CA2239868A1 - Compositions and method for stimulating antibody release by b lymphocytes - Google Patents

Abstract

The invention describes compositions comprising adding granulocyte-macrophage colony stimulating factor (GM-CSF), interleukin-3 (IL-3), or a combination thereof useful for stimulating the release of antibody by B cells. Methods of using the compositions, pharmaceutical compositions, vaccines, and vaccines adjuvants are also described. In addition, this invention describes an assay system useful for identifying compounds capable of stimulating the release of antibody by B cells.

Description

W O 97/20g40 PCT~US96/19327 COMPOSITIONS AND METHOD FOR STIMULATING
ANTIBODY RELEASE BY B LYMPHOCYTES

CROSS-REFFRFl~CF, TO RELAT~n APPLICATION
This application is a continuation-in-part of prior pending application Serial No. 08/467,146, filed June 6, 1995, which is a continuation of application Serial No. 08/315,492, filed September 30, 1994, and which is a continuation-in-part ofapplication Serial No. 08/150,510, filed November 10, 1993, hereby incorporated by ~f~ ce.
GO~,~NMENT INTEREST
The invention described herein may be m~m-f~ctllred, licensed, and used for gov~:rnmen~l purposes without the payment of any royalties to the inventors or ~ign~e FJF,l,l> OF THE INVF,NTION
This invention relates to compositions comprising granulocyte-macrophage colony stim~ ting factor (GM-CSF), interleukin-2 (IL-2), interleukin-3 (IL-3), and a T cell stimlll~ting peptide, such as a universal T cell epitope (TCE), either alone or in combination, and compositions of such factors, either alone or in combination, along ~,vith h,~lr~ n-y (IFN-y). The compositions are useful for stim~ ting the release of antibodies by m~mm~ n B lymphocytes. This invention also relates to an in vitro assay system for identifying compositions that stiml~l~f~ the release of antibodies by B lymphocytes.
Stimulation of antibody release by B Iymphocytes is useful in the battle to prevent, treat, and/or ameliorate the deleterious effects of infection and disease. This usefulness extends to adjuvants for bolstering m:~mm~ n immune responses under normal conditions and under immnn~suppressed or immllnncompromised conditions. The novel compositions can also be used in conjunction with other immlln~therapies to bolster the human immllne system.
B~CKG~OUND OF T~ INVENTION
The human immune system comprises numerous dirr~ types of cells having overlapping functions which together act to protect the human body against sickness and disease. The cells of the immllne system have complex multiple functions and interconnecting relationships.
GM-CSF, IL-2, IL-3, and IFN-y are all cytokines. "Cytokines" are a class of compounds that regulate responses of cells of the immllne system, such as B and T lymphocyte cells ("B cells" and "T cells") and natural killer ("NK") cells. A "cytokine" is a generic term for a non-antibody protein released by certain cell populations on contact with an inducer and which acts as an intercellular m~ or. A "lymphokine" is a soluble substance released by sen~iti7f~cl Iymphocytes on contact with specific antigen or other stimuli which helps effect cellular or humoral immllnity.
The terms "cytokine" and "Iymphokine" have become interchangeable. In an attempt to simplify the nomenclature of these compounds, a group of participants at the Second Tnt~rn:~tional Lymphokine Workshop held in 1979 proposed the term "interleukin,"
abbreviated "IL," to develop a ullir~lln system of nomencl~hlre based on the ability of the proteins to act as communication signals between different populations of leukocytes.
To date, 21 difr~lellt cytokines, most but not all of which are produced by T cells, have been identified. Each has a distinct molecular configuration and pc.~lllls a dirreL~
task. A number of the known cytokines have been shown to have a demonstrable activity on B cells. In vitro, the lymphokines IL-1, IL-2, IL-4, IL-5, IL-6, IL-10, IFN-y, and TGF-,~
(transforrning growth factor ,(~) have been shown to enh~nce B cell proliferation, imml-nnglobulin secretion, or to otherwise play a role in infll-~ncing the subclass of secreted Ig. Depending on the system being studied, addition of either one or a number of the above Iymphokines has been shown to increase in vivo antibody production or to alter the isotype (i.e., IgG, IgM, IgE, IgA, etc.) of secreted antibody. Arnong the Iymphokines reported to influence B cell proliferation include IL-l, IL-2, IL-4, and IL-10, and those reported to influence B cell differentiation and Ig secretion include IL-2, IL-5, IL-6, TGF-,~, and IFN-y.
None of the reported cytokines which enhance Ig secretion in vifro have been shown to play a prominent role in vivo. Thus, infusion of monoclonal antibodies specific for IL-2, IL-5, IL-6, or IFN-y does not significantly ~iU~ i antigen-stimulated antibody W O 97/20940 PCTrUS96/19327 production. This suggests that, under physiologic conditions, B cell differentiation depends solely on direct T cell interaction, or that other as yet unknown cytokines me~ te this step.
While immlme responses to antigens that stimulate T cell activation (the so-called T dependent antigens or "TD antigens") could rely on direct T cell interactions with B cells to effect Ig secretion, this is not the case with antigens that are unable to induce T cell activation. Antigens that are T cell independent ~"TI antigens") induce high levels of antibody production in the ~ks~-nce of direct or even indirect T cell help. Thus, the events that regulate B cell dirre~ liation and immlmoglobulin secretion to TI antigens must rely on other as yet nn~l~fin~-l p~hw~ys. Since B cell differentiation leading to immlmoglobulin secretion is the final event which underlies a competent humoral antibody system, defining the events or cytokines which regulate this step is invaluable in deeigning methods for amplifying or ~UplJlG~illg an immune response.
To facilitate a quick appreciation of the invention, the following provides a brief description of the ~,limal y known functions of immlln~globulin~ antibodies, Iymphocytes, B cells, T cells, and NK cells as background. Also provided is a brief surnmary of the known activities of the cytokines reported to influence B cell proliferation or antibody secretion, namely IL-l, IL-2, IL-4, IL-5, IL-6, IL-1 O, TGF-,~, and IFN-y . A sumrnary of the known activities of GM-CSF and IL-3 is also provided. Reference m~teri;~ includeFlmcl:~merltal Immunolo~y, Second Edition, William E. Paul, M.D., ed. (Raven Press, New York 1989); Funtl~ment~l Immunology. Third Edition, Williain E. Paul, M.D., ed. (Raven Press, New York 1993); Tnlr~re~on: Principles ~ncl Medical Applications. S. Baron et al., eds. (The University of Texas Medical Branch at Galveston, Galveston, Texas 1992); The Cytokine ~ndbook, Angus Thomson, ed. (Ac~ nnic Press Inc., San Diego, CA 1992);
and The Cytokine Handbook Second Edition, Angus Thomson, ed. ~Academic Press Inc., San Diego, CA 1994), all of which are specifically incorporated by re~erence.
~ mm~l~, including man, are confronted on a daily basis with a myriad of org~ni~m~ A major component of the immlm~ system and playing an essential role in protecting the host against infection by these org~ni~m~, is the humoral antibody.
Antibodies are protein molecules, also known as immunoglobulins, which have exquisite specificity for the foreign particle which stim~ tt~s their production. For example, W O 97/20940 PCTnJS96/19327 systemic infection with "bacteria A" will induce antibodies that bind with a high avidity to "bacteria A" but not to "bacteria B." Similarly, "bacteria B" will induce anti-"bacteria B"
antibodies that do not cross-react with "bacteria A."
Tmmlmoglobulin (Ig) is a class of structurally related proteins consisting of two pairs of polypeptide chains, one pair of light (L) [low molecular weight] Kchains~K or ~), and one pair of heavy (H) chains (~ , ,u, o, and ~), all four linked together by flienlfide bonds. Both H and L chains have regions that contribute to the binding of antigen and that are highly variable from one Ig molecule to another. In addition, H and L chains contain regions that are nonvariable or constant. On the basis of the structural and antigenic properties of the H chains, Ig's are classified as IgG, IgA, IgM, IgD, and IgE isotypes.
Subclasses of IgG's, based on differences in the H chains, are referred to as IgGl, etc.
Lymphocytes are white blood cells formed in lymphatic tissues throughout the body, such as lymph nodes, spleen, thymus, tonsils, Peyer's patches (small intt?etinP tissue3, and sometimes in bone marrow. Individual lymphocytes are speci~li7~l in that they are c~ .od to respond to a limited group of structurally related antigens. This comrnitment, which exists prior to the first contact of the imm-me system with a given antigen, is expressed by the presence of antigen-specific receptors (i.e., immlm()globulin) on the Iymphocyte membrane. The ability of an organism to respond to virtually any antigen is achieved by the ~xie~nce of a very large number of dirr~,.ellt clones of Iymphocvtes, each bearing receptors specific for distinct ~ntig~n~ In consequence, lymphocytes are an enormously heterogeneous collection of cells.
Lymphocytes differ from one another not only in the specificity of their lect;pLol~
but also in their functional plopelLies. Two broad classes of lymphocytes are recognized:
the B lymphocytes and the T lymphocytes. In addition to these two classes, lymphoid cells that mediate certain "nonspecific" cytotoxic responses are known. These include natural killer (NK) cel}s.
B Iymphocytes, also kno~,vn as "B cells," are a type of Iymphocyte that derive from hematopoietic stem cells by a complex set of dirrel~llLiation events that are only partially understood. B cells are precursors of antibody-secreting cells and thus are responsible for the production of immunoglobulins The cell-surface receptor of B cells is an antibody or W O 97/20940 PCTrUSg6/19327 immlm--globulin (Ig) molecule specialized for expression on the cell surface. Newly dir~-ellLiated B cells initially express surface Ig solely of the IgM class. Associated with maturation of a B cell is the appearance of other immllnoglobulin isotypes on the surface of the B cell.
To release antibody in response to cytokines, the B cells must first be activated.
There are many ways to activate B cells, including cross-linkage of membrane Ig molecules by the antigen (cross-linkage-dependent B cell activation), direct encounter with T cells (helper T cells or helper T cell-associated molecules, such as, for example, CD40 ligand), or encounter with mitogens. In such encounters, the antigen presents epitopes recognized by the B cell's cell-surface Ig.
Because each B cell bears multiple membrane Ig molecules with identical variableregions, optimal membrane-Ig mediated cross-linkage activation is achieved by a high level of cross-linkage of the cell-surface receptors, which requires that the antigen present more than one copy of the epitope that the cell-surface Ig recognizes. Although many simple protein antigens do not have this potential, such a requirement is fulfilled by polys~ch~ri(les and other antigens with .c~!e~Ling epilopes, such as surfaces of microbes and DNA. Among these antigens are the ç~ps~ r polys~c~h~rides of many medically illll,olL~L microorg~ni~m~, such as pneumococci, streptococci, and meningococci.There are much data to show that cross-linkage of membrane Ig can also lead to elimin~tif~n or inactivation of B cells. In general, it is believed that certain types of receptor cross-linkage events, if they occur in the absence of specific sfim~ tory signals, lead to inactivation rather than activation. The highly repetitive epitopes e~ ed on polys~cch~rides may lead to activation in the absence of co-stimnl~tion, possibly because of the m~gnil~ e of the lece~JLol-mediated stimul~tion.
T Iymphocytes, or "T cells," are thymocyte derived, o~imml]nological importance that is long-lived (months to years), and are responsible for cell-mediated immunity.
T cells consist of functionally different populations, known as "helper T cells," "suppressor T cells," and "killer T cells." T cells involved in delayed hypersensitivity and related immlme phenomena are also known.

W O 97/20940 PCT~US96/19327 Natural killer cells, or "NK cells," are lymphoid cells that mediate certain "nonspecific" cytotoxic responses. Such nonspecific cytotoxic responses kill certain forms of tumor cells using recognition systems that are dirre~ L from those used by T or B cells.
Killing of one cell type by another through contact inter~rtion COllS~ S a major effector arm of self-defense of the immllne system.
In addition to these cells, other compounds, the cytokines, play a significant role in protecting a host. One group of cytokines is the interleukins.
IL-1 is primarily an infl~mm~tf~ry cytokine, whereas IL-2 and other cytokines are primarily growth factors for Iymphocytes. IL-1 is a polypeptide hormone synth~i7~d by monocytes. During infl~mm~tion, injury, immlln~logical challenge, or infection~ IL-1 is produced and, because of its multiple biological properties, this cytokine appears to affect the pathogenesis of the disease. In ~nim~l~, IL-l is a potent inducer of hypotension and shock. IL-1 acts on the hypoth~l~mll~ to induce fever and directly on skeletal muscle to promote protein catabolism IL-2, also known as T cell growth factor, is a lymphokine and polypeptide hormone produced by both T helper and suppressor Iymphocytes. This cytokine has direct effects on the growth and di~elcllliation of T cells, B cells, NK cells, lymphokine-activated killer (LAK) cells, monocytes, macrophages, and oligodendrocytes.
IL-3, also known as multicolony stim~ ting factor, acts on numerous target cellswithin the hemopoietic system. This cytokine has the broadest target specificity of any of the h~m~topoietic growth factors (HPGFs), and can stimlll~te the generation and liation of hemopoietic stem cells (i.e., precursors of blood cells), which give rise to macrophages, neutrophils, eosinophils, basophils, mast cells, megakaryocytes, and erythroid cells.
The relationship between IL-3 and B cells was unclear prior to the invention. Infact, as of 1994, it was believed that the range of target cells of IL-3 did not include cells committed to the T- and B-lymphoid lineages, and there was no compelling evidence that IL-3 had a significant, direct effect on B-cell development. J. W. Schrader, "Chapter 5:
Interleukin-3," The Cytokine Handbook~ 2nd Ed., Angus Thomson, ed., page 84 (Academic Press, New Yorlc, 1994).

-W O 97/20940 PCTrUS96/~9327 Secretion of IL-3 by B cells has not been reported, although IL-3 is synthe~i7~cl by T cells and mast cells. Several reports demonstrated that IL-3 could induce a modest enh~nef~ment of Ig secretion by human B cells activated with SAC (a polyclonal activator) and IL-2. For example, Xia et al., "Human Recombinant IL-3 is a Growth Factor for Normal B Cells," J. of Immunology~ 148~ 491-497 (1992), reported that IL-3 enh~nee-l the proliferation of a population of cells enriched in B cells. Similarly, Tadmori et al., "Human Recombinant IL-3 Stim~ te~ B Cell DirfelellLiaLion," J. of Tmmllnol., 142, 1950-1955 (1989), reported that IL-3 stimulated IgG secretion from tonsillaF cells cont~ining B cells or in a population of peripheral blood-derived enriched B cells activated by bacterial antigen.
In addition, Matsumoto et al., "Induction of Igl~ Synthesis in Anti-IgM-Activated Nonatopic Human B Cells by Recombinant Lnterleukin-3," Tnt Arch. Allerg~v Apl~ln.
Tmml-nol. 89, 24-30 (1989), reported that hurnan recombinant IL-3 ~ngmentt~l IgEsynthesis by normal B cells or nli~Lules of T and B lymphocytes, and that IL-1, IL-2, IL-5, IL-6, GM-CSF, G-CSF, M-CSF, and IFN-~y failed to induce IgE synthesis. Matsumoto et al. also note that they could not conclusively identify IL-3 as the factor in the T cell ~u~ responsible for inducing IgE synthesis because the activity could not bereversed by the addition of anti-IL-3 antibody.
These results were attributed to an IL-3-mediated çnh~n~ement in cell growth. Inthese studies, B cells were not electronically sorted and thus were not highly purified.
Thus, the Ig enhancing effect may reflect the action of IL-3 on many cont~min7~ting non-B, non-T cells in the population, and it is therefore not possible to ~letermine from these ";,.~ent~ whether IL-3 was acting directly on the B cell. Further, in prior experiments, B cells were not fractionated according to size. Thus, a possible role for the prior activational state of the B cell was not addressed.
In further contrast to the present invention, Kimoto et al., "Recombinant MurineIL-3 Fails to Stim~ te T or B Lymphopoiesis In Vivo, But Fnh:~n~e~ Immune Responses To T Cell-Dependent Antigens," J. of Immunology, 140. 1889-1894 ~1988~, reported that mice bearing osmotic mhn~ s loaded with murine recombinant IL-3 showed no increase in the lymphoid organs of the total number of B and T cells. Furthermore, Kimoto et al.
~lgge~ 1 that IL-3 does not act directly on lymphocytes or their precursors. but may potentiate the humoral immllne response to T eell-dependent antigens, presumably by aeting on aeeessory eells.
IL-4 is a glyeo~ tehl also known as B eell sfim~ tin~ factor 1 (BSF-l) and B eell di~rcl~llliation faetor. It funetions to co-st;m~ te B eell growth, Ig elass switehing, T eell growth and differentiation, maerophage aetivation, regulate mast cell growth, and to co-stimnl~te hematopoietic preeursor cells.
IL-5, also known as B eell growth faetor II (BGF-II~, T cell replacing factor, IgA-enhaneing factor, and eosinophil colony stiml-l~tin~ faetor, is a glycoprotein produced by T lymphoeytes and mast eells. This eytokine has the dual funetions of a eolony Stimlll~tinp faetor, as well as promoting the dirrciellLiation of eosinophilie eolonies in bone marrow. IL-5 induees speeifie in vitro antibody produetion by B eells primed with antigen in vivo. While IL-5 serves as a di~clcllli~Lion faetor in vitro, it does not appear to aet as a dirr.,lcll~iation faetor in vivo.
IL-6, also known BSF-II, hybridom~plasmaeytoma growth faetor, illLclrcl~ (32, and hepatoeyte stimulator.v factor, is a glyeoprotein produeed by both lymphoid and nonlymphoid eells. This eytokine regulates immlme responses, aeute-phase reaetions, and hemopoiesis. IL-6 aets on B eell lines at the mRNA level and induees biosynthesis of seeretory-type Ig. In addition to IL-5, IL-6 has also been shown under very restrieted eonditions to funetion as a differentiation faetor. All other known T eell or maerophage derived faetors that have been tested eannot induee activated B cells to secrete Ig in the absence of added growth factors.
IL-10, also known as eytokine synthesis inhibitory faetor, is produced by T eells, macrophages, and other cell types. This cytokine inhibits several macrophage functions, including cytokine synthesis and some microbial activities, in addition to enhancing or stimulating mast eells and B cells. IL-l 0 eauses strong proliferation of human B eells aetivated by anti-CD40 antibodies or eross-linking of the antigen reeeptor.
In addition to the interleukins, other eytokines have been eharaeterized. Colonystim~ ting faetors (CSFs) are a group of faetors primarily eoneerned with hematopoiesis.
They are defined as proteins whieh ~timlll~tP the elonal growth of bone-marrow eells in vitro.

W O 97/20940 PCTnUS96/19327 Granulocyte-macrophage colony stimnl~tin~ factor (GM-CSF) is a glycoprotein growth factor that modulates the growth or differentiation of hemopoietic cells. This growth factor can be produced by a number of different cells under different circumstances, including T cells, macrophages, endothelial cells, stromal cells, fibroblasts, mast cells, and others. The major actions of GM-CSF involve the regulation of survival, differentiation, and proliferative and functional activities in granulocyte-macrophage populations. There are no reports prior to the invention indicating that GM-CSF can stim~ te the release of antibody by B cells.
Finally, another class of cytokines that function in the body's immllne system is interferons (IFNs). IFNs are major contributors to the first line of antiviral defense by inhibiting virus replication, in addition to exerting many other important effects on cells.
IFNs do not act directly to protect cells from infection. Rather, they stimulate production of a protein in neighboring cells that stops the growth of the virus, thus protecting the cells from infection.
IFNs are classified into three groups, alpha, beta, and g~nnm~, based on the cells of origin and method of induction. The production of IFN-oc and IFN-,B is not a specialized cell function, and probably all cells of the organism are capable of producing these IFNs.
In contrast to IFN-cc and IFN-,~ synthesis, which can occur in any cell, production of IFN-y is a function of T cells and NK cells. All IFN-y inducers activate T cells either in a polyclonal (mitogens or antibodies) or in a clonally restricted, antigen-specific manner.
Human IFN-y promotes proliferation of activated human B cells and, in cultures of human B cells, can act synergistically with IL-2 to enhance immnn~globulin light-chain synthesis.
*****
The brief discussion describing functions of various human immune cells, and theknown activities of IL-l, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, GM-CSF, and IFN-yexemplifies the extreme diversity of the human immllne system. Despite this level of knowledge, however, there is no complete understanding of the intricacies of the imm~me system. Tremendous gaps remain. For example, it is not possible to make general stzltem~nt~ about the properties of cytokines, except that they act as intercellular mediators by regulating responses of cells of the immune system. Thus, there remains a need in the WO 97~0940 PCT~US96/19327 art for a greater understanding of the imml-ne system and for the provision of additional and superior methods of treating imml-n~ disorders.
S~IlVl~RY OF TR~li', Tl~ TION
The present invention fulfills a need in the art for new and improved imm--n-)therapies. The novel compositions and methods, employing IL-2,IL-3, G M-CSF, a universal TCE, and IFN-y, enable improved and new tre~tment.~ for immun~ disorders, as well as adjuvants for current imm~-notherapies.
This invention is directed to a composition of G M-CSF a I d IL-2 or IL-3, either alone or in combination, present in an amount effective for the stiml-l~tion of antibody release by B cells. Another object of the invention is directed to compositions of GM-CSF
and IL-2 or IL-3, either alone or in combination, present in an amount effective for the stiml-l~tion of antibody release by B cells, wherein the resultant complex is conjugated to a protein, and wherein that complex is additionally conjugated to a polysacrh~ri(le. The invention also embodies combinations of other lymph-)kin~ s that can enh~n~e imml-ne~
reactivity by increasing the responsiveness of cells participating in the illlllllll.~. response, as established by T or B cell proliferation in vitro or by in vifro antibody form~tinn Molecularly enginP~red fra~ment~ of GM-CSF, IL-2 or IL-3 that retain GM-CSF, IL-2 or IL-3 activity, respectively, can also be employed in the invention.
Another object of this invention is a composition of GM-CSF, IL-2 and IL-3, either alone or in combination, along with IFN-y, all of which are present in amounts effective for the stimulation of antibody release by B cells.
Another object of the invention is a composition comprising a U~ ,al TCE
non-antigen specific conjugated to a protein, wherein that complex may also be further conjugated to a polysaccharide. In another embodiment, a universal TCE is directly conjugated to a polysaccharide.
A further object of the invention is a ph~rrn~entical composition comprising thenovel compositions and a ph~rrn~elltically acceptable carrier.
The use of the novel compositions comprising GM-CSF, IL-2,IL-3,IFrN-y, and the universal TCE to stimulate the release of antibody by B cells is also encompassed by the invention. A method of stimlll~t;n~ the release of antibody by B cells can be used to W O 97/20940 PCT~US96/19327 bolster m~mm~ n immlme responses to, for example, vaccination under conditions of normal or immlmclcu~ u~ised conditions.
A further object of this invention is the use of the novel compositions as adjuvants for vaccines. For example, many vaccines are currently ~lm;ni~tered h~ vt;.lously or intramllscnl~rly to allow a rapid stimlll~tion of immllne cells present in the blood system.
By combining the novel compositions, either as a fusion protein covalently linked to a carrier molecule, admixed, or any other combination, with a vaccine to be ~-lmini.~tiored, the magnitude of the antibody response can be increased both at the systemic and local levels.
~ nother object of this invention is neutralization of GM-CSF, IL-3, and IFN-~under conditions where the production of antibody is pathogenic, such as in autoimmlml?
disorders.
The compositions can also be used to optimize monoclonal antibody production in vitro or in vivo. For example, an animal can be s~n~iti7~d with antigen and the compositions in vivo. Alternatively, in vitro stimlll~tion or s~n~iti7~tion of Iymphocytes to produce antibody can be f?nh~n~e~ in the presence of the novel compositions. This is particularly useful for the production of human monoclonal antibodies.
This invention is also directed to a novel assay system that allows the identification of compositions useful for stim~ ting antibody release by B cells. This assay system, which mimics in vivo antibody stimulation, comprises dextran-conjugated anti-Ig antibodies and highly purified B cells. The anti-Ig-dextran conjugate effectively and polyclonally activates the B cells via membrane Ig by a mechanism comparable to activation of B cells in~ ced by antigen in vivo.
Other objects and advantages of the present invention will be set forth in part in the description which follows, and in part will be obvious from this description, or may be learned from the practice of this invention. The accompanying drawings and tables~ which are incorporated in and constitute a part of this specification, illustrate and, together with this description, serve to explain the principle of the invention.

W O 97/20940 PCT~US96/19327 BRIEF I~ESCI~TPTION OF T~ DRAWINGSFi,~nre 1: IgM secretion (ng/ml) by B cells in various mediums was measured.
Culture mediums are listed along the left side of the chart. The compositions tested showed the highest IgM secretion in medium co"~ ing dextran-conjugated anti-IgD antibody, IL-l, and IL-2. To this me~ m was added a supe~ LL~l~ from a cell culture described in the parent application, RA5-SN, and GM-CSF or IL-3. A control composition was also measured. The RA5-SN supern~nt having both GM-CSF and IL-3, showed significant IgM secretion by B cells as compared to other tested compositions.~~gure 2: B cells were activated with dextran-conjugated anti-IgD antibodies (3 ng/ml) plus IL-1 (150 U/ml) + IL-2 (150 U/ml) in the presence or absence of varying concentrations of IL-3 or GM-CSF. IgM secretion was measured by ELISA 6 days after initiation of culture.~~ure 3: IgM secretion (ng/ml) by activated B cells was measured with the addition of varying concentrations of GM-CSF and IL-3. ~;lx;~ l IgM secretion of about 4650 ng/ml was obtained with the addition of GM-CSF at 10 U/ml and IL-3 at about 3 U/ml.~i~ure 4: B cells were activated with dexkan-conjugated anti-IgD antibodies (3 ng/ml) plus IL-1 (150 U/ml) + IL-2 (150 U/ml) in the presence or absence of IL-3 (10 U/ml) or GM-CSF (10 U/ml). In addition, anti-IL-3 (10 ~Lg/ml) and/or anti-GM-CSF ~10 ~lg/ml) antibodies were added. A control with no added anti-IL-3 or anti-GM-CSF was also prepared. IgM concentrations were measured by ELISA 6 days after initiation of culture.~~llre ~: IgM secretion (ng/ml) by activated B cells was measured for dextran-conjugated anti-IgD antibody-activated B cells in AF7 supçrn~t~nt.
IgM secretion was measured in various mediums com~ri~ing two or more of the following: IL-l, IL-2, aIL-3, and aGM-CSF.~~ure 6: B cells were activated with dextran-conjugated anti-IgD antibodies (3 ng/ml) plus IL-I (150 IJ/ml) + IL-2 (150 U/ml) in the presence or absence of IL-3 W O 97t20940 PCT~US96/19327 (100 U/ml) and/or GM-CSF (100 U/ml). Viable cells (those that excluded trypan blue) were enumerated 4 days after initiation of culture using a hemocytometer. IgM concentrations in replicate cultures were measured by ELISA 6 days after initiation of culture. Data is ~ ;s~ d as mean +/-standard error of the mean of duplicate cultures.
F~nre 7: B cells were activated with dextran-conjugated anti-IgD antibodies (3 ng/rnl) plus IL-1 (150 U/ml) + IL-2 (150 U/ml) in the presence or absence of IL-3 (100 U/ml) + GM-CSF (100 U/ml). IgM concentrations in replicate cultures were measured by ELISA for a control composition, a composition to which IFN-y was added at 0 h, and for a composition to which IFN- y was added at 24 h. Data is lc:~lesellLed as mean +/- standard error of the mean of duplicate cultures.
Fi~ure 8: S~ m~tic of yeast t;~l,rei,~ion vector Fu~ure 9: Coomassie blue stained gel Co~ protein A-IL2 (PA-IL2) product prior to and after purification on nickel agarose column and sizing gel.
F~ure 10: Immunoblot of protein A-IL2 (PA-IL2, blotted with anti-IL2 (left hand gel) and anti-prote;n A (PA) (right hand gel). IL2 and PA are run as controls;
protein A runs at a higher molecular weight than fusion product because fusion was made with a truncated version of protein A.
Fi~ure 11: F~ln~tional activity of IL2 in fusion product. Sx103 HT2 cells, a CllJL line e~l,oll~ive to IL2 was with IL2 or protein A-IL2 and thymidine incorporation was measured 48 hours later.
Fu~ure 12: Size exclusion HPLC (Beckman SEC 2000) column profile of: A)-purified protein A-IL2 (PA-IL2) fusion product B)-non purified concentrated supern~t~nt~ cont~ining PA-IL2 from protease free yeast. No purification was done on the supern~t~nt~, demonstrating that a "cleaner" product with less lower m.w breakdown products was obtained in protease free transfected yeast.
F~ure 13: Purified protein D (lanes 1 and 2) and molecularly engineered universal TCE-protein D (TCE-protein D, lane 4) and protein A (as a control~ lane 5) W O 97/20940 PC~US96/19327 were run on SDS PAGE and stain by Coomassie blue (upper panel). A
second gel was immllnoblotted with anti-protein D ~lower panel). The figure shows the proteins and TCE-protein D were purified to homogeneity.
D~.~CRIPT~ON OF PREFERRFl) EM~ODIMFNTS
The invention describes compositions of cytokines which individually and in combination lead to 100 fold enhancement in antibody secretion by B cells. The == compositions comprise IL-2, IL-3 and GM-CSF, either alone or in combination, in an amount effective for stim~ tin~ the release of antibody by B cells. Also encomp~cce~l by the invention are compositions comprising IL-2, IL-3 and GM-CSF, either alone or in combination, and IFN-~, all present in amounts effective for stim~ ting the release of antibody by B eells. More preferably, the compositions additionally eomprise CD40 ligand (CD40L), or at least one other cytokine, or a combination thereof. Also more preferably, the eompositions additionally comprise CD40 ligand (CD40L), a universal TCE, or at least one other cytokine, or a eombination thereof. By ulPi~L~l TCE, is meant a non-antigen specific, strongly T cell ctim~ tory peptide, pocc~ccing no B cell epitopes. Strongly T eell stiml-l~tory means able to potently stimu~ T eells as reflected in T cell proliferation assays known to those ordinarily skilled in the art. By peptide is meant a molecule of about 8 to 20 amino acid residues, preferably from 11 to 18 rçcid~les Non-antigen specific means that the peptide stim~ tes T cells across a wide variety of antigenic specificities and of divers genetic backgrounds. pocceccing no B cell epitopes means that the epitopes are not recognized by B cells and, therefore, will not induce antibody formation specific to the epitope. The peptide may be chemieally eonjugated to a protein or a fusion protein may be constructed according to the methods well known to those ordinarily skilled in the art. The peptide may be conjugated to a polysaecharide according to methods known to those ordinarily skilled in the art, including CDAP.
Most preferably, the compositions comprise GM-CSF, IL-3, or a eombination thereof, IFN-y, CD40L, and IL-l + IL-2. In another most preferable embodiment, the compositions comprise GM-CSF, IL-3, or a combination thereof, IFN-y, CD40L, universal TCE and IL-l + IL-2. Also most preferably, the compositions comprise GM-CSF, IL-3, IFN-y, CD40L, universal TCE, and IL-1 ~ IL-2, or a combination of some or all of the above conjugated to a protein and further conjugated to a polysaccharide.
The enh~nl~çment in Ig secretion mer1;~ted by IL-2 or IL-3 or GM-CSF typically ranges between 10-50 fold, and the combination of IL-3 and GM-CSF induces enhancement of up to 100 fold. Preferably, GM-CSF and IL-3 are present at from about 1 to about 10 U/ml, in vitro. I~ vivo amounts are scaled up accordingly, as is well known in the art. More preferably, GM-CSF and IL-3 are present at from a~out 10 up to about 100 U/ml, and in particular, at about 100 U/ml.
IDENTIFICATION OF IL-3, GM-CSF, AND IFN-~
AS B CELL STIl~ULATORY AGI~NTS
Prior to the invention, it was not known that GM-CSF or IL-3 act directly on theB cell to stim~ te the release of antibody, or that GM-CSF and IL-3 act synergistically and directly on the B cell to stim~ tP the release of antibody. Moreover, GM-CSF has not been previously implicated in directly regulating mature B cell function. It was also not known that addition of IFN- y 24 hours after ~timnl~tion could by itself stimul~t~ optimal antibody secretion, nor that it could further enhance the activity of IL-3 and GM-CSF, either alone or in combination.
The composition described in parent application Serial No. 08/150,510 comprised IL-3 and GM-CSF. Early e~ nt.~ rlPtt~rmined that the Ig secretory response of electronically sorted highly purified B cells was significantly lower than B cell enriched populations that contained small numbers of "non-T, non B" cells. Based on thesefinrling~, the first application showed that NK cells and/or NK cell-derived cytokines could enhance Ig secretion in anti-Ig-dextran stin~ te~l B cells.
The parent application also disclosed supernatants derived from T cell clones (THl or TH2~ that çnh~nred Ig secretion by B cells. Experiments clct~ rmin~-l that this enhancement was not due to IL-l, IL-2, I3_-5, IL-6, or IL-l0. Surprisingly, it was discovered that the dirr~ Liation-inducing activity was due to the presence of GM-CSF
and IL-3. When GM-C~F and IL-3 were added to anti-Ig-dextran stimulated cells, Ig secretion was in~ cefl Conversely, when anti-GM-CSF and/or anti-IL-3 were added, the stimulatory activity of the T cell supernatants was ~limini~hP~1 It was also found that Ig secretion could be further enhanced by the addition of IL-2, the addition of IFN-y 24 hours after culture, or the addition of IFN-y + IL-2. M~imllm levels of Ig secretion were in~1uced when IL-2, IL-3, and GM-CSF were added to anti-Ig-dextran stimulated cells at the onset of culture, and IFN-y was added one day later (24 hr period after st;mlll~tion with anti-Ig-dextran). IFN-y added im met1i~tely following B cell activation does not enhance the stim~ tQry effect, or minim~lly enh~n~es the ~timlll~tc~ry effect, of IL-3 or GM-CSF. Thus, the timing of the addition of IFN-y is important for further enhancing the release of antibody by B cells. This discovery has not been reported prior to the invention.
To deterrnine whether the activity of IL-3 or GM-CSF, or IL-3 + GM-CSF, could account for the differenti~tin~ activity of the THl or TH2 derived SUpC~ .2~t, the effect of adding neutralizing qn~ntiti~ of anti-IL-3 or anti-GM-CSF antibody on Ig secretion was analyzed. While each antibody m.oflisltecl significant ~u~cs~ion of Ig secretion, the combination of anti-IL-3 and anti-GM-CSF in~luced a greater than 80% suppression of Ig secretion in anti-Ig-dextran stimlll~te-l cells in the presence of TH1- or TH2-derived :iu~)CI llnl~
The fin~1inf~ that IL-3 and GM-CSF can enhance Ig secretion by B cells, and thatanti-IL-3 + anti IL-GM-CSF can suppress Ig secretion, were completely unexpected.
Moreover, it was also surprising that IFN-y added 24 hours after the onset of culture enh~need the stim~ tory effect of IL-3 and GM-CSF. These f;n~ling~ have not beenpreviously reported.
P~Rl~ACl~,UTICAI, COMPOSITIONS
ph~rm~re~ltie~l compositions are also encompassed by the invention. Such compositions comprise an effective arnount of IL-3 and GM-CSF, either alone or in combination, and a ~h~rm~eutically acceptable carrier. Also encomr~ed by the invention are rh~rm~eutical compositions comprising an effective amount of IL-3 and GM-CSF, either alone or in combination, and an effective arnount of IFN-y, along with a ph~rm~eeutically acceptable carrier.
Treatment comprises ~lmini~tering the ph~rm~eeutical composition by intravenous,intraperitoneaL intracorporeal injection, intra-articular, intraventricular, intrathecal~

W O 97/20940 PCTrUS96/19327 intr~mll~c~ r, ~ub~;u~uleous, intranasally, intravaginally, orally, or by any other suitable method of a~lmini~tration. The composition may also be given locally, such as by injection to the particular area, either intramuscularly or subcutaneously.
Any rh~rm~r~eutically accèptable carrier can be employed for GM-CSF, IL-3, and IFN-y. Carriers can be sterile liquids, such as water, oils, including petroleum oil, animal oil, vegetable oil, peanut oil, soybean oil, mineral oil, sesame oil, and the like. With intravenous ~lmini~tration~ water is a preferred carrier. Saline solutions, aqueous dexkose, and glycerol solutions can also be employed as liquid carriers, particularly for injectable solutions. Suitable pharm~eutical carriers are described in Remi~ton's Pharm~-~Putical Sciences. 1 8th Edition (A. Gennaro, ed., Mack Pub., Easton, Pa., 1990), incorporated by reference.
Vaccine adjuvants comprising the compositions This invention also encomp~ec vaccine adjuvants comprising the compositions of the invention. Many vaccines are ~;ullGlllly a-lmini~tered intravenously or il~ ""~cs;ulz~rly to allow a rapid stim~ tion of immllne cells present in the blood system. By combining the novel compositions with a drug to be ~f1mini~t~red, the m~gnitlltle of the antibody response is increased, both locally and systemically.
In the in vitro assay, the anti-IgM or anti-IgD antibody when presented in a multivalent form, such as dextran, acts efficiently to activate all B cells. This high level of activation, coupled with the use of highly purified B cells, allows the identification of compounds that stimlll~tf~ the release of antibody by B cells. However, this response is not desired in vivo. In a patient, the goal is to activate only a small number of B cells having receptors specific for the antigen. The specific antigen of the vaccine acts to cross-link specific B cell receptors. In contrast, the in vitro model employing anti-Ig-dextran acts to cross-link all antigen receptors.
Preferably, when used as a vaccine adjuvant, the compositions of the invention are conjugated to a multivalent carrier molecule, such as dextran or a capsular polys~ççh~ri~l~
of a bacteria. Pneumococci, streptococci, and meningococci capsular polysaccharides are .e~L.~d. GM-CSF, IL-2, or IL-3, molecularly engineered fr~gm~nt~ of GM-CSF, IL-2, or Il,-3 that retain GM-CSF or IL-3 activity, or a combination thereof. can be independently conjugated to the multivalent carrier. Alternatively, GM-CSF, IL-2, and IL-3 can be fused together or fused to another protein to form a fus;on protein, which can also be bound to the multivalent carrier. Other vaccine variations will be a~ l to one of skill in the art.
This application encomp~çs the use of anti-cytokine-cytokine complexes which allow for the slow but prolonged delivery of the cytokine. The complexes can be ~mini~tered as a mixture with the antigen of a vaccine, or the complexes can be bound to the antigen of a vaccine.
In still another embodiment, the vaccine adjuvant can comprise CD40L, one or more cytokines other than GM-CSF, IL-3, or IFN-~, or a combination thereof. CD40L and the one or more cytokines, or a universal TCE can also be bound to the multivalent carrier.
Compositions used in a conjugate vaccine To form a conjugate vaccine, the antigen of the vaccine and the compositions of the invention can be covalently conjugated to a multivalent carrier molecule, such as dextran or a capsular poly~ h~rid~ of a b~rtf ri~ Pneumococci, streptococci, and meningococci capsular polysacch~ritles are ~r~ c,d.
The antigen is a peptide or protein specific for the disease to be vaccinated against.
To further optimize the humoral immllne response upon ~fimini.~tration of the vaccine, CD40 or at least one other cytokine, or a universal TCE, or a combination thereof, can be conjugated to the multivalent carrier.
Table III shows exemplary vaccines employing the compositions of the invention.
As noted in the Table, several of the vaccines are conjugate vaccines. Methods of conjugation are well known to those of oldillaly skill in the art, and include the heteroligation techniques of Brunswick et al., J. Tmmunol.~ 140:3364 (1988); Wong, S.S., Chemi.~try of Protein Conjugates and Crosslinkin~ CRC Press, Boston (1991); and Brenkeley et al., "Brief Survey of Methods for Preparing Protein Conjugates With Dyes, Haptens and Cross-r~inking Agents," Bioconj~ t~ Chemistry~ 3, No. 1 (Jan. 1992),specifically inco-~o-~l~;d by reference. A preferred method of covalent conjugation is set forth in application Serial No. 08/482,661, filed June 7, 1995, which is a continll~ti~n-in-part of application Serial No. 08/408,717, filed March 22, 1995, which is a continuation-in-part of application Serial No. 07/124,491, filed September 23, 1993, the so-called "CDAP"

W O 97/20940 PCT~US96/19327 conjugation method, the disclosures of which are specifically incorporated herein by reference.
MF,THODS OF USING THE COMPOSITIONS
A further object of this invention is the use of the compositions comprising IL-2, IL-3 and GM-CSF, either alone or in combination, and compositions comprising IL-2, IL-3 and GM-CSF, either alone or in combination, along with IFN-~, to stimUl~te the release of antibody by B cells.
Suitable hosts for trç~tmçnt include any suitable m~mm~J Preferred hosts are hllm~n~, including neonates, adults, and immlmodeficient patients.
The compositions can be ~rlminictered employing any suitable ~lmini~tration method. Preferable methods of ~lmini~t~rin~ the compositions include subcutaneously, hlllave~ously~ nasally, mucosal routes, orally, intramuscularly, or a combination thereof.
The dosage of the compounds employed varies depending upon age, individual di~elGllces, ~yll~ s, mode of ~11mini~tr~tion, etc., and can be readily determinPcl by one of skill in the art. Exemplary dosages of GM-CSF and IL-3 are given in J. N~
"Granulocyte-macrophage-colony-stim~ ting factor: a review from preclinical development to clinical application," Iransfusion, ~, 70-83 (1993); Lieschke et al., "Granulocyte Colony-Stimlll~ting Factor and Granulocyte-Macrophage Colony-Stim~ ting Factor (First of Two Parts)," The N. Fn,~. J. of Med.. 327, 28-35 (1992); Lieschke et al., "Granulocyte Colony-Stimtll~ting Factor and Granulocyte-Macrophage Colony-Stimlllsltin~
Factor (Second of Two Parts)," Th~ N. Eng. J. of Med., ;~Z, 99-106 (1992); Schulz et al., "Adjuvant Therapy with Recombinant Interleukin-3 and Granulocyte-Macrophage Colony-Stim~ ting Factor," ph~rm~(~ Ther., ~, 85-94 (1992); Hoelzer et al., "Interleukin 3 Alone and in Combination with GM-CSF in the Treatment of Patients with Neoplastic Disease," Semin~rs in Hematology~ ~, suppl. 2 (April), 17-24 (1991).
NFUTRAT.~7,~NG lVlF,THODS OF USING THE ~OMPOSITIONS
Another ob~ect of this invention is methods of neutralizing the activity of GM-CSF, IL-3, and IFN-~ under conditions where the production of antibody is pathogenic. such as in autoimmune disorders, such as lupus, systemic lupus erythematosus ~SLE), idiopathic thrombocytopenic purpura (ITP), vasculitis, Graves' disease, allergic reactions, etc.

W O 97/20940 PCTrUS96/19327 For the construction of a neutrali_ing vaccine, an antibody against the cytokine is made. Those of oldi~ skill in the art are familiar with the well established methods of obtaining specific antibody. The antibody is then ~1mini~t~red to a patient. The antibody can be bound to a carrier to increase the half-life of the antibody.
METHODS OF USING THE COMPOSITIONS TO
OPTJMT7,F. MONQCLONAL ANTTRODY PROI~UCTION
The compositions can also be used to U~llillli;C~ monoclonal antibody production in vitro or in vivo. For example, an animal can be s~.n~iti7~1 by in~ecting a solution ~;v~ g the antigen of interest and the compositions in vivo. Because the compositions stim~ tP
the release of antibody by B cells, the zl-lmini~tration of the compositions in conjunction with the specific antigen will optimize the production of antibody against the specific antigen.
Those of ordinary skill in the art would be familiar with techniques for such immlmi7~tion, as well as the dosages of the antigen and compositions needed to elicit the antibody and to stimul~t~ its release in light of the te~hings in this specification.
All~lllali~/t;ly, in vitro stimlll~tion or s~ 1 ion of Iymphocytes to produceantibody can be ~nh~nce~ in the presence of the novel compositions. The te~hing~ for such stimulation and sen.~iti7~tion are well within the routine skill of those in the art, as is the deterrnin~tion of the a~lo~,liate amounts of antigen and compositions to apply in light of the te~ching.c in this specification. This method is particularly useful for the production of human monoclonal antibodies.
ASSAY SYSTEM FOR IDENTIFYING COMPOSITIONS USEFUL
FOR STIMUT,~TING T~F, RELEASE OF ANTIBODY BY B C~.T.T,.~
A further embodiment of the invention is the development of a novel in vitro assay system that mimics in vivo antibody stim~ tion~ The assay system allows the i(l~ntific~tion of compositions that stim~ te the release of antibodies by B cells.
Because resting B cells do not release antibody, they must first be activated before antibody release can be measured. In the present invention, anti-IgM or ant;-IgD antibody is covalently conjugated to high molecular weight dextran (i.e., MW = 2.0 x 106) to create a multivalent antigen on a polysaccharide carrier, as set forth in Snapper et al.? "Cul~lpa,a~ e W O 97/20940 PCT~US96/19327 In Vitro Analysis of Proliferation, Ig Secretion, and Ig Class Switching by Murine Marginal Zone and Follicular B Cells," J. of Tmmunology~ 150, 2737-2745 (1993), specifically incorporated by reference. This procedure converts the bivalent anti-Ig molecule into an extremely stimulatory multivalent conjugate which can induce persistent and repetitive ~ign~ling via B cell membrane Ig, even at picomolar concentrations. The anti-Ig-dextran conjugate stim~ tec high levels of B cell proliferation at concentrations as low as 1 pg/m1, a concentration 10,0~0 fold lower than that stiml]l~t~d by unconjugated anti-Ig. The activation of B cells occurs irrespective of antigen specificity. Anti-Ig-dextran does not stim~ te the release of antibody by resting B cells in the absence of cytokines. With the addition of a cytokine that sfimlll~t~c the release of antibody by B cells, high levels of Ig secretion are observed.
Previously, it was known that activation of B cells through the antigen receptor can be optimally achieved using an anti-imml-n~ globulin (Ig) antibody conjugated to a large molecular weight polysaccharide, e.g., dextran. However, it was not possible to screen for compositions that stiml-lz3t~ the release of antibody by B cells because, previously, the B cells employed in the assays were not highly purified. Cont~min~ting cells in the B cell ~,u~ IL often secreted sufficient amounts of cytokines that could stim~ t~ the release of antibody by B cells. Therefore, it was not possible to determine whether the stim~ tQry effect of a tested composition was due to the added composition or to a Cont~min~nt.
In contrast, the present invention employs highly purified B cells in the assay system. Preferably, the B cells are electronically sorted to produce highly purified B cells.
With the use of highly purified B cells, it is possible to measure the stim~ t- ry effects of sllbst~r~es on t~he re1ease of antibody by B cells activated through multivalent mIg cross-linking. This system was not described prior to the invention. In fact, it was not possible to test compositions for the sole characteristic of stim~ tQry activity on the release of antibody by B cells prior to the discovery of this assay system.
* * * * *
The unexpected effect of the present invention is demonstrated in the following experiments and is depicted in Figures 1-12.

W O 97/20940 PCTnJS96/19327 Having generally described the invention, a more complete underst~n-1in~ can be obtained by reference to specific examples, which are provided herein for purposes of illustration only and are not int~?n~ to be limitin~.
~,Y~ntple 1 This example shows that the B cell antibody stim~ tc ry activity of the cell rn~ts.nt reported in the parent application was not due to one known cytokine.
Materials and Methods: Female DBA/2 mice were obtained from the National Cancer Institute (Frederick, MD) and were used at 7-10 weeks of age. Culture medium used was RPMI 1640 (Biofluids, Rockville, MD) supplemented with 10% fetal bovineserum (Sigma, St. Louis, MO), L-glutamine (2 mM), 2-mercaptoethanol (0.05 mM), penicillin (50 ,ug/ml) and streptomycin (50 ,ug/ml).
Dextran-conjugated anti-IgD antibody was ~e~t,d by conjugation of Ho'l/l (monoclonal mouse IgG2b (b allotype)) anti-mouse IgD (a allotype) to a high molecular weight dextran (2 x 1 o6 MW). Approximately 9 dextran-conjugated anti-IgD antibodies were conjugated to each dextran molecule. FITC-anti-CD3~ mAb (2C11) was ~ ,hasedfrom Ph~rrningen (San Diego, CA).
PE-labeled affinity-purified polyclonal goat anti-mouse IgM antibody was purchased from Southern Biotechnology Associates (Birmingh~nn, AL). Murine recombinant IL-l and IL-2 were kind gifts from Dr. Stephanie Vogel (USUHS. Bethesda, MD) and Dr. Maurice Gately (EIoffman-La Roche, Nutley, NJ), respectively. Recombinant murine IL-3 and GM-CSF were purchased from Ph~rrningen. Polyclonal goat anti-mouse IL-3 and GM-CSF antisera were both obtained from R & D Systems (Minneapolis, MN).
Functional assays were carried out in 96-well flat-bottom Costar plates (Costar,Cambridge, MA). Cultured cells were incuh~t~o~l at 1 x 105 cells/ml in a total volume of 200 ~LL at 37OC in a hllmi~lified atmosphere cont~inin~ 6% CO2.
Polyclonal IgM concentrations were measured by ELISA. Qn~ntit~tion was achieved by comparison with IgM standard curves employed in every assay.
Preparation and culture of B cells: Enriched populations of B cells were obtained from spleen cells. T cells were eliminz~ted by treatment with rat anti-Thy-l, anti-CD4, and anti-CD8 monoclonal antibodies, followed by monoclonal mouse anti-rat Igk and W O 97/20940 PCT~US96/19327 complement. Cells were fractionated on the basis of their density over discontinuous Percoll gradients (Pharmacia, Pi~c~law~, NJ) consisting of 70, 65, 60, and 50% Percoll solutions (with densities of 1.086, 1.081, 1.074, and 1.062 g/ml, respectively). The high density (small, resting) cells were collected from the 70 to 65% interface and cnn~i~ted of ~90% B cells. Highly purified B cells were then obtained by electronic cell sorting of membrane (m)IgM+CD3- cells on an EPICS Elite cytometer (Coulter Corp., Hialeah, FL) after staining with FITC-anti-CD3~ + PE-anti-IgM antibodies. Sorted cells were ;m m~ tf~1Y re-analyzed and found to be con~ict.ontly >99% B cells. These cells were used in all experiments.
TH1 mouse cell lines were cloned and used for the e~ç~ nt~ As described in the parent application, a composition which stim~ tecl the release of antibody by B cells was isolated. To determine the identity of the components of the composition, the activity of all known cytokines was system~tic~l~y ~u~ d by adding monoclonal antibodies specific for IL- 1, IL-2, IL-3, etc., to the composition, and then measuring activity.
Because the composition compri~ed two independent B cell antibody release stim~ tQrs, IL-3 and GM-CSF, ~x~G-h~lents ~u~plessillg each known cytokine one by one did not result in completely suppressing the measured activity.
F,Y~nlPIe2 This example ~l-ot~rmines the relative stim~ tory effects of a supçrn~f~nt of a T cell culture i(1entified in the parent application, GM-CSF, and IL-3.
IgM secretion (ng/ml) by B cells in various mediums was measured. Culture mediums are listed along the left side Fig. l. Culture mediums were anti-IgD-dextran and IL-l + IL-2, anti-IgD-dextran and IL-4; soluble CD40 ligand and IL-l + IL-2; soluble CD40 ligand and IL-4, membrane CD40 ligand, membrane CD40 ligand and IL-l + IL-2, and membrane CD40 and IL-4. The compositions tested were a s~ i of a cell culture described in the parent application, RA5-SN, GM-CSF, and IL-3. A controlcomposition was also tested for each medium.
As set forth in Fig. 1, compositions tested showed the highest IgM secretion in medium comprising dextran-conjugated anti-IgD antibody, IL-1, and IL-2. The RA5-SN

CA 02239868 l998-06-08 W O 97/20940 PCTrUS96/19327 supern~t~nt, having both GM-CSF and IL-3, showed significant IgM secretion by B cells as co~ ucd to other tested compositions.
F,Y~n~ e 3 This example shows that IL-3 and GM-CSF stim~ te Ig secretion by activated B cells.
B cells were activated with dextran-conjugated anti-IgD antibodies (3 ng/ml) plus IL-1 (150 U/ml) + IL-2 ~150 U/ml) in the presence or absence of varying concentrations of IL-3 or GM-CSF. IgM secretion was measured by ELISA 6 days after initiation of culture.
Electronically sort-purified resting B cells prolif~ r;~t~d but failed to secrete Ig in response to activation by dextran-conjugated anti-IgD antibody or dextran-con3ugated anti-IgD antibody plus IL-1 + IL-2. As reported in the parent application, sUpPrn~t~nt~
obtained from anti-CD3-activated CD4+ TH1 and TH2 clones induce strong Ig secretory responses by B cells stim-ll~tf d with dextran-conjugated anti-IgD antibody plus IL-1 +
IL-2. This stimlll~ticln is IL-4 and IL-5-independent.
Because TH1 and TH2 clones share the capacity to secrete IL-3, GM-CSF, and TNF-a upon activation, these cytokines were tested for their ability to induce Ig secretion.
Cytokines were titered in log increments from 0.1 to 1000 U/ml final concentration.
TNF-a had no effect on Ig synthesis. In contrast, IL-3 and GM-CSF stimlll~ted cignific~nt enhanct-mf?nt~ in Ig secretion over that observed with dextran-conjugated anti-IgD antibody - plus IL-l + IL-2 alone (Fig. 2). Both IL-3 and GM-CSF ~tim~ t~ optimal Ig secretory responses at 100 U/ml, producing a 19- and 9-fold enhancement, respectively. Significant, though lower, }evels of induction of Ig synthesis in response to IL-3 or GM-CSF were still observed at 1-10 U/ml.
In multiple ~ hllents the degree of çnh~nrement in Ig secretion mediated by IL-3and GM-CSF typically ranged between 10-50 fold (Fig. 3). No significant dirr~ ces were observed between IL-3 and GM-CSF for either dose response or optimal level of induction of Ig secretion.

W O 97/20940 PCTrUS96/19327 F,Y~n~VIe4 This example demonstrates that anti-IL-3 and anti-GM-CSF antibodies specificallyinhibit the in(1~ tif)n of Ig secretion in response to IL-3 and GM-CSF, lG.,~e~;~ively.
B cells were activated with dextran-coniugated anti-IgD antibodies (3 ng/ml) plus IL-1 (150 U/ml) + IL-2 (150 U/ml) in the presence or absence of IL-3 (10 U/ml) or GM-CSF (10 U/ml). In addition, anti-IL-3 (10 llg/ml) andlor anti-GM-CSF (10 ~lg/ml) antibodies were added. A control with no added anti-IL-3 or anti-GM-CSF was also~r~aled. IgM concentrations were measured by ELISA 6 days after initiation of culture.
Neutralizing anti-IL-3 and anti-GM-CSF antibodies specifically and completely inhibited the respective Ig-inducing activities of IL-3 and GM-CSF (Fig. 4). These antibodies, either alone or in combination, had no effect on the Ig secretory response to dextran-conjugated anti-IgD antibody + IL-5 activation (data not shown). Results also demonstrated that the combination of anti-IL-3 and anti-GM-CSF antibodies could significantly reduce the IL-4 + IL-5-independent Ig-intl~lcin~ activities of supernslt~nt~ from either a CD4+ TH1 or TH2 clone. It was also demonstrated that IL-3 and GM-CSF could not induce Ig secretion by B cel}s activated through the CD40-mP~ t. c~ ~iqn:~lling pathway.
li'Ys-~ple 5 This example demonstrates that IL-3 and GM-CSF induce Ig secretion by B cells activated with dextran-conjugated, but not unconjugated, anti-IgO antibodies. This example also shows that IL-3 and GM-CSF act synergistically with IL- 1 + IL-2.
Multivalent, but not bivalent, mIg cross-linkage co-stim~ t~ cytokine-mediated Ig secretion and class ~wi~chillg. In this example, dextran-conjugated anti-IgD (Ho~/1 mAb) was compared with unconjugated anti-IgD (Hoa/l mAb) for co-stimlll~tion of Ig secretion in response to IL-3 or GM-CSF.
B cells were first cultured with AF7 sup~ L, ccIL-4, and aIL-5. For AF7 supern~t~nt KLH-specific, lab-restricted, CD4+ T cell clone, derived from C57BLl6 mice, was established and m~inf~ined by standard methodologies. AF7 is a TH2 clone on the basis of its production of IL-4 and lack of production of IL-2 and IFN- y.
Cytokine-co~ .g supern~t~nt was obtained from cultures of AF7 cells in the following marmer: Tissue culture wells were incubated with anti-CD3 mAb (2C11) at W O 97/20940 PCTfUS96/19327 1 0,ug/ml in PBS for 3 hr at 370C and then washed 3x in fresh PBS. AF7 cells, which were allowed to return to their resting state, 7 days after sfim~ tion with antigen, APCs, and IL-2, were added to anti-CD3-coated plates at 1 x 106 cells/ml for various times, upon which cell-free supern~t~nt.c were obtained and either stored at -200C or 40C. In the latter case, supern:~t~nt was used in cellular assays within 1-2 weeks of having been hal ~e~led.
B cells were then cultured with dextran conjugated (Ho~/l-dex, 3 ng/ml) or unconjugated (Hoa/1,30 ~Lg/ml) anti-IgD antibodies in the presence or absence of IL-l (150 U/ml) + IL-2(15~ U/ml). A control composition did not contain conjugated orunconjugated dextran. IL-3 (100 ml) and/or GM-CSF (100 U/ml) were added, and control compositions with no addition of IL-3 or GM-CSF were prepared (Fig. 5). IgM
concentrations were measured by ELISA 6 days after initiation of culture.
In contrast to conjugated dextran (H8a/1-dex), unconjugated dextran (Ho~/l) was ineffective as a co-stimulator of Ig secretion in the presence of IL-3 or GM-CSF (Table 1).
These results were obtained despite the ability of unconjugated dextran to induce early B cell activation events, such as increases in cell size and MHC class II induction. IL-3 or GM-CSF alone were unable to induce resting B cells to secrete detectable Ig (Table 1).
However, addition of IL-3 and GM-CSF to cultures of dextran-conjugated anti-IgD
antibody-activated cells, in the absence of exogenous IL-l + IL-2, led to 7.4- and 5.4-fold inductions, respectively, of Ig secretion relative to stim~ tion by dextran-conjugated anti-IgD antibody alone. The addition of IL-I + IL-2 to cells activated witn dextran-conjugated anti-IgD antibody alone did not further increase Ig secretion.
The combination of IL-3 or GM-CSF with IL-l + IL-2 led to 4g-fold and 75-fold enh~n~ements in Ig secretion relative to that observed for B cell cultures activated with dextran-conjugated anti-IgD antibody plus IL-l + IL-2 alone. Thus, IL-l + IL-2is strongly synergistic with IL-3 or GM-CSF for induction of Ig secretion by dextran-conjugated anti-IgD antibody-activated resting B cells.

W O 97/20940 PCTrUS96/19327 T~RT,h' I

M~flillm Level of Stimulation Ho~ dex 65 IL-3 <24 GM-CSF <24 Ho~/l-dex + IL-3 480 Hon/1-dex + GM-CSF 350 Ho~/l-dex + IL-l + IL-2 57 Ho~/l-dex + IL-l + IL-2 + IL-3 2750 Ho~/1-dex + IL-l + IL-2 + GM-CSF 4250 Ho~/-l <24 Ho"/-l + IL-l + IL-2 <24 H~a/-l + IL-l + IL-2 + IL-3 <24 Ho~/-l + IL-1 + IL-2 + GM-CSF <24 FY~n~P1eG
This example demonstrates that both IL-3 and GM-CSF act primarily through stim~ tion of B cell differentiation to Ig secretion.
To determine the m.-~h~ . " by which IL-3 and GM-CSF ~h~n~e Ig secretion, the effects of IL-3 and GM-CSF on cell oul~lowlh as colllpal~d to Ig secretion were tl~t~rmin~l B cells were activated with dextran-conJugated anti-IgD antibodies (3 ng/ml) plus IL-l (150 U/ml) + IL-2 (150 U/ml) in the presence or absence of IL-3 (100 U/ml) and/or GM-CSF (100 U/ml). Viable cells (those that excluded trypan blue) were enumerated 4 days after initiation of culture using a hemocytometer. IgM concentrations in replicate cultures were measured by ELISA 6 days after initiation of culture. Data is represented as mean +/- standard error of the mean of duplicate cultures.

WO 97nog40 PCT~US96/19327 IL-3 stim~ t~d a significant but less than 2-fold enhancement in viable cell outgrowth, whereas GM-CSF had no significant effect on the degree of cell expansion (Fig. 6). Studies ~ ec~ing DNA synthesis on the basis of 3H-thymidine incorporation are c~-n~i~tent with these observations.
The induction of Ig secretion of replicate cultures in response to IL-3 and GM-CSF
was 13.8-fold and 13.6-fold, respectively. These resu~ts inrliç~tç that the primary effect of IL-3 and GM-CSF is to stimlllz3te a strong increase in the average amount of Ig secreted per cell, and not enh~n~.t?ment in the total of number of cells present in culture. Thus, the data strongly suggest that IL-3 and GM-CSF are differentiation factors for resting B cells activated through multivalent antigen receptor cross-linking. The combined action of IL-3 and GM-CSF, lltili7.ing optimal doses of each cytokine, con~i~t~ntly led to greater than additive Ig secretory .G~onses (Fig. 6). These data in~lic~te that IL-3 and GM-CSF act synergistically. A similar degree of synergism was observed with the combined use of IL-3 and GM-CSF on cells activated with dextran-conjugated anti-IgD antibody in the absence of IL- l + IL-2 (data not shown).
.~ple 7 This example demonstrates the stim~ tQry effect on antibody secretion of B cellswhen CD40L is added to the B cell composition.
B cells were activated with dextran-conjugated anti-IgD antibodies (3 ng/ml or 0.3 ng/ml) plus IL-l (150 U/ml) + IL-2 (150 U/ml). IgM con.~Pntr~tions in replicate cultures were measured by ELISA for a control composition, for a composition with IL-3 +
GM-CSF, and for a composition with IL-3 + GM-CSF + CD40 ligand (CD40L) (Table II).
IL-3 and GM-CSF were present at 100 U/ml. CD40L was in the form of a recombinantsoluble CD8-CD40 ligand fusion protein which was added at a final concentration of 10 ~Lg/ml. The fusion protein was a kind gift of Dr. M. Kehry, Boehringer Ingelheim ph~ ceuticals, Inc., Ridgefield, CT.

W O97/20g40 PCT~US96/19327 TAB~E II
~1~ Secretion Cn~/ml) IL-l + IL-2 Medium IL-3 + GM-CSF IL-3 + GM-CSF + CD40L
+ medium <12 <12 115 + anti-IgD-dex1,060 21,250 61,250 (3 ng/ml) + anti-IgD-dex37 3,575 42,125 (0.3 ng/ml) The most significant results (61,250 ng/ml) were obtained with anti-IgD-dex at 3 ng/ml and the composition con~i~tin~ of IL-3 + GM-CSF + CD40L. However, even at low }evels of the multivalent activator (anti-IgD-dex at 0.3 ng/ml), significant IgM
secretion was obtained (42,125 ng/ml). Thus, even at low levels of conjugate, extremely potent antibody release responses can be obtained by including CD40L on a carrier.
~ple 8 This example demonstrates the stim~ tory effect on antibody secretion of B cellswhen IFN- y is added 24 hours after stimulation of the B cells.
B cells were activated with dextran-conjugated anti-IgD antibodies (3 ng/ml) plus IL-l (150 U/ml) + IL-2 (150 U/ml) in the presence or absence of IL-3 (100 U/ml) +
GM-CSF (100 U/ml). IgM concentrations in replicate cultures were measured by ELISA
for a control composition, a composition to which IFN-~ was added at 0 h, and for a composition to which IFN-~ was added at 24 h. Data is leplesel~led as mean +/- standard error of the mean of duplicate cultures (Figure 7).
The composition consisting of anti-IgD-dex-activated B cells and IL-I + IL-2 showed minim~l IgM secretion with no added INF-~ (885 ng/ml) and minim~l secretion with IFN-~ added at 0 hours (975 ng/ml). When IFN-~ was added 24 hours after stimulation of the B cells, antibody secretion jumped to 11,250 ng/ml.
Even more significant results were obtained with a composition consisting of anti-IgD-dex-activated B cells, IL-l + IL-2, and IL-3 + GM-CSF. In medium without W O 97/20940 PCTnJS96/19327 IFN-y, antibody secretion was measured at 21,250 ng/ml. Surprisingly, the addition of IFN-y at 0 hours ~pl~ssed antibody secretion to 7,250 ng/ml. However, with the addition of IFN-~ 24 hours after stim~ tion of the B cells, a synergistic reaction occurred, producing antibody secretion of almost 200,000 ng/ml.
F.Y~nQFIe9 This example provides exemplary vaccines and vaccine adjuv~lL~ employing the compositions of the invention.
Table III shows exemplary vaccines employing the compositions of the invention.
As noted in the Table, several of the vaccines are conjugate vaccines. Methods of conjugation are well known to those of ordinary skill in the art, and include the heteroligation techniques of Brunswick et al., J. Tmmlm--l., 140:3364 (1988); Wong, S.S., Chemi~try of Protein Co~uFates and Cro~linkin~ CRC Press, Boston (1991); and Brenkeley et al., "Brief Survey of Methods for Preparing Protein Conjugates With Dyes, Haptens and Cross-T.inkin~ Agents,'~ Bioco33Jugate Ch~mi.ctry. 3, No. 1 (Jan. 1992), specifically incorporated by lererc;llce.
The multiva~ent carrier replaces the multivalent conjugate ~i.e., anti-IgD-dex) employed in vitro. Such a carrier can be, for example, a poly~r.çh~ri~l~ such as dextran, or a capsular polysaccharide from a bacteria, such as pneumococci, streptococci, ormeningococci. The polysaccharide may or may not be medically relevant depending on the use envisioned for the vaccine.
The cytokine can be GM-CSF, IL-3, IFN-~, or a combination thereof.
This example is not int(~:n~le~ to be limitin~, and other types of vaccines will be L to those skilled in the art from consideration of the specification and practice of the invention.

- = . =

W O 97/20940 PCTrJS96/19327 TABLE III
Vaccine Structure and Components 1. cytokine~ + antigen (admixed); co~-lmini~ation in aqueous solution, slow release particles, adjuv~ll4 etc.

2. cytokine + antigen + multivalent carrier; the cytokine, antigen, or both can be directly conjugated to the carrier, i.e.:
cytokine cytokine antigen antigen l l or I + antigen; or I + cytokine multivalent carrier 3. cytokine-antigen; direct conjugation via covalent bonding 4. cytokine-antigen (covalent bonding) bound to a multivalent carrier, i.e.:
antigen I

cytokine multivalent carrier 5. cytokine-antigen (fusion protein) 6. cytokine-antigen (fusion protein) bound to a multivalent carrier 7. peptide-cytokine (fusion protein) + antigen 8. peptide-cytokine (fusion protein) + antigen, with the fusion protein, antigen, or both bound to a multivalent carrier 9. cytokine-multivalent carrier, i.e.:

I

or multivalent multivalent carrier carrier 10. antibody complex (i.e., IL-3 + anti-IL-3) + antigen CA 02239868 l998-06-08 W O 97/2~940 PCT~US96/19327 11. antibody complex (i.e., IL-3 + anti-IL-3) + antigen + multiva}ent carrier (the cytokine, antigen, or both can be conjugated to the carrier) 12. anti-cytokine antibody; this is a neukalizing vaccine 13. anti-cytokine antibody + multivalent carrier; this is a neutralizing vaccine14. Vaccine examples 1 through 13 can be further modified by the addition of CD40L, either ~rlmixe(l or bound to the multivalent carrier 15. Vaccine examples 1 through 14 can be further modified by the addition of one or more cytokines other than GM-CSF, IL-3, or INF- y, such as IL- 1 + IL-2, either admixed or bound to the multivalent carrier The cytokine can be GM-CSF, IL-3, IFN- y, or a combination thereof.
* * * * * *
In sum, the ~ ,hllental results show that IL-3 and/or GM-CSF stim~ tt? the release of antibody by B cells and that, in combination, these cytokines act synergistically to stim~ tt~ the re}ease of antibody by B cells. This in~luction is specifically inhibited by anti-IL-3 and anti-GM-CSF antibodies, le~e.;~;vely. The stimulatory effect of IFN-r added 24 hours after culture is also shown.
The effects of IL-3 and GM-CSF are enh~n~eA by multivalent antigen lc;c~L ~r cross-linkage, as mediated by dexkan-conjugated anti-Ig antibody in the experimental model. The effects of IL-3 and GM-CSF may also be f nh~nce~l with antigen that does not induce high levels of membrane Ig cross-linking. Although both dextran-conjugated anti-Ig antibody and IL-1 + IL-2 are required for optimal IL-3- and GM-CSF-me~ tf -l Ig secretion, both IL-3 and GM-CSF stim~ te a modest Ig secretory response by cellsactivated with dextran-conjugated anti-Ig antibody alone. This is the first report demon~L.alillg the ability of IL-3 and GM-CSF to act directly as differentiation factors for B cells.
In addition, while msn~im~l antibody secretion is obtained with the use of GM-CSF, IL-3, IL-l + IL-3, and IFN-y added 24 hours after culture, a modest increase in antibody secretion was also observed with the sole addition of IFN-~ 24 hours after culture.

W O 97/20940 PCT~US96/19327 FY~IelO
This exarnple shows the prc;~Lion of cytokine-protein fusion products.
A. Plasmid Construction. The YEpFlag-l expression plasmid (Kodak Scientific Tmslging systems) was used for the cloning and expression of the fusion proteins.
The YEpFlag-l vector contains: an ADH2 promoter for regulated expression in yeast, tryptophan marker for selection in yeast, and oc factor for protein secretion in yeast. The vector also contains a multiple cloning site, the gene for ampicillin resistance and a Flag peptide sequence. The insertion of the fusion protein DNA withill the Kpnl and BamHl restriction sites removed the Flag peptide sequence and 14 bp's offof the o~ Factor peptide sequence. To f~cilitS~te purification of the fusion protein 6 hi.cti~lin~ residues were added to its C t~rmin~l end, thus enabling binding to and acid elution from a Ni-NTA resin. This method is superior to Flag dependent purification since it does not require subsequent enzymatic removal of the Flag peptide and provides for larger ~ ily purification. The 14 bp's removed from the 3' end of the Alpha Factor extracellular secretion peptide sequence were replaced by PCR.
Prior to con~ ;Lion of the fusion proteins, individual proteins were first s~ dl~;ly cloned. Protein D, pneurnococcal surface protein A (protein A) and pneumolysin were cloned according to the methods described in Son et al.. Infect. and Tmmlm, Vol. 63, No. 2, 696-699 (Feb. 1995); T .~n~ermann et al.. J. Exp. Med., Vol. 180, pp 2277-2286 (Dec.
1994); and Walker et al., Infect. and Tmmnn, Vol. 55, No. 5, pp 1184- 1189 (May 1987).
The fusion proteins were constructed with the cytokine on the N 1~,. . ";. ,..~i end linked by 6 glycine residues to the antigenic protein co..~ g a 6 hi~titline tail on the C t~ ....i..~i end.
B. IL2, GM-CSF Plasmid Construction. Human IL-2 or murine GM-CSF
cDNA cont~ining plasmids were used as a template for PCR. The 5' primer contained a Kpnl restriction site and the removed alpha factor bp's, and the first 18 b~es for the N-termini of the protein. The 3' primer contained the l~t 18 bases for the C-termini of the protein, three glycine codons between the cytokine sequence and the Smal restriction site as a linker. The PCR products and the YEpFlag plasmid were digested with Kpnl and Smal restriction endonucleases and ligated. E. Coli DH lOB colllp~Lel,l cells were transformed W O 97/20940 PCT~US96/19327 with the ligation mixture and ampicillin resistant clones were selected and pl~mid DNA
was purified from several clones. The presence of IL-2 and GMCSF genes was (letermined by PCR. The sequences were checked by using the Applied BioSystems DNA sequencing system and the recombinant plasmids were identified as pYIL2 and pYGM.
C. Pneumococcal surface protein A (PA), Protein D, and Pneumolysin (Ply) plasmid construction. Genomic DNA from S. pneumonia Rx1 was used as the template for both PA and Ply. The plasmid (pHIC 348) with cloned gent for Tmmllnnglobulin D-Binding Protein (PD) of Haemophilus influenzae was used as a template for PCR to obtain PD-gene fragment with convenient restriction sites at the ends (Sma I and BarnHI). 3' primer contained 6 condons for His and stop-condon. The His tail at the C-end of protein serves for purification using Ni-NTA resin (Qiagen). PCR product was cloned between Sma I and Bam HI sites of p YEP FLAG-l expression vector (Kodak Scientific Tm~ing System), which provides ~ es~ion and secretion in yeast culture.
Plasmid, C~ g correct PD-gene copy was identified as p YPD. The gene for PA was truncated to 888 bp and codes for that part of PA which retains its B cell antigenic epitopes as well as epitopes for T-cell activation (obtained from Dr. Larry McDaniel, Univ. of Al~hzlm~). The PCR fr~ment contains a Smal-restriction site, 3 glycine codons as a linker between the c~ po~ of the fusion protein, PA sequence (Ply sequence), 6 hi~ti~line codons (His 6 sequence to be used in protein purification) a stop codon and BamHI
restriction site at the 3' termini. The PCR products were cloned into YEpFlagl between the Smal and BamHl sites. After transformation clones were selected by Am pr. The presence of the PA or Ply gene was cletecte~l by PCR. The plasmids were clçsi~n7/~cl pYPA and Ply, and the accuracy of cloned sequence was verified by sequence analysis on the Applied Biosystems Model 373A DNA sequencing system.
D. Fusion-Protein Plasmid Construction. The Smal/BamHI fragment c~nt:~inin~ the sequences for PA and PLY were cut from their respective plasmids and ligated with either pYIL2 or pYGM digested with Smal and BamHl. The new fusion protein e~ .,;,sion plasmids were sequenced and named pIL2PA (IL-2+PA), pIL2Ply (IL-2+PIy) pGMPA (GMCSF+PA), pGMPly (GMCSF+Ply).

E. Cloning of the single sequence for Pan DR epitope peptide ~PDREP)-"unive~sal" epitope for MHC class II-r~ ed T cells. One of the PDREP's, as described in the article of Alex~n(1~r et al.~ T,..,~ y, Vol 1, 751-761 (1994), has a structure aKFVAAWTLKAAa. In our constructs this PDREP, which forms part of fusion protein, has the same structure, but L-alanine residues at the ends replace D-alanine.
To produce ~gm.-nt~ of DNA cont~ining the sequence of PDREP, a mixture of the two complem~nt~ry synthetic oligonucleotides was heated to 90~ and slowly cooled down.
Where n~ce~ , 5' ends of the oligonucleotides were phosphorylated before ~nnto~ling Agarose-gel electrophoresis showed the presence of double-stranded fragment (~60 b.p.). Sense strand is 53b and anti-sense is 57b. At the S'ter~nini the fragment had cohesive end corresponding with the Kpnl restriction site, a sequence for ~c:~ol~lion of the o~ leader peptide codons which was excluded at the time of cloning, 39 b.p. for PDREP and blunt 31 termini. This fragment was ligated with PYPA digested with Kpnl and Smal.
DNA from ampicillin resistant clones is analyzed by PCR and by sequencing.
To produce a triple PDREP sequence DNA fr~gmPnt a double sequence PDREP
fr~grn~nt of 78 bases was produced by ~nnP~Iing the respective synthetic oligonucleotides.
After a few steps of ligation, we obtained a construct, cont~ining between the Kpnl and the BamHI sites, a full size o~ leader peptide sequence, three 39 b.p. sequences for PDREP and a PA truncated gene.
F. Yeast Transformation. Starter cultures of BJ 3505 Tryptophan(-) Yeast were cultured at 30O overnight from frozen stocks in 25 ml YPD media. 100 ml fresh YPD
media was inoculated from the overnight culture and incubated overnight to an O.D of .3-.5 at 2~0 nm. The yeast and fusion proteins plasmid were then prepared for electroporation according to published protocols. Electroporation was pelro,llled using the BioRad Gene pulser set at 1.5KeV, 200 ohms, 25 uF. Electroporated yeast was plated on selective media plates with transformed colonies appearing in 3-5 days.
G. Isolation of Fusion Protein E~pressing Yeast. Fusion protein secreting clones were selected by growing transformed colonies on ~ ,cs~ion plates. ~xpression plates were prepared by placing nitrocellulose and then a cellulose acetate filter on top of expression media agar plates. After colonies were grown on the plates for 3-5 days the W O 97/20940 PCT~US96/193Z7 nitrocellulose was removed and immunoblotted. Membranes were washed/blocked in 1 x TBS tween 20 (TBS20~ for 2 hrs and then in~ b~t~cl for 1 hr with the respective anti-protein antibody. Membranes were washed for 1 hr in lx TBS20 and subsequently incubated for 1 hr in rabbit anti-mouse IgG antibody and then washed for an additional hr in lx TBS20. Membranes were incubated for 1 hr in alkaline phosphatase labeled goat anti-rabbit lg, washed for 1 hr in TBS20 and then developed with BCIP. Colonies that were positive for protein secretion were selected and used to inoculate 25 ml of expansion media which was incubated overnight and then frozen in 60% glycerol.
H. Batch Fusion Protein Production. 25 ml starter cultures of transformed yeast were prepared from frozen stocks. 10 ml from the starter culture were added to 1 liter of expansion media, incubated at 30O C shaking at 240 revolutions/minute. Cultures were harvested 3 days after inoculation, S~ J ,1~ 1 was sterile filtered after collection by centrifugation at 2000 rev/min. The media was then concentrated 5 times and dialyzed with 10 times the volume of 2x PBS using an amicron stirred cell.
I. Purification. Fusion proteins Co~ irlg 6 residues of hi~if1in~ were purified by incubating the concentrated media with I ml. Ni-NTA resin (Qiagen~ per liter original culture volume for 2 hr on a rotating shaker. The media and Ni-NTA were added to a small 10 cm. column and the Ni-NTA was washed with I liter lx PBS. The protein was eluted using 2.5 M sodium acetate pH 4.5.
Fractions were collected and the O.D. was measured at 280 nm to cictt--rrnine relative protein concentrations, fractions with O.D. greater than 0.1 were pooled and dialyzed against lx PBS pH 7.4. Media samples from the original media, conce~ led media, and purified protein were electrophoresed on an SDS denaturing gel and then either stained with comassie blue or immunoblotted. Gels for immllnoblots were transferred to nitrocellulose using Pharmacia's NovaBlot system, and immlm~blotting was performed as described above. Controls for protein identification were media from yeast with pYEP
plasmid without fusion protein, and positive controls for IL-2 and PSP-A. Protein bands on gels stained with coomassie blue were compared to molecular weight markers to estim~t~
fusion protein size. Recombinant proteins without a histidine tag were isolated by ion exchange on an 5100HR column. Purities were assayed by SDS PAGF and W O 97/20940 PCT~US96/19327 immllnnblotting for each component of the fusion protein. Samples may also be e~min~cl by analytical HPLC on a Beckman Ultraspherigel 2000 gel filtration column. The final products were tested for cytokine activity.
J. Conjugation of protein or cyto~ine protein to polys:~c~h~ride.
Pnl4 (SmithKline or ATCC) is derivatized with amines using 1-cyano-4-dimethylaminopyridinium tetrafluoroborate (CDAP), as set forth above. In brief, 1 ml of Pnl4 (10 mg/ml in water) was activated by the addition of 30 ul of CDAP (100 mg/ml in acetonitrile), followed 30 seconds later by 30 ul of 0.2 M triethylamine (TEA). At 2 min, 0.5 ml of 0.5 M hç~c~n~ mine in 0.75 M HEPES, pH 7.5 was added. 1.5 hr later, the product is desalted on P6 cartridge (BioRad), equilibrated with saline, concentrated by ultrafiltration and df~lted again. The amine content was assayed by the method of Vi~l~l et ~, J. Imm. Methods, 1986, 86, 155 and the polysaccharide by the method of Mo}~ y et L, Anal. Chem. 1988, 175, 525.
Protein was made up at 5-20 mg/ml in 0.15 M HEPES, pH 7.5 and thiolated with a 20 fold molar excess of N-Succinimidyl S-acetylthioacetate (SATA, ProChem, Rockford, IL) from a 0.1 M stock in DMF. After 2 hr incubation in the dark at room temp, the sample was desalted using a P6 cartridge (BioRad), equilibrated with 10 mM NaAc. 0.1 M NaCl, 2 mM EDTA, pH 5 and concentrated to 10-20 mg/ml using a Centricon 30 device (Amicon).
The NH2-Pnl4 was iodoacetylated with 10 ul of 0.1 M Nhydroxys~lcçinimidyl io~ et~te (SIA, ProChem, Rockford. IL) per mg Pnl4 and after 2 hr ~lec~lt~d and concentrated as above.
The thiolated protein and iodoacetylated Ps was combined and the reaction begun by the addition of l/9 volume of 0.5 M hydroxylamine in 0.75 M HEPES, pH 7.5. After an overnight reaction, the reaction was quenched by making the solution 0.2 mM in mercaptoethanol for 1 hr, followed by a 10 min incubation with 10 mM iodo~-~et~mide.
The conjugate was passed over a S400~R gel filtration column, equilibrated with PBS.
The high molecular weight material was pooled and sterile filtered by passage through a Millex GV filter (Millipore). The conjugate was assayed for protein using the BioRad protein assay and for polysaccharide by the method of Monsigny et al., noted above.

W O 97/20940 PCT~US96/19327 K. Conjugation of "universal" T cell epitope to polys~crharide.
The universal T cell epitope peptides c- ntzlining a N t~rrnin~l cysteine were reduced with DTT, ~t s~lt~d on a G10 column, equilibrated with 10 rnM sodium acetate, pH 5 and reacted with iodoacetylated polysaccharide in HEPES buffer, pH 7.5. ~lt~rn~tively, a one step process may be used. Peptides were synth.oci7~d with an N t~rmin~l hydrazide and directly reacted with CDAP-activated polysaccharide at pH 5. By either method, after an overnight reaction, the conjugate may be dialyzed into saline and sterile filtered. The peptide was quantitated from its absorbance spectra and the polys~c-~h~ri(le assayed by the method of Monsigny et al.
L. Deterr ir~tion of enhancement of anti-protein and anti-polys ~crh~ride antibody responses Mice were injected s.c. with .01-10 ~Lg of protein, protein-cytokine or protein-cytokine-polys~c~h~ride~ and anti-protein and anti polysaccharide responses determined 14, 28 and 52 days later. Antibody responses were determined by ELISA. The t?nh~nred anti-protein and anti-polysaccharide responses obtained are shown in Table IV.

W O 97/20940 PCT~US96/19327 Table IV

INDUCTION OF ENHANCI~:D ANTI-PROTEIN AND ~NTI-POLYSACCHARIDE
RESPONSES WITH PspA-IL2-Pnl4 VACCINE CONST:RUCT

F.~ T.SA~ TITF~S
vaccin~ construct dose (llg) IgGI anti-proteinA I~GI anti-Pnl4 Protein A-IL2 5 >40,000 .5 10,721 ~05 ~10 Protein A-IL2-Pnl4 5 4~690 9,708 .5 24,495 >40,000 .05 17,845 >40,000 Pnl4 5 <10 <10 Protein A 5 <10 - - -M. Dete.~ .ation o~T Cell Proliferation in Vitro in Response to Vaccine Constructs Engineered with Universal T Cell Epitope and Conjugated Pneumococcal Pol~ Ir~ ride t~vpe 14.
T cells from draining Iymph nodes were purified from mice having received control or ~ clhllental vaccines. Cells were cultured at 2.5 x 104 - 2 x 10~/well with 5 x 104 milolllycill C treated spleen cells as a source of presenting cells. The results are shown in Table V.

WO 97~0940 PCTfUS96/19327 Table V

T-Cell proliferation in vitro in response to vaccine constructs en~si..re~ ~d with universal T cell epitope (TCE) Tmmnni7.inE~ Antigen Antigen in vitro ~0.5 ~ moll~P) (,u~/ml!

TCE ~ PT~/TCF
Thymidine incol~,o-~L~on (cpm) media 1,400 1,400 18,000 21,000 1 8,0001 5,000 0.10 13,000 7,500 0.01 7,000 3,500 Pr~-TC~
media 1,500 1,500 24,000 28,000 22,000 17,000 0. 10 12,000 9,300 0.01 6,000 4,000 Pn-TCE-PN14 media 2,000 1,000 7,000 14,000 5,000 7,000 0.10 3,000 5,000 0.01 3,000 3,000 Mice were immlmi7.-~l with universal T cell epitope (TCE), protein D-TC~, or protein D-TCE-pneumococcal polysaccharide type 14 in CFA. Ten days later draining lymph nodes were separated and cultured with the various stimuli. Thymidine incorporation was ~le.termin~.cl 4 days later. T cell proliferation was somewhat lower in the group jmmlmi7.
with PD-TCE-Pnl4 probably as a consequence of the chemical conjugation of PD-TCE to the polysaccharide, which may have affected a number of important epitopes.

Having now fully described the invention, it will be ~palell~ to one of ordinary skill in the art that many changes and modifications can be made thereto without departing from the spirit or scope of the invention as set forth herein. The appended claims are not int.-nrlPcl to be limiting.

Claims (13)

We claim:

1. A composition useful for stimulating the release of antibody by B cells comprising an effective amount of granulocyte-macrophage colony stimulating factor (GM-CSF), interleukin-3 (IL-3), or a combination thereof.

2. A conjugate vaccine comprising a) granulocyte-macrophage colony stimulating factor (GM-CSF), interleukin-3 (IL-3), or a combination thereof, wherein GM-CSF, IL-3, or a combination thereof are covalently bound to a multivalent carrier;
b) antigen of the vaccine, wherein the antigen is covalently bound to the multivalent carrier.

3. The vaccine of claim 2, wherein GM-CSF and IL-3 are fused together, or separately fused to another suitable protein, to form a fusion protein, which is then bound to the multivalent carrier.

4. The vaccine of claim 2, further comprising interferon-V (IFN-.gamma.).

5. The vaccine of claim 2, further comprising CD40 ligand or at least one cytokine other than GM-CSF, IL-3, or IFN-.gamma., or a combination thereof.

6. The vaccine of claim 5, wherein the at least one cytokine is a combination ofinterleukin-1 (IL-1) and interleukin-2 (IL-2).

7. The vaccine of claim 5, wherein the CD40 ligand or at least one cytokine other than GM-CSF, IL-3, or IFN-.gamma., or a combination thereof, are covalently bound to the multivalent carrier.

8. The vaccine of any one of claims 2, 3, 4, 5, 6, 7, or 8, wherein the multivalent carrier is selected from the group consisting of a bacterial capsular polysaccharide, dextran, and genetically engineered vectors.

9. The vaccine of claim 8, wherein the bacterial capsular polysaccharide is from pneumococci, streptococci, or meningococci.

10. A neutralizing vaccine adjuvant comprising one or more antibodies selected from the group consisting of:
a) an antibody against granulocyte-macrophage colony stimulating factor (GM-CSF);

b) an antibody against interleukin-3 (IL-3); and c) an antibody against interferon-.gamma. (IFN-.gamma.).

11. The adjuvant of claim 10, wherein at least one of the one or more antibodies is bound to a multivalent carrier.

12. The adjuvant of claim 11, wherein the multivalent carrier is selected from the group consisting of a bacterial capsular polysaccharide, dextran, and genetically engineered vectors.

13. The adjuvant of claim 12, wherein the bacterial capsular polysaccharide is from pneumococci, streptococci, or meningococci.