Acyl and CO Ligands in the [Fe]-Hydrogenase Cofactor Scramble upon Photolysis.

[Fe]-hydrogenase harbors the iron-guanylylpyridinol (FeGP) cofactor, in which the Fe(II) complex contains acyl-carbon, pyridinol-nitrogen, cysteine-thiolate and two CO as ligands. Irradiation with UV-A/blue light decomposes the FeGP cofactor to a 6-carboxymethyl-4-guanylyl-2-pyridone (GP) and other components. Previous in vitro biosynthesis experiments indicated that the acyl- and CO-ligands in the FeGP cofactor can scramble, but whether scrambling occurred during biosynthesis or photolysis was unclear. Here, we demonstrate that the [18 O1 -carboxy]-group of GP is incorporated into the FeGP cofactor by in vitro biosynthesis. MS/MS analysis of the 18 O-labeled FeGP cofactor revealed that the produced [18 O1 ]-acyl group is not exchanged with a CO ligand of the cofactor, indicating that the acyl and CO ligands are scrambled during photolysis rather than biosynthesis, which ruled out any biosynthesis mechanisms allowing acyl/CO ligands scrambling. Time-resolved infrared spectroscopy indicated that an acyl-Fe(CO)3 intermediate is formed during photolysis, in which scrambling of the CO and acyl ligands can occur. This finding also suggests that the light-excited FeGP cofactor has a higher affinity for external CO. These results contribute to our understanding of the biosynthesis and photosensitive properties of this unique H2 -activating natural complex.

Introducing Aliphatic Fluoropeptides: Perspectives on Folding Properties, Membrane Partition and Proteolytic Stability.

A de novo designed class of peptide-based fluoropolymers composed of fluorinated aliphatic amino acids as main components is reported. Structural characterization provided insights into fluorine-induced alterations on ß-strand to α-helix transition upon an increase in SDS content and revealed the unique formation of PPII structures for trifluorinated fluoropeptides. A combination of circular dichroism, fluorescence-based leaking assays and surface enhanced infrared absorption spectroscopy served to examine the insertion and folding processes into unilamellar vesicles. While partitioning into lipid bilayers, the degree of fluorination conducts a decrease in α-helical content. Furthermore, this study comprises a report on the proteolytic stability of peptides exclusively built up by fluorinated amino acids and proved all sequences to be enzymatically degradable despite the degree of fluorination. Herein presented fluoropeptides as well as the distinctive properties of these artificial and polyfluorinated foldamers with enzyme-degradable features will play a crucial role in the future development of fluorinated peptide-based biomaterials.

Hydrophobicity of Self-Assembled Monolayers of Alkanes: Fluorination, Density, Roughness, and Lennard-Jones Cutoffs.

The interplay of fluorination and structure of alkane self-assembled monolayers and how these affect hydrophobicity are explored via molecular dynamics simulations, contact angle goniometry, and surface-enhanced infrared absorption spectroscopy. Wetting coefficients are found to grow linearly in the monolayer density for both alkane and perfluoroalkane monolayers. The larger contact angles of monolayers of perfluorinated alkanes are shown to be primarily caused by their larger molecular volume, which leads to a larger nearest-neighbor grafting distance and smaller tilt angle. Increasing the Lennard-Jones force cutoff in simulations is found to increase hydrophilicity. Specifically, wetting coefficients scale like the inverse square of the cutoff, and when extrapolated to the infinite cutoff limit, they yield contact angles that compare favorably to experimental values. Nanoscale roughness is also found to reliably increase monolayer hydrophobicity, mostly via the reduction of the entropic part of the work of adhesion. Analysis of depletion lengths shows that droplets on nanorough surfaces partially penetrate the surface, intermediate between Wenzel and Cassie-Baxter states.

Electrostatics, proton sensor, and networks governing the gating transition in GLIC, a proton-gated pentameric ion channel.

The pentameric ligand-gated ion channel (pLGIC) from Gloeobacter violaceus (GLIC) has provided insightful structure-function views on the permeation process and the allosteric regulation of the pLGICs family. However, GLIC is activated by pH instead of a neurotransmitter and a clear picture for the gating transition driven by protons is still lacking. We used an electrostatics-based (finite difference Poisson-Boltzmann/Debye-Hückel) method to predict the acidities of all aspartic and glutamic residues in GLIC, both in its active and closed-channel states. Those residues with a predicted pKa close to the experimental pH50 were individually replaced by alanine and the resulting variant receptors were titrated by ATR/FTIR spectroscopy. E35, located in front of loop F far away from the orthosteric site, appears as the key proton sensor with a measured individual pKa at 5.8. In the GLIC open conformation, E35 is connected through a water-mediated hydrogen-bond network first to the highly conserved electrostatic triad R192-D122-D32 and then to Y197-Y119-K248, both located at the extracellular domain-transmembrane domain interface. The second triad controls a cluster of hydrophobic side chains from the M2-M3 loop that is remodeled during the gating transition. We solved 12 crystal structures of GLIC mutants, 6 of them being trapped in an agonist-bound but nonconductive conformation. Combined with previous data, this reveals two branches of a continuous network originating from E35 that reach, independently, the middle transmembrane region of two adjacent subunits. We conclude that GLIC's gating proceeds by making use of loop F, already known as an allosteric site in other pLGICs, instead of the classic orthosteric site.

Diversifying Metal-Ligand Cooperative Catalysis in Semi-Synthetic [Mn]-Hydrogenases.

The reconstitution of [Mn]-hydrogenases using a series of MnI complexes is described. These complexes are designed to have an internal base or pro-base that may participate in metal-ligand cooperative catalysis or have no internal base or pro-base. Only MnI complexes with an internal base or pro-base are active for H2 activation; only [Mn]-hydrogenases incorporating such complexes are active for hydrogenase reactions. These results confirm the essential role of metal-ligand cooperation for H2 activation by the MnI complexes alone and by [Mn]-hydrogenases. Owing to the nature and position of the internal base or pro-base, the mode of metal-ligand cooperation in two active [Mn]-hydrogenases is different from that of the native [Fe]-hydrogenase. One [Mn]-hydrogenase has the highest specific activity of semi-synthetic [Mn]- and [Fe]-hydrogenases. This work demonstrates reconstitution of active artificial hydrogenases using synthetic complexes differing greatly from the native active site.

The Bacterial [Fe]-Hydrogenase Paralog HmdII Uses Tetrahydrofolate Derivatives as Substrates.

[Fe]-hydrogenase (Hmd) catalyzes the reversible hydrogenation of methenyl-tetrahydromethanopterin (methenyl-H4 MPT+ ) with H2 . H4 MPT is a C1-carrier of methanogenic archaea. One bacterial genus, Desulfurobacterium, contains putative genes for the Hmd paralog, termed HmdII, and the HcgA-G proteins. The latter are required for the biosynthesis of the prosthetic group of Hmd, the iron-guanylylpyridinol (FeGP) cofactor. This finding is intriguing because Hmd and HmdII strictly use H4 MPT derivatives that are absent in most bacteria. We identified the presence of the FeGP cofactor in D. thermolithotrophum. The bacterial HmdII reconstituted with the FeGP cofactor catalyzed the hydrogenation of derivatives of tetrahydrofolate, the bacterial C1-carrier, albeit with low enzymatic activities. The crystal structures show how Hmd recognizes tetrahydrofolate derivatives. These findings have an impact on future biotechnology by identifying a bacterial Hmd paralog.

Dioxygen Sensitivity of [Fe]-Hydrogenase in the Presence of Reducing Substrates.

Mono-iron hydrogenase ([Fe]-hydrogenase) reversibly catalyzes the transfer of a hydride ion from H2 to methenyltetrahydromethanopterin (methenyl-H4 MPT+ ) to form methylene-H4 MPT. Its iron guanylylpyridinol (FeGP) cofactor plays a key role in H2 activation. Evidence is presented for O2 sensitivity of [Fe]-hydrogenase under turnover conditions in the presence of reducing substrates, methylene-H4 MPT or methenyl-H4 MPT+ /H2 . Only then, H2 O2 is generated, which decomposes the FeGP cofactor; as demonstrated by spectroscopic analyses and the crystal structure of the deactivated enzyme. O2 reduction to H2 O2 requires a reductant, which can be a catalytic intermediate transiently formed during the [Fe]-hydrogenase reaction. The most probable candidate is an iron hydride species; its presence has already been predicted by theoretical studies of the catalytic reaction. The findings support predictions because the same type of reduction reaction is described for ruthenium hydride complexes that hydrogenate polar compounds.

CuI and H2 O2 Inactivate and FeII Inhibits [Fe]-Hydrogenase at Very Low Concentrations.

[Fe]-Hydrogenase (Hmd) catalyzes reversible hydride transfer from H2 . It harbors an iron-guanylylpyridinol as a cofactor with an FeII that is ligated to one thiolate, two COs, one acyl-C, one pyridinol-N, and solvent. Here, we report that CuI and H2 O2 inactivate Hmd (half-maximal rates at 1 µM CuI and 20 µM H2 O2 ) and that FeII inhibits the enzyme with very high affinity (Ki =40 nM). Infrared and EPR studies together with competitive inhibition studies with isocyanide indicated that CuI exerts its inhibitory effect most probably by binding to the active site iron-thiolate ligand. Using the same methods, it was found that H2 O2 binds to the active-site iron at the solvent-binding site and oxidizes FeII to FeIII . Also it was shown that FeII reversibly binds away from the active site iron, with binding being competitive to the organic hydride acceptor; this inhibition is specific for FeII and is reminiscent of that for the [FeFe]-hydrogenase second iron, which specifically interacts with H2 .

Surface-enhanced infrared absorption spectroscopy (SEIRAS) to probe monolayers of membrane proteins.

Surface-enhanced infrared absorption spectroscopy (SEIRAS) represents a variation of conventional infrared spectroscopy and exploits the signal enhancement exerted by the plasmon resonance of nano-structured metal thin films. The surface enhancement decays in about 10nm with the distance from the surface and is, thus, perfectly suited to selectively probe monolayers of biomembranes. Peculiar to membrane proteins is their vectorial functionality, the probing of which requires proper orientation within the membrane. To this end, the metal surface used in SEIRAS is chemically modified to generate an oriented membrane protein film. Monolayers of uniformly oriented membrane proteins are formed by tethering His-tagged proteins to a nickel nitrilo-triacetic acid (Ni-NTA) modified gold surface and SEIRAS commands molecular sensitivity to probe each step of surface modification. The solid surface used as plasmonic substrate for SEIRAS, can also be employed as an electrode to investigate systems where electron transfer reactions are relevant, like e.g. cytochrome c oxidase or plant-type photosystems. Furthermore, the interaction of these membrane proteins with water-soluble proteins, like cytochrome c or hydrogenase, is studied on the molecular level by SEIRAS. The impact of the membrane potential on protein functionality is verified by monitoring light-dark difference spectra of a monolayer of sensory rhodopsin (SRII) at different applied potentials. It is demonstrated that the interpretations of all of these experiments critically depend on the orientation of the solid-supported membrane protein. Finally, future directions of SEIRAS including cellular systems are discussed. This article is part of a Special Issue entitled: FTIR in membrane proteins and peptide studies.

Kinetics, Thermodynamics, and Structural Effects of Quinoline-2-Carboxylates, Zinc-Binding Inhibitors of New Delhi Metallo-ß-lactamase-1 Re-sensitizing Multidrug-Resistant Bacteria for Carbapenems.

Carbapenem resistance mediated by metallo-ß-lactamases (MBL) such as New Delhi metallo-ß-lactamase-1 (NDM-1) has become a major factor threatening the efficacy of essential ß-lactam antibiotics. Starting from hit fragment dipicolinic acid (DPA), 8-hydroxy- and 8-sulfonamido-quinoline-2-carboxylic acids were developed as inhibitors of NDM-1 with highly improved inhibitory activity and binding affinity. The most active compounds formed reversibly inactive ternary protein-inhibitor complexes with two zinc ions as proven by native protein mass spectrometry and bio-layer interferometry. Modification of the NDM-1 structure with remarkable entropic gain was shown by isothermal titration calorimetry and NMR spectroscopy of isotopically labeled protein. The best compounds were potent inhibitors of NDM-1 and other representative MBL with no or little inhibition of human zinc-binding enzymes. These inhibitors significantly reduced the minimum inhibitory concentrations (MIC) of meropenem for multidrug-resistant bacteria recombinantly expressing blaNDM-1 as well as for several multidrug-resistant clinical strains at concentrations non-toxic to human cells.

Biosynthesis of the iron-guanylylpyridinol cofactor of [Fe]-hydrogenase in methanogenic archaea as elucidated by stable-isotope labeling.

[Fe]-hydrogenase catalyzes the reversible hydride transfer from H(2) to methenyltetrahydromethanoptherin, which is an intermediate in methane formation from H(2) and CO(2) in methanogenic archaea. The enzyme harbors a unique active site iron-guanylylpyridinol (FeGP) cofactor, in which a low-spin Fe(II) is coordinated by a pyridinol-N, an acyl group, two carbon monoxide, and the sulfur of the enzyme's cysteine. Here, we studied the biosynthesis of the FeGP cofactor by following the incorporation of (13)C and (2)H from labeled precursors into the cofactor in growing methanogenic archaea and by subsequent NMR, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS), electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI-FT-ICR-MS) and IR analysis of the isolated cofactor and reference compounds. The pyridinol moiety of the cofactor was found to be synthesized from three C-1 of acetate, two C-2 of acetate, two C-1 of pyruvate, one carbon from the methyl group of l-methionine, and one carbon directly from CO(2). The metabolic origin of the two CO-ligands was CO(2) rather than C-1 or C-2 of acetate or pyruvate excluding that the two CO are derived from dehydroglycine as has previously been shown for the CO-ligands in [FeFe]-hydrogenases. A formation of CO from CO(2) via direct reduction catalyzed by a nickel-dependent CO dehydrogenase or from formate could also be excluded. When the cells were grown in the presence of (13)CO, the two CO-ligands and the acyl group became (13)C-labeled, indicating either that free CO is an intermediate in their synthesis or that free CO can exchange with these iron-bound ligands. Based on these findings, we propose pathways for how the FeGP cofactor might be synthesized.

pH-induced insertion of pHLIP into a lipid bilayer: In-situ SEIRAS characterization of a folding intermediate at neutral pH.

The pH low insertion peptide (pHLIP) is a pH-sensitive cell penetrating peptide that transforms from an unstructured coil on the membrane surface at pH > 7, to a transmembrane (TM) α-helix at pH < 5. By exploiting this unique property, pHLIP attracts interest as a potential tool for drug delivery and visualisation of acidic tissues produced by various maladies such as cancer, inflammation, hypoxia etc. Even though the structures of initial and end states of pHLIP insertion have been widely accepted, the intermediate structures in between these two states are less clear. Here, we have applied in situ Surface-Enhanced Infrared Absorption spectroscopy to examine the pH-induced insertion and folding processes of pHLIP into a solid-supported lipid bilayer. We show that formation of partially helical structure already takes place at pH only slightly below 7.0, but with the helical axis parallel to the membrane surface. The peptide starts to reorientate its helix from horizontal to vertical direction, accompanied by the insertion into the TM region at pH < 6.2. Further insertion into the TM region of the peptide results in an increase of inherent α-helical structure and complete secondary structure formation at pH 5.3. Analysis of the changes of the carboxylate vibrational bands upon pH titration shows two distinctive groups of aspartates and glutamates with pKa values of 4.5 and 6.3, respectively. Comparison to the amide bands of the peptide backbone suggests that the latter Asp/Glu groups are directly involved in the conformational changes of pHLIP in the respective intermediate states.

Monitoring the Progression of Cell-Free Expression of Microbial Rhodopsins by Surface Enhanced IR Spectroscopy: Resolving a Branch Point for Successful/Unsuccessful Folding.

The translocon-unassisted folding process of transmembrane domains of the microbial rhodopsins sensory rhodopsin I (HsSRI) and II (HsSRII), channelrhodopsin II (CrChR2), and bacteriorhodopsin (HsBR) during cell-free expression has been investigated by Surface-Enhanced Infrared Absorption Spectroscopy (SEIRAS). Up to now, only a limited number of rhodopsins have been expressed and folded into the functional holoprotein in cell free expression systems, while other microbial rhodopsins fail to properly bind the chromophore all-trans retinal as indicated by the missing visible absorption. SEIRAS experiments suggest that all investigated rhodopsins lead to the production of polypeptides, which are co-translationally inserted into a solid-supported lipid bilayer during the first hour after the in-vitro expression is initiated. Secondary structure analysis of the IR spectra revealed that the polypeptides form a comparable amount of α-helical structure during the initial phase of insertion into the lipid bilayer. As the process progressed (>1 h), only HsBR exhibited a further increase and association of α-helices to form a compact tertiary structure, while the helical contents of the other rhodopsins stagnated. This result suggests that the molecular reason for the unsuccessful cell-free expression of the two sensory rhodopsins and of CrChR2 is not due to the translation process, but rather to the folding process during the post-translational period. Taking our previous observation into account that HsBR fails to form a tertiary structure in the absence of its retinal, we infer that the chromophore retinal is an integral component of the compaction of the polypeptide into its tertiary structure and the formation of a fully functional protein.

Scanning near-field IR microscopy of proteins in lipid bilayers.

We use infrared near-field microscopy to chemically map the morphology of biological matrices. The investigated sample is built up from surface-tethered membrane proteins (cytochrome c oxidase) reconstituted in a lipid bilayer. We have carried out infrared near-field measurements in the frequency range between 1600 and 1800 cm(-1). By simultaneously recording the topography and chemical fingerprint of the protein-tethered lipid bilayer with a lateral resolution of 80 nm × 80 nm, we were able to probe locally the chemical signature of this membrane and to provide a local map of its surface morphology.

Near-Infrared Activation of Sensory Rhodopsin II Mediated by NIR-to-Blue Upconversion Nanoparticles.

Direct optical activation of microbial rhodopsins in deep biological tissue suffers from ineffective light delivery because visible light is strongly scattered and absorbed. NIR light has deeper tissue penetration, but NIR-activation requires a transducer that converts NIR light into visible light in proximity to proteins of interest. Lanthanide-doped upconversion nanoparticles (UCNPs) are ideal transducer as they absorb near-infrared (NIR) light and emit visible light. Therefore, UCNP-assisted excitation of microbial rhodopsins with NIR light has been intensively studied by electrophysiology technique. While electrophysiology is a powerful method to test the functional performance of microbial rhodopsins, conformational changes associated with the NIR light illumination in the presence of UCNPs remain poorly understood. Since UCNPs have generally multiple emission peaks at different wavelengths, it is important to reveal if UCNP-generated visible light induces similar structural changes of microbial rhodopsins as conventional visible light illumination does. Here, we synthesize the lanthanide-doped UCNPs that convert NIR light to blue light. Using these NIR-to-blue UCNPs, we monitor the NIR-triggered conformational changes in sensory rhodopsin II from Natronomonas pharaonis (NpSRII), blue light-sensitive microbial rhodospsin, by FTIR spectroscopy. FTIR difference spectrum of NpSRII was recorded under two different excitation conditions: (ⅰ) with conventional blue light, (ⅱ) with UCNP-generated blue light upon NIR excitation. Both spectra display similar spectral features characteristic of the long-lived M photointermediate state during the photocycle of NpSRII. This study demonstrates that NIR-activation of NpSRII mediated by UCNPs takes place in a similar way to direct blue light activation of NpSRII.

Molecular impact of the membrane potential on the regulatory mechanism of proton transfer in sensory rhodopsin II.

Metabolism establishes a potential difference across the cell membrane of every living cell which drives and regulates secondary ion and solute transfer across membrane proteins. Unraveling the effect of the membrane potential on the level of single molecular groups of the membrane protein was long hampered by the lack of appropriate analytical techniques. We have developed Surface Enhanced Infrared Difference Absorption Spectroscopy (SEIDAS), a highly sensitive vibrational technique for surface analysis, for the study of solid-supported monolayers of orientated membrane proteins. Here, we present spectroscopic data on vibrational changes of sensory rhodopsin II from Natronomonas pharaonis (NpSR II). The application of the electrode potential provides a voltage drop across the NpSR II monolayer through the Helmholtz double layer that mimics the cellular membrane potential. IR difference spectra indicated a shift of the photostationary equilibrium from an M and O mixture toward an M dominant equilibrium. The shift of positive to negative potential exhibited similar effects on the light-induced SEIDA spectra as the increase in pH. This effect is explained in terms of local pH change raised by the compensation of excess charge from the electrode. As we have shown earlier (Jiang, et al. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (34), 12113-12117), the application of an electric field opposite to the physiological proton transfer from the retinal Schiff base to its counterion Asp75 leads to the selective halt of the latter. However, when the solution pH is much higher than 5.8, that is, when the proton releasing group at the extracellular side is ionized, proton transfer of Asp75 becomes insensitive to the electric field exerted by the electrode. We infer that the deprotonation of the proton release group creates a local polar environment surrounding Asp75 as a consequence of hydrogen-bonding rearrangements that exceeds the energy of the external dipole. Our results reveal a molecular model for the physiological regulation of the photocycle of NpSR II by the potential drop across the membrane which came about by the interplay between the change in local pH at the membrane surface and the external electric field.

Thinner, smaller, faster: IR techniques to probe the functionality of biological and biomimetic systems.

New techniques in vibrational spectroscopy are promising for the study of biological samples as they provide exquisite spatial and/or temporal resolution with the benefit of minimal perturbation of the system during observation. In this Minireview we showcase the power of modern infrared techniques when applied to biological and biomimetic systems. Examples will be presented on how conformational changes in peptides can be traced with femtosecond resolution and nanometer sensitivity by 2D IR spectroscopy, and how surface-enhanced infrared difference absorption spectroscopy can be used to monitor the effect of the membrane potential on a single proton-transfer step in an integral membrane protein. Vibrational spectra of monolayers of molecules at basically any interface can be recorded with sum-frequency generation, which is strictly surface-sensitive. Chemical images are recorded by applying scanning near-field infrared microscopy at lateral resolutions better than 50 nm.

A catalytically active [Mn]-hydrogenase incorporating a non-native metal cofactor.

Nature carefully selects specific metal ions for incorporation into the enzymes that catalyse the chemical reactions necessary for life. Hydrogenases, enzymes that activate molecular H2, exclusively utilize Ni and Fe in [NiFe]-, [FeFe]- and [Fe]-hydrogeanses. However, other transition metals are known to activate or catalyse the production of hydrogen in synthetic systems. Here, we report the development of a biomimetic model complex of [Fe]-hydrogenase that incorporates a Mn, as opposed to a Fe, metal centre. This Mn complex is able to heterolytically cleave H2 as well as catalyse hydrogenation reactions. The incorporation of the model into an apoenzyme of [Fe]-hydrogenase results in a [Mn]-hydrogenase with an enhanced occupancy-normalized activity over an analogous semi-synthetic [Fe]-hydrogenase. These findings demonstrate a non-native metal hydrogenase that shows catalytic functionality and that hydrogenases based on a manganese active site are viable.

Disc Antenna Enhanced Infrared Spectroscopy: From Self-Assembled Monolayers to Membrane Proteins.

Plasmonic surfaces have emerged as a powerful platform for biomolecular sensing applications and can be designed to optimize the plasmonic resonance for probing molecular vibrations at utmost sensitivity. Here, we present a facile procedure to generate metallic microdisc antenna arrays that are employed in surface-enhanced infrared absorption (SEIRA) spectroscopy of biomolecules. Transmission electron microscopy (TEM) grids are used as shadow mask deployed during physical vapor deposition of gold. The resulting disc-shaped antennas exhibit enhancement factors of the vibrational bands of 4 × 104 giving rise to a detection limit <1 femtomol (10-15 mol) of molecules. Surface-bound monolayers of 4-mercaptobenzoic acid show polyelectrolyte behavior when titrated with cations in the aqueous medium. Conformational rigidity of the self-assembled monolayer is validated by density functional theory calculations. The membrane protein sensory rhodopsin II is tethered to the disc antenna arrays and is fully functional as inferred from the light-induced SEIRA difference spectra. As an advance to previous studies, the accessible frequency range is improved and extended into the fingerprint region.

Structure formation during translocon-unassisted co-translational membrane protein folding.

Correctly folded membrane proteins underlie a plethora of cellular processes, but little is known about how they fold. Knowledge of folding mechanisms centres on reversible folding of chemically denatured membrane proteins. However, this cannot replicate the unidirectional elongation of the protein chain during co-translational folding in the cell, where insertion is assisted by translocase apparatus. We show that a lipid membrane (devoid of translocase components) is sufficient for successful co-translational folding of two bacterial α-helical membrane proteins, DsbB and GlpG. Folding is spontaneous, thermodynamically driven, and the yield depends on lipid composition. Time-resolving structure formation during co-translational folding revealed different secondary and tertiary structure folding pathways for GlpG and DsbB that correlated with membrane interfacial and biological transmembrane amino acid hydrophobicity scales. Attempts to refold DsbB and GlpG from chemically denatured states into lipid membranes resulted in extensive aggregation. Co-translational insertion and folding is thus spontaneous and minimises aggregation whilst maximising correct folding.