The Escherichia coli lac repressor (LacI) transmits the signal of inducer binding at its core-domain pocket to the distant DNA-binding domain with almost no conformational change, making it a paradigm of allostery without structural change. Hydrogen-deuterium exchange mass spectrometry (HDX/MS) across six functional states, combined with computational prediction of structural water molecules, supported a dynamic model in which ligand-specific differential rigidification of core-domain secondary structures — partly via water-mediated hydrogen bonds in the ligand pocket — shifts the conformational ensemble between mutually exclusive functional states (Glasgow et al., 2023). Subsequent high-resolution NMR work argued instead that LacI allostery can be described by a two-state Monod–Wyman–Changeux conformational pre-equilibrium and explicitly disputed the dynamic-model interpretation (Romanuka et al., 2023). Later studies delivered higher-resolution ensemble energetics from HDX/MS, hydration-linked molecular-dynamics simulations, and validated mutation–landscape causality tests at a rheostat position (a residue where different amino-acid substitutions tune functional output over a graded range rather than simply abolishing it), but none adjudicated between the two mechanisms or tested the predicted water network experimentally. The open problem is to establish the causal mechanism of LacI allostery: whether the ligand-specific changes in dynamics and the predicted water-mediated interactions cause the allosteric switch; what the temporal order of molecular events is during inducer-triggered operator release; and how the mechanism depends on oligomeric state and DNA occupancy.

Background

The Escherichia coli lac repressor (LacI) is the paradigm allosteric transcription factor: it represses the lactose-utilization genes by binding its operator DNA sequence, and binding of a sugar inducer at a pocket in the core domain lowers its affinity for the operator several-hundred-fold, turning transcription on. In vivo LacI is a homotetramer in which each dimer binds one operator; each protomer consists of an N-terminal DNA-binding domain (the headpiece) connected through a hinge helix to the core domain, which is built from N- and C-terminal subdomains and harbors the inducer pocket. A long-standing puzzle is the mechanism of this long-range allostery: X-ray crystal structures of the free, inducer-bound, and operator-bound repressor are nearly superposable in the core domain, so the signal must be transmitted without a large conformational change. Hydrogen-deuterium exchange measured by mass spectrometry (HDX/MS) reports local flexibility: backbone amide hydrogens in each peptide exchange with solvent deuterium faster where the peptide is flexible and less protected by stable hydrogen bonds. For most LacI peptides the exchange follows the EX2 regime, in which repeated brief local unfolding events precede exchange, so the uptake curves report equilibrium populations within the conformational ensemble rather than slow cooperative unfolding. Glasgow et al. (2023) applied peptide-level HDX/MS to a dimeric LacI construct (truncated before the tetramerization helix) across six functional states — apo, bound to the inducers IPTG (isopropyl-β-D-thiogalactoside) or TMG (thiomethyl-β-D-galactoside), bound to the anti-inducer ONPF (a ligand that strengthens operator binding), bound to operator DNA, and the ONPF–DNA ternary complex — together with a computational solvation protocol in the Rosetta macromolecular-modeling suite that places 'structural waters', defined as water molecules making three or four hydrogen bonds to protein or ligand atoms. They found ligand-specific differential rigidification: regions uniquely rigidified in IPTG-bound LacI cluster near the pocket and the interface between core subdomains, where predicted water-mediated hydrogen-bond networks form upon inducer binding, whereas ONPF rigidifies distal N-terminal subdomain regions that interface with the DNA-binding domain. The rigidified regions correlate strongly with mutational phenotypes from classic saturation-mutagenesis screens of 4000 LacI point mutants (positions whose mutation abolishes induction versus positions whose mutation causes constitutive transcription). On this basis they proposed a dynamic model in which differential rigidification of secondary-structure elements — partly water-mediated — shifts the ensemble between mutually exclusive low-energy functional states. Shortly afterward, Romanuka et al. (2023) used high-resolution NMR spectroscopy on dimeric LacI constructs — a K84M mutant dimer lacking the tetramerization domain (residues 1-333) and, for the ligand and anti-inducer titrations, an isolated core dimer (residues 60-333) — rather than the wild-type dimeric LacI(1-331) measured by HDX/MS, and reached the opposite interpretation: the free repressor is a fast-exchange equilibrium between DNA-bound-like and inducer-bound-like conformations, and allostery follows the Monod–Wyman–Changeux (MWC) framework, in which a protein pre-exists in equilibrium between two conformations that bind the two effectors (DNA and inducer) with opposite preference, each effector shifting the equilibrium toward the state it favors through a concerted, symmetry-conserving transition. They explicitly described the dynamic model as contradicting their findings and attributed the discrepancy to crystallography missing solution-state conformations. Molecular-dynamics studies of hydration-linked substates and of substitutions that reshape the conformational landscape are compatible with population-shift pictures but have not discriminated the mechanisms. Thus the causal question — whether ligand-specific dynamics, including the predicted structural-water networks, cause the allosteric switch or merely accompany a two-state conformational pre-equilibrium — remains open, along with the temporal sequence of events during induction and the dependence of the ensemble shifts on oligomeric state and DNA occupancy.

Problem Statement

Establish the causal mechanism of long-range allostery in the E. coli lac repressor, adjudicating between the dynamics-mediated mechanism and the two-state conformational pre-equilibrium. The central objective is the causal adjudication in item 1, which can be resolved without also resolving the separate question of the predicted structural waters treated in item 2; items 2–4 are related open questions that delimit the scope in which the problem remains open:

  1. Causal test of the dynamics-mediated mechanism. Determine whether the ligand-specific differential rigidification of core-domain secondary structures observed by HDX/MS — attributed in part to water-mediated hydrogen-bond networks predicted to form in the inducer-bound pocket — causally mediates transmission of the allosteric signal from the ligand-binding pocket to the DNA-binding domain, or whether the same observations are fully accounted for by a pre-existing equilibrium between DNA-bound-like and inducer-bound-like conformations (Monod–Wyman–Changeux allostery).

  2. Role of the predicted structural waters. Whether specific predicted water-mediated hydrogen bonds in the ligand-binding pocket (for example those involving residues D149 and N125, which form water-mediated contacts to IPTG in crystal structures and whose mutation lowers IPTG affinity) contribute causally to the ligand-specific ensemble shifts and the allosteric response remains unresolved. The water network is a computational prediction of the Rosetta solvation protocol rather than a direct observation — only one high-resolution LacI structure (PDB 2P9H) includes electron density for water molecules in the ligand-binding site — and it has not been experimentally determined across ligand states or perturbation-tested for an allosteric role; the reported D149/N125 mutation effects concern IPTG affinity rather than a demonstrated role in allosteric transmission.

  3. Temporal order of events during induction. The sequence of molecular events during inducer-triggered release from the operator — changes at the monomer-dimer interface, ligand-pocket rearrangement, loss of core-headpiece interdomain contacts, hinge-helix disordering, and DNA release — is undetermined: existing accounts are equilibrium descriptions of the end states, and the source model 'does not include information on an order of events for release from the DNA operator upon IPTG binding'.

  4. Architectural and occupancy scope. The existing HDX/MS ensemble data were collected on dimeric LacI(1-331) 'without the tetramerization domain' at the DNA concentrations used, and the source notes that 'higher DNA concentrations in the experiment would impact these data', whereas in vivo LacI is a homotetramer; how the observed ensemble shifts and the allosteric response depend on oligomeric state (the dimeric construct versus the full-length tetramer) and on operator DNA concentration and occupancy therefore remains open.

The common objective is a causally tested mechanistic account of LacI allostery.

References

  1. Glasgow A., Hobbs H.T., Perry Z.R., Wells M.L., Marqusee S., Kortemme T. Ligand-specific changes in conformational flexibility mediate long-range allostery in the lac repressor. Nature Communications 14, 1179 (2023). DOI: 10.1038/s41467-023-36798-1. https://doi.org/10.1038/s41467-023-36798-1
  2. Glasgow A., Hobbs H.T., Perry Z.R., Wells M.L., Marqusee S., Kortemme T. Author Correction: Ligand-specific changes in conformational flexibility mediate long-range allostery in the lac repressor. Nature Communications 14, 4208 (2023). DOI: 10.1038/s41467-023-39918-z. https://doi.org/10.1038/s41467-023-39918-z
  3. Romanuka J., Folkers G.E., Gnida M., Kovačič L., Wienk H., Kaptein R., Boelens R. Genetic switching by the Lac repressor is based on two-state Monod–Wyman–Changeux allostery. Proceedings of the National Academy of Sciences USA 120(49), e2311240120 (2023). DOI: 10.1073/pnas.2311240120. https://doi.org/10.1073/pnas.2311240120
  4. Kariyawasam N.L., Ploetz E.A., Swint-Kruse L., Smith P.E. Simulated pressure changes in LacI suggest a link between hydration and functional conformational changes. Biophysical Chemistry 304, 107126 (2024). DOI: 10.1016/j.bpc.2023.107126. https://doi.org/10.1016/j.bpc.2023.107126
  5. Kariyawasam N.L., Sivchenko A., Swint-Kruse L., Smith P.E. Substitutions at rheostat position 52 of LacI have long-range effects on the LacI conformational landscape. Biophysical Chemistry 320-321, 107414 (2025). DOI: 10.1016/j.bpc.2025.107414. https://doi.org/10.1016/j.bpc.2025.107414
  6. Lu C., Wells M.L., Reckers A., McBride S.K., Glasgow A. Site-resolved energetic information from HX-MS experiments. Nature Chemical Biology 22(2), 307-317 (2026). DOI: 10.1038/s41589-025-02049-1. https://doi.org/10.1038/s41589-025-02049-1
  7. Yuan J., Lüking M., Zikrin S., Sen B.C., Marklund E., van der Spoel D., Fange D., Elf J. LacI strikes a balance between stability and inducibility. Nucleic Acids Research 54(6), gkag296 (2026). DOI: 10.1093/nar/gkag296. https://doi.org/10.1093/nar/gkag296
  8. Kandavalli V., Zikrin S., Elf J., Jones D. Anti-correlation of LacI association and dissociation rates observed in living cells. Nature Communications 16, 764 (2025). DOI: 10.1038/s41467-025-56053-z. https://doi.org/10.1038/s41467-025-56053-z
  9. Montserrat-Canals M., Cordara G., Krengel U. Allostery. Quarterly Reviews of Biophysics 58, e5 (2025). DOI: 10.1017/S0033583524000209. https://doi.org/10.1017/S0033583524000209