WO1990012813A1 - Regulatory proteins - Google Patents

Abstract

The present invention relates to regulatory proteins which bind to the same antigenic determinants on human cytokines as auto-antibodies to human cytokines, the use thereof for producing a pharmaceutical preparation for regulation of immunological mechanisms in humans, the use of cytokine, cytokine analogues or cytokine fragments exhibiting a human functional determinant, for producing a preparation for regulation of the auto-antibody activity in humans, an assay kit containing as a functional constituent such regulatory protein, a pharmaceutical preparation containing such regulatory protein, and a process for the preparation of such regulatory protein.

Description

Title: Regulatory proteins

The present invention relates to regulatory proteins for use in therapeutical or prophylactic treatment of the immunological mechanisms in humans, a process for the preparation thereof as well as an assay kit for determining the cytokine or auto-antibody level in body fluids.

Hormones, which are formed by immune competent cells, cytokines, are important mediators in the response of the organism to pathological changes: inflammation. In addition to coordinating the local inflammatory response, it has now also become known that the cytokines are directly responsible for the systemic effects of the inflammation: fever, indisposition, metabolic changes, etc. Thus, the cytokines have influence on most cells in the organism, and some cytokines, f. ex. γ-interferon, tumor necrosis factor, interleukin 1, a. o., have been found to damage normal healthy cells and tissue directly, in concentrations wh.ich theoretically may be obtained in vivo.

Consequently, there must be a fine "im uno-physiological" balance between the useful effects of the cytokines, especially in the local inflammatory area, and their simultaneous potent systemic/toxic effects. The underlying homeostatic mechanisms are partly unknown. It has for some years been known to use interferons in human cancer therapy. It has in this context been known that certain patients develop antibodies to interferon in connection with such therapy. However, the reported results vary extensively.

S.J. Jacobs et al., Minimal Antigenecity of Intron A in Human Recipients Demonstrated by three Analytical Methods, J. Biol. Response, Mod. vol. 7, pp 447-456, 1988, state that out of 101 α-interferon- treated cancer patients only a few developed antibodies. Thus, biological assay showed positive result in one case only, while 7 positive results were obtained in an ELISA-assay.

In contrast thereto, R.G. Steis et al., Resistance to Recombinant Interferon alpha-2a in Hairy-cell Leukemia Associated with Neutralizing Anti-interferon Antibodies, N. Eng. J. Med. vol. 318, pp 1409-1413 (1988), detected antibodies in 31 out of 51 hairy-cell leukemia patients in interferon treatment.

R.A. Figlin et al., Recombinant Interferon alpha-2-a in Metastatic Renal Cell Carcinoma: Assessment of Antitumor Activity and Anti- interferon Antibody Formation, 0. Clin. Oncol, vol. 6, pp. 1604-1610 (1988) proved that patients with metastatic renal cell carcinoma developed antibodies in different degrees.

In spite of the varying results with respect to the formation of antibodies in connection with interferon therapy, it appears to have been proven that antibodies may be developed by exogenously induced interferon.

Various theories regarding antibody conditions have been tested by many internationally recognized scientists. Thus, Panem et al., Alpha Interferon and Antibody to alpha- Interferon in a Patient with Systemic Lupus Erythematosus, 1982, J. Immunol. 129, p. 1-3 (1982) examined 200 patients with different autoimmune diseases with respect to neutralizing antibodies to interferon. None of the patients had previously been treated with interferon, and the applied neutralization assay showed a positive result for one patient.

In Patient with Circulating Antibodies to alpha-Interferon, Lancet, 2, pp. 1227-1228 (1981) Mogensen et al. describe an otherwise healthy 77-yr old man who developed Herpes-zoster localized to the 4th lumbar nervous fegion and who in connection with interferon treatment appeared to have circulating antibodies of the IgG-type for 11 months.

Trown et al., Antibodies to Human Leucocyte Interferon in Cancer Patient, Lancet. 1. pp. 81-84 (1983) examined 76 cancer patients, of which antibodies were found in 2; 60 diabetes patients in which no antibodies were found, and, most interesting, 200 healthy donors in which no antibodies were found either.

These results show that in a few patients suffering from the above- mentioned diseases auto-antibodies may occur, and it has not been possible to explain the reason therefor. De Maeyer et al . describe in Natural Antibodies to Interferon and Interferon-? are a Common Feature to Inbred Mouse Strains, J. Immun. vol. 136, No. 5 (1986), the natural occurrence of antibodies to interferon-α and -β in inbred mice, while De Maeyer et al . at the same time points out the striking contrast to the lack of antibodies in humans. De Maeyer et al . cannot explain this difference.

Literature thus describes that as part of a sort of immune response, antibodies to interferons are in different degree developed by exogenous addition of interferons. At the same time, literature describes a few cases of naturally occurring antibodies in patients who have not been treated with interferon. It has, however, not been possible to explain the latter formation of antibodies.

Furthermore, it is a commonly accepted phenomenon that in human serum and other body fluids substances are present which interfere with immunochemical as well as biological methods of measurement for cytokines. These substances show an inhibitory effect on cytokines by said method of measurement.

In spite of the view until now that healthy humans do not have antibodies to cytokines, it has now surprisingly been found that in healthy humans there is a number of circulating, naturally occurring antibodies to cytokines. In this connection it is important to discern between naturally occurring antibodies, i.e. auto-antibodies, and antibodies formed as a consequence of exogenous addition of antigen, f. ex. in the form of cytokines.

The recognition of the occurrence of auto-antibodies to cytokines in humans changes the understanding of immune regulatory mechanisms in the cytokine network and implies important diagnostic and therapeutic progresses. The total number (200) of serum donors who have been examined exhibit the occurrence of auto-antibodies to cytokines. This circulating amount of neutralizing antibodies to cytokines regulates and controls the cytokine activity level. It is assumed that there is a correlation between certain groups of diseased humans having a very high or very low auto-antibody concentration and the ability of these humans to regulate cytokine activities. The recognized occurrence of auto-antibodies to cytokines may explain some of the disappointing results obtained by the use of recombinant as well as native cytokines for pharmaco-therapeutic purposes. Before commencing therapy with f. ex. interferon it will now according to the invention be possible to examine the auto-antibody titer to interferon of the patient, and possibly regulate the antibody production or antibody function. Furthermore, the recognition of the formation of auto-antibodies opens the possibility of charting hereditary tendencies to certain diseases, f. ex. certain forms of cancer and certain immunologic sufferings which may be predicted. In addition, many acute, serious diseases have been caused by the cytokine effect and may, therefore, be alleviated by the addition of neutralizing antibodies. The auto-antibodies are purified according to the invention and are applied directly therapeutically in the treatment of diseases or for immunochemical analyses as f. inst. radioimmunological assaying (RIA), ELISA or immunofluorescence, etc. The immunochemical analyses may be based on affinity purified, labeled auto-antibodies. The single cases of auto-antibody activity to interferon in connection with illness or malregulation described in literature are very well in agreement with the occurrence of auto-antibodies in all humans, as now recognized, as the development of auto-antibodies is assumed to function as a buffer against cytokines. Illness or malfunction causes an increase of the auto-antibody level, with the results reported in the above-mentioned literature.

It is a purpose of the present invention to provide means for regulating the i munologic mechanisms in the cytokine network in humans. This is surprisingly obtained by the regulatory proteins according to the invention.

These regulatory proteins may be mono- or polyclonal antibodies of the type which reacts with the same antigenic determinants on cytokines, such as α-interferon, /}-interferon, γ-interferon, interleukin 1-6, tumor necrosis factor, lymphotoxin, etc., as human auto-antibodies to such cytokines. Furthermore, the regulatory proteins may be cytokines, non- toxic cytokine analogues or cytokine fractions with antigenic determinants.

The regulatory proteins in the form of antibodies may be used for the treatment of acute cases, where it is desired to inhibit the cytokine effect, f. ex. the cysteine effect. This may for instance be the case in cerebral malaria, meningococcus meningitis, etc., where the patient, even after the infection has been controlled by antibiotics, may die of the effect of the high cytokine production on the cells of the body. Furthermore, the auto-antibodies according to the invention may possibly be used in a specific or unspecific inhibition of the auto- antibody production. A specific inhibition of the auto-antibody production may for example be obtained by exogenous supply of antibodies according to the invention, in a manner f. ex. being known from therapy against rhesus immunization. Such inhibition may f. ex. be of importance in the interferon treatment of cancer diseases. The antibodies according to the invention may furthermore possibly be used for the treatment of immune diseases, such as f.ex. chronic polyarthritis, serious viral and bacterial diseases as those mentioned above, allergic cases, acute shocks, cases of poisoning, etc.

The cytokines, cytokine analogues, or cytokine fragments with antigenic determinants may in a hitherto unknown manner be used for increasing the auto-antibody production and function in humans, f. ex. possibly as acute phase vaccination for particularly exposed or susceptible persons. Finally, the antibodies according to the invention may be used in immunochemical and immunobiological assays. The cytokine activity in body fluids is presently being determined by means of antibodies from animals, and an expression of the number of the free epitopes or determinants on the cytokine molecule which may react with animal antibodies is hereby obtained. However, these epitopes are not necessarily the biologically active sites. In contrast hereto, the use of human monospecific cytokine antibodies according to the invention provides specificity to biologically active determinants on human cytokines. By labeling of such monospecific human antibodies according to the invention, it will be possible to obtain a better measurement of the cytokine activity in immunochemical assays. The use of the human cytokine antibodies according to the invention gives a measure for the number of biologically active determinants on the cytokine molecules and thereby a better measure of the cytokine activity.

The antibodies may f. ex. be used in ELISA-techniques, radioimmunoassay techniques (RIA-techniques), immunofluorescence techniques, etc.

ELISA-techniques may be used in various ways for the detection of antigens or antibodies in biologic materials. ELISA-techniques have been developed for many purposes and in many variations, also including competitive assays of the type which are also carried out by RIA-assays. Usually, microtiter plates are used as solid phase in ELISA-techniques and, according to the purpose of the assay, one or more layers may be bound to the solid phase to increase the sensitivity and otherwise optimize the specific assay. The principle of the ELISA-technique will be known to those skilled in the art and will, therefore, only be briefly mentioned. The principle consists of enzyme labeling of a specific reaction component being applied in an antigen-antibody system which by application under appropriate in vitro conditions enables assays which combine high sensitivity and immunologic specificity. The RIA-technique will also be known to those skilled in the art, and the assay is in principle carried out in the same way as by the ELISA- technique, apart from the fact that the antibodies are labeled with radioactive isotopes instead of enzymes. The immunofluorescence technique will also be known to those skilled in the art, and the antibodies are in this case labeled with appropriate chemical compounds (fluorochroraes), giving visible fluorescence by radiation with f. ex. ultraviolet light. This is a very sensitive technique which makes it possible, by microscopy f. ex., to see directly where antibody has reacted with antigen. The most common fluorochromes are derivatives of fluorescein and rhodanine. These give green and red fluorescence by ultraviolet radiation, respectively, and thereby make it possible to see several antigens at the same time. There are direct and indirect immunofluorescence techniques. These techniques may furthermore be applied fdr studying biopsy material.

According to the present invention an assay kit is provided for the above-mentioned techniques. This assay kit may be used for determining the cytokine activity in body fluids, as the human monospecific cytokine antibody component according to the invention is bound directly or indirectly to the solid phase. As a variation, the assay kit may contain a cytokine component, cytokine analogue component or a component comprising a cytokine fragment with an antigenic determinant. The cytokine, cytokine analogue or cytokine fragment is bound either directly or indirectly to the solid phase and may be used for qualitative as well as quantitative analysis of auto-antibodies to cytokines in humans. As a consequence of the useful pharmacological properties, the regulatory proteins according to the invention may be manipulated, as peptide sequences are formulated for achievement of various pharmaceutical forms for administration purposes. For the production of pharmaceutical preparations according to the present invention an effective amount of the specific compound as active ingredient is combined in intimate mixture with a pharmaceutically acceptable carrier, which carrier may be in many different forms, depending on the desired form of the preparation for administration. These pharmaceutical preparations are desirable in unit dosage form, which especially is suitable for percutaneous, intramuscular, intravenous or parenteral injection or rectal administration. The carrier will, at least to a large extent, usually comprise sterile water, even though other ingredients may form part of the preparation. The carrier may thus comprise a saline solution, glucose solution, or a mixture of saline and glucose solution. Furthermore, injectable suspensions may be prepared, in which case suitable liquid carriers, suspension agents or the like may be used. Reference being made to the drawings,

Fig. 1 graphically shows the result of a neutralization test, using recombinant α-interferon, EMC-virus (A. Meager, Natl. Lab. Biol. Control., London) and human serum together with cell control and virus control , Fig. 2 graphically shows the result of a neutralization test, using CIF (Crude Human Leukocyte Interferon), EMC-virus and human serum, together with cell control and virus control,

Fig. 3 graphically shows the result of titration in a neutralization test, using vesicular stomatitis virus (VSV), EMC-virus and CIF,

Fig. 4 graphically shows the result of titration in a neutralization test, using VSV, EMC-virus, and recombinant α-interferon,

Fig. 5 graphically shows the result of titration in a neutralization test, using cell line A549 (human lung carcinoma cell line described in J. Nat., 51:1417-1423 (1973)), EMC-virus, CIF, recombinant a-interferon, Fll-medium (Eagle's minimal essential medium with Earle's salt and 2% and 1% of Ultroser^®, respectively,

Fig. 6 graphically shows the result of affinity chromatography on a column with rabbit anti-human IgG and rabbit anti-human IgM, Fig. 7 graphically shows the result of an antiviral neutralization bio-assay of eluates and wash from a combined rabbit anti-human IgG- and rabbit anti-human IgM-column, using CIF,

Fig. 8 graphically shows the result of an antiviral neutralization bio-assay of eluates and wash from a combined rabbit anti-human IgG- and rabbit anti-human IgM-column, using recombinant α-interferon,

Fig. 9 graphically shows the result of an antiviral neutralization bio-assay of eluates and wash from a combined rabbit anti-human IgG- and rabbit anti-human IgM-column, using native }-interferon,

Fig. 10 graphically shows the result of an antiviral neutralization bio-assay of eluates and wash from a combined rabbit anti-human IgG- and rabbit anti-human IgM-column, using recombinant γ-interferon,

Fig. 11 graphically shows the result of affinity chromatography on a column with rabbit anti-human IgM, Fig. 12 graphically shows the result of the antiviral neutralization bio-assay of wash and eluate from a column with rabbit anti-human IgM, using CIF,

Fig. 13 graphically shows the result of the antiviral neutralization bio-assay of wash and eluate from a column with rabbit anti-human IgM, using recombinant α-interferon,

Fig. 14 graphically shows the result of the antiviral neutralization bio-assay of wash and eluate from a column with rabbit anti-human IgM, using J-interferon, Fig. 15 graphically shows the result of the antiviral neutralization bio-assay of wash and eluates from a column with rabbit anti-human IgM, using γ-interferon,

Fig. 16 graphically shows the result of affinity chromatography on a column with protein A-Sepharose^® 4B, Fig. 17 graphically shows the result of an antiviral neutralization bio-assay of wash and eluates from a column with protein A-Sepharose^® 4B, using CIF,

Fig. 18 graphically shows the result of an antiviral neutralization bio-assay of wash and eluates from a column with protein A-Sepharose® 4B, using recombinant α-interferon,

Fig. 19 graphically shows the result of an antiviral neutralization bio-assay of wash and eluates from a column with protein A-Sepharose® 4B, using β-interferon,

Fig. 20 graphically shows the result of an antiviral neutralization bio-assay of wash and eluates from a column with protein A-Sepharose® 4B, using γ-interferon,

Fig. 21 graphically shows the result of the antiviral neutralization bio-assay in comparison with wash and eluate from affinity chromatography on T-gel, using CIF, Fig. 22 graphically shows the result of the antiviral neutralization bio-assay in comparison with wash and eluate from affinity chromatography on T-gel, using recombinant α-interferon,

Fig. 23 graphically shows the result of the antiviral neutralization bio-assay in comparison with wash and eluate from affinity chromatography on T-gel, using β-interferon,

Fig. 24 graphically shows the result of the antiviral neutralization bio-assay in comparison with wash and eluate from affinity chromatography on T-gel, using γ-interferon, Fig. 25 graphically shows the result of affinity chromatography on a column packed with recombinant interferon-α,

Fig. 26 graphically shows the result of affinity chromatography on a column, wherein interferon is bound directly on matrix, Fig. 27 graphically shows the result of a neutralization test with lymphotoxin,

Fig. 28 graphically shows the result of a neutralization test with tumor necrosis factor,

Fig. 29 graphically shows the result of a biological assay with serum HNI IgG from Statens Seruminstitut (SSI), Copenhagen, Denmark, using native α-interferon,

Fig. 30 graphically shows the result of the titration in a neutralization test with serum, using EMC-virus and recombinant α- interferon, Fig. 31 graphically shows the result of the titration in a neutralization test, using IgG, EMC-virus and i-interferon,

Fig. 32 graphically shows the result of the titration of serum against γ-interferon, using EMC-virus,

Fig. 33 graphically shows the result of a toxicity control and anti-viral control of human normal immunoglobulin fraction (HNI-IgG), and

Fig. 34 graphically shows the result of an anti-viral neutralization bio-assay, using Fab-fragments and α-interferon.

In all assays A549 cells are used, a human cell line being sensitive to all types of human interferon. Several types of virus may be used against this cell line. In the present case Encephalomyocarditis virus (EMC-virus) was used, which very clearly and rapidly kills unprotected A549 cells. The obvious advantage of using human cells is that the specificity of the auto-antibodies can be demonstrated directly with these cells. Thus, the auto-antibodies only react with epitopes in the interferon molecule being biologically active in the human system. It has thus not been possible to trace naturally occurring auto- antibodies to interferons by using f. ex. bovine cells (MDBK), being a highly recognized cell line which is sensitive to human α-interferon. The present invention is described in more detail in the following.

Serum Production

Donor serum is obtained from healthy donors whose sex and age are recorded for statistical reasons. Blood is obtained from these donors and is allowed to stand at room temperature until coagulation. Serum is thereafter withdrawn with a micropipette. Any turbid serum is discarded. All serum is heat-treated before use (30 min. at 56°C).

Neutralization test

A human cell line A549 (human lung carcinoma cell line described in J. MAT. Cancer Inst. 51:1417-1423, (1973)) is inoculated and cultured in 96-well microtiter trays (Nunclone) in Eagle's minimal essential medium containing 1% L-glutamine, 1% penicillin/streptomycin, and 3% NaHC0₃ and enriched with 5% foetal calf serum (FCS). 2% Ultroser^® Gibco BRL may be used instead of FCS for enrichment of the growth medium, which does not change the test result. For inoculation is used the lowest number of cells per well which under the given growth conditions result in maximum production of formazan (blue-colouring of live cells) by colouring with 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) by the socalled MTT-test in the course of 12 h. The cells are cultured in the microtiter wells for 18 h at 37°C and under an atmosphere containing 5% C0«. Serum from healthy donors is diluted (double series dilution) in the above-mentioned enriched culture medium which contains an amount of interferon just sufficient to secure significant protection of the above-mentioned cultured^" cells against indicator virus under the infection step described herebelow. The serum dilutions with interferon are incubated for 1 h at 37°C. The original culture medium is absorbed from the wells, and the serum dilutions containing interferon are added, and additional incubation takes place for abt. 18 to 24 h at 37°C under an atmosphere containing 5% CC This medium is thereafter removed from the wellSj and indicator virus (EMC-virus from A. Meager National Lab. Biol. Control, London) in the above-mentioned culture medium, yet only enriched with 2% FCS (foetal calf serum), is added. This virus is added with a titer securing total cytopathogenic effect on unprotected cells after 24 h infection. After incubation for abt. 20 h at 37°C under an atmosphere containing 5% CC MTT is added, and after abt. 2 hrs. the cells are lysed, and cell viability is determined spectrophotometrically at 570 run in an ELISA-scanner.

Concurrently a serum toxicity test is carried out as a control, using serum dilutions without interferon and without subsequent addition of virus. A concurrent control of a possible anti-viral effect of serum is carried out, using serum dilutions, but with subsequent addition of virus. Furthermore, a control of the effect of interferon is carried out without the use of serum dilutions in the above-mentioned test. A control of the effect of indicator virus is carried out by omitting serum dilutions and interferon from the above-mentioned test. Finally, the cell viability is controlled by omitting serum dilutions, interferon, and virus from the above-mentioned test. In this test it is observed that the addition of serum dilutions inhibits the protecting effect of interferon on said human cell line against viral attacks. A neutralization test for determining antibodies having a neutralizing effect on interferons implies a very fine balance between cells, virus, and interferon concentrations. Prior to the above- mentioned test, the necessary titer adjustments must be carried out to ensure optimal sensitivity and specificity. These titer adjustments are illustrated in more detail in Figures 3-5.

Neutralization test for tumor necrosis factor or lymphotoxins

A very sensitive cell line WEHI 164 clone 13 (T. Espevik and H. Nissen-Meyer, A Highly Sensitive Cell Line WEHI 164 clone 13, for Measuring , J. Immunol. Methods, 1988) is cultured as described above in microtiter trays. Twofold dilutions of serum in the culture medium used in the test described above are incubated with an amount of tumor necrosis factor or lymphotoxin (TNF/LT), just sufficient for killing the WEHI-cells. After incubation for 1 h the serum-TNF/LT-test medium is transferred to the microtiter trays, and the cell viability is determined by the MTT-method as described above. Control tests show that the cells are killed if serum is omitted from the test medium, while they survive if the test medium contains serum.

Affinity chromatography

Isolation of the immunoglobulin fraction.

2% (v/v) acetate buffer and 250 mg/ml ammonium sulphate is added to a volume of serum. The mixture is allowed to stand in the dark at room temperature for 24 h and is thereafter centrifuged for 30 min. at 6000 rpm. The precipitate is washed in 1.75 M aqueous ammonium sulphate solution, using a volume of ammonium sulphate solution corresponding to approx. half of the initial volume of serum. The ammonium sulphate solution containing the precipitate is centrifuged for 30 min. at 6000 rpm. The precipitate is dissolved in distilled water, using for dissolution one volume of water corresponding to one half of the initial serum volume. The aqueous solution is thereafter dialysed for 12 h vs. distilled water, 12 h vs. acetate buffer, pH 5, 12 h vs. distilled water and 12 h vs. a volume of acetate buffer, pH 5, corresponding to the double of the serum volume. The precipitated microproteins are removed by centrifuging for 30 min. at 6000 rpm. The supernatant is led through a DEA A50 Sephadex® column, whose bed volume is 25% of the initial serum volume. The acetate ions are thereby removed. The eluate is concentrated by repeated salting out with ammonium sulphate as mentioned above, and the precipitate is dissolved in an amount of water corresponding to half the initial serum volume, and the solution is dialysed for 2 h vs. distilled water. The solution is next dialysed vs. PBS (phosphate buffer-adjusted brine).

Anti IgG- + IgM-column

10 ml rabbit anti-human IgG (3010C, Kem-En-Tec Aps. Biotechnology Corp., Copenhagen) and 2 ml rabbit anti-human IgM (3090Y, Kem-En-Tec Aps. Biotechnology Corp., Copenhagen) are used in mixture for the column. The affinity chromatography column thus contains rabbit- antibodies directed specifically toward human immunoglobulins of the IgG- and IgM-type.

The column is loaded as follows: 3 g lyophilized gel powder (CNBr-activated Sepharose^® 4B,

Pharmacia) is weighed, corresponding to a gel volume of 10.5 ml. The gel powder is precipitated in 600 ml 1 mM HC1 and is allowed to stand for 15 min. at room temperature. The gel is next washed on a sintered glass filter (porosity G3). The above-mentioned rabbit anti-human antibody (RAH) is dissolved in 0.1 M NaHC0₃ solution (pH 8.3) containing 0.5 M NaCl, said solution in the following called the coupling buffer. Next, 105 mg/ml RAH-Igl + IgM is dissolved in the coupling buffer, the binding capacity being stated to be abt. 5-10 g antibody/ml gel (Pharmacia). The antibody solution is added to the gel suspension and is allowed to stand for 2 h at room temperature. The gel suspension is filtered, and the content of protein in the supernatant is determined to 17.85 mg/ml at A 280. As mentioned, 3 g gel powder is used which during the gel application swells to approx. 14 ml. After adsorption, 21.7 mg is detected as residue in the buffer, to which a total of 31.85 mg has thereafter been bound (59.5%). The gel is transferred to a blocking buffer in the form of 0.2 M glycine (pH 8) and is alowed to stand for 16 h at 4°C. Excessive absorbed protein is washed out with the above- mentioned coupling buffer followed by an acetate buffer (0.1 M, pH 4, containing 0.5 M NaCl) and the gel is again washed with the coupling buffer. Ion-bound ligands are thereby removed. Possible residual blocking buffer is washed out with coupling buffer. The Sepharose®-gel with bound antibodies is stored at 4°C in the presence of antibiotics. Before use, the gel is precipitated in approx. 20 ml PBS and is transferred to the column. The column is washed with PBS until 0D- (optical density) monitoring shows that the base line has been reached.

Affinity chromatography is carried out on the above-mentioned column with serum globulin volumes from 300 μl to 10 ml. The flow, in the following called wash, is collected in fractions of 900 μl by means of a time-volume counter and a fraction collector. The column is eluted with acetic acid, pH 2.8, and 200 μl of TRIS-buffer, pH 9.0, is added to eluate receptacles as neutralizing agent. En eluate fraction volume of totally 1100 μl is hereby obtained. The eluate fraction is dialysed vs. PBS for removal of possible toxicity, and before use in the biological neutralization test, pH is controlled in eluates and wash.

Anti-IgM-column

The column is loaded in the same way as the above-mentioned combined column, whereby, though, rabbit anti-human-IgM (no-nonsense- antibody, Kem-En-Tec) is employed instead of the combination of RAH-IgG and RAH-IgM. The antibody is bound to approx. 3 ml of gel corresponding to 13 mg (approx. 70% binding). Besides, the process is identical to that described for the combined column.

Protein A-column

8 ml protein A-Sepharose^® 4B (Pharmacia) is employed for this column. This column is widely used for affinity chromatography of IgG. The column is manufactured in accordance with instructions in "Affinity Chromatography Principles and Methods", Pharmacia. For eluation 1 M of acetic acid (pH 3) is used. TRIS-buffer (pH 9) is used under cooling to neutralization, and wash and eluates are dialysed vs. PBS. T-gel «

The T-gel method is widely used for coarse purification of immunoglobulins. For employment of the gel, Kem-En-Tec's instructions are followed. The T-gel is equilibrated with 0.75 N ammonium sulphate. Next, the serum sample and then ammonium sulphate is added to a concentration of 0.75 M. 10- ml T-gel is used for 5 ml of undiluted serum. The gel is washed with 0.75 M ammonium sulphate until 0D- monitoring shows that the base line has been reached. The gel is eluted with 0.1 NaCl. The eluate must contain IgG, IgM and traces of α-2- acroglobul n. The eluate is dialysed vs. PBS, and pH is controlled.

Recombinant interferon column

The principle in using this column is to bind the antigen, in this case pure recombinant α-interferon, to CNBr-activated Sepharose^® 4B. Serum and globulin fraction samples containing auto-antibodies to α- interferon are passed through the column, whereby the auto-antibodies * are bound to the antigen. The column has been prepared in accordance with the manufacturer's instructions ("Affinity Chromatography ; ^" Principles and Methods", Pharmacia).

300 mg lyophilized powder is used for the column, and this gives a final gel volume of 1.17 ml. The column is loaded with 2 ml recombinant α-interferon (Intron-A) in a concentration of 4.48 mg/ml (totally 8.96 mg). 61.5%, corresponding to 5.51 mg, is bound to the column. Affinity chromatography is carried out in the same way as with the above- mentioned combined column.

Radial immunodiffusion in gel ad modum Mancini

This method is used in documentation of the quality of the affinity chromatography. The agar gel (Nor-Partigen, Hoechst) is used in accordance with the manufacturer's instructions. The assay is carried out as a one-point calibration with a serum pool as serum control, including the manufacturer's own standard control. 5 μl of sample is added to each well. Concentrations are determined by means of Nor- Partigen Table of Reference Values.

As it is difficult to make precise dilution calculations on the affinity chromatographies, and as the detection of auto-antibodies beforehand is qualitative, the results of the Mancini test are semi- quantitative, the results being stated as + (positive), trace, - (negative).

Concentration Concentration Sample IgG (mg/ml) IgM (mg/ml)

Control serum (serunι pool) + +

Wash (IgG + IgM) trace -

Eluate No. 1 (IgG + IgM) + trace

Eluate No. 2 (IgG + IgM) + trace

Wash (IgM) + -

Eluate (IgM) trace trace

Wash (protein A) - +

Eluate (protein A) + trace

T-gel (wash) trace +

T-gel (eluate) + -

Ammonium sulphate precipitated glo¬ bulin + +

Control serum Ig determined on basis of "Mancini Reference Table" from Hoechst: IgG: 12.3 g/1 and IgM: 1.6 g/1. At the same time, a pool of serum from 195 healthy individuals is used as control serum. This corresponds to normal values stated in literature and implies the validity of the Mancini-test.

Biologic assay of immunoglobulins from Statens Seruminstitut

For further documentation that the cytokine inhibiting effect of serum is attributable to immunoglobulins (in this case IgG) a human normal immunoglobulin fraction (HNI) from Statens Seruminstitut,

Copenhagen, Denmark, has been assayed. The fraction, IgG-10-1-890213, has been stabilized with glycine and contains 16% (w/v) NHI and 2% (w/v) glycine. The test is carried out as a neturalization test against all types of interferons, and HNI-concentrations from 15 mg/ml to 0.1 mg/ml are used, as described under said neutralization test.

Dot-blot analysis

This analysis is a qualitative immunochemical method for detecting antigen/antibody reaction, α-interferon being used as antigen in this case and auto-antibodies directed against α-interferon as antibody. 3 μl of pure α-interferon (concentration 10 mg/ml) is used, which is applied onto nitrocellulose paper (NCP). NCP is blocked by treatment for 10 min. with 2% Tween/TRIS, pH 7.4. NCP is thereafter blocked by treatment for 15 min. with 1% BSA/0.3% Tween/TRIS, pH 7.4. Serum is diluted 1:5 in 0.5% BSA/TweenΛRIS, pH 7.4, and incubated with NCP overnight. NCP is then washed in 0.3% TRIS, pH 7.4, for 10 min., this last-mentioned step being carried out three times. Then follows incubation with peroxidase- conjugated RAH-IgG (Dako P 212, 1:50 in 0,5% BSA/0.3% Tween/TRIS, pH 7.4), and NCP on a tilting bath for 4 h. Washing is then carried out 2 x 10 min. in TRIS, pH 7.4, and for 10 min. in TRIS, pH 7.4, diluted 1:5 in distilled water. Finally, dyeing with naphtol in methanol/TRIS is carried out, pH 7.4 (1:5 in distilled water) with an admixture of hydrogen peroxide. The dyeing reaction is stopped by washing 5 times in distilled water. The method is described in more detail by H. Towbin and J. Gordon in Immunoblotting and Dot Immunobinding - Current Status and Outlook, J. Immunol. Methods 72, 313-340, 1980.

Preparation of Fab-fragments

An antibody may be cleaved into Fab/Fc or Fab2/Fc'-fragments by using papain or pepsin, respectively. The Fab-fragments are used for proving that a specific antigen-antibody reaction has taken place and not only a Fc-binding. Fab/Fc-fragments are prepared by a standard procedure whereby papain Sepharose® is admixed to purified IgG from healthy donors, whereafter incubation takes place for 24 h at room temperature. The supernatant, isolated after centrifugation, is subjected to affinity-chromatography on protein A Sepharose^® gel, whereby the Fc portion will be bound in the column, and the Fab portion will be found in the wash. The effect of the Fab-fragments is assayed by the above-mentioned dot-blot analysis and by an anti-viral neutralization bio-assay.

Results Neutralization test of serum from healthy donors

This test showed that the neutralizing effect on recombinant as well as native α-, β- , γ-interferon and tumor necrosis factor and lymphotoxin was found in the total number of examined healthy donors (195). The donors are equally represented with respect to sex and age (15-65 yr). It was also found that the individual variation of the inhibitory activity against cytokines in serum was less than 40% from the mean value (cf. Fig. 1 and 2, which illustrate 10 donors, chosen at random, and tested against CIF and recombinant α-interferon, respectively). The figures have been drawn on the basis of arithmetic mean values of threefold determinations (standard deviations of less than 10% are not shown). Serum dilutions (final dilution 1:20) have been examined vs. 0.13, 0.25, and 0.50 IU/ml (final concentration) of interferon. It appears that all donor sera significantly inhibit the added amount of interferon. Titrations with identical results have been carried out vs. recombinant α-, β- , and γ-interferon as well as native α- and β-interferon.

Biologic assay of the globulin fraction

It was found that the cytokine inhibiting effect exists in ammonium sulphate-precipitated globulin fractions vs. all the tested cytokines. The neutralization assay was carried out in the same way as the assay with serum, and using the same dilutions. It was found that 90% of the interferon-inhibiting effect had been retained in the globulin fraction. After each of the examined donors had been found positive, serum from 115 donors was combined for the purpose of affinity chromatography. Affinity chromatography of the ammonium sulphate-precipitated globulin fraction from the above serum pool was carried out by means of specific rabbit anti-human (RAH) immunoglobulins. It clearly appeared that the interferon inhibiting effect of the globulin fraction could be removed by passing this fraction through the mentioned columns and examine the flow (wash) of the biologic assay. This is very well in agreement with the finding that IgG and IgM, respectively, bind to the respective columns and consequently are not found in the flow. At the same time, the inhibiting effect could be restored by eluting the columns with acetic acid, dialyse the eluate and test it in the biologic assay.

It is hereby proved that serum of healthy individuals contains neutralizing auto-antibodies to interferons, tumor necrosis factor, and lymphotoxin, and it must furthermore be considered to have been rendered probable that likewise auto-antibodies to other cytokines, including interleukines, occur. The results show that in any case naturally occurring antibodies of both the IgG- and IgM-type are present. It has furthermore been found that the serum and globulin fraction may be diluted at least 80 times and still inhibit the cytokine effect.

Furthermore, the IgG- and IgM-concentration in the flow (wash) and eluates has been tested by radial immune diffusion (Mancini), and it was thereby found that there was a clear coherence between the antibody concentration and the cytokine inhibiting effect. It is f. ex. not possible to detect IgG and IgM in the flow of the combined column (RAH IgG + IgM) when the column is not overloaded. There is, however, a strong concentration of IgG in the eluate of the protein A column (cf. Fig. 18).

10 ml serum globulin fraction was transferred to the recombinant α- interferon column to secure binding of many antibodies. The eluate was found to possess high anti-interferon activity to α-interferon (cf. Figs. 25 and 26). Lyophilized human immunoglobulin from Statens Seruminstitut was also found to contain biologic anti-cytokine effect (cf. Figs. 29-32). The presence of the auto-antibodies, expressed by a dose/response curve, in the biologic assay appears clearly from these Figures.

Finally, a dot-blot analysis of the above-mentioned serum pool has been carried out vs. recombinant α-interferon. This analysis had a positive result in confirmation of the presence of the specific antibodies.

A corresponding dot-blot analysis with Fab-fragments gave a distinct precipitation, which shows that a specific binding of antibody to interferon is taking place.

The following designations are used in the figures:

CC: Cell control, the result of an MTT-assay with healthy non- infected cells;

VC: Virus control, the result of an MTT-assay with virus-infected cells;

TC: Toxicity control, the result of an MTT-assay with healthy non- infected cells to which an antibody-containing sample has been added; IC: Interferon control, showing the protection afforded by the given amount of interferon against indicator virus;

W: Effect of flow or wash, i.e. the portion of the serum globulin fraction passing through the column, in an MTT-assay of cells in the presence of interferon and virus;

E : Effect on the eluate, i.e. the portion of the serum globulin fraction specifically retained by the rabbit-antibodies, corresponding to human antibody classes, the eluate being collected in 1-3 fractions, pH neutralized (TRIS), and dialysis performed

(vs. PBS), in an MTT-assay in the presence of interferon and virus;

Fab: Effect of Fab-fragments in an MTT-assay in the presence of interferon and virus.

Figures 3 and 4 illustrate neutralization tests carried out to find the proper challenge virus for these tests and at the same time the proper amount of virus for infection. It appears clearly that EMC (dilution 10 ) has a good effect, as an increased amount of interferon causes an increased OD-signal (MTT-method). In other words, this means that an increased amount of interferon will afford an increased protection of the cells (A549) against challenge virus, and thereby a higher rate of survival. At very low interferon concentrations the curve approaches the virus control curve, and this means that such low interferon concentration (0.25 U) does not afford any appreciable protection to the cells. At the other end of the curve is seen that at 2

U CIF and 2 U Intron A, the curve approaches the cell control curves, indicating that interferon affords a good protection against challenge virus. At the same time it appears that VSV (vesicular stomatitis virus)

_2 is not nearly as active in this system, and an EMC (dilution 10 ) is therefore chosen for the further tests.

Fig. 5 shows interferon titrations performed with EMC (dilution

10 ) as challenge virus on CIF and Intron in concentrations from 5 U/ml to 0.05 U/ml. It is found that the system is sensitive with interferon concentrations of 1- and 2 U/ml. This counts for CIF and Intron A, and similar titrations have been performed vs. β- and γ-interferon. The beta-system is also sensitive at 1- and 2 U/ml, whereas the γ-interferon system is better at 2 and 4 U/ml. These interferon values are hereafter used. Interferon values are stated in accordance with international reference units, by 69/19 B "Standard Reference for HuIFN-alpha(Le), from Mill Hill, MRC, UK." Herebelow follow the results of experiments proving that auto- antibodies to interferon, and thereby cytokines, are found in humans.

Figure 6:

Affinity chromatography on column with RAH IgG and RAH IgM. During the test, OD-monitoring is performed, so that the proper fractions containing the immunoglobulins can be taken out. Samples are collected from the OD-peak and are tested for interferon activity in a neutralization test. The first peak of the curve corresponds to wash, and the second peak corresponds to the eluate.

Figure 7:

The bar chart illustrates the result of the antiviral neutralization bio-assay, with eluates and wash from the combined RAH IgG + IgM column. It is seen that the eluate has an interferon- neutralizing effect against CIF in concentrations of 1 and 2 IU/ml , respectively. A slight inhibiting effect is observed in both washes as an expression of supersaturatiσn of the column. Supersaturation means that a larger amount of antibodies has been added, in the form of serum globulin fraction, than the RAH-column is able to bind. Tests have also been carried out in which the column is not supersaturated, and in these tests the effect totally disappears from the wash; in return, the interferon-inhibiting effect of the eluate decreases, as smaller quantities of sample must be used in order to ensure that the column is not supersaturated. It is seen that the effect of wash vs. 2 U corresponds to the result of the interferon control, and that the effect of the eluate lies in the vicinity of the result of the virus control. This means that the interferon-inhibiting effect of the serum globulin fraction by means of the affinity chromatography described, has been removed from the wash and thereafter is found in the eluate. It is seen that the eluate has a strong interferon-neutralizing effect both vs. 1 U and 2 U CIF, which shows that the column has bound the auto-antibodies which are thereafter found in the eluate. The fact that both eluates show an effect close to the result of the virus control indicates that a sufficient amount of antibodies to interferon is present in the eluate for neutralizing the effect of 2 U interferon (in this case CIF). These findings must be considered to be heavy arguments for the belief that serum from normal healthy donors contains auto-antibodies which are able to neutralize the anti-viral effect of native α-interferon.

Figure 8:

Figure 8 shows the same as Figure 7. Interferon = recombinant α- interferon 2A = Intron A. It can thus be concluded that there also exist naturally occurring antibodies which can neutralize the effect of recombinant α-interferon (intron A). It may seem illogical, that healthy humans have auto-antibodies to recombinant interferon, but it has previously been found that patients being treated with recombinant interferon develop antibodies to recombinant as well as native interferon. On reflection, it is therefore understandable that healthy donors having auto-antibodies to native α-interferon will also have naturally occurring antibodies with neutralizing effect vs. recombinant interferon-alpha. The figure shows a weak effect in both washes which is due to a slight supersaturation of the column (see above). It is also seen that there is sufficient antibody effect to neutralize the effect of 2 U Intron A.

Figure 9: Figure 9 shows the same as Figure 7, except that native β- interferon is used here. The interferon concentration is final 1 and 2 U/ml, respectively. There is a strong interferon-inhibiting effect in both eluates, and this is quite in agreement with antibody titer assays carried out as controls of serum, where it has been possible to dilute serum at least 160 times and still observe a clear inhibiting effect vs. ?-interferon. There is also a clear effect in the wash, which means that the small amount of immunoglobulin being present in the wash (cf. Figure 18) may neutralize a part of the interferon effect.

It is thus extremely clear that there is a natural occurrence of large amounts of antibodies having a neutralizing effect vs. β- interferon. Figure 10:

Figure 10 shows the same as Figure 7 vs. recombinant γ-interferon. The interferon concentration is final 2 U/ml, resp. 4^~ U/ml . It is seen that the effect of the wash lies slightly above the result of the interferon controls.

At the same time, it appears that there is a clear γ-interferon- inhibiting effect in the wash, which means that there is a sufficient amount of antibodies to neutralize the effect of 4 U γ-interferon. It may thus be concluded that the affinity chromatography on the RAH- column, directed against human IgG and IgM, shows that there exist naturally occurring antibodies of the IgG- and IgM-type against all examined interferons in sera from healthy donors. Eluates and washes have at the same time been controlled by radial immune diffusion ad modum Mancini as an additional control (see above).

Figure 11:

Figure 11 shows OD-monitoring of affinity chromatography on an RAH IgM column.

Figure 12:

The bar chart shows the result of the anti-viral neutralization bio-assay with wash and eluate from the RAH IgM column (Fig. 11). It is seen that the eluate has interferon-neutralizing effect vs. native α- interferon (final concentration 1 and 2 U/ml). At the same time it is observed that there is a neutralizing effect in the wash which is due to the presence of naturally occurring neutralizing antibodies of the IgG type. (As the effect in eluates is more significant in the combined RAH IgG + IgM column, indicating that auto-antibodies of the IgG type enhance the neutralizing effect vs. interferon, and eluates from the protein A column show a clear interferon-inhibiting effect, it appears that these auto-antibodies are of the IgG type). On the whole, it can be concluded that there exist naturally occurring antibodies of the IgM type which inhibit the effect of native α-interferon. At the same time it should be mentioned that the major part of the auto-antibodies are immunoglobulins of the IgG type.

Figure 13:

The bar chart shows the result of the anti-viral neutralization bio-assay. It is seen that the eluate has an interferon-neutralizing effect vs. recombinant α-interferon (final concentration 1 and 2 U/ml). Again, however, there is an interferon-inhibiting effect in the wash. This is presumably due to naturally occurring auto-antibodies to recombinant α-interferon of the IgG type (cf. explanation above).

Figure 14:

Figure 14 shows the same as Figs. 12 and 13, but vs. β-interferon. The interferon concentration is final 1 and 2 U/ml, resp. Yet, the inhibiting effect is not as clear in this experiment, but the figure still indicates auto-antibody activity. The weak interferon-inhibiting effect which is seen here, compared to the strong interferon-inhibiting effect seen in eluates from the combined RAH IgG + IgM column and protein A column, indicates that the naturally occurring antibodies to ^-interferon are mainly of the IgG type.

Figure 15:

Figure 15 shows the same as Figs. 12-14, but vs. γ-interferon. The interferon concentration is final 2 U/ml and 4 U/ml. There is effect in eluate as well as in wash (cf. explanation above).

Figure 16:

Figure 16 shows OD- onitoring of affinity chromatography on a protein A-Sepharose® 4B column.

Figure 17:

The bar chart shows the result of the anti-viral neutralization bio-assay with wash and eluate from the protein A column (Fig. 16). It is seen that the eluate has interferon-neutralizing effect vs. native α- interferon (final concentration 1 and 2 U/ml). At the same time it is observed that there is trace of neutralizing effect in wash. This could be due to the presence of neutralizing IgM auto-antibodies, and this has in fact been proved in the Mancini test (see Fig 18). Protein A rather selectively binds antibodies of IgG 1,2,4 and small amounts of IgM and IgA. The following results which are influenced by the strong interferon-inhibiting effect in eluate from this column clearly indicate the presence of auto-antibodies. Figure 18:

The bar chart shows the result of the anti-viral neutralization bio-assay. It is seen that the eluate has interferon-neutralizing effect vs. recombinant α-interferon (final concentration 1 and 2 U/ml). At the same time a trace of neutralizing effect vs. Intron 2 U/ml is observed in wash (see explanation above).

Figure 19:

Figure 19 shows the same as Figs. 17 and 18, but using β- interferon. The interferon concentration is final 1 and 2 U/ml, respectively. Again, a strong interferon-inhibiting effect is observed, and this is in good agreement with previous results, as the protein A column mainly binds immunoglobulins of the IgG type.

Figure 20:

Figure 20 shows the same as Figs. 17-19, but using γ-interferon. The interferon concentration is final 2 U/ml and 4 U/ml. It can, therefore, be concluded that protein A column affinity chromatography, being a highly recognized method, fully confirms the remaining results, and that there exist auto-antibodies with interferon-inhibiting effect of IgG type. It cannot be excluded that some of the few IgA antibodies which are bound by a protein A column, contribute to the interferon- inhibiting effect.

Figure 21:

Figure 21 shows the effect of T-gel wash and eluate in a neutralization assay vs. CIF 1- and 2 U/ml. It is seen that the effect of wash vs. 1 U/ml lies slightly above the result of the interferon control, and that the effect of the eluate lies in the vicinity of the result of the virus control. This means that the effect from the immunoglobulin fraction has been removed from the wash and is now present in the eluate.

The effect of wash vs. 2 U/ml CIF is higher than vs. 1 U/ml, which shows that the interferon effect comes better through here. The effect of eluate vs. 2 U/ml CIF also lies in the vicinity of the interferon control, showing that antibodies being bound to the gel may neutralize the effect of 2 U/ml CIF.

It can be concluded that those antibodies which are bound to the T- gel and thereafter eluted, have a clearly neutralizing effect vs. CIF.

Figure 22:

Figure 22 shows the same as Fig. 21, but vs. Intron 1 and 2 U/ml . It can be concluded that those antibodies which are bound to the T-gel and then eluted, have a clearly neutralizing effect vs. Intron.

Figure 23:

Figure 23 shows the same as Fig. 22, but vs. J-interferon 1 and 2 U/ml. It should be mentioned that the differences are smaller, but still significant.

It can be concluded that those antibodies which are bound to the T- gel and then eluted, have a clearly neutralizing effect vs. β- interferon.

Figure 24:

Figure 24 shows the same as Fig. 21, but vs. γ-interferon 2 and 4 U/ml. It should be noted that the effect of both washes lies slightly below the result of the interferon control. Considering standard deviation, the difference is very small.

It can be concluded that those antibodies which are bound to the T- gel and then eluted, have a clearly neutralizing effect vs. γ- interferon.

COMMENT: It should be mentioned that the T-gel has shortcomings, as it may bind α-2-macroglobulin. Yet, it is often used for purifying antibodies and is here used as an additional control.

Figure 25: Figure 25 shows OD-monitoring of affinity chromatography performed on a column loaded with recombinant α-interferon.

Figure 26:

The principle of affinity chromatography on the recombinant α- interferon column is that interferon is bound directly to matrix. Antibodies recognizing epitopes on the interferon molecule will be caught during chromatography.

It is seen that interferon-neutralizing effect is found in the eluate, indicating that human auto-antibodies to α-interferon with interferon-inhibiting /-neutralizing effect have been retained on the column.

Figure 27:

This figure is drawn through the median of pentaplicates. The standard deviation is not shown, but is everywhere lower than 8%. The figure clearly shows the presence of naturally occurring auto-antibodies with biologically neutralizing effect vs. Lymphotoxin in serum pool. The lowermost curve is the recombinant lymphotoxin standard curve. This curve shows that recombinant lymphotoxin kills the cells and that the effect is concentration-dependent (the higher lymphotoxin concentration, the more cell killings / lower OD-signal). This is a commonly known phenomenon, but it is here desired to prove that the present samples, in this case serum pool 1:18 and eluate from affinity chromatography on a protein A column, contain auto-antibodies with neutralizing effect vs. lymphotoxin. The middle curve is the eluate from the protein A column. The cells are seen to be partly protected when using the eluate from this column. This must be taken as an argument for the presence of neutralizing antibodies of IgG type in this eluate. Serum in dilution 1:18 protects the WEHI-cells even more. This is presumably due to the presence of antibodies of the IgM class with lymphotoxin-neutralizing effect, in addition to the antibodies of IgG type seen in the eluate from the protein A column. As seen, the protein A column preferably binds antibodies of the IgG class. It is noted that eluate and serum per se do not have any toxic effect on the WEHI-cells. Serum as well as the protein A eluate may be diluted so much that the effect ceases, and this will also happen in a usual titration of lymphotoxin.

Figure 28:

This figure is drawn through the median of hexaplicates. The standard deviation is not shown, but is everywhere lower than 10%. The figure shows the presence of naturally occurring auto-antibodies with biologically neutralizing effect vs. tumor necrosis factor (TNF) in normal serum pool. The lowermost curve is the recombinant TNF standard curve. The middle curve is the eluate from the protein A column. The cells are seen to be protected when using the eluate from this column. This must be taken as an argument for the presence of neutralizing antibodies of IgG type in this eluate. Serum in dilution 1:18 protects the cells even more. This is presumably due to the presence of antibodies of IgG as well as IgM class with TNF-neutralizing effect. As seen, the protein A column preferably binds antibodies of the IgG class. It is again noted that eluate and serum per se do not have any toxic or growth-stimulating effect on the WEHI-cells. Serum as well as the protein A eluate may be diluted so much that the effect ceases.

Figure 29: The figure shows that serum HNI IgG from Statens Seruminstitut contains auto-antibodies directed against native α-interferon. A distinct dose/response course is noted, both vs. CIF 2 U/ml and CIF 1 U/ml. For IgG concentrations lower than 0.25 mg/ml, the auto-antibody activity cannot be traced, while concentrations above this level significantly (p<0.01, rank/sum test) inhibit the anti-viral effect of α-interferon. It is seen that for the IgG concentration greater than or equal to 15 mg/ml, the inhibitory activity of the auto-antibodies is total both vs. CIF 1 U/ml and 2 U/ml, as the OD-signal corresponds to the virus control signal. The curves are drawn through medians of pentaplicates. Standard deviation was everywhere less than 12% (not shown).

Figure 30:

Serum is also seen to have inhibiting effect vs. recombinant α- interferon. The titration corresponds quite well to Figure 29. It can be concluded that it is a question of relatively small amounts of auto- antibodies, as the curve for Intron 2 U/ml shows saturation at 0.25 mg/ml HNI IgG. This means, in other words, that such IgG fraction only contains antibodies sufficient to neutralize 2 U/ml Intron, which must be said to be a low interferon concentration. However, this corresponds quite well with the fact that it is a question of auto-antibodies which in healthy individuals occur in small concentrations. The curve shows a dose/response course. The sharp rise of the OD-signal for the Intron 2U curve at abt. 2 mg/ml HNI IgG may indicate a threshold value as a consequence of the fact that the neutralizing antibodies have been exhausted, whereby the interferon effect becomes very strong and approaches the result of the interferon control. 28

Figure 31 :

When the IgG concentration exceeds 1 mg/ml, an inhibitory effect of the antiviral activity of 0-interferon is noted. The phenomenon is concentration-dependent. The weak depression of the OD-signal noted at concentrations lower than 0.25 mg/ml cannot be explained.

Figure 32:

The auto-antibody activity vs. γ-interferon is not very distinct in this test. This may possibly be due to the use of a higher interferon concentration in this experiment. The weak effect may also be a result of the serum quality used.

Figure 33: This experiment has been carried out to ascertain whether the samples (HNI IgG from Statens Seruminstitut, Copenhagen, Denmark) should be toxic to the cells of the biologic neutralization assay. At the same time, it is controlled whether there should be any anti-viral activity in the samples which could disturb the interferon reaction.

Figure 34:

The bar chart shows the result of the anti-viral neutralization assay with Intron A (1 U/ml and 2 IU/ml) and wash from Fab/Fc affinity chromatography on a protein A Sepharose^® gel column. The Fc portion is bound to the column, whereas the Fab portion is found in the wash. It appears from the chart that the Fab fragments neutralize the interferon effect, which substantiates that the auto-antibodies according to the invention exhibit a specific antigen-antibody reaction. The toxicity control shows that the employed amount of Fab fragments was slightly toxic. Immune-depleted wash from affinity chromatography on T-gel did not exhibit any neutralizing effect.

Claims

1. A regulatory protein, CHARACTERIZED in that it binds to the same antigenic determinants on human cytokines as auto-antibodies to human cytokines.

2. A regulatory protein according to Claim 1, CHARACTERIZED in being of the type which may be formed as auto-antibody to human cytokines and which i) in a neutralization test binds cytokine, and ii) by affinity chromatography is bound to an anti-immunoglobulin column.

3. A regulatory protein according to Claims 1-2, CHARACTERIZED in being of the type which is found in body fluids of healthy individuals as auto-antibodies to cytokines.

4. A regulatory protein according to Claims 1-4, CHARACTERIZED in that it is isolated as auto-antibody to cytokines from body fluids of healthy individuals.

5. A regulatory protein according to Claim 1, CHARACTERIZED in that it is isolated by binding to an antigenic determinant on a human cytokine, cytokine analogue or a cytokine fragment exhibiting a human functional antigenic determinant.

6. A regulatory protein according to Claims 1-5, CHARACTERIZED in being of the IgG- or IgM-type.

7. A regulatory antibody according to Claims 1-6, CHARACTERIZED in that it exhibits the same characteristics as such auto-antibodies to cytokines which may be isolated from serum and body fluids from humans.

8. The use of a regulatory protein which binds to the same antigenic determninants on human cytokines as auto-antibodies to human cytokines, for the production of a pharmaceutical preparation for regulation of immunological mechanisms in humans.

9. A solution containing a regulatory protein, CHARACTERIZED in that the regulatory protein exists as auto-antibody to human cytokines.

10. The use of a cytokine, cytokine analogue or a cytokine fragment exhibiting a human functional determinant for the production of a pharmaceutical preparation for regulation of the auto-antibody activity in humans.

11. Use according to Claim 10, CHARACTERIZED in that the cytokine, cytokine analogue or cytokine fragment is selected from the group consisting of interferons, interleukines, tumor necrosis factor and lymphotoxin.

12. An assay kit, CHARACTERIZED in containing as a functional constituent a compound as claimed in Claims 1-7.

13. A pharmaceutical preparation, CHARACTERIZED in containing as an acitve ingredient a compound as claimed in Claims 1-7 and optionally a pharmaceutical carrier.

14. A process for the preparation of regulatory proteins,

CHARACTERIZED by isolating from the antibody fraction of serum or body fluids such antibodies which bind to human functional determinants on cytokines.