoflourescent assay, are not found in the particularly preferred ELISA method. However, as discussed hereinafter in Section I D, such assay methods can be used in the method of the present invention. The enzyme linked as an indicating means may be coupled to antigens or antibodies by various cross-linking agents, particularly glutaraldehyde, and dimaleimide, as is well known in the art. The particular enzyme utilized is not thought to be critical to the ELISA, as long as it is soluble, stable, and not present in biological fluids used in the assay in a form that would interfere with the assay determinations.

Enzyme labels also require a substrate and some require co-factors. Exemplary substrates and co-factors for enzymes such as alkaline phosphatase and horseradish peroxidase are well known in the art and include p_-nitrophenyl phosphate, and hydrogen peroxide and o-phenylenediamine, respectively. The enzyme and its substrate and co-factors, where used, are broadly referred to herein as an indicating means or label. For a discussion of such enzymes and substrates, see generally Maggio, Enzyme Immunoassay, CRC Press, Boca Raton, Florida, (1980) . D. Additional Assay Methods Another assay that can be utilized in this diagnostic method is radioimmunoassay. The general methodology of radioimmunoassay is outlined below.

Antigen such as one of the individual members of the class of histones is adsorbed onto tubes or covalently linked to a solid matrix. In the coated method, antigen is simply adsorbed onto polystyrene tubes, and free antibody is then removed by washing. An unlabelled aliquot of liquid human body sample, such as serum is then added followed by a radiolabelled second antibody that binds to the anti-histone antibodies in the serum; i.e., an_. anti-human IgG or IgM antibody. After separately maintaining the body sample-solid matrix admixture for a predetermined time for antibodies to react with the histones (incubating) , separation of the solid and liquid phases, and admixture of the radiolabelled antibody and maintenance of the admixture for a predetermined time sufficient for the second antibodies to react, the tubes are washed and dried, and radioactivity is measured in a gamma spectrometer.

A further alternative method of conducting the assay of this invention is fluorescent immunoassay system. The measurement of the anti-histone antibodies is accomplished by competitive binding of fluorescent, labelled, specific antiserum for free and solid phase antigens.

The free, anti-histone antibodies are present in an aliquot of human body sera to be sampled. The bound individual member of the class of histones is fixed to a polymeric hydrophobic surface such as the solid support defined before. The histone, after being fixed (adsorbed or coated) to the polymer surface in solid phase, is incubated with test sera as described before. A second flourescent anti-immunoglobulin reagent (anti-Ig) is then bound to the first antigen-antibody immunoreaction product following the beforedescribed procedures, and the amount of bound flourescence is measured flourometrically.

The label utilized in the assay can be any suitable label that allows for identification and quantification. Such labels include enzymes, flourochrome dyes or radioactive isotopes such as Iodine 125 or 131, Hydrogen 3 and Sulfur 35.

Suitable flourochrome dyes, that can be linked for labelling by immunoflourescence include beta-anthracene, rhodamine and flourescein that are bound to an antibody or antigen by a linking group such as a reacted isocyanate group.

The enzymes chosen should be relatively stable, have a relatively long shelf life, be readily available and inexpensive. The activity of the enzyme should also be easily measurable using simple colorimetric or flourimetric methods with small amounts of enzyme being detectable. Therefore, the enzymes should have a high substrate turnover number, and not be affected by biological components in the test sample. Such enzymes include alkaline phosphatase, horseradish peroxidase, or catalase and glucose oxidase, and their substrates, as was discussed hereinbefore.

^E« Diagnostic Assay System

The assay method for determining activity of the anti-histone antibodies, individually. particularly the anti-histone H2A, anti-histone H2B and anti-histone-H2A-H2B complex, can be practiced through use of a- diagnostic assay system containing the necessary materials in kit form. Generally, such a system comprises a first specific binding agent and a second specific binding agent that includes the label. Where the label is an enzyme, an enzyme substrate and co-factor can also be provided.

In the kit, the several specific binding agents and enzyme substrates are preferably packaged separately. The first specific binding agent such as the individual, purified histones H2A, H2B, as well as the H2A-H2B complex can be provided in solution to be immobilized on a solid support before use or can be coupled to the solid phase.

The second, labelled, specific binding agent such as enzyme-labelled anti-human IgG and IgM antibodies and their substrate, if needed, can also be provided in dry form or in one or more solutions as are necessary to simplify use. While an antibody ^■ dispersed in liquid may not be a true solution in the strict sense of the word, for purposes of this discussion it will be considered as a solution.

For ease of use the first specific binding agent is preferably provided coupled and immobilized in a solid matrix or support, such as in the wells of a microtitration plate having a plurality of wells. An optimal arrangement includes the first specific binding agent immobilized in the lower or bottom portion of the wells. A blocking protein such as bovine serum albumin (BSA) or gelatin may also be coated on the solid support to block non-specific protein binding sites on the solid support.

The microtitration plate is preferably made of any suitable material that is clear and adsorbs or permits linking of the first specific binding agent so that the binding agent is retained in the wells even during washing. Such materials include polystyrene, polyvinyl chloride, polypropylene, polycarbonate or treated glass. The wells preferably have flat bottoms to provide an appropriate optical surface through which the amount of label may be read spectrophotometrically.

The plate is preferably divided into two sections, a sample section, having a plurality of sample wells in which the aliquots of body sample to be assayed are placed, and a standard section or row, having a plurality of standard wells in which standard control samples have been or will be placed. The wells in the sample section are provided with a known, predetermined amount of first specific binding agent (histone) such as about 1-3 micrograms/milliliter, an amount in a molar excess of the expected amount of first anti-histone component in the samples to be assayed. The wells in the standard section contain varying known quantities of second antibody components. II. MATERIALS AND METHODS

^A« Reagents and Buffers Total histones were purified from calf thymus or obtained from CalBiochem Behring, Inc. (La Jolla, CA) . Thymus total histones were purified and prepared as described by Rubin, et al.. Biochemistry, 14 (1978) . Briefly, the isolation of the individual histones used herein as antigens was accomplished by gel filtration chromatography following the teachings of Bohm, Ed. E.L., et al., FEBS Letters, 3-4:217-221 (1973) , and yielded pure factions of the histones. The complex H2A-H2B was prepared by mixing 100 micrograms of each histone. The complex H2A-H2B was diluted to 100 micrograms/ml of each histone H2A and

H2B in 10-X PBS (10 fold-concentrated PBS). The solution so prepared was maintained for a period of 1-2 hours at 5 degrees Celsius before dilution to 2 micrograms per milliliter in PBS and addition to microtiter plates.

Bovine serum albumin (BSA) was obtained from

Sigma (St. Louis, MO). Horseradish peroxidase-linked goat anti-human polyvalent, anti-IgG and an i-IgM antibodies were obtained from Tago (Berlinga e, CA) .

PBS contained 10 millimolar (mM) phosphate and 150 mM

NaCl, and had a pH value of 7.4.

B. Human Sera Sera from patients were obtained from persons being treated in the Division of Cardiovascular Disease, Scripps Clinical and Research Foundation. Some patients had additional cardiac problems in addition to the ventricular arrhthymias for which they were being treated at the Scripps Clinic. Those patients with additional problems presented congestive heart failure, chronic obstructive pulmonary disease and arterioschlerosis and were taking additional medications including Digoxin, Furosi ide, Nifidipine, Allopurinol, and occasionally a low-dose of Prednisone. The bulk of the patients were treated with sustained release Procainamide (Procan-SR, Parke-Davis, Morris Plaines, NJ) at 2-3 gm/day to suppress ventricular arrhythmias. In addition, two patients with hydralzine-induced lupus were evaluated.

C. ELISA Conditions

Histone purity was greater than 95% as j udged by polyacrylamide gel electrophoresis. To each Immulon 2 plate , (Dynatech , Alexandria, Virginia) were added 0.2 ml histone solution at 2.5 micrograms/milliliter when total histone was used, 2 ug/ml when H2A-H2B complexes were used, or 1 ug/ml when individual histones were used. Contact between the solid support and histone-containing solutions was maintained at 5 degrees Centigrade for six hours with agitation or 18 hours without agitation to form the coated solid support. After that contact period, the wells were post-coated with 1 mg gelatin, incubated in the antigen coated plate for 2 hours to bind the gelatin to non-specific binding sites on the wells. After incubation the wells were washed with PBS-TWEEN 20.

In the ELISA utilized to provide the above results, sera were diluted one to two hundred (1:200) in phosphate buffered saline (PBS), containing 0.05 percent TWEEN [polyoxyethylene (20) sorbitan monolaurate] , bovine serum albumin (BSA) at 1 mg/ml, gelatin at 1 mg/ml and 0.75 mg/ml bovine gamma globulin (BGG) . A 0.2 ml aliquot of the diluted serum was maintained in contact (incubated) with the antigen coated plate for two hours. Bound antibody activity was then determined using horseradish peroxidase linked anti-human immunoglobulin (Ig) , diluted 1:500 in PBS-TWEEN-20 containing 5 mg/ml BSA, 1 mg/ml BGG and 1 mg/ml thimerisol (Sigma, St. Louis, Mo.). The anti-human immunoglobulin was either goat anti-human immunoglobulin (IgG) and anti-human immunoglobulin (IgM). A 0.2 ml aliquot of the peroxidase conjugated anti-human immunoglobulin solution was incubated in the wells for 1 to 2 hours with agitation.

After the incubation, the wells were rinsed with PBS-TWEEN-20 and twice with PBS. A 0.2 ml aliquot of substrate solution containing 1 mg/ml 2,2'-azino-di-(-3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) (Sigma, St. Louis, Mo.) and 0.005 percent H-O- in 0.1 molar Mcllvaine's buffer at pH value of 4.6 was added to the well. Optical density (OD) was determined after 30 minutes^" and after 2 hours using an MR-600 automated spectrophotometer (Dynatech Laboratories, Alexandria, Va.) at 410 nanometers.

The validity of the ELISA for measuring anti-histone antibodies was established by comparison with results using a solid phase radioimmunoassay and immunoflourescence assay, as described in Rubin, et al., Scand. J. Immunol., 15:3 (1982).

Claims

WHAT IS CLAIMED IS:

1. A method for assaying the presence of antibodies associated with drug-induced lupus erythematosus in humans comprising the steps of: (a) determining the activity of anti-histone H2A-H2B in an aliquot of a liquid antibody-containing body sample from a human to be assayed;

(b) determining the activity of anti-histone H2A in an aliquot of a liquid antibody-containing body sample from said human;

(c) determining the activity of anti-histone H2B in an aliquot of a liquid antibody-containing body sample from said human; and (d) determining the difference between the activity of anti-histone H2A-H2B antibody and of the activities of anti-histone H2A and H2B antibodies, an activity of said anti-histone H2A-H2B antibody greater than the sum of activities of anti-histone H2A and H2B antibodies indicating the presence of said antibody associated with drug-induced lupus erythmetosus.

2. The method according to claim 1 wherein said activity determinations are carried out separately using separate aliquots of body sample.

3. The method according to claim 2 wherein said separate activity determinations are carried out for each of said individual anti-histone antibodies by

(i) coating a solid support with each of histone H2A, histone H2B and a histone H2A-H2B complex to form coated solid phase supports;

(ii) admixing an aliquot of a liquid antibody-containing body sample to be assayed with said coated solid supports to form solid-liquid phase admixtures; (iii) maintaining said admixtures for a predetermined time sufficient for anti-histone H2A, H2B or complex anti-histone H2A-H2B antibodies to immunoreact with each said coated solid support; (iv) separating said solid and liquid phases; and

(v) determining the activity of anti-histone H2A, H2B and H2A-H2B complex antibodies that immunoreacted with said coated solid supports.

4. The method according to claim 3 wherein each of the coated solid supports of step (i) is contacted with a solution containing a protein that binds to said solid support to block non-specific protein binding but does not interefere with the remaining steps of said method.

5. The method according to claim 3 wherein said activities of anti-histone H2A, H2B and H2A-H2B antibodies that immunoreacted with said solid supports are determined by the additional steps of: (A-l) admixing an aqueous solution of second antibodies containing a linked indicating means with the solid phase obtained after step (iv) to form second solid-liquid phase compositions, said second antibodies immunoreacting with said anti-histone H2A, H2B and H2A-H2B antibodies, and said indicating means providing a means for assessing the amount of said second antibodies that reacted with said anti-histone H2A-H2B antibodies;

(A-2) maintaining said admixture for a predetermined time sufficient for said second antibodies to immunoreact with said anti-histone H2A, H2B and H2A-H2B antibodies;

(A-3) separating the solid and liquid phases of said second solid-liquid phase compositions; and (A-4) determining the activity of said second antibodies that immunoreacted with said anti-histone H2A, H2B and H2A-H2B antibodies, and thereby the activity of said immunoreacted anti-histone H2A, H2B and H2A-H2B antibodies.

6. The method according to claim 5 wherein said linked indicating means is an enzyme.

7. The method according to claim 5 wherein said second antibodies are anti-human Ig antibodies.

8. The method according to claim 7 wherein said anti-human Ig antibodies are anti-human IgG antibodies.

9. The method according to claim 7 wherein said anti-human Ig antibodies are anti-human IgM antibodies.

10. A method of assaying for the presence of antibodies related to drug-induced lupus erythematosus in humans comprising the steps of:

(a) separately determining the activity of antibodies to histone H2A, histone H2B and the histone H2A-H2B complex in substantially identical aliquots of a single, liquid antibody-containing body sample from a human; said separate activity determinations being carried out for each of said antibodies by

(i) coating separate polystyrene microtiter plate wells with solutions containing known, predetermined amounts of each of said histone H2A, histone H2B and the histone H2A-H2B complex to form coated solid supports;

(ii) admixing said aliquots of body sample with each of said solid supports to form solid-liquid phase admixtures;

(iii) maintaining said admixtures for a predetermined time sufficient for anti-histone H2A, anti-histone H2B and anti-histone H2A-H2B complex antibodies to immunoreact with histones H2A, H2B and the H2A-H2B complex, respectively, of the coated solid supports; (iv) separating each of said solid and liquid phases;

(v) admixing separate aqueous solutions of second, anti-human Ig antibodies linked to indicating means with said separated solid phases to form second solid-liquid phase compositions, said indicating means comprising an enzyme;

(vi) maintaining said second compositions for a predetermined time sufficient for said anti-human antibodies to immunoreact with said anti-histone H2A, H2B and H2A-H2B complex antibodies;

(vii) separating the solid and liquid phases of said second compositions; and

(viii) determining the activities of each of said anti-human antibodies that immunoreacted with said anti-histone antibodies, and thereby determining the activities of each of said anti-histone antibodies; and

(b) determining the difference between the activity of the anti-histone H2A-H2B complex antibody activity and the activities of anti-histone H2A and H2B antibody activities, an activity of said anti-histone H2A-H2B complex antibody greater than the sum of activities of anti-histone H2A and H2B indicating the presence of said antibody associated win drug-induced lupus erythematosus in said body sample.

11. The method according to claim 10 wherein said body sample is serum.

12. The method according to claim 10 wherein said anti-human Ig antibodies are anti-IgG antibodies.

13. A method of predicting the appearance of symptoms of drug induced lupus erythematosus in humans undergoing therapy with a drug selected from the group consisting of procainamide, hydralazine, and isoniazide comprising the steps of:

(a) determining the activity of anti-histone H2A-H2B complex in an aliquot of a liquid antibody-containing body sample from a human to be assayed; (b) determining the activity of anti-histone H2A in an aliquot of a liquid antibody-containing body sample from said human;

(c) determining the activity of anti-histone H2B in an aliquot of a liquid antibody-containing body sample from said human; and

(d) determining the difference between the activity of anti-histone H2A-H2B antibody and the activities of anti-histone H2A and H2B antibodies, an activity of said anti-histone H2A-H2B greater than the sum of activities of anti-histones H2A and H2B antibodies being predictive of the appearance of symptoms of drug induced lupus erythematosus.

14. A method for assaying the presence of antibody associated with drug-induced lupus erythematosus in humans comprising the steps of: (a) determining the activity of anti-histone H2A-H2B antibodies in an aliquot of a liquid antibody-containing body sample from a human to be assayed; and (b) comparing said activity to the anti-histone H2A-H2B antibodies in normal patients, an activity of said anti-histone H2A-H2B antibodies in said sample that is more than two standard deviations above the normal activity indicating the presence of said antibody associated with drug-induced lupus erythematosus.