CA2105969A1 - Murine monoclonal antibodies recognizing polymorphic determinants of hla - Google Patents

Abstract

Novel monoclonal antibodies are provided herein. Such monoclonal antibodies comprise a set of monoclonal antibodies that react with epitopes on DR4 molecules.

Description

W092/1~698 PCT/CA92/0010~

2 ~

MURINE MONOCLONAL ANTI~ODIES RECOGNIZING
POLYMOhPHIC DETERMINANTS OF HLA

2) Backqround of the Invention fi) Field of the Invention This invention relates to the production and use of a set of monoclonal antibodi~es to HLA and its subtypes and to the set of monoclonal antibodies so produced.
Such monoclonal antibodies are specific for antigens coded by the HLA gene complex and therefore are useful in the typing of human tissue that is to be used in organ transplants.
(ii) Prior Art The human system involves the production of serum proteins, known as antibodies, by the lymphoid cell series capable of reacting with antigenic determinants-which trigger their production. Since the conventional response of the immune system to an antigen with many antigenic determinants is the production of antibodies to each determinant, the anti~erum produced is heterologous in nature and polyclonal, or produced by many different cells each producing antibodies to a specific determinant. Antigenic determinants may be referred to as epitopes when more than one occurs on a single molecule and particularly when each elicits an antibody developing, immune response. A single antibody molecule is specific for a unique antigenic determinant or epitope.
The introduction of foreign material (antigenic material) into the body of a vertebrate animal provokes an immune reaction, the intent of which is to prevent the antigenic material causing damage to the body and to facilitate the rlemoval of such material from the body.
The immune system achieves this by producing immunoglobulin molecules ~hereafter referred to as antibodies) which have the property of selectively . ": '' ,: .
'.

Wo92/l5698 PCT/CA92/OOlOS
2io~969 '.

recognising and binding to characteristic sites on the antigenic material. These sites are known as determinants and an antiyen may possess one or more such determinants. Antibodies generated by the immune system each have specificity to only one determinant but a number of different antibodies may be produced if the antigenic material against which antibodies are raised possesses more than one determinant.
The primary function of antibodies is to protect the body from harmful foreign material, by agglutinating it, thereby assisting the normal body processes to remove the material.
Antibodies are proteins that have the ability to combine with and recognize other molecules, known as antigens. Monoclonal antibodies are no different from other antibodies and, except that they are very uniform in their properties, recognize only one antigen or a portion of an antigen known as a determinant.
In the case of cells, the determinant recognized is an antigen on or in the cell which reacts with the antibody. It is through these cell antigens that a particular antibody recognizes, i.e. reacts with, a particular kind of cell. Thus the cell antigens are markers by which the cell is identified.
These antigenic markers may be used to observe the normal process of cell differentiation and to locate abnormalities within a given cell system. The process of differentiation is accompanied by changes in the cell surface antigenic phenotype, and antigens that distinguish cells belonging to distinct differentiation lineages or distinguish cells at different phases in the same differentiation lineage may be observed if the correct antibody iS available.
Human antibodies have been used both for diagnostic and therapeutic purposes for a number of years.
Diagnostic techniques include blood typing for ' ' ' W092/15698 PCT/CA92/OOlOS
21~9g9 !

transfusion or transplantation. The oldest is isolation from immune serum. However, the small concentration of the antibody of desired specificity among those which are generally present in serum presents a serious drawback.
Conventional antisera, produced by immuni~ing animals with tumour cells or other antigens, contain a myriad of different antibodies differing in their specificity and properties. In 1975 Kohler and Milstein introduced a procedure which lead to the production of quantities of antibodies of precise and reproducible specificity. The Kohler-Milstein procedure involved the fusion of spleen cells (from an immunized animal) with an immortal myeloma cell line. By antibody testing of the fused cells (hybridomas), clones of the hybridomas were selected that produced antibody of the desire specificity. Each clone continued to produce only that - one antibody, monoclonal antibody (mAb). As hybridoma cells can be cultured indefinitely (or stored frozen in liquid nitrogen), a constant supply of antibody with uniform characteristics is assured.
The preparation of hybridoma cell lines can be successful or not depending on such experimental factors as nature of the inoculant, cell growth conditions, hybridization conditions etc. Thus~-it is not always possible to predict successful hybridoma preparation of one cell line although success may have been achieved with another cell line.
The usefulness of monoclonal antibodies has been confirmed not only in the field of immunology but also in many other ~ields. They are hence employed widely.
However, since these antibodies are produced primarily with hybridoma originated from mouse cell, certain inherent limitations are obviously imposed on their application for the diagnosis and treatment of human.
Upon formation of a human monoclonal antibody, it is necessary to obtain a cell which has been challenged by WO92/15698 PCT/CA92/00l05 a desired antigen and can produce a human antibody specific to the antigen. The in-vivo antigenic stimulation is not feasible in human except for a certain class of antigens and there have not been established any method applicable to various antigens.
It is attempted to obtain permanently-established human cells by immortalizati.on of antibody-producing cells, for example, by their fusion with human myeloma cells, their transformation with an Epstein-Barr virus (EBV) or the like and then to obtain a monoclonal antibody from the permanently-established human cells.
Unlike in mouse lines, hybridomas or transformed cells having the ability of stable production of antibody have not been obtained in human lines for the time being.
15Monoclonal antibodies are uniform antibodies directed to a single determinant or epitope on the antigen molecule which may be repeated at several sites of the molecule. Obviously, to produce such monoclonal antibodies in vitro requires selecting a homogeneous antibody having the desired specif.ications from numerous antibodies elicited in a conventional polyclonal response.
More recently, the production of human monoclonal antibodies has become possible and these may serve as -tools in diagnostic testing and in therapy. Two major approaches are new possible for the production of mAB
direct immortalization of immunized lymphocytes with Epstein-Barr Virus tEBV) and mab production by hybridomas formed between immortalized human B cell lines (EBV~, lymphoblastoid, or human or murine myelomas, and human B
lymphocytes from an immunized host. Neither of these approaches has proved entirely satisfactory.
It is common,experience among practitioners in the art that EBV transformation, while successful in forming 35 Mab-secreting cultures, will often fail to provide ~ -antigen specific E~3V transfor~ed cells which have W O 92/15698 PC~r/CA92/0010~
2~ Q5969 s sufficiently long life spans to provide reliable sources of the desired antibodies. This method fails to provide reliably for antibody production over extended periods.
Previously produced hybridomas between immunized human B
cells and appropriately drug marked murine or human myeloma or human lymphoblastoid cell lines have suffered from low frequency of hybrid formation in the case of human-human hybridizations. Mur.ine-murine hybridomas are stable, but the antibodies produced are immunogenic if used in passive therapy.
An immunized experimental animal can sometimes serve as a source for specific antibody secreting B cells to provide the immunized lymphoid member of the hybridoma.
This method cannot be used, however, to provide reagents for HLA or other blood type testing since when human antigens are injected, the plethora of antibodies elicited is mostly immunoreactive to antigens common to all humans, and the desired antigen-specific antibody is formed only as a very small percentage of the total response. Further, these non-human antibodies can themselves result in an adverse immune response if injected for human therapy.
There are two principal classes of lymphocytes involved in the immune system of humans and animals. The first of these (the thymus-derived cell or T cell) is differentiated in the thymus from haemopoietic stem cells. While within the thymus, the differentiating cells are termed "thymocytes". The mature T cells emerge from the thymus and circulate between the tissues, lymphatics, and the bloodstream. These T cells form a large proportion of the pool of recirculating small lymphocytes. They have immunological specificity and are directly involve~ in cell-mediated immune responses (such as graft rejection) as effector cells. Although T cells do not secrete humoral antibodies, they are sometimes required for the secretion of these antibodies by the : ~ .

WO92/1~698 PCT/CA92/00105 2~ 6 second class of lymphocytes discussed below. Some types of T cells play a regulating function in other aspects of the immune system. The mechanism of this process of cell cooperation is not yet completely understood.
The second class of lymphocytes (the bone marrow- ;
derived cells or B cells) are those which secrete antibody. They also develop from haemopoietic stem c~lls, but their differentiation is not determined by the thymus. In birds, they are differentiated in an organ analogous to the thymus, called the Bursa of Fabricius.
In mammals, however, no equivalent organ has been discovered, and it is thought that these B cells differentiate within the bone marrow.
It is now recognized that T cells are divided into at least several subtypes, termed "helper", "suppressor", and "killer" T cells, which have the function of (respectively) promoting a reaction, suppressing a reaction, or killing (lysing) foreign cells. These -subclasses are well understood for murine systems, but they have only recently been described for human systems.
The ability to identify or suppress classes or sub-classes of T cells is important for diagnosis or treatment of various immunoregulatory disorders or conditions.
2S Previous suppressor factors have been reported in the literature. Such factors have potential use for example, in the treatment of patients with cancer, graft versus host diseases, autoimmune diseases and lympho-proliferative malignancy disorders, e.g. leukemia. -The discovery of the human major histocompatibility complex (hereinafter "MHC") was when leukoagglutinating antibodies were first found in the sera of multiply transfused patie~ts in a pattern that suygested the - antisera were detecting allo-antigens, antigens present on the cells of some individuals of a given species which are products of a polymorphic genetic locus. The role of these antigens in determining the success of tissue and organ transplants was soon appreciated and provided the initial studies of genes that determine human leucocyte antigens (hereinafter termed "HLA").
The HLA system is extremely polymorphic, having multiple different alleles at each known genetic locus.
Based on their tissue distribution and structure, HLA
antigens have been divided into two broad classes: Class I antigens which include the HLA A, HLA-B, and HLA-C
antigens found on virtually every human cell and which have counterparts in other mammalian cells including the murine system; and Class II antigens including the HLD-, DR, DQ, and DP antigens found chiefly on the surface of immunocompetent cells including macrophages/monocytes, activated T-lymphocytes, and B lymphocytes. Class II
antigens also have counterparts in other mammalian systems such as murine mammals. The presence of these Class I and Class II antigenic molecules plays a major role in the functional heterogeneity of p~ripheral T-cells.
The different re~ulatory and effector functions ofT-cells are mediated by different subpopulation of cells which can be distinguished by differences in their phenotypes and antigenic determinants (identifiable by different monoclonal antibodies). This has led to the typing T cell functional subsets in accordance with the expression of specific surface molecul~s which are commonly designated by the letter "T" followed by a number. Based on functional differences between T4 and T8 cells, the peripheral blood T-cells can be broadly divided into two populations: one population constituting approximately 65% of peripheral blood T-cells is T4+; thelother constituting approximately 35% of all peripheral blood T-cells is T8+. The T8+ cell may be activated to become a cytolytic T lymphocyte (hereinafter termed "CTL cell") which functions as a cytotoxic W092/1~698 PC~/CA92/00105 ~9~ :

effector cell and plays an important role in the hosts' defense against foreign bodies. In combination with natural killer cells (hereinafter termed "NK cells") and lymphokine activated killer cells (hereinafter tPrmed "LAK cells"), these cells respond to protect the body against invasion by foreign c:ells and substances. The role of the T4~ cell has been traditionally viewed as an inducer cell for the activation of other T-cell subpopulation. This role is achieved in combi~ation with an accessory cell or antigen presenting cell (hereinafter termed "APC") which bears Class II MHc molecules on its surface and is able to take up and process an identifiable antigen. The antigen presented by an APC
bearing Class II molecules activates specific T4+ cells.
The activated T4+ cells in turn secrete a variety of lymphokines to initiate the effector and cytolytic functions of other T-cell lymphocytes. It is noteworthy that with the increasing use of lymphocyte-mediated immunotherapies including those directed against tumours, all such immunotherapies utilize only those activated lymphocytes equipped with cytolytic effector function, e.g. CTL cells, NK cells, and LAK cells. T-cells of the inducer phenotype are traditionally viewed as lacking the necessary cytolytic activity and therefore have not been considered useful for treatment of tumours as immunotherapeutic lymphocytes.
The human system of multiple T-cell subpopulation has ~ direct counterpart in the murine system. There are two major functional subsets of T-lymphocytes in the murine system. The L3T4+ subset of T-lymphocytes has inducer or helper functions and is generally activated by APCs that bear exogenous antigen and express Class II
molecules (Ia) of the MHC. This subset is equivalent to the T4+ lymphocyte subpopulation in humans. The second major T-cell subset expresses Lyt-2 determinants and possesses either suppressor or cytolytic functions.
.
.
r WO92/15698 PCT~CA92/00105 21 0~969 These are equivalent to T8+ lymphocytes in human~. When activated, Lyt-2 cells become cytolytic T-lymphocytes (CTL cells) which generally lack the L3T4+ antigenic marker and which recognize Class I molecules of the MHC.
SAs in the human system, it has been traditionally viewed that L3T4+ inducer T cells help initiate the effector functions of other T-lymphocytes, but do not demonstrate any cytolytic effect themselves. Very recently however, several investigators observed an effector function for selected L3T4~, antigen-specific, Ia-restricted T-cell clones. These investigations comprised in-vitro experiments in which selected L~T4+ -clones appeared to be cytolytic in short-term (less than 6 hours) chromium release assays for Ia-bearing B-cell hybridoma targets in the presence of antigen. The primary thrust of each report dealt with the specificity and the killing mechanisms for the observed cytotoxicity.
To date, therefore, there is little knowledge or ; `
appreciation as to: whether inducer T-cells generally in murine and human systems are able to express cytolytic effector function; whether all major types of antigen presenting cells are sensiti~e to such cytolytic activity; whether such cytolytic activity can be ~`
maximally expressed and, if so, under what conditions; ~ `
25 and whether such inducer T-cell cytolytic activity can be "
utilized in-vivo for any therapeutic purpose.
Numerous methods have been already proposed for detection of HLA antibodies. Among these methods, -techniques measuring the release of radioisotopes, e.g. ;
5ICr from labelled target cells or of fluorescent compounds were said to have a high degree of accuracy.
Unfortunately, they require too large a number of cells `-and serum and areltoo time consuming to be considered for routine or emergency HLA typing.
3SThe presently most frequently used method for determination of complement-dependent cytotoxicity ~ "

WO9~i/lS698 PCTiCA92/00105 2`~9~

mediated by anti-HLA antibodies is the microcytotoxicity test (designated by the abbreviation MCT) which was first proposed by Mittal et al. This method, which consists in measuring the uptake of supraYital dyes by lysed target cells, has been found to be a simple, reliable and ' accurate method for performing routine HLA typing or cross-match test and has been universally accepted by most histocompatibility laboratories. These laboratories have standardized the originally described process with only minor changes from one to another.
Most of the successive technical steps in the MCT
process have now been automatised with the exception of the final step which consists in the re~ding of the percentage of dead cells. Since this reading should be performed under microscope, it depends on the use of the technician to appreciate the uptake of the supravital dyes and consequently could not be automatised. In addition, this MCT process is often considered to be inadequate for precisely measuring the extent of cell damage. Thus, for these two reasons, it could be of a great interest for histocompatibility centres, to provide a more precise and easily quantifiable method for measuring anti-HLA mediated complement-dependent cytotoxicity. Many attempts have been made in order to 25 find such a method. -Transplantation of human organs, particularly the kidney, has become a relatively common procedure.
Ideally, the donor organ is obtained from an identical twin since the antigens of the donor and recipient in such a case are identical and no histoincompatibility exists. Therefore, no immune response to the graft occurs in such a transfer, known as an isograft.
However, most tra!nsplants are between two less closely related inclividuals of the same species and histocompatibility differences in such an allograft may be strong or weak, depending on the individuals. The W092/15698 PCT/CA92/00l0~
21~96~

fate of transplanted tissues and org~ns depends on a number of factors, but the recipient's immune response to graft antigens is the central event. Definition of antigenic systems which serve as strong barriers to transplantation has therefore become a major investigational interest, having both practical application in clinical transplantation and theoretical value in understanding the natural role of the histocompatibility antigens in immunobiology.
A single chromosomal gene complex codes for the major histocompatibility antigens in each vertebrate species investigated so far. In humans, the - histocompatibility antigens are produced by the HLA gene complex. This complex occupies a portion of the short arm of the human C6 chromosome and consists of several series of paired alleles which are inherited from generation to generation in a dominant fashion, segregating randomly from other important antigens such as the ABH red blood cell type groups.
Antigens of the HLA system are divided into two classes. Each class I antigen consists of an 11.500-dalton ~2-microglobulin sub-unit and a 44.OCO-dalton heavy chain which carries the antigenic specificity. Three gene loci (A, B and C) are recognized for the class I
antigens. There are over sixty clearly defined A and B
specificities while 8C locus specificities are known.
Evidence that this gene complex plays the major role in the transplantation response comes from the fact that haplotype-matched sibling donor-recipient combinations show excellent re~ults in kidney transplantation, in the vicinity of 85% to 90~ long term survival.
A non-serologically defined antigen, responsible for the mixed lymphocyte response r is caused by a distinct locus called D. Although D-locus antigens are not as yet clearly identifiable by serotyping techniques, serologically defined specificities closely related to -- . .... , ~,, WO92/15698 PCT~CA92/00105 '2105969 12 the D-locus have been defineclO These have the special property of that being expressed on platelets or unstimulated T lymphocytes. These specificities are termed class II having two glycoprotein chains of 29.(XX) (~) and 34.(XX~ (~) daltons and lacking ~2 globulin.
These antigens are also termed HLA-DR (D-related) and are important in tissue typing. -Tissuè typing is currently being carried out using sera obtained from multiparous women. A major problem exists because of the unreliability of this source. only a limited amount of antibody is available from any one woman. Accordingly, it is necessary continually to replace standard antibodies with new antibodies which must be standardized and checked against the previously existing ones. Furthermore, because of the heterogeneous nature of antibodies obtained in this fashion, cross-reactivity is a major problem. Accordingly, there have been many attempts to produce antibodies more suitable for crossmatching by immunological techniques. Several such attempts involve xenoimmune (cross-species) sera to detect allospecificity in the antigen-donor species.
Examples include rabbit anti-human antibodies and monkey anti-human antibodies. One specific ~nti-HLA serum appears to be a rabbit anti-A9 serum prepared by immunization with A9 antigen purified from human serum or urine. ~owever, A9 is actually a common determinate of A23 and A24, and thus possibly the allele-specific major epitope is not the only target with which this antibody reacts.
The adven~ of hybridomal techniques has brought about the possibility of producing homogeneous populations of highly specific antibodies against a variety of antigèns. Antibodies have been produced by somatic cell hybrids between myeloma cells and spleen or lymph cells that are specific for malignant tumours.
Continuous cell lines have been produced of genetically-2~.as969 stable fused-cell hybrids capable of producing large amounts of IgG antibodies against specific viruses.
Hybridomas have been provided which produce monoclonal IgG antibodies against tetanus toxin.
Monoclonal antibodies have also been described against human tumour cells. ~n acldition, there have been disclosures of the use of rat-mouse hybridomas and their application to studies of the histocompatibility complex in various species. Several rat-mouse hybridomas producing rat, anti-rat antibodies have been found which were reactive with determinants on cells from other species, e.g. humans. Some hybridoma antibodies found which were reactive with peripheral blood lymphocytes from randomly chosen humans in a complement-mediated cytotoxicity assay. However, such antibodies were not useful in identifying private determinants of the HLA
locus since no correlation between specific HLA antigens and the hybridomally produced antibodies occurred.
Matching for antigens determined by the HLA genes is an important component of the whole process of transplantation of organs and tissues. In the case of bone `marrow transplantation this tissue typing and matching becomes critical; certain mismatches may lead to the death of the patient. Genes determine the production of at least six types of molecule of immunogenetic importance which fall into two classes, I and II. Class I molecules are called H~A-A, HLA-B and HLA-C while Class II molecules are called HLA-DR, HLA-DQ and HLA-DP, with some further subdivisions within these. All the molecules are extraordinarily polymorphic and so typing patients and donors is complex. Finding matches outside of family groupings is usually difficult, since fully matched pairs in the general population are rare.
Typing for Class I molecules and for HLA-DR is well 35 advanced. HLA-DQ is less well served with reagents and ! ~' ' HLA-DP is for all practical purposes not typed by routine ~..

..

WO92/15698 PCT/CA92/00~05 2~05 g~9 14 tissue typing laboratories. The reason for the lack of DP typing is the unavailability of antibodies against the H~A-DP polymorphism.
The patent literature is replete with patents which 5 relate to monoclonal antiboclies and cell lines and methods of use thereof for therapy and typing. For example, U.S. Patent No. 4,314,026 patented February 2, 1982 to B. Descamps-Latscha provided a process for determining the complement-dependent cytotoxicity 10 mediated by anti-HLA antibodies by means of ~TP
determination and device for ATP determination. That patent provided such a process which included measuring the loss of intra-cellular ATP after addition of complement to anti-HLA-coated target cells. Cytolysis 15 was determined by measuring the intracellular ATP content of human lymphoid cells (target cells) after these latter had been incubated with antiserum (anti-HLA antibody) and complement (rabbit serum). When both target cells and serum shared the same HLA specificities, cell lysis was 20 observed and exteriorised by a dramatic loss of its intracellular ATP content. r U.S. Pater.t No. 4,517,289 patented May 14, 1985 by E. L. Milford et al, provided monoclonal antibodies for human tissue cross-matching. That patent provided an 25 immortal, antibody-producing, hybridomally-produced clone and an antibody produced thereby. The antibody was an immunoglobulin specific for an antigenic determinant encoded by an HLA gene complex in humans. The clone was produced by an immortal cell line fused with a lymphocyte 30 obtained from a first rat immunized against cells obtainDd from a second rat having a different histocompatibility antigen. That patent therefore also provided novel hybridoma cell lines, novel monoclonal antibodies against a human HLA antigen, the antibody 35 having been produced by a novel hybridoma cell line and wo92/ls69B PCT/CA92/00105 21~9~9 a tissue-crossing assay kit, including a monoclonal antibody produced by a novel cell line and a dye.
U.S. Patent No. 4,634,66~ patented January 6, 1987 by E. G Engleman et al, provided an ideal fusion partner for specific B-lymphoid cell lines, producing triomas that secreted specific antibodies of human character.
The immortalizing, non-secreting hybridoma having human characteristics was prepared by fusing mouse myeloma cells with human B lymphocytes and selecting the fusion product for stable immunoglobulin secretion and HLA
surface antigen production, followed by treating the selected fusion product with mutagen and selecting the mutated product for non-secretion of immunoglobulin but retention of ~LA antigen production. That invention also provided the products of fusing the immortalizing hybridomas with suitable human immunized lymphoid cells.
Such triomas are useful sources of desired Mab's. That invention also provides human monoclonal antibodies which are produced by the triomas and their diagnostic and -therapeutic compositions and uses. Thus, the above patent provided a specifically recited immortalizing fusion partner for use in producing a trioma cell line capable of secreting a human monoclonal antibody specific against a selected antigen, when fused with a non-malignant b-lymphoid cell derived from a human donor exposed to such antigen. It also provided a trioma cell line capable of secreting a normal human monoclonal antibody specific against a selected antigen. The cell line was the fusion product of a mouse myeloma/non-malignant human B-lymphocyte hybridoma fusion partner which expressed HLA surface antigens, did not secrete immunoglobulins, and was deficient in hypoxanthine phosphoribosyl transferase, as evidenced by the inability of the fusion partner to grow in hypoxanthine-35~ aminopterin-thymidine or azaserin-hypoxanthine medium, WO92/15698 PCT/CA92/00]05 2~59~9 and a non-malignant B-lymphoicl cell derived from a human donor exposed to the selected antigen.
U.S. Patent No. 4,657,760 patented April 14, 1987 by P.C. Kung, provided methods and compositions using monoclonal antibodies to human T cells. This patentee provided a novel hybridoma which was capable of producing a monoclonal antibody against an antigen found on essentially all normal human peripheral T cells. The antibody so produced was monospecific for a single determinant on normal human T cells and contained essentially no other anti-human immuneglobulin. The patentee also provided a novel hybridoma producing antibody to an antigen found on essentially all normal human T cells, the antibody itself, and diagnostic and therapeutic methods employing the antibody.
U.S. Patent No. 4,681,760 patented July 21, 1987 provided a method of conferring immunotolerance to a specific antigen. That patent provided a method for suppressing undesired immune responses, e.g. allergic reactions, to antigens whose administration to the subject was either desired or inevitable but otherwise harmless. It also provided a method for inducing tolerance to tissue transplants. The patented method involved the co-administration of the antigen for which immunotolerance is sought and an antibody which is specific for the "L3T4-equivalent" differentiation antigen on T cells, thus preventing these helper T cells from participating in the immune response otherwise concurrently mounted against the particular co-inject~d or co-administered antigen.
U.S. Patent ~o. 4,692,405 patented September 8, 1987 by A. Freedman et al, provided monoclonal antibodies to antigen on activated human B-cells and assays therefor, protein antigenic determinants therefor and methods of making same. That invention provided a monoclonal antibody recognizing an antigenic determinant on .. "~.

WO92/15698 PCT/CA92/OOlOS
2~9~9 activated human B-cells. That invention also provided a substantially pure protein having an antigenic determinant or determinants substantially identical to determinants of a single-chain polypeptide having an apparent molecular weight of approximately 75,000 daltons under reducing conditions and 67,000 daltons under non-reducing conditions, the single-chain polypep~ide being a protein on the surface of activated human B-cells.
That invention also provided a specifically recited 10 process for preparing the antigenic protein. That `
invention also provided kits useful for assaying a '~
biological sample for the presence of cells expressing the antigen of the invention and for assaying a biological sample for the presence of antibody to the cells expressing the antigen of the invention. These kits contained one or more containers, each holding separately detectably labelled or unlabelled antibody or antigen of the invention, and in another compartment, a means for detecting the formation of immunocomplexes.
U.S. Patent No. 4,708,930 patented November 24, 1987 to K. H. Kartright et al. provided murine, monoclonal antibodies specific to a unique antigenic determinant on the surface and in the cytoplasm of human neoplastic`
tissue are produced. The unique antigenic determinant ',~
25 was designated the "KC-4 antigen" which was capable of ' eliciting an antibody which bonded selectively only to neoplastic carcinoma cells and not to normal human tissues. ''~
U.S. Patent No. 4,710,4S7 patented December 1, 1987 -by Bo Dupont e_ al, provided monoclonal antibodies for human haematopoietic glycoproteins. The patentee also provided a specifically recited method for differentiating between human B cells and T cells in a human haematopoietic specimen.
U.S. Patent No. 4,743,681 patented May 10, 1988 by P.C. Kang et al, provided a hybrid cell line for W092/15698 PCT/CA92/OOlOS
2~05~g~ , producing monoclonal antibody to a human T cell antigens and antibodies. The patentee provided a hybridoma (designated OICTll) which was capable of producing monoclonal antibodies against an antigen found on essentially all normal human peripheral T cells and on approximately 95% of normal human thymocytes, but not on normal human B cells or null cells. The antibody so produced was monospecific for a single determinant on essentially all normal human peripheral T cells and contained essentially no other anti-human immune globulin.
U.S. Patent No. 4,843,004 patented June 27, 1989 by C. Platsoucas provided a specifically recited method for the production of human T-T cell hybrids and production suppressor factor by human T-T cell hybrids. The patented method was developed for the production of human haematopoietic cell hybrids especially T-T cell hybrids as determined by HLA typin~. Some of these T-T cell hybrids pr~duce factors useful for biotherapy or exhibiting specific-immunological functions. This is accomplished by fusing cells from human T cell lines with appropriately sensitized or induced human T cells exhibiting specific immunological function or producing the desired factors.
U.S. Patent No. 4,861,589 patented August 29, 1989 by S. T. Ju, provided a method for therapeutically treating abnormal cells expressing a major histocompatibility complex class II antigen using cytolytic inducer T4 cells. That patent provided a specifically recited method for treating a subject afflicted with tumour cells expressing a major histocompatibility complex Class II antigen either constitutively or~inductively.
U.S. Patent No. 5,009,995 patented April 23, 1991 by A. Albino, provided monoclonal antibodies to melanoma cells. The patent related to monoclonal antibodies r .- . ' ~. : . , ' ~:............ . , . . : .
- : . : : . . .

WO92/15698 PCT/CA92/0010~ ~

recognizing the gpl30 antigen of human cells. Monoclonal antibodies which recognize di~;tinct determinants on this antigen and methods of detecting the determinants by immunoassay with the monoclonal antibodies which recognize them are also disclosed. Hybridoma cell lines which produced such monoclonal antibodies were also disclosed. The monoclonal antibodies are useful in the detection of the gpl30 antigerl and human cells including melanoma which contain this antigen.
lo In spite of the technical literature and the patents described above there still exists a need ~or monoclonal antibodies suitable for tissue matching.
It is common knowledge that attempts to prevent unwanted immune responses have not been particularly successful. For example, efforts are made to match transplant recipients with donors so as to minimize the amount of immunogenic response to foreign materials.
Only in the case of identical twins c~n reasonable success be certain. The limitations of such an approach are so apparent as to warrant no further comment.
Alternatively, brute force efforts to suppress the immune system in general, such as administration of anti-mitotic agsnts may prevent rejection at the expense of the recipient's life due to the resulting susceptibility to infection.
An alternate approach applicable only to preventing tissue rejection is passive immunization of recipients with antibodies directed against the histocompatibility antigens. Other approaches also applicable only to the transplant rejec$ion problem have employed treatment of the donor tissue. These are based on the assumption that the rejectiGn response is caused by the histocompatibilitly antigens on the surface of passenger leukocytes carried on the transplant which leukocytes are not an essential part of the desired tissue Per se. In-vivo culture of the donor transplant tissue has been used , ,," '"', WOg2/l5698 PCT/CA92/00105 S9~

to eliminate passenger leukocytes. The donor tissue has also been treated directly with suitable antibodies~
the use of immunotoxins formed by conjugating antibodies with a cytotoxic moiety has also been suggested for pretreatment of donor tissue.
Methods to prevent imm~me responses to soluble antigens have been largely ~onfined to avoidance of exposure. Patients allergic to certain drugs are treated with alternative formulations when available; hay fever sufferers attempt to stay away from the immunogenic pollen. If avoidance is impossible, one must resort to treating the symptoms.
The major histocompatability complex (MHC) of human is a cluster o~ genes occupying a region located on the sixth chromosome. This complex, denoted HLA (Human Leukocyte Antigen), is currently divided into five major gen loci, which according to World Health organization nomenclature are designated HLA-A, HLA-B, HLA-C, HLA-D, and HLA-DR. The products of the HLA genes are commonly called ~'antigens~. The genes of the A, B, and C loci encode the classical transplantation antigens whereas the genes of the D and DR loci most probably encode antigens that control immune responsiveness. HLA antigens are present in the membranes of human cells. Some are present in most cells of the body whereas others are present only in specific kinds of cells. For instance, HLA-DR antigens have been identified in B cells but not in resting T cells.
The HLA antigens are categorized into types that vary from individual to individual. HLA typing is used in paternity determinations, transplant and transfusion compatibility testing, blood component therapy, anthropological studies, and in disease association correlation to diagnose diseases or to predict susceptibility to disease. Current HLA-DR typing techniques consist of two basic methods. One involves , ~, .. . ' .. , , ' ' . , .. , .' '~ . , ' ' WO92/1~698 PCT/CA92/00l05 2~ ~3~

separating B cells from a tot:al lymphocyte sample, e.g.
peripheral blood lymphocytes (PBL), treating the B cells with anti-DR sera and complement, and reading the resultant cytotoxicity as an index of reactivity. The B
cells are separated from the total lymphocyte population because DR antigens are present only in B cells and B
cells constitute only a small proportion, typically 10%
to 25%, of PBL. cytotoxicity of such a small proportion of cells would be difficult to discern accurately.
Numerous methods have been used previously to separate B
cells from PBL. The most common method takes advantage of the reaction of T cells with sheep erythrocytes (SRBC) to form rosettes that can be centrifuged through a layer of Ficol-Hypaque, leaving the B cells at the top of the gradient. Other methods take advantage of the affinity of B cells for various materials such as nylon wool, Degalon beads, and anti-human F(ab')2 reagent. These methods suffer from various combinations of being time consuming or technically dif~icult, yielding impure preparations ti.e. contamination with non-B lymphocytes), providing poor absolute yields of testable B cells or, yielding separated B cells that have poor viability.
Monoclonal antibodies against HLA-DR antigens have been used to separate B cells from PBL for use in HLA-DR
typing tests.
The second basic HLA-DR typing method s the two colour fluorescence technique. In this method, to label B cells with an immunofluorescent marker, a PBL -preparation is incubated with a fluorochrome labelled anti-human Ig, washed, and then dispensed in tissue typing trays. Following sequential incubations with DR ~- -antisera and complement the test results are read by --determining the percent of viable B cells remaining by adding a fluorescent vital dye and measuring percent 35 viability only of those cells having ring -immunofluorescence. Although this method avoids a B cell ', r~

separation step, it requires that the cells be stained with anti-human Ig. It also is practical only when read under high power microscopy and, therefore, has a more demanding reading step than the B cell separation method.
The serologically-defined HLA-DR4 specificity is complex and has recently been reported to have eight allelic variants or subtypes. These subtypes have been defined mainly by T cell recognition methods and have been confirmed by DNA typing techniques. From analysis 10 of sequence data it is apparent that the serologically !
-defined DR4 specificity can be attxibuted to amino acid differences in the first and second hypervariable regions of the first domain of the DR4 molecule, whereas the subtypic differences are all located in the third hypervarible region. Some subtypes vary by as little as one amino acid and at the most by three; yet these differences are enough to be recognized by T cells.
It has been known for some time that DR4 is associated with susceptibility to developing Rheumatoid Arthritis (RA) and that this association is mainly with the subtypes Dw4 and Dw14 in caucasians and ~lacks.
These two subtypes and the Dw15 subtype which is associated with RA in the Oriental populàtion, vary by only one to three conservative amino acid substitutions at positions 71 and 86 for Dw4 and Dw14, and at positions 57 and 71 for Dw4 and Dw15. What is most intriguing is that two non-DR4 alleles, DRl and DR14 (subtype Dw16) which have major differences in the first and second hypervariable regions from each other and from DR4, have almost complete sequence homogology with Dw14 in the third hypervariable region (1). Both these specificities have been shown to be associated with predisposition to RA in different ethnic groups, for example, DR1 in the Jewish population and DR14 (Dw16) in the Yakima Indians (7,8). Although one can sperulate on the role of a putative epitope formed by these residues, which is - ' : "~ ,, ,,." .~ ,.." ,, ",, .,~ ", "
.. . , .,. ." . , .. , . .~

WO92/15698 PCT/CA92/0010~

23 2~
located on the alpha helix of the peptide binding site, in antigen presentation to T-cells in RA individuals, the ;
relevance of the association still remains a puzzle.

5 3 ! Summary of the Invention (i) Aims of the Invent:ion -: -What is desired is a specific immuntolerance with respect to a particular antigen, leaving the general competence of the immune system intact. None of the foregoing approaches, described in the technical and patent literature above, achieved such a selective immunosuppression of the subject. Treatments employed to prevent transplant rejection which are directed toward the host Der se generally depress the entire system;
15- treatments of the donor tissue alter the nature of the foreign material introduced. In the case of allergic responses to drugs or to environmental antigens, alteration of the foreign material is either undesirable or impractical. In the present invention, the immune system of the host is selectively and specifically suppressed with respect to a particular immunogen without impairing general immunocompetence.
It is also desirable to provide a process which is suitable for HLA typing of human lymphoid cells, and which is also appropriate for anti-HLA antibodies detection in ~he serum from subjects sensitized against --histocompatiblity antigens(-polytransfused patients, multiparous women, organ graft recipients~.
Accordingly, it is an object of this invention to ~ -~
provide a hybridoma cell line and an antibody produced thereby useful for the tissue typing of human tissues.
It is a further object of this invention to provide a method of prod~cing such antibodies in an efficient manner.
A principal object of the present invention is to provide a simple and effective HLA-DR typing technique WO92/15698 PCT/CA92~00105 ~59~9 that: (1) does not involve a B cell separation step or a lymphocyte staining step; and ~2) is based on cytotoxicity function such that the sera and Gomplement used in available lymphocytotoxicity tests may be used in the invention method.
Another object of this invention is to provide murine monoclonal antibodies recognizing polymorphic determinants of HLA-DP.
rii~ ~Statements of Invention lo since a substantial body of work has already been done using T cell cloning techniques to try to understand the role of DR4 in RA, it was decided to put efforts intG
producing monoclonal antibodies (moab) to XLA-DR4 and its subtypes. Such antibodies would enable the testing of whether non-DR4 individuals who develop RA have conformationally-equivalent epitopes to Dw4. If proven correct, antibodies that recognize such epitopes would be useful in identifying individuals at risk for RA and additionally, may also prove useful in specifically blocking peptide presentation to T cells. At the very least such antibodies would simplify DR4 subtyping for tissue matching for organ transplantation.
The DR4 specificity was originally defined by alloantisera derived from multiparous females but attempts to subtype with such reagents have been mostly unsatisfactory. Attempts to make murine monoclonal antibodies to HLA antigens generally proved to be more difficult than had been anticipated. This is thought to be due to the type of immunogen, usually whole cells, which express an enormous array of different molecules including at least six different HLA antigens= The murine immune system recognizes most of these molecules as foreign and ev~n when purified HLA molecules are used, the bulk of the antigen-specific cells will be against ~5 the species-specific or monomorphic determinants present on the histocom~atibility molecules.

WO92tl~698 PCT/C~92/00~05 2~

The development of mousQ transfectant cell lines expressing human histocompati~ility molecules seemed to - be a tremendous advance in this technology, particularly with respect to an anti-DP moab made using a transfectant as an immunogen. Theoretically, the only foreign molecule expressed on the surface of the transfectant should be a HLA molecule. Therefore, the bulk of the ~;~
antigen-specific B cells should be directed to HhA
molecules and some of these should be directed to polymorphic determinants. The ability to differentially screen hundreds of hybridoma supernatants for specific antibody in a short time as was provided by transfectants, was also an important advance.
The present invention provides a set of monoclonal antibodies that react with epitopes on DR4 molecules.
Specifically the present invention provides monoclonal antibodies which are specific for HLA-DR4 molecules.
The present invention also provides the use of the murine monoclonal antibodies to detect subtypes of DR4.
It also provides the use of the murine monoclonal antibodies which react with putative RA-susceptibility determinants for the study of rheumatoid arthritis.
The present invention also provides for the production of 14 such monoclonal antibodies, and for the 25 characterization of the properties thereof. -The present invention also provides for the producing and of analyzing the specificities of moabs to the subtypes of HLA-DR4 using transfectants.
The present invention also provides two other antibodies with DR4 subtypic specificity, that were produced from mice immunized with human molecules.
(iii) Other Features of the Invention Embodiments of such antibodies include the following: NFLD.Dl which binds to all DR4 molecules;
NFLD.D12, wh:ich binds only to the Dw4 subtype of DR4;
NFLD.D14, which binds to Dw4 and Dw14 subtypes; NFLD.D7, 2~ ~963 26 which binds to all DR4 and DR2 molecules but also, less strongly with several non-DR4 molecules; NFLD.D2, NFLD.D3, NFLD.D4, NFLD.D8 and NFLD.D9, which bind strongly to Dw4 and Dw14, but not at all to the subty~e of DR4 called DwlO, which give moderate to low reactions with some other DR4 subtypes, and also which react with DRl, DR2, and DRl4 (Dw16); and NFLD.DlO which reacts with the Dw9 subtype of DRl4 as well as binding weakly to some of the DR3-, DR7-, and DR9- typed B cell lines.

4) Brie~ Description of the Drawinqs In the accompanying drawings, Figure l is a histogram which shows the reactions in CELISA of the antibody NFLD.DlO. Each bar represents the reaction against a particular transfectant line, whosa specificities are shown at the bottom of the figure. The heights of the bars represent the adjusted optical densities, which have (a) had the background subtracted and then (b) been converted into a percentage figure with reference to a positive control antibody, in this ca~e L243, a~ antibody-reactive with all DR molecules;
Figure 2 is a histogram which shows the reaction in CELISA of the antibody NFLD.D7; and Figure 3 is a diagram which shows the di~fering specificities of this series of antibodies. Each bar represents the reactions of one antibody, which identity is given on the vertical axis. On the horizontal axis the relevant DR subtypes are each assigned one interval;
below these are given the DR grouping in which the subtypes are contained; thus DR4 contains Dw4, Dw14, DwlO, Dwl3, and Kt. Filled parts of the bars indicate strong reactions of an antibody with the DR or Dw type shown; latch~ed ~ndicates smaller but still significant reactions. White indicates negative reactions.

:

27 :
5~ Descrl~tion of Preferred Embodiments The following describe various procedures according to the present invention.
Transfectants as Immunoqens for Producinq Monoclonal Antibodies Transfectants:
The DR4 transfectants used for the immunizations and ~.
most of the analysis are now included in the transfectants distributed by the organizers of the 11th IHW (see Table l below).

TABLE 1 HLA-DR Expressing Transfectants Obtained fr~m the 11th International Histocompatibility ~ :
Workshop and Used for Specificity Analysis.
: .. - .
.: . ..
11th IHW HLA
Number Specificity Local Name Contributor ' .:
8103 DRl DAP3 DRl~ R. Sekaly/E. Long 8104 DR Bon DR BON P. Claude~C. Thomaaen B107 DR2aDw2 DAP-3DR2a* D.Jaraguemad/E. Long 8109 DR2bDw2 DAP-3DR2b~ D. Jaraquemad/E. Long B110 DR2aDwl2 L-DR2-Dw12 T. Saaazuki 8111 DR2bDw12 LARBl H. Inoko 8112 DR3 L168.2~ R. Karr/J. Silver 8115 DR4Dw4 DAP-3DR4~ R.Sekaly/E.Long/R.Karr 8116 DR4Dw10 L164.11~ R. Karr/J. Silver 8118 DR4Dw14 L165.6~ R. Karr/J. Silver 8122 DR4DwKT2 B18 R. Lechler ' B123 DR4DwrAS L89.2~ R. Xarr/J. Silver B124 DR4Dw~AS Bl9 R. Lechler B125 DRwllDw5 L91.7*~ R. Karr/J. Silver B126 DRw14Dw9 L167.2~' R. ~arr/J. Silver B127 DR14Dw16 LlB2.1~ R. Karr/J. Silver B131 DRw10 29ØC.27 H. Peter B132 DR52aDw24 LR6.2/52a B. Mach B134 DRw52bDw25 DAP-3DR52b~ P.. Sekaly/E. Long B13B DRw53 ~DR4Dw15) L17.B~ R. Karr/J. Silver WO92/lS698 PCT/CA92/00105 2~
The transfectants as described in the Table 1 above include L89.2. (Dw13); L164.11 (Dwlo); L165.6 (Dw14); Dw4 transfectant (DAP3DR4); two other transfectants, L24306 (DR4Dw4) and L259.1 (DR4Dw13). All transfectants were grown in Dulbecco's modified Eagles medium (DMEM) containing 10% fetal bovine serum (FBS~, 5 x 103 mM 2-mercaptoethanol, penicillin and streptomycin (Flow Laboratories). The cells were grown on either 10 cm dishes (FALCONTM) or 75 cm flasks (LINBROTM) and were harvested in log phase using trypsin (Flow Laboratories) and left in standard type bacteriological petri dishes for one to three days. Expression was assayed by CELISA
or FACS analysis using ~he moabs Tu39 or GSP4.1 prior to immunization.
Immunizations: , Cells expressing high levels of HLA were washed three times with phosphate-buffered saline (PBS) and in all immunization procedures 1 x 107 cells were injected.
Essentially three strategies were tried (as shown below in Tables 2 - 4).

S96~ ''' TA~LE 2 Summary of Data o~l Fusions Derived from C3H .-~`
Mice given Indiscriminate Immunizations With HLA-Class II Transfectants i,' ,. ;.
~ ~,,,.,",.

::.
IMMW~L~ATIo~ SCHF~LE
~wks) zations ~ Hybri~s~
: ~008t splenocyte9) h Ab ~d ific ~.
__ . . _ .. _ ... . :
10 1~ DRl IP DRl-IP 75 R5 M 12 2~ DRl IP 5 week8 14 3~ DRl IP after last boost ( 214) 1 DR
monomorph 2 1~ DR4~CFA

3 o R12 M 16 2~ Dw4 IP 4 week~ 1 DR
32 3 Dw4 IP last boost 1478) monomorph polymorphs :
35 R13 F 28 2~ DwlO IP DwlO-IV 1015 2 DR
monomorphs - after (423) last boost 1~ DW14 IP Dw14-IV 204 1 DR
R14F 28 2~ Dw14 IP 3 wk~ aft~r monomorph last boosS

R15F 2'3 2~ Dw14 IP 3 wks after ~230) NONE
' ~ Total numb~r of hybrid from the fusion.
~ Number of hybrids per 101 spleen cell~.

wo 92/1~698 PCJ/CA92/00105 '21~S~9 TAsLE 3 Res~lts of Fusions Done Using Neonatally Tolerized C3H Mice and a Control Group IMMUNIZAI~ON~ LkS
Fusion Sex Age Toleri- Immuni- ~oost ~ Hybri~ds~ Selected zat~on zation Iper 10 Hybri'ds splenocytes)~
R6 M <24hrs non-2 wks Dw4 IP

6 wks Dw4IP 412 polymorph ~ CFA (25a) 1 DR
14 wks Dw4 IP polymorph Same as Dw14 R7 M 8 wks for R6 . ~ CFA Dw4 IS 273 2 weak 14 wks Dw14 IP ~273) . _ _ NONE 1' Dw4 R8 16 k ~CFA Dw4 IP 758 1 monomorph w 8 2 Dw4 IP ~ 20 wks ~469~1 weak polymorph NONE 1~ Dw4 R9 M 8 wks ~ CFA Dw4'IP 750 1 monomorph 1 wks 2 Dw4 IV Q 20 wks ~444~1 weak polymorph - 35 R10 M 24 hrs Dw10 Dw13 - IP ~ CFA Dw13 - 7 16 wks ~ 28 wks 1250 - 20 wks Dw13 IP . ~28a~
NONE
R11 M 16 wks Dw13 Dw 13 boost 600 1 anti-~ CFA ~ 2a wks DR4 20 wks Dw13 IP ~462)3 mono-morphs Total number-of hybrid from the fusion.
~ Number of hybrids per 10; spleen cells. ~ ' 5 0 ;

W092/15698 PCT/CA92tO0105 9 ~ ~ :
3:L
TAB~E ~ Results of Fusio3ls Done to Compare I3nmunization -.:

, ' . ' Po3t- IMMUNIZATION l! Hybrid~ Number of (per 103 ~ybrid~ Fusion splenocytes3~Selected Serum Titer Time Fusion Primary Interval Boost (Weeks) 1128(434) 7 <1/100 336 (480~ 2 1/1600 R16 IP - 4 IS238(170~ 5 1/400 R27 65 467 ~275~ 3 1/800 R29 7 X 542~3403 X = 4.3 R18 IP 4 IP864(455~ 0 1/800 R26 6 546~455~ 1 1/800 X s 483~317~ X 0.

R25 S/C 6 75B (237) 1 >1/1600 R30 7 38 (23) 3 >1~1600 409(340) ` 4 1/1600 592 (423) 4 1~3200 X = 449(-258) X = 3 R19 4 812(427) 18 1/3200 R21 CFA 5 IS672(354~ 0 >1/3200 R24 S~C 6 560~467~ 9 >1~800 R28 7 7041440) 7 >1/32Q0 X = 68~(422~ X = 11 -Total number of hybrid from the fu~ion.
Number of hybrid3 per 103 spleen cell~.
:~
IP = Intraperitoneal. CFA = Complete Freunds Adjuvent S~C = Subcutaneou~. IS = Intranpleinic.
X Average.

W092/15698 PCT~CA92/OOloS

2~96~

The first approach consisted of indiscriminate standard-type immunizations where young adult C3H mice were immunized twice intraperitoneally ~IP~ followed by a final boost intravenously (IV) or IP three days prior to fusion. In the second approach, neonatal tolerization was attempted. Finally, the third group consisted of 10 wk old femal mice which were primed either IP in saline - or subcutaneously in complete Freund's adjuvant ~CFA), left for 4 to 6 weeks and then boosted either IP or intraspenically (IS) three days prior to fusion. In several of the experiments serum was collected before and/or at the time of fusion.
Fuslon:
All fusions were done three days after the final boost and were carried out using the fusion partner SP2/0-Ag~4, (see Shulman M, Wilde CD & Kohler G. "A
better cell line for making hybridomas secreting specific antibodies". Nature 1978: 276: 269.). Hybridization was done according to a previously described method, (see 20 Drover S, Marshall WH & Younghusband HB. "A mouse `

W092/15698 2 ~ ~ 5 9 6 gPCT/CA92/00105 monoclonal antibody with HLA-DR4 associated specificity''O
Tissue Antigens 1985: 26: 340-343.). Cells were seeded at various concentrations and using different feeder cells in order to establish optimal conditions.
Selection ~as done using standard HAT (hypoxanthine and thymidine reagents) in DMEM containing 20% FBS and ;
supplements as described above.
Screeninq & S~ecif icity Testlng:
All supernatants were tested approximately lO days after fusion using CELISA as previously described (see Morris RE, Thompson PT & Hong R. "Cellular enzyme-linked immunospecific assay (CELISA) I. A new micromethod that detects antibodies to cell surface antigens". Hum Immunol 1982: 5: 1-19; and Drover S & Marshall WH.
"Glutaraldehyde fixation of target cells to plastic for ELISA assays of monoclonal anti-HLA antibodies produces artefacts". J Immunol M~thods 19~6: 90: 275-2~1.).
For the first screen, the supernatants were tested against the immunizing cells and all positive were differentially screened on the following day against the immunizing cell and non-transfected L cells. Those that were positive only with the immunizing cells were selected for further testing against a small panel of transfected cells, including those expressing DP, DQ and Z5 informative DR. Hybridomas were then selected for ---cloning and further analyzed on both transfectants and B
cell lines. ~ -Use of Human B Cell Lines as Immunoqens The production of NFLD.D12 and NFLD.D13:
This was achieved in a fusion made from Balb/c -spleen cells. The mice had been primed with affinity purified HLA molecules extracted from a lysate of the B
cell line SAVC' (lOth Workshop ~9034), using beads (DYNAL~M) that had been coated with two antibodies: first, anti-mouse IgG had been coated by the manufacturer;
secondly the beads wer~ coated with a mouse IgG1 WO92/lS698 PCT/CA92/00105 ,9~5 34 monoclonal antibody made in t:his laboratory (NFLD.M67) that detects a monomorphic determinant on HLA-DP
molecules. Beads loaded with two antibodies and class II
molecules absorbed onto them from the lysate were injected subcutaneously in CFA in four sites. The boost injection given three days bef~re splenectomy and fusion consisted of lo million SAVC cells injected intravenously.
Transfectant Cell Lines The lines used for these experiments are listed below in Table 5.

~A~LE 5. ~LA-DP Transfectant Cell Lines 11th w/S Number Other N~me DP Type Provided bY:
8301 L3.6.2 DPal~0201 J. G. 3Odmer 8302 DAP-3 DPw2 DP31'0201 R. SekalyiE~. Long 8303 L11.3 DP31'0401 J. G. aodmer - 8304 LPP 3-6 DPw4 H. Inoko 8305 L25.4 DP31~0402 R. Karr 8306 LAP 4108-6 Cp H. Inoko ~DPal~osol) IDPAl'0201) : . ., ::
.
~';' ~' : , . ..
' ~ ~ .

-W O 92~15698 2 ~ ~/CA92/00105 flasks (LINBROTM) and were harvested when in log phase using trypsin. Following harvest, they were cultured in standard bacteriological PETRI dishes (to which they do not stick) for one to three days. Culture medium consisted of Dulbecco's modified Eagles medium (DMEM) containing 10% fetal bovine ~erum (FBS), 5 x 10-5 mM 2-mercaptoethanol, penicillin and streptomycin. Expression of HLA-DP was assessed by flow cytometry or by OELISA
assay, using either the monoclonal antibody B7/21 or one of our own antibodies against monomorphic DP
specificities.
Immunization Protocols varied from experiment to expériment, but a typical protocol is as follows: 107 transfectant cells were injected subcutaneously, dividing the dose between four sites on the back, together with Freunds complete adjuvant, 0.1 ml per site. After a wait of 4-8 weeks, the mice were boosted by 107 cells, either gi~en introperitonealy or intrasplenically, and the spleen removed three days later. In some experiments, a primary immunization with human B cell line cells or with DP
transfectants was made by intravenous injection; three days later the spleen was removed and a fusion performed.
Fusions Fusions were performed three days after the last injection of antigen and were carried out with the fusion partner SP2/0-Agl4 (Shulman M, Wilde CD, Kohler G. Nature 1978: 276: 289.). Fusions in the presence of polyethylene glycol were done according to a standard method, (Drover S, Marshall WH, Youghusband HB. Tissue Antigens 1985: 26: 340.). Usually 2 x 105 cells per well were plated in a-96 well plate with flat bottomed wells.
.

.,: : . ~ . .. .. . .
- : - .. . .. ~ . .. . . ~

WO92/15698 PCT/CA92/0010~
~ ~5~9 Screeninq and Specificity Testinq All supernatants were tested approximately ten days after fusion, using the CE`LISA assay, (Morris RE, Thompson PT, Hong R. Hum Immunol 1982: 5: 1; Drover S, Marshall WH. J Immunol Met l9El6: 90: 275.).
In the first screen, the :immunizing cell was used as a target. In a second screen of positive derived from the first screen, differential testing was done on the transfectant that had been used as immunogen and on L
cells. In the case of experiments where human B cell lines were the immunogen, the second screen was done on several B cell lines plus a human T cell line that fails to express class II HLA molecules (Molt/4).
Later studies of specificity were done using human B cell lines in CELISA assays. The majority were the homozygous cell lines collected during the 10th IHW, (see -~
Yang SY, Milford E, Hammerling ~, Dupont B. In "Immunobiology of HLA, Vol. 1, Histocompatibility Testing 1987" B Dupont Editor 1989: p. 11.) and these were `
supplemented by a few others obtained from other laboratories ~here DNA sequencing had been done on them.

6! _ Operation of Preferred Embodiments Results In some of the earlier fusions, hybrids were lost through what appeared to be overcrowding and lysis due to cytotoxic T cells. This problem was partially alleviated by plating the fused cells at a maximum densi`ty of 2 x 105 cells per wall and eliminating spleen cells or thymocytes as feeder cells.
In the first set of experiments, summarized in Table 6, five fusions were derived from mice that were randumly immunized. Over 2500 hybrids were assayed on the immunizing cells; after differentially screening the positives, six were fuxther analyzed for polymorphic activity. As can be seen from the data, no hybrids made WO92/1~698 PCT/CA92/00105 antibody with a short specificity to polymorphic epitopes; only two antibodies reacted to a polymorphism and 5 hybrids produced antibody to class II monomorphic determinants.
An attempt was made at neonatally tolerizing C3H
mice by injecting them with non-DR4 transfeatants at various times from age 24 hours to 6 weeks. Prior to immuni~.ation, serum samples were obtained from these mice as well as from non-tolerized litter mates. The sera were titered in CELISA on the tolerizing cells and on non-transfected L cells. The CELISA data showed evidence of antibody activity to the tolerizing cells, indicating that tolerance to DR had not been achieved. It was decided to use some of the mice for fusions and, at age 8 to 16 weeks, they and some of their non-tolerized litter mates were immunized with DR4-expressing transfectants. It is obvious from the data shown~in Table 6 that very few hybrids were selected for further studies despite screening over 4000 hybrids. After limited specificity analysis, it appeared that one of the hybrids was making an antibody specific for DR4. This antibody derived from the Rll fusion and now designated NFLD.D1, is described in more detail below.
Fifteen fusions done to compare immunization schedules were very productive. The data summarized in Table 7 clearly show that the worst immunization strategy was an IP primary followed by an IP boost. The best strategy was immunizing subcutaneously along with CFA
followed by an IP boost, while the other two strategies immunizing IP in saline or subcutaneously with CFA
followed by an IS boost were equally good. Titering of sera was not always predictive of the number of specific hybrids. For example, the mouse (fusion R21) that produced the highest titer against the immunizing cells, although producing a large number of hybrids, did not yield a single one making antibody specific for the , .. , ... . . ~ , .. ~ . . - . ~ . .

WO92/1~698 PCT/CA92/00l0~
2~ ~59~9 immunizing cell. On the other hand the mouse (fusion R16) which had the lowest titer produced 7 hybrids making specific antibody.
From the above experiments in which a total of fifteen fusions were done, yielding over 8000 hybrids, sixty-four hybrids were selected for further analysis. -Thirteen produced antibody to polymorphic determinants;
eight of these are described below. `
".

NFLD.D Monoclonal Antibodies The monoclonal antibodies derived from mice immunized with transfectants were all isotyped as IgG1.
This is a non-complement fixing subclass so all specificity analysis has been done using CELISA.
Homozygous B cell lines from the 10th IHW were used for specificity analysis on eight monoclonal antibodies produced from mice immunized with DR4-expressing 20 transfectants. These tests were done using optimally- -diluted supernatants from cloned hybridomas. In addition two monoclonal antibodies from uncloned hybridomas resulting from mice immunized with human B cell lines were also studied (see below). The antibodies were also 25 tested on a panel of L-cell transfectants expressi~g ~ -various DR4 and non-DR4 molecules. A summary of the data presented below in Tables 6 and 7 and Figure 3 shows there are six different patterns of reactivity.
Preliminary- testing shows that the antibodies are reactive with B lymphocytes from some individuals and a full analysis will be done when the subtyping-of the DR
specificities is available.

.

WO92/15698 PCT/CA92/0010~
2 ~ 0 ~

TA~LE 6 Reactivity of NFLD.D Monoclonal An~ibodies Produced to HLA-DR4 with Human B Cell Lines.

HLA-CLASS II~ REACTIVI~Y FOR NFLD-~
Cell Line~ DR Dw DQ DP Dl D2 D3 D4 D7 D8 D9 D10 D12 D14 9034 4,53 4 8 10 8 10 6 3 a 6 8 10 10 10 0 9029 4,53 4 8 2.1 10 10 6 3 8 8 10 10 10 10 9031 4,53 4 a 4.1 10 10 6 8 6 8 10 10 10 10 -9025 4,53 4 7 4 10 10 6 6 6 0 10 10 10 10 9027 4,53 4 7 4 10 10 6 6 4 0 10 10 10 8 9028 4,53 14 ~ 4 10 10 6 8 6 0 10 10 ~ 4 S098 4,53 14 8 ? 8 10 6 8 6 6 8 10 4 2 LS40 4,53 14 3 3.6 10 1 1 1 4 1 1 1 1 2 9026 4,53 10 8 4.1 10 1 1 1 6 0 1 1 2 4 TS10 4,53 10 3 ? 8 4 4 6 4 6 6 8 9030 4,53 13 7 3 8 6 4 6 6 6 6 8 0 0 9024 4,53 XI 8 5 1 4 4 6 1 6 8 10 9002 1 20 5 4.1 1 4 4 4 1 6 6 8 2 0 9014 15 2 6 4.1 1 6 8 6 8 4 6 8 9010 15 2 6 4.2 0 2 0 4 6 1 2 6 0 0 9013 15 2 6 4.2 1 6 6 6 6 6 6 10 4 9011 15 12 6 ? 1 8 6 6 6 6 6 10 0 0 9009 16 21 1 4.1,14 1 6 6 6 6 6 8 10 0 0 9016 16 22 7 4.2 1 1 1 1 2 9023 17,52 3,24 2 1 1 1 1 1 1 1 1 2 9022 17,52 3,24 2 3 1 1 1 1 2 1 1 4 4 2 9018 17,52 3,25 2 3 1 1 1 1 1 1 1 1 1 2 9019 17,52 3,25 2 2.2 9021 18.52 NE~,24 4 3 9037 11,52 5.25 7 4.2 9043 11,52 5,25 7 10 1 1 1 1 1 0 3 9032 11,52 DB2 25 5 2 1 1 1 1 1 1 1 1 0 9038 12,52 Da6 25 7 2.1 1 I 1 1 l l 1 1 o 9060 13,52 18,25 5 19 1 1 1 1 1 1 1 1 0 0 9059 13,52 19,2~ 6 3 1 1 1 1 1 1 1 1 1 4 9055 13,52 19,26 6 5 9063 13,52 19,26 6 16 35 9056 13;14,52 9/19,25 5 2.1,13 1 1 1 1 1 0 1 8 1 2 9054 14,52 9,25 5 4.2 0 1 0 1 1 1 1 2 0 0 9057 14,52 9,25 1 4 1 1 1 1 1 1 1 6 0 0 9064 14,52 16,24 7 13 1 8 6 8 1 6 8 10 9052 7,53 11 9 4.1 0 1 0 1 1 1 ~ 1 0 0 9049 7,53 17 2 1 1 1 1 1 1 1 1 2 2 9047 7,53 DB1 2 17 l 1 1 1 1 0 1 2 1 0 4 0 9096 7,53 Dal 2 15 1 1 1 1 1 1 1 6 9069 8,52 8.1 4 4.1 0 1 0 1 1 1 1 1 0 0 9071 8,52 8.2 4 3 1 1 1 1 1 0 1 1 1 4 r-9066 8,52 8.3 6 3 9074 9,53 23 9 ? 0 1-- 0 1 1 1 1 1 0 0 9075 9,53 23 9 4.1 1 1 1 1 1 1 1 2 0 0 45 9076 9-,53 ? 3 1 1 1 1 1 1 1 4 Refers to the 10th International Histocompatibllity numbers designated forlthe homozygous B cell lines Ireference 22).
~ The HLA cla~s II types and splits were obtained from references 23 to 26.
The CELISA data converted to conventional serology scores: 1 0-10% binding: 2, 11-20%; 4, 21-40%; 6, 41-80% 8, 81-100%, 10, > 100%; 0, not done.

. ' ' . ' : ,:
.

S 9 ~ 9 ~ABLE 7 Reactivity' of NE'LD Monoclonal Anti~odies Produced Ayainst HLA-DR4 Using HLA-DR
Expressing Transfe~tants.
. :.
Trans- NF1D MONOCLONAL ANTI~ODIES
DR Gene ExpressedD1 D2 D3 D4 D7 DB D9 10 - -- - . . __ ,:
Dap3 DR4118~ 45 38 7D 93 121 92 125 DR4 ~w4) (.90)~1) (1) ~1) ~1) tl~
L243.6128 58 34 71 96 12a 94 142 DR4 ~w4~ ~.94) ~1.28~ ~.89~ ~l.Q1~ ~1.03~ ~1.06) ~1.02) ~1/14) L165.6111 51 40 7292 112 98 117 .
DR4 ~w14) ~.84) ~1.13) ~1.05) ~1.03) ~.99) ~.93) ll.06)~.94) L259.1131 6 10 40100 73 47 141 DR4 ~w13) ~1.0) ~.13) ~.26) ~.53) ~1.08~ l.60) ~.51)~1.13) L164.11 118 0 0 2 25 0 0 0 DR4 ~w10) ~.90~ ~.03~~.27) L167.20 0 13 L182.1O 4 21 49 0118 73 192 DR14 ~w16) ~.09)~.55)~.70) ~.97) ~.79~1.54~
Dap3DR1 0 27 19 48 0 87 57 132 DRl ~wl~ 1.60~~.50~~.69~ ~.72l ~.62~1.06~
Dap3 DR2a 0 7 44 28107 68 3a 14 DR2 ~w2) t.l6~ 11.16~ t.40) ~1.15) ~.56~ ~.41~
Dap3 DR2b 0 0 0 0 0 0 0 0 .
' Reactivity OD value of test calculated as a percentage of the pos1t ve control.
The numbers in brackets refers to the ratio of ~ ~eactivity for cach cell divided by the ~ reactivity for the immunizing substype DW13 for NFLD.Dl and DW4 for all the other ~oabs.

, :: ~ .

~ ~. .....

. - . ,. , . . .. : - . , - ~ . . . . . .. . -; , . .. .. . , .: , .. .... .

: :

WO92/l~S9~ PCT/CA92/00105 ~l 21~59~9 NFLD.D1:
This moab was d~rived from a mouse (R11) immunized with DR4-Dw13 expressing transfectants as shown in Table 3. It appears completely monospecific for the DR4 specificity since it reacts with all the subtypes, although Dw15 has not so far been tested (Table 6). This specificity was confirmed ~y testing on a small panel of transfectants as is shown in Table 7. In addition, testing supernatant from the uncloned hybrid against additional transfectants provided by the 11th I~ (data not shown) revealed no extra reactivity.
NFLD.D2. NFLD.D3, NFLD.D4. NFLD.D8, and NFLD.D9:
All exsept NFLD.D3 were obtained from different microculture plates of the same fusion (R19, see Table 4) and are believed to be derived from different clones although their specificities are similar. The specificity of NFLD.D3 which was derived from a completely different fusion (R17, Table 4) is remarkably similar. It is apparent from the data presented in ~able `
that all five moabs react most strongly with DR4 subtypes Dw4 and Dw14, to a lesser extent with Dw13 and KT, and not at all with DwlO. In addition they also react with DR1, DR14 (Dw16) and DR2 (all` its subtypes) but with no other DR molecules that were tested. The pattern of reactivity for the cell lines has been confirmed by testing transfectants (Table 6) with a few exceptions. The weaker pattern for NFLD.D2 with the DR4 (Dw13), DRl, DR2 and DR14 (Dw16) compared to the DR4 (Dw4 and w14) homozygous, typing cells is striking on the transfectant cells, particularly for those expressing Dw16 and DR2. NFLD.D3 also gives reduced binding to the DR1 and Dw16 transfectants, but reacts quite well with the DR2A trans'fectant, suggesting that these two antibodies are different. The other three are extremely similar and may have been derived from one clonally expanded B celi in the spleen of the R19 mouse.

2~9~ 42 NFLD.D10:
This antibody, derived from a different fusion, R23 (Table 4), has a similar activity to those derived from R19 but reacts more strongly with DW13, DRl, DR2 and DR14 (w16). Unlike the preceding antibodies, it also reacts with the DR14 subtype Dw9 and weakly with some DR17, DR7 and DR9 cells. From data obtained by testing undiluted supernatant on the 11th IXW transfectants (Figure 1), this pattern ~as essentially confirmed. However, the weak reactivity observed with DR3 was not apparent when the appropriate DR3 transfectant was tested. No DR7 transfectant with good expression was available for testing. In addition the antibody reacted weakly with the DRlO transfectant.
NFLD.D7:
It is obvious from the data presented in Tables 6 and 7 that this moab has a different pattern of reactivity than all of the above. In addition to reacting with all the DR4 subtypes, it reacts strongly with all the DR2 subtypes but not with DR1. It also reacts weakly with some DR3 lines. From the CELISA data on the 11th IHW transfectants (Figure 2) in which undiluted supernatant was used it appears to crossreact with DR52 specificity.
NFLD.D12 and NFLD.D13:
These two antibodies, obtained from the same mouse, were selected out in preliminary screening as they were reactive with the SAVC cell line that had been used for the immunizations but negative with a contrasting human B cell line H0301, (lOth Workshop number is 9055). Both were isotyped as IgM. They were then tested on a minimum of 33 B cell lines, most beiny from the well characterized col'lection of homozygous cell lines studied in the 10th International HLA Workshop. The CELISA
results are summarized in Table 4.

.. , . . ,, .,, , . ., - ., ~.. .. .. . , ...... . :

W092/15698 PCT/CA92/OOlOS
21~9~9 It can be seen that NFLD.D12 give positive results with five cell lines that express the Dw4 subtype of DR4.
Reactions with other types are small except for modest binding to the Dw12 variant of DR15. In the case of NFLD.Dl3, Dw4 is again a target molecule but with this antibody there are significant reactions also with one of the four lines expressing the Dw14 subtype of DR4. Minor reactions are also noted with the remaining two members of the Dw14 subtypes as well as with the DwlO subtype.
Reactions with the remaining cells in the panel were mostly either negative or trivial.
Some 76 fusions have been performed using a variety of experimental designs and with C3H mice, Balbtc mice and C3H x Balb/c F1 hybrids. Mice have been immunized with and without adjuvant, by various routes (subcutaneous, intraperitoneal, intravenous, intrasplenic) and with varying antigen doses. The antigenic material has mostly been in the form of L-cell transfectants expressing HhA-DP molecules, but some immunizations have been done with EBV-transformed human B cell lines; a few immunizations have been done with affinity-purified HLA-DP molecules and with synthetic peptides designed to reproduce small polymorphic parts of the HLA-DP molecules. About the only clear conclusion reached in terms of immunization protocols is th~t immunization with peptides has been a fruitless procedure.
The antibodies described here and summarized in Tables 8a and 8b, represent those selected after exhaustive screening and testing of many thousands of hybridomas, typically on the order of 1000 per fusion (one fusion means one mouse spleen).

:, , . : ~ - , . , WO 92/15698 PCr/CA92/00105 2i~9~ 44 -U~ _ = _ . :, C ~ U ~ ~ U
:1~ I -~ _C ~- O _C ,~ ~ ~ X '3 CO ~

3 ~ -- ~ -- ~

~ ~ ._.
U~ U) >~ C~l (::1 _ G _ ~ _ _ ~J : .
OC~ ~1 . _ : -' ~ C ~ O O ~

E~ u~ o ~ ~o ~D 0 '1:) ~D ~ .
E 1~ J o 1~ o z _ .

''~

S~5BSTITUTE SHEET: `

WO 92/15698 PCr/CA92tO0105 9 ~ 9 i~ ~

b - L~l `

~ L~ ~- ~ ~ . h. I

SUB~ITUTE SHEE~
~ .. .. .. .. .....

WO 92/l~i698 PCr/CA92/0010~ ~ .
2~0ri969 -` .
. . 45 ~ ~ r~

O ~ X ~ O ~ ~ a H ~ C O ~ ._ ~ _ O C C X ~ .

_ n _ v _ n ) ~ Ln _ N _ C . ~ _ . I ~ _ C ~ _ ~ QJ ._ O ~ . : ' .

~c- -~u ~ ,~d ,_ .~

u ~ v. . _ 9 L N . ' o ~ c c c o c ~ ,~, 3 N
O O O O O--~O = ~ - O C~ -_ v~t~l v~ ~ r3 O o 3 ~ c~l C ;
o c o al c~ c ~ c ^-- z o O C~ .__ _ _ - _ ~7 ~ _ _ o~ cn ~ :., D ~ N ~ ~S o ~ ~ t s ~ ~ : ~

o ~o o o C~ o .
= = L~ ~ ~
. .. ~ .
~ . . .
SU~STilrlJTE SHEET. -:

-WO92/15698 PCT/CA92/0010~

21~969 Antibodies Recoqnizina M onomorphic Determinants Four antibodies recognizi~g monomorphic determinants have been selected for listing, from a larger number. OL
those which appear to recognize a monomorphic de~erminant that is only found on H~-DP molecules, one is IgG1 subclass (NFLD.M67), one ls 1g~2 (~FLD.M68) and a third is IgM (NFLD.m6O), thus ma~ing an interesting set for use in a variety of experiments. The four~h, IgG1 (NFLD.M~s), appears to recognize all D~ and all DR
molecules but is ne~ative Cor those DQ molecules Ihat have so far ~een tested, na~.ely the DQ t~ansfec~ants distributed to participantc OI the 11th International U~st~Co~D2t~ t`~ W~r'cc~.o~ !T~W! -Antibodies Recoqnizinq Polvmorphic Determinants One antibody, N~LD.M5~, was produced in an earlyexperiment. This antibody showed a striking resemblance to two other published antibodies. In our laboratory the same specificity has been found again, either with a , SUE~ST~TUTE SHEET

WO92/1~698 PCT/CA9~/00105 , .

210~969 46 strong resemblance (NFLD.M60 and NFLD.M64) or with relatively minor variations (NFLD.M73, NFLD.M74 and NFLD.M75). NFLD.M58 is considered to be recognizing an epitope requiring the amino acid sequence DE at positions 55 and 56 on the HLA-DP beta c:hain. Tests to prove this conclusion are planned, using mutated beta chain molecules.
Anothar specificity found in our collection of antibodies is shown by NFLD.M63. It is sufficiently different from M58 to warrant description separately. It binds best to HLA-DPw3 molecules as well as to HLA-DRwll molecules (including a DRwll transfectant). There are weaker reactions with other DP molecules, as shown in the :~;
table, which may be possibly lost if the antibody concentration is reduced by dilution.
A different specificity is displaced by NFLD.M66.
This antibody reacts with virtually all HLA-DP nolecules except for the DPw2 and DPw4 molecules (coded by DPBlI0201, DPB1*0202, DPB1*0401 and DPBl*0402 genes).
This pattern suggested that it might be recognizing a polymorphism due to the amino acid sequence DEAV at positions 83-86 on the beta chain. Further testing supported this hypothesis because it failed also to recognize molecules coded by the DPB1*1501 gen which carries VGPM at these positions (DPw2 and DPw4 molecules are GGPM at that location. This hypothesis is currently being tested using beta chains deliberately mutated at these four positions. -~
A further specificity is showed by NFLD.M70. This cytotoxic antibody has been tested on most of the homozygous celi lines from the loth IHW and only -recognizes tho~e which carry the less common alpha chain, coded by the DPAl*0201 gene. This chain, which we can call the alpha 2 chain for short, is found almost exclusively with beta chains coding for DPwl and DPw5, as well as with molecules (not yet officially named by WHO

WO92/1~698 PCT/CA92/00105 47 2i~ 969 nomenclature committee) coded by DPBl*0901, DPBl*1001 and DPBl*l901. That NFLD.M70 is truly dependent on the presence of the alpha 2 chain is shown by its failure to bind to transfectant molecules containing the alpha 1 chain in combination with a DPB1*0901 coded chain (marshall et al., unpublished observation); a similar transfectant made with the alpha ~ chain was fully reactive.
A fourth antibody, different from all the others is NFLD.M770 This antibody binds to cells that Pxpress DP
molecules containing the amino acid sequence "QL" at positions 10 and 11 on the beta chain. The single exception tG this is that cells expressing the DPB1*1301 gene are not recognized by this antibody. Since there are no available examples of homozygous cell lines expressing DPB1.1101, the antibody has not been evaluated for its reaction to the product of DP81*1101; according to the present interpretation it should bind, unless the DP molecular structure is influenced by polymorphic sequences in the adjacent chain, which may be the case for the non-binding DPBl*1301 product. In preliminary studies the NFLD.M77 antibody binds also, as predicted, to transfectant cells expressing the DPB1*0901 gene.
These results make it clear (Figure 3) that there are six main patterns of reactivity. The first two patterns show that two monoclonal antibodies (NFLD.D12 &
13) have the potential to differentiate between Dw4 &
Dw14 at least on B cell lines since NFLD.D12 only reacts with Dw4 cells whereas NFLD.D13 reacts with Dw4 and some but not all Dw14 cells. In the case of the second antibody Dw14 specificity is not totally clear since we do not know whether the Dw14 cell lines are 14.1 or 14.2.
Both variants differ from Dw4 by an arginine substitution for lysine at position 71 and they differ from each other by a glycine for valine substitution at position 86. Dw4 has glycine at this position. It is also interesting to ,: . , , , . . . : .: ::, - :, : - , , - ., : , :.

WO92/1~698 PCT/CA92/00l05 note that neither of these antibodies reacted with DR1 (Dw20) cell line, (9002) which is identical to Dw14.2 in the third hypervariable region, nor with D~l (Dwl) cell line, (9003) which is identical to Dw14.1 in the same region. The DR14 (Dw16) cell line (9064) which is identical to DR1 (Dwl) and ~)w14.1 in this region was -clearly negative. This suggests that if indeed the NFLD.D14 antibody can differentiate between Dw14.1 and Dw14.2, the apitope must also be influenced by amino acids outside the third hypervariable region; otherwise one would expect it to react with either DR1 or Dw16 in the same way as does the CCCL20 moab (Dejelo CL, Braun WE, Zachary AA, Teresi, GA, Smerglia AR & Clark LV. Hum Immunol 1986: 17: 135-136.) which reacts mainly with Dw14, Dw4, DR1 and DR14 (Dw16). The epitope for this moab has recently been mapped to positions 67, 70 and 71 and its activity does not appear to be influenced by -residues in the rest of the molecule (Hiraiwa A, Yamanaka K, Kwok WW, Mickelson EM, Masewicz S, Hansen JA, Radka SF
& Nepom GT. "Structural requirements for recognition of the HLA-DW14 class II epitope: a key HLA determinant associated with rheumatoid arthritis". Proc Natl Acad Sci USA 1990: 87: 8051-8055.). To our kno~ledge, this is the first example of monoclonal antibodies with the ability to discriminate between two conservative amino acid substitutions in a manner similar to T cell ~-recognition.
An interesting and unexplained finding in the case of NFLD.D12 and NFLD.D13, is that these-two antibodies, used at the same concentration as for B cells lines, faiIed to bind significantly to L cell transfectants expressing Dw4 or other DR4 subtypes. It is speculated that this could b~ due either to a low molecular density ;
on the L cells which could prevent multiple attachments of the IgM antibody binding sites, or an altered glycosylation pattern coded by the mouse cell, or perhaps - WO92/15698 PCT/CA92~00105 9:
it is due to the presence of a different peptide in the class II groove than is found in human B cell lines.
The moab NFLD.Dl shown as the third pattern recognized all DR4 cells tested and no other cells. It 5 is expected that this moab r~acts with an epitope found ~ -on all DR4 subtypes but not on non-DR4 molecules.
Whether or not the epitope is the same as that recognized by the HLA-DR4 moab, GS359-13FlO, (Alber CA, Watts R, Xlohe EP, Drover S, Marshall WH, Radka SF & Karr W.
"Multiple regions of HLA-DR~l chains determine polymorphic epitopes recognized by monoclonal antibodies". J Immunol 1989: 143: 2248-2254.) will be determined by epitope mapping at a later time.
The reactivity pattern for D7 is considerably more complex (Table ~, Figures 2 and 3~. In addition to reacting with all DR4 cells tested, it also reacts moderately with DR2 cells (all subtypes). At the dilution used for specificity analysis on the cell lines (Table 6) it reacted weakly ox not at all with numerous cells expressing DR52. However, testing on transfectants using undiluted supernatant from an uncloned culture was positive for the two D~52 transfectants tFigure 2).
Since the DRB8 gene, which encodes the DR52 specificity, is constitutively expressed at lower levels than the DRB1 gene, this simply may be a dilution problem. It is also possible that the culture from which D7 was~derived was not clonal. More testing on the transfectants using supernatant from a cloned culture, as well as testing the cell lines with antibody in excess, should clarify this.
The fifth pattern shown in Figure 3 is for five moabs, but the specificities are not quits as simple as portrayed in the figure, due to graduations of reactivity. Thi~ is particularly apparent in the data in Table 6. Both NFLD.D2 and NFLD.D3 are considerably less 35 reactive with the Dw13 expressing transfectant (L259.1) than are NFLD.D4, NFLD.D8, and NFLD.D9 moabs, all of -2105969 .

which show the same degree of reactivity. However, NFLD.D2 differs from the others in that it reacts poorly with the Dw16 transfectant tLl82.1) and with the DR2a transfectant lines.
The final pattern shown in Figure 3, produced by NFLD.Dlo is similar to the pattern produced by NFLD.D4, NFLD.D8, and NFLD.D9 but it binds Dw13 more strongly. In addition it also binds to DR14 (Dwg) molecules and gives weak reactions with some DR3, DR7, and DRs molecules.
When used undiluted on the transfectants, it is also weakly bound to DRwlO (Figure 1).
The fact that antibodies react with DR2 is an interesting feature. The DR2 specificity is even more complex than DR4 in that there are two serologically-defined specificities (DR15 and DR16), each with twosubtypes. Also each DR2 haplotype contains two expressed genes (DRBl and DRB5). Analysis of the sequences and transfectant data (Figure 2) suggest that in the DR15 speci~icity the positive molecules are coded for by DRB1, whereas in the D~16 specificity the positive molecules are coded for by DRB5. Another interesting feature is that all the reactive molecules for pattern 5 have glutamine at position 70 and all except DR2 have lysine or aryinine at position 71, both considered to be the --critical amino acids involved in the susceptibility to RA. However, not all molecules with these residues confer susceptibility to RA. A final interesting point is that NFLD.D10 which also reacts with molecules containing these residues, cross reacts with other molecules containing arginine at position 70, such as DRl4 (Dw~), DRs and DRlo. It has been speculated by Gregerson et al. (see Gregersen PK, Silver J & Winchester RJ. "The shared epitope hypothesis. An approach to understanding the molecular genetics cf susceptibility to rheumatoid arthritis". Arth ~ Rheum 1987: 30 1205-1213.) that this residue in some specificities, e.g. DR53 - . , - ;,, ., ,., , . . ~ ~ .

21 0~969 and DRlo may be conformationally equivalent to the RA-susceptibility epitope on Dw4. It has also recently been reported that DR10 is highly associated with RA in the Spanish population.
The HLA-DP system, discovered in a remarkable series of experiments by Shaw et al. (see Shaw S, Johnson AH, Shearer GM. J Exp Med 1980: ~52: 565.) was revealed by a primed lymphocyte test (PLT) procedure. In a PLT, the polymorphism ~s recognized by T-lymphocytes and not by antibodies. By 1984, PLT had revealed six probable alleles. An uncertainty wlth the DP system was that it might not be accessible to classical serology. However, various observations have contributed to showing that it is accessible. A monoclonal antibody was made by Heyes et al. (see Heyes J, Austin P, Bodmer J, et al. Proc Natl Acad Sci 1986: 83: 3417.) using a transfectant cell line as immunogen, that reac~ed primarily with DPw4 and to a lesser extent with DPw2 cells. Second, Johnson et al.
(see Johnson AH, Thorsby E, Nakatsuji T, et al. Hum Immunol 1986: 17: 21.) found an antiserum that had been raised by planned immunization that reacted with DPwl positive cells once a confounding anti-DR2 antibody had been removed by absorption. Third, Park et al. (Park MS, Tonai R, Terasaki PI, et al. Abstract for ASHI (American Society for Histocompatibility and Immunogenetics meeting in 1986) found, by screening 5000 pregnancy sera, a few which reacted reasonably well with either DPwl, DPw3 and DPw4. So by 1986 it was becoming clear that the DP
system was accessible to serology.
The antibodies provided by the present invention takes this conclusion further. There is now no reason to think there should not be other monoclonal antibodies produced, eithe~ to other polymorphisms that have been revealed by DNA sequencing of DP alleles (see Bodmer JG, 35 Marsh SGE, Albert ED et al. Tissue Antigens 1991: 37:
1.) or to reciprocal epitopes ~rom those for which W092/15698 PCT/CA92/OOlV5 ~ ~ .
2los969 .-antibodies have been producecl e.g. non-DE at position 55-56, GGPM at positions 83-86, or to the common alpha chain (DPAl*O101), or to sequences at positions 10 and 11 other than QL.
Eventually there should be a series of anti-DP
antibodies available that will allow DP typing to be carried out by serological methods in a useful, valid fashion to compare the matching of donor and recipient in critical situations such as in bone marrow transplantation. The monoclonal antibodies may not be good at recognizing alleles, since there is ~o much sharing of polymorphic portions of the molecule between alleles, bllt they should be excellent at detecting epitopes, which after all is what are important in provoking immune responses, either of graft rejection or of sraft versus host disease.
Finally, it should be noted that two of the monoclonal antibodies described here are of the IgGl subclass and cannot be used in complement-dependent cytotoxicity assays. If these can be converted to IgG2 or IgG3 by the heavy chain "switchingl' procedure then all of them may be available for use in routine hospital tissue typing laboratories; those laboratories rely most heavily upon complement dependent cytotoxicity as a tissue typing technique.
Fourteen murine monoclonal antibodies have been provided herein. These were chosen because -of their specificity for HLA-DR4 molecules. There are 6 patterns of reactivity among these ten antibodies (5 are very -30 similar). NFLD.D1 binds to all DR4 molecules, whilst others bind only to subtypes of DR4. The shortest, NFLD.D12, binds only to the Dw4 subtype of DR4; NFLD.D14 binds to Dw4 and~Dwl4 subtypes; NFLD.D7 binds to all DR4 and DR2 molecules but also, less strongly with several non-DR4 molecules. NFLD.D2, D3, D4, D8 & D9 have approximately the same pattern as each other, they all WO92/15698 PCT/CA92/00]05 2~05969 bind strongly to Dw4 and Dw14, but not at all to the subtype of DR4 called DwlO; they give moderate to low reactions with some other DR4 subtypes; they also react with DR1, DR2, and DR14 (Dw16). The final pattern, that of NFLD.DlO resembles the one just described; in addition, it reacts with the Dw9 subtype of DR14 as well as binding weakly to some of the DR3-, DR7-, and DR9-typed B cell lines.
These antibodies, in combination, can be used to detect subtypes of DR4 and will dc this more quickly and simply than is possible with T cell technology. In addition, they have relevance to studies of rheumatoid arthritis, particular those which react with putative RA-susceptibility determinants.
As used in the present invention,"cell line" refers to various embodiments including, but not limited to individual cells, harvested cells and cultures containing cells so long as these are derived from cells of the cell line referred to. By "derived" is meant progeny or issue. It is, further, known in the art that spontaneous or induced changes can take place in karyotype during storage or transfer. Therefore, cells derived from the cell line referred to may not be precisely identical to the ancestral cells or cultures, and any cell line referred to includes such variants.

WO92~15698 PCT/CA92/00105 -.~
i21 0~969 REFERENc2s 1. Bodmer JG, Marsh SGE, Parham P, et al Nomenclature For Factors of the HLA system", 1989 Tissue Antigens 1990: 35: 1-8.
:. .
2. Stastny P. "Mixed Lymphocyte Cultures in Rheumatoid Arthrltls". J Clin Invest 1976: 57: 1148-1157.

3. Nepom GT, Seyfried CE, ~olbeck SL, Wilske KR ~ Nepom BS. "Identification of HLA-Dw14 Genes in DR4+
Rheumatoid Arthritis". Lancet 1987: I: 1002-1004.
. ~
4. Karr RW, Rodey GE, Lee T, & Schwartz BD. ~. ;
"Assoclation of HLA-DRw4 With Rheumatoid Arthritis ln Black and Whit~ Patients". 1980: 23: 1241.

5. Martell RW, Stein M, Davis P, West G, Emmanuel J &
du Toit ED. "The Association Between HLA and Rheumatoid Arthritis in Zimbabwean Blacks". TissUe Antigens 1990: 36: 125-126.

6. Ohta N, Nishimura YK, Tanimoto R, Horiuchi Y, Abe C
Shiokawa Y, Abe T, Katagiri M, Yoshiki T & Sasazuki T. "Association Between HLA and Japanese Patients ~ith Rheumatoid Arthritis". Hum Immunol 1982: 5:
123-132.

7. Christiansen FT, Kelly H ~ Dawkins RL. I'Rheumatoid Arthritis". In: Albert ED, Baur MP, Mayr WR, eds.
'Hlstocompatibility Testing" 1984. Heidelberg:
Sprlnger Verlag, 1984: 378-379.
.

8. Willkens RF, Nepom GT, Marks CR, Nettles JW ~ Nepom BS. "Association of HLA-Dw16 With Rheumatoid Arthri*~s in Yakima Indians". Arth & Rheum 1991:
34: 43-47.

9. Duquesnoy RJ,, Marrari M, Hackbarths S & Zeevi A
"Serological and Cellular Definition of a New HLA-D~
Associated Determinant, MC1, and its Association With Rheumatoid Arthritis". Hum Immunol 1984: 10:
165-176.

WO92/1~698 PCT/CA92/00105 21~a9~ -lO. Weyand CM, Goronzy J & Fathman CG. "Human T-cell Clones Used to Define ~Functional Epitopes on ~LA
Class II Molecules". Proc Natl Acad Sci USA 1986-83: 762-766.

11. Winchester RJ & Gregersen PK. Ilrrhe Molecular Basis of Susceptibility to Rheumatoid Arthritis: The Conformational Equivalence Hypothesis". Springer Semin Immunopathol 1988: 10: 119-139.

12. Heyes J, Austin P, Bodmer J, Bodmer W, Madrigal A, Mazzilli MC & Trowsdale J. "Monoclonal Antibodies to HLA-DP-Transfected Mouse L Cells". Proc Natl Acad Sci 1986: 83: 3417-3421.

13. Shulman M, Wilde CD & Kohler G. "A Better Cell Line For ~aXing Hybridomas Secreting Speci~ic Antibodies". Nature 1978: 276: 269.
, 14. Drover S, Marshall WH & Younghusband HB. "A Mouse Monoclonal Antibody With HLA-DR4 Associated Specificity~'. Tissue Antigens 1985: 26: 340 343.

15. Morris ~E, Thompson PT & Hong R. "Cellular Enzyme-linked Immunospecific Assay (CELISA) I. A New Micromethod That Detects Antibodies to Cell Surface Antigens". Hum Immunol 1982: 5: 1-19. ~ -: -"
16. Drover S. & Marshall WH. "Glutaraldehyde Fixation of Target Cells to Plastic For ELISA Assays of Monoclonal Anti-HLA Antibodies Produces Artefacts".
J Immunol Methods 1986: 90: 275-281.

17. Dejelo CL, Braun WE, Zachary AA, Taresi, GA
Smerglia ~R ~ Clark LV. Hum Immunol 1986: 17: 135-136.

18. Hiraiwa A, Yamanaka K, Kwok WW, Mickelson XM, Masewicz S, Hansen JA, Radka SF & Nepom GT. ~
"Structural Requirements for ~ecognition of the HLA- . -Dw14 Class' II Epitope: A Key HLA Determinant Associated With Rheumatoid Arthristis". Proc Natl Acad Sci USA l99O: 87: 8051-8055. ~;

~' ' . : .
:., , 21~5969 19. Alber CA, Watts R, Klohe EP, Drover S, Marshall WH, Radka SF & Karr W. "Multiple Regions of HLA-DR~l Chains Determine Polymorphic Epitopes Recognized By Monoclonal Antibodies". J Immunol 1989: 143: 2248-20. Gregersen PK, Silver J & Winchester RJ. "The Shared Epitope Hypothesis. An Approach To Understanding th~ Molecular Genetics of Susceptibility to Rheumatoid Arthritis". Arth & Rheum I987: 30: 1205-. .
21. Sanchez B, Moreno I, Magarino R, Garzon M, Gonzales ~.
MF, Garcla A ~ Nunez-Roldan A. "HLA-DRwlO Confers the Hlghest Susceptibility to Rheumatoid Arthritis n a Spanlsh Population". Tissue Antigens 1990: 360 22. Young Yank S, Milford E, Hammerling U ~ ~upont B.
'Description of the Reference Panel of ~-Lymphoblastoid Cell Lines for Factors of the ~A
System: The B Cell Panel Designed For the Tenth Internatlonal Histocompatibility Workshop" In-Dupont 8, ed. "Immunobiology of HLA. Vol.
Hlstocompatibility Testing" 1987. New York- -Springer-Verlag 1989: 420-425.

23. Knowles R~, Adluri V, Xilaru S, Sirotina A and Small TN. "Binding Patterns of the Monoclonal Antibodies in the 10th Workshop 2-D Gel Analysis of Class II
Antlgens". In: Dupont ~, ed. "Immunobiology of HLA. Vol. 1. Histocompatibility Testing" 1987 New York: Springer-Verlag 1989: 420-425.

24. Bugawan TL, Horn GT, Long CM, et al. "Analysis of HLA-DP Allelic Sequence Polymorphism Using the In Vitro Enzymatic DNA Amplification of DP-~ and DP-~
Locl'. J Immunol 1988: 141: 4024-4030. `

2S. Angelini G, Bugawan TL, Delfino L, Erlich HA &
Ferrara GB. "HLA-DP Typing by DNA Amplification and Hyrldization With Specific Oligonucleotides". Hum Immunol 1989,: 26: 169-177.
: ~:
26. ~ernandez-Vina M, Moraes ME & Stastny P. "DNA
Typing For Class II HLA Antigens With Allele- `
Specific or Group-Specific Amplification III.
' W092/1~698 PCT/CA92/00~0~
21~59~
.

26. Fernandez-Vina M, Moraes M~ & Stastny P. ~'DNA
Typ~ n~ For Class II HLA Antigens With Allele-Speclfic or Group-Specific Amplification III. Typing or 24 Alleles of HLA-DP". Hum Immunol l99l: 30:

Claims (11)

1. A set of monoclonal antibodies that react with epitopes on DR4 molecules.

2. Murine monoclonal antibodies which are specific for HLA-DR4 molecules.

3. As an antibody of claim 2, NFLD.D1 which binds to all DR4 molecules.

4. As an antibody of claim 2, NFLD.D12, which binds only to the Dw4 subtype of DR4.

5. As an antibody of claim 2, NFLD.D14, which binds to Dw4 and Dw14 subtypes.

6. As an antibody of claim 2, NFLD.D7, which binds to all DR4 and DR2 molecules but also, less strongly with several non-DR4 molecules.

7. As antibodies of claim 2, NFLD.D2, NFLD.D3, NFLD.D4. NFLD.D8 and NFLD.D9, which bind strongly to Dw4 and Dw14, but not al all to the subtype of DR4 called Dw10; which give moderate to low reactions with some other DR4 subtypes; and also react with DR1, DR2, and DR14 (Dw16).

8. As an antibody of claim 2, NFLD.D10 which reacts with the Dw9 subtype of DR14 as well as binding weakly to some of the DR3-, DR7-, and DR9- typed B cell lines.

9. The use of the murine monoclonal antibodies of claim 2 to detect subtypes of DR4.

10. The use of the murine monoclonal antibodies of claim 2 which react with putative RA-susceptibility determinants for the study of rheumatoid arthritis.

11. Murine monoclonal antibodies recognizing polymorphic determinants of HLA.