Reproducible axis of symmetry between the constant domains $C_L$ and $C_{H1}$ of the immunoglobulin Fab, and robustness of their geometric invariant core

Stoyanov, Kister, Gelfand, Kulikowski and Chothia (2000) identified superposition-free geometric invariant cores for the constant immunoglobulin domains $C_L$ and $C_{H1}$ from 53 Fab structures, using thresholds chosen empirically from dispersion distributions, and the paper explicitly left open whether a reproducible axis of symmetry exists between $C_L$ and $C_{H1}$ (analogous to the affine/twofold symmetry computed for the $V_L$–$V_H$ pair), whether the empirical thresholds and resulting core compositions are robust to dataset composition, resolution and expansion, and whether residue conservation at the core positions holds when resolved by species or lineage. The problem remains open: construct (or refute) a reproducible, coordinate-free axis of symmetry for the constant module, and establish the robustness of the constant-domain geometric cores (threshold sensitivity, dataset expansion, lineage-resolved residue conservation) on which such a construction rests.

Background

Immunoglobulin (antibody) molecules consist of paired heavy and light polypeptide chains, each divided into domains of roughly 100–110 amino-acid residues. The antigen-binding fragment (Fab) comprises four such domains: the variable domains $V_L$ (light chain) and $V_H$ (heavy chain), which carry the antigen-binding loops, and the constant domains $C_L$ and $C_{H1}$, which pair with each other to form the constant module of the Fab. Every domain adopts the immunoglobulin fold, a beta-sandwich of two antiparallel beta-sheets; in the convention of this literature, residue positions are labeled by the strand or loop they belong to (strands A through G, with loop positions such as AB and BC), so a residue is identified by a secondary-structure position such as B5 rather than by a species-specific sequence number.

Classically, structural comparisons within a protein family are made by least-squares superposition — rotating and translating one structure onto another — but the resulting measures depend on which atoms are fitted and on the coordinate frame. The Gelfand–Kister program instead sought coordinate-free, superposition-free invariants of antibody structure. Gelfand, Kister and Leshchiner (1996) constructed an invariant system of coordinates for immunoglobulin molecules based on the geometrical twofold symmetry inherent to the $V_L$–$V_H$ complex, finding 76 positions whose C-alpha atoms (the backbone carbon of each residue) are symmetrical about that axis. Gelfand, Kister, Kulikowski and Stoyanov (1998) then formalized a distance-dispersion algorithm: for a family of structures, compute for every pair of secondary-structure positions the dispersion (root-mean-square deviation across structures) of the C-alpha–C-alpha distance between them; a small mutually consistent seed set of low-dispersion positions (termed FOLDREP) is expanded hierarchically under numeric thresholds ($T_1$ for seed self-consistency, $T_2$ for expansion, $T_3$ for core self-consistency) until a geometrical core — a set of positions whose mutual C-alpha–C-alpha distances are conserved across the family — stabilizes. Applied to 53 immunoglobulin structures, this produced cores for the $V_L$ and $V_H$ domains, and an affine symmetry (a correspondence implemented by an affine transformation — a linear map followed by a translation) between $V_L$ and $V_H$ was defined and computed for each structure.

Stoyanov, Kister, Gelfand, Kulikowski and Chothia (2000) extended this analysis to the constant domains. Using the same 53 high-resolution structures (all with kappa-type light chains; antibodies carry one of two light-chain types, kappa and lambda), the authors identified FOLDREP seeds of six positions per domain and, with thresholds chosen empirically by inspecting the dispersion distributions ($T_1 = 0.20$ Å, $T_2 \le 0.25$ Å, $T_3 = 0.35$ Å for $C_L$; $T_1 = 0.25$ Å, $T_2 \le 0.30$ Å, $T_3 = 0.40$ Å for $C_{H1}$), obtained the cores $CORE_{T_L}$ (85 positions) and $CORE_{T_H}$ (59 positions). A residue-conservation analysis at these core positions, based on roughly 130 heavy-chain and 85 kappa-chain constant-domain sequences from the Kabat database (the era's canonical curated antibody sequence collection), found weaker conservation than in the variable domains, where Chothia, Gelfand and Kister (1998) had described a 76-residue common main-chain core with eight nearly invariant sites.

The paper explicitly left several issues unresolved; these constitute the problem. Its context is that the geometry relating the two domains of each module matters functionally: the elbow angle — the angle between the pseudo-two-fold axes of the variable and constant modules, each axis being the approximate twofold rotation relating that module's two domains — varies across Fabs and is coupled to antibody activity. Later structural immunology has characterized the variable–constant mutual orientation extensively by superposition-based means — Stanfield, Zemla, Wilson and Rupp (2006) computed elbow angles for 365 Fab fragments and found that lambda light-chain Fabs span a wider range than kappa ones; Landolfi et al. (2001) identified the ball-and-socket joint (residues 11, 110 and 112 of $V_H$ with residues 148–149 of $C_{H1}$) that couples the V–C orientation to activity; Sela-Culang, Alon and Ofran (2012) documented binding-related elbow-angle changes; Fernández-Quintero et al. (2020, 2022) analyzed $V_H$–$V_L$ and $C_{H1}$–$C_L$ interdomain-orientation dynamics by simulation. What none of this provides is an answer, in the coordinate-free terms of the geometric-core program, to whether a reproducible axis of symmetry exists for the constant module — and whether the cores and empirically chosen thresholds on which such a construction would rest are themselves stable. In the source's broader phrasing, that open question also concerns the mutual orientation between the variable and constant domains of the Fab.

Problem Statement

The 2000 study by Stoyanov, Kister, Gelfand, Kulikowski and Chothia identified the geometric invariant cores of the immunoglobulin constant domains $C_L$ and $C_{H1}$ from 53 high-resolution Fab structures but explicitly left the following interlocking questions unresolved; the problem is to resolve them.

1. Axis of symmetry. For the variable domains, the same program defined and computed a geometric (affine, approximately twofold) symmetry between $V_L$ and $V_H$, and the 1996 invariant coordinate system rested on the twofold axis of the $V_L$–$V_H$ complex. Determine whether an analogous, reproducible axis of symmetry exists for the constant module: define a symmetry element relating $C_L$ to $C_{H1}$ from coordinate-free data, compute it independently in each structure of a nonredundant, quality-filtered set of Fab structures, and establish whether it reproduces — within explicitly stated and justified tolerances — the same geometric object across the family, or whether the constant module's geometry varies too much for any reproducible axis.

2. Robustness of the cores and thresholds. The thresholds $T_1$, $T_2$, $T_3$ were chosen empirically by inspecting dispersion distributions of the 53-structure sample. Determine how sensitive the composition and self-consistency of the reported cores $CORE_{T_L}$ (85 positions) and $CORE_{T_H}$ (59 positions) are to these threshold values, to the composition of the structure set (nonredundancy criteria, experimental resolution, species and chain-type composition), and to expansion of the sample to the much larger set of Fab structures available today; establish whether the reported cores are stable, and how they change if they are not.

3. Lineage-resolved conservation. The original residue-conservation analysis pooled Kabat database sequences across species and chain types. Perform a systematic, species- or lineage-resolved assessment of residue conservation at the geometrical core positions of $C_L$ and $C_{H1}$, separating structural from functional conservation where the two can be distinguished.

These components serve one objective: to establish (or refute) a validated, coordinate-free geometric characterization of the constant-domain module of immunoglobulins. The reproducibility of the symmetry element and of the cores is to be established under explicitly specified datasets, position-correspondence conventions, symmetry definition and tolerances, so that the claim is independently testable.

References

1. Ognyan Stoyanov, Alexander Kister, Israel Gelfand, Casimir Kulikowski, Cyrus Chothia (2000). Geometric Invariant Core for the C_L and C_H1 Domains of Immunoglobulin Molecules. Journal of Computational Biology 7(5):673-684. DOI: 10.1089/106652701446143. https://doi.org/10.1089/106652701446143 (primary source)
2. Israel M. Gelfand, Alexander E. Kister, Dmitry Leshchiner (1996). The invariant system of coordinates of antibody molecules: prediction of the "standard" C alpha framework of VL and VH domains. Proceedings of the National Academy of Sciences USA 93(8):3675-3678. DOI: 10.1073/pnas.93.8.3675. https://doi.org/10.1073/pnas.93.8.3675
3. Israel Gelfand, Alexander Kister, Casimir Kulikowski, Ognyan Stoyanov (1998). Algorithmic Determination of Core Positions in the VL and VH Domains of Immunoglobulin Molecules. Journal of Computational Biology 5(3):467-477. DOI: 10.1089/cmb.1998.5.467. https://doi.org/10.1089/cmb.1998.5.467
4. Israel Gelfand, Alexander Kister, Casimir Kulikowski, Ognyan Stoyanov (1998). Geometric invariant core for the V(L) and V(H) domains of immunoglobulin molecules. Protein Engineering 11(11):1015-1025. DOI: 10.1093/protein/11.11.1015. https://doi.org/10.1093/protein/11.11.1015
5. Cyrus Chothia, Israel Gelfand, Alexander Kister (1998). Structural determinants in the sequences of immunoglobulin variable domain. Journal of Molecular Biology 278(2):457-479. DOI: 10.1006/jmbi.1998.1653. https://doi.org/10.1006/jmbi.1998.1653
6. Robyn L. Stanfield, Adam Zemla, Ian A. Wilson, Bernhard Rupp (2006). Antibody elbow angles are influenced by their light chain class. Journal of Molecular Biology 357(5):1566-1574. DOI: 10.1016/j.jmb.2006.01.023. https://doi.org/10.1016/j.jmb.2006.01.023
7. Nicholas F. Landolfi, Archana B. Thakur, Helen Fu, Max Vásquez, Cary Queen, Naoya Tsurushita (2001). The integrity of the ball-and-socket joint between V and C domains is essential for complete activity of a humanized antibody. Journal of Immunology 166(3):1748-1754. DOI: 10.4049/jimmunol.166.3.1748. https://doi.org/10.4049/jimmunol.166.3.1748
8. Vladimir Potapov, Vladimir Sobolev, Marvin Edelman, Alexander Kister, Israel Gelfand (2004). Protein-protein recognition: juxtaposition of domain and interface cores in immunoglobulins and other sandwich-like proteins. Journal of Molecular Biology 342(2):665-679. DOI: 10.1016/j.jmb.2004.06.072. https://doi.org/10.1016/j.jmb.2004.06.072
9. Jessica Shapiro, Douglas Brutlag (2004). FoldMiner: structural motif discovery using an improved superposition algorithm. Protein Science 13(1):278-294. DOI: 10.1110/ps.03239404. https://doi.org/10.1110/ps.03239404
10. Lawrence Shapiro, James Love, David R. Colman (2007). Adhesion molecules in the nervous system: structural insights into function and diversity. Annual Review of Neuroscience 30:451-474. DOI: 10.1146/annurev.neuro.29.051605.113034. https://doi.org/10.1146/annurev.neuro.29.051605.113034
11. Pierre P. Eleniste, Heike Hofstetter, Oliver Hofstetter (2013). Expression and characterization of an enantioselective antigen-binding fragment directed against alpha-amino acids. Protein Expression and Purification 91(1):20-29. DOI: 10.1016/j.pep.2013.06.010. https://doi.org/10.1016/j.pep.2013.06.010
12. Anirban Banerji (2014). Attractors in residual interactions explain the differentially-conserved stability of Immunoglobulins. arXiv:1404.0449. https://arxiv.org/abs/1404.0449
13. Inbal Sela-Culang, Shahar Alon, Yanay Ofran (2012). A systematic comparison of free and bound antibodies reveals binding-related conformational changes. Journal of Immunology 189(10):4890-4899. DOI: 10.4049/jimmunol.1201493. https://doi.org/10.4049/jimmunol.1201493
14. Raheleh Toughiri, Xiufeng Wu, Diana Ruiz, Flora Huang, John W. Crissman, Mark Dickey, Karen Froning, Elaine M. Conner, Thomas P. Cujec, Stephen J. Demarest (2016). Comparing domain interactions within antibody Fabs with kappa and lambda light chains. mAbs 8(7):1276-1285. DOI: 10.1080/19420862.2016.1214785. https://doi.org/10.1080/19420862.2016.1214785
15. Maximilian Bönisch, Carolin Sellmann, Daniel Maresch, Claudia Halbig, Stefan Becker, Lars Toleikis, Björn Hock, Florian Rüker (2017). Novel CH1:CL interfaces that enhance correct light chain pairing in heterodimeric bispecific antibodies. Protein Engineering Design and Selection 30(9):685-696. DOI: 10.1093/protein/gzx044. https://doi.org/10.1093/protein/gzx044
16. Lucas J. Bailey, Kimberly M. Sheehy, Pawel K. Dominik, Wenguang G. Liang, Huan Rui, Michael Clark, Mateusz Jaskolowski, Yejoon Kim, Dawid Deneka, Wei-Jen Tang, Anthony A. Kossiakoff (2018). Locking the Elbow: Improved Antibody Fab Fragments as Chaperones for Structure Determination. Journal of Molecular Biology 430(3):337-347. DOI: 10.1016/j.jmb.2017.12.012. https://doi.org/10.1016/j.jmb.2017.12.012
17. Mathieu Dondelinger, Patrice Filée, Eric Sauvage, Birgit Quinting, Serge Muyldermans, Moreno Galleni, Marylène S. Vandevenne (2018). Understanding the Significance and Implications of Antibody Numbering and Antigen-Binding Surface/Residue Definition. Frontiers in Immunology 9:2278. DOI: 10.3389/fimmu.2018.02278. https://doi.org/10.3389/fimmu.2018.02278
18. Monica L. Fernández-Quintero, Katharina B. Kroell, Martin C. Heiss, Johannes R. Loeffler, Patrick K. Quoika, Franz Waibl, Alexander Bujotzek, Ekkehard Moessner, Guy Georges, Klaus R. Liedl (2020). Surprisingly Fast Interface and Elbow Angle Dynamics of Antigen-Binding Fragments. Frontiers in Molecular Biosciences 7:609088. DOI: 10.3389/fmolb.2020.609088. https://doi.org/10.3389/fmolb.2020.609088
19. Monica L. Fernández-Quintero, Patrick K. Quoika, Florian S. Wedl, Clarissa A. Seidler, Katharina B. Kroell, Johannes R. Loeffler, Nancy D. Pomarici, Valentin J. Hoerschinger, Alexander Bujotzek, Guy Georges, Hubert Kettenberger, Klaus R. Liedl (2022). Comparing Antibody Interfaces to Inform Rational Design of New Antibody Formats. Frontiers in Molecular Biosciences 9:812750. DOI: 10.3389/fmolb.2022.812750. https://doi.org/10.3389/fmolb.2022.812750
20. Zirui Zhu, Hossein Ashrafian, Navid Mohammadian Tabrizi, Emily Matas, Louisa Girard, Haowei Ma, Edouard C. Nice (2025). Antibody numbering schemes: advances, comparisons and tools for antibody engineering. Protein Engineering Design and Selection 38:gzaf005. DOI: 10.1093/protein/gzaf005. https://doi.org/10.1093/protein/gzaf005
21. Boris Reva, Alexander Kister, Stephen Topiol, Israel Gelfand (2002). Determining the roles of different chain fragments in recognition of immunoglobulin fold. Protein Engineering 15(1):13-19. DOI: 10.1093/protein/15.1.13. https://doi.org/10.1093/protein/15.1.13