Alternating-access membrane transporters switch between metastable end-states through large, transient, nonlinear conformational transitions that are difficult to observe experimentally and expensive to compute. The cold-inbetweening method (Yorke and Ginn, 2025) generates cheap athermal (zero-temperature) minimum-energy trajectories in torsion-angle space (parametrizing the protein by rotations about single bonds) between experimental end-state structures, yielding specific mechanistic predictions for three transporter superfamilies: supporter-arm helix unwinding during elevator transport in the MalT sugar transporter, outward-gate closure before inward-gate opening in the DraNramp metal transporter, and obligatory N-terminal helix rewinding that transiently plugs the binding cavity in the MATE exporter. None of these predictions has been tested, and the method's central claim - that a zero-temperature minimum-energy path approximates the route that real finite-temperature trajectories follow - is asserted but not established. The open problem is to determine these transition pathways and establish their validity, and to characterize the conditions under which athermal pathway computation yields general, verifiable descriptions of nonlinear functional conformational transitions.

Background

Membrane transporters move ions and solutes across lipid bilayers by the alternate-access mechanism: the protein cycles through metastable states in which a substrate-binding site is exposed to only one side of the membrane at a time, so that a bound ligand can be translocated without free solvent flow. Structural biology has determined many end-state structures (outward-open, occluded, inward-open), but the transitions between them - large-scale, nonlinear conformational changes lasting microseconds to milliseconds - are fleeting and remain difficult to observe experimentally or to compute. The mechanistic repertoire is conventionally distilled into three idealized modes: the elevator mechanism, in which a transport domain slides past a fixed scaffold; and the rocking-bundle and rocker-switch mechanisms, in which one or both helix bundles rock to shift the barrier between the two sides of the membrane (Drew and Boudker, Annu. Rev. Biochem. 2016; Nature 2024).

Molecular dynamics (MD) simulations, which propagate atomic positions under heat-driven equations of motion, can in principle follow these transitions, but the relevant timescales are far beyond unbiased MD for membrane-embedded transporters. A dedicated review of the enhanced-sampling techniques used instead - biased sampling such as umbrella sampling (restraining the simulation to successive windows along a chosen coordinate) and metadynamics (adding a repulsive bias that discourages revisiting already-sampled states), Markov state models (discrete kinetic models of conformational populations and transition rates), transition path sampling (rare-event Monte Carlo over trajectories), and path-based methods - documents both their reach and their substantial cost and coverage limitations (Pant et al., J. Phys. Chem. B 2025). Path methods compute a minimum-energy or minimum-free-energy path connecting two known end-states: the nudged elastic band and string methods relax an initial interpolation toward the path, with variants such as the string method with swarms of trajectories (Pan, Sezer and Roux, J. Phys. Chem. B 2008) and finite-temperature string formulations that estimate thermal mean forces along the path (Kulshrestha et al., Soft Matter 2022). Landmark applications show what such campaigns achieve system by system: an atomic-scale characterization of the elevator transition of the aspartate transporter GltPh required more than two thousand coupled simulation replicas (Thangapandian et al., PNAS 2025).

Cold-inbetweening (Yorke and Ginn, Commun. Chem. 2025; bioRxiv 2024) enters from the opposite end of the cost spectrum. The name borrows 'inbetweening' from animation - filling in frames between keyframes - with 'cold' signalling that the model of heat is deliberately omitted. The method parametrizes the protein by its torsion angles (rotations about single bonds), anchors the experimentally determined start and end structures, and numerically minimizes the kinetic and potential energy needed to complete the transition, using a standard molecular-mechanics force field without solvent or membrane. It thereby generates a smooth, athermal minimum-energy trajectory at very low computational cost. Applied to three transporter superfamilies, it produced specific mechanistic hypotheses: in the MalT sugar transporter from Bacillus cereus (a member of the EIIC family, the membrane component of the bacterial sugar phosphotransferase system), elevator transport proceeds with unwinding of a supporter-arm helix that keeps the cavity wide enough for maltose; in the DraNramp manganese transporter from Deinococcus radiodurans (a member of the Nramp family of natural-resistance-associated macrophage proteins), the outward gate closes before the inward gate opens, with shearing of helix TM1a (a segment of transmembrane helix 1; TM denotes a transmembrane helix, numbered from the N-terminus, with letter suffixes naming the segments of a split helix) against TM6b and TM8 and a transient displacement of TM6b consistent with prior biochemical labeling data (Bozzi et al., J. Biol. Chem. 2020); and in the MATE (multidrug and toxic compound extrusion) exporter from Pyrococcus furiosus, the conformational switch requires obligatory rewinding of the N-terminal helix TM1, which is unwound in the inward-open structure and rewound in the outward-open structure (Tanaka et al., Nature 2013; Zakrzewska et al., PNAS 2019), transiently plugging the cavernous internal cavity mid-transition.

The open question is the epistemic status of such trajectories. The authors themselves state that cold-inbetweening produces hypotheses that could be evaluated experimentally, for example by mutants chosen to test them in transport or activity assays or by structural determination; that no timescale can be assigned; and that omitting thermal motion precludes direct comparison with MD. The general claim - that a zero-temperature minimum-energy path runs close to the route that real, finite-temperature trajectories follow, with heat adding only noise - is asserted but not established, and earlier interpolation and morphing tools in structural bioinformatics were criticized as aesthetic rather than analytical devices. Whether these computed pathways can be validated and generalized into trustworthy descriptions of nonlinear functional transitions is the unresolved target.

Problem Statement

Determine the transient conformational pathways by which alternating-access membrane transporters switch between their experimentally determined metastable end-states, and establish whether computed minimum-energy pathways are valid descriptions of these nonlinear functional transitions.

The problem has one overarching objective - turning computed pathway hypotheses into verified descriptions of the functional transition - with three jointly supporting components.

(1) Pathway determination and testing for the cold-inbetweening case studies (cold-inbetweening is the athermal, torsion-space minimum-energy-path computation of Yorke and Ginn). For the MalT elevator transition (MalT: the Bacillus cereus sugar transporter of the EIIC family, the membrane component of the bacterial phosphotransferase system; predicted supporter-arm/TM1 unwinding maintaining the substrate cavity, with the V20/V353 solvent gate), the DraNramp rocking-bundle transition (DraNramp: the Deinococcus radiodurans Nramp-family manganese transporter; predicted outward-gate closure preceding inward-gate opening, TM1a-b unwinding and shearing against TM6b and TM8, transient TM6b displacement, where TM denotes a transmembrane helix and TM1a-b the two segments of the split helix TM1), and the Pyrococcus furiosus MATE rocker-switch transition (MATE: multidrug and toxic compound extrusion; predicted obligatory rewinding of the N-terminal TM1 helix with transient plugging of the binding cavity), determine whether the predicted mechanistic features genuinely occur on the functional transition route.

(2) The general relationship between athermal and thermal pathways. Establish, by theory or systematic computation across systems, under what conditions a zero-temperature minimum-energy path in torsion space approximates the dominant finite-temperature transition-path ensemble of a large conformational change: which features of the cold path (event ordering, helix unwinding and rewinding, transient plugging intermediates) are robust to thermal fluctuations, and where entropic, solvent, and membrane effects reorder or replace them, thereby defining the domain of validity of cold-pathway predictions.

(3) Generality beyond the case studies. Determine whether the validated approach extends to other proteins with high-quality, fully modelled end-state structures - that is, whether cheap athermal pathway computation combined with a defined validation standard yields generally trustworthy mechanistic descriptions of nonlinear functional conformational transitions, rather than merely plausible morphs between end-states.

For each of the three case-study transporters (MalT, DraNramp, and MATE) and for components (2) and (3), the question is which mechanistic predictions are confirmed, refuted, or refined, and on what evidence.

References

  1. Yorke, Briony A.; Ginn, Helen M. Calculation of minimum energy pathways in transport proteins. Communications Chemistry 8, 377 (2025). DOI 10.1038/s42004-025-01754-1. https://doi.org/10.1038/s42004-025-01754-1
  2. Yorke, Briony A.; Ginn, Helen M. Calculation of minimum energy pathways in transport proteins. bioRxiv preprint (2024). DOI 10.1101/2024.08.07.607056. https://doi.org/10.1101/2024.08.07.607056
  3. Drew, David; Boudker, Olga. Shared molecular mechanisms of membrane transporters. Annual Review of Biochemistry 85, 543-572 (2016). DOI 10.1146/annurev-biochem-060815-014520. https://doi.org/10.1146/annurev-biochem-060815-014520
  4. Drew, David; Boudker, Olga. Ion and lipid orchestration of secondary active transport. Nature 626, 963-974 (2024). DOI 10.1038/s41586-024-07062-3. https://doi.org/10.1038/s41586-024-07062-3
  5. Pan, Albert C.; Sezer, Deniz; Roux, Benoit. Finding transition pathways using the string method with swarms of trajectories. The Journal of Physical Chemistry B 112, 3432-3440 (2008). DOI 10.1021/jp0777059. https://doi.org/10.1021/jp0777059
  6. Kulshrestha, Avijeet; Punnathanam, Sudeep N.; Ayappa, K. Ganapathy. Finite temperature string method with umbrella sampling using path collective variables: application to secondary structure change in a protein. Soft Matter 18, 7593-7603 (2022). DOI 10.1039/d2sm00888b. https://doi.org/10.1039/d2sm00888b
  7. Pant, Shashank; Dehghani-Ghahnaviyeh, Sepehr; Trebesch, Noah; Rasouli, Ali; Chen, Tianle; Kapoor, Karan; Wen, Po-Chao; Tajkhorshid, Emad. Dissecting large-scale structural transitions in membrane transporters using advanced simulation technologies. The Journal of Physical Chemistry B 129, 3703-3719 (2025). DOI 10.1021/acs.jpcb.5c00104. https://doi.org/10.1021/acs.jpcb.5c00104
  8. Thangapandian, Sundar; Fakharzadeh, Ashkan; Moradi, Mahmoud; Tajkhorshid, Emad. Conformational free energy landscape of a glutamate transporter and microscopic details of its transport mechanism. Proceedings of the National Academy of Sciences 122, e2416381122 (2025). DOI 10.1073/pnas.2416381122. https://doi.org/10.1073/pnas.2416381122
  9. Zhao, Xiaoyu; Gao, Feng. Rocker-switch mechanism of choline transmembrane transport in FLVCR2: dynamic latch residues and cooperative regulation of transmembrane helices. JACS Au 5, 4936-4955 (2025). DOI 10.1021/jacsau.5c00892. https://doi.org/10.1021/jacsau.5c00892
  10. Ahn, Do-Hwan; Alleva, Claudia; Reichenbach, Tom; Gulati, Ashutosh; Ruda, Alessandro; Bonaccorsi, Marta; Silberberg, Jakob M.; Claesson, Magnus; Suades, Albert; Delemotte, Lucie; Widmalm, Goran; Drew, David. A two-step mechanism for sugar translocation. Nature Structural and Molecular Biology 33, 652-663 (2026). DOI 10.1038/s41594-026-01784-w. https://doi.org/10.1038/s41594-026-01784-w
  11. Shugaeva, Tatiana; Howard, Rebecca J.; Haloi, Nandan; Lindahl, Erik. Modeling cryo-EM structures in alternative states with AlphaFold2-based models and density-guided simulations. Communications Chemistry 8, 317 (2025). DOI 10.1038/s42004-025-01751-4. https://doi.org/10.1038/s42004-025-01751-4
  12. Bozzi, Aaron T.; Gaudet, Rachelle. Molecular mechanism of Nramp-family transition metal transport. Journal of Molecular Biology 433, 166991 (2021). DOI 10.1016/j.jmb.2021.166991. https://doi.org/10.1016/j.jmb.2021.166991
  13. Bozzi, Aaron T.; McCabe, Anne L.; Barnett, Benjamin C.; Gaudet, Rachelle. Transmembrane helix 6b links proton and metal release pathways and drives conformational change in an Nramp-family transition metal transporter. Journal of Biological Chemistry 295, 1212-1224 (2020). DOI 10.1074/jbc.ra119.011336. https://doi.org/10.1074/jbc.ra119.011336
  14. Licht, Jacob A.; Berry, Samuel P.; Gutierrez, Michael A.; Gaudet, Rachelle. They all rock: a systematic comparison of conformational movements in LeuT-fold transporters. Structure 32, 1528-1543 (2024). DOI 10.1016/j.str.2024.06.015. https://doi.org/10.1016/j.str.2024.06.015
  15. Tanaka, Yoshiki; Hipolito, Christopher J.; Maturana, Andres D.; Ito, Koichi; Kuroda, Teruo; Higuchi, Takashi; Katoh, Takayuki; Kato, Hideaki E.; Hattori, Motoyuki; Kumazaki, Kaoru; Tsukazaki, Tomoya; Ishitani, Ryuichiro; Suga, Hiroaki; Nureki, Osamu. Structural basis for the drug extrusion mechanism by a MATE multidrug transporter. Nature 496, 247-251 (2013). DOI 10.1038/nature12014. https://doi.org/10.1038/nature12014
  16. Zakrzewska, S.; Mehdipour, A. R.; Malviya, V. N.; Nonaka, T.; Koepke, J.; Muenke, C.; Hausner, W.; Hummer, G.; Safarian, S.; Michel, H. Inward-facing conformation of a multidrug resistance MATE family transporter. Proceedings of the National Academy of Sciences 116, 12275-12284 (2019). DOI 10.1073/pnas.1904210116. https://doi.org/10.1073/pnas.1904210116