OF B I O ~ I C A L C m m r
`ItE J ~ ~ ~ R N A L
`Q 1994 by The American Society for Biochemistry and Molecular Biology, Inc.
`
`Vol. 269, No. 1, Issue of January pp. 734-738,1994
`Printed in U S A .
`
`Antigen Recognition by an Antibody Light Chain"
`
`(Received for publication, August 9, 1993, and in revised form, September 16, 1993)
`
`Mei Sun, Lan Li, Qing Sheng Gao, and Sudhir Pad$
`From the Departments of Anesthesiology, Internal Medicine and Pathology, University of Nebraska Medical Centel;
`Omaha, Nebraska 68198
`
`A monoclonal antibody to vasoactive intestinal poly-
`the heavy chain makes the major contribution in antigen bind-
`peptide (VIP) was reduced and alkylated and its light
`ing, with the light chain serving as a relatively nonspecific
`and heavy chains were purified by denaturing gel filtra-
`partner.
`tion. Following renaturation, the light chain displayed
`The variable regions of several light chains
`possess se-
`sequence-specific binding of VIP. The specific VIP bind-
`quences notably similar to active site sequences found in some
`ing activity of several fractions spanning the light chain
`serine proteases (15). The demonstration of VIP' hydrolysis by
`peak recovered from the gel filtration column was con-
`light chains purified from human autoantibodies (16) suggests
`stant, the light chain was electrophoretically homogene-
`the potential utility of these light chains as proteases, either
`ous, the VIP binding activity was precipitated by anti-
`alone or in combination with noncatalytic heavy chains of a
`light chain antibody but not anti-heavy chain antibody
`different specificity, These reagents can be predicted to display
`and the activity remained associated with a light chain
`the
`at least in part on
`substrate specificities dependent
`fraction recovered by resolutive chromatography on a
`strength and specificity with which the light chains bind VIP.
`hydroxylapatite column. N-terminal amino acid
`se-
`We undertook, therefore, a detailed study of the binding prop-
`quencing of the light and heavy chain fractions con-
`erties of a light chain purified from a monoclonal antibody to
`firmed the purity of these proteins and suggested that
`the V, and VH regions belonged to rc-family I1 and y-fam-
`VIP. We observed that this light chain displays sequence-spe-
`ily III, respectively. The VIP-binding affinity of the light
` binding of VIP. This observation suggests
`cific and high e
`t
`y
`that it should be possible to develop novel light chain reagents
`chain was only &fold lower than that of the parent an-
`with a n t i g e n - s ~ f i c activity, including a catalytic activity, and
`tibody and the light chain did not bind unrelated pep-
`tides. These observations suggest that light chains dis-
`permits conception of an antigen-specific biological function for
`free light chains in vivo.
`for high
`play structural characteristics necessary
`aEnity antigen binding.
`
`EXPERIMENTAL PROCEDURES
`Antibody Subunits-Monoclonal antibody c23.5 was raised by immu-
`nization of mice with VIP conjugated via N H 2 groups to keyhole limpet
`hemocyanin and purified from ascites fluid by ammonium sulfate pre-
`cipitation and affinity chromatography on protein G-Sepharose chro-
`matography (17). Nonimmune I&,,,
`K secreted by a myeloma served as
`the control protein (UPC10, Sigma; the isotype of the control antibody
`and the anti-VIP antibody are identical). The two types of antibodies
`were reduced and alkylated by a method similar to that described in
`Ref. 4. Briefly, 5 mg of antibody was reduced with Z - m e ~ a p t ~ t ~ n o l
`(0.2 M, 3 h) followed by i ~ o a c e ~ m i d e
`(0.3 M, 15 minf in 0.05 M Tris-HC1,
`0.15 M NaCl with the pH maintained at 8.0 using Tris base. The reaction
`mixture was concentrated (Centriprep-10, Amicon) and immediately
`separated by high performance gel filtration in 6 M guanidinium chlo-
`ride, pH 6.5, on two Superose-12 columns (Pharmacia) connected in
`series (flow rate, 0.4 dmjn). The fractions were dialyzed against 50 m~
`Tris-HCI, 0.02% sodium azide, pH 7.3 (12-14 kDa cut-off), for 4 days
`(4 "C) with four buffer changes. Assuming complete equilibration across
`the dialysis membrane, the final guanidinium concentration in the re-
`natured protein solutions was 1 w. Chromatography of renatured light
`chains on hydroxylapatite (Bio-Gel HPHT 100 x 7.8 mm, Bio-Rad) was
`done at pH 6.8 using a gradient of sodium phosphate (10-300 m)
`calcium chloride, O.Oh% sodium azide, pH 6.8. Elec-
`dissolved in 10
`trophoresis was on gradient polyacrylamide gels (&25%) using a Phast
`system (Pharmacia LXB Biotechnology Inc.). Immunobfotting of the
`g& was with rabbit anti-mouse heavy (Fct chain (Axelif and light ( K )
`chain antibodies (Cappel) followed by staining with a goat anti-rabbit
`IgG-peroxidase conjugate (Cappel) (18).
`Protein Sequencing-The purified antibody subunits (10 1.18 each)
`were adsorbed on polyvinylidine difluoride membranes (Pro%tt car-
`tridges; Applied Biosystems) according to instructions supplied by the
`manufacturer and their N-terminal amino acid sequence was deter-
`mined using a pulsed liquid phase sequenator with on-line phenylthio-
`hydantoin-derivative detection (Applied Biosystems, model 477A). With
`P-lactalbumin as standard, the initial yield for sequencing was 53%,
`
`Antigen binding sites in antibodies are formed by the vari-
`able regions of light and heavy chains. Delineation of antigen
`interactions with the individual subunits of antibodies is of
`interest for several reasons, including: (i) the presence of free
`light chains in B-lymphocytes (1) and in the circulation of pa-
`tients with myelomas (B-lymphocyte tumors) @), and lii) deri-
`vation of high atlinity antibodies in vitro from randomly com-
`bined libraries of light and heavy chains is dependent on
`pairing of subunits that display appropriate interactions with
`each other and the antigen (3). Purified heavy chains (4-8) and
`heavy chain variable regions can bind antigens (9) with affini-
`ties approaching those of native antibodies. Light chains puri-
`fied from polyclonal antibodies, in contrast, display little ( 6 - 8 )
`or no antigen binding activity (4, 5). Dimers of light chain
`secreted by myeloma cells are known to bind haptens. However,
`their hapten-binding affinities are several orders of magnitude
`smaller than native antibodies (10-12) and the binding is USU-
`ally interpreted to be fortuitous and unrelated to the activity of
`light chains in vivo. Study of crystals of antigen-antibody com-
`plexes has shown that the number and area of antigen contacts
`at light and heavy chain residues is comparable (131, but it has
`been suggested that the subset of residues that contribute to
`the free energy of binding are located primarily in the heavy
`chain (14). These observations have lead to suggestions that
`
`* This work was supported by Grants HL 44126, AI 31268, and K
`02217 from the National Institutes of Health and a contract with IGEN,
`Inc. The costs of publication of this article were defrayed in part by the
`payment of page charges. This article must therefore be hereby marked
`to
`"advertisement" in accordance with 18 U.S.C. Section 1734 solely
`1 The abbreviations used are: VIP, vasoactive intestinal peptide;
`indicate this fact.
`$ To whom correspondence should be addressed. Fax: 402-559-5592.
`CDRs, complementarity determining regions.
`734
`
`Lassen - Exhibit 1051, p. 1
`
`

`

`Antigen Binding by Light Chains
`B 7
`A
`
`735
`
`0.31
`
`Heavy
`
`h E
`0 0.2-
`W cu
`
`v
`W
`0
`Z
`4 m
`ul
`0
`6
`
`0.1.
`
`0 0
`
`4
`
`60
`RETENTION TIME, MIN
`FIG. 1. Separation of the heavy and light chaine from anti-VIP antibody by gel filtration. Monoclonal antibody c23.5 (5 mg) was reduced
`and alkylated and chromatographed in 6 M guanidinium chloride, pH 6.5, on two Superose-12 gel filtration columns (Pharmacia) connected in series
`(A). Pooled fractions corresponding to the heavy chain peak (retention time: 54-56 min) and the light chain peak (retention time: 60-66 min) were
`electrophoresed (8-25% polyacrylamide gels) under reducing conditions and stained with silver (lanes 2 and 3, respectively) ( B ) . Lune 1 shows a
`silver-stained gel of the parent antibody electrophoresed under reducing conditions. Lanes 4 and 5 are immunoblots of the light chain fraction
`stained with antibody to mouse light chain and heavy chain, respectively.
`
`I
`80
`
`and the repetitive yield was >go%. The sequence data reported here are
`based on amino acid yields >20 pmol in individual sequencing cycles.
`[1251-T)r'0JVlP Binding-Synthetic VIP (HSDAVETDNWRLRKQ-
`MAVKKYLNSILN-NH2; peptide content 81%, Bachem) was radioiodin-
`ated using chloramine-T and ['251-Tyr10MP was purified and identified
`as described (19). Binding of [lwI-Tyrlo]VIP (approximately 0.05 m) by
`intact antibody or renatured antibody subunit fractions was measured
`in duplicate according to Ref. 20 with the following modifications: (i)
`aRer incubation with radiolabeled peptide, 25 pl of human y-globulins
`(100 pg; Sigma) was added as carrier protein, and (ii) protein-bound VIP
`was precipitated by addition of polyethylene glycol (M, 8,000; Sigma) to
`20% (w/v). In some assays, bound [1251-Tyr'01VIP was precipitated with
`rabbit antibodies to mouse light chains (K) (Cappel) or Fc (Axell) (50 pl
`of a 40 pg/ml dilution) (20). The binding data were corrected for non-
`specific binding (<5% of available radioactivity) determined by incuba-
`tion in the presence of 1 1.1~ unlabeled VIP.
`
`RESULTS
`VIP Binding by Purified Antibody Subunits-A monoclonal
`antibody to VIP (clone c23.5) was reduced and acetylated in a
`nondenaturing buffer, solvent conditions described to favor re-
`duction of intersubunit s-S bonds and minimize reduction of
`intrasubunit S-S bonds (4). Since heavy and light chains form
`dimers and higher order aggregates by noncovalent interac-
`tions (211, they were separated by gel filtration in a denaturing
`solution (6 M guanidinium chloride) (Fig. 1). SDS-electrophore-
`sis showed a single silver-stained 60-kDa band in the heavy
`chain fraction and a 25-kDa band in the light chain fraction.
`f i r a second round of gel filtration, guanidinium chloride was
`removed by dialysis and saturable binding of [1251-Tyr10JVIP
`(binding displaced by excess unlabeled VIP) by the refolded
`light and heavy chains was measured. Both proteins displayed
`binding activity. The specific binding activity was essentially
`constant across the width of the light chain peak recovered
`from the gel filtration column (fractions 58-61; 4964, 4830,
`4210, and 4967 cpdpg of Gchains, respectively) (Fig. 21, sug-
`gesting that the activity is attributable to the light chain. Simi-
`larly, fractions spanning the heavy chain peak displayed com-
`parable specific activities (fractions 52-55; 1504, 2026, 1800,
`and 1487 cpdpg).
`
`Since strong antigen binding by light chains has not been
`described previously, additional control experiments were per-
`formed. Immunoblotting of SDS gels revealed staining of the
`25-kDa light chain band with anti-light chain antibody. Stain-
`ing with anti-heavy chain antibody was undetectable. An anti-
`body to mouse light chain (K) precipitated the [1251-Tyr10]VIP
`binding activity (2,325 2 149 cpm) of the light chain (1.2 pg/
`assay), but essentially no binding activity was detected by pre-
`cipitation with an equivalent concentration of antibody to the
`heavy chain. [1251-Tyr10]VIP complexed with the parent anti-
`body (0.15 m) was precipitated effectively by both types of
`antibodies (anti-light chain antibody, 4782
`580 cpm; anti-
`heavy chain antibody, 3056 * 660 cpm). We concluded that the
`binding was not due to trace contamination with heavy chains
`or heavy-light chain complexes.
`A single protein sequence was detected in the antibody sub-
`unit preparations by N-terminal amino acid sequencing. The
`sequence of the N-terminal27 residues of the light chain was:
`Asp-Val-Val-Met-Thr-Gln-Thr-Pro-Leu-Thr-Leu-Ser-Val-Thr-
`Ile-Gly-Gln-Pro-Ala-Ser-Ile-Ser-X-Lys-Ser-Ser-Gln,
`and of the
`N-terminal 29 residues of the heavy chain, Glu-Val-Lys-Leu-
`Val-Glu-Ser-Gly-Gly-Gly-Leu-Val-Lys-Pro-Gly-Gly-Ser-Leu-
`Lys-Leu-Ser-X-Ala-Ala-Ser-Gly-Phe-Thr-Phe (X, unidentified
`residues). The amino acids on the C-terminal side of light chain
`residue 27 (Gln) and heavy chain residue 29 (Phe) could not be
`identified with certainty due to diminished amino acid yields
`(<lo pmol). A comparison of the N-terminal sequences with
`sequence data in Ref. 22 suggested that the light chain was a
`member of K-chain family 11. The anti-VIP light chain contains
`all 10 invariant residues found at corresponding positions in
`the N-terminal region of this family and its remaining 16 resi-
`dues are also found among other members of this family. Simi-
`lar analysis suggested that the heavy chain variable region
`belonged to the y-chain family IIIA. As in the case of the light
`chain, all of the N-terminal amino acids of the anti-VIP heavy
`chain are found at corresponding positions among members of
`family IIIA. The X residues at positions 23 in the light chain
`
`Lassen - Exhibit 1051, p. 2
`
`

`

`736
`
`Antigen Binding by Light Chains
`
`A 5oI
`
`FIG. 2. Saturable binding of ['"I-
`T~r'~]vIp by purified light chains ( A )
`and heavy chain8 (R). Pooled light
`chains (500 pg) and heavy chains (135 pg)
`from Fig. 1 were rechromatographed in 6
`u panidinium chloride on a gel filtration
`column (Superose-12). The fractions were
`renatured by dialysis and binding of ['251-
`Tyr"'lVIP (0.1 nu) by duplicate aliquota
`(light chains. 20 pl; heavy chains, 50 pl) of
`the column fractions was measured (see
`'Experimental Procedures").
`
`RETENTION
`
`"",,',
`
`z.
`. _.
`
`.
`
`.
`
`
`
`tz .::
`
`and 22 in the heavy chain may correspond to acetylated Cys,
`since Cys is invariant at these positions.
`DISCUSSION
`Chromatography of the light chain fraction on hydroxylapa-
`tite revealed two poorly resolved major components (labeled B
`Several observations indicated that the VIP binding activity
`and C in Fig. 3) and at least one well-resolved minor protein
`to contamination with
`of the light chain fraction is not due
`heavy chains or complexes of heavy and light chains. including
`component (labeled A). Identical amounts (3 pg of protein) of
`the three peaks were subjected to N-terminal sequencing (11
`demonstration of the electrophoretic purity of the light chains,
`detection of a single protein sequence by N-terminal sequenc-
`cycles). The deduced sequence was identical in all three peaks
`(Asp-Val-Val-Met-Thr-Gln-Thr-Pro-Leu-Thr-Leu),
`ing, a constant specific activity across the width of the light
`correspond-
`
`ing to the N-terminal sequence of the light chain loaded on the chain peak recovered by gel filtration, precipitation of the VIP
`column. The major light chain peaks ( R and C ) were analyzed
`binding activity by antibody to light chains but not antibody
`to
`further. As expected, a single 25-kDa light band was observed
`heavy chains, and retention of the VIP binding activity in a
`light chain fraction purified by hydroxylapatite chromatogra-
`in both peaks by nonreducing SDS-electrophoresis and silver
`phy. We concluded that the light chain is capable of independ-
`staining. Determination of VIP binding by pooled fractions cor-
`[','I-
`responding to the two peaks showed that the specific
`e n t recognition of VIP.
`Tyr'"1VIP binding activity of the light chain in peak C (26,451
`Excess albumin as well as several short and mid-sized pep-
`tides unrelated to VIP did not inhibit the binding of radioiw
`c p d p g protein) was approximately 12-fold greater than that of
`peak B (2,231 c p d y g protein). Since peaks B and C were not
`dinated VIP by the light chains. A nonimmune light chain did
`not bind VIP. The VIP-binding affinity of the light chain was
`base-line resolved, contamination with high activity C could be
`responsible for the low-level activity of B.
`nearly equivalent to that of the high affinity component in the
`heavy chain preparation and only 5-fold lower than that of the
`Afflnity, Specificity, and Binding Capacity-The
`binding af-
`parent antibody. These observations indicate that the binding
`finities of light and heavy chains purified by gel filtration (see
`
`above) were estimated from Scatchard plokq of binding of [""I-
`' l j ~ ' ~ ] v I P mixed with increasing concentrations of unlabeled
`VIP (Fig. 4). Linear plots were evident for the light chain ( r =
`0.98) and the parent intact antibody ( r = 0.97). with apparent
`& values 10.1 and 1.9 nM, respectively. The heavy chain data
`suggested two binding components with Kd 6.8 and 58.3 nv ( r
`> 0.9 for each component). VIP binding capacities
`(per pg of
`protein) deduced from the x-interceph of Scatchard plokq were,
`light chain, 0.8 pmol; heavy chain, 0.05 and 0.13 pmol; and
`intact antibody, 6.4 pmol. In each case, the slope of a Hill plot
`of the data was close to unity, suggesting an absence of coop-
`erativity. These data show that the light and heavy chains can
`bind VIP independently with affinities only 3-5-fold lower than
`that of intact antibody.
`The VIP binding was observed in the presence of excess al-
`Ro. 3. H y h ~ y l . p t i b c h r o m m p h y
`of light chrirar: ~ p -
`bumin (0.5%, wh). Peptides unrelated in sequence to VIP (a-
`ration of a high activity rubpopulation. Light chains (325 pg) pu-
`human atrial natriuretic peptide, neurotensin. bombesin, and
`rified by gel filtration (see FIR. 1 1 were chromatographed on a high
`eledoisin, 1 p ~ ) were without detectable effect on the VIP bind-
`performance hydroxylapatite column using a gradient of sodium phos-
`phate, pH 6.8 ( 10-300 mu. 45 min; dotted line). Aliquob of pooled
`ing activity of the light or heavy chain. In comparison, >90'7 of
`fractions corresponding to the two major A*,,,, peaks (solid line) were
`[1'L51-Tyr'olVIP bound by these proteins was displaced competi-
`assayed for I""I-Tyr"'lVIP binding (hatched bar). Data for the binding
`of
`tively by 1 p~ unlabeled VIP. Since unfolding and folding
`are normalized for protein content to permit direct comparison of bind-
`light chains could expose nonspecific binding
`sites. the light
`ing activity. 1met shows a silver-stained SDS-polyacrylamide gel ( %
`25%) of the pooled fractions corresponding to peaks R and C (lanes 2 and
`I&,,,
`chain of a control nonimmune antibody (myeloma
`K ) was
`1 , respectively).
`purified and assayed for VIP binding under conditions identical
`to those used for the anti-VIP light chain. The control light
`chain did not display detectable VIP binding.
`
`j
`
`6 0
`
`Lassen - Exhibit 1051, p. 3
`
`

`

`
`
`
`
`Antigen Binding
`
`w
`cr
`W
`L L
`\
`0 z
`3
`0 m
`
`O i l
`0 0'
`0 0 0 2
`
`\*
`
`0 4 0 6
`
`BOUND, nM
`
`\
`I
`0 8 1.0
`
`
`
`0.0 0
`
`6
`2
`4
`BOUND. nM
`h ~ . 4. VIP-binding affinity of intact c23.6 antibody (A) and ita
`purified heavy (0) and light chain subunits (0) ( E ) . Binding of
`[lasI-Tyrlo]VIP
`(0.16 m) by antibody (1.7 m), light chain (221 m), or
`heavy chain (417 m) in the presence of increasing concentrations of
`unlabeled VIP (0.2 m to 1 px) was determined. Antibody subunits were
`purified as described in the legend to Fig 2. Data (means of duplicates)
`were analyzed using LIGAND (Elsevier BiosoR).
`
`by Light Chains
`
`
`737
`chains derived from polyclonal antisera, formation of heterolo-
`gous dimers is likely to be the predominant reaction, which
`could account for low antigen binding activity. In the present
`study, the VIP binding activity was associated with the light
`chain monomer peak (25 kDa) identified by gel filtration (Su-
`perose-12 column) in 6 M guanidinium chloride (Fig. 2) and in
`nondenaturing solvent (50 l l l ~ "is, pH 7.7, 0.15 M NaC1,
`0.025% Tween-20; not shown). These data do not permit evalu-
`ation of the relative levels of activity in the monomer and dimer
`forms of the light chain, since noncovalent dimerization could
`occur following column separation and exposure to VIP in the
`binding assay could also promote dimerization (23). In the case
`of the heavy chain of the anti-VIP antibody, overt aggregation
`was observed in the nondenaturing solvent; approximately 90%
`of the protein was recovered at the void volume of the gel
`filtration column (exclusion limit 2 x lo6 kDa) and only 10% as
`the 50-60 kDa monomer (not shown). This type of aggregation
`may explain the apparent heterogeneity of VIP binding by the
`heavy chain preparation (Fig. 4).
`Antigen binding entails contact with amino acid residues of
`both antibody subunits. The lengths of heavy and light chain
`CDRs in different antibodies are variable (22, 24), as are the
`relative extents of antigen contacts with heavy and light chains
`in different antigen-antibody complexes (13). Although the
`number and aggregate surface area of heavy chain contacts are
`generally somewhat greater, contacts at the light chain can
`approach 50% of the total interactions. Heavy chain prepara-
`tions consistently show lower binding activity than the parent
`antibodies (1-6) and interactions at heavy chain residues alone
`are unlikely to account fully for high affinity antigen-antibody
`binding. Light chain contributions in antigen binding may be
`particularly important when favorable contacts with the heavy
`chain CDRs are limited, for example, in antibodies expressing
`short heavy chains CDRs.' In view of these considerations, we
`consider it unlikely that the anti-VIP light chain is unique in its
`high affinity antigen binding activity.
`B-lymphocytes synthesize light chains in excess over heavy
`chains and secretion of free light chains by these cells has been
`demonstrated (1). Large amounts of light chains accumulate in
`the extracellular fluids and tissues of patients with light chain
`secreting tumors (2). There is compelling evidence that light
`chains can mediate peptide-bond cleavage
`(16), activate
`complement components (251, and suppress antibody synthesis
`(26). Observation of high affinity antigen binding by a light
`chain warrants study of the hypothesis that free light chains
`can simulate the antigen-specific functions of antibodies.
`AckmwZedgments-We are grateful to Professor T. T. Wu for helpful
`suggestions. Amino acid sequencing was done at the University of Ne-
`braska Protein Structure Core Facility (B. Bagenstoss).
`
`is sequence-specific and can be attributed to the variable re-
`gions of the light chain.
`The results of hydroxylapatite chromatography of light
`chains suggested that proper refolding of the protein was a
`mqjor factor governing antigen binding. This column resolved
`the light chains into two major species (labeled B and C in Fig.
`3), one with a high specific binding activity and another with
`little or no activity. N-terminal sequencing confirmed that both
`peaks were composed of the same light chain. The two peaks
`must represent,
`therefore, differently refolded light chains.
`Possible explanations for this observation are: (i) the two light
`chain states may be chemically different because of incomplete
`reduction of disulfide bonds, or (ii) chemically homogenous
`light chains may adopt alternative conformations. The possi-
`bility that B and C represent different aggregation states of the
`protein is unlikely, since establishment of equilibrium between
`the two states following chromatographic separation can be
`predicted to produce similar levels of binding activity. In view of
`the presence of two forms of the light chain, the binding capac-
`ity value deduced from Fig. 4 must underestimate the stoichi-
`ometry of VIP binding by the high activity species. On the other
`hand, computation of the binding afEnity of the high activity
`species is not compromised, since VIP binding by the second
`species is apparently too low to produce a deviation of the
`Scatchard plot h m linearity.
`Improper refolding could help explain reports describing
`minimal antigen binding by light chains purified from poly-
`clonal antibodies (4-8). For example, recovery of antigen bind-
`ing activity in light chains denatured with SDS (5) or propionic
`acid (4,6,8) may be difficult compared to guanidinium chloride
`used in the present study. A second factor is the possibility of
`inactive aggregate-formation in concentrated solutions of anti-
`body subunits (21). In the case of concentrated solutions of light
`
`REFERENCES
`1. Hopper, J. E., and Papagiannes, E. (1986) Cell. Immuml. 101,122-131
`2. Stevens, F. J., Solomon, A,, and Schiffer, M. (1991) Biochemistry 30,6803-6805
`3. Gherardi, E., and Milstein, C. (1992) Nature 367,201-202
`4. Fleischman, J. B., Porter, R. R., and Press, E. M. (1963) Biochem. J. 88,
`22&228
`5. Utaumi, S., and Karush, F. (1964) Biochemistry 3,1329-1342
`6. Edelman, G . M., Olins, D. E., Cally, J. A,, and Zinder, N. D. (1963) Proe. Natl.
`had. Sci. U. S. A. 50, 75S761
`
`cDNA for the light and heavy chain variable regions of antibody
`c23.5 were obtained by the reverse transcriptase-polymerase chain re-
`action method, cloned into a phagemid vedor, and sequenced by the
`dideoxynucleotide chain termination method (27). The deduced amino
`acid sequences suggest an aggregate of 26 residues in heavy chain
`CDRa and 32 residues in light chain CDRs (Q. S. Gao and S. Paul,
`unpublished). CDR3 in this heavy chain is short (4 amino acids) com-
`pared to its length in most mouse antibodies (mean, 8.7 residues; >95%
`of known heavy C D B s are composed of 5 or more residues, Ref. 24).
`
`Lassen - Exhibit 1051, p. 4
`
`

`

`Antigen Binding by Light Chains
`738
`18. Paul, S., Sun, M., Mody, R., Eklund, S. H., Beach, C. M., Massey, R. J., and
`7. fianek, F., and Nezlin, R. S. (1963) Biokhimiya 28,193
`8. Roholt, 0.. Onoue, K, and Pressman, D. (1963) P m . Natl. Acad. Sci. U. S. A.
`Hamel, F. (1991) J. Biol. Chem. 266, 16128-16134
`51, 173-178
`19. Paul, S., Volle, D. J., Beach, C. M., Johnson, D. R., Powell, M. J., and Massey,
`9. Ward, E. S., Gussow D., Griffiths, A. D., Jones, P. T., and Winter, G. (1989)
`R. J. (1989) Science 2-44, 115g1162
`20. Paul, S., Said, S. I., Thompson, A, Volle, D. J., Agrawal, D. K, Foda, H., and
`Nature 941,544"
`10. Schiffer, M., Girling, R. L., Ely, K R., and Edmundson, A. B. (1973) Biochem-
`De la Rocha, S. (1989) J. Neuroimmunol. 23, 133-142
`istry 12,4620-4631
`21. Bjork, I., and Tanford, C. (1971) Biochemistry 10, 1289-1295
`22. Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. (1991)
`11. Edmundson A. B., Ely, K R., Girling, R. L., Abola, E. E., Schiffer, M., West-
`Sequences of Proteins of Immunological Interest, Fifth Ed., U. S. Depart-
`holm, F. A,, Fausch, M. D., and 'Deutsch, H. A. (1974) Biochemistry 1%
`3816-3827
`ment of Health and Human Services. NIH Publication 91-3242
`12. Stevens, F., Westholm, F.,
`23. King, D. J., Byron, 0. D., Mountain, A., Weir, N., Harvey, A, Lawson, A. D. J.,
`R., and
`Panagiotopoulos, N., SchifFer, M., Popp,
`Proudfoot, K. A., Baldock, D. E., Harding, S. E., Yarranton, G. T., Owens, R.
`Solomon, A. (1981) J. Mol. Bid. 147, 185-193
`J. (1993) Biochem. J. 2SO.723-729
`13. Davies, D. R., Padlan, E. A, and Sheriff, S. (1990) Annu. Reu. Biochem. 59,
`24. Wu, T. T., Johnson, G., and Kabat, E. A. (1993) Proteins Struct. Funct. Genet.
`43-74. and references therein
`16, 1-7
`14. Novotny, J., Bruccoleri, R. E., and Saul, F. A (1989) Biochemistry 28,4735-
`25. Men, S., Koistinen, V., Miettinen, A, lbrnroth T., and Seppala, I. J. T. (1992)
`4749
`15. Erhan, S., and Greller, L. D. (1974) Nature 251,353-355
`J. Exp. Med. 175,939-950
`26. Ioannidis, R. A., Joshua, D. E., Warburton, P. T., Francis, S. E., Brown, R. D.,
`16. Mei, S., Mody, B., Eklund, S. H., and Paul, S. (1991) J. Biol. Chem. 286,
`15571-15574
`Gibson, J., and Kronenberg, H. (1989) Hematol. Pathol. 3, 169-175
`17. Paul, S., Sun, M., Mody, R., %wary, H. K, Mehmtra, S., Gianferrara, T.,
`27. Hoogenboom, H. R., Griffths, A. D., Johnson, K S., Chiswell, D. J., Hudson, P.,
`and Winter, G. (1991) Nucleic Acids Res. IS, 4133-4137
`Meldal, M., and Tramontano, A. (1992) J. Biol. Chem. 267,13142-13145
`
`Lassen - Exhibit 1051, p. 5
`
`

`

`Antigen Binding by Light Chains
`738
`18. Paul, S., Sun, M., Mody, R., Eklund, S. H., Beach, C. M., Massey, R. J., and
`7. fianek, F., and Nezlin, R. S. (1963) Biokhimiya 28,193
`8. Roholt, 0.. Onoue, K, and Pressman, D. (1963) P m . Natl. Acad. Sci. U. S. A.
`Hamel, F. (1991) J. Biol. Chem. 266, 16128-16134
`51, 173-178
`19. Paul, S., Volle, D. J., Beach, C. M., Johnson, D. R., Powell, M. J., and Massey,
`9. Ward, E. S., Gussow D., Griffiths, A. D., Jones, P. T., and Winter, G. (1989)
`R. J. (1989) Science 2-44, 115g1162
`20. Paul, S., Said, S. I., Thompson, A, Volle, D. J., Agrawal, D. K, Foda, H., and
`Nature 941,544"
`10. Schiffer, M., Girling, R. L., Ely, K R., and Edmundson, A. B. (1973) Biochem-
`De la Rocha, S. (1989) J. Neuroimmunol. 23, 133-142
`istry 12,4620-4631
`21. Bjork, I., and Tanford, C. (1971) Biochemistry 10, 1289-1295
`22. Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. (1991)
`11. Edmundson A. B., Ely, K R., Girling, R. L., Abola, E. E., Schiffer, M., West-
`Sequences of Proteins of Immunological Interest, Fifth Ed., U. S. Depart-
`holm, F. A,, Fausch, M. D., and 'Deutsch, H. A. (1974) Biochemistry 1%
`3816-3827
`ment of Health and Human Services. NIH Publication 91-3242
`12. Stevens, F., Westholm, F.,
`23. King, D. J., Byron, 0. D., Mountain, A., Weir, N., Harvey, A, Lawson, A. D. J.,
`R., and
`Panagiotopoulos, N., SchifFer, M., Popp,
`Proudfoot, K. A., Baldock, D. E., Harding, S. E., Yarranton, G. T., Owens, R.
`Solomon, A. (1981) J. Mol. Bid. 147, 185-193
`J. (1993) Biochem. J. 2SO.723-729
`13. Davies, D. R., Padlan, E. A, and Sheriff, S. (1990) Annu. Reu. Biochem. 59,
`24. Wu, T. T., Johnson, G., and Kabat, E. A. (1993) Proteins Struct. Funct. Genet.
`43-74. and references therein
`16, 1-7
`14. Novotny, J., Bruccoleri, R. E., and Saul, F. A (1989) Biochemistry 28,4735-
`25. Men, S., Koistinen, V., Miettinen, A, lbrnroth T., and Seppala, I. J. T. (1992)
`4749
`15. Erhan, S., and Greller, L. D. (1974) Nature 251,353-355
`J. Exp. Med. 175,939-950
`26. Ioannidis, R. A., Joshua, D. E., Warburton, P. T., Francis, S. E., Brown, R. D.,
`16. Mei, S., Mody, B., Eklund, S. H., and Paul, S. (1991) J. Biol. Chem. 286,
`15571-15574
`Gibson, J., and Kronenberg, H. (1989) Hematol. Pathol. 3, 169-175
`17. Paul, S., Sun, M., Mody, R., %wary, H. K, Mehmtra, S., Gianferrara, T.,
`27. Hoogenboom, H. R., Griffths, A. D., Johnson, K S., Chiswell, D. J., Hudson, P.,
`and Winter, G. (1991) Nucleic Acids Res. IS, 4133-4137
`Meldal, M., and Tramontano, A. (1992) J. Biol. Chem. 267,13142-13145
`
`Lassen - Exhibit 1051, p. 6
`
`