Recombinant Cys-Protein L

Researchers purifying antibody fragments (Fab, scFv) or non-IgG isotypes face failure with Protein A/G resins. Recombinant Cys-Protein L binds exclusively via kappa light chains, capturing fragments and IgM, IgA, IgE, IgD. The engineered N-terminal cysteine enables site-directed thiol conjugation for oriented immobilization, improving dynamic binding capacity 1.6-fold vs. random amine coupling. • Captures Fab, scFv, Dab fragments - structurally impossible with Protein A/G • Binds IgM, IgA, IgE, IgD with strong affinity • N-terminal cysteine for site-directed conjugation & oriented immobilization • ≥95% purity by SDS-PAGE; lyophilized; stored at -20°C

Recombinant Cys-Protein L: Kappa Chain Affinity Ligand

Recombinant Cys-Protein L is a genetically engineered immunoglobulin-binding protein derived from Peptostreptococcus magnus, expressed in Escherichia coli as a single non-glycosylated polypeptide chain fused with an N-terminal cysteine residue . The protein comprises five IgG-binding domains (B1–B5), totaling 366 amino acids with a molecular mass of 40.6 kDa (apparent ~46 kDa on SDS-PAGE), from which cell wall binding, cell membrane binding, and albumin binding regions have been removed to maximize specific IgG binding [1]. Distinct from Protein A and Protein G, which target the Fc region of heavy chains, Cys-Protein L binds exclusively through kappa (κ) light chain variable domain interactions without interfering with the antibody's antigen-binding site . The engineered N-terminal cysteine enables site-specific thiol-directed conjugation and oriented immobilization onto solid supports, a capability absent in native Protein L or standard recombinant Protein A/G products .

Binding Target Kappa light chain variable domain Distinct from Fc-binding Protein A/G

Immobilization Oriented via N-terminal cysteine Thiol-directed, site-specific coupling

Fragment Coverage Fab, scFv, Dab (Fc-independent) Enables purification workflows impossible with Protein A/G

• [1] ProSpec. Cys-Protein L Recombinant (PRO-1933). Produced in E. coli, single non-glycosylated polypeptide chain fused with Cys at N-terminus, 5 IgG-binding regions, 366 amino acids, 40.6 kDa. View Source

Irreplaceability of Recombinant Cys-Protein L in Kappa-Chain Workflows

Immunoglobulin-binding proteins A, G, and L exhibit fundamentally distinct binding targets, species cross-reactivity profiles, and antibody class coverage, making generic substitution scientifically untenable [1]. Protein A and Protein G bind primarily to the Fc region of IgG heavy chains and cannot capture antibody fragments lacking Fc domains (Fab, scFv, Dab), nor do they bind IgM, IgA, IgD, or IgE with useful affinity [1][2]. Standard recombinant Protein L lacks the engineered N-terminal cysteine residue, precluding site-directed thiol conjugation and oriented immobilization—capabilities that directly affect binding capacity, assay sensitivity, and conjugate homogeneity [1]. Furthermore, SPR-based kinetic analyses have demonstrated that Protein A and Protein A/G capture surfaces interfere with antibody Fab regions, causing up to 489-fold and 2,826-fold attenuation of measured antigen binding affinities, respectively, while Protein L surfaces preserve antigen binding to a significantly greater extent [3]. The quantitative evidence below substantiates why Recombinant Cys-Protein L occupies a non-substitutable niche in antibody fragment purification, immunoassay development, and site-specific bioconjugation workflows.

Binding target mismatch Protein A/G target Fc region, not kappa light chain; cannot capture Fab, scFv, IgM, IgA, IgE, IgD.

Oriented immobilization absent Standard Protein L lacks N-terminal cysteine; random amine coupling may reduce binding capacity and conjugate homogeneity.

Antigen-binding interference Protein A and A/G surfaces interfere with Fab region, causing up to 2,826-fold attenuation in SPR; Protein L preserves binding closer to solution reference.

• [1] GenScript. Protein A, G, and L Binding Characteristics by Species and Antibody Subclass. View Source
• [2] Sviatenko OV, Gorbatiuk OB, Vasylchenko OA. Application of Immunoglobulin-Binding Proteins A, G, L in the Affinity Chromatography. Biotechnologia Acta. 2014;7(2):34-45. doi:10.15407/biotech7.02.034. View Source
• [3] Kamat V, Rafique A, Huang T, Olsen O, Olson W. The impact of different human IgG capture molecules on the kinetics analysis of antibody-antigen interaction. Anal Biochem. 2020;593:113580. doi:10.1016/j.ab.2020.113580. PMID: 31926892. View Source

Recombinant Cys-Protein L: Quantified Differentiation Evidence

SPR Antigen-Binding Preservation: Protein L vs. Protein A

In a direct head-to-head SPR study across six unique human monoclonal antibody–antigen interactions and eight capture surfaces, antibodies captured on Protein L surfaces exhibited 219-fold weaker antigen binding affinities compared to the KinExA reference standard, whereas Protein A surfaces caused 489-fold weaker binding and Protein A/G surfaces caused 2,826-fold weaker binding [1]. Protein L thus preserved antigen binding 2.2-fold better than Protein A and 12.9-fold better than Protein A/G. The study authors attributed Fab-region interaction as the mechanism for interference by Protein A and Protein A/G, a structural liability absent in Protein L which binds κ-light chains remote from the paratope [1].

SPR binding preservation

Head-to-head

Protein L surface: 219-fold weaker vs KinExA
Protein A: 489-fold; Protein A/G: 2,826-fold

Reported 2.2× better binding preservation than Protein A; 12.9× better than Protein A/G

6 hmAbs, picomolar affinity; SPR comparison to KinExA

Evidence Dimension	Antigen binding affinity attenuation vs. KinExA reference (fold-weaker KD)
Target Compound Data	Protein L surface: 219-fold weaker antigen binding
Comparator Or Baseline	Protein A surface: 489-fold weaker; Protein A/G surface: 2,826-fold weaker; Protein G surface: 32-fold weaker; Goat AHC pAb: 3–7-fold weaker (reference standard)
Quantified Difference	Protein L preserves antigen binding 2.2× better than Protein A; 12.9× better than Protein A/G; but 6.8× worse than Protein G
Conditions	SPR (Biacore); six hmAbs with picomolar affinity; eight anti-human IgG capture surfaces; comparison to kinetic exclusion assay (KinExA)

Why This Matters

For immunoassay and biosensor applications where capture surface integrity directly governs quantitative accuracy, Protein L offers measurably superior antigen-binding preservation compared to the industry-standard Protein A and Protein A/G surfaces [1].

• [1] Kamat V, Rafique A, Huang T, Olsen O, Olson W. The impact of different human IgG capture molecules on the kinetics analysis of antibody-antigen interaction. Anal Biochem. 2020;593:113580. doi:10.1016/j.ab.2020.113580. PMID: 31926892. View Source

Antibody Class and Fragment Binding Breadth

Cross-study comparison of binding specificities shows that Recombinant Cys-Protein L binds all five human immunoglobulin classes (IgG, IgM, IgA, IgE, IgD) as well as Fab, scFv, and Dab fragments containing kappa light chains, whereas Protein A shows no binding to human IgD, weak binding to IgM and IgA, and weak-to-no binding to scFv; Protein G shows no binding to human IgM, IgA, IgE, or IgD, and no binding to scFv [1][2]. Specifically, Protein L binds human IgM with strong affinity (+++ on semi-quantitative scale), while Protein A binds IgM weakly (+) and Protein G shows no binding (−) [1]. For scFv fragments, Protein L provides strong binding (+++), whereas Protein A binds only weakly (+) and Protein G shows no binding (−) [1].

Ig class & fragment binding

Cross-study

Protein L: IgG+++, IgM+++, IgA+++, IgE+++, IgD+++, Fab+++, scFv+++
Protein A: IgM+, IgD−, scFv+
Protein G: IgM−, IgA−, IgE−, IgD−, scFv−

Only Protein L covers all five Ig classes and kappa-containing fragments

Semi-quantitative binding data; vendor/literature compilation

Evidence Dimension	Binding to human immunoglobulin classes and antibody fragments (semi-quantitative: +++ strong, ++ moderate, + weak, − none)
Target Compound Data	Protein L: IgG +++, IgM +++, IgA +++, IgE +++ (strong), IgD +++, Fab +++, scFv +++
Comparator Or Baseline	Protein A: IgG +++, IgM +, IgA +, IgE medium, IgD −, Fab +, scFv +; Protein G: IgG +++, IgM −, IgA −, IgE −, IgD −, Fab +, scFv −
Quantified Difference	Protein L binds all five Ig classes and all kappa-containing fragments vs. Protein A (binds 3/5 Ig classes, weak scFv) and Protein G (binds 1/5 Ig classes, no scFv binding)
Conditions	Compiled vendor and literature binding specificity data for human antibodies; GenScript and Biosyn reference tables

Why This Matters

For laboratories purifying non-IgG antibody isotypes (IgM, IgA, IgE) or antibody fragments (Fab, scFv, Dab) lacking Fc regions, Protein L is the only viable single-ligand affinity solution among the three canonical antibody-binding proteins [1][2].

• [1] GenScript. Protein A, G, and L Binding Characteristics by Species and Antibody Subclass. View Source
• [2] Bio-Synthesis Inc. Binding Comparison between Protein A, Protein G, Protein A/G, and Protein L. View Source

Site-Directed Cysteine Conjugation Capacity Advantage

The N-terminal cysteine engineered into Recombinant Cys-Protein L enables thiol-directed, site-specific immobilization onto solid supports—a strategy whose quantitative benefit has been demonstrated in a controlled study of Fab fragment affinity resins. Among multiple C-terminal tags tested (PPKPPK, FLAG, Cys), only the C-terminal Cys tag achieved site-directed immobilization, yielding a 1.6-fold increase in dynamic binding capacity at 100% breakthrough (DBC100%) compared to Fab wild-type immobilized via random amine coupling on CNBr-activated Sepharose 4 FF and Actigel ALD Superflow at equivalent ligand density [1]. Furthermore, site-directed Cys immobilization on sulfhydryl-reactive resin achieved the highest absolute DBC100% of 11 mg/mL, representing a 13.8-fold total improvement over mAb-based affinity resin (0.8 mg/mL) when combined with pore/ligand size ratio optimization [1]. The Cys-Protein L product specifically provides this thiol-directed conjugation capability via its N-terminal cysteine residue, which is absent in standard recombinant Protein L [2].

Cys-directed DBC gain

Class-level

1.6-fold higher DBC_100% vs random amine coupling; up to 11 mg/mL absolute DBC

Site-directed thiol immobilization improves binding capacity per ligand

Fab fragment data; equal ligand density; sulfhydryl-reactive resin

Evidence Dimension	Dynamic binding capacity at 100% breakthrough (DBC100%)
Target Compound Data	Cys-directed immobilization: 1.6-fold higher DBC100% vs. random amine coupling; absolute DBC100% up to 11 mg/mL (Fab Cys on sulfhydryl-reactive resin)
Comparator Or Baseline	Fab wt via random amine coupling on CNBr-activated Sepharose 4 FF: baseline DBC100%; mAb on CNBr Sepharose: 0.8 mg/mL DBC100%
Quantified Difference	1.6-fold DBC100% increase attributable specifically to Cys-mediated site-direction; 13.8-fold total improvement vs. mAb resin (combined site-direction + fragment optimization)
Conditions	Fab (~50 kDa) and scFv fragments; CNBr-activated Sepharose 4 FF, Actigel ALD Superflow, Sulfhydryl-reactive resin; equal ligand density across conditions

Why This Matters

Site-directed immobilization via the unique N-terminal cysteine translates directly into higher functional binding capacity per unit of immobilized ligand, reducing resin volume requirements and cost per purification cycle [1][2].

• [1] Sakhnini LI, Pedersen AK, Azevedo AM, et al. Designing monoclonal antibody fragment-based affinity resins with high binding capacity by thiol-directed immobilisation and optimisation of pore/ligand size ratio. J Chromatogr A. 2016;1468:143-153. doi:10.1016/j.chroma.2016.09.035. PMID: 27663728. View Source
• [2] The Biotek. Cys-Protein L Product Page. N-terminal Cys enables site-specific labeling or immobilization on solid supports. Recombinant production in E. coli, 366 amino acids, 40.6 kDa. View Source

Kappa Light Chain Binding Affinity Characterization

The binding affinity of Protein L for the human kappa light chain variable domain has been quantitatively established since the foundational characterization by Nilson et al. (1992), with an affinity constant Ka = 1.5 × 10⁹ M⁻¹ for native kappa I light chain and Ka = 0.9 × 10⁹ M⁻¹ for an isolated VκI fragment [1]. This affinity is comparable in magnitude to reported values for Protein A (Ka = 1.4 × 10⁸ M⁻¹ toward IgG Fc [2]) and Protein G (Ka = 6.7 × 10⁹ M⁻¹ toward IgG [2]), though direct cross-comparison is cautioned due to differing measurement conditions [2]. Critically, Protein L binding is structurally dependent on the tertiary integrity of the kappa chain VL domain, with complete loss of binding upon extensive reduction and alkylation of the kappa chain, confirming conformational rather than linear epitope recognition [1].

κ chain affinity

Class-level

K_a = 1.5 × 10⁹ M⁻¹ (native κI); 0.9 × 10⁹ M⁻¹ (VκI fragment)

Conformation-dependent binding to kappa VL domain

Nilson et al. 1992; direct cross-comparison to Protein A/G not standardized

Evidence Dimension	Affinity constant (Ka) for target immunoglobulin domain
Target Compound Data	Protein L: Ka = 1.5 × 10⁹ M⁻¹ (native κI light chain); Ka = 0.9 × 10⁹ M⁻¹ (isolated VκI fragment)
Comparator Or Baseline	Protein A: Ka = 1.4 × 10⁸ M⁻¹ (IgG Fc); Protein G: Ka = 6.7 × 10⁹ M⁻¹ (IgG); values not directly comparable due to different measurement conditions
Quantified Difference	Protein L κ-chain affinity is ~10.7-fold higher than Protein A Fc affinity by nominal Ka comparison; Protein G Fc affinity is ~4.5-fold higher than Protein L κ-chain affinity by nominal Ka comparison
Conditions	Native and isolated VL domain preparations; equilibrium binding; Protein L from P. magnus (Nilson et al., J Biol Chem, 1992); Protein A/G Ka values from Choe et al., Materials 2016, 9(12):994

Why This Matters

The well-characterized, conformation-dependent binding mechanism provides predictable and reproducible purification performance for antibodies and fragments bearing kappa light chain variable domains, a biophysical property validated across three decades of literature [1].

• [1] Nilson BH, Solomon A, Björck L, Åkerström B. Protein L from Peptostreptococcus magnus binds to the kappa light chain variable domain. J Biol Chem. 1992;267(4):2234-2239. PMID: 1733926. View Source
• [2] Choe W, Durgannavar TA, Chung SJ. Fc-Binding Ligands of Immunoglobulin G: An Overview of High Affinity Proteins and Peptides. Materials. 2016;9(12):994. doi:10.3390/ma9120994. Table 1. Ka values: Protein A 1.4×10⁸ M⁻¹, Protein G 6.7×10⁹ M⁻¹, Protein L 5.69×10⁷ M⁻¹. View Source

Kappa Light Chain Subtype Specificity

Protein L binding is restricted to three of the four major human kappa light chain variable region subgroups: VκI, VκIII, and VκIV bind Protein L, whereas VκII and all lambda (λ) light chain subgroups show no detectable binding [1]. In mice, Protein L binds kappa 1 subtype [2]. Since approximately 60% of human serum IgG antibodies carry kappa light chains and the majority of these fall within the VκI, VκIII, and VκIV subgroups, Protein L can capture a substantial fraction of polyclonal IgG from human serum [3]. In contrast, Protein A and Protein G bind through heavy chain Fc regions and exhibit no kappa/lambda or Vκ subtype selectivity [1]. This subtype specificity is a double-edged differentiation: it enables selective enrichment of kappa-chain-containing antibodies and fragments but precludes capture of lambda-chain or VκII antibodies—a limitation absent in Protein A/G [1].

Vκ subtype specificity

Class-level

Protein L binds VκI, VκIII, VκIV; no binding to VκII or any lambda chains

Captures ~60% human IgG; requires target κ subtype verification

Nilson et al. 1992; CIM disk data; lambda antibodies need alternative

Evidence Dimension	Human kappa light chain variable region subgroup binding
Target Compound Data	Protein L: VκI (+), VκIII (+), VκIV (+); VκII (−); all lambda subgroups (−); estimated ~60% human IgG captured from serum
Comparator Or Baseline	Protein A: no Vκ subtype restriction (binds Fc of heavy chain); Protein G: no Vκ subtype restriction (binds Fc/Fab of heavy chain); no lambda/kappa discrimination
Quantified Difference	Protein L captures ~3 of 4 human Vκ subgroups; captures antibodies representing ~60% of human serum IgG; zero capture of lambda-chain antibodies vs. Protein A/G which capture both kappa and lambda IgG
Conditions	Purified human Vκ and Vλ subgroup proteins; binding assays (Nilson et al., J Biol Chem, 1992); CIM Protein L disk characterization data

Why This Matters

Users must verify that their target antibody or fragment contains a Protein L-binding kappa subtype (VκI, III, or IV); for lambda-chain antibodies or VκII kappa antibodies, Protein A or G remain necessary alternatives, making subtype awareness essential for procurement decisions [1][2].

• [1] Nilson BH, Solomon A, Björck L, Åkerström B. Protein L from Peptostreptococcus magnus binds to the kappa light chain variable domain. J Biol Chem. 1992;267(4):2234-2239. PMID: 1733926. View Source
• [2] Luna Nanotech Canada. Protein L, Biotinylated Product Page. Protein L binds kappa 1, 3, 4 subtypes. View Source
• [3] BIA Separations (now Sartorius). CIM Protein L Disk Characterization. Protein L binds up to 60% of IgG antibodies from human serum sample. View Source

Product Purity and Endotoxin Specifications

Recombinant Cys-Protein L is supplied at >95% purity as determined by SDS-PAGE analysis, with endotoxin levels controlled to less than 0.1 EU/μg as measured by the LAL method [1]. The endotoxin specification of <0.1 EU/μg positions this product at the threshold for in vivo applicability, where generally endotoxin levels below 0.1 EU/mL are considered acceptable [1]. The lyophilized formulation contains no additives, preservatives, or carrier proteins, providing maximum flexibility for downstream conjugation and immobilization workflows . For context, many E. coli-derived recombinant protein products report endotoxin levels in the range of 0.1–1.0 EU/μg; ultra-low endotoxin products (<0.01 EU/μg) are available at premium pricing for specialized in vivo applications .

Purity & endotoxin

Lot attribute

Purity >95% (SDS-PAGE); endotoxin <0.1 EU/μg (LAL)

Supports reproducible conjugation and cell-based assays

Lyophilized, additive-free; supplier specification

Evidence Dimension	Purity and endotoxin level of lyophilized recombinant product
Target Compound Data	Purity >95% (SDS-PAGE); Endotoxin <0.1 EU/μg (LAL method); lyophilized without additives
Comparator Or Baseline	Industry typical E. coli recombinant proteins: endotoxin 0.1–1.0 EU/μg; ultra-low endotoxin grade: <0.01 EU/μg; ProSpec Cys-Protein L: >95% purity
Quantified Difference	Endotoxin specification meets threshold for general research and potential in vivo use; ±additive-free formulation enables direct conjugation without interference
Conditions	SDS-PAGE under reducing conditions; Limulus Amebocyte Lysate (LAL) endotoxin assay; lyophilized from sterile aqueous solution

Why This Matters

The combination of high purity and controlled low endotoxin reduces batch-to-batch variability in downstream immunoassays, conjugation efficiency, and cell-based experiments, while the additive-free formulation avoids interference with thiol-specific conjugation chemistry .

• [1] ProSpec. Cys-Protein L Recombinant (PRO-1933). Purity >95.0% as determined by SDS-PAGE; lyophilized without any additives. View Source

Recombinant Cys-Protein L: Application Scenarios

Purification of Fc-Deficient Antibody Fragments

Recombinant Cys-Protein L is the ligand of choice for single-step affinity purification of antibody fragments that lack the Fc region and therefore cannot be captured by Protein A or Protein G [1]. As demonstrated by binding specificity data, Protein L binds Fab fragments (+++), scFv (+++), and Dab fragments containing kappa light chains with strong affinity, whereas Protein A and Protein G show only weak or no binding to these fragment classes [1][2]. The engineered N-terminal cysteine further enables oriented covalent immobilization onto chromatography resins, which literature evidence shows can improve dynamic binding capacity by 1.6-fold compared to random amine coupling at equivalent ligand density [3]. For laboratories producing recombinant Fab or scFv in E. coli or mammalian systems, Cys-Protein L enables a purification workflow that is structurally impossible with Protein A/G resins [1].

Oriented Capture Ligand Immobilization for Immunoassays

The N-terminal cysteine of Recombinant Cys-Protein L enables thiol-directed, site-specific conjugation to maleimide-activated surfaces, gold nanoparticles, or sulfhydryl-reactive biosensor chips, achieving uniform ligand orientation that random amine coupling cannot provide [1]. Evidence from SPR studies shows that capture surface chemistry profoundly impacts measured antibody-antigen binding kinetics: Protein L surfaces preserved antigen binding with 2.2-fold less interference than Protein A (219-fold vs. 489-fold weaker) and 12.9-fold less interference than Protein A/G (219-fold vs. 2,826-fold weaker) [2]. Site-directed immobilization via cysteine thiol chemistry produces more pronounced, sigmoid-shaped breakthrough curves with higher utilization yield (82–85% at optimal ligand density), translating to more efficient use of expensive capture ligand [3]. This scenario is particularly relevant for Biacore SPR, BLI, and ELISA platforms where oriented ligand presentation directly governs assay sensitivity and reproducibility [1][2].

Purification of Non-IgG Antibody Isotypes

For researchers isolating IgM, IgA, IgE, or IgD antibodies—whether polyclonal from serum or recombinant from cell culture—Recombinant Cys-Protein L offers a unique affinity capture solution. Binding specificity data show that Protein L binds human IgM (+++), IgA (+++), IgE (strong), and IgD (+++), whereas Protein A binds these isotypes weakly (+) or not at all (−), and Protein G shows no binding (−) to any of these four isotypes [1][2]. Protein L chromatography is therefore explicitly recommended by manufacturers as the affinity method for antibody-related molecules that do not bind to Protein A, including IgM and IgA [3]. The caveat is that only kappa-light-chain-containing antibodies of these isotypes will bind; lambda-chain isotypes require alternative strategies [4].

Site-Specific Bioconjugate Production

The single, defined N-terminal cysteine of Recombinant Cys-Protein L provides a unique chemical handle for stoichiometrically controlled bioconjugation via thiol-maleimide, thiol-iodoacetamide, or thiol-disulfide exchange chemistries [1]. This site-specific conjugation strategy yields homogeneous conjugates with defined drug/ligand-to-protein ratios, in contrast to random lysine-directed conjugation which produces heterogeneous product mixtures with variable activity [1]. Evidence from cysteine-tagged Fab fragment studies demonstrates that site-directed thiol immobilization achieves not only higher binding capacity (1.6-fold improvement) but also more efficient capacity utilization (sigmoid breakthrough curves) compared to random amine immobilization [2]. For ADC developers, immuno-PCR assay builders, or researchers requiring oriented Protein L-coated magnetic beads, the N-terminal Cys enables conjugation workflows that are incompatible with standard recombinant Protein L, Protein A, or Protein G products lacking engineered cysteine residues [1].

Application

Selection Property

Validation Focus

Fc-deficient fragment purification

Kappa light chain binding (Fab, scFv, Dab)

Binding capacity for target fragment vs. random immobilization

Oriented immunoassay surface

N-terminal cysteine for thiol-directed conjugation

Antigen-binding preservation in SPR/BLI vs. Protein A/G surfaces

Non-IgG isotype purification

Broad Ig class kappa binding (IgM, IgA, IgE, IgD)

Binding confirmation for target isotype; lambda exclusion check

Site-specific bioconjugate production

Single defined Cys residue for stoichiometric coupling

Conjugate homogeneity and retained binding activity

• [1] GenScript. Protein A, G, and L Binding Characteristics by Species and Antibody Subclass. View Source
• [2] Bio-Synthesis Inc. Binding Comparison between Protein A, Protein G, Protein A/G, and Protein L. View Source
• [3] Sakhnini LI, Pedersen AK, Azevedo AM, et al. Designing monoclonal antibody fragment-based affinity resins with high binding capacity by thiol-directed immobilisation and optimisation of pore/ligand size ratio. J Chromatogr A. 2016;1468:143-153. doi:10.1016/j.chroma.2016.09.035. PMID: 27663728. View Source
• [4] Nilson BH, Solomon A, Björck L, Åkerström B. Protein L from Peptostreptococcus magnus binds to the kappa light chain variable domain. J Biol Chem. 1992;267(4):2234-2239. PMID: 1733926. View Source