A residue-50 crystallographic B-factor scalar tracks catalytic-efficiency ordering across HG Kemp eliminase evolution
Year: 2026
Type: preprint
Abstract: Active-site pre-organisation is a well-characterised contributor to enzyme catalysis, and progressive rigidification of catalytic residues during HG-series Kemp eliminase evolution — including decreasing structure-normalised B-factor Z-scores at the position-50 residue — has already been reported directly from room-temperature crystallography (Broom et al., 2020). The contribution here is therefore not that observation, but a bounded quantitative test of how completely a single compact scalar recovers catalytic-efficiency ordering within an HG Kemp eliminase series, together with an explicit account of where that scalar fails to transfer. In the primary HG dataset — five room-temperature structures from the directed-evolution series plus the separately engineered variant HG4 — the residue-50 side-chain scalar correlated with log10(kcat/KM) at r = −0.9899 (p = 1.52 × 10-4) and reproduced the catalytic-efficiency ordering of the archived scalar values exactly (Spearman ρ = −1.000; exact two-sided permutation p = 0.0028). Restricting the analysis to the five evolutionary-series variants gave r = −0.9893. The association is robust to the choice of published kinetic dataset. In four independently designed HG variants the multi-residue active-site scalar was directionally negative (r = −0.7759) but weak on a rank basis (ρ = −0.400; exact two-sided p = 0.75) and is treated as neither replication nor support of comparable weight. Application of the same extraction logic to an independent KE07 scaffold failed a pre-committed robustness criterion and is reported as a negative transfer result. Two distinct sensitivities are separated there: a normalization correction, and an additional dependence on which deposited structure represents each variant across all 48 pre-declared combinations. We do not claim that the specificity of the primary single-residue scalar has been established: the archived negative controls were computed on a four-residue active-site average, not on the primary scalar, and their generating workflow is not fully archived. B-factors are treated throughout as operational crystallographic observables rather than as measurements of solution dynamics. The result is a narrowly scoped, reproducible descriptive reanalysis of public data with a clearly documented limit on transfer — not a causal mechanism, a specificity demonstration, or a generalisation to other reaction classes.
Source: ChemRxiv
Author: Mahesh babu Sodalagunta
Language: English
Subfield: Structural Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7213464345