WO1991008487A1

WO1991008487A1 - Method of determining anti-xanthine oxidase antibody

Info

Publication number: WO1991008487A1
Application number: PCT/GB1990/001872
Authority: WO
Inventors: Roger Harrison
Original assignee: University Of Bath
Priority date: 1989-12-01
Filing date: 1990-12-03
Publication date: 1991-06-13
Also published as: AU6883591A; EP0506682A1

Abstract

Anti-(xanthine oxidase) IgM antibodies, whose presence is related to heart disease, may be measured using xanthine oxidase or a polypeptide fragment thereof capable of competing with xanthine oxidase for binding to the antibody and an anti-IgM immunoglobulin.

Description

"METHOD OF DETERMINING ANTI-XANTHINE OXIDASE ANTIBODY"

The present invention relates to a method of determining anti-xanthine oxidase antibody, and a test kit suitable for use in such a method. In the 1970's a study was made of the possible association of anti-(whole cows milk) antibodies and heart disease. The results of this study by Davies et al were published in the Lancet i , 1012-1014 (1974) . This article claimed that patients who had suffered a myocardial infarction (MI) had significantly higher levels of such antibodies compared with age- and sex-matched controls who had not suffered MI. The article also indicated that there was an increased chance of death in MI patients who had "high" antibody levels. In the late 1960's and early 1970's a series of papers was published promoting the hypothesis that xanthine oxidase, ingested in cows milk, directly initiated MI. This hypothesis was based on the idea that xanthine oxidase (XO) , a large protein, could be absorbed from the intestine intact and remain enzymatically active. In support of this assertion it was reported that anti-XO antibodies occur in normal human serum and that their levels are elevated in patients with coronary heart disease. These articles include Oster et aj., Am. Lab. 41-47 (1974) . However, this hypothesis was the subject of a Food and Drug Administration report (Carr et aj., Contract No. FDA 223-75-2020 (Div. of Nutrition, Bureau of Foods) FASEB Life Sciences Research Office, Bethesda, US, 1975) . This report attacked many aspects of the Oster paper. As a result of this report and a number of other publications which disputed the absorption of intact XO, the hypothesis was discredited.

The correlation between anti-(whole cows milk) antibodies and heart disease was also discredited in the late 1970"s. Scientific and medical studies which failed to find any correlation were reported by Toivanen et al. Lancet, j L, 205-207 (1975) , Scott et al, Lancet, ii, 125-126 (1976) and Gibney et al, Atherosclerosis 22 151-155 (1980) . Thus the idea of a link between anti-(whole cows milk) antibodies and heart disease was abandoned.

Anti-XO antibodies were reported in the serum of healthy humans by a group in Heidelberg, . Germany. A review of this work is reported by Jarasch et al_, Acta Physiol. Scand. Suppl. 548 39-46 (1986). Anti-XO antibodies were described as being exclusively of the IgG class and as being present in high amounts (up to 8% of total IgG) . However, no association with any disease state could be found.

Work on the anti-(whole cows milk) antibodies showed that the antibodies were directed essentially towards the membrane surrounding the fat droplets in the milk. This membrane, the bovine milk fat globule membrane (BMFGM) is a true biological membrane, derived from the mammary cell. However, studies comparing levels of anti-BMFGM antibodies in male MI patients and in age-matched male controls do not show consistent elevation of anti-BMFGM antibodies in MI patients.

It has now suprisingly been found that anti-BMFGM antibodies are almost exclusively directed towards the enzyme xanthine oxidase, a major protein component of BMFGM. Moreover, it has suprisingly been found that MI patients do not show any elevation of anti-XO antibodies of IgA or IgG type, but that MI patients show a statistically significantly higher level of IgM anti-XO antibodies. Serial assays of IgM anti-XO antibody levels in patients, following MI, indicate that elevated levels of such antibodies are not a consquence of MI. Measurement of the levels of anti-XO IgM antibody in a subject may therefore provide an indicator of the risk of his suffering a myocardial infarction.

The present invention therefore provides a method of determining anti-(xanthine oxidase) IgM antibody (anti-XO IgM) in a sample, comprising carrying out a said determination using xanthine oxidase or a polypeptide fragment thereof capable of competing with xanthine oxidase for binding to the antibody, and an anti-IgM immunoglobulin. The term "XO" will be used hereinafter to denote xanthine oxidase or the polypeptide fragment thereof. XO preferably but not necessarily, shows enzymatic activity in converting xanthine to uric acid, as is shown by xanthine oxidase enzyme denoted by the E.C. No. 1.1.3.22. The xanthine oxidase may be intact enzyme obtained from any cell source e.g., bovine enzyme (SDS-PAGE band 150kD) or enzyme from human or other vertebrate. Cloned intact xanthine oxidase is also included in this defnition. The polypeptide fragment may be derived from intact enzyme or chemically synthesised. Cloned fragments are also capable of being used as xanthine oxidase in the present invention. A polypeptide fragment is sold commercially by Sigma, UK, and this has no SDS-PAGE bands higher than 50kD.

The polypeptide fragment should be capable of binding to the antibody, i.e., it should be capable of competing with xanthine oxidase denoted by the E.C. No. 1.1.3.22 for binding to the antibody. This ability of the fragment may be tested using known techniques e.g., by carrying out comparative assays of a sample known to contain the antibody using xanthine oxidase in the absence and in the presence of polypeptide fragment.

The invention further provides a test kit suitable for use in determining anti-XO IgM, which kit comprises XO, an anti-IgM immunoglobulin and means for determining anti-XO which binds to the XO and which is bound by the anti-IgM immunoglobulin.

It is preferred to carry out the method by contacting the sample with the XO and providing an anti-IgM immunoglobulin so that a XO/anti-XO IgM/anti-IgM immuno¬ globulin conjugate forms if the sample contains anti-XO IgM, and determining whether a said conjugate has formed. The method can be carried out quantitatively, semi-quantitatively or qualitatively. In other words, the level of anti-XO IgM in a sample can be measured. The amount of the XO/anti-XO IgM/ anti-IgM immunoglobulin conjugate can be determined. Alternatively the presence or absence of anti-XO IgM in a sample can be determined. ny appropriate sample from a human patient may be assayed. The sample is typically a serum sample. The sample may be from a patient suspected of being at risk from suffering a MI. The sample may be from a patient who has suffered a MI. Samples may be taken from all individuals in a population, healthy or otherwise, as part of a screening procedure for coronary heart disease. Since myocardial infarction is a common form of heart disease in our society, the applicant's research has been directed mainly with this disease in mind. However the present invention may be used in the diagnosis, prognosis or prediction of a wide variety of other coronary heart diseases such as angina or arrhythmia. Generally samples are taken from men as being more at risk of suffering a MI.

The anti-IgM immunoglobulin is typically antibody specific for the u chain of human IgM. Such antibody may be monoclonal or polyclonal. It is typically antibody from another species, for example an animal such as a goat, rabbit, rat or mouse anti-human IgM immunoglobulin. The anti-IgM immunoglobulin is generally labelled.

The types of assay in which xanthine oxidase is used to capture the anti-XO IgM from a sample involve sandwiching the anti-XO IgM between XO and an anti-IgM immunoglobulin antibody (anti-IgM) . Solution phase assays may be used. Generally, however, the XO or anti-IgM is immobilised on a solid surface and contacted with the test sample^'and with whichever of the XO and anti-IgM is not immobilised on the solid surface. The test sample and whichever of the XO and anti-IgM is not immobilised may be added sequentially or simultaneously. In particular the assay typically involves immobilisation of a first sandwich material, which is either XO or anti-IgM, onto a solid surface. This surface should be capable of being washed. The types of surfaces which may be used include polymers of various types (moulded into microtitre wells; beads; dipsticks; aspiration tips; electrodes and optical devices) , particles (for example latex; stabilized red blood cells (erythrocytes) ; bacterial or fungal cells; spores; gold or other metallic or metal containing sols; organic sols; and proteinaceous colloids; with the usual size of the particle being from 0.005 to 5, for example from 0.1 to 5, microns), membranes (for example of nitrocellulose; paper; cellulose acetate; chemically activated membranes such as Millipore Immobilon (Trade Mark) or Pall Biodyne (Trade Mark) ; and high porosity/high surface area membranes of an organic or inorganic material) .

The attachment of the XO or anti-IgM to the surface can be passive adsorption from a solution of optimum composition which may include surfactants, solvents, salts ^* and/or chaotropes; or by active chemical bonding. Active bonding may be through a variety of reactive or activatable functional groups which may be exposed on the surface (for example condensing agents; active esters;- acid halides; anhydrides; amino, hydroxyl, or carboxyl groups; sulphydryl groups; carbonyl groups; diazo groups; unsaturated groups). After contacting (reacting) the surface bearing the XO or anti-IgM with a test sample, allowing time for reaction and, where necessary, separating or removing the excess of the sample by any of a variety of means (washing, centrifugation, filtration, application of a magnetic field, capillary action) , the captured anti-XO IgM is then reacted with a second sandwich material which is whichever of XO and anti-IgM is not the first sandwich material. The excess of the second sandwich material may be removed if necessary by any of a variety of means (washing, centrifugation, filtration, application of a magnetic field, capillary action) . The second sandwich material is then detected by any means which will give a detectable signal. For example, this may be achieved by use of a labelled molecule, in particular antibody, or particle as described above which will react with the second sandwich material.

The detectable signal may be optical, radioactive or physico-chemical and may be provided either directly by labelling the molecule or particle, especially the anti-IgM, referred to with for example a dye, radiolabel, electroactive species, magnetically resonant species or fluorophore, or indirectly by labelling the molecule or particle with an enzyme itself capable of giving rise to a measurable change of any sort. Alternatively the detectable signal may be due to, for example, agglutination, diffraction effect or birefringent effect occurring if any of the surfaces referred to is in the form of particles. A preferred assay format is the enzyme-linked immunosorbent assay (ELISA) , for example carried out in microtitre plates. Generally a solid support such as the wells of a microtitre plate, is coated with XO and, if necessary, washed with a buffer. The sample under test is then added and incubated with the coated support. Any anti-XO antibodies in the sample bind to the XO. The sample is then removed by washing. The presence of anti-XO IgM is detected by further incubation of the solid support with anti-IgM. Excess anti-IgM is removed. The presence of any anti-IgM which has bound to anti-XO IgM, which has itself bound to the immobilised XO, is detected. Typically the anti-IgM is labelled with an enzyme such as an alkaline phosphatase or horse radish peroxidase. A substrate for the enzyme is added. The presence or absence of the anti-XO IgM is thus revealed.

Other formats which may be employed are any of those suitable for immunoassays including:

(1) A dipstick ELISA with an antibody-coated dipstick, and (2) Sandwich assays using small magnetic particles coated with the first sandwich material, together with the second sandwich material labelled either with coloured particles, or particles with the potential of colour development, or with an enzyme or a fluorescent moiety.

Test kits suitable for use in determining anti-XO IgM antibody in a sample comprise xanthine oxidase or a polypeptide fragment thereof capable of competing with xanthine oxidase for binding to the antibody, an anti-IgM immunoglobulin antibody and means for detecting anti-XO IgM which binds to the xanthine oxidase or polypeptide fragment thereof and which is bound by the anti-IgM immunoglobulin. Specific components of the kits may be as described above. The kits may also comprise one or more components selected from a control, buffer and diluent. Typically the kits comprise a solid support for, for example, the XO. The solid support may be provided coated or uncoated with XO. A kit for use in an enzyme-immunoassay typically includes an enzyme-labelled reagent and a substrate for the enzyme. The enzyme may either be bound to the XO or the anti-IgM or be bound to polyclonal or monoclonal antibody capable of binding to the XO or anti-IgM. The assay of the present invention may be used diagnostically, prognostically or predictively. For example, in the case of MI, the assay may be used as further confirmation that MI has occurred. The applicants have also found indications that the likelihood of death within 6 months of MI correlates with levels of antibody in a sample from a person who has suffered MI. A prognostic use of the assay is therefore contemplated. It is also envisaged that the assay would be used predictively e.g. by screening samples taken from individuals in all states of health, the levels of antibody in a sample giving an indication of the likelihood that the sample donor will in the future suffer coronary heart disease.

The following Example illustrates the invention. In the accompanying drawings. Figures 1 to 3 show the distribution of anti-(xanthine oxidase) antibody titres in 107 male MI patients (40-65 years old) compared with 86 age-matched male controls. EXAMPLE 1. Raτnp1_._. Serum samples from male MI patients and controls, were obtained from two studies.

In Study 1, samples were collected, in the course of the Caerphilly Collaborative Heart Disease project (J. Epidemiol.Commun.Health, 1984; .38., 259-262) and other Medical Research Council studies (Burr et al. Lancet, 1989; ii, 757-761) from 100 male patients with MI (49-64 years) living in Cardiff, Newport and the South Wales Valleys and from 100 male controls (49-64) living in Caerphilly. In the case of patients with MI, samples were taken 3-5 weeks following MI. In Study 2, samples were taken, within 24 h, from 107 male patients with MI (40-65 years) and from 86 male controls (40-65 years) consecutively admitted to the Coronary Care Unit, Royal United Hosptials, Bath. Control subjects were admitted with non-cardiac chest pains.

In both studies, MI was defined on the basis of clinical history, serial electroradiograms and cardiac enzymes; controls showed no present or past evidence of coronary heart disease. All serum samples were stored at -80°C prior to assay.

2. Antigens

BMFGM was prepared as described by Mather and Keenan (J. Membrane Biol. .21 65-85 (1975)) and a crude lipid extract was obtained essentially as described by Feizi and Childs (Biochem J. 1978: 173 245-254). Polar and non-polar lipid fractions were prepared from the crude lipid extract according to the method of Yu and Ledeen (J. Lipid Res. 1972; .13. 680-686) and incorporated into phosphatidylcholine-cholesterol liposomes as described by Feizi and Childs.

Xanthine oxidase was prepared from bovine milk according to the method of Nakamura and Yamazaki (J. Biochem. 1982; £2 1279-1286) and showed a single major ban (Mr 150kD) together with a minor (5%) band (Mr 145 kD) on sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis (SDS-PAGE) . This purified product can be stored frozen for several weeks with no change in its effectiveness in an assay. Alternatively, commercial enzyme (XO-2, Biozyme Laboratories, Blaenavon, Wales) was further purified by passage down a folate affinity column.

3. ELISA'S

BMFGM (0.2mg/ml), xanthine oxidase (0.02mg/ml) or dried milk powder (0.15mg/ml) in 50 mM sodium carbonate, pH 9.6, were added (100 ul) to each well of a polystyrene microtitre plate and incubated overnight at 4°C. Wells were washed by incubation at room temperature (20 in) with three successive portions (250 ul/well) of phosphate-buffered saline (PBS) containing 0.1% Tween 20 and then incubated with serum samples (100 ul, diluted 100-fold in PBS) and 0.05% Tween 20 for lh at 37°C. The plates were washed as above and goat anti-[human Ig (G,A or M)]IgG-horse radish peroxidase conjugate (diluted 1000 fold in PBS - 0.05% Tween) was added to each well (100 ul/well) . The wells were incubated for 2h at room temperature, washed three times as above and incubated with staining buffer (100 ul/well) for 20-30 min until a blue colour developed. The reaction was stopped by addition of 1M sulphuric acid (50 ul) and, after 10 min, absorbance was recorded at 450nm. Staining buffer contained 1% 3,3' ,5,5'-tetramethylbenzidine in dimethylsulphoxide, diluted 100-fold in 0.1M sodium acetate/citric acid, pH 6.0; 30% hydrogen peroxide (0.lul/ml) was added immediately before use.

Each plate (96 wells) included 10 dilutions of a standard pooled reference serum, so allowing construction of a standard curve. Relation of absorbance to this curve gives the titre of a test sample as a percentage of the standard. Each plate also included a sample of a second standard whose titre is, say, 50% of the first standard; this titre must be essentially (+8%) constant for the plate to be valid.

Titres of test samples are thus always relative to the laboratory pooled standard serum.

Titres are therefore quoted as percentage absorbance shown by standard high titre pooled human serum, assayed on the same plate. All values are means of triplicate determinations and, in repeat experiments, showed variations less than 8% standard. All comparisons of MI and control samples were done blind, at the same time, by a single operator. In inhibition experiments, standard serum was incubated with potential inhibitor for 2h at 37°C prior to assay. In the case of lipid inhibitors, these were incorporated into liposomoses prior to incubation (Feizi & Childs) . The liposomes were removed by centrifugation

(100,000g) immediately before assay. Liposomes alone served as a control for studies of lipid inhibition.

4. Electrophoresis and T-nmnnnblottinq

SDS-PAGE (10% acrylamide) was performed according to Laemmli (Nature 1970; 227 680-85) and the separated proteins were transferred to nitrocellulose as described by Tsang et al (Methods in Enzymology 1983; .92. 377-91). Interaction of human immunoglobulins with the transferred proteins was detected according to the method of Towbin et al. (Proc. Natl. Acad. Sci. USA 1979; 76 4350-54) using goat anti-(human immunoglobulin) IgG, conjugated to horse radish peroxidase, as the second antibody.

5. Statistical Analyses

Antibody levels in MI and control sera were compared by using the Kolmogorov-Smirnov two tailed test. This test compares cumulative frequencies of a given parameter between two samples of population. It identifies the point of maximum difference between the samples and, from this difference, estimates the probability that the null hypothesis holds true; ie. that the two samples have been drawn from the same population.

The correlation between two different antibody titres measured on a single set of serum samples was judged by the Rank correlation test. 6. Results

The identification of BMFGM as the major antigenic target in whole dried cows' milk was confirmed by comparing, by ELISA, levels of IgG anti-BMFGM antibodies and anti-(cows milk) antibodies in 35 human serum samples chosen at random from the control subjects in Study 1. Correlation (r=0.98; p<0.001) between the two sets of titres was highly significant.

In an attempt to define further the target(s) of human anti-BMFGM antibodies, polar and non-polar lipid extracts of BMFGM were tested as competitive inhibitors in the ELISA for anti-BMFGM antibodies. Inhibition was, however, much less than that shown by BMFGM itself (Table

1).

In view of the relative lack of antigenicity of the lipid component, the proteins of BMFGM were investigated. Electrophoresis (SDS-PAGE) of BMFGM proteins, followed by immunoblotting showed that, of the major BMFGM proteins, only those with Mr 150kD and 145 kD, corresponding to the enzyme, xanthine oxidase, bound human immunoglobulin.

The identification of xanthine oxidase as the major target of human anti-BMFGM antibodies was confirmed by using the purified enzyme as inhibitor in the ELISA for anti-BMFGM antibodies, as was done for BMFGM lipid fractions. As can be seen in Table 1 below, the tested concentration of xanthine oxidase showed 80% inhibition of the assay, comparable with the inhibition shown by BMFGM itself and very much greater than that shown by the other milk proteins, casein or bovine serum albumin. Further confirmation of the relationship of anti-BMFGM and anti-(xanthine oxidase) antibodies was obtained by comparing titres in 198 of the 200 serum samples of Study 1. Correlations between the two sets of titres were highly significant (p<0.001) for IgG, IgA and IgM antibodies.

Figures 1, 2 and 3 show the distributions of anti- (xanthine oxidase) antibody titres in serum samples from the male MI Patients and controls collected in the course of Study 2. Differences between the two populations were not significant (p > 0.05) for IgG or IgA antibodies (Figs 1 and 2) but IgM anti-(xanthine oxidase) antibody titres (Figure 3) were significantly (p < 0.001) higher in the MI patients. Two further comparisons of male MI patients and controls (57 v 57 and 50 v 50 respectively, randomly selected from the above subjects) both showed significant (p<0.001) elevation of IgM anti-(xanthine oxidase) antibodies in MI patients but no significant differences (p<0.05) in the corresponding IgG or IgA antibodies.

The elevated levels of IgM anti-(xanthine oxidase) antibodies in patients with MI, compared to controls, do not simply reflect higher overall levels of IgM. This was shown by assay of total IgM in a random sample of 50 pairs of subjects from Study 2. There was no significant difference (p > 0.05) between MI samples and controls, nor was there any correlation (p > 0.05) between IgM anti-(xanthine oxidase) antibody levels and total IgM levels in either the MI samples or in the controls.

Assay of IgM anti-(xanthine oxidase) antibodies in serial serum samples, taken up to 6 months following MI from 20 patients chosen at random from Study 2, showed no consistent patterns of change, suggesting that the higher antibody levels in MI patients are not a consequence of the MI itself.

In the study of 107 patients in Bath, 12 died within 6 months of MI and 81 are known to have survived. Thus the median level of IgM anti-(xanthine oxidase) antibodies in the patients who died (68% standard) was higher than that of the survivors (49% standard) . These values are to be compared with median values of 58% standard for all 107 patients with MI and 29% standard for the 86 control subjects. This points to an association of IgM anti-(Xanthine oxidase) antibody level with increased mortality following MI and suggests that the present assay is likely to be of use as a clinical test of prognostic importance. Preparation and use, as antigen, of human xanthine oxidase

Xanthine oxidase was also prepared from human milk. Human milk was treated with detergent, fractionated using ammonium sulphate and stored at -20^βC as described by Zikakis et al [In: Instrumental analysis of foods" (ed. G. Charalambous) pp 243-303 Vol 2, Academic Press, Orlando FL, 1983]. The resultant 12225g supernatant was dialysed against Buffer A (0.2M Na₂HP0 , ImM EDTA and 1.25mM sodium salicylate, pH 6.00) and applied to a column of calcium phosphate gel equilibrated at 4^βC, in the same buffer. The column was washed with Buffer A until no more protein (280 n ) was eluted, after which xanthine oxidase was eluted with Buffer A containing 5% (w/v) ammonium sulphate. Enzyme-containing fractions were dialysed against a suitable buffer, overnight, to remove ammonium sulphate.

Purified enzyme showed two major bands (Mr 160 kD and 155 kD) on SDS-PAGE, a ratio of A so to A ₅₀ of 5.1 ± 0.2 (n=3) and specific activity of (1.42 ± 0.32) X lθ~³ μmol uric acid.min.mg.

Use of human enzyme, as antigen, in ELISA of anti- (xanthine oxidase) antibodies, in a series of samples of human serum, showed titres that ranked exactly with those obtained by using bovine enzyme as antigen.

TABLE 1

INHIBITION OF ELISA FOR ANTI-BMFGM ANTIBODIES IN POOLED HUMAN SERUM BY BMFGM COMPONENTS

INHIBITOR % INHIBITION + SEM(n) (lmg/ml)

BMFGM

Polar BMFGM lipids + liposomes Non-polar BMFGM lipids + liposomes Liposomes alone Xanthine oxidase Casein Bovine serum albumin

Claims

1. A method of determining anti-(xanthine oxidase) IgM antibody (anti-XO IgM) in a sample, comprising carrying out a said determination using xanthine oxidase or a polypeptide fragment thereof capable of competing with xanthine oxidase for binding to the antibody (XO) , and an anti-IgM immunoglobulin.

2. A method according to claim 1 comprising contacting the sample with XO and providing an anti-IgM immunoglogulin so that a XO/anti-XO IgM/anti-IgM immunoglobulin conjugate forms if the sample contains anti-XO IgM, and determining whether a said conjugate has formed.

3. A method according to claim 1 or 2 in which the anti-IgM immunoglobulin is labelled.

4. A method according to any one of the preceding claims, wherein an enzyme-linked immunosorbent assay for the anti-XO IgM is performed.

5. A method according to any one of the preceding claims, wherein the sample is a sample from a patient suspected of being at risk from suffering a myocardial infarction.

6. A method according to any one of claims 1 to 4, wherein the sample is a sample from a patient who has suffered a myocardial infarction.

7. A method according to any one of the preceding claims, in which the level of anti-XO IgM thus determined is compared with the level of anti-XO in a reference sample.

8. A test kit suitable for use in determining anti-XO IgM, which kit comprises XO, an anti-IgM immunoglobulin and means for determining anti-XO IgM which binds to the XO and which is bound by the anti-IgM immunoglobulin.

9. A test kit according to claim 8, wherein the said means is an enzyme label on the anti-IgM immunoglobulin.

10. A test kit according to claim 8 or 9, wherein the XO is bound to a solid, support.