Sensitivity of single membrane-spanning alpha-helical peptides to hydrophobic mismatch with a lipid bilayer: effects on backbone structure, orientation, and extent of membrane incorporation

Energetics of the Transmembrane Peptide Sorting by Hydrophobic Mismatch

Hydrophobic mismatch between a lipid membrane and embedded transmembrane peptides or proteins plays a role in their lateral localization and function. Earlier studies have resolved numerous mechanisms through which the peptides and membrane proteins adapt to mismatch, yet the energetics of lateral sorting due to hydrophobic mismatch have remained elusive due to the lack of suitable computational or experimental protocols. Here, we pioneer a molecular dynamics simulation approach to study the sorting of peptides along a membrane thickness gradient. Peptides of different lengths tilt and diffuse along the membrane to eliminate mismatch with a rate directly proportional to the magnitude of mismatch. We extract the 2-dimensional free energy profiles as a function of local thickness and peptide orientation, revealing the relative contributions of sorting and tilting, and suggesting their thermally accessible regimes. Our approach can readily be applied to study other membrane systems of biological interest where hydrophobic mismatch, or membrane thickness in general, plays a role.

Implicit model to capture electrostatic features of membrane environment

Membrane protein structure prediction and design are challenging due to the complexity of capturing the interactions in the lipid layer, such as those arising from electrostatics. Accurately capturing electrostatic energies in the low-dielectric membrane often requires expensive Poisson-Boltzmann calculations that are not scalable for membrane protein structure prediction and design. In this work, we have developed a fast-to-compute implicit energy function that considers the realistic characteristics of different lipid bilayers, making design calculations tractable. This method captures the impact of the lipid head group using a mean-field-based approach and uses a depth-dependent dielectric constant to characterize the membrane environment. This energy function Franklin2023 (F23) is built upon Franklin2019 (F19), which is based on experimentally derived hydrophobicity scales in the membrane bilayer. We evaluated the performance of F23 on five different tests probing (1) protein orientation in the bilayer, (2) stability, and (3) sequence recovery. Relative to F19, F23 has improved the calculation of the tilt angle of membrane proteins for 90% of WALP peptides, 15% of TM-peptides, and 25% of the adsorbed peptides. The performances for stability and design tests were equivalent for F19 and F23. The speed and calibration of the implicit model will help F23 access biophysical phenomena at long time and length scales and accelerate the membrane protein design pipeline.

In Search of a Molecular View of Peptide-Lipid Interactions in Membranes

Lipid bilayer membranes are often represented as a continuous nonpolar slab with a certain thickness bounded by two more polar interfaces. Phenomena such as peptide binding to the membrane surface, folding, insertion, translocation, and diffusion are typically interpreted on the basis of this view. In this Perspective, I argue that this membrane representation as a hydrophobic continuum solvent is not adequate to understand peptide-lipid interactions. Lipids are not small compared to membrane-active peptides: their sizes are similar. Therefore, peptide diffusion needs to be understood in terms of free volume, not classical continuum mechanics; peptide solubility or partitioning in membranes cannot be interpreted in terms of hydrophobic mismatch between membrane thickness and peptide length; peptide folding and translocation, often involving cationic peptides, can only be understood if realizing that lipids adapt to the presence of peptides and the membrane may undergo considerable lipid redistribution in the process. In all of those instances, the detailed molecular interactions between the peptide residues and the lipid components are essential to understand the mechanisms involved.

Implicit model to capture electrostatic features of membrane environment

Membrane protein structure prediction and design are challenging due to the complexity of capturing the interactions in the lipid layer, such as those arising from electrostatics. Accurately capturing electrostatic energies in the low-dielectric membrane often requires expensive Poisson-Boltzmann calculations that are not scalable for membrane protein structure prediction and design. In this work, we have developed a fast-to-compute implicit energy function that considers the realistic characteristics of different lipid bilayers, making design calculations tractable. This method captures the impact of the lipid head group using a mean-field-based approach and uses a depth-dependent dielectric constant to characterize the membrane environment. This energy function Franklin2023 (F23) is built upon Franklin2019 (F19), which is based on experimentally derived hydrophobicity scales in the membrane bilayer. We evaluated the performance of F23 on five different tests probing (1) protein orientation in the bilayer, (2) stability, and (3) sequence recovery. Relative to F19, F23 has improved the calculation of the tilt angle of membrane proteins for 90% of WALP peptides, 15% of TM-peptides, and 25% of the adsorbed peptides. The performances for stability and design tests were equivalent for F19 and F23. The speed and calibration of the implicit model will help F23 access biophysical phenomena at long time and length scales and accelerate the membrane protein design pipeline.

In Situ Assembly of Transmembrane Proteins from Expressed and Synthetic Components in Giant Unilamellar Vesicles

Reconstituting functional transmembrane (TM) proteins into model membranes is challenging due to the difficulty of expressing hydrophobic TM domains, which often require stabilizing detergents that can perturb protein structure and function. Recent model systems solve this problem by linking the soluble domains of membrane proteins to lipids, using noncovalent conjugation. Herein, we test an alternative solution involving the in vitro assembly of TM proteins from synthetic TM domains and expressed soluble domains using chemoselective peptide ligation. We developed an intein mediated ligation strategy to semisynthesize single-pass TM proteins in synthetic giant unilamellar vesicle (GUV) membranes by covalently attaching soluble protein domains to a synthetic TM polypeptide, avoiding the requirement for detergent. We show that the extracellular domain of programmed cell death protein 1, a mammalian immune checkpoint receptor, retains its ligand-binding function at a membrane interface after ligation to a synthetic TM peptide in GUVs, facilitating the study of receptor-ligand interactions.

The Electrostatic Basis of Diacylglycerol Pyrophosphate—Protein Interaction

Diacylglycerol pyrophosphate (DGPP) is an anionic phospholipid formed in plants, yeast, and parasites under multiple stress stimuli. It is synthesized by the phosphorylation action of phosphatidic acid (PA) kinase on phosphatidic acid, a signaling lipid with multifunctional properties. PA functions in the membrane through the interaction of its negatively charged phosphomonoester headgroup with positively charged proteins and ions. DGPP, like PA, can interact electrostatically via the electrostatic-hydrogen bond switch mechanism but differs from PA in its overall charge and shape. The formation of DGPP from PA alters the physicochemical properties as well as the structural dynamics of the membrane. This potentially impacts the molecular and ionic binding of cationic proteins and ions with the DGPP enriched membrane. However, the results of these important interactions in the stress response and in DGPP’s overall intracellular function is unknown. Here, using ³¹P MAS NMR, we analyze the effect of the interaction of low DGPP concentrations in model membranes with the peptides KALP23 and WALP23, which are flanked by positively charged Lysine and neutral Tryptophan residues, respectively. Our results show a significant effect of KALP23 on the charge of DGPP as compared to WALP23. There was, however, no significant effect on the charge of the phosphomonoester of DGPP due to the interaction with positively charged lipids, dioleoyl trimethylammonium propane (DOTAP) and dioleoyl ethyl-phosphatidylcholine (EtPC). Divalent calcium and magnesium cations induce deprotonation of the DGPP headgroup but showed no noticeable differences on DGPP’s charge. Our results lead to a novel model for DGPP—protein interaction.

Contribution of Membrane Lipids to Postsynaptic Protein Organization

The precise subsynaptic organization of proteins at the postsynaptic membrane controls synaptic transmission. In particular, postsynaptic receptor complexes are concentrated in distinct membrane nanodomains to optimize synaptic signaling. However, despite the clear functional relevance of subsynaptic receptor organization to synaptic transmission and plasticity, the mechanisms that underlie the nanoscale organization of the postsynaptic membrane remain elusive. Over the last decades, the field has predominantly focused on the role of protein-protein interactions in receptor trafficking and positioning in the synaptic membrane. In contrast, the contribution of lipids, the principal constituents of the membrane, to receptor positioning at the synapse remains poorly understood. Nevertheless, there is compelling evidence that the synaptic membrane is enriched in specific lipid species and that deregulation of lipid homeostasis in neurons severely affects synaptic functioning. In this review we focus on how lipids are organized at the synaptic membrane, with special emphasis on how current models of membrane organization could contribute to protein distribution at the synapse and synaptic transmission. Finally, we will present an outlook on how novel technical developments could be applied to study the dynamic interplay between lipids and proteins at the postsynaptic membrane.

Membrane Activity of a DNA-Based Ion Channel Depends on the Stability of Its Double-Stranded Structure

DNA nanotechnology has emerged as a promising method for designing spontaneously inserting and fully controllable synthetic ion channels. However, both insertion efficiency and stability of existing DNA-based membrane channels leave much room for improvement. Here, we demonstrate an approach to overcoming the unfavorable DNA-lipid interactions that hinder the formation of a stable transmembrane pore. Our all-atom MD simulations and experiments show that the insertion-driving cholesterol modifications can cause fraying of terminal base pairs of nicked DNA constructs, distorting them when embedded in a lipid bilayer. Importantly, we show that DNA nanostructures with no backbone discontinuities form more stable conductive pores and insert into membranes with a higher efficiency than the equivalent nicked constructs. Moreover, lack of nicks allows design and maintenance of membrane-spanning helices in a tilted orientation within the lipid bilayer. Thus, reducing the conformational degrees of freedom of the DNA nanostructures enables better control over their function as synthetic ion channels.

Single-molecule fluorescence vistas of how lipids regulate membrane proteins

The study of membrane proteins is undergoing a golden era, and we are gaining unprecedented knowledge on how this key group of proteins works. However, we still have only a basic understanding of how the chemical composition and the physical properties of lipid bilayers control the activity of membrane proteins. Single-molecule (SM) fluorescence methods can resolve sample heterogeneity, allowing to discriminate between the different molecular populations that biological systems often adopt. This short review highlights relevant examples of how SM fluorescence methodologies can illuminate the different ways in which lipids regulate the activity of membrane proteins. These studies are not limited to lipid molecules acting as ligands, but also consider how the physical properties of the bilayer can be determining factors on how membrane proteins function.

Carbon nanotube transmembrane channel formation and single-stranded DNA spontaneous internalization: a dissipative particle dynamics study

Single-walled carbon nanotube (SWCNT) transmembrane channel formation in a pure 1,2-dimyristoyl-sn-glycero-3-phosphorylcholine (DMPC) bilayer, and the spontaneous internalization of single-stranded DNA (ssDNA) into the formed pore were simulated. A combination of computational techniques, Dissipative Particle Dynamics-Monte Carlo hybrid simulations and quantum mechanical calculations at the hybrid-DFT level, was used as a new proposal to perform DPD simulations granting specific chemical identity to the model particles. The simulated transmembrane channels showed that, in the case of pristine SWCNTs and upon increasing the nanotube length, a higher tilt angle with respect to the bilayer normal is observed and more time is needed for the nanotube to stabilize. On the other hand, for SWCNTs with polar rims an almost perpendicular orientation is preferred with less than 15° of tilt with respect to the bilayer normal once the nanotubes have pierced both monolayers. These findings are supported by experimental observations where CNTs of average inner diameters of 1.51 ± 0.21 nm and lengths in the 5-15 nm range were inserted in DOPC membranes [J. Geng, et al., Nature, 2014, 514(7524), 612-615]. Moreover, the narrower the SWCNTs, the slower the spontaneous internalization of ssDNA becomes, and ssDNA ends hydrophobically trapped inside the artificial pore. A dependence on the nucleotide content is found indicating that the higher the presence of adenine and thymine in the ssDNA chains the slower the internalization becomes, in agreement with the experimental [A. M. Ababneh, et al., Biophys. J., 2003, 85(2), 1111-1127] and predicted solvation tendency in water for nucleic acid bases.

A Paradigm for Peptide Hormone-GPCR Analyses

Work from our laboratories over the last 35 years that has focused on Ste2p, a G protein-coupled receptor (GPCR), and its tridecapeptide ligand α-factor is reviewed. Our work utilized the yeast Saccharomyces cerevisiae as a model system for understanding peptide-GPCR interactions. It explored the structure and function of synthetic α-factor analogs and biosynthetic receptor domains, as well as designed mutations of Ste2p. The results and conclusions are described using the nuclear magnetic resonance interrogation of synthetic Ste2p transmembrane domains (TMs), the fluorescence interrogation of agonist and antagonist binding, the biochemical crosslinking of peptide analogs to Ste2p, and the phenotypes of receptor mutants. We identified the ligand-binding domain in Ste2p, the functional assemblies of TMs, unexpected and interesting ligand analogs; gained insights into the bound α-factor structure; and unraveled the function and structures of various Ste2p domains, including the N-terminus, TMs, loops connecting the TMs, and the C-terminus. Our studies showed interactions between specific residues of Ste2p in an active state, but not resting state, and the effect of ligand activation on the dimerization of Ste2p. We show that, using a battery of different biochemical and genetic approaches, deep insight can be gained into the structure and conformational dynamics of GPCR-peptide interactions in the absence of a crystal structure.

Modification and Targeted Design of N-Terminal Truncates Derived from Brevinin with Improved Therapeutic Efficacy

Antimicrobial peptides (AMPs) are a class of molecules that play an essential role in innate immune regulation. The Brevinin-1 family are AMPs that show strong pharmacological and antimicrobial potential. A novel peptide, B1A, was designed based on the primary structure of brevinin-1PLb and brevinin-1PLc. Subsequently, a synthesised replicate was subjected to a series of bioassays and was found to display antimicrobial activity. However, it also displayed high levels of haemolysis in a horse red blood cell haemolytic assay, suggesting potential toxicity. Therefore, we rationally designed a number of B1A analogues with aim of retaining antimicrobial activity, lowering toxicity, and to explore the structure–activity relationship of its N-terminus. B1A and its analogues still retained the “Rana Box” and the FLP-motif, which is a feature of this subfamily. However, the introduction of Lys and Trp residues into the peptide sequences revealed that antimicrobial activity of these analogues remained unchanged once the hydrophobicity and the charge reached the threshold. Hence, the idea that the hydrophobicity saturation in different situations is related to antimicrobial activity can be understood via the structure–activity relationship. Meanwhile, it could also be the starting point for the generation of peptides with specific antimicrobial activity.

Conserved Luminal C-Terminal Domain Dynamically Controls Interdomain Communication in Sarcolipin

Sarcolipin (SLN) mediates Ca²⁺ transport and metabolism in muscle by regulating the activity of the Ca²⁺ pump SERCA. SLN has a conserved luminal C-terminal domain that contributes to its functional divergence among homologous SERCA regulators, but the precise mechanistic role of this domain remains poorly understood. We used all-atom molecular dynamics (MD) simulations of SLN totaling 77.5 μs to show that the N- (NT) and C-terminal (CT) domains function in concert. Analysis of the MD simulations showed that serial deletions of the SLN C-terminus do not affect the stability of the peptide nor induce dissociation of SLN from the membrane but promote a gradual decrease in both the tilt angle of the transmembrane helix and the local thickness of the lipid bilayer. Mutual information analysis showed that the NT and CT domains communicate with each other in SLN and that interdomain communication is partially or completely abolished upon deletion of the conserved segment Tyr29-Tyr31 as well as by serial deletions beyond this domain. Phosphorylation of SLN at residue Thr5 also induces changes in the communication between the CT and NT domains, which thus provides additional evidence for interdomain communication within SLN. We found that interdomain communication is independent of the force field used and lipid composition, which thus demonstrates that communication between the NT and CT domains is an intrinsic functional feature of SLN. We propose the novel hypothesis that the conserved C-terminus is an essential element required for dynamic control of SLN regulatory function.

A Trifunctional Linker for Palmitoylation and Peptide and Protein Localization in Biological Membranes

Attachment of lipophilic groups is an important post-translational modification of proteins, which involves the coupling of one or more anchors such as fatty acids, isoprenoids, phospholipids, or glycosylphosphatidyl inositols. To study its impact on the membrane partitioning of hydrophobic peptides or proteins, we designed a tyrosine-based trifunctional linker. The linker allows the facile incorporation of two different functionalities at a cysteine residue in a single step. We determined the effect of the lipid modification on the membrane partitioning of the synthetic α-helical model peptide WALP with or without here and in all cases below; palmitoyl groups in giant unilamellar vesicles that contain a liquid-ordered (Lo ) and liquid-disordered (Ld ) phase. Introduction of two palmitoyl groups did not alter the localization of the membrane peptides, nor did the membrane thickness or lipid composition. In all cases, the peptide was retained in the Ld phase. These data demonstrate that the Lo domain in model membranes is highly unfavorable for a single membrane-spanning peptide.

Peptide-bicelle interaction: Following variations in size and morphology by a combined NMR-SAXS approach

Changes in membrane properties occurring upon protein interaction are key questions in understanding membrane protein function. To report on the occurring size and shape variation we present here a combined NMR-SAXS method performed under physiological conditions using the same samples, enabling determination of a global parameter, the hydration radius (r_H) and estimating the bicelle shape. We use zwitterionic (DMPC/DHPC) and negatively charged (DMPC/DHPC/DMPG) bicelles and investigate the interaction with model transmembrane and surface active peptides (KALP23 and melittin). ¹H NMR measurements based mostly on the translational diffusion coefficient D determination are used to characterize cmc values of DHPC micelles under the investigated conditions, to describe DHPC distribution with exact determination of the q (long chain/short chain) lipid ratio, to estimate aggregation numbers and effective r_H values. The scattering curve is used to fit a lenticular core-shell model enabling us to describe the bicelle shape in terms of ellipsoidal axis length parameters. For all studied systems formation of oblate ellipsoids is found. Even though the r_G/r_H ratio would be an elegant way to characterize shape variations, we show that changes occurring upon peptide-bicelle interaction in the "effective" size and in the measure on the anisometry - morphology - of the objects can be described by using r_H and the simplistic ellipsoidal core-shell model. While the influence of the transmembrane KALP peptide is significant, effects upon addition of surface active melittin peptide seem negligible. This synergy of techniques under controlled conditions can provide information about bicellar shape modulation occurring during peptide-bicelle interactions.

Manipulation of Bacterial Signaling Using Engineered Histidine Kinases

Two-component systems allow bacteria to respond to changes in environmental or cytosolic conditions through autophosphorylation of a histidine kinase (HK) and subsequent transfer of the phosphate group to its downstream cognate response regulator (RR). The RR then elicits a cellular response, commonly through regulation of transcription. Engineering two-component system signaling networks provides a strategy to study bacterial signaling mechanisms related to bacterial cell survival, symbiosis, and virulence, and to develop sensory devices in synthetic biology. Here we focus on the principles for engineering the HK to identify unknown signal inputs, test signal transmission mechanisms, design small molecule sensors, and rewire two-component signaling networks.

Topological analysis of the gp41 MPER on lipid bilayers relevant to the metastable HIV-1 envelope prefusion state

The membrane proximal external region (MPER) of HIV-1 envelope glycoprotein (gp) 41 is an attractive vaccine target for elicitation of broadly neutralizing antibodies (bNAbs) by vaccination. However, current details regarding the quaternary structural organization of the MPER within the native prefusion trimer [(gp120/41)₃] are elusive and even contradictory, hindering rational MPER immunogen design. To better understand the structural topology of the MPER on the lipid bilayer, the adjacent transmembrane domain (TMD) was appended (MPER-TMD) and studied. Membrane insertion of the MPER-TMD was sensitive both to the TMD sequence and cytoplasmic residues. Antigen binding of MPER-specific bNAbs, in particular 10E8 and DH511.2_K3, was significantly impacted by the presence of the TMD. Furthermore, MPER-TMD assembly into 10-nm diameter nanodiscs revealed a heterogeneous membrane array comprised largely of monomers and dimers, as enumerated by bNAb Fab binding using single-particle electron microscopy analysis, arguing against preferential trimeric association of native MPER and TMD protein segments. Moreover, introduction of isoleucine mutations in the C-terminal heptad repeat to induce an extended MPER α-helical bundle structure yielded an antigenicity profile of cell surface-arrayed Env variants inconsistent with that found in the native prefusion state. In line with these observations, electron paramagnetic resonance analysis suggested that 10E8 inhibits viral membrane fusion by lifting the MPER N-terminal region out of the viral membrane, mandating the exposure of residues that would be occluded by MPER trimerization. Collectively, our data suggest that the MPER is not a stable trimer, but rather a dynamic segment adapted for structural changes accompanying fusion.

Wavelength-Selective Fluorescence of a Model Transmembrane Peptide: Constrained Dynamics of Interfacial Tryptophan Anchors

WALPs are prototypical, α-helical transmembrane peptides that represent a consensus sequence for transmembrane segments of integral membrane proteins and serve as excellent models for exploring peptide-lipid interactions and hydrophobic mismatch in membranes. Importantly, the WALP peptides are in direct contact with the lipids. They consist of a central stretch of alternating hydrophobic alanine and leucine residues capped at both ends by tryptophans. In this work, we employ wavelength-selective fluorescence approaches to explore the intrinsic fluorescence of tryptophan residues in WALP23 in 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) membranes. Our results show that the four tryptophan residues in WALP23 exhibit an average red edge excitation shift (REES) of 6 nm, implying their localization at the membrane interface, characterized by a restricted microenvironment. This result is supported by fluorescence anisotropy and lifetime measurements as a function of wavelength displayed by WALP23 tryptophans in POPC membranes. These results provide a new approach based on intrinsic fluorescence of interfacial tryptophans to address protein-lipid interaction and hydrophobic mismatch.

Structural characterisation of a full-length mitochondrial outer membrane TOM40 preprotein translocase: implications for its interaction with presequence peptides

Tom40, the central component of the preprotein translocase of the mitochondrial outer membrane (TOM complex), forms the import pore that facilitates the translocation of preproteins across the outer membrane. Though the function of Tom40 has been intensively studied, the details of the interactions between presequence peptides and Tom40 remain unclear. In this study, we expressed rat Tom40 in Escherichia coli and purified it from inclusion bodies before investigating the refolded protein by fluorescence spectroscopy and circular dichroism (CD) spectroscopy. The far-UV CD spectra of the refolded Tom40 in various concentrations of urea revealed that the refolded protein has a well-defined structure consisting mainly of β-sheet. Moreover, the specific binding of presequence peptides to Tom40, which was demonstrated by fluorescence quenching, showed that the refolded purified protein is functional and that the interaction between Tom40 and presequence peptides is mainly electrostatic in nature.

Amide I SFG Spectral Line Width Probes the Lipid-Peptide and Peptide-Peptide Interactions at Cell Membrane In Situ and in Real Time

The balance of lipid-peptide and peptide-peptide interactions at cell membrane is essential to a large variety of cellular processes. In this study, we have experimentally demonstrated for the first time that sum frequency generation vibrational spectroscopy can be used to probe the peptide-peptide and lipid-peptide interactions in cell membrane in situ and in real time by determination of the line width of amide I band of protein backbone. Using a "benchmark" model of α-helical WALP23, it is found that the dominated lipid-peptide interaction causes a narrow line width of the amide I band, whereas the peptide-peptide interaction can markedly broaden the line width. When WALP23 molecules insert into the lipid bilayer, a quite narrow line width of the amide I band is observed because of the lipid-peptide interaction. In contrast, when the peptide lies down on the bilayer surface, the line width of amide I band becomes very broad owing to the peptide-peptide interaction. In terms of the real-time change in the line width, the transition from peptide-peptide interaction to lipid-peptide interaction is monitored during the insertion of WALP23 into 1,2-dipalmitoyl- sn-glycero-3-phospho-(1'- rac-glycerol) (DPPG) lipid bilayer. The dephasing time of a pure α-helical WALP23 in 1-palmitoyl-2-oleoyl- sn-glycero-3-phospho-(1'- rac-glycerol) and DPPG bilayer is determined to be 2.2 and 0.64 ps, respectively. The peptide-peptide interaction can largely accelerate the dephasing time.

Azide-Modified Membrane Lipids: Synthesis, Properties, and Reactivity

In the present work, we describe the synthesis and the temperature-dependent behavior of photoreactive membrane lipids as well as their capability to study peptide/lipid interactions. The modified phospholipids contain an azide group either in the middle part or at the end of an alkyl chain and also differ in the linkage (ester vs ether) of the second alkyl chain. The temperature-dependent aggregation behavior of the azidolipids was studied using differential scanning calorimetry (DSC), Fourier-transform infrared (FTIR) spectroscopy, and small-angle X-ray scattering (SAXS). Aggregate structures were visualized by stain and cryo transmission electron microscopy (TEM) and were further characterized by dynamic light scattering (DLS). We show that the position of the azide group and the type of linkage of the alkyl chain at the sn-2 position of the glycerol influences the type of aggregates formed as well as their long-term stability: P10AzSPC and r12AzSHPC show the formation of extrudable liposomes, which are stable in size during storage. In contrast, azidolipids that carry a terminal azido moiety either form extrudable liposomes, which show time-dependent vesicle fusion (P15AzPdPC), or self-assemble in large sheet-like, nonextrudable aggregates (r15AzPdHPC) where the lipid molecules are arranged in an interdigitated orientation at temperatures below T_m (L_βI phase). Finally, a P10AzSPC:DMPC mixture was used for photochemically induced cross-linking experiments with a transmembrane peptide (WAL-peptide) to demonstrate the applicability of the azidolipids for the analysis of peptide/lipid interactions. The efficiency of photo-cross-linking was monitored by attenuated total reflection infrared (ATR-IR) spectroscopy and mass spectrometry (MS).

Heterogeneous dielectric generalized Born model with a van der Waals term provides improved association energetics of membrane-embedded transmembrane helices

The heterogeneous dielectric generalized Born (HDGB) implicit membrane formalism is extended by the addition of a van der Waals dispersion term to better describe the nonpolar components of the free energy of solvation. The new model, termed HDGBvdW, improves the energy estimates in the hydrophobic interior of the membrane, where polar and charged species are rarely found and nonpolar interactions become significant. The implicit van der Waals term for the membrane environment extends the model from Gallicchio et al. (J. Comput. Chem. 2004, 25, 479) by combining separate contributions from each of the membrane components. The HDGBvdW model is validated with a series of test cases ranging from membrane insertion and pair association profiles of amino acid side chain analogs and transmembrane helices. Overall, the HDGBvdW model leads to increased agreement with explicit membrane simulation results and experimental data. © 2016 Wiley Periodicals, Inc.

Influenza M2 Transmembrane Domain Senses Membrane Heterogeneity and Enhances Membrane Curvature

Targeting host cell membranes by M2 of influenza A virus is important for virus invasion and replication. We study the transmembrane domain of M2 (M2TM) interacting with mica-supported planar bilayers and free-standing giant unilamellar vesicles (GUVs). Using solution atomic force microscopy (AFM), we show that the size of M2TM oligomers is dependent on lipid composition. The addition of M2TM to lipid bilayers containing liquid-ordered (Lo) and liquid-disordered (Ld) phases reveals that M2TM preferentially partitions into the Ld phase; phase-dependent partitioning results in a larger rigidity of the Ld phase. We next use fluorescence microscopy to study the effects of M2TM on phase-coexisting GUVs. In particular, M2TM is found to increase GUVs' miscibility transition temperature Tmix. The augmented thermodynamic stability can be accounted for by considering an enhanced energy barrier of lipid mixing between coexisting phases. Our GUV study also shows that M2TM can elicit an array of vesicle shapes mimicking virus budding. M2TM enhanced membrane curvature is consistent with our AFM data, which show altered membrane rigidity and consequently line tension at domain edges. Together, our results highlight that in addition to conducting protons, M2TM can actively regulate membrane heterogeneity and augment membrane curvature.

Multiple locations of peptides in the hydrocarbon core of gel-phase membranes revealed by peptide (13)C to lipid (2)H rotational-echo double-resonance solid-state nuclear magnetic resonance

Membrane locations of peptides and proteins are often critical to their functions. Solid-state rotational-echo double-resonance (REDOR) nuclear magnetic resonance is applied to probe the locations of two peptides via peptide (13)CO to lipid (2)H distance measurements. The peptides are KALP, an α-helical membrane-spanning peptide, and HFP, the β-sheet N-terminal fusion peptide of the HIV gp41 fusion protein that plays an important role in HIV-host cell membrane fusion. Both peptides are shown to have at least two distinct locations within the hydrocarbon core of gel-phase membranes. The multiple locations are attributed to snorkeling of lysine side chains for KALP and to the distribution of antiparallel β-sheet registries for HFP. The relative population of each location is also quantitated. To the best of our knowledge, this is the first clear experimental support of multiple peptide locations within the membrane hydrocarbon core. These data are for gel-phase membranes, but the approach should work for liquid-ordered membranes containing cholesterol and may be applicable to liquid-disordered membranes with appropriate additional analysis to take into account protein and lipid motion. This paper also describes the methodological development of (13)CO-(2)H REDOR using the lyophilized I4 peptide that is α-helical and (13)CO-labeled at A9 and (2)Hα-labeled at A8. The I4 spins are well-approximated as an ensemble of isolated (13)CO-(2)H spin pairs each separated by 5.0 Å with a 37 Hz dipolar coupling. A pulse sequence with rectangular 100 kHz (2)H π pulses results in rapid and extensive buildup of REDOR (ΔS/S0) with a dephasing time (τ). The buildup is well-fit by a simple exponential function with a rate of 24 Hz and an extent close to 1. These parameter values reflect nonradiative transitions between the (2)H spin states during the dephasing period. Each spin pair spends approximately two-thirds of its time in the (13)CO-(2)H (m = ±1) states and approximately one-third of its time in the (13)CO-(2)H (m = 0) state and contributes to the ΔS/S0 buildup during the former but not the latter time segments.

Membrane interactions in small fast-tumbling bicelles as studied by 31P NMR

Small fast-tumbling bicelles are ideal for studies of membrane interactions at molecular level; they allow analysis of lipid properties using solution-state NMR. In the present study we used 31P NMR relaxation to obtain detailed information on lipid head-group dynamics. We explored the effect of two topologically different membrane-interacting peptides on bicelles containing either dimyristoylphosphocholine (DMPC), or a mixture of DMPC and dimyristoylphosphoglycerol (DMPG), and dihexanoylphosphocholine (DHPC). KALP21 is a model transmembrane peptide, designed to span a DMPC bilayer and dynorphin B is a membrane surface active neuropeptide. KALP21 causes significant increase in bicelle size, as evidenced by both dynamic light scattering and 31P T2 relaxation measurements. The effect of dynorphin B on bicelle size is more modest, although significant effects on T2 relaxation are observed at higher temperatures. A comparison of 31P T1 values for the lipids with and without the peptides showed that dynorphin B has a greater effect on lipid head-group dynamics than KALP21, especially at elevated temperatures. From the field-dependence of T1 relaxation data, a correlation time describing the overall lipid motion was derived. Results indicate that the positively charged dynorphin B decreases the mobility of the lipid molecules--in particular for the negatively charged DMPG--while KALP21 has a more modest influence. Our results demonstrate that while a transmembrane peptide has severe effects on overall bilayer properties, the surface bound peptide has a more dramatic effect in reducing lipid head-group mobility. These observations may be of general importance for understanding peptide-membrane interactions.

Transferring the PRIMO Coarse-Grained Force Field to the Membrane Environment: Simulations of Membrane Proteins and Helix-Helix Association

An extension of the recently developed PRIMO coarse-grained force field to membrane environments, PRIMO-M, is described. The membrane environment is modeled with the heterogeneous dielectric generalized Born (HDGB) methodology that simply replaces the standard generalized Born model in PRIMO without further parametrization. The resulting model was validated by comparing amino acid insertion free energy profiles and application in molecular dynamics simulations of membrane proteins and membrane-interacting peptides. Membrane proteins with 148–661 amino acids show stable root-mean-squared-deviations (RMSD) between 2 and 4 Å for most systems. Transmembrane helical peptides maintain helical shape and exhibit tilt angles in good agreement with experimental or other simulation data. The association of two glycophorin A (GpA) helices was simulated using replica exchange molecular dynamics simulations yielding the correct dimer structure with a crossing angle in agreement with previous studies. Finally, conformational sampling of the influenza fusion peptide also generates structures in agreement with previous studies. Overall, these findings suggest that PRIMO-M can be used to study membrane bound peptides and proteins and validates the transferable nature of the PRIMO coarse-grained force field.

Solution NMR studies of cell-penetrating peptides in model membrane systems

Cell-penetrating peptides (CPPs) are a class of short, often cationic peptides that have the capability to translocate across cellular membranes, and although the translocation most likely involves several pathways, they interact directly with membranes, as well as with model bilayers. Most CPPs attain a three-dimensional structure when interacting with bilayers, while they are more or less unstructured in aqueous solution. To understand the relationship between structure and the effect that CPPs have on membranes it is of great importance to investigate CPPs at atomic resolution in a suitable membrane model. Moreover, the location in bilayers is likely to be correlated with the translocation mechanism. Solution-state NMR offers a unique possibility to investigate structure, dynamics and location of proteins and peptides in bilayers. This review focuses on solution NMR as a tool for investigating CPP-lipid interactions. Structural propensities and cell-penetrating capabilities can be derived from a combination of CPP solution structures and studies of the effect that the peptides have on bilayers and the localization in a bilayer.

A putative transmembrane leucine zipper of agrobacterium VirB10 is essential for t-pilus biogenesis but not type IV secretion

The Agrobacterium tumefaciens VirB/VirD4 type IV secretion system is composed of a translocation channel and an extracellular T pilus. Bitopic VirB10, the VirB7 lipoprotein, and VirB9 interact to form a cell envelope-spanning structural scaffold termed the "core complex" that is required for the assembly of both structures. The related pKM101-encoded core complex is composed of 14 copies each of these VirB homologs, and the transmembrane (TM) α helices of VirB10-like TraF form a 55-Å-diameter ring at the inner membrane. Here, we report that the VirB10 TM helix possesses two types of putative dimerization motifs, a GxxxA (GA4) motif and two leucine (Leu1, Leu2) zippers. Mutations in the Leu1 motif disrupted T-pilus biogenesis, but these or other mutations in the GA4 or Leu2 motif did not abolish substrate transfer. Replacement of the VirB10 TM domain with a nondimerizing poly-Leu/Ala TM domain sequence also blocked pilus production but not substrate transfer or formation of immunoprecipitable complexes with the core subunits VirB7 and VirB9 and the substrate receptor VirD4. The VirB10 TM helix formed weak homodimers in Escherichia coli, as determined with the TOXCAT assay, whereas replacement of the VirB10 TM helix with the strongly dimerizing TM helix from glycophorin A blocked T-pilus biogenesis in A. tumefaciens. Our findings support a model in which VirB10's TM helix contributes to the assembly or activity of the translocation channel as a weakly self-interacting membrane anchor but establishes a heteromeric TM-TM helix interaction via its Leu1 motif that is critical for T-pilus biogenesis.

Formation of raft-like assemblies within clusters of influenza hemagglutinin observed by MD simulations

The association of hemagglutinin (HA) with lipid rafts in the plasma membrane is an important feature of the assembly process of influenza virus A. Lipid rafts are thought to be small, fluctuating patches of membrane enriched in saturated phospholipids, sphingolipids, cholesterol and certain types of protein. However, raft-associating transmembrane (TM) proteins generally partition into Ld domains in model membranes, which are enriched in unsaturated lipids and depleted in saturated lipids and cholesterol. The reason for this apparent disparity in behavior is unclear, but model membranes differ from the plasma membrane in a number of ways. In particular, the higher protein concentration in the plasma membrane may influence the partitioning of membrane proteins for rafts. To investigate the effect of high local protein concentration, we have conducted coarse-grained molecular dynamics (CG MD) simulations of HA clusters in domain-forming bilayers. During the simulations, we observed a continuous increase in the proportion of raft-type lipids (saturated phospholipids and cholesterol) within the area of membrane spanned by the protein cluster. Lateral diffusion of unsaturated lipids was significantly attenuated within the cluster, while saturated lipids were relatively unaffected. On this basis, we suggest a possible explanation for the change in lipid distribution, namely that steric crowding by the slow-diffusing proteins increases the chemical potential for unsaturated lipids within the cluster region. We therefore suggest that a local aggregation of HA can be sufficient to drive association of the protein with raft-type lipids. This may also represent a general mechanism for the targeting of TM proteins to rafts in the plasma membrane, which is of functional importance in a wide range of cellular processes.

The cell membrane is composed of a wide variety of lipids and proteins. Until recently, these were thought to be mixed evenly, but we now have evidence of the existence of “lipid rafts” — small, slow-moving areas of membrane in which certain types of lipid and protein accumulate. Rafts have many important biological functions in healthy cells, but also play a role in the assembly of influenza virus. For example, after the viral protein hemagglutinin is made inside the host cell, it accumulates in rafts. Exiting virus particles then take these portions of cell membrane with them as they leave the host cell. However, the mechanism by which proteins associate with lipid rafts is unclear. Here, we have used computers to simulate lipid membranes containing hemagglutinin. The simulations allow us to look in detail at the motions and interactions of individual proteins and lipids. We found that clusters of proteins altered the properties of nearby lipids, leading to accumulation of raft-type lipids. It therefore appears that aggregation of hemagglutinin may be enough to drive its association with rafts. This helps us to better understand both the influenza assembly process and the properties of lipid rafts.

Dynamic Heterogeneous Dielectric Generalized Born (DHDGB): An implicit membrane model with a dynamically varying bilayer thickness

An extension to the heterogeneous dielectric generalized Born (HDGB) implicit membrane formalism is presented to allow dynamic membrane deformations in response to membrane-inserted biomolecules during molecular dynamic simulations. The flexible membrane is implemented through additional degrees of freedom that represent the membrane deformation at the contact points of a membrane-inserted solute with the membrane. The extra degrees of freedom determine the dielectric and non-polar solvation free energy profiles that are used to obtain the solvation free energy in the presence of the membrane and are used to calculate membrane deformation free energies according to an elastic membrane model. With the dynamic HDGB (DHDGB) model the membrane is able to deform in response to the insertion of charged molecules thereby avoiding the overestimation of insertion free energies with static membrane models. The DHDGB model also allows the membrane to respond to the insertion of membrane-spanning solutes with hydrophobic mismatch. The model is tested with the membrane insertion of amino acid side chain analogs, arginine-containing helices, the WALP23 peptide, and the gramicidin A channel.

Analyzing the effects of hydrophobic mismatch on transmembrane α-helices using tryptophan fluorescence spectroscopy

Hydrophobic matching between transmembrane protein segments and the lipid bilayer in which they are embedded is a significant factor in the behavior and orientation of such transmembrane segments. The condition of hydrophobic mismatch occurs when the hydrophobic thickness of a lipid bilayer is significantly different than the length of the membrane spanning segment of a protein, resulting in a mismatch. This mismatch can result in altered function of proteins as well as nonnative structural arrangements including effects on transmembrane α-helix tilt angles, oligomerization state, and/or the formation of non-transmembrane topographies. Here, a fluorescence-based protocol is described for testing model transmembrane α-helices and their sensitivity to hydrophobic mismatch by measuring the propensity of these helices to form non-transmembrane structures. Overall, good hydrophobic matching between the bilayer and transmembrane segments is an important factor that must be considered when designing membrane proteins or peptides.

Membrane protein structure determination: back to the membrane

NMR spectroscopy enables the structures of membrane proteins to be determined in the native-like environment of the phospholipid bilayer membrane. This chapter outlines the methods for membrane protein structural studies using solid-state NMR spectroscopy with samples of membrane proteins incorporated in proteoliposomes or planar lipid bilayers. The methods for protein expression and purification, sample preparation, and NMR experiments are described and illustrated with examples from OmpX and Ail, two bacterial outer membrane proteins that function in bacterial virulence.

Residue-specific membrane location of peptides and proteins using specifically and extensively deuterated lipids and ¹³C-²H rotational-echo double-resonance solid-state NMR

Residue-specific location of peptides in the hydrophobic core of membranes was examined using (13)C-(2)H REDOR and samples in which the lipids were selectively deuterated. The transmembrane topology of the KALP peptide was validated with this approach with substantial dephasing observed for deuteration in the bilayer center and reduced or no dephasing for deuteration closer to the headgroups. Insertion of β sheet HIV and helical and β sheet influenza virus fusion peptides into the hydrophobic core of the membrane was validated in samples with extensively deuterated lipids.

Sterol affinity for phospholipid bilayers is influenced by hydrophobic matching between lipids and transmembrane peptides

Lipid self-organization is believed to be essential for shaping the lateral structure of membranes, but it is becoming increasingly clear that also membrane proteins can be involved in the maintenance of membrane architecture. Cholesterol is thought to be important for the lateral organization of eukaryotic cell membranes and has also been implicated to take part in the sorting of cellular transmembrane proteins. Hence, a good starting point for studying the influence of lipid-protein interactions on membrane trafficking is to find out how transmembrane proteins influence the lateral sorting of cholesterol in phospholipid bilayers. By measuring equilibrium partitioning of the fluorescent cholesterol analog cholestatrienol between large unilamellar vesicles and methyl-β-cyclodextrin the effect of hydrophobic matching on the affinity of sterols for phospholipid bilayers was determined. Sterol partitioning was measured in 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC), 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) bilayers with and without WALP19, WALP23 or WALP27 peptides. The results showed that the affinity of the sterol for the bilayers was affected by hydrophobic matching. An increasing positive hydrophobic mismatch led to stronger sterol binding to the bilayers (except in extreme situations), and a large negative hydrophobic mismatch decreased the affinity of the sterol for the bilayer. In addition, peptide insertion into the phospholipid bilayers was observed to depend on hydrophobic matching. In conclusion, the results showed that hydrophobic matching can affect lipid-protein interactions in a way that may facilitate the formation of lateral domains in cell membranes. This could be of importance in membrane trafficking.

Properties of membrane-incorporated WALP peptides that are anchored on only one end

Peptides of the "WALP" family, acetyl-GWW(LA)(n)LWWA-[ethanol]amide, have proven to be opportune models for investigating lipid-peptide interactions. Because the average orientations and motional behavior of the N- and C-terminal Trp (W) residues differ, it is of interest to investigate how the positions of the tryptophans influence the properties of the membrane-incorporated peptides. To address this question, we synthesized acetyl-GGWW(LA)(n)-ethanolamide and acetyl-(AL)(n)WWG-ethanolamide, in which n = 4 or 8, which we designate as "N-anchored" and "C-anchored" peptides, respectively. Selected (2)H or (15)N labels were incorporated for solid-state nuclear magnetic resonance (NMR) spectroscopy. These peptides can be considered "half"-anchored WALP peptides, having only one pair of interfacial Trp residues near either the amino or the carboxyl terminus. The hydrophobic lengths of the (n = 8) peptides are similar to that of WALP23. These longer half-anchored WALP peptides incorporate into lipid bilayers as α-helices, as reflected in their circular dichroism spectra. Solid-state NMR experiments indicate that the longer peptide helices assume defined transmembrane orientations with small non-zero average tilt angles and moderate to high dynamic averaging in bilayer membranes of 1,2-dioleoylphosphatidylcholine, 1,2-dimyristoylphosphatidylcholine, and 1,2-dilauroylphosphatidylcholine. The intrinsically small apparent tilt angles suggest that interactions of aromatic residues with lipid headgroups may play an important role in determining the magnitude of the peptide tilt in the bilayer membrane. The shorter (n = 4) peptides, in stark contrast to the longer peptides, display NMR spectra that are characteristic of greatly reduced motional averaging, probably because of peptide aggregation in the bilayer environment, and CD spectra that are characteristic of β-structure.

The determinants of hydrophobic mismatch response for transmembrane helices

Hydrophobic mismatch arises from a difference in the hydrophobic thickness of a lipid membrane and a transmembrane protein segment, and is thought to play an important role in the folding, stability and function of membrane proteins. We have investigated the possible adaptations that lipid bilayers and transmembrane α-helices undergo in response to mismatch, using fully-atomistic molecular dynamics simulations totaling 1.4 μs. We have created 25 different tryptophan-alanine-leucine transmembrane α-helical peptide systems, each composed of a hydrophobic alanine-leucine stretch, flanked by 1-4 tryptophan side chains, as well as the β-helical peptide dimer, gramicidin A. Membrane responses to mismatch include changes in local bilayer thickness and lipid order, varying systematically with peptide length. Adding more flanking tryptophan side chains led to an increase in bilayer thinning for negatively mismatched peptides, though it was also associated with a spreading of the bilayer interface. Peptide tilting, bending and stretching were systematic, with tilting dominating the responses, with values of up to ~45° for the most positively mismatched peptides. Peptide responses were modulated by the number of tryptophan side chains due to their anchoring roles and distributions around the helices. Potential of mean force calculations for local membrane thickness changes, helix tilting, bending and stretching revealed that membrane deformation is the least energetically costly of all mismatch responses, except for positively mismatched peptides where helix tilting also contributes substantially. This comparison of energetic driving forces of mismatch responses allows for deeper insight into protein stability and conformational changes in lipid membranes.

Transmembrane domains are critical to the interaction between platelet glycoprotein V and glycoprotein Ib-IX complex

BACKGROUND

The glycoprotein (GP) Ib-IX-V complex, the von Willebrand factor receptor on the platelet surface, is critically involved in hemostasis and thrombosis. The GPV subunit interacts with GPIb-IX to form the GPIb-IX-V complex, but the underlying molecular basis remains unclear. It was observed earlier that efficient expression of GPV in the plasma membrane requires co-expression of GPIb-IX.

OBJECTIVES AND METHODS

Hypothesizing that GPIb-IX stabilizes GPV through direct interaction and consequently enhances GPV surface expression, we aim in this study to identify structural elements in the complex that mediate the interaction between GPV and GPIb-IX by analyzing mutational effects on GPV surface expression in transfected Chinese hamster ovary cells.

RESULTS

Enhancement of GPV surface expression by GPIb-IX requires transmembrane domains of both GPV and GPIbα, as replacing the GPV transmembrane domain with an unrelated poly-leucine-alanine sequence abolished the enhancing effect of GPIb-IX. Additional mutagenesis analysis of the GPV transmembrane helix identified three helical sides containing conserved polar residues as critical to efficient GPV surface expression. Similarly, replacing residues in three sides (Gly495/Ala502/Leu509, Phe491/Trp498/Val505, and Y492/L499/L506) of the GPIbα transmembrane domain with leucines preserved the surface expression level of GPIb-IX but significantly altered that of GPV.

CONCLUSIONS

Our results demonstrate for the first time the importance of transmembrane domains for efficient surface expression of GPV and suggest that GPV and GPIbα transmembrane domains interact with each other, contributing to assembly of the GPIb-IX-V complex.

Membrane physical properties influence transmembrane helix formation

The pHLIP peptide has three states: (I) soluble in aqueous buffer, (II) bound to the bilayer surface at neutral pH, and (III) inserted as a transmembrane (TM) helix at acidic pH. The membrane insertion of pHLIP at low pH can be used to target the acidic tissues characteristic of different diseases, such as cancer. We find that the α-helix content of state II depends on lipid acyl chain length but not cholesterol, suggesting the helicity of the bound state may be controlled by the bilayer elastic bending modulus. Experiments with the P20G variant show the proline residue in pHLIP reduces the α-helix content of both states II and III. We also observe that the membrane insertion pKa is influenced by membrane physical properties, following a biphasic pattern similar to the membrane thickness optima observed for the function of eukaryotic membrane proteins. Because tumor cells exhibit altered membrane fluidity, we suggest this might influence pHLIP tumor targeting. We used a cell insertion assay to determine the pKa in live cells, observing that the properties in liposomes held in the more complex plasma membrane. Our results show that the formation of a TM helix is modulated by both the conformational propensities of the peptide and the physical properties of the bilayer. These results suggest a physical role for helix-membrane interactions in optimizing the function of more complex TM proteins.

Multiple Pathway Relaxation Enhancement in the System Composed of Three Paramagnetic Species: Nitroxide Radical-Ln(3+)-O2

Longitudinal relaxation of nitroxide spin-labels has been measured for a membrane-incorporated α-helical polypeptide in the presence and absence of residual amounts of membrane-dissolved O2 and paramagnetic Dy(3+) ions. Such a model system, containing three different types of paramagnetic species, provides an important example of nonadditivity of two different relaxation channels for the nitroxide spins.

Aggregation of model membrane proteins, modulated by hydrophobic mismatch, membrane curvature, and protein class

Aggregation of transmembrane proteins is important for many biological processes, such as protein sorting and cell signaling, and also for in vitro processes such as two-dimensional crystallization. We have used large-scale simulations to study the lateral organization and dynamics of lipid bilayers containing multiple inserted proteins. Using coarse-grained molecular dynamics simulations, we have studied model membranes comprising ∼7000 lipids and 16 identical copies of model cylindrical proteins of either α-helical or β-barrel types. Through variation of the lipid tail length and hence the degree of hydrophobic mismatch, our simulations display levels of protein aggregation ranging from negligible to extensive. The nature and extent of aggregation are shown to be influenced by membrane curvature and the shape or orientation of the protein. Interestingly, a model β-barrel protein aggregates to form one-dimensional strings within the bilayer plane, whereas a model α-helical bundle forms two-dimensional clusters. Overall, it is clear that the nature and extent of membrane protein aggregation is dependent on several aspects of the proteins and lipids, including hydrophobic mismatch, protein class and shape, and membrane curvature.

Lateral sorting in model membranes by cholesterol-mediated hydrophobic matching

Theoretical studies predict hydrophobic matching between transmembrane domains of proteins and bilayer lipids to be a physical mechanism by which membranes laterally self-organize. We now experimentally study the direct consequences of mismatching of transmembrane peptides of different length with bilayers of different thicknesses at the molecular level. In both model membranes and simulations we show that cholesterol critically constrains structural adaptations at the peptide-lipid interface under mismatch. These constraints translate into a sorting potential and lead to selective lateral segregation of peptides and lipids according to their hydrophobic length.