WO1989005976A1

WO1989005976A1 - Diagnostic method for primary biliary cirrhosis and antibodies suitable to be used in the method

Info

Publication number: WO1989005976A1
Application number: PCT/SE1988/000698
Authority: WO
Inventors: Buck Dean Nelson; Ib Oluf Mendel-Hartvig
Original assignee: Pharmacia Ab
Priority date: 1987-12-22
Filing date: 1988-12-21
Publication date: 1989-06-29
Also published as: AU2910589A; SE8705137D0

Abstract

Immunoassay for detecting anti-mitochondrial autoantibodies employing as the target antigen for the autoantibodies a substantially pure form of the 75 KD subunit of Complex I. Also described are monoclonal antibodies specifically directed against the subunit.

Description

This invention relates to an immunoassay in which circulating anti-mitochondrial autoantibodies (AMA) in a sample from a biological fluid of a patient are detected and correlated to primary biliary cirrhosis (PBC) . The basis of our invention is that the target antigen for the AMA appearing in PBC has been identified by us as being a subunit of Complex I (NADH-ubiquinone reductase) .

Circulating -anti-mitochondrial autoantibodies are an almost exclusive diagnostic feature of PBC (Doniach et al. 1966) . Several attempts have been made to identity the mitochondria! antigen against which AMA are directed in PBC (Mendel -Plartvig et al. 1985, 1986 and 1987; Frazer et al. 1985; Schulrhfiss et al. 19S3; Ishii et al . 1985; and Ba m & Palmer 1985) . To date, the only polypeptide with a known structure and function to be identified as a itochondrial antigen in this disease is the adenine nucleotide translocator protein (ANT) (Schultheiss et al. 1983) . However, antibodies against ANT are not unique to PBC (Mendel-Hartvig et al. 1986) .

Though not identified, the major PBC-associated antigen, classified as the M-2 antigen by immunofluorenscence (Berg et al. 1981 and 1982) , has been characterized in some detail (Mendel-Hartvig et al. 1985 and 1987; Frazer et al. 1985; Baum & Palmer 1985; Berg et al. 1982; Berg et al. 1969; Baum & Berg 1981; and Karlsson-Parra et al. 1987) . This major mitochondrial antigen has been assigned an apparent molecular weight of 70 kD (Mendel-Hartvig et al. 1985 and 1987) or 75 kD (Frazer et al. 1985) . This antigen is sensitive to sulfydryl reagents (Mendel-Hartvig et al. 1985 and 1987; and Sayers et al. 1981) and trypsin (Frazer et al. 1985) , and thus fulfills the criteria of the -M-2 antigen (Mendel- Hartvig et al. 1985; Baum & Palmer 1985; Berg et al. 1981; Berg et al. 1982; Berg et al. 1969; and Baum & Berg 1981) . The antigen is an inner membrane polypeptide, but its location within the inner membrane has not been established.

Complex I has been isolated and characterized quite a long time ago (Hatefi et al. 1967; and Ragan et al. 1982) . By passing a crude mitochondrial membrane preparation through an immunoadsorbent column of im unoglobulins from PBC patients ' sera coupled to Sepharose 4B and subsequent elution it has been possible to obtain an AMA antigen preparation 100 times more active in a complement fixation test (Ben-Yoseph et al. (1974)) .

Recently a test for AMA has become commercially available (Pharmacia PBC-IgG/IgM EIA, Pharmacia AB, Sweden. Brochure available cn request) . The test utilizes a F-. -ATPase prep¬ aration containing low amounts of the hitherto unidentified AMA target antigen as an impurity. The AMA levels found with the test correlate quite well with the clinical findings for PBC, although it is a matter of course that the value of the test would be improved after an identification and sub¬ sequent employment of the pure form of the target antigen.

For a recent review on the PBC antigen see Baum, H. & Palmer, C. (1985) .

We have now proved that the major PBC antigen consists of one single peptide that is identical to the 75 kD subunit of Complex I. This invention accordingly proposes as its main characteristic feature that in an immunoassay for AMA a protein preparation substantially enriched in the 75 kD sub¬ unit is employed as the target antigen for AMA. In other words, the invention proposes to use Complex I or the 75 kD subunit in a form that is substantially free from other antigenic polypeptides. In particular it is important for the preparation not to contain non-mitochondrial poly¬ peptides that are known to give rise to circulating antibodies in man. In the preparation used according to the invention the 75 kD subunit amounts to more than 4 %, such as more than 80 _ or more than 95 %, of other antigenic mitochondria! proteins and polypeptides in the preparation (percentage given as w/w) . Fragments and derivatives of the 75 kD subunit exhibiting PBC-specific determinants can be used equivalently in the invention. In order to establish whether or not a certain fragment or derivative can be used, the tests required will be analogous to those for the 75 kD subunit of Complex I in the experimental part of this specification.

The target antigen of sufficient purity can be prepared as in the patent examples. Target antigens of extremely high purity are obtainable by means of affinity purification using antibody preparations (polyclonals as well as monoclonals) specifically directed against the 75 kD subunit.

The term "derivative" above relates to insolubilized forms of the 75 kD subunit or forms in which the 75 kD subunit has been covalently coupled to other compounds facilitating the performance of an immunoassay. Such compounds are well-known in the immunoassay field and encompass i.a. compounds bestowing improved detectability on the subunit.

Accordingly, the invention is a method for the specific measurement of anti-mitochondrial autoantibody (= AMA) in a human biological fluid containing humoral antibodies by the use of an immunoassay comprising the steps of

(i) incubating the 75 kD subunit of Complex I in a substan¬ tially pure form with a sample of said fluid suspected of containing said autoantibody for the formation of an immune complex between said autoantibody and said subunit in a way such that the amount of complex formed is a measure of the amount of said autoantibody in the sample, and (ii) measuring directly or indirectly the amount of the immune complex formed by the use of a labelled reactant capable of being incorporated in said complex, and

(iii) correlating of the amount of immune complex formed with the amount of AMA in the sample; an increased amount of AMA relative to the mean for samples from the normal population indicating that the individual from whom the sample is derived suffers from PBC.

A direct measurement means that the amount of the complex is quantified. An indirect method means that after complex formation the amount of uncomplexed 75 kD subunit is quantified,

The sample in question may be a blood sample, such as a whole blood, plasma or serum sample. Of course the sample under investigation is compared with samples treated in the same way.

A large number of general types of immunoassay methods are known per se. The artisan who is acquainted with these methods will readily recognize those among them to which the measurement of AMA can be applied.

The method of the invention utilizes the 75 kD subunit of Complex I as a target antigen for the formation of an immune complex containing the anti-mitochondrial autoantibody and the subunit. The complex formation constitutes qualitative and quantitative indication means for demonstrating in the sample the presence and amount of the anti-mitochondrial autoantibody. Detection and quantification are facilitated by the use of an immune reactant labelled with an analytically detectable group and capable of being incorporated in the complex. The amounts of reactants used are chosen such that the amount of labelled reactant bound to its immunological counterpart or the amount of labelled reactant remaining free, in a non-complexed state, will be indicative of the amount of the antibody sought. Immunoassay methods may be subdivided into "homogeneous" and "heterogeneous" methods. In the case of the homogeneous methods, determination of (assay for) a labelled reactant is carried out without any physical separation of complex-bound labelled reactant from the non-complex-bound form. The homogeneous methods use markers which will undergo a change in their activity depending on whether or not they are complex-bound; in this manner it is possible to measure the signal from a reaction mixture containing the marker in both forms, and to draw conclusions from the value obtained as to the amount of the substance looked for. The heterogeneous methods involve physical separation of complex-bound labelled reactant from the non-complex-bound form; there is thus no requirement that the marker should undergo any changes in activity. The separation is feasible because one of the two forms of labelled reactant has been or is being bound to a solid phase which is readily separable from the liquid phase. Assays for the analytically detectable group are then carried out on one or both of the two phases.

From another point of view, immunoassay methods may be subdivided into "competitive" and "non-competitive" methods. In a competitive method the arrangement is such that two reactants having a common epitope are allowed to compete for an insufficient number of homologous binding sites on an immunological counterpart. Usually the systems are chosen such that competition occurs between the substance assayed for and a variant form thereof which is labelled or bound to a solid phase. The amount that binds to the immunological counterpart is a measure of the substance to be detected. In a non-competitive method, the reactants chosen are such that no competition can occur. As examples of non-competitive methods may be mentioned in particular the so-called "sand¬ wich" systems. According to a third mode of subdivision, the methods comprise precipitation methods on one hand and non-precipi¬ tation methods on the other hand. When precipitation methods are carried out the first immune reactions performed will proceed in a homogeneous liquid phase, whereupon the resultant immune complex is precipitated with the aid of a precipitant, e.g. polyethylene glycol, antiserum or solid-phase-bound antibody (care being taken that said antiserum or antibody is not directed against the reactant which is labelled) .

A fourth mode of subdivision classifies the methods according to their marker group; thus there are radio-, enzyme-, fluorescence-, chemiluminescence-, enzyme-substrate-immuno- assay etc. methods, including biotin-avidin reagents as the marker group.

With specific reference to the invention, particularly useful systems are heterogeneous sandwich type assays employing insolubilized anti-antibodies directed against the Fc-portions of human immunoglobulins, such as in IgG or IgM, or insolubilized 75 kD subunit. The insoluble anti-antibody is combined with labelled 75 kD subunit and the insoluble subunit with labelled anti-antibody. Insolubilization means that the reactant in question is physically adsorbed or covalentl or biospecifically bound to a substance that is insoluble in the aqueous media in which the antigen-antibody reaction is to take place (so-called solid phase) . The bond between the immune reactant and the solid phase has to resist the normal washing procedure employed in immunoassay methods.

In connection with the scientific work related to the invention, and by techniques known per se, we have produced monoclonal antibodies specifically directed against the 75 kD subunit. Such monoclonals are new and can be used as immune reagents in the invention and as purification means for the target antigen. Most probably they will give rise to high quality tests for anti-mitochondrial autoantibodies. They can be used in a soluble or insoluble form, depending on the assay system chosen. See above. These monoclonal antibodies are one aspect of the invention.

The reaction conditions selected are those commonly employed for immune reactions, i.e. aqueous media buffered to a pH which will normally be within the range of 5.0-9.0. Tempera¬ tures are usually maintained in the range of from +4 to +40 C. Additions may be made of buffers and detergents tl will not interfere with the immune reaction or its result,

The invention is defined by the appended claims which are an integral part of the specification. The invention will now be illustrated by a relevant part of the scientific work on which it is based.

Example 1 - Identification

A. Materials and methods

Protein preparations. Complex I and the iron protein sub- fraction of Complex I (Ragan et al . 1982; Hatefi & P.ieski 1967) , beef heart mitochondria (Lee & Ernster 1967) and submitochondrial particles (Lee & Ernster 1967) were prepared as described. Protein was determined by the method of Petersson 1977 or by the BCA protein assay (Pierce, USA) .

Immunosorbent purification

A high titer AMA serum was precipitated at 50 % saturation of ammonium sulfate for 1 h at room temperature. The pellet formed by centrifugation at 15 000 g for 5 min was dissolved in PBS (phosphate buffered saline pH 7.4) and the precipi¬ tation step was repeated. The final pellet was dissolved in coupling buffer (0.5 M NaCl, 0.1 M NaHCO-) and the immuno- globulin fraction was desalted on a PD-10 column (Pharmacia AB, Uppsala, Sweden) equilibrated with the same buffer. Coupling of the immunoglobulin fraction to CNBr activated Sepharose was conducted as described by the manufacturer (Pharmacia AB, Uppsala, Sweden) . Mitochondria (2.6 mg/ml) were dissolved in 1 % Zwittergent 3-14 (Pierce, USA) in PBS, 0.5 M NaCl overnight at 4 C prior to loading the i munosorbent column. The column was washed with 5 volumes of PBS, 0.5 M NaCl, 0.1 % Zwittergent 3-14 and the bound proteins were eluted with 0.5 % SDS in PBS, 0.5 M NaCl. The yield was 2 ,ug/mg of mitochondrial protein.

Electrophoresis. SDS-polyacrylamide gel electrophoresis was run on gradient (7-18 % pol acrylamide) gels using -the buffers of Laemmli 1970. Samples (1 mg/ml) were boiled 1 min in 2 % SDS, 2 % 2-mercaptoethanol and 10 M Tris-HCl pH 6.8 prior to electrophoresis. In some experiments 2-mercapto- ethanol was excluded (see B. Conclusions based on the results) . Two dimensional electrophoresis was run as described by C'Farreii 1975. The molecular weights were calculated from known standards: lactalbumin 14.4 kD, soybean trypsin inhibitor 20.1 kD, carbonic anhydrase 30 kD, ovalbumin 43 kD, bovine serum albumin 67 kD, phosphorylase b 94 kD and alpha unit of ATPase 55 kD.

Immunoblots. Immunoblots (western blots) were performed essentially as described by Towbin et al. 1979. The dried nitrocellulose sheets were exposed at 70 C to Cronex 4 X-ray film using a Cronex extra life intensifying screen. Preparation of rabbit antisera against the 75 kD subunit of Complex I was performed as described by Cleeter et al. 1986.

Enzyme linked immunoassay (ELISA) . Microtiter wells (Nunc, Denmark) were coated overnight at room temperature with 100 ,ul of SMP (25 ,ug/ml) in PBS pH 7.4. The wells were washed four times with washing buffer (Pharmacia PBC-IgG/IgM EIA kit, Pharmacia AB, Sweden) . A high titer PBC-AMA serum was preincubated over night (room temperature) with increasing amounts of SMP or purified Coniplex I. 100 ,ul of the incubate or patient sera diluted 1/1000 was added to the wells and the ELISA was run as described by the manufacturer (Pharmacia AB)

Patients and controls. Sera from 8 patients with liver biopsy-proven PBC and fulfilling the criteria of De Groote et al. 1968 were analyzed. All sera were positive for AMA with the conventional indirect immunofluorenscence assay (Munoz et al. 1981) . Sera from 2 healthy blood donors, 2 patients with chronic active hepatitis (CAH) and 2 patients with systemic lupus erythematosus (SLE) served as controls.

B. Conclusions based on the results

Several types of evidence support the conclusion that the 75 kD PBC antigen is a subunit of Complex I. First, the 75 kD PBC antigen is present in SMP, Complex I and the IP subfraction of Complex I . Thus all preparations retain the antigen during purification. Secondly, the 75 kD PBC antigen in all three preparations is reactive in western blotting only in the presence of mercaptoethanol. This unusual property suggests a strong similarity, if not identity, of the antigen in the three preparations. It also indicates that the antigen in the three preparations fulfills the definition of the PBC-specific M-2 antigen (Mendel-Hartvig et al. 1985; Baum & Palmer 1985; Berg et al. 1981; Berg et al. 1982; Berg et al. 1969; and Baum & Berg 1981) . Thirdly, the isoelectric point (6.4) of the 75 kD PBC antigen in SMP and in isolated Complex I is identical to that of the major subunit of Complex I and of the IP subfraction. Furthermore, rabbit antiserum raised against the purified 75 kD subunit of Complex I reacts with the same antigen as the PBC serum, both in two- and one-dimensional gels. Finally, rabbit anti-75 kD antiserum also reacted with immunoaffinity purified PBC-antigen. Competition ELISA experiments did not detect any enrichment of the PBC antigen in the IP subcoroplex which was commensurate with the purification of the complex, even though complete cross reaction between the antigens in submitochondrial particles and the IP complex was obtained. It is, of course, possible that ELISA and western blotting do not detect the same epitopes. However, since a good correlation was obtained between ELISA and the amount of the 75 kD antigen detected by western blotting of the different inner membrane protein complexes (Mendel-Hartvig et al. 1985) , a more reasonable explanation for the lack of purification is a partial loss of antigenicity in the IP subcomplex. Indeed, as shown here and elsewhere (Mendel-Hartvig et al. 1985 and 1987) , immuno- reactivity of the 75 kD antigen is dependent upon the oxidation/reduction state of sulfydryl groups and is probably sensitive to conformational alterations in the polypeptide.

Studies of the PBC antigen by immunoblot analysis have so far mainly used mitochondrial fractions denatured by heating in SDS (Mendel-Hartvig et al. 1985; Mendel-Hartvig et al. 1987; Frazer et al. 1985; Baum & Palmer 1985; Baum & Berg 1981; Ishii et al. 1985; Karlsson-Parra et al. 1987) . If boiling was omitted an otherwise completely undetectable antigen at 70 kD appeared as the major AMA binding antigen. It was reactive with sera from all eight PBC patients but not with sera from patients with SLE, CAH, RA and normal blood donors. Thus, the 70 kD antigen seems to be PBC-specific.

The heat-sensitive, 70 kD antigen also needs to be reduced in order to retain its immunoreactivity in immunoblots. This similarity in behaviour of the 70 kD antigen and the 75 kD antigen suggests a close relationship between the two proteins . The most probable explanation is that the smaller antigen is a proteolytic digest of the 75 kD subunit. Binding of anti-75 kD rabbit antiserum to the heat-sensitive 70 kD antigen further supports this conclusion. Two other antigens occasionally binding AMA at 60 kD and 40 kD also reacted with the rabbit anti-75 kD antiserum, thus suggesting two additional digestion products from the 75 kD subunit. The two latter antigens co-purified v/ith the 75 kD antigen during i munosorbent chromatography and were visible upon analysis by SDS-electrophoresis. Despite the strong immune reaction at 70 kD, no protein was stained by Coo assie blue at this molecular weight. This may have two explanations; firstly the 70 kD protein stains poorly with Coomassie blue, and secondly, there was only a minute amount of protein present but the immunoreactivity was drastically increased as a result of the proteolysis.

Table 1 suggests the possible relationship between the different digestion products of the 75 kD subunit and also indicates the different antibody reactions. The table views most, if not all, of the PBC-specific antigens as degradation products of the parental 75 kD antigen, i.e. the 75 kD subunit of Complex I .

Table 1

Reactivity of different fragments of the PBC antigen (75 kD) with AMA in patient sera and with rabbit antibodies against the 75 kD subunit of Complex I.

Fragment Antibody binding

75 kD (native polypeptide) most AMA sera, rabbit anti-75 kD

70 kD (highly immunoreactive) most AMA sera, rabbit anti-75 kD

60 kD some AMA sera, rabbit anti-75 kD

40 kD some AMA sera, rabbit anti-75 kD

25-33 kD rabbit anti-75 kD

Example 2 - Immunoassay utilizing the 75 kD subunit of Complex I in pure form as the target antigen

Microtiter wells were coated as in example 1 except that SMP was replaced with the immunoscrbent purified 75 kD subunit (approx. 0.7 ,ug/ml) of Complex I.

The wells were coated with 100 ,ul aliquots of material overnight at +4 C, then washed four times with 250 ,ul of PBS-0.5 % Tween 20. Sera were diluted 1/1000 with PBS-0.5 % Tween and 100 ,ul aliquots were added for 60 min at room temperature. The wells were then washed four times with PBS-0.5 % Tween 20, whereupon 100 ,ul aliquots of beta- galactosidase-conjugated rabbit anti-human IgG/IgM (Pharmacia, Sweden) were added for 60 min at room temperature. The wells were washed four times with PBS-0.5 % Tween 20 and 100 ,ul of substrate solution containing ortho-nitrophenyl-beta- galactoside (Pharmacia, Sweden) was added for 30 min. The reaction was stopped with 100 ,ul of 0,66 M Na-CO- and the absorbance read at 405 nm.

The test procedure was essentially the same as Pharmacia PBC-IgG/IgM EIA (Pharmacia AB, Sweden) .

Table 2

Results obtained with pure form of the 75 kD subunit of Complex I as the target antigen.

Sera/sample

Buffer only Normal human serum Serum (SLE) Serum-2 (PBC) Serum-4 (PBC)

The results show that the 75 kD subunit of Complex I clearly discriminates PBC-sera from other sera. Preparation of monoclonal antibodies

BALB/C mice were immunized i.p. and s.c. with 100 ,ug of SMP in Freund's complete adjuvant and boosted i.p. with three times 100 ,ug in saline over a period of five weeks. Three days after the last injection, spleen cells were harvested and fused with Sp 2/0 myeloma cells according to Galfrie et al. (1977) with some minor modifications. Spleen cells were fused with myeloma cells at a ratio of 1:2 using PEG 1500 (BDH limited pool, England) . After fusion the cells were distributed to 500 wells of microculture plates and selected by HAT medium. Culture supernatants were screened for binding activity by enzyme linked immunoadsorbent assay (ELISA) (Carlsson et al. 1972) using antigen coated microtiter plates (Mendel-Hartvig et al. 1985) . Positive wells were then cloned by limiting dilution as described elsewhere (Carlsson et al. 1985) . Monoclonal antibodies were purified from the cell free supernatant by a two step procedure utilizing cation-exchange chromatography and gel filtration (Carlsson et al. 1985) .

REFERENCES

Baum, H., and P.A. Berg. 1981. Semin. Liver Disease. 1: 309.

Baum, H., and C. Palmer. 1985. Molec. Aspects Med. 8: 201.

Ben-Yoseph, Y. et al. 1974. Immunology 26: 311.

Berg, P.A. et al. 1969. Clin. exp. Immunol. 4: 511.

Berg, P.A. et al. 1981. Lancet ii: 804.

Berg, P.A. et al. 1982. Lancet ii: 1423.

Carlsson, H.E. et al. 1972. Infect. Immunity 6:703. Carlsson, M. et al. 1985. J. Immun. Meth. 79:89.

Cleeter, M.W.J. et al. 1986. Biochem. J. 227: 467.

DeGroote, J. et al. 1968. Lancet ii: 626.

Doniach, D. et al. 1966. Clin. exp. Immunol. 1: 237.

Frazer, I.H. et al. 1985. J. Immunol. 135: 1739.

Galfrie, G. et al. 1977. Nature, Lond. 266:550.

Hatefi, Y., and J.S. Rieske. 1967. Methods Enzymol. 10: 235.

Ishii, H. et al. 1985. Immunol. Letters. 9: 325.

Karlsson-Parra, A. et al. 1987. Int. Arch. Allergy. Appl. Immunol. 83: 256.

Laemmli, U. 1970. Nature 227: 680.

Lee, C.P., and L. Ernster. 1967. Methods Enzymol. 10: 543.

Mendel-Hartvig, I. et al. 1985. Clin. exp. Immunol. 62: 371.

Mendel-Hartvig, I. et al. 1986. Clin. exp. Immunol. 66: 399.

Mendel-Hartvig, I. et al. 1987. Int. Arch. Allergy. Appl. Immunol. 83: 265.

Munoz, L.E. et al. 1981. Is mitochondrial antibody diagnostic of primary biliary cirrhosis? Gut. 22: 136.

O'Farrell, P.H. 1975. J. Biol. Che . 250(10) : 4007.

Peterson, G.L. 1977. Anal. Biochem. 83: 346. Ragan, C.I. 1982. Biochemistry (USA) 21: 2518.

Sayers, et al. 1981.

Schultheiss, H-P et al. 1983. Clin. exp. Immunol. 54: 648.

Towbin, H. et al. 1979. Proc. Natl. Acad. Sci. USA. 76: 4350.

Claims

1. Method for detecting circulating anti-mitochondrial autoantibodies (= AMA) in a biological fluid of a patient by incubating a sample of said fluid with a target antigen (reactant 1) against which the AMA are directed, to thus form an immune complex containing said target antigen and said AMA, the formation of said immune complex being indicative of the presence of AMA in said sample and said immune complex being measured by the use of a labelled immune reactant capable of being incorporated into the complex, said method being c h a r a c t e r i z e d in the target antigen being the 75 kD subunit of Complex I in a substantially pure form.

2. Method according to claim 1, c h a r a c t e r i z e d in that the 75 kD subunit, when being incorporated in the immune complex, binds to an anti-mitochondrial autoantibody which in turn is reacted with an anti- i munoglobulin antibody (reactant 2) specific for said autoantibody, either said subunit or said anti-immuno- globulin antibody being labeled or capable of being labelled so that said complex can be detected.

3. Method according to claim 2, c h a r a c t e r i z e d in that said subunit or said anti-immunoglobulin antibody is bound or is capable of being bound to a solid phase while the other is labelled or capable of being labelled so that the complex can be physically separated from the remaining reactants and quantitated with the aid of the labelled reactant.

4. Antibody specifically directed against the 75 kD subunit of Complex I and prepared by monoclonal technique