STATEMENT OF RIGHTS TO INVENTIONS MADE UNDER FEDERALLY SPONSORED RESEARCH

This work was performed in part with U.S. government funding under grants from the National Institute of Health (NIH) Nos. GM 4683 1 and T32 GM 08270 and from a grant from the University of Michigan Multipurpose Arthritis Center (NIH Grant AR 20557). Accordingly, the United States government may have rights in this invention.

TECHNICAL FIELD

This invention relates to novel antibodies which bind ssDNA and dsDNA.

BACKGROUND

Systemic Lupus Erythematosus ("SLE") is a human autoimmune disease mediated by pathogenic immune complexes. The immune complexes believed to play a key role in the renal pathology of systemic lupus erythematosus comprise anti-DNA auto-antibodies (hereinafter "anti-DNA") which localize within the basement membrane of the glomerulus (1-4). However, not all anti-DNA cause renal damage and establishing a relationship between the physical properties of these antibodies, their affinity and mechanisms of antigen binding, and pathogenesis remains a central goal of SLE research (47).

The presence of serum anti-DNA autoantibodies is one of the clinical features of SLE. The involvement of anti-double stranded DNA antibodies (anti-dsDNA) in the renal pathology of SLE has been inferred from the correlation of disease activity with the presence anti-dsDNA and the isolation of anti-dsDNA from renal eluates. However, a clear understanding of the series of events that lead from the production of anti-dsDNA to their site of tissue localization has not been elucidated. To this end, early efforts in this area were concentrated to defining the binding properties and physical characteristics of anti-DNA. Polyclonal sera obtained from

both patients afflicted with SLE and murine models of this disorder were employed (10, 11). A great deal of the early research effort was to elucidating the determinants involved in self recognition and the interactions that stabilize DNA*anti-DNA interactions (16). For a variety of reasons, the general features of dsDNA epitopes and the parameters that distinguish high and low affinity anti-dsDNA are not well defined. This point is particularly significant since anti-dsDNA levels seem to correlate with disease activity (3). Indirect evidence from comparative sequence alignments of anti- dsDNA (15), and competition binding experiments using chemically modified dsDNA duplexes (21) suggests that anti-dsDNA may recognize epitopes on the bases in both the major and minor groove, as well as on the phosphate backbone of dsDNA. Binding of dsDNA may also involve more electrostatic interactions than recognition of single- stranded DNA ("ssDNA") (22).

Thus, a need exists to further elucidate the pathological interaction of DNA with autoantibodies and to relate this interaction to inflammatory glomerulonephritis and SLE. This invention provides the reagents and methods to this end.

DISCLOSURE OF THE INVENTION

This invention provides an antibody with high affinity for single-stranded DNA, low or no affinity for double-stranded DNA, and capable of specifically binding a DNA hairpin. The antibody is either polyclonal or monoclonal. Also provided herein is a hybridoma cell line which produces the monoclonal antibodies.

A chimeric mouse comprising the hybridoma cell lines and a histocompatible mouse is further provided.

A method for screening for an agent which will inhibit anti-DNA antibody DNA binding is provided herein. The method requires providing an anti- DNA antibody bound to a solid support; contacting the agent to be tested with the receptor bound support under conditions favoring binding of antibody to DNA and contacting detectably-labeled DNA to the solid support of under conditions favoring binding of DNA to anti-DNA. The presence of any complex formed between the antibody and the DNA is then detected, the absence of complex being indicative that the agent inhibits binding of antibody to DNA.

Finally, a method of inhibiting the binding of an anti-DNA antibody to its DNA ligand in a sample by contacting the sample with an effective amount of a benzodiazepine derivative is encompassed by this invention.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows SDS-PAGE of fully purified anti-DNA. Samples (3-5 μg) were electrophoresed on 10% polyacrylamide-SDS gels and stained with commassie blue. The minor band above the heavy chain in lanes containing IgG2b anti-DNA is the result of asymmetric glycosylation of the heavy chain (38).

Figure 2 shows the sequence of model DNA ligands. The synthesis of these sequences has been described previously (26). Incorporation of disulfide crosslinks into these ligands increases their thermodynamic stability without disrupting the native geometry of the structure (26, 55). (.) represent hydrogen bonds.

Figure 3 shows cation release in anti-DNA complexes with dT ₂₁ . The apparent dissociation constant (__ _d ) was measured as a function of sodium ion [Na+] concentration as described in the experimental section below. The number of released cations was calculated by solving the equation ln[K^ - Zψln[Na+] + ln[__°] for Z using a value of 0.71 for ψ and the slope derived from a plot of ln[__ _d ] vs ln[Na ⁺ ] (58). Plots of In[__ _d ] vs ln[Na ⁺ ] are shown for 15D8 (Zψ = 0.04, r = 0.54) and 4B2 (Zψ = 0.79, r = 0.98) and are representative of anti-DNA that release zero (IgG2b mAbs and 11F8) or one cation (IgG2a mAbs), respectively, in complexes with dT ₂₁ . Similar results were obtained using dT ₅ with 9F11 and 11F8; however, binding by the IgG2a mAbs was too weak to measure so this analysis could not be performed.

Figures 4A and 4B are representative fluorescence quenching titration of anti- DNA with poly(dT). In Figure 4A, fluorescence titration was obtained for 9F11. The mAb (0.87 μM) was excited at 295 nm and the emission was monitored from 300-

450 nm as a function of DNA concentration (O-2 μM). The spectra shown are baseline corrected. All mAbs exhibit fluorescence quenching in the presence of poly(dT). Figure 4B is a calculation of the number of bases occluded upon binding. The percent fluorescence quenching was quantified by integrating the fluorescence spectra at each titration point relative to the free protein. The molar concentration of bases was determined using the extinction coefficient for phosphate (8.1 x 10 M ^"1 cm ^"1 ) as

described by Kim et al. (38). The initial and final slopes are determined from best fits to the first and last five data points and extrapolated to the point of intersection. The intersection of the initial and final slopes represent the number of nucleotides occluded in the binding site (36). Similar results are obtained for all mAbs studied when using a ten- fold lower protein concentration, suggesting that protein binding is noncooperative (63). Due to their relatively low affinity, mAbs 4B2, 7B3 and 10F4 do not yield a distinct saturation point. Based on the data for these antibodies, it is estimated that between four to six consecutive nucleotides are occluded upon binding.

Figure 5 is a representative gel shift binding isotherm. Each data point represents the average from three different binding titrations. These data were fit by non-linear least squares regression to a single-site binding isotherm and afford a K _d of 939 ± 81 nM (x ^{2 *} = 0.02).

Figures 6 A to 6C are potassium permanganate footprinting of the anti-DNA panel. Anti-DNA or normal mouse IgG (nlgG) (10 - 20 μM) was complexed to ligand (2) (0.5 nM), reacted with KMnO ₄ for 30 s, and treated as previously described (18).

Equivalent amounts of DNA (as judged by liquid scintillation counting of samples) were applied to the sequencing gel. Figure 6A is an autoradiography of cleavage products separated by denaturing gerelectrophoresis. Lanes 1 and 2, G+A and T sequencing reactions, respectively; lanes 3 and 4, KMnO ₄ modification in the absence (lane 3) or presence (lane 4) of nlgG; lanes 5-12, KMnO ₄ modification in the presence of mAbs

4B2 (lane 5), 7B3 (lane 6), 8D8 (lane 7), 10F4 (lane 8), 9F11 (lane 9), 15B10 (lane 10), 15D8 (lane 11), and 11F8 (lane 12). Figure 6B is representative densitometry traces comparing KMnO ₄ reactivity of DNA in the presence of nlgG and mAbs 4B2, 9F11 and 11F8. Figure 6C shows the difference probability maps calculated from the densitometry plots in B and presented in the same order. The difference in KMnO ₄ modification at each thymidine is represented by bars and is calculated by subtracting the logarithm of the probability of modification at each position in the free DNA (Fn) from the same position in the complex (P _n ) according to Rhodes (64). A larger negative value of ln(P _n ) - ln(F _n ) indicates greater protection from modification. Figures 7 A and 7B show the anti-DNA V _H chain nucleotide and deduced amino acid sequence. Sequences were aligned using the SAW software. Nomenclature and

numbering were according to Kabat. Sequences begin after the N _H primer at position 23 in codon 8. Identities are indicated by (,) and gaps are indicated by (-). The absence of an amino acid is indicated by a blank space.

Figures 8 A and 8B show the anti-DΝA N _κ chain nucleotide and deduced amino acid sequence. Sequences are displayed according to the legend of Figure 7. Sequences begin after the _κ primer at position 28 in codon 9.

Figure 9 is the putative consensus N _H and N _L nucleotide (Panel A and B, respectively) and amino acid sequences (Panel C) for the 9F11 group, 4B2, and 11F8. The sequences were derived using the SAW software by aligning homologous anti-DΝA whose sequences have been published (found in Table I) as well as several antibodies of other specificities obtained from a search of the GenBank database. The statistically most represented nucleotide at each position was defined as the consensus nucleotide at that position. In searching the GenBank database, care was taken to avoid using multiple sequences from the same source. The GenBank accession numbers for unrelated antibodies is as follows. For 9F11 N _H : S59838, X64998, Z29586, X00160, K00706; 9F11 _κ : Z22096, X69859, S53109, U00929, X01431; 4B2 N _H : U09593, L22535, X80958, X13188, J00507; 4B2 _κ : M18239, D14630, L09009, J00575, U01851; 11F8 N _H : K00723, S50914, S51594, X15471, M36228, L14742; 11F8 _κ : X16955, L18942, M34633, M31270, S64047. Sequences are displayed as described in the legend of Figure 7. The genealogy of the 9F11 group could not be determined due to the lack of sufficient parallel and unique somatic mutations that can be used to elucidate branchpoints in the evolution of these mAbs.

Figures 10A and 10B show V(D)J construction in CDR3. HCDR3 (Figure 10A) and LCDR3 (Figure 10B) are displayed according to Radic and Weigert. The parent gene fragment from which the functional CDR3 is derived is provided above the sequence. The boxed sequence represents the portion of the germline gene incorporated into the CDR3 construct. Positions that have mutated away from the putative germline gene are highlighted in boldface. Νontemplate-derived sequences are indicated by Ν. Figure 11 is map of phagemid pCAΝTAB 5E. The scF _v is constructed NH _r region (C-terminus) to the first framework region of V _L (Ν-terminus) using (Gly ₄ Ser) ₃ . The E-tag peptide (GAPNP YPDPLEPR) for affinity purification is located at the C- terminus (3 '-end) of the scF _v . The g3p signal sequence

(MKKLLFAIPLNNPFYAAQPA) directs transport into the periplasm. In suppressor strains of E. coli transformed with a recombinant vector, translation continues through the amber stop codon to produce scF _v -fusion protein that is packaged and displayed on the phage tip. In nonsuppressor strains, the amber stop codon is recognized and only the scF _v is produced and accumulates in the periplasmic space.

Figure 12 shows SDS-PAGE analysis after induction of E. coli HB2151 cells transformed with pCAΝTAB 5 E-4B2. Lane 1 shows whole cell extract. Lane 2 shows periplasmic extract TCA concentrated 20: 1. Lane 3 are the molecular weight markers. The band in Lane 2 migrates as expected for the 4B2 scF _v (28kDa) and is detected with an anti-E-Tag antibody in Western blots. 4B2 scF _v was not detected in pelleted cell debris or in the growth medium.

Figure 13 is a schematic of the synthesis of the 1,4 benzodiazepine derivatives useful to inhibit antibody*DΝA complex formation.

MODE(S) FOR CARRYING OUT THE INVENTION

Definitions: The terms "proteins", "peptides" and "polypeptides" are used interchangeably and are intended to include molecules containing amino acids linearly coupled through peptide bonds. The amino acids of can be in the L or D form so long as the biological activity of the polypeptide is maintained. These also include proteins which are post-translationally modified by reactions that include glycosylation, acetylation and phosphorylation. Such polypeptides also include analogs, alleles and allelic variants which can contain amino acid derivatives or non-amino acid moieties that do not affect the biological or functional activity of the protein as compared to wild-type or naturally occurring protein. The term amino acid refers both to the naturally occurring amino acids and their derivatives, such as TyrMe and PheCl, as well as other moieties characterized by the presence of both an available carboxyl group and an amine group. Non-amino acid moieties which can be contained in such polypeptides include, for example, amino acid mimicking structures. Mimicking structures are those structures which exhibit substantially the same spatial arrangement of functional groups as amino acids but do not necessarily have both the α-amino and α-carboxyl groups characteristic of amino acids.

"Muteins" are proteins or polypeptides which have minor changes in amino acid sequence caused, for example, site-specific mutagenesis or other manipulations; by errors in transcription or translation; or which are prepared synthetically by rational design. These minor alterations result in amino acid sequences wherein the biological activity of the protein or polypeptide is altered as compared to wild-type or naturally occurring polypeptide or protein.

As used herein, the term "peptide bond" or "peptide linkage" refers to an amide linkage between a carboxyl group of one amino acid and the α-amino group of another amino acid. As used herein, the term "hydrophobic" is intended to include those amino acids, amino acid derivatives, amino acid mimics and chemical moieties which are non- polar. Hydrophobic amino acids include Phe, Nal, Trp, He and Leu. As used herein, the term "positively charged amino acid" refers to those amino acids, amino acid derivatives, amino acid mimics and chemical moieties which are positively charged. Positively charged amino acids include, for example, Lys, Arg and His.

"Purified" when referring to an antibody, protein or polypeptide, are distinct from native or naturally occurring antibodies, proteins, polypeptides or antibodies because they exist in a purified state. These "purified" proteins or polypeptides, or any of the intended variations as described herein, shall mean that the compound or molecule is substantially free of contaminants normally associated with the molecular compound in its native or natural environment.

"Native" polypeptides, proteins, antibodies or nucleic acid molecules refer that those recovered from a source occurring in nature or "wild-type".

A "composition" is intended to mean a combination of active agent and another compound or composition, inert (for example, a detectable agent or solid support) or active, such as an adjuvant.

A "pharmaceutical composition" is intended to include the combination of an active agent with a carrier, inert or active, making the composition suitable for diagnostic or therapeutic use in vitro, in vivo or ex vivo. As used herein, the term "pharmaceutically acceptable carrier" encompasses any of the standard pharmaceutical carriers, such as a phosphate buffered saline solution, water, and emulsions, such as an oil/water or water/oil emulsion, and various types of

wetting agents. The compositions also can include stabilizers and preservatives. For examples of carriers, stabilizers and adjuvants, see Martin, Remington's Pharm. Sci.. 15th Ed. (Mack Publ. Co., Easton (1975)).

The term "nucleic acid" or "polynucleotide" means single and double stranded DNA, cDNA and RNA, as well as the positive and negative strand of the nucleic acid which are complements of each other, including anti-sense RNA. A "nucleic acid molecule" refers to a polymeric form of nucleotides of any length, either ribonucleotides or deoxyribonucleotides, or analogs thereof. An "analog" of DNA, RNA or a polynucleotide, refers to a macromolecule resembling naturally occurring polynucleotides in form and/or function (particularly in the ability to engage in sequence-specific hydrogen bonding to base pairs on a complementary polynucleotide sequence) but which differs from DNA or RNA in, for example, the possession of an unusual or non-natural base or an altered backbone. See for example, Uhlmann et al. (1990) Chemical Reviews 90:543-584. "Isolated" when referring to a nucleic acid molecule, means separated from other cellular components normally associated with native or wild-type DNA or RNA intracellularly or in serum.

An "antisense" copy of a particular polynucleotide refers to a complementary sequence that is capable of hydrogen bonding to the polynucleotide and can therefor, be capable of modulating expression of the polynucleotide. These may be DNA, RNA or analogs thereof, including analogs having altered backbones, as described above. The polynucleotide to which the antisense copy binds may be in singe-stranded form or in double-stranded form.

As used herein, the term "operatively linked" means that the DNA molecule is positioned relative to the necessary regulation sequences, e.g., a promoter or enhancer, such that a promoter will direct transcription of RNA off the DNA molecule in a stable or transient manner.

"Vector" means a self-replicating nucleic acid molecule that transfers an inserted nucleic acid molecule into and/or between host cells. The term is intended to include vectors that function primarily for insertion of a nucleic acid molecule into a cell, replication vectors that function primarily for the replication of nucleic acid and

expression vectors that function for transcription and/or translation of the DNA or RNA. Also intended are vectors that provide more than one of the above functions.

"Host cell" is intended to include any individual cell or cell culture which can be or have been recipients for vectors or the incorporation of nucleic acid molecules and/or proteins. It also is intended to include progeny of a single cell, and the progeny may not necessarily be completely identical (in morphology or in genomic or total DNA complement) to the original parent cell due to natural, accidental, or deliberate mutation.

An "antibody" is an immunoglobulin molecule capable of binding an antigen. As used herein, the term encompasses not only intact immunoglobulin molecules, but also anti-idiotypic antibodies, mutants, fragments, fusion proteins, humanized proteins and modifications of the immunoglobulin molecule that comprise an antigen recognition site of the required specificity.

An "antibody complex" is the combination of antibody (as defined above) and its binding partner or ligand.

Significant advances to understanding the involvement of anti-DNA in SLE have been made in the last decade. However, the factors involved in initiating and sustaining the anti-DNA response remain elusive. Analysis of the N-region genes of monoclonal antibodies derived from lupus-prone mice has revealed that anti-DΝA are oligoclonal and exhibit patterns of somatic mutation to residues like arginine and asparagine that can interact with and improve affinity to DΝA (13-15). These results are consistent with an autoantigen driven response in which stimulation by DΝA or a DΝA-containing complex results in affinity maturation toward DΝA antigens. Although the specific molecule(s) that drives this response in vivo is still unknown, free DΝA and protein-DΝA complexes such as nucleosomal particles are believed to be likely candidates (41). This hypothesis, however, is not consistent with the observation that dsDΝA is not immunogenic (71) which in turn has led to the proposal that in SLE, tolerance defects exist that permit autoreactive B cells to escape the mechanisms normally in place to anergize or eliminate them (72). In support of this latter hypothesis, differential expression of regulatory genes in mice (e.g., Fas/Apol and bcl2) has been implicated in the etiology of autoimmunity by facilitating the survival of autoreactive T and B cells (73, 74). While identifying the defects that lead to this

breakdown in tolerance is necessary to gain a full understanding of SLE, constructing a conceptual framework to explain ssDNA and dsDNA binding in anti-DNA is an important first step for understanding immune self recognition and ultimately pathogenicity. This invention provides the reagents and methods for elucidating this complex pathology. Also provided by this invention are methods to inhibit the pathological complex formation implicated in this disease state thereby providing a therapy for its eradication.

Antibodies and Cell Lines: This invention provides polyclonal and monoclonal antibodies with high affinity for single-stranded DNA and single-stranded oligo-dT, low or no affinity for double-stranded DNA. The antibodies specifically bind to DNA hairpin structure. A panel of antibodies with these specificities has been generated. They are useful in the diagnosis of disorders that are associated with the pathological complexation of DNA, such as inflammatory glomerulonephritis and SLE, as well as for the generation of reagents to screen for pharmaceutical agents and therapies for the treatment and prevention of these diseases.

As used herein, an "antibody" means in one embodiment, a monoclonal antibody which is produced in an animal that reacts with DNA with an effective specificity and affinity for its intended purpose. The animal is sacrificed and its spleen cells are removed for fusion with an immortalized cell line such as a heteromyeloma to produce a clonal cell line. All monoclonal antibodies derived from the clone are chemically and structurally identical, and specific for a single antigenic determinant. The hybridoma cell lines producing the monoclonal antibodies also are within the scope of this invention. These monoclonal antibodies can be from any species, e.g., mice, rat, rabbit or a human monoclonal antibody.

As used herein, the term "monoclonal antibody" or "antibody" also intended to include antibody fragments (variable and framework regions) and recombinant antibodies having a variable and/or framework region of an antibody of this invention combined with the same from another source. Further intended by the use of this term is antibodies which have changes in their primary amino acid sequences.

As used herein, the term "high affinity" shall mean an apparent equilibrium dissociation values of less than 1 micromolar and "low affinity" shall mean an apparent

equilibrium dissociation values of more than 1 micromolar. Examples of high affinity antibodies include, but are not limited to the antibodies designated herein 9F11, 15B10, 15D8, and 11F8. Specifically excluded are the prior art antibodies Bv04-01 and HedlO. Examples of low affinity antibodies include, but are not limited to 4B2, 7B3, 8D8, 10F4 and 5F3. Specifically excluded is the prior art antibody 3H9.

The hybridoma cell line 11F8 has been deposited with the American Type Culture Collection (ATCC) 12301 Parklawn Drive, Rockville Maryland, U.S.A. 20852 on May 18, 1995 under the provisions of the Budapest Treaty for the International Recognition of the Deposit of Microorganisms for the Purposes of Patent Procedure. It was accorded Accession No. HB 11890.

Laboratory methods for producing the antibodies of this invention also are provided below, as well as methods for deducing their corresponding nucleic acid sequences. Additional methodologies for producing monoclonal antibodies and deducing their sequences, are known in the art, see Harlow and Lane, (1988) Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory, New York and

Sambrook et al. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory. Briefly, the monoclonal antibodies of this invention can be biologically produced by introducing an immunological nucleic acid molecule into an animal, e.g., a mouse or a rabbit. The antibody producing cells in the animal are isolated and fused with myeloma cells from the same or different species (mouse, rabbit or human) or heteromyeloma cells to produce hybrid cells or hybridomas. Accordingly, the hybridoma cells producing the monoclonal antibodies of this invention also are provided. As noted above, these hybridoma cells can be the result of mouse-mouse fusion, mouse-human fusion or human-human fusion methods. The immortalized cell lines (fusion partners) are commercially available from a variety of sources, including the ATCC.

This invention also provides biologically active fragments of the antibodies described above, e.g., an isolated polypeptide consisting essentially of the framework or variable region of a monoclonal antibody of this invention. These "antibody fragments" retain some ability to selectively bind with its DNA antigen or immunogen, in this case ssDNA, dsDNA or both. Such antibody fragments can include, but are not

limited to:

(1) Fab, the fragment which contains a monovalent antigen-binding fragment of an antibody molecule produced by digestion with the enzyme papain to yield an intact light chain and a portion of one heavy chain; (2) Fab', which traditionally has been defined at the fragment of an antibody molecule obtained by treating with pepsin, followed by reduction, to yield an intact light chain and a portion of the heavy chain; two Fab' fragments are obtained per antibody molecule;

(3) F(ab') ₂ , the fragment of the antibody that is obtained by treating with the enzyme pepsin without subsequent reduction; F(ab') ₂ is a dimer of two Fab' fragments held together by two disulfide bonds;

(4) Fv, defined as a genetically engineered fragment containing the variable region of the light chain and the variable region of the heavy chain expressed as two chains; and (5) SCA, defined as a genetically engineered molecule containing the variable region of the light chain, the variable region of the heavy chain, linked by a suitable polypeptide linker as a genetically fused single chain molecule. A specific examples of "biologically active antibody fragment" include the CDR regions of the antibodies. Methods of making these fragments are described below and in Harlow and Lane, (1988) supra.

As noted above, the antibodies of this invention also can be modified to create chimeric antibodies and humanized antibodies (Oi, et al. (1986) BioTechniques 4(3):214). Chimeric antibodies are those in which the various domains of the antibodies' heavy and light chains are coded for by DNA from more than one species. Accordingly, further provided by this invention are recombinant antibodies containing the variable region from an antibody, wherein the antibody has high affinity for single- stranded DNA, low or no affinity for double-stranded DNA, and capable of specifically binding DNA hairpin structure. In one embodiment of this invention, the antibodies specifically bind an adjacent stem duplex without unwinding the stem duplex, e.g., the recombinant antibodies designated 4B2 and 9F11HCDR3. In a further embodiment they have high affinity for ss oligo-dT.

The isolation of other hybridomas secreting monoclonal antibodies with the specificity of the monoclonal antibodies of the invention can also be accomplished by one of ordinary skill in the art by producing anti-idiotypic antibodies (Herlyn, et al. (1986) Science.2___:100). An anti-idiotypic antibody is an antibody which recognizes unique determinants present on the monoclonal antibody produced by the hybridoma of interest. These determinants are located in the hypervariable region of the antibody. It is this region which binds to a given epitope and, thus, it is responsible for the specificity of the antibody. The anti-idiotypic antibody can be prepared by immunizing an animal with the monoclonal antibody of interest. The immiiriized animal will recognize and respond to the idiotypic determinants of the immunizing antibody by producing an antibody to these idiotypic determinants. By using the anti-idiotypic antibodies of the second animal, which are specific for the monoclonal antibodies produced by a single hybridoma which was used to immunize the second animal, it is now possible to identify other clones with similar idiotypes as the antibody of the hybridoma used for immunization.

Idiotypic identity between monoclonal antibodies of two hybridomas demonstrates that the two monoclonal antibodies are the same with respect to their recognition of the same epitopic determinant. Thus, by using antibodies to the epitopic determinants on a monoclonal antibody it is possible to identify other hybridomas expressing monoclonal antibodies of the same epitopic specificity.

It is also possible to use the anti-idiotype technology to produce monoclonal antibodies which mimic the DNA binding area of the nucleic acid binding partner of the antibody. These anti-idiotypic antibodies are useful in an immunochemical screen for the presence of the antibodies ofthis invention. For example, an anti-idiotypic monoclonal antibody made to a first monoclonal antibody will have a binding domain in the hypervariable region which is the mirror image of the epitope bound by the first monoclonal antibody.

Also provided by this invention are compositions having the antibody (recombinant, polyclonal, monoclonal, fragments or anti-idiotypic) described above and a carrier. The antibodies ofthis invention also can be combined with various liquid phase carriers, such as sterile or aqueous solutions, pharmaceutically acceptable carriers, suspensions and emulsions. Examples of non-aqueous solvents include propyl

ethylene glycol, polyethylene glycol and vegetable oils. When used to prepare anti- idiotypic antibodies, the carriers also can include an adjuvant which is useful to non- specifically augment a specific immune response. A skilled artisan can easily determine whether an adjuvant is required and select one. However, for the purpose of illustration only, suitable adjuvants include, but are not limited to Freund's Complete and Incomplete, mineral salts and polynucleotides.

Equivalent antibodies to the antibodies specifically described herein are within the scope ofthis invention. To deteπriine whether antibodies are "equivalent" one of skill in the art will know to test whether the test antibody binds with the same specificity to the DNA target. If they do, they are equivalent. It also is possible to determine without undue experimentation, whether an antibody has the same specificity as the antibody ofthis invention by determining whether the antibody being tested prevents an antibody ofthis invention from binding it DNA target with which the monoclonal antibody is normally reactive. If the antibody being tested competes with the antibody of the invention as shown by a decrease in binding by the antibody ofthis invention, then it is likely that the two antibodies bind to the same or a closely related epitope. Alternatively, one can pre-incubate the antibody ofthis invention with the target DNA with which it is normally reactive, and determine if the monoclonal antibody being tested is inhibited in its ability to bind the DNA. If the antibody being tested is inhibited then, in all likelihood, it has the same, or a closely related, epitopic specificity as the monoclonal antibody ofthis invention.

The term "antibody" also is intended to include antibodies of all isotypes. Particular isotypes of a monoclonal antibody can be prepared either directly by selecting from the initial fusion, or prepared secondarily, from a parental hybridoma secreting a monoclonal antibody of different isotype by using the sib selection technique to isolate class switch variants using the procedure described in Steplewski et al. (1985) Proc. Natl. Acad. Sci. 82:8653 or Spira et /. (19841 J. Immunol. Methods 74:307.

Also encompassed by this invention are proteins or polypeptides that have been recombinantly produced, biochemically synthesized, chemically synthesized or chemically modified, that retain the ability to bind DNA with the same affinity and specificity defined herein.

The antibodies ofthis invention can be linked to a detectable agent or a hapten. The complex is useful to detect the DNA in a sample or detect agents which interfere with antibody-DNA binding, using the methods described below. Examples of types of immunoassays which can utilize antibodies of the invention are competitive and non- competitive immunoassays in either a direct or indirect format. Examples of such immunoassays are the enzyme linked immunoassay (ELISA) radioimmunoassay (RIA) and the sandwich (immunometric) assay. Detection methods using the antibodies of the invention include immunoassays which are run in the forward, reverse, or simultaneous modes, including immunohistochemical assays on physiological samples. Those of skill in the art will know, or can readily discern, other immunoassay formats without undue experimentation.

Greater sensitivity can be achieved by coupling the antibodies to low molecular weight haptens. These haptens can then be specifically detected by means of a second reaction. For example, it is common to use such haptens as biotin, which reacts avidin, or dinitropherryl, pyridoxal, and fluorescein, which can react with specific anti-hapten antibodies. See Harlow and Lane (1988) supra.

The antibodies of the invention can be bound to many different carriers. Thus, this invention also provides compositions containing the antibodies and another substance, active or inert. Examples of well-known carriers include glass, polystyrene, polypropylene, polyethylene, dextran, nylon, amylases, natural and modified celluloses, polyacrylamides, agaroses and magnetite. The nature of the carrier can be either soluble or insoluble for purposes of the invention. Those skilled in the art will know of other suitable carriers for binding antibodies, or will be able to ascertain such, using routine experimentation. There are many different labels and methods of labeling known to those of ordinary skill in the art that are suitably bound to the antibodies ofthis invention. Examples of the types of labels which can be used in the present invention include enzymes, radioisotopes, fluorescent compounds, colloidal metals, chemiluminescent compounds, and bioluminescent compounds. Those of ordinary skill in the art will know of other suitable labels for binding to the antibody, or will be able to ascertain such, using routine experimentation. Furthermore, the binding of these labels to the antibody of the invention can be done using standard techniques common to those of

ordinary skill in the art. The antibodies also can be combined with pharmaceutical carriers and adjuvants for the preparation of anti-idiotypic antibodies. The fusion or hybridoma cell lines which produce the antibodies also can be combined with pharmaceutical carriers for administration to immunohistocompatible mice to generate a chimeric mouse useful to screen and test new therapies for the prevention and treatment of pathologies like glomerulonephritis that are associated with anti- ssDNA*DNA complexes.

The antibodies and antibody fragments ofthis invention can biologically produced from hybridoma cell lines alone or in combination with chemical digestion to yield the fragments thereof or they can be obtained by chemical synthesis using a commercially available automated peptide synthesizer such as those manufactured by Applied Biosystems, Inc., Model 430A or 431 A, Foster City, CA and the amino acid sequence of the antibodies or the fragments as provided in Figures 7 through 10. The synthesized protein or polypeptide can be precipitated and further purified, for example by high performance liquid chromatography (HPLC). Accordingly, this invention also provides a process for chemically synthesizing the antibodies and fragments thereof by providing the sequence of the antibody or fragment and reagents, such as amino acids and enzymes and linking together the amino acids in the proper orientation and linear sequence. Alternatively, the antibodies can be obtained by well-known recombinant methods as described, for example, in Sambrook et al. (1989) supra, using the host cell and vector systems described below. Thus, this invention further provides a process for recombinantly producing an antibody or fragment thereof, by growing a host cell containing a nucleic acid molecule encoding the antibody or fragment thereof, the nucleic acid molecule being operatively linked to a promoter of RNA transcription.

The host cell is grown under suitable conditions such that the nucleic acid is transcribed and translated into protein and purifying the protein so produced.

Also provided by this application are the antibodies and fragments thereof described herein conjugated to a detectable agent for use in diagnostic methods. For example, detectably labeled anti-idiotypic antibodies or a fragment containing a conformationally stabilized fragment having similar specificity and avidity as the native anti-idiotypic antibody, can be bound to a column and used for the detection and

purification antibodies ofthis invention. It is preferable that the serum sample contacted with more that one antibody ofthis invention to determine its binding pattern. The binding pattern then can be correlated to determine whether the patient from whom the serum sample was obtained is predisposed to or has a pathology associated with the presence of anti-ssDNA*DNA complexes.

The antibodies or fragments thereof which contain a conformationally stabilized binding site having similar biding specificity and avidity as native antibody also can be used (as above or in any of the standard immunochemical techniques such as ELISA) to test for agents which interfere with or prevent the binding of DNA, preferably ssDNA, to its antibody.

Nucleic Acids: Further provided by this invention is a nucleic acid molecule coding for a monoclonal antibody ofthis invention or a polypeptide corresponding to the variable or framework region of a monoclonal antibody, wherein the monoclonal antibody is characterized by having high affinity for single-stranded DNA, low or no affinity for double-stranded DNA, and being capable of specifically binding a hairpin DNA. As used herein, the term "high affinity" shall mean an apparent equilibrium dissociation values of less than 1 micromolar and "low affinity" shall mean an apparent equilibrium dissociation values of more than 1 micromolar. Also intended to be encompassed by this invention are the complements (DNA and RNA) to these nucleic acid molecules.

Examples of such nucleic acid molecules are the nucleic acids coding for the antibodies designated 9F11, 15B10, 15D8, and 11F8. Specifically excluded are nucleic acid molecules coding for the prior art antibodies Bv04-01 and HedlO. Examples of nucleic acids coding for low affinity antibodies include, but are not limited to the nucleic acids coding for the antibodies designated 4B2, 7B3, 8D8, 10F4 and 5F3. Specifically excluded is the nucleic acid molecule coding for the prior art antibody 3H9.

In a further embodiment, the nucleic acid molecules coding for unrelated antibodies having the following GenBank Accession numbers: For 9F11 N _H : S59838, X64998, Z29586, X00160, K00706; 9F11 _κ : Z22096, X69859, S53109, U00929, X01431; 4B2 N _H : U09593, L22535, X80958, X13188, J00507; 4B2 _κ : M18239, D14630, L09009, J00575, U01851; 11F8 V _H : K00723, S50914, S51594, X15471,

M36228, L14742; 11F8 _κ : X16955, L18942, M34633, M31270, S64047; are specifically excluded.

With the nucleic acid molecules, one of skill in the art can recombinantly reproduce the nucleic acid molecules or the antibodies, anti-idiotypic antibodies or polypeptides encoded by them using methods known to those of skill in the art and described in Sambrook et al. (1989) supra, with the host cell and vector systems described below. This invention further provides a process for producing an antibody, anti-idiotypic antibody or fragments thereof by growing a host cell containing a nucleic acid molecule encoding the antibody, anti-idiotypic antibody, protein or fragments thereof, the nucleic acid being operatively linked to a promoter of RNA transcription. The host cell is grown under suitable conditions such that the nucleic acid is transcribed and translated. In one embodiment, the antibody product is purified.

As noted above, nucleic acid molecules and isolated nucleic acid molecules which encode amino acid sequences corresponding to an antibody fragment or an antibody ofthis invention, as well as complements of these sequences, are fiirther provided by this invention. In addition to these sequences, the sequences shown in the Figures 7 through 10 corresponding to fragments of the antibodies ofthis invention. Also provided by this invention are the anti-sense polynucleotide stand, e.g. antisense RNA. One can obtain an antisense RNA using the sequence of the antibodies or the sequences of the fragments provided in the Figures 7 through 10 and the methodology described in Vander Krol et al. (1988) BioTechniques 6:958. In one aspect ofthis invention, the nucleic acid molecule encoding a fragment of the antibody is defined to be any of the sequence or parts thereof shown in Figure 7 through 10. Expression of functional fragments in bacterial cells (4BscF _v ) has been accomplished with E. coli HB2151 cells (ATCC) by adding plasmid containing the nucleic acid fragment to E. coli cells in culture. The sample is incubated on ice for 45 minutes and then heat shocked for 2 minutes at 42°C, and then chilled on ice for 5 minutes. An aliquot of the transformation reaction (lOOμL) was added to LBG media (900μL), incubated for 1 hour at 37°C and then plated on selective media containing antibiotics. After growth for approximately 12 hours at 30°C, plasmid DNA was isolated and digested with Sfi I and Not I as described above. Restriction enzyme analysis can be used to confirm presence of insert.

Suitable culture for the bacteria is inoculated with bacterial culture containing the insert and grown at 30°C for 12 hours. Additional media is added and the culture is grown for an additional 1 hour with vigorous shaking. Culture is removed and pelleted by centrifugation at 1500g. For isolation of the polypeptide, the cells were resuspended in PBS (about 0.5 mL containing lmM EDTA) incubated on ice for 10 minutes and then centrifuged to pellet the cell debris. The whole cell extract was prepared by resuspending the cell pellet in PBS (0.5mL) boiling for 5 minutes and centrifugation to pellet the cell debris. Aliquots of the supernatant, periplasm and whole cell extract were analyzed by SDS-PAGE. A band migrating at 28kDa was observed in the periplasmic extract indicating soluble expression of the polypeptide.

Periplasmic extracts are bound to a Sepharose-anti-E Tag antibody column (5mL bed volume; 1 mg anti-E Tag/mL of Sepharose) at a flow rate of lmL/minute. After washing with PBS (50mL) and 0.1 M glycine (25mL, pH 5.0), the scF _v 's will be eluted with 0.1 M glycine buffer (pH 2.8) into 0.2 M TRIS-HCl, pH 8.0 (175 μL; 2mL fractions.) The purity of each scF _v is assessed by SDS-PAGE, IEF and ion exchange chromatography using a DEAE matrix. If the samples are <95% homogenous, they are further purified by affinity chromatography over agarose-ssDNA and if needed by ion exchange chromatography. This latter ion exchange step may prove critical for mutant scF _v 's that do not bind to the agarose-ssDNA column. Although each scF _v is expected to be primarily localized in the periplasm in soluble form, it is possible that some proteins may form inclusion bodies or co- precipitate with the cell pellet. In either of these cases it will be necessary to denature and re-fold the scF _v prior to chromatographic purification. As a starting point for these experiment, the conditions outlined below are used. Briefly, cell paste will be isolated by centrifugation (11000^) and suspended in 10 volumes of buffer (50 mM TRIS-HCl, 1 mM EDTA, 0.1 mM PMSF, pH 8). The mixture will be passed through a French press apparatus (10,000 psi) and the homogenate will be pelleted by centrifugation. After repeating the homogenization, the protein pellet is washed with TRIS buffer and then dissolved in denaturation buffer (6 M guanidinium hydrochlorine, 50 mM TRIS- HCl, 50 mM KCl, pH 8 to a total protein concentration of 10 mg/mL). The dissolved scF _v will then be diluted rapidly (1 :200) into renaturation buffer at 7°C (50 mM TRIS- HCl, 50 mM KCl, 10 mM CaCl ₂ , 0.1 mM PMSF, pH 8) and allowed to sit for 24 hours

without agitation. Following filtration (Millipore MiniTAN apparatus using a 10kD _a cutoff membrane), concentration (Centriprep concentrators), and dialysis against glycine buffer, the scF _v are chromatographically purified as described above. If significant quantities of the scF _v 's leak into the growth medium (as might be caused by over-expression) they are concentrated and purified chromatographically as described herein.

The invention also encompasses nucleic acid molecules which differ from that of the nucleic acid molecules described above, but which produce the same phenotypic effect, such as the allele. These altered, but phenotypically equivalent nucleic acid molecules are referred to "equivalent nucleic acids." This invention also encompasses nucleic acid molecules characterized by changes in non-coding regions that do not alter the phenotype of the polypeptide produced therefrom when compared to the nucleic acid molecules that encode the naturally occurring antibody or anti-idiotypic antibody. This invention further encompasses nucleic acid molecules which hybridize to the nucleic acid molecules of the subject invention under stringent conditions.

The nucleic acid molecules can be conjugated to a detectable marker, e.g., an enzymatic label or a radioisotope for detection of nucleic acid and/or expression of the gene encoding the antibodies in a cell or serum. Briefly, this invention further provides a method for detecting a single-stranded nucleic acid molecule encoding an amino acid sequence which is at least a portion of an antibody ofthis invention by contacting single-stranded nucleic acid molecules with a labeled, single-stranded nucleic acid molecule (a probe) which is complementary to a single-stranded nucleic acid molecule encoding an amino acid sequence which is at least a portion of an antibody under conditions permitting hybridization (preferably stringent hybridization conditions) of complementary single-stranded nucleic acid molecules. Hybridized nucleic acid molecules are separated from single-stranded nucleic acid molecules. The hybridized molecules are detected using methods well known to those of skill in the art and set forth, for example, in Sambrook (1989) supra.

The nucleic acid molecules ofthis invention can be replicated and isolated using the recombinant technique described above or replicated using PCR (Perkin-Elmer) as described below. For example, the sequence can be chemically replicated using PCR (Perkin-Elmer) which in combination with the synthesis of oligonucleotides, allows

easy reproduction of DNA sequences. The PCR technology is the subject matter of United States Patent Nos. 4,683,195, 4,800,159, 4,754,065, and 4,683,202 and described in PCR: The Polymerase Chain Reaction Mullis et al. eds, Birkhauser Press, Boston (1994) and references cited therein. Alternatively, one of skill in the art can use the sequences provided herein and a commercial DNA synthesizer to replicate the DNA. Accordingly, this invention also provides a process for obtaining the polynucleotides ofthis invention by providing the linear sequence of the polynucleotide, nucleotides, appropriate primer molecules, chemicals such as enzymes and instructions for their replication and chemically replicating or linking the nucleotides in the proper orientation to obtain the polynucleotides. In a separate embodiment, these polynucleotides are further isolated. Still further, one of skill in the art can insert the nucleic acid into a suitable replication vector and insert the vector into a suitable host cell for replication and amplification. The DNA so amplified can be isolated from the cell by methods well known to those of skill in the art. A process for obtaining nucleic acid molecules by this method is further provided herein as well as the nucleic acid molecules so obtained.

RNA can be obtained by using the isolated DNA and inserting it into a suitable cell. A suitable cell for this purpose includes but is not limited to a bacterial cell, a yeast cell, or a mammalian cell. The DNA can be inserted by any appropriate method, e.g., by the use of an appropriate insertion vector or by electroporation. When the cell replicates and the DNA is transcribed into RNA; the RNA can then be isolated using methods well known to those of skill in the art, for example, as set forth in Sambrook et al. (1989) supra.

The invention further provides the isolated nucleic acid molecule operatively linked to a promoter of RNA transcription, as well as other regulatory sequences for replication and or transient or stable expression of the DNA or RNA. As used herein, the term "operatively linked" means positioned in such a manner that the promoter will direct transcription of RNA off the DNA molecule. Examples of such promoters are SP6, T4 and T7. In certain embodiments, cell-specific promoters are used for cell- specific expression of the inserted nucleic acid molecule. Vectors which contain a promoter or a promoter/enhancer, with termination codons and selectable marker sequences, as well as a cloning site into which an inserted piece of DNA can be

operatively linked to that promoter are well known in the art and commercially available. For general methodology and cloning strategies, see Gene Expression Technology. Goeddel ed., Academic Press, Inc. (1991) and references cited therein and Vectors: Essential Data Series Gacesa and Ramji, eds., John Wiley & Sons, N.Y. (1994), which contains maps, functional properties, commercial suppliers and a reference to GenEMBL accession numbers for various suitable vectors. Preferable, these vectors are capable of transcribing RNA in vitro or in vivo.

As noted above, an isolated nucleic acid molecule ofthis invention can be operatively linked to a promoter, either an inducible or non-inducible promoter, of RNA transcription. Accordingly, this invention also provides a vector (insertion, replication or expression vector) having inserted therein an isolated nucleic acid molecule described above, for example, a viral vector, such as bacteriophage, baculovirus and retrovirus, or cosmids, plasmids, YACS, yeast and other recombinant vectors. Nucleic acid molecules are inserted into vector genomes by methods well known in the art. For example, insert and vector DNA can both be exposed to a restriction enzyme to create complementary ends on both molecules that base pair with each other and which are then joined together with a ligase. Alternatively, synthetic nucleic acid linkers can be ligated to the insert DNA that correspond to a restriction site in the vector DNA, which is then digested with a restriction enzyme that recognizes a particular nucleotide sequence. Additionally, an oligonucleotide containing a termination codon and an appropriate restriction site can be ligated for insertion into a vector containing, for example, some or all of the following: a selectable marker gene, such as neomycin gene for selection of stable or transient transfectants in mammalian cells; enhancer/promoter sequences from the immediate early gene of human cytomegalovirus (CMV) for high levels of transcription; transcription termination and RNA processing signals from SV40 for mRNA stability; SV40 polyoma origins of replication and ColEl for proper episomal replication; versatile multiple cloning sites; and T7 and SP6 RNA promoters for in vitro transcription of sense and anti-sense RNA. An additional example of a vector construct ofthis invention is a bacterial expression vector including a promoter such as the lac promoter and for transcription initiation, the Shine-Dalgarno sequence and the start codon AUG (Sambrook et al. (1989) supra). Similarly, a eucaryotic expression vector is a heterologous or

homologous promoter for RNA polymerase II, a downstream polyadenylation signal, the start codon AUG, and a termination codon for detachment of the ribosome. Such vectors can be obtained commercially or assembled using the sequences described herein. Expression vectors containing these nucleic acids are useful to obtain host vector systems to produce the antibodies of the invention. It is implied that these expression vectors must be replicable in the host organisms either as episomes or as an integral part of the chromosomal DNA. Suitable expression vectors include viral vectors, including adenoviruses, adeno-associated viruses, retroviruses, cosmids, etc. When a nucleic acid is inserted into a suitable host cell, e.g., a procaryotic or a eucaryotic cell and the host cell replicates, the protein can be recombinantly produced. Suitable host cells will depend on the vector and can include mammalian cells, animal cells, human cells, simian cells, insect cells, yeast cells, and bacterial cells constructed using well known methods. See Sambrook et al. (1989) supra. In addition to the use of viral vector for insertion of exogenous nucleic acid into cells, the nucleic acid can be inserted into the host cell by methods well known in the art such as transformation for bacterial cells; transfection using calcium phosphate precipitation for mammalian cells; or DEAE-dextran; electroporation; or microinjection. See Sambrook et al. (1989) supra for this methodology. Thus, this invention also provides a host cell, e.g. a mammalian cell, an animal cell (rat or mouse), a human cell, or a bacterial cell, containing a nucleic acid molecule encoding an antibody or a polypeptide ofthis invention.

A Chimeric Mouse: The antibodies ofthis invention provide a novel reagent for a screen for pharmaceuticals and methods to treat or prevent pathological DNA- antibody binding, especially ssDNA-antibody binding which leads to severe inflammatory glomerulonephritis, inflammatory glomerulonephritis associated with SLE and nephritis associated with other diseases. The hybridoma cell lines which produce these antibodies can be administered to an immunohistocompatible non- autoimmune mouse. Suitable mice for this purpose include, but are not limited to BALB/c mice, p57 black mice or the mice described below. These animals are commercially available from Jackson Laboratories. Accordingly, also provided by this invention is a chimeric mouse having a hybridoma cell line ofthis invention. These

mice are very suitable for screening for agents which inhibit pathological DNA anti- DNA interactions.

Compositions: The antibodies, fragments, nucleic acid molecules and hybridoma cell lines ofthis invention also can be combined with various liquid phase carriers, such as sterile or aqueous solutions, pharmaceutically acceptable carriers, suspensions and emulsions. Examples of non-aqueous solvents include propyl ethylene glycol, polyethylene glycol and vegetable oils. When used to prepare anti-idiotypic antibodies, the carriers also can include an adjuvant which is useful to non-specifically augment a specific immune response. A skilled artisan can easily determine whether an adjuvant is required and select one. However, for the purpose of illustration only, suitable adjuvants include, but are not limited to Freund's Complete and Incomplete, mineral salts and polynucleotides.

Thus, this invention also provides pharmaceutical compositions containing the antibodies, fragments, nucleic acid molecules and hybridoma cell lines and a pharmaceutically acceptable carriers. These are useful for the preparation of medicaments for the diagnosis and treatment of pathologies associated with antibody- DNA complex formation.

Industrial Applicability: The antibodies ofthis invention are useful not only for the generation of anti-idiotypic antibodies and recombinant antibodies, they also provide reagents for a cell-free screen for agents that inhibit or prevent the binding of antibody to DNA, and preferably ssDNA. This is significant since it is the presence of complex which is deposited in the kidneys and leads to severe inflammatory glomerulonephritis.

To perform the cell free screen, an effective amount of an DNA, preferably heat denatured calf thymus DNA, is first bound to a solid support (for example, glass, polystyrene, polyethylene, dextran, nylon, natural and modified celluloses, polyacrylamides, agaroses or coated onto microtiter plate). Those skilled in the art will know of other suitable carriers for this purpose. They are deposited in a suitable concentration, e.g., between about 5 μg/ml to about 12 μg/ml, and more preferably between about 6 μg/ml and about 10 μg/ml. Then, approximately 10 μg/ml of agent is contacted with the DNA under suitable conditions which favor the binding of antibody to DNA. Subsequently or simultaneous with the previous contacting step, a

diagnostically effective amount of anti-DNA antibody to be is contacted under suitable conditions which favor binding of the DNA to the antibody. Procedures for the detecting of complex are then performed. Simultaneously a control with no agent is performed. The amount of bound antibody is then compared relative to the control. If the agent prevented the formation of an antibody-DNA complex by greater than 50% at a concentration of lOμM or less as compared to control, it is a candidate pharmaceutical.

The anti-idiotypic antibodies ofthis invention are similarly useful to screen for the presence of the antibodies ofthis invention which have been correlated with the presence of severe inflammatory glomerulonephritis and nephritis associated with other diseases. Using standard immunochemical techniques, a serum sample is obtained from the patient and contacted with a detectably labeled anti-idiotypic antibody under suitable conditions which favor the formation of antibody-anti-idiotypic complex formation and determining whether any complex was formed. The presence of complex being a positive indication that the patient has or is predisposed to developing a disease caused by the presence of these complexes. This method also can be used to screen for drugs which will interfere or prevent the formation of these complexes and therefore are useful to treat or prevent the formation of complex and its associated disease. Accordingly, this invention also provides a kit to perform the screens described above. The kits comprise the above reagents, control reagents and instruction for use to accomplish the objective of the screen.

This invention further provides administering a hybridoma cell line, preferably the cell line which produces antibody 11F8 to an immunocompatible non-autoimmune mouse and the chimeric mouse having these cells. The mouse is a powerful animal model to screen for agents which are an effective therapy to treat disorders associated with the formation of anti-DNA#DNA complex in an individual. This method comprises the steps of providing a mouse as identified above and administering an effective amount of a cell line which produces an antibody ofthis invention. The mouse is then maintained for an additional three to six days. The agent is then administered to the mouse and compared to the control, and if lives longer than the control, it is an agent effective or potential agent for treating these disorders.

As used herein, disorders associated with the formation of anti-DNA*DNA complex in an individual include but are not limited to severe inflammatory glomerulonephritis associated with SLE and nephritis associated with other diseases.

The agents or drugs identified by this method also are provided by this invention. One class of drugs having this pharmaceutical efficacy is 1,4- benzodiazepine derivatives having the formula:

In the above formula, R ¹ is a moiety comprising from 1 to about 30 hydrogen atoms and from 0 to about 15 carbon atoms. In certain embodiments, R ¹ may further comprise from 1 to about 3 nitrogen atoms; from 1 to about 3 oxygen atoms; and from 1 to about 3 sulfur atoms. As described below, R is derived from the α-amino acid used during synthesis. A group of preferred R ¹ moieties includes -CH ₂ CH ₂ COOH (derived from glutamic acid), -CH ₂ COOH (derived from aspartic acid), -CH ₂ -C ₆ H ₅ (derived from phenylalanine), -CH ₂ CH ₂ -C ₆ H ₅ (derived from homophenylalanine), and - CH(C ₂ H ₅ )CH ₂ -(2-naphthyl) (derived from 2-naphthyl-isoleucine).

Also in the above formula, R ² is a moiety comprising from 1 to about 30 hydrogen atoms and from 0 to about 15 carbon atoms. In certain embodiments, R further comprises from 1 to about 3 halogen atoms (such as iodine, bromine, chlorine, or fluorine); from 1 to about 3 nitrogen atoms; from 1 to about 3 oxygen atoms; and from about 1 to about 3 sulfur atoms. As described below, R ² is either -H or is derived from the optional alkylating agent used during synthesis. In one group of preferred

1 9 embodiments, R is simply -H. Another group of preferred R moieties includes 4- bromo-benzyl and -CH ₂ C(=O)O-C(CH ₃ ) ₃ ).

Presently, the most prefened 1,4-benzodiazepines derivatives are the p-chloro- p-hydroxy-l,4-benzodiazepines shown below; the observed per-cent inhibition observed for each of these derivatives is also shown.

Compound R ¹ R ² % Inhibition

I -CH ₂ CH ₂ COOH -CH ₂ C(-=O)OC(CH ₃ ) ₃ 50

II -CH ₂ COOH p-bromobenzyl 40

III -CH ₂ CH ₂ COOH p-bromobenzyl 50

IV -CH(C ₂ H ₅ )CH ₂ -(2-naphthyl) -H 95

V -CH ₂ -C ₆ H ₅ -H 80

VI -CH ₂ CH ₂ -C ₆ H ₅ -H 80

Improved solid-phase synthetic methods for the preparation of a variety of 1,4-benzodiazepine derivatives with very high overall yields have been reported in the literature. See, for example, Bunin and Ellman, J. Am. Chem. Soc, 1992, Vol. 114, pp. 10997- 10998. Using these improved methods, the 1 ,4-benzodiazepines derivatives are constructed on a solid support from three separate components: 2- aminobenzophenones, α-amino acids, and (optionally) alkylating agents, as shown in the reaction scheme of Figure 13.

Prefened 2-aminobenzophenones include the substituted 2- aminobenzophenones, for example, the halo-, hydroxy-, and halo-hydroxy- substituted 2-aminobenzophenones, such as 4-halo-4'-hydroxy-2-aminobenzophenones. A prefened substituted 2-aminobenzophenone is 4-chloro-4'-hydroxy-2- aminobenzophenone.

Prefened α-amino acids include the 20 common naturally occurring α-arnino acids as well as α-amino acid mimicking structures, such as homophenylalanine, homotyrosine, and thyroxine. A prefened group of α-amino acids is represented by the formula H ₂ N-CH(R')-COOH.

Prefened alkylating agents, represented by the formula R ² -X, wherein X is -I, - Br, or -Cl, include both activated and inactivated electrophiles, of which a wide variety

are well known in the art. Prefened alkylating agents include the activated electrophiles p-bromobenzyl bromide (i.e., where X is -Br and R is p-bromo-benzyl) and t-butyl-bromoacetate (i.e., where X is -Br and R ² is -CH ₂ C(=O)O-C(CH ₃ ) ₃ ).

In the first step of such a synthesis, the 2-aminobenzophenone derivative, (1), is attached to a solid support, such as a polystyrene solid support, through either a hydroxy or carboxylic acid functional group using well known methods and employing an acid-cleavable linker, such as the commercially available

[4-(hydroxymethyl)phenoxy]acetic acid, to yield the supported 2-aminobenzophenone, (2). See, for example, R.C. Sheppard and B. J. Williams, Int. J. Pept. Protein Res., 1982, Vol. 20, pp. 451-454. The 2-amino group of the 2-aminobenzophenone is preferably protected prior to reaction with the linking reagent, for example, by reaction with FMOC-C1 (9-fluorenylmethyl chloroformate) to yield the protected amino group 2*-NHFMOC.

In the second step, the protected 2-amino group is deprotected (for example, the -NHFMOC group may be deprotected by treatment with piperidine in dimethylformamide (DMF)), and the unprotected 2-aminobenzophenone is then coupled via an amide linkage to an l-amino acid (the amino group of which has itself been protected, for example, as an -NHFMOC group) to yield the intermediate (3). Standard activation methods used for general solid-phase peptide synthesis may be used (such as the use of carbodiimides and hydroxybentzotriazole or pentafluorophenyl active esters) to facilitate coupling. However, a prefened activation method employs treatment of the 2-aminobenzophenone with a methylene chloride solution of the of α- N-FMOC-amino acid fluoride (i.e., FMOC-NH-CH(R ¹ )-C(=O)F) in the presence of the acid scavenger 4-methyl-2,6-di-tert-butylpyridine yields complete coupling via an amide linkage. This prefened coupling method has been found to be effective even for unreactive aminobenzophenone derivatives, yielding essentially complete coupling for derivatives possessing both 4-chloro and 3-carboxy deactivating substituents.

In the third step, the protected amino group (which originated with the amino acid) is first deprotected (for example, -NHFMOC may be converted to -NH ₂ with piperidine in DMF), and the deprotected compound is reacted with acid, for example, 5% acetic acid in DMF at 60°C, to yield the supported 1,4-benzodiazepine derivative, (4). Complete cyclization has been reported using this method for a variety of

2-aminobenzophenone derivatives with widely differing steric and electronic properties.

In an optional fourth step, the 1 ,4-benzodiazepine derivative is alkylated, by reaction with a suitable alkylating agent and a base, to yield the supported fully derivatized 1,4-benzodiazepine, (5). Standard alkylation methods, for example, an excess of a strong base such as LDA (lithium diisopropylamide) or NaH, may be used; however, such methods may result in undesired deprotonation of other acidic functionalities and over-alkylation. Prefened bases, which may prevent over-alkylation of the benzodiazepine derivatives (for example, those with ester and carbamate functionalities), are those which are basic enough to completely deprotonate the anilide functional group, but not basic enough to deprotonate amide, carbamate or ester functional groups. An example of such a base is lithiated 5-(phenylmethyl)-2- oxazolidinone, which may be reacted with the 1,4-benzodiazepine in tetrahydrofuran (THF) at -78°C. Following deprotonation, a suitable alkylating agent, as described above, is added. In the final step, the fully derivatized 1,4-benzodiazepine, (6), is cleaved from the solid support. This may be achieved (along with concomitant removal of acid-labile protecting groups), for example, by exposure to a suitable acid, such as a mixture of trifluoroacetic acid, water, and dimethylsulfide (85:5:10, by volume).

Accordingly, this invention further provides a method of preventing or treating disorders associated with formation of DNA*anti-DNA complexes in an individual by administering to the individual an effective amount of benzodiazepine from the class identified above to prevent or treat the disorder. One of skill in the art can determine when such a pathology is prevented or treated by kidney tissue biopsy or suitable clinical chemical test such as protein urea concentration. As used herein, the term "administering" means providing the individual with an effective amount of the agent or benzodiazepine derivative effective to inhibit complex formation. Methods of administering pharmaceutical compositions are well known to those of skill in the art and include, but are not limited to, microinjection, intravenous or parenteral administration. The compositions are intended for systemic or local administration as well as intravenously, subcutaneously, or intramuscularly. Administration can be effected continuously or intermittently throughout the course of treatment. Methods of determining the most effective means and dosage of

administration are well known to those of skill in the art and will vary the condition of the individual, including age weight and general health. Single or multiple administrations can be carried out with the dose level and pattern being selected by the treating physician. For example, the compositions can be administered to an individual already suffering from an inflammatory disease or condition. In this situation, an effective "therapeutic amount" of the composition is administered to prevent or at least partially inhibit further kidney damage.

However, the compositions can be administered to subjects or individuals susceptible to or at risk of developing severe glomerulonephritis (which can be predicted by the appearance of complex in the patient's serum) to prevent further complex formation and disease progression. In these embodiments, a "prophylactically effective amount" of the composition is administered.

The compositions ofthis invention also can be used in immune tolerance therapy. As used herein "immune tolerance" is a permanent form of immune suppression which keeps individuals from reacting with their own tissues. A failure of this mechanism is the cause of SLE. Thus, if one can re-establish this immune suppression, the disease can be effectively reversed. The use of the DNA ligands to the antibodies or the binding site of the anti-idiotypic antibodies ofthis invention can be used to re-establish immune tolerance. The DNA ligand or binding sites can be administered alone or conjugated to a carrier using methods known to those of skill in the art and describe in U.S. Patent No. 5,276,013. A method of treating SLE in an affected individual comprises adrninistering a therapeutically effective amount of the antibody or conjugate described above.

It is to be understood that while the invention has been described in conjunction with the above embodiments, that the foregoing description and the following examples are intended to illustrate and not limit the scope of the invention. Other aspects, advantages and modifications within the scope of the invention will be apparent to those skilled in the art to which the invention pertains.

Throughout this application are publications referenced by arabic numerals within parenthesis and by authors' name. Full citations for these publications may be found at the end of the specification immediately preceding the claims. The disclosures of these publications, as well as issued U.S. patents, published patent specifications and

publications referenced in this application are hereby incorporated by reference into the present disclosure to more fully describe the state of the art to which this invention pertains.

EXAMPLES Generation and Characterization of a Panel of Anti-DNA Antibodies

Monoclonal Antibody Preparation: A single unmanipulated 8-week-old female MKL-lpr mouse was obtained from Jackson Laboratories (Bar Harbor, ME) and housed in the University of Michigan Unit for Laboratory Animal Medicine in a pathogen-free environment. Serum samples were screened by ELISA (infra) every four weeks for the presence of anti-DNA, until the titer showed high levels of both anti- ssDNA and anti-dsDNA. The mouse was sacrificed by CO ₂ asphyxiation at 26- weeks of age and the spleen removed. Spleen cells were fused with nonsecreting myeloma Sp2/O cells in a 5:1 ratio in PBS containing 15% DMSO and 42% PEG 4000 at 37°C for 30 seconds (23). The solution was slowly diluted with serum free media followed by addition of unstimulated peritoneal cells in HAT media and seeded in 96- well microtiter plates. After 21 days, 620 viable hybridomas were observed (65% of 960 starting wells) of which 137 reacted with heat denatured calf thymus DNA. These hybridomas were subcloned by limiting dilution and reassayed for anti-DNA reactivity. Several hybridomas died and others stopped producing antibody during this time (presumably due to gradual chromosome loss and segregation of the genes required for antibody synthesis; 24) leaving eleven viable cell lines. The isotype of the anti-DNA produced by each hybridoma was determined using an isotyping kit (Boehringer Mannheim, Indianapolis, IN) according to the manufacturer's instructions. All of these anti-DNA possess K light chains.

Large quantities of monoclonal antibodies ("mAbs") for characterization were produced in ascites by interperitoneal injection of -10 hybridoma cells into pristane- primed retired female BALB/c breeders. Ascites fluid was clarified by centrifugation and then chromatographed over protein A-agarose (Pierce, Rockford, IL) at 4°C using the ImmunoPure buffer system (Pierce) according to manufacturer's specifications. Eluted immunoglobulin was exchanged into DNA binding buffer (50 mM K ₂ HPO ₄ ,

pH 8, 150 mM NaCl, 1 mM EDTA) using Centricon 30 microconcentrators (Amicon, Beverly, MA). The IgG fraction (1-2 mg) was then chromatographed over ssDNA- agarose (GEBCO-BRL, Gaithersburg, MD; 2 mL bed volume) equilibrated with DNA binding buffer at 4°C. After washing with 10 column volumes of binding buffer (flow rate ~ 0.15 rnL/min), mAbs 4B2, 7B3, 10F4 were eluted with DNA binding buffer containing 2 M NaCl (elution buffer), whereas 8D8, 9F11, 11F8, 15B10, and 15D8 were eluted with a steep gradient of urea (to 4M dissolved in elution buffer). Anti- DNA eluted with urea retained full activity as judged by complete retention of purified protein samples on the ssDN A-agarose column. In addition, the affinity of the Ab (as judged with the gel shift assay; infra) was not altered after repetitive elution with these low concentrations of urea.

Anti-DNA samples were purified to homogeneity by high-performance ion exchange chromatography on DEAE. Briefly, the mAb were exchanged into Tris buffer (20 mM Tris-HCl, pH 7.5) and about 1-2 mg of protein loaded onto a Protein Pak DEAE 5PW column (Waters, Marlborough, MA) equilibrated with Tris buffer at 4°C. The protein was eluted with a linear gradient of Tris buffer (pH 8) containing 1 M NaCl at a flow rate of 1 mL/min. The desired fractions were pooled, concentrated, and exchanged into binding buffer. Protein concentrations were determined using the BCA protein assay reagent (Pierce, Rockford, IL) using normal mouse IgG of known ' concentration as a standard.

F(ab) Preparation: Crude IgG was isolated from clarified ascites fluid by protein A chromatography as described above. Intact IgG was digested with immobilized papain using the ImmunoPure F(ab) preparation kit (Pierce) according to manufacturer's specifications. Isolated F(ab) was exchanged into DNA binding buffer and chromatographed over ssDNA-agarose. F(ab) purified using this strategy was > 95% pure by SDS-PAGE. The pi of the F(ab) was determined by comparing to pi markers (Pharmacia, Piscataway, NJ) on EEF 3-9 gels using the PhastGel system (Pharmacia).

Nucleic acids: Calf thymus DNA (10 mg; Calbiochem, San Diego, CA) was dissolved in Tris buffer (60 mM Tris, pH 8.0, 100 mM NaCl, 2 mM CaCl ₂ , to 1 OmL total volume), deproteinized by chloroform/isoamyl alcohol extraction, and precipitated with ethanol. The DNA was then briefly incubated with micrococcal nuclease

(40 U/mg DNA) and purified by gel filtration on BioGel A1.5M (BioRad, Hercules, CA) according to Stollar (25).

Single-stranded DNA and RNA homopolymers, poly(dA)*poly(dT) and poly(dG)*poly(dC) were purchased from Pharmacia and used without further purification. Oligodeoxyribonucleotides were chemically synthesized and 5' end- labeled with [γ- ³² P]-ATP as previously described (18, 26).

For screening assays employing a small DNA ligand, the oligonucleotide was first conjugated to BSA to facilitate binding to the microtiter plate (27). Briefly, a 21- base-long oligonucleotide that folds into a stem-loop structure was synthesized with a 5'-hexylaminolinker and purified as previously described (18, 26). The hairpin was dissolved in water (0.5 mL) containing l-ethyl-3-(3-dimethyl- aminopropyl)carbodiimide (~ 50 equiv) and the pH was adjusted to 5.0. BSA was then added (0.25 equiv, Boehringer-Mannheim) and the conjugation was allowed to proceed at 23 °C maintaining the pH at 5.0. After 24 hours the reaction was quenched with sodium acetate (3 M, pH 4.5) and the conjugate was exhaustively dialyzed against PBS and used without ^• further purification. The ratio of the OD _{260 280} indicated that about three hairpins are conjugated to each BSA molecule.

ELISA: Immulon II microtiter plates (Dynatec, Chantilly, VA) were coated with heat denatured calf thymus DNA (100°C for 12 minutes, then 0°C for 20 min, 10 μg/mL), hairpin-BSA conjugate (5 μg/mL), or polynucleotides (10 μg/mL) in TBS (10 mM Tris, pH 7.4, 150 mM NaCl) for 18 hours at room temperature (28). To prepare wells containing only dsDNA, wells coated with calf thymus DNA were treated with SI nuclease (0.2 U/mL) in nuclease buffer (100 mM sodium acetate, pH 4.6, 100 mM NaCl, 0.1 mM mM ZnCl ₂ , 0.043% glycerol) at 37°C for 2 hours (29). After blocking with PBS containing 3% BSA for 18 hours at room temperature, the wells were washed with PBS containing 0.1% Tween-20 (PBS-Tween). Anti-DNA samples for assay were diluted in PBS containing 1% BSA and 0.05% Tween-20 (PBT), added to the appropriate wells, and incubated for 2 hours at room temperature. After washing with PBS-Tween, goat anti-mouse Ig-alkaline phosphatase conjugate (Boehringer- Mannheim) diluted 1 : 1000 in PBT was added and incubated for 2 hours. After washing, bound anti-DNA was visualized by the addition of /?mfrophenylphosphate

(Sigma- 104, 1 mg/mL) in Na ₂ CO ₃ (100 mM, pH 9.6). The absorption at 405 nm was measured using a microtiter plate reader (Biotek Instruments, Winooski, VT).

Reactivity to Sm, Sm/nRNP, SS-A, SS-B, histone, and Scl-70 autoantigens was determined by direct ELISA using a commercially available kit (INCSTAR, Stillwater, MN). Briefly, anti-DNA samples (50 μL of 5 μg/mL solutions) were added to wells that were precoated with antigen (as provided by the manufacturer) and incubated for 30 minutes at 25°C. After washing, bound anti-DNA was detected using a goat anti- mouse Ig-alkaline phosphatase conjugate and visualized by addition of p- nitrophenylphosphate as described above. Positive controls provided by the manufacturer were run in parallel.

Binding to cardiolipin (CL; Fluka, Ronkonkoma, NY), fibronectin (FN; Fluka), laminin (Lam; ICN, Costa Mesa, CA), heparin sulfate (HS; Fluka), phosphatidyl serine (PS; Fluka), and collagen type IV (ColIV; Fluka) was assessed by direct ELISA. Briefly, solutions of CL, PS in ethanol (100 μL of 50 μg/mL solutions) were added to Immulon II microtiter plates and evaporated over 18 hours at 25°C (30). Lam, ColIV, HS (50 μL of solutions that are 2 μg/mL, 5 μg/mL, and 50 μg/mL, respectively, in PBS) and FN (50 μL of a 10 μg/mL solution in 0.1 M sodium carbonate, pH 9.6) were coated 18 hours at 25°C (31-34). After blocking with PBS containing 3% BSA, anti-DNA (50 μL of 5 μg/mL solutions) were incubated with the antigens for 2 hours at 25°C. Bound antibody was detected as described above.

Gel shift assay and DNA footprinting: Anti-DNA affinity for oligonucleotide ligands was measured by gel shift assay as described by Stevens et al. (35) except that the sodium ion concentration was adjusted with NaCl (to 150 mM). The number of cations released upon DNA binding was assessed by performing a series of binding titrations as a function of sodium ion concentration. Footprinting with KMnO ₄ , diethyl- pyrocarbonate (DEPC), dimethyl sulfate (DMS) was performed as previously described by Swanson et al. (18).

Fluorescence quenching studies: Fluorescence measurements were performed on a Perkin-Elmer LS-50 luminescence spectrometer equipped with a thermostated cell block maintained at 23°C. All measurements were carried out in phosphate buffer (50 mM K ₂ HPO ₄ , pH 8.0, 150 mM NaCl, 1 mM EdTA) following the procedures described by Kelly et al. (36) and Kim et al. (37). Briefly, solutions of monoclonal

antibody (~1 μM) were excited at 278 or 295 nm and the emission spectra was monitored for changes in fluorescence as a function of varying DNA ligand concentration. To minimize inner filter effects, the optical density at 295 nm was kept under 0.1 AU. The binding site size was approximated by measuring the number of molar equivalents of nucleotide necessary to achieve complete saturation of the binding sites as indicated by fluorescence quenching (36).

Experimental Results

Isolation and purification of monoclonal anti-DNA: A panel of anti-DNA hybridomas was generated from an unmanipulated MPL-lpr mouse for the presence of antibodies that bind to the ssDNA regions on heat denatured calf thymus DNA and to the dsDNA epitopes present on native calf thymus DNA. Positive clones also was screened against a DNA hairpin stem-loop construct to determine (a) the frequency with which small DNA ligands are bound by anti-DNA and (b) the extent to which the single stranded and double stranded regions on hairpins can model the ssDNA and dsDNA epitopes recognized by anti-DNA (18, 35). Indeed, greater than 85% of anti- DNA reactive with either ssDNA or dsDNA recognized the hairpin construct by ELISA, suggesting that small oligonucleotide constructs can model epitopes on ssDNA and dsDNA ligands. Those that do not bind the hairpin construct either prefer a different sequence or may bind weakly, requiring longer antigens for multivalent recognition (avidity). Application ofthis hairpin ligand in other binding experiments is described later.

Purification of large quantities of mAbs from endogenous Ig present in ascites fluid was achieved by affinity chromatography on protein A-agarose followed by affinity chromatography using an ssDNA-agarose column. Additional purification of each mAb by preparative ion-exchange chromatography on a DEAE matrix afforded samples >98% pure as judged by analytical ion exchange chromatography and SDS- PAGE (Figure 1). Each mAb is more acidic than the conesponding F(ab) fragments and the pi values of the IgG2a F(ab)'s are generally higher than those generated from the IgG3 or IgG2b mAb. Moreover, F(ab)'s generated from these anti-DNA hybridomas that bind dsDNA (1OF4 and 4B2, IgG2a) possess higher pi values than all

of the other anti-DNA. These observations suggest the IgG2a mAb may possess a greater number of cationic residues than the other anti-ssDNA in this panel and is consistent with binding data suggesting that DNA recognition by the IgG2a mAb involves more electrostatic interactions than the other anti-ssDNA (infra). Specificity of monoclonal anti-DNA: While monoclonal antibodies generally do not crossreact with many different antigens, the reactivity of anti-DNA mAbs can be wider, including recognition of both cellular and nuclear antigens (39, 40). Precisely defining anti-DNA specificity is important because recognition of antigens other than DNA may be involved in both the evolution of anti-DNA and their pathogenicity in vivo (40-45). Although there have been several studies addressing this issue, there is conflicting evidence on the nature and extent of anti-DNA crossreactivity (6, 30, 42, 43, 45, 46-48). To help determine the extent to which crossreactivity is a feature of anti- DNA recognition, the ability ofthis panel of mAbs to bind antigens identified previously as having crossreactive epitopes (30-34) was assessed. These molecules include nuclear proteins, ribonucleoproteins, phospholipids, proteoglycans, and extracellular matrix components normally present within normal glomerular basement membrane. As shown in Table 1, only 7B3 shows crossreactivity to any of these antigens. These data support recent studies (30, 45, 46) suggesting the specificity of anti-DNA may be nanower than has been described (31, 33, 42, 43, 47). To gain a better understanding of the specificity that these anti-DNA display toward nucleic acid ligands, their reactivity toward several different polynucleotides was assessed. The data indicate that each mAb strongly reacts with poly(dT) which is consistent with previous observations that poly(dT) contains immunodominant epitopes (Table 2) (1, 19, 20, 49). Several mAb, including 9F11, 15B10, 15D8 and 11F8 only recognize poly(dT) whereas the others bind one or more of the other polydeoxyribonucleotides. In contrast, none of the mAb bind significantly to polyribonucleotides, a result also consistent with previous studies (1, 20, 49). These findings may either suggest that anti-DNA can discriminate between the two different sugar moieties, or that more complicated secondary or tertiary structures exist in these polymers that limit access to the nucleotide bases in RNA relative to DNA (50). Given that these anti-DNA bind oligo(dU), the latter alternative seems more likely (infra).

Relative affinity of anti-DNA for oligonucleotide ligands: Although the initial direct ELISA experiments reveal some information about the specificity of these anti- DNA, they do not provide a reliable assessment of the relative differences in affinity between different anti-DNA (51, 52). To address this point, binding to small DNA oligomers was assessed using a gel shift assay which directly measures binding in solution at equilibrium (53). Twenty-one (21)-base-long oligomers were selected for the test antigens because they are long enough to retain the features of polymeric DNA (i.e., they should contain at least one epitope for binding), but too short to be bridged by both antibody combining sites simultaneously thereby avoiding the possibility of artifacts caused by avidity. The reactivity patterns determined by ELISA generally reflect the gel shift data (Table 3). Specifically, all mAb have the highest relative affinity for dT ₁ , regardless of isotype. However, the relative affinity for dT ₂₁ varies with the isotope, with IgG2b mAb having the highest affinity. dA ₂₁ is not recognized by any of mAb on this panel, and the affinity toward either dG ₁₅ or dC ₂₁ varies among the different antibodies, especially in the IgG2a subclass. Although the IgG2b and IgG3 anti-DNA do not bind to poly(dl) by ELISA, the gel shift data clearly show that the order of base specificity for the IgG2b and IgG3 mAb is dT»dG>dC≥dA.

Other groups have observed similar trends in the specificity of monoclonal anti- DNA. For example, Tetin et al. (54) using fluorescence quenching experiments find that BV04-01 (IgG2b, K) binds dT ₆ with about 5 times greater affinity than dG ₆ (I = 0.13 vs 0.71 μM, respectively) and does not exhibit appreciable affinity for either dC ₆ or dAg (54). Similarly, Lee et al. (20) reported that an F(ab) fragment of HEdlO (IgG2a, K) displays much higher affinity toward oligo(dT) than either oligo(dC) or oligo(dA) (K _d = 80 nM vs > 2 μM, respectively) (20). Both BV04-01 and HEdlO are derived from (NZBxNZW)Fj mice. The data presented here suggest these trends are also representative of anti-DNA derived from MRL-lpr mice and is consistent with variable region sequence analysis demonstrating a high degree of homology between anti-DNA generated from different mouse strains (14, 15).

In addition to single-stranded homopolymers, anti-DNA affinity for dsDNA was assessed with a disulfide cross-linked dodecamer (1, Figure 2). This ligand was chosen because (a) high resolution structural information is available for this sequence (56), (b) the short length of the sequence precludes binding by both antigen binding sites

simultaneously, and (c) the disulfide cross-links prevents terminal fraying and keeps the duplex from adopting alternate conformations without compromising the native helical geometry (26). These latter two points are particularly important since alternate conformations may present a variety of "complex" epitopes, including regions of ssDNA. Although these anti-DNA may possess sequence selectivity and hence may not bind duplex optimally, this sequence nevertheless provides the basic geometric requirements of B-DNA and is therefore a good starting point with which to examine the binding properties of anti-dsDNA. The data in Table 3 indicate that only two of the IgG2a anti-DNA, 4B2 and 10F4, have an appreciable affinity for dsDNA. Interestingly, these anti-DNA also bind dT ₂₁ , although the relative affinity for ssDNA is much weaker than seen with the IgG2b anti-DNA.

To address the contribution of ion pair formation in stabilizing anti-DNA-DNA complexes, a gel shift assay was employed to measure __ _d values for each mAb as a function of [Na+]. Plotting the ln[__ _d ] against ln[Na+] should afford a line whose slope represents the number of sodium cations released upon binding (57, 58). Assuming that the released cations were originally bound to the phosphates on DNA, and neglecting anion effects, the number of released cations represent the number of phosphates that have exchanged a sodium ion for a positively charged residue on the antibody (59). Binding of dT ₂₁ by the IgG2b mAb (e.g. 15D8) and by 11F8 is not accompanied by release of bound cations, whereas the other IgG2a mAbs (e.g., 4B2) appear to release one cation upon complex formation (Figure 3). These data suggest that recognition of dT ₂₁ does not involve ion pair formation for the IgG2b and IgG3 mAbs, whereas the IgG2a anti-DNA may form one salt bridge upon binding. This observation is in general agreement with studies of other anti-ssDNA. For example, Lee and coworkers have found that two phosphates may form ion pairs in the complex between an F(ab) fragment of HEdlO and poly(dT), whereas only one ion pair is observed in BV04-01- ssDNA complexes (17-20). Thus, it seems that electrostatic interactions play a relatively minor role in recognition of ssDNA.

The affinity of 4B2 for the cross-linked duplex as a function of sodium ion concentration also was measured. The data suggest one cation is released upon formation of the 4B2-dsDNA complex, which is also the same number observed in the binding of 4B2 to dT ₂₁ . Similar analysis could not be performed for mAb 10F4

because at high [Na+], binding was too weak to measure. The cation release observed for 4B2 suggests that the molecular interactions which mediate binding of both ssDNA and dsDNA by this mAb may not be just electrostatic. Although the presence of cationic residues like lysine and arginine in antiDNA (HCDR3 in particular) may be associated with the greater affinity for dsDNA (60), they are not necessarily indicative of ion pair formation. Indeed, cationic residues are also involved in hydrogen bonding, van der Waals contacts, and stabilizing CDR architectures through inter- and intra- chain hydrogen bonds and salt bridges (17, 61, 62).

Fluorescence measurements, binding site size, and intrinsic affinity: DNA recognition can result in the quenching of the intrinsic fluorescence of tryptophan and/or tyrosine in DNA binding proteins (36, 37). In anti-DNA, these phenomena have been attributed to changes in microenvironment upon ligand binding and reflects a static mechanism of quenching (54). Studies by both Tetin et al. (54) and Lee et al. (20) show anti-ssDNA undergo fluorescence quenching in the presence of ssDNA ligands. Each of the mAbs ofthis panel display fluorescence quenching when titrated with poly(dT) (Figure 4). This effect is observed at excitation wavelengths of either 278 nm or 295 nm, implicating tryptophans, and perhaps tyrosines in DNA binding (37). In control experiments using poly(dA), fluorescence quenching of these anti- DNA is not observed, indicating that complex formation is required for this effect. Also, when normal mouse IgG is titrated with poly(dT), no quenching is found which indicates that nonspecific binding is not responsible for changes in the fluorescence intensity. Lastly, the other DNA homopolymers do not induce fluorescence quenching when titrated with these mAbs. These results suggest that the other DNA polymers may not be bound in the same manner as poly(dT) and that the difference in affinity may be reflected in these different modes of recognition.

The number of consecutive nucleotides occluded upon binding (i.e., epitope size) has been determined for several single-stranded DNA binding proteins using fluorescence quenching methods (36, 37). This analysis is possible if the affimty for a polynucleotide is sufficiently high to approximate a linear reduction in fluorescence intensity as a function of nucleotide concentration. After a stoichiometric amount of ligand is added, all of the available binding protein binding sites are saturated and titrating with additional ligand has no effect on the protein fluorescence. Performing

this experiment with this panel of mAbs using poly(dT) as the ligand reveals that about five nucleotides are occluded upon binding (Figure 4). These data are in general agreement with the site size determined for both BV04-01 and HEdlO (17, 20).

Because the fluorescence quenching experiments suggest that the epitope size of this panel mAbs is about five nucleotides long, determining the "intrinsic" affinity for a DNA ligand that just fills the binding site was next determined. The intrinsic affinity of the mAbs ofthis panel for dT ₅ , dG ₅ and dC ₅ was measured using the gel shift assay. A representative equilibrium binding isotherm is shown in Figure 5. Significant binding is only observed with dT ₅ and only for about half of the mAbs (Table 4). One important finding is that the intrinsic affimty for dT ₅ is lower than the apparent affinity observed for dT ₂₁ which suggests that the apparent affinity for dT _2! (and all the polymer ligands) is largely derived from additivity effects resulting from the presence of multiple overlapping binding sites (36). While this observation is not surprising, one implication of these results is that the binding data obtained using high molecular weight nucleic acid polymers does not reflect the intrinsic affinity for a single binding site on the ligand and may complicate affinity/specificity analysis.

To examine the contribution of the C5 -methyl group of thymine in stabilizing anti-DNA-oligo(dT) complexes, the deoxyuridine analogs of the thymidylate ligands was synthesized and binding titrations were performed using the gel shift assay and fluorescence spectroscopy (Table 4). The fluorescence spectra of anti-DNA ofthis panel are quenched by the addition of oligo(dU), suggesting that the mode of recognition of oligo(dU) is similar to oligo(dT). However, the affinity for dU5 is about 5-fold lower for dT ₅ for both 9F11 and 11F8. This increased affinity for oligo(dT) relative to oligo(dU) may be explained if the C5 methyl group of thymine is oriented into the antigen binding site where it can participate in favorable van der Waals contacts. Indeed, this hypothesis is consistent with the results of KMnO ₄ footprinting experiments (infra).

The mode of dsDNA recognition with fluorescence binding titrations was assayed and found that neither poly(dG)-poly(dC), poly(dA)-poly(dT), nor the model dodecamer duplex afford changes in the fluorescence spectra of 4B2. One possibility is that binding occurs at a site other than the antigen binding cleft. However, dT ₂₁ competes with the duplex in complexes with anti-DNA and vise versa, indicating that

both ligands are bound in the same site. Thus, although fluorescence quenching is observed upon titration with poly(dT), binding of dsDNA has no effect on the fluorescing chromophore. This observation, though indirect, indicates that for this mAb, the mode of binding dsDNA is different from the manner of ssDNA recognition. Quenching of the chromophore in ssDNA binding may involve direct contact with ligand or conformational changes in the antigen binding site associated with complexation that alters the microenvironment of chromophore (54). Binding dsDNA may utilize a different set of residues (14, 15) or require less movement to accommodate the ligand. By measuring the helix-coil transition of poly(dA)-poly(dT) in the presence of different concentrations of anti-dsDNA Jel241, Braun and Lee (22) estimated that this mAb recognizes about six base pairs, which is consistent both with the dimensions of the known antibody combining site size and the binding site size obtained here for ssDNA ligands. Defining the autoreactive epitopes on ssDNA and dsDNA antigens:

Potassium permanganate selectively modifies the C5-C6 double bond of thymine bases that are not Watson-Crick hydrogen bonded and can be used to probe the accessibility ofthis base in protein-DNA complexes, including those with anti-DNA (18). Since each of the mAbs ofthis panel has a preference for binding poly(dT) and a binding site size of about five nucleotides in length, their interaction with a 21 -base-long oligomer containing a central dT ₅ segment flanked by a nonbinding (dA) sequence was examined to aid in sequencing (oligomer 2, Figure 2). When compared to normal mouse IgG as a control, the pattern of permanganate reactivity of 2 is marginally increased when complexed to any of the IgG2a mAbs (Figure 5). However, the reactivity of the thymine bases is quite different when bound by the IgG2b and the IgG3 mAbs.

Quantitative measurements of the difference in probability of modification clearly show that in the presence of 9F11 and 11F8 the thymine bases in 2 exhibit both hypeneactivity and reduced reactivity relative to nlgG (18; Figure 4). Furthermore, the asymmetry of the KMn0 ₄ reactivity pattern suggests that ssDNA is bound with a specific polarity, however, the orientation of the DNA (5' to 3' or 3' to 5') in the antibody binding site cannot be ascertained from these data alone. Those positions that are hypermodified (e.g., T _n and T ₉ for 9F11) may result from a DNA conformation in

which the C5-C6 double bond of the base projects away from the binding site where it is exposed to KMnO ₄ modification. In contrast, if the base is oriented in a pocket so that the C5 methyl groups faces the antigen binding site or the DNA adopts a conformation in the complex that shields the base relative to the free DNA, the C5-C6 double bonds may be rendered less accessible to KMnO ₄ modification (e.g. T ₁₀ and T for 9F11). The minimal increases in permanganate sensitivity seen in DNA ligands bound the IgG2a mAbs coupled with their relatively low affinity suggests that DNA binding may be nonspecific.

Another mechanism used by antibodies for antigen recognition is induced fit (18, 35, 65, 66). For example, it was previously demonstrated by the inventors that concomitant with binding the single-stranded loop in DNA hairpins, anti-ssDNA BV04-01 partially unwinds the adjacent stem duplex (18). To determine if the anti- DNA described here induce conformational changes in DNA ligands, mAb complexes of hairpin 3 were footprinted (Figure 2) with KMnO ₄ and sensitivity was examined at sites adjacent to the single-stranded pentathymidine loop. None of the mAbs reported here induce significant structural changes in the duplex region ofthis hairpin construct. This observation suggests that either the energetic penalty for partially denaturing the duplex is greater than the available binding free energy or that these mAbs do not require conformational changes in the duplex to achieve optimal binding (67). Since the affinity of the 21 IgG2b anti-DNA and BVO4-01 for 3 is similar (-O.5 μM) the latter alternative is favored.

Footprinting duplex 1 (Figure 2) complexed to mAbs 4B2 and 10F4 with KMnO ₄ , DMS, and hydroxyl radical does not result in any detectable pattern of protection or provide evidence of conformational changes in the ligand; similar results were also obtained with longer duplexes, both crosslinked and unmodified (68). The lack of a distinct footprint suggests that either the target of the chemical probes are equally accessible in both the bound and free DNA or that the binding is not specific enough to yield discernible footprints. Because duplex 1 may not be the optimal sequence for binding, the latter alternative may be more likely (it may also be possible that anti-DNA are not sequence selective DNA binding proteins). In support ofthis hypothesis, La Baer and Yamamoto reported that while the relative affinity of the glucocorticoid receptor for its consensus binding sequence containing a single base-pair

substitution is reduced only five-fold relative to the wild-type sequence, the footprint generated by copper phenanthroline is abolished (69).

DISCUSSION

Antibodies that bind ssDNA constitute the largest population of anti-DNA (1) and their reactivity and mode of ligand recognition is better understood than anti- dsDNA (16). Most of what is known about the mechanism of anti-ssDNA binding is based on the crystal structure of anti-ssDNA BVO4-01 complexed to d(pT3), along with models of BVO4-01-DNA complexes based on binding/footprinting studies (17, 18). The results from these experiments have shown that (a) base stacking of thymine between aromatic residues stabilizes binding of ssDNA, (b) ion pair formation is limited, and (c) hydrogen bonding to the phosphate backbone as well as to substituents on the DNA bases are important for binding. Studies of the anti-ssDNA HEdlO by Lee and coworkers support and extend these findings (20). These studies show that tryptophans are involved in binding poly(dT), four consecutive residues are recognized, and two ion pairs may be formed in the complex. Contact to the pyrmdine ring, specifically the 5-methyl, 4-carbonyl, and 3-imino groups, was infened through affinity measurements to poly(dU), poly[d(brU)] and poly[d(brC)].

The four highest affinity mAbs in this study, 9F11, 15B10, 15D8 and 11F8, comprise one group of anti-ssDNA and share many similarities with both BVO4-01 and HEdlO. These similarities include specificity for thymidine, limited ion effects, involvement of tryptophans and tyrosines in binding, and a binding site size that is about five nucleotides long. Importantly, KMnO ₄ footprinting offers clues to common features of recognition in this group of anti-DNA. For example, these experiments show that two thymine bases are protected from KMnO ₄ modification and two bases are hypermodified by KMnO ₄ in the presence of9Fl l, 15B10 and 15D8, whereas in the presence of 11F8, only one base is protected while the other four are hypermodified. These observations are consistent with a mechanism in which the thymine base is specifically bound in a subsite on the protein surface, possibly sandwiched between two aromatic residues as revealed in X-ray and modeling studies of BVO4-01 and predicted for HEdl 0 (17, 18, 85). The driving force for this mode of recognition is not clear, but since binding is not accompanied by cation release, it may involve hydrophobic effects

including van der Waals contacts and the entropic benefit of releasing bound water molecules from the protein surface (86, 87).

Since the pattern of reactivity in 9F11, 15B10 and 15D8 is similar, it is possible that these mAbs are clonally related. The small differences in affimty in these three mAbs may reflect somatic mutations that subtly alter the complementarity of the binding site to the ligand without changing the basic mode of recognition. That 11F8 protects one less base from modification than the other three mAbs may account for the lower affinity to oligo(dT), but the overall mechanism of binding appears conserved among this group of anti-ssDNA. Taken together, these data suggest that "high affinity" anti-ssDNA possess a limited repertoire of motifs to bind ssDNA. The hallmark ofthis group of anti-ssDNA may be the ability to bind nucleotide bases in subsites in the antibody binding cleft. Differences in affinity may be due to the number of subsites available to accommodate base stacking or variations in the CDRs that affect the overall molecular complementarity of the binding site to DNA. A second group of anti-ssDNA, including 4B2, 7B3, 8D8 and 10F4, have lower affinity for oligo(dT) relative to the first group, and 4B2 and 10F4 also bind dsDNA. Anti-DNA in this group may form an ion pair in complexes with oligo(dT), as evidenced by the release of one cation upon complex formation, and show no pattern of KMnO ₄ protection. The lack of KMnO ₄ protection suggests the low affinity of these anti-ssDNA may be due to a different geometry of the antigen binding site in which pockets that accommodate a thymine base are not present and may be similar to the shallow binding site topology observed in crystallographic studies of anti-dsDNA Jel72, which was obtained by immunization with poly(dG)-poly(dC) (88).

Anti-ssDNA share a number of features with other ssDNA binding proteins, like T4 bacteriophage gene 32 protein and the E. coli single strand binding (SSB) protein (37, 89). These include specificity for and high affinity binding of poly(dT) (e.g., apparent K _ά to dT ₁₆ of 0.1 μM by SSB-1 protomer, and 14 nM by gene 32, both in 50 mM [Na+]), and involvement of tryptophan and tyrosine residues in binding, particularly through base stacking. However, the SSB and T4 bacteriophage gene 32 proteins also have several important differences, including cooperativity, several modes of DNA binding, and large ion effects consistent with extensive electrostatic interactions (89, 90). In this regard, anti-ssDNA may represent a relatively simple

model of ssDNA binding proteins. In principle, understanding the mode of anti-ssDNA recognition may help illuminate the molecular basis for the intrinsic affinity of other ssDNA binding proteins to poly(dT). Indeed, structural studies of the bacteriophage fl gene V ssDNA binding protein suggest that flexible loops within a framework of β- sheets may be a common motif used to bind ssDNA (91).

One anti-DNA in this group, 4B2, has affinity for a duplex DNA comparable to the anti-dsDNA H241 (21) and stronger than Jel241 (22). In competition experiments, it was shown that dsDNA is bound in the same binding site as is ssDNA. Binding of dsDNA may be stabilized by the formation of one ion pair and ligand-induced fluorescence quenching was not observed. The number of cations released upon binding dsDNA is the same between mAb 4B2 and anti-dsDNA BV17-45 (1), but less than the four reported for anti-dsDNA Jel241 (22). Thus, the number of putative ion pairs formed may not strictly conelate with affinity, even though residues like arginine that potentially can form ion pairs are selected by anti-dsDNA and appear to contribute to dsDNA binding. Taken together, these data suggest formation of hydrogen bond networks to the backbone may play an important role in stabilizing complexes with dsDNA.

Sequence specificity of dsDNA binding proteins often results from protein secondary structures that can insert into the major groove of dsDNA to form a network of hydrogen bonds to the DNA bases (91). Whether CDR loops positioned on a framework of β-sheets can impart sequence-specific recognition of dsDNA is not known. Although several sequence-specific dsDNA binding proteins constructed from β-sheet architecture exist (92), dsDNA binding proteins which model the β-sheet-CDR loop interface in their binding site have not been reported. Stollar has proposed that the antibody binding site could straddle the phosphate backbone allowing the CDR loops to project into the major and minor grooves of dsDNA (16). Whether this anangement can provide the basis to discriminate among related sequences is unclear. Studies in this laboratory using in vitro selection techniques are cunently underway to address this point. With an appropriate consensus DNA sequence, footprinting may provide insight into the autoreactive epitopes on dsDNA recognized by anti-dsDNA.

To summarize, these studies provide detailed mechanistic insight into the recognition of ssDNA and dsDNA by anti-DNA autoantibodies. These data indicate

that the range of anti-DNA reactivity is quite nanow and that anti-DNA may be grouped according to shared modes of ligand recognition. If the mechanism of binding is conserved in these groups, it may be possible to exploit this fact in the design of molecules that inhibit binding to classes of anti-DNA (8). Antagonists derived from knowledge of the molecular basis of anti-DNA-DNA interactions would, in principle, be more selective for anti-DNA than non-specific immunosuppressive agents. Although recent studies have provided cogent evidence that recognition of DNA by anti-DNA is required for renal tissue damage (7), this relationship is still poorly understood. While serum levels of anti-dsDNA are conelated with disease progression (3), antibodies to both ssDNA and dsDNA can induce nephritis in nonautoimmune mice (5, 6). Characterizing the features of DNA binding that distinguish pathogenic from nonpathogenic anti-DNA will hopefully aid in revealing the molecular basis for pathogenic self-recognition. To address these issues, the V-region genes of these mAbs were sequenced and examined their in vivo properties to provide a framework for future studies of anti-DNA structure, binding, and pathogenicity

Cloning and Characterization of the Variable Regions of the Antibodies

Elucidating the structural features of anti-DNA that are related to their activity both in vitro and in vivo remains a key goal of SLE research. Although the genes that encode anti-DNA have been extensively studied, the molecular basis of DNA recognition by anti-DNA and the relationships between anti-DNA-DNA complex formation in vitro, and disease pathogenesis in vivo are poorly understood. One problem is that relatively few monoclonal anti-DNA have been studied for their capacity to induce nephritis, so that features which distinguish pathogenic from nonpathogenic anti-DNA are not defined. Furthermore, little is known about the ligand binding properties of nephritogenic anti-DNA compared to anti-DNA that do not induce kidney damage. To address these issues, a panel of anti-DNA from an MRL-lpr mouse and characterized their affinity, specificity, mode of DNA binding, genetic origins, and ability to induce nephritis in vivo.

Nucleic acid: Primers were synthesized on an Expedite Nucleic Acid Synthesizer (Milligen, Framingham, MA) using β-cyanoethyl phosphoramidite

chemistry. Oligomers were deprotected and purified using OPC columns (ABI, Foster City, CA) according to the manufacturer's specifications. The preparation of ssDNA and dsDNA has been described previously.

Cloning and sequencing of V _H and V _L chain genes: Poly(A+) RNA was isolated from 10 hybridoma cells by adsorption to poly(dT) cellulose using the FastTrack mRNA isolation system (Invitrogen, San Diego, CA). Full length first- strand cDNA was obtained by reverse transcription of poly(A+) RNA using an oligo(dT) primer and avian myeloma reverse transcriptase using the cDNA Cycle kit (Invitrogen). Anti-DNA V _H and V _L genes were amplified by PCR from first strand cDNA using AmpliTaq polymerase (Perkin Elmer Cetus, Norwalk, CT). A degenerate set of primers was used to amplify the V _H genes whereas a single primer set was used to amplify the V _L sequences. The sequence of the primers is as follows:

VL(κ) GTGCCAGATGTGAGCTCGTGATGACCCAGTCTCCA (3') C _L (κ) TCCTTCTAGATTACTAACACTCTCCCCTGTTGAA

C _H 1 [γ2a2b] GATATCACTAGTGGGCCCGCTGGGCTC C _H 1 [γ3] TGGGCAACTAGTACCTGGGGGGGTACTGGGCTTGG V _H AGGTCCAGCT(T/G)CTCGAGTC(T/A)GG

Briefly, 10-100 ng of cDNA was added to PCR buffer (10 mM Tris-HCI pH

8.3, 50 mM KCl, 1.5 mM MgCl ₂ , 0.01% gelatin) containing dNTPs (each 250 μM), primers (each 1 μM) and AmpliTaq polymerase (5 U) to a final volume of 100 μL. The samples were overlaid with liquid wax (80 μL, MJ Research, Watertown, MA), denatured (94°C, 5 min) and annealed (54°C, 5 min) before being subjected to 40 cycles of extension (72°C, 3 min), denaturation (93°C, 1.5 min) and annealing (54°C, 2.5 min) using a thermal cycler (MJ Research). The length of the PCR products was verified on a 1% agarose gel.

PCR products were ligated into the precut sequencing vector pCRII (Invitrogen) using T4 DNA ligase (4 U, 15°C, 18 hours) and the construct was used to transform competent Escherichia coli INVαF' cells (Invitrogen) according to the manufacturer's protocol. Transformants were spread on LB agar plates containing both ampicillin (50 μg/mL) and X-Gal (10 mg/plate), and were grown overnight at 37°C. Colonies

containing an insert (identified by blue-white color selection) were then grown to log phase at 37 °C in liquid LB media (5 mL) containing 50 μg/mL ampicillin. Plasmid DNA was isolated using the Wizard Miniprep kit (Promega, Madison, WI) and analyzed by restriction analysis with EcoRI for an insert of the conect length. Plasmid DNA (3-5 μg) was sequenced directly with a Sequenase 2.0 kit (USB, Cleveland, OH) following the manufacturer's alkaline denaturation protocol. Each nucleotide sequence was determined from at least two independent bacterial colonies.

Adoptive Transfer Experiments: Ten days after pristane priming, hybridoma cells (~10 ⁷ ) from 4B2 (IgG2a), 9F11 (IgG2b), 11F8 (IgG3), and IF/12, an IgG2a secreting control (obtained from the American Type Culture Collection, Rockville, MD), were injected into the peritoneal cavity of 6-week old (AKR x DBA/2)F ₁ mice (three mice per cell line). Levels of proteinuria, hematuria and antiDNA activity were measured prior to adoptive transfer and before sacrifice. Proteinuria was measured using a Chemstrip 6 (Boehringer Mannheim, Indianapolis, IN) and hematuria was quantified. At the onset visible ascites (between 7-21 days), mice were tail bled and serum and ascites samples were assayed for anti-DNA activity by ELISA as previously described. Mice testing positive for anti-DNA in both serum and ascites were sacrificed by anesthesia. Kidney sections were stained with either trichrome or hematoxylin/eosin and analyzed by light microscopy. Immune complex deposition was examined by direct immunofluorescence and electron microscopy as previously described. The IgG concentration in serum and ascites was determined by ELISA using a commercially available kit (Boehringer Mannheim).

In vitro binding assays: Binding of mAbs to protein and DNA antigens was assessed in two sets of experiments as described by Ohnishi et al. (45). First, Immulon II microtiter plates were coated with either histone (Boehringer Mannheim), collagen type IV (Fluka, Ronkonkoma, NY), fibronectin (Fluka) or laminin (ICN, Costa Mesa, CA) (100 μL of 5 μg/mL solutions in PBS) and then blocked with PBS containing 3% BSA. MAbs 9F11 and 11F8 (50 ng) were preincubated with either poly(dT) (50 ng) or native ctDNA, or heat denatured ctDNA (50 ng) in PBS for 1 hour at 25°C and then for 18 hours at 4°C. These complexes were then added to wells coated with the proteins described above and incubated for 2 hours at 25°C. After washing wells with PBS containing 0.1 % Tween-20, bound mAb was detected with alkaline phosphatase

conjugated goat anti-mouse Ig followed by addition of 7-nitrophenyl phosphate substrate as described. In the second set of experiments, preformed protein-DNA complexes were coated onto microtiter plates before addition of anti-DNA mAbs. Briefly, DNA [50 ng of either poly(dT), native ctDNA, or heat denatured ctDNA] was preincubated with either histone, laminin or collagen (50 ng) for 18 hours at 4°C and then coated onto microtiter plates. MAbs 9F11 and 11F8 (50 μL of 1 μg/mL solutions in PBS containing 1% BSA and 0.05% Tween-20, PBT) were added to the appropriate wells and incubated for 2 hours at 25°C. Bound anti-DNA was detected as described above. Binding of mAb-DNA bistone complexes to collagen IV, laminin, fibronectin and heparin sulfate (Fluka) was performed as described by Ohnishi et al. (45). Briefly, 9F11 or 11F8 (50 ng) were preincubated with histone (100 ng) and DNA [100 ng of either poly(dT), native ctDNA, or heat denatured ctDNA] in PBS for 1 hour at 25°C and then 18 hours at 4°C. These complexes were then transfened to appropriate wells precoated with either collagen IV, laminin, fibronectin ( 100 μL of 10 μg/mL solutions of each in PBS) or heparin sulfate (100 μL of a 25 μg/mL solution in PBS). After incubating for 2 hours at 25°C, wells were washed and bound anti-DNA was detected as described above.

MAbs 9F11 and 11F8 were examined for their ability to bind glomerular antigens present in isolated kidney. Briefly, 2-μm thick cryostat sections of naive AKR mouse kidney were mounted on glass slides as previously described. Kidney sections were then treated with deoxyribonuclease I (Sigma; immersed in 50 mL of buffer containing 20 mM Tris-Cl, pH 8, 10 mM MgCl ₂ , 100 μg/mL DNase I) for 1 hour at 37°C. After washing with PBS, either normal mouse IgG, 9F11, or 11F8 (100 μL of 10 μg/mL solutions in PBS) were added to kidney slices and incubated for 1 hour at 25°C. Bound antibody was detected as described previously. In a second group of experiments, the antibody of interest was preincubated with dT ₂₁ (0.8 μM, 1 hour at 25°C) prior to incubating with the kidney section.

The binding of mAbs 4B2, 9F11 and 11F8 to mouse complement C3 was assessed following the protocol of Oshnishi et α/.(45). Briefly, heat denatured calf thymus DNA (100 μL of a 10 μg/mL solution in TBS) was coated onto Immulon II microtiter plates for 18 hours at 25°C. After blocking with PBS containing 3% BSA,

purified mAbs (50 μL of a 1 μg/mL solution in PBT) were added to appropriate wells and incubated for 1 hour at 25°C. After washing, freshly pooled BALB/c serum diluted 1 :25 in veronal buffered saline (either with or without preheating at 60°C for 30 minutes) was added and incubated at 37°C for 1 hour. Bound complement was visualized by addition of anti-mouse C3 peroxidase conjugate (Cappel, Durham, NC) followed by ABTS substrate (Boehringer Mannheim). A second group of experiments were performed as described above except that the diluted serum was incubated with mAbs that were precoated onto microtiter plates (50 μL of 1 μg/mL solutions in PBS) for 18 hours at 25°C and blocked with PBS containing 3% BSA. Serum and ascites were tested for the presence of anti-idiotype antibodies prior to, and following adoptive transfer of 11F8 hybridomas by ELISA. Briefly, the F(ab) of 11F8 was prepared and purified according to the procedure of Swanson et al. (18). Microtiter plates were coated with 11F8 F(ab) (50 μL of a 1 μg/mL solution in PBS) for 18 hours at 25°C and blocked with PBS containing 3% BSA. Serum and ascites diluted 1 : 100 in PBT was added and incubated for 2 hours at 25°C. At this dilution, high reactivity to ssDNA is observed. Bound IgG was detected using anti-mouse Fc alkaline phosphatase conjugate (Cappel) followed by addition of 7-nitrophenyl phosphate substrate (Sigma, St. Louis, MO).

RESULTS Analysis of anti-DNA V genes: The complete V _H and V _L sequences and their conesponding gene families were determined for nine anti-DNA (Figures 7 and 8). Clones 9F11, 15B10, 15D8, and a low avidity anti-DNA 5F3, share identical V _H DJ _H and VKJK junctional regions, providing strong evidence for their clonal relatedness. The remaining anti-DNA are all single isolates. Eight of the V _H genes are derived from the J558 family while one clone uses V _H Q52. This finding is consistent with previous genetic analysis showing that the J558 family accounts for the majority of anti-DNA V _H genes. Radic and Weigert (15) have described ten homology subgroups that account for most of the V _H J558 derived anti-DNA. It was determined that 7B3, 8D8, and 1OF4, belong to subgroups 4, 7, and 8, respectively, while the other six clones are sufficiently different to make assignment ambiguous.

In contrast to the recunent usage of V _H J558 genes, five different VK genes are represented among these anti-DNA. Of the anti-DNA whose V _H J558 subgroup could be assigned, V _H J558-V _K pairings shown previously to be recunent are not observed. Comparison of the gene structures of our anti-DNA to published anti-DNA V _H and V _κ genes reveal that the most closely related sequences are between 86-100% homologous. The high degree of sequence similarity to published anti-DNA sequences and the recunent usage of the V _H J558 gene family suggests that this panel of mAbs are representative of antiDNA commonly expressed in lupus-prone mice.

Previous analysis of anti-DNA V genes provides evidence for recunent structural motifs that may be associated with DNA binding. These include, among others, the presence of asparagine at position 35 in HCDRl, tyrosine at position 100 in HCDR3, arginines in HCDR3 and at position 96 of LCDR3 (particularly in association with Jκl), and the sequence tyr-tyr-gly-ser-ser in HCDR3. Inspection of the V-region gene structures of these anti-DNA show that several mAbs possess motifs that may confer specificity to DNA. Specifically, the HCDRl of 7B3 encodes an asparagine at position 35 and two anti-DNA, 1OF4 and 4B2, have tyrosines at position 100 of HCDR3. The group of clonally related mAbs contain two arginines at positions 96 and 98 of HCDR3 and mAbs 4B2 and 8D8 encode an arginine at position 96 of LCDR3.

The structural basis for the reactivity of three representative anti-DNA from this panel, 4B2 (IgG2a, K), 9F11 (IgG2b, K) and 11F8 (IgG3, K), was investigated further. These mAbs were chosen based on their DNA binding activity, gene usage, and physical properties that may be associated with pathogenicity. Clone 4B2 shares several features common to pathogenic anti-DNA like isotype and reactivity to dsDNA. Moreover, this mAb is homologous to anti-dsDNA 3H9, a mildly pathogenic anti-DNA characterized by Weigert and coworkers. MAb 9F11 represents a relatively high affinity anti-ssDNA that is part of a group of clonally related antibodies. The last mAb, 11F8, was chosen because IgG3 production has been implicated in the development of glomerular nephritis in MRL-lpr mice. In addition, its mode of ligand recognition is similar to 9F11, thereby providing an opportunity to test the relationship between the mode of ligand binding, gene structure, and pathogenicity. Because the germline gene sequences in MRL-lpr mice that encode anti-DNA are not characterized, we constructed putative consensus sequences for these mAbs using nucleotide sequences

from both other anti-DNA, and antibodies with different binding specificities (Figure

9).

There are few differences between the V-region of 4B2 and its consensus nucleotide sequence. Specifically, a GGA to AGA conversion at codon 53 in HCDR2 results in a gly to arg replacement. This mutation occurs in other anti-DNA like 3H9 and is thought to impart specificity for dsDNA, an observation that is consistent with the binding data obtained for 4B2. HCDR3 is apparently constructed from a D _SP27 gene element with short N sequences at the 5' and 3' ends (Figure 10). The light chain is identical to the derived consensus sequence for this mAb and suggests that the 4B2 VK gene may be near germline in origin. A TGG to CGG conversion occurs in the Jκl gene that gives rise to arg at position 96 in LCDR3. Whether this mutation originates from junctional diversity or somatic mutation is unknown. Interestingly, it was previously reported that this mutation is conelated with dsDNA specificity. Although 4B2 reacts with dsDNA, the absence of dsDNA reactivity in clone 8D8, which also possesses the same JK mutation, suggests this conelation is not strict.

Clones 9F11, 15B10, 15D8, and 5F3 have >99% sequence identity with each other, and show little divergence from the derived consensus sequence. 9F11 differs from the V _H consensus sequence at codons 34 and 55, substituting ile for met in HCDRl and asn for ser in HCDR2. Construction of HCDR3 appears to involve fusion of D _Q52 and D _SP162 D _H genes (Figure 10). The codons formed from the D _H -D _H junction and the N addition 3 'of the D _SP162 segment encode arginines at positions 96 and 98 of HCDR3. Both of these mechanisms have been described previously to explain the origin of arg in HCDR3 of anti-DNA. The presence of arg in HCDR3 is one characteristic feature of anti-DNA, and their selection in these mAbs may imply an important role of arg in defining their specificity. The VK sequence of this group is homologous to the derived consensus sequence. Clones 9F11 and 5F3 are identical, whereas mAbs 15B1O and 15D8 differ in only one codon. Specifically, 15B1O has a conservative val for ala mutation in FR1, and 15D8 loses a hydrogen bond donor/acceptor in an thr to ile mutation in LCDR2. The observation that 5F3 possesses only low avidity to ssDNA whereas 9F 11 , 15B 10 and 15D8 bind ssDNA with relatively high affinity suggests that a mutation in 5F3 strongly interferes with ligand recognition. The only mutation unique to 5F3 is a ser to arg mutation in HCDRl. Although arg

mutations are generally thought to improve affinity to DNA, this mutation may disrupt the geometry of the antigen combining site by altering the conformation of the HCDRl. The structural basis for this observation is not resolved, but preliminary molecular modeling studies suggest that the ser is involved in a hydrogen bond with the carbonyl oxygen of thr 27 which may be sterically precluded by replacement with arg.

Furthermore, mutagenesis experiments by Weigert offer precedence for loss of DNA reactivity by the incorporation of arg by showing that a single lys to arg mutation in HCDR2 of 3H9 completely abrogates DNA binding activity.

Two relatively conservative replacement mutations occur in the framework regions of the heavy chain of clone 11F8. One substitutes val for gly in FR1 and the other is a substitution of lys for arg at the end of FR3. HCDR3 is constructed from the fusion of an inverted D _SP2527 and a D _SP2527 gene (Figure 10). In the light chain, a gly replaces ala at position 50 in LCDR2 and this is the only residue altered from the consensus sequence. Both the J _H and the J _κ genes are unmutated. That 11F8 may be encoded by a near germline sequence is particularly significant given previous studies showing that this mAb recognizes ssDNA with relatively high affinity and results showing this mAb is unusually pathogenic in vivo (infra).

Pathogenicity of anti-DNA: Hybridomas of mAbs 4B2, 9F11, and 11F8, as well as an irrelevant IgG2a secreting mAb (HOPC IF/12) were administered i.p. to 6- week old non-autoimmune (AKR x DBA/2)Fj mice to determine whether these mAbs induce nephritis in vivo. This method of adoptive transfer was pursued over i.v. injection of purified mAbs because it may provide levels of circulating Ig, cytokines and inflammatory factors that more closely resemble conditions associated with active disease. All animals developed ascites within about 10 days and showed evidence of proteinurea and hematuria (Table 6). Levels of anti-DNA in serum and ascites are very similar. However, the total level of IgG produced varies among groups of mice.

The amount of circulating IgG is not conelated to the induction of nephritis in nonautoimmune mice, since mice administered with HOPC IF/ 12 have larger quantities of circulating antibody than those inoculated with 11F8 but do not develop nephritis. Light microscopy of kidney sections of animals treated with 11F8 show enlarged glomerulus with cellular proliferation and infiltration of polymorphonuclear leukocytes (Figure 11). Immunofluorescence and electron microscopy studies provide evidence of

diffuse mesangial and subendothelial immune deposits with the presence "wire loop" formation. Taken together, these observations are indicative of diffuse proliferative glomerular nephritis. In contrast, the staining patterns of kidney sections from animals treated with antiDNA 4B2 and 9F11 were not remarkably different from the negative control HOPC IF/12 which appeared normal. The difference between the pathogenicity of 11F8 and 9F11 is striking, given the high degree of similarity in their specificity and mode of DNA binding.

Potential mechanisms of anti-DNA pathogenicity: To explore the possibility that anti-DNA bound to histone-DNA complexes may localize to the glomeruli via interactions between histone and the GBM, the question of whether ternary complexes consisting of histone, mAb, and DNA [either poly(dT), ssDNA, or dsDNA] bind to heparan sulfate, collagen IV, laminin, and fibronectin was examined. Both 9F11 and 11F8 form complexes with poly(dT) and histone that bind to laminin and fibronectin, but not to heparan sulfate or collagen IV. Importantly, the levels of antibody binding to laminin and fibronectin were higher in the presence of both histone and DNA than just DNA alone. Complexes containing ssDNA or dsDNA did not bind to the ECM components tested. These results show that binding of ternary immune complexes to the GBM is a plausible mechanism for the localization of anti-DNA to the glomeruli. To determine whether complex or uncharacterized antigens exist in vivo that were not tested in the ELISA experiments, the binding of nlgG, 9F11, and 11F8 to isolated naive kidney sections by immunofluorescence was examined. As expected, nlgG does not bind either unmanipulated or DNase I treated kidney sections. In contrast, both 9F11 and 11F8 bind to unmanipulated kidney sections and are localized to the glomeruli, and 11F8 exhibits a more intense staining pattern than 9F11. Importantly, the binding of both mAbs is abrogated by pretreatment of the kidney sections with DNase I as well as by preincubation of the mAbs with dT ₂₁ , which binds both mAbs with an apparent dissociation constant in the low nanomolar range. Collectively, these data provide strong evidence that anti-DNA specifically recognizes DNA epitopes localized to the glomeruli of the kidneys. This hypothesis is consistent with earlier studies demonstrating that binding of MRL-lpr serum to GBM antigens is inhibited by DNAse I pretreatment and reports that anti-DNA affinity helps determine the pattern and degree of glomerular immune-complex deposition. Although 11F8

displays stronger IF than 9F11 to sections of normal kidney, the difference is not as dramatic as in kidney sections obtained in adoptive transfer studies. The small difference in in vitro anti-DNA binding may be amplified in vivo by repetitive cycles of binding, tissue damage, and release and localization of nuclear antigens in the GBM. Several other factors that may potentially influence disease severity independent of antigen recognition were also investigated. To examine whether disease severity is conelated to the ability of the antibody to fix complement, the binding of C3 by these anti-DNA was measured and found to be similar for all three mAbs. This observation is consistent with data obtained in other laboratories and suggests that pathogenic and nonpathogenic anti-DNA cannot be distinguished by their ability to fix complement. The possibility that adoptive transfer of 11F8 initiates an anti-idiotype antibody cascade that itself mediates the pathogenicity associated with 11F8 also was explored. Although there are elevated levels of anti-DNA in the serum of mice inoculated with 11F8 hybridoma cells, there is no evidence of anti-idiotype antibodies. Lastly, the hypothesis of Izui and coworkers which proposes that IgG3 anti-DNA from MRL-lpr mice may possess cryoglobulin activity resulting in immune complex formation through Fc-Fc interactions was explored. However, purified 11F8 does not precipitate from solution, even after extended periods at 4°C, which suggests this mAb is not a cryoglobulin. Cloning and Expression of Recombinant Antibodies: Recombinant antibodies were produced using the generalized procedure described below. Antibody 4B2 is used as an example. The technology is essentially that developed by Winter as implemented in the Pharmacia Recombinant Antibody serum. mRNA Isolation, cDNA Synthesis and Primary PCR Amplification: polyA ⁺ mRNA was isolated from approximately 10 hybridoma cells using the FastTrack mRNA isolation kit (Invitrogen) following the manufacturer's protocols. First strand cDNA synthesis was conducted using the reagents supplied with Pharmacia Antibody System. Briefly, two identical reactions were performed, one for the V _H and one for the V _L . The reaction mixtures (3 μL polyA ⁺ mRNA, 180 U moloney leukemia virus reverse transcriptase, primer) were incubated in the supplied buffer for 1 hour at 37°C. The mixtures were then used for primary PCR amplification of the V _H and V _L sequences. For the light chain, V _L Primer Mix (2 μL) was added to the cDNA, while

V _H primer 1 (2 μL) and V _H primer 2 (2 μL) were added to the V _H cDNA (sequences of the V-region primers are proprietary). AmpliTaq DNA polymerase (1 μL, 5 U) was added to both reactions and subjected to 40 cycles of PCR amplification (94°C, 1 min; 55°C, 2 min; 72°C, 2 min). The PCR products were electrophoresed on a 1% agarose gel and stained with ethidium bromide. Product bands (~350 base-pairs long) were visualized by long wave UV inadiation, excised from the gel, and purified using a Sephaglas band prep kit.

PCR Joining and Secondary Amplification: The scF _v construct was built from the V _H and V _L genes as follows. Equimolar aliquots of the V _H and V _L PCR products (1 μL) along with the linker primer (0.5 μL) were dissolved in buffer containing dNTP's and AmpliTaq polymerase (2.5 U). Assembly of the three DNA fragments was conducted by 10 cycles of PCR (94°C, 1 min; 63 °C, 4 min). Restriction sites for cloning (Sfi I and Not I) were then introduced in another PCR reaction; after addition of the appropriate primers, dNTP's and AmpliTaq polymerage (2.5 U), the mixture was amplified over 40 cycles (94°C, 1 min; 55°C, 2 min, and 72°C, 2 min). The product was electrophoresed on a 1% agarose TAE gel and stained with ethidium bromide. The scF _v band ( approximately 750 base-pairs long) was visualized by UV inadiation, excised from the gel, and purified using a Sephacryl microspin column.

Restriction Digest ofscFy, Ligation intopCANTAB 5 E, and Transformation: The termini of the scF _v gene were then cleaved with the appropriate restriction enzymes for ligation into pCANTAB 5 E (Figure 11). Briefly, Sfi (20 U) was added to the scF _v gene (350 ng) and incubated for 4 hours at 50°C. Once the Sfi I digest was completed, the mixture was cooled to 37°C and Not I (40 U) was added. After incubation for 4 hours at 37°C, the DNA was purified using a microspin column. The scF _v was then ligated into the predigested phagemid using T4 DNA ligase (5 U) for 1 hour at 16°C. The ligase was heat denatured for 10 min at 70°C and the mixture was chilled on ice. For the transformation reaction, 1 mL of competent E. coli TGI cells (K12Δ(lac-pro), supE, thi, hsdD5/F', traD36, proAB, laclq, lacZΔM15) were pipetted into a centrifuge tube and the ligation reaction mixture from above was added. Following incubation on ice for 45 min, the mixture was heat shocked at 42°C for 2 min, and placed on ice for 5 min. As a control, an aliquot ofthis mixture (100 μL) was incubated in LBG media (1 mL) for 1 hour at 37°C and was then plated on SOBAG plates (SOB media

containing 100 μg/mL ampicillin and 111 mM glucose). Approximately 100 colonies were observed after an 18 hour incubation at 30°C which indicated that the scF _v gene was properly ligated into the phagemid.

Phage Rescue, Panning, and Reinfection: An aliquot of the 4B2 scFv transformation reaction mixture from above (800 μL) was pipetted into a sterile tube containing 2xYT-G media (15 mL of 2xYT containing 2% glucose). The sample was incubated at 37°C with shaking until growth of OD ₆₀₀ ™0.5 was reached (1 hour). At that point, ampicillin (75 μL of a 20 mg/mL solution) and M13KO7 helper phage (10 ¹⁰ pfu) were added to the cells resuspended on 2xYT-AK (10 mL of 2xYT containing 100 μg/mL ampicillin and 50 μg/mL kanamycin). After growth at 37°C for 18 h, the sample was centrifuged and the supernatant containing the recombinant phage was removed and transfened to a sterile tube for storage. The supernatant was filtered through a 0.45 m filter and stored at 4°C.

One round of panning was conducted to enrich for recombinant phage expressing the 4B2 scF _v gene. A sterile 25 cm tissue culture was coated with heat denatured calf thymus DNA (5 mL of a 10 μg/mL solution) overnight. After washing with TBS, the flask was blocked with 2% dry milk in PBS, pH 7.4 for 18 hours. After the appropriate washes, the supernatant from the phage rescue (8 mL) and blocking solution (8 mL) were added to the flask. The recombinant phage were allowed to bind to the antigen-coated surface for 3 hours at 37°C. After washing, competent TGI cells (10 mL of a culture grown to OD ₆₀₀ -0.3) was added to the flask and incubated for 1 hour at 37°C. An aliquot of the newly infected TGI cells was withdrawn and diluted 1 : 1 , 1 : 10, 1 : 100, and 1 : 1000. The diluted cells (100 μL) were grown on SOBAG plates at 30°C for 18 hours. Approximately 100 colonies grew on the 1:1 plate, 10 colonies on the 1 : 10 plate, and no colonies were observed on the other plates. Five colonies from the 1 : 1 plate were used to inoculate LBG overnight cultures (25 mL). Plasmid DNA was isolated from the cultures using the Promega Wizard Kit and restriction digest showed that four of the five colonies contained the 4B2 insert (which was then verified by DNA sequencing). Transformation of Non-SupE Bacteria for Expression: Approximately 2 ng pCANTAB 5 E-4B2 (recovered from TGI cells) was added to 1 mL of competent E. coli HB2151 cells [K12Δ(lac-pro), ara, nal ^r , thi/F* proAB, laqlq lacZΔM15] and

incubated on ice for 45 min. The sample was heat shocked for 2 minutes at 42°C, then chilled on ice for 5 min. An aliquot of the transformation reaction (100 μL) was added to LBG media (900 μL), incubated for 1 hour at 37°C, then plated (100 μL) on SOBAG-N media (SOB containing 100 μg/mL ampicillin, 100 μg/mL nalidixic acid, and 111 mM glucose). Roughly 50 colonies were observed after incubation for

18 hours at 30°C. To confirm that the HB2151 cells contained pCANTAB 5 E-4B2, four colonies from the SOBAG-N plates were used to inoculate SB-AG media (25 mL of SB media containing 100 μg/mL ampicillin and 2% glucose). After growth at 30°C for 12 hours, plasmid DNA was isolated and digested with Sfi I and Not I as described above. Analysis of the restriction digest on a 1% agarose gel showed as insert 750 base-pairs long along with the digested pCANTAB 5 E-4B2 running 4300 base-pairs in length. Glycerol stocks of these cultures were prepared for further use.

Re-growth and Induction: An SB-AG culture (5 mL) was inoculated with a sample from one of the glycerol stocks described above and grown at 30°C. After 12 hours, the culture was added to additional SB-AG media (50 mL) and the sample was incubated for 1 hour at 30°C with vigorous shaking. The sample was then pelleted at 1500g (15 min), and the pellet was resuspended in SB-AI media (50 mL of SB containing 100 μg/mL ampicillin and 1 mM IPTG). The culture was grown for 3 hours at 30°C with vigorous shaking, divided into two portions, and centrifuged at 1500g. The supernatant was removed and filtered through a 0.8 μm filter. One cell pellet was used to prepare a periplasmic extract while the other was used for the whole cell extract. For the periplasmic prep, the cells were resuspended in PBS (0.5 mL containing 1 mM EDTA), incubated on ice for 10 minutes, and then centrifuged to pellet the cell debris. The whole cell extract was prepared by resuspending the cell pellet in PBS (0.5 mL), boiling for 5 min, and centrifugation to pellet the cell debris. Aliquots of the supernatant, periplasm, and whole cell extract were analyzed by SDS- PAGE. A band migrating at 28 kDa was observed in the periplasmic extract indicating soluble expression of 4B2 scF _v (Figure 12).

DISCUSSION

As reported previously, mAbs in this panel are specific for DNA; they do not bind RNA, ribonucleoproteins, phospholipids, proteoglycans and ECM components present in normal glomerular basement membrane. These mAbs possess a high base selectivity for thymine, but the affinity and mode of DNA recognition differ among the mAbs. Specifically, anti-DNA with the highest affinity for oligo(dT) (e.g. 9F11 and 11F8) may bind ssDNA by recognizing the base moiety (in subsites) on the surface of the antigen combining site. Based on epitope mapping studies, however, 9F11 and 11F8 appear to subtly differ in the number and position of these putative subsites. Anti-DNA 4B2 binds both ssDNA and dsDNA, but recognizes ssDNA more weakly than 9F11 and 11F8. In addition, 4B2 appears to use distinct and different mechanisms to bind ssDNA and dsDNA.

Structural basis of anti-DNA specificity: To explore the relationship between gene structure and DNA specificity, the V genes of these anti-DNA were cloned and sequenced. MAbs 9F 11 , 15B 10, and 15D8 appear clonally related which is consistent with biophysical data showing these anti-DNA share very similar DNA binding properties. Although 11F8 has similar affimty and specificity for ssDNA as does 9F11, this mAb is encoded by an entirely different set of V genes. Two other well characterized thymine specific anti-ssDNA, BVO4-01 and HEdlO, possess ligand binding characteristics comparable to 9F11 and 11F8. Since BVO4-01 and HEdlO are encoded by V _H 10 and V _H J606, respectively, an interesting question is how DNA specificity and recognition can be so similar between the four anti-DNA despite the usage of distinct V region genes. One possibility is that V _H genes encode a limited repertoire of structurally conserved three dimensional motifs that form the basis of ssDNA specificity. This hypothesis is supported by recent crystallographic and modeling studies of anti-DNA which suggest that the shape of the antigen combining site may be conelated to the specificity for DNA.

There are several examples of anti-DNA that are homologous with the mAbs in this panel (Table 6). The V genes utilized by these mAbs occur in anti-DNA obtained from both autoimmune and nonautoimmune strains of mice. Importantly, both the

9F11 group of mAbs and 11F8 are encoded by V _H genes similar to those found in anti-

DNA obtained from normal mice immunized with bacterial DNA or protein-DNA complexes, respectively. The structural similarity between autoimmune and immunization-induced anti-DNA has been noted previously and suggests that the development of anti-DNA in normal and lupus-prone mice may occur by a common mechanism in which DNA or a protein-DNA complex stimulates appropriate B cells to differentiate in a receptor mediated response. The observation that mice immunized with Fusl-DNA complexes, from which bfd89 was derived, also develop mild nephritis supports this hypothesis and suggests that anti-DNA generated by immunization may have qualitatively similar features as autoantibodies with respect to renal pathology. However, mechanisms exist in normal individuals that may moderate the severity of the disease. Another interesting similarity is found between mAb 4B2 and 3H9, a related anti-DNA obtained from an MRL-lpr mouse. Both mAbs share the same V _H and V _κ genes and also recognize both ssDNA and dsDNA. These data suggest the clonal precursor that independently gave rise to these two anti-DNA may be abundant in MRL-lpr mice, or is preferentially selected during the course of the anti-DNA response. The fact that they share a common somatic mutation from gly to arg in HCDR2 indicates that their selection may be driven by the same antigen, and lends support to the latter possibility.

Although the V gene usage and structure of anti-DNA have been extensively analyzed, knowledge of the primary amino acid sequence alone yields little information about the involvement of specific residues in ligand recognition. When combined with affinity measurements and epitope mapping studies, however, these data together may help illuminate the relationship between anti-DNA structure and reactivity. For example, previous ligand binding studies of 9F11 suggest that aromatic residues, particularly trp, may by involved in ssDNA recognition through base stacking interactions, similar to that seen in studies of BVO4-01 by Voss et al. (1). Genetic analysis of 9F11 verifies the presence of a single (germline encoded) trp residue at position 33 of HCDRl . Although the trp residue in 9F11 is located in a different position than BVO4-01, the similarity in the mode of DNA recognition between 9F11 and BVO4-01 suggests that tip 33 in HCDRl of 9F11 may be involved in a similar interaction. Indeed, this hypothesis is supported by preliminary modeling studies. The importance of aromatic residues, particularly tip and tyr, in mediating DNA recognition

may be implied by the prevalence of these residues in the CDRs of anti-DNA. Specifically, 40% of the V _H J558 subgroups defined by Radic and Weigert encode a trp at position 33. In many cases, this is the only solvent accessible trp residues encoded in the V _H CDRS, opening the possibility that these anti-DNA may share a similar mode of DNA recognition as 9F 11.

The recunence of arg in anti-DNA through somatic mutation and V(D)J joining implies that it may contribute to antigen binding and has stimulated interest in defining the mechanism by which this residue mediates DNA recogmtion. Sequence analysis can identify the location of arg residues, but not the mechanism by which they participate in antigen recognition. In contrast, binding studies can describe the number of ion pairs formed upon complex formation and the specific phosphates on DNA that interact with anti-DNA, but cannot distinguish between the cationic residues arg and lys. Together, these studies provide the opportunity to define both the location and the putative role of arg residues in mediating DNA binding. For example, the cumulative evidence suggests that arg 53 in HCDR2 of 4B2 may contact DNA electrostatically. In contrast to 4B2, DNA binding by 9F11 apparently does not involve ion pair formation even though genetic analysis reveals the presence of arg at positions 96 and 98 in HCDR3. Other possibilities consistent with the binding data are that the arg form a hydrogen bonding network (either directly or water-mediated) to the DNA ligand or that they serve to enforce a specific conformation in the antigen combining site through intra- or interchain CDR contacts. Preliminary modeling studies support the latter alternative and suggest that arg may contribute directly or indirectly to antigen recognition and this role may depend on its location in the antigen combining site.

Table 1

MAb Isotype Pi Sm Sm/nRNP SS-A SS-B Histone Scl-70

7B3 IgG2a 7.1

8B8 IgG2a 6.9

4B2 IgG2a 6.9

10F4 IgG2a 6.7

9F11 IgG2b 7.0

15B10 IgG2b 6.8

15D8 IgG2b 6.8

11F8 IgG3 7.0

Table 1 (continued)

Crossreactivity of anti-DΝA. Binding to Sm, Sm nRΝP, SS-A, SS-B, histone and Scl-70 was assessed by direct ELISA as described in the experimental section. After 2 hours background reactivity of mAbs to wells containing no antigen was 0.089 AU at 405 nm. Serum samples containing autoantibodies to these antigens was used as a positive control. ( — ), <2x background. Crossreactivity to FΝ, ColIN, Lam, HS, PS, and CL was determined by direct ELISA as described herein. Background reactivity of mAbs to wells containing no antigen did not exceed an OD ₄₀₅ value of 0.08 AU after 2 hours. Monoclonal antibodies specific for FΝ, ColIN, Lam, and HS were obtained commercially and served as a positive control in these experiments. Similar results were obtained at different concentrations of anti-DΝA and antigen. ( — ), <2x background; (+), 2-5x background; (++), 5-10x background; (+++), >10x background.

Table 2

Reactivity of anti-DNA to nucleic acid polymers. Binding to RNA and DNA polymers was determined by direct ELISA as described herein. Data represents the mean of triplicate measurements taken after 2 hours at 405 nm. The backgroimd reactivity of mAbs to wells containing no antigen did not exceed 0.067 AU. See the legend to Table 1 for a description of the symbols.

Table 3

MAb dC ₂₁ dG ₁₅ α dT ₂₁ Duplex* 1

7B3 11000 4300

8D8 — 16600 96 —

4B2 4270 26400 817 7170

10F4 45500 7130 1470 -77000

9F11 «— 480 1 —

15B10 — 750 1.9 —

15D8 — 1860 5.5 —

11F8 3230 44 —

Duplex = Ligand

Relative affinity of anti-DNA for DNA oligomers. Relative affinity was measured using a gel shift assay as described herein. The fraction of free DNA was quantified from by densitometry from triplicate experiments and the protein concentration required to reach half saturation of the ligand was determined by linear interpolation of the data. The values are normalized to 9F11 (0.3 x 10 ^"9 M, per IgG

binding site). 9F11 is the tightest binder so that the larger values for the other mAbs represent weaker binding. While binding of more than one antibody to a single oligomer is theoretically possible, modeling studies suggest that the binding of consecutive antibodies is not sterically feasible. Moreover, the formation of multiple anti-DNA-DNA complexes has not been observed in the gel shift assay, and the relative affinity for the mAbs and their conesponding F(ab) fragments (per antigen binding site) are within enor (data not shown). The affimty of the mAbs for dT ₅ « dT ₆ which suggests that dT ₅ fills the binding site (data not shown). ( — ), no binding is detected at the highest concentration tested (>20 μM).

Table 4

Intrinsic affinity of anti-DNA. The intrinsic affimty was measured using a gel shift assay as described in the experimental section. The fraction of free DNA was quantified by densitometry from triplicate experiments and fit by nonlinear least squares regression to a single site binding isotherm. A representative example of the data and fit is shown in Figure 5. The K _d presented (10 ^" M) is per IgG binding site. ( — ), no binding detected at the highest protein concentration tested (20 μM); NM, not measured.

Table 5

Clone Isotype J558* D _H ^C J ^d V ^v κ ^e V ^v κ J ^s Subgroup family Subgroup

9F11 IgG2b J558 NA Q52, FL16.2 4 12, 13 NA 4

15B1 IgG2b J558 NA Q52, FL16.2 4 12, 13 NA 4

0 15D8 IgG2b J558 NA Q52, FL16.2 4 12, 13 NA 4

5F3 IgG2a J558 NA Q52, FL16.2 4 12, 13 NA 4

4B2 IgG2a J558 NA SP 2.7 3 5 5 1

7B3 IgG2a J558 4 SP 2, Q52 2 21 * 4

8D8 IgG2a J558 7 Q52r 4 5 5 1

10F4 IgG2a J558 8 SP 2.7 3 1 1 1

11F8 IgG3 Q52 ~ SP 2.5, 2.7 4 21 * 1

"Designation based on > 80% homology to members within N _H gene family described by Brodeur and Riblet (93).

* Assignment according to Radic and Weigert (15). ΝA; not assignable. ^C D gene assignments are based on homology to sequences determined by Kurosawa and

Tonegawa (94). D genes followed by (r) are inverted. ^Classification according to Sakano et al. (95). ^Designation according to Potter et al. (96).

"subgroups defined by Radic and Weigert (15). ΝA, not assignable; (*), assignment ambiguous. ^Classification according to Sakano et al. (97).

Table 6

Clone Related Anti-DΝA % Nucleotide Strain* Isotype Reactivity ¹⁷ Binding ^d

V _H Homology ⁰

9F11 SEC F2/3 93% BALB/c IgG2a,κ ss ECdsDNA ++

163-100 91% NZB/R IgM,κ ssDNA +++

4B2 bxwl4 99% NZB/W IgM,κ ssDNA +++

3H9 86% MRL-lpr IgG2b,κ ss/dsDNA +++

10F4 202-61 98% NZB W IgM,κ ssDNA -H-r

7B3 H161 100% MRL-lpr IgG3,κ ssDNA ++*

8D8 25-12m 98% NZB W IgM,κ ssDNA ++4-

11F8 bfd89 99% BALB/c IgM,κ ssDNA* +++

9F11 17-S166 97% NZB W IgM,κ ssDNA +++

4B2 3H9 99% MRL-lpr IgG2b,κ ss/dsDNA +++

10F4 MRL-10 99% MRL-lpr IgM,κ ssDNA +++

7B3

8D8 3H9 98% MRL-lpr IgM,κ ssDNA +-H-

11F8 SECF5/1 97% BALB/c IgGl,κ ECDNA* + nucleotide homologies are determined for codon 9 through codon 94. *NZB/R, (NZWxSWR)Fj; NZB/W, (NZWxNZW)F,; MRL-lpr, MRL MpJ Ipr/lpr ^c ssDNA, single-stranded DNA; dsDNA, double-stranded DNA; ECdsDNA, E. coli dsDNA;

^Assessed by competition or quantity of mAb needed to achieve given OD.

+++, <1 μg/mL; ++, 1-10 μg/mL; +, > 10 μg/mL.

*, obtained by immunization; %, nephritogenic in non-autoimmune mice.

Table 7

Hybridoma Isotype Serum anti- Serum anti- Serum IgG Ascites

DNA" DNA cone* IgG cone

Pre-injection Post-transfer (mg/mL) (mg/mL)

(Mean OD ₄₀₅ ) (mean OD ₄ o ₅ )

HOPC IF/12 IgG2a 0.108 0.151 20.9 -50

4B2 IgG2a 0.113 1.65 5 7.5

9F11 IgG2b 0.165 2.31 10.4 9.1

11F8 IgG3 0.141 1.34 7.4 5.6

Hybridoma C fixation ⁰ Proteinuria'' Hemanuria Kidney mg/dL erythrocytes Deposition* per μL

HOPC IF/12 NM 100 10-50 none 0

4B2 2.25 (0.117) 100 50-250 mes (weak) 1

9F11 2.36 (0.106) 100-500 50-250 mes (weak) 1

11F8 1.4 (0.116) 100-500 50-250 mes/subend 4+ o (strong)

^a Serum or ascites was diluted 1:100 and measured for binding to heat denatured CT DNA by ELISA. The value presented is the mean of triplicate measurements from the mice tested. The background using BALB/c sera diluted 1:100 did not exceed 0.2 AU. ^έ The concentration of IgG was determined using a commercially available ELISA kit

(Boehringer Mannheim).

Complement fixation as assayed by binding of murine C'3 in ELISA. Presented are means of triplicate measurements taken at 405 nm after 15 min. The values obtained using heat denatured samples are shown in parentheses. NM, not measured.

^Proteinuria was estimated using Chemistrip 6. ^e Mes, mesangial; subendo, subendothelial.

^■ ^As assessed by direct immunofluorescence with FITC conjugated goat anti-mouse IgG and electron microscopy.

References

I . Noss, E.W. Jr. 1990 Anti-DΝA antibodies in SLE. E. W. Noss, Jr. ed. CRC Press, Boca Raton, FL. 2. Tan, E. M. 1989. Antinuclear antibodies: diagnostic markers for autoimmune diseases and probes for cell biology. Adv. Immunol. 44:93.

3. Koffler, D., et al. 1967. Immunological studies concerning the nephritis of systemic lupus erythematosus. J. Exp. Med. 126:607.

4. Foster, M. H. et al. 1993. Νephritogenic autoantibodies in systemic lupus erythematosus: inununochemical properties, mechanisms of immune deposition, and genetic origins. Lab. Invest. 69:494.

5. Nlakos, D.N. et al. 1992. Anti-DΝA antibodies form immune deposits at distinct glomerular and vascular sites. Kidney Int. 41:1690.

6. Tsao, B. P. et al. 1990. Structural characteristics of the variable regions of immunoglobulin genes encoding a pathogenic autoantibody in murine lupus. J. Clin.

Invest. 85:530.

7. Katz, J. B. et al. 1994. Mutational analysis of an autoantibody: differential binding and pathogenicity. J. Exp. Med. 180:925.

8. Ben-Chetrit, E. et al. 1988. Specific inhibition of the DΝAanti-DΝA immune reaction by low molecular weight anionic compounds. Immunology 65:479™.

9. Jones, D. S. et al. 1994. Conjugates of double-stranded oligonucleotides with polyethylene glycol) and keyhole limpet hemocyanin: a model for treating systemic lupus erythematosus. Bioconjugate Chem. 5:390.

10. Stollar, D., L. et al. 1962. The antigenic determinants of denatured DΝA reactive with lupus erythematosus serum. Proc. Natl. Acad. Sci. USA 48:874.

I I . Koffler, D. et al. 1971. Antibodies to polynucleotides in human sera: antigenic specificity and relation to disease. J. Exp. Med. 134:294.

12. Eilat, D. 1982. Monoclonal antibodies: an approach to studying autoimmune disease. Mol. Immunol. 19:943. 13. Taki, S. et al. 1992. Somatically mutated IgG anti-DΝA antibody clonary related to germ-line encoded IgM anti-DΝA antibody. Eur. J. Immunol. 22:987.

14. Marion, T. N. et al. 1992. Selection of inimunoglobulin variable regions in autoimmunity to DNA. Immunol. Reviews 128:123.

15. Radic, M. Z., and M. Weigert, M. 1994. Genetic and structural evidence for antigen selection of anti-DNA antibodies. Ann. Rev. Immunol. 12: 487 and references therein.

16. Eilat, D., and W. F. Anderson. 1994. Structure -function conelates of autoantibodies to nucleic acids. Lessons from immunochemical, genetic and structural studies., Mol. Immunol. 184:1377.

17. Henon, J. N. et al. 199 1. An autoantibody to single-stranded DNA: comparison of the three-dimensional structures of the unliganded Fab and a deoxynucleotide-Fab complex. Proteins 11:159.

18. Swanson, P.C., B. C. Cooper, and G. D. Glick. 1994. High resolution epitope mapping of an anti-DNA autoantibody using model DNA ligands. J. Immunol. 152: 2601. 19. Smith, R. G. et al. 1989. Structural features of a murine monoclonal anti- ssDNA autoantibody. J. Indian Inst. Sci. 69:25.

20. Lee, J. S., D. F. Dombroski, and T. R. Mosmann. 1982. Specificity of autoimmune monoclonal Fab fragments binding to single-stranded deoxyribonucleic acid. Biochemistry 21:4940. 21. Stollar, B. D., G. Zon, and R. W. Pastor. 1986. A recognition site on synthetic helical oligonucleotides for monoclonal anti-native DNA autoantibody. Proc. Natl.

Acad Sci. USA. 83:4469.

22. Braun R. P., and J. S. Lee. 1-987. Equilibrium binding parameters of an autoimmune monoclonal antibody specific for double-stranded DNA. J. Immunol. 139:175.

23. Oi, V. T., L. A. Herzenberg. 1980. In Selected methods in cellular immunology. B. B. Mishell and S. M. Shiigi, eds. W. H. Freeman, New York, NY. p. 29 1.

24. Margulies, D. H., W. M. Kuehl, and M. D. Scharff. 1976 Somatic cell hybridization of mouse myeloma cells. Cell 8:405. 25. Papalian, M., E. Lafer, R. Wong, and B. D. Stollar. 1980 Reaction of systemic lupus erythematosus antinative DNA antibodies with native DNA fragments from 20 to 1200 base pairs. J. Clin. Invest. 65:469.

26. Glick, G. D. et al. 1992. Trapping and isolation of an alternate DNA conformation. J. Am. Chem. Soc. 1 4:5447.

27. Harlow, E. and D. Lane. 1988. In Antibodies: A laboratory manual Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. p. 84. 28. Zoali, M., and B. D. Stollar. 1986. A rapid ELISA for measurement of antibodies to nucleic acid antigens using UV-treated polystyrene microplates. J. Immunol. Methods 90:105.

29. Eaton, R. B. et al. 1983. Enzyme immunoassay for antibodies to native DNA. Arthritis Rheum. 26:52. 30. Smeenk, R. J. et al. 1988. Reaction patterns of monoclonal antibodies to DNA. J. Immunol. 140-3786.

31. Sabbaga, J., S. R. Line, P. Potocnjak, M. P. Madaio. 1989. A murine nephritogenic monoclonal anti-DNA autoantibody binding directly to mouse laminin, the major noncollagenous component of the glomerular basement membrane. Eur. J. Immunol. 19:137.

32. Gay, S., M. J. Losman, W. J. Koopman, and E J. Miller. 1985. Interaction of DNA with connective tissue matrix proteins reveals preferential binding to type V collagen. J. Immunol. 129:1097.

33. Faaber, P. et al. 1986. Crossreactivity of human and murine anti-DNA antibodies with heparan sulfate, the major glycosaminoglycan in glomerular basement membranes. J. Clin. Invest. 77:1824.

34. Lake, R. A. et al. 1985. A key role for fibronectin in the sequential binding of native dsDNA and monoclonal anti-DNA antibodies to components of the extracellular matrix: its possible significance in glomerulonephritis. Immunology 54:389. 35. Stevens, S. Y et al. 1993. Evidence for induced fit in antibody-DNA complexes. J. Am. Chem. Soc. 115:1585.

36. Kelly, R. C, D. E. Jensen, and P. H. von Hippel. 1976. DNA melting proteins.

IV. Fluorescence measurements of binding parameters for bacteriophage T4 gene 32- protein to mono-, oligo-, and polynucleotides. J. Biol. Chem. 251:7240. 37. Kim, Y. T. et al. 1992. Purification and characterization of the bacteriophage

T7 gene 2.5 protein. J. Biol. Chem. 267:15022.

38. Andrew, S. M. and J. A. Titus. 1991. Antibody detection and preparation. In Current Protocols in Immunology, Vol. 1. J. E. Coligan, A. M. Kruisbeek, D. H. Marguilies, E. M. Shevach, and W. Strober, eds. John Wiley and Sons, New York, NY.

39. Eilat, D. 1985. Cross-reactions of anti-DNA and the central dogma of lupus nephritis. Immunology Today 6:123.

40. Schwartz, R. S., and B. D. Stollar. 1985. Origins of anti-DNA autoantibodies. J. Clin. Invest. 75:321.

41. Mohan, C. et al. 1993. Nucleosome: a major inimunogen for pathogenic autoantibody-inducing T cells of lupus. J. Exp. Med. 177.1367. 42. Madaio, M. P. et al. 1987. Murine monoclonal anti-DNA antibodies bind directly to glomerular antigens and form immune deposits. J. Immunol. 138:2883.

43. Pankewycz, 0. G., P. Mgliorini, and M. P. Madaio. 1987. Polyreactive autoantibodies are nephritogenic in murine lupus nephritis. J. Immunol. 139:3287.

44. Termaat, J-H. et al. 1992. Anti-DNA antibodies can bind to the glomerulus via two distinct mechanisms. Kidney Intl. 42:1363.

45. Ohnishi, K. et al. 1994. Comparison of pathogenic and non-pathogenic murine antibodies to DNA: antigen binding and structural characteristics. Int. Immunol. 6:817.

46. Edberg, J. C, and R. P. Taylor. 1986. Quantitative aspects of lupus anti-DNA autoantibody specificity. J. Immunol. 136:4581. 47. Costello, P. B., and F. A. Green. 1988. Binding affinity of serum immunoglobulin G to cardiolipin and other phospholipids in patients with systemic lupus erythematosus and syphilis. Infection and Immunity 56:1738. 48. Brinkman, K et al. 1990. Anti-DNA antibodies and lupus nephritis: the complexity of crossreactivity. Immunology Today 11:232. 49. Andrejewski, A., Jr. et al. 1981. Antigen-binding diversity and idiotypic cross- reactions among hybridoma autoantibodies to DNA. J. Immunol. 126.226.

50. Amott, S. 1970. The geometry of nucleic acids. Prog. Biophys. Molec. Biol. 21:265.

51. Pesce, A. J., and J. G. Michael. 1992. Artifacts and limitations of enzyme immunoassay. J. Immunol. Methods 150:111.

52. Sanford, D. G. and B. D. Stollar. 1992. Assay of anti-DNA antibodies. Methods EnzymoL 212:355.

53. Fried, M. G. and D. M. Crothers. 1981. Equilibria and kinetics of lac repressor -operator interactions by polyacrylamide gel electrophoresis. Nucleic Acids Res. 9:6505.

54. Tetin, S. Y. et al. 1993. Elucidation of anti ssDNA autoantibody BVO4-01 binding interactions with homooligonucleotides. Biochemistry 32:9011.

55. Wang, H. et al. 1994. Solution structure of a disulfide crosslinked DNA hairpin. J. Am. Chem. Soc. 116:5022.

56. Wing, R. et al. 1980. Crystal structure analysis of a complete turn of B-DNA. Nature 287:755. 57. Weeks, K. M., and D. M. Crothers. 1992. RNA binding assays for Tat-derived peptides: implications for specificity. Biochemistry 31:10281.

58. Record, M. T. Jr., T. M. Lohman, and P. de Haseth. 1976. Ion effects on ligand- nucleic acid interactions. J. Mol. Biol. 107:145.

59. Lohman, T. M., P. L. de Haseth, and M. T. Record. 1980. Pentalysine- deoxyribonucleic acid interactions: a model for the general effects of ion concentrations on the interactions of proteins with nucleic acids. Biochemistry 19:3522.

60. Radic, M. Z. et al. 1993. Residues that mediate DNA binding of autoimmune antibodies. J. ImmunoL 150:4966.

61. Chothia, C. and A. M. Lesk. 1987. Canonical structures for the hypervariable regions of immunoglobulins. J. Mol. Biol. 196:901.

62. Stanfield, R. L. et al. 1990. Crystal structure of an antibody to a peptide and its complex with peptide antigen at 2.8 A. Science 248:712.

63. Kowalczykowski, S. C. 1986. Cooperative and noncooperative binding of protein ligands to nucleic acid lattices: experimental approaches to the determination of thermodynamic parameters. Biochemistry 25:1226.

64. Rhodes, D. 1989. Analysis of sequence-specific DNA binding proteins. In Protein Function: A Practical Approach. T.E. Creighton, ed. IRL Press, Oxford, p. 177.

65. Rini, J. M. et al. 1992. Structural evidence for induced fit as a mechanism for antibody-antigen recognition. Science 255:959.

66. Getzoff, E. D. et al. 1988. The chemistry and mechanism of antibody binding to protein antigens. Adv. Immunol. 43:1.

67. Draper, D. E. 1993. Protein-DNA complexes: the cost of recognition. Proc. Natl. Acad. Sci. USA 90:7429.

68. Lesser, D. R., M. R. Kurpiewski, and L. Jen- Jacobson. 1990. The energetic basis of specificity in the Eco RI endonuclease-DNA interaction. Science 250.776. 69. La Baer, J. and K. R. Yamamoto. 1994. Analysis of the DNA-binding affinity, sequence specificity and context dependence of the glucocorticoid receptor zinc finger region. J. Mol. Biol. 239:664.

70. Isenberg D. A. et al. 1994. The origin, sequence, structure and consequences of developing anti-DNA antibodies. A human perspective. Arthritis Rheum. 37:169. 71. Madaio, M. P. et al. 1984. Responsiveness of autoimmune and normal mice to nucleic acid antigens. J. Immunol. 132:872.

72. Erikson, J. et al. 1991. Expression of anti-DNA immunoglobulin transgenes in non-autoimmune mice. Nature 349:331.

73. Watanabe-Fukunaga, R., C. et al. 1992. Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 296:314.

74. Strasser, A., et al. 1991. Enforced bcl2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease. Proc. Natl. Acad. Sci. USA 88:8661.

75. Polymenis, M. and B. D. Stollar. 1994. Critical binding site amino acids of anti-Z-DNA single chain Fv molecules. Role of heavy and light chain CDR3 and relationship to autoantibody activity. J. Immunol. 152:5318.

76. Rumbley, C. A., et al. 1993. Construction, characterization, and selected site- specific mutagenesis of an anti-single-stranded DNA single-chain autoantibody. J. Biol. Chem. 268:13667. 77. Sharon, J. 1990. Structural conelates of high antibody affinity: three engineered amino acid substitutions can increase the affinity of an anti-p-azophenylarsonate antibody 200-fold. Proc. Natl. Acad. Sci. USA 87:4814.

78. Schillbach, J. F., et al. 1993. Modulation of antibody affinity by a non-contact residue. Protein Science 2:206. 79. Shefner, R, G. Kleiner, A. Turken, L. Papazian, and B. Diamond. 1991. A novel class of anti-DNA antibodies identified in BALB/c mice. J. Exp. Med. 173:287.

80. Foster, M. H., M. P. Madaio, and K. J. Baffett. 1992. Variable region sequence analysis of anti-DNA antibodies: evidence for a family of closely related germ-line VH genes encoding lupus autoantibodies. DNA Cell Biol. 11:175.

81. Sanger, W. and U. Heinemann. 1989. Protein-Nucleic Acid Interaction. CRC Press, Boca Raton, FL.

82. Stevens, S. Y., P. C. Swanson, and G. D. Glick. 1994. Application of the gel shift assay to study the affinity and specificity of anti-DNA antibodies. J. Immunol. Methods 177:185.

83. Polymenis, M. and B. D. Stollar. 1995. Domain interactions and antigen binding of recombinant anti-Z-DNA antibody variable domains. The role of heavy and light chains measured by surface plasmon resonance. J. Immunol. 154:2198.

84. Sanford, S. G., and B. D. Stollar. 1990. Characterization of anti-Z-DNA antibody binding sites on Z-DNA by nuclear magnetic resonance spectroscopy. J. Biol. Chem. 265:18608. 85. Cygler, M., et al. 1987. Crystallization and structure of an autoimmune anti- poly(dT) immunoglobulin Fab fragment at 3.0 A resolution. J. Biol. Chem. 262:643. 86. Record, M. T., Jr. and R. S. Spolar. 1990. Some thermodyna™mic principles of nonspecific and site specific protein-DNA interactions. In The biology of nonspecific DNA-protein interactions. A. Rezvin, ed. CRC Press, Boca Raton, FL, p. 33. 87. Bhat, T. N., et al. 1994. Bound water molecules and conformational stabilization mediate an antigen-antibody association. Proc. Natl. Acad. Sci. 91:1089.

88. Mol, C. D., et al. 1994. Structure of an immunoglobulin Fab fragment specific for poly(dG)-ρoly(dC). J. Biol. Chem. 269:3605.

89. Lohman, T. M. and W. Bujalowski. 1990. Escherichia coli single strand binding protein: multiple single-stranded DNA binding modes and cooperativities. In The biology of nonspecific DNA-protein interactions. A. Rezvin, ed. CRC Press, Boca Raton, FL, p. 131.

90. Karpel, R. L. 1990. T4 bacteriophage gene 32 protein. In The biology of nonspecific DNA-protein interactions. A. Revzin, ed. CRC Press, Boca Raton, FL, p. 103.

91. Skinner, M. M., et al. 1994. Structure of the gene V protein of bacteriophage fl determined by multiwavelength x-ray diffraction on the selenomethionyl protein. Proc. Natl. Acad. Sci. USA 91:2071.

92. Pabo, C. O., and R. T. Sauer. 1992. Transcription factors: structural families and principles of DNA recognition. Annu. Rev. Biochem. 61:1053.

93. Brodeur and Riblet. 1984. The immunoglobulin heavy chain variable region (Igh-V) locus in the mouse. Eur. J. Immun. 14:922-930.

94. Kurosawa and Tonegawa. 1982. Organization, structure, and assembly of immimoglobulin heavy chain diversity DNA segments. J. Exp. Med. 155:201-288. 95. Sakano et al. 1980. Two types of somatic recombination are necessary for the generation of complete heavy-chain genes. Nature 286:676-683. 96. Potter et al. 1982. Classification of mouse VK groups based on the partial amino acid sequence to the first invariant tryptophan: impact of 14 new sequences from IgG myeloma proteins. Mol. Immunol. 19:1619-1630. 97. Sakano et al. 1979. Sequences at the somatic recombination sites of immimoglobulin light-chain genes. Nature 280:288-294.