Structure and function in rhodopsin. Measurement of the rate of metarhodopsin II decay by fluorescence spectroscopy

Single-cysteine substitution mutants at amino acid positions 306-321 in rhodopsin, the sequence between the cytoplasmic end of helix VII and the palmitoylation sites: sulfhydryl reactivity and transducin activation reveal a tertiary structure

As sensors for structure at the cytoplasmic face of rhodopsin, single-cysteine substitution mutants have been previously studied in the regions connecting helices III and IV and helices V and VI. In this paper we report on single-cysteine substitution mutants at amino acid positions 306-321, comprising the cytoplasmic sequence between helix VII and the palmitoylation sites in rhodopsin. The cysteine opsin mutants were expressed in COS-1 cells and on treatment with 11-cis-retinal all formed the characteristic rhodopsin chromophore. Cysteines at positions 306-316 and 319 reacted in the dark with the thiol-specific reagent 4, 4'-dithiodipyridine (4-PDS) but showed a wide variation in reactivity. Cysteines at positions 317, 318, 320, and 321 showed no reaction with 4-PDS. The mutants on illumination also showed wide variations in activating GT. The mutant Y306C showed almost no GT activation, I307C and N310C were poor, and the activity of the mutants M309C, F313C, and M317C was also reduced relative to WT. The results suggest that the region comprising amino acids 306-321 is a part of a tertiary structure and that specific amino acids in this region on light-activation participate in the interaction with GT.

Structure and function in rhodopsin: kinetic studies of retinal binding to purified opsin mutants in defined phospholipid-detergent mixtures serve as probes of the retinal binding pocket

In the current standard procedure for preparation of mammalian rhodopsin mutants, transfected COS-1 cells expressing the mutant opsin genes are treated with 5 microM 11-cis-retinal before detergent solubilization for purification. We found that binding of 11-cis-retinal to opsin mutants with single amino acid changes at Trp-265 (W265F,Y,A) and a retinitis pigmentosa mutant (A164V) was far from complete and required much higher concentrations of 11-cis-retinal. By isolation of the expressed opsins in a stable form, kinetic studies of retinal binding to the opsins in vitro have been carried out by using defined phospholipid-detergent mixtures. The results show wide variation in the rates of 11-cis-retinal binding. Thus, the in vitro reconstitution procedure serves as a probe of the retinal-binding pocket in the opsins. Further, a method is described for purification and characterization of the rhodopsin mutants after retinal binding to the opsins in vitro.

Structure and function in rhodopsin: further elucidation of the role of the intradiscal cysteines, Cys-110, -185, and -187, in rhodopsin folding and function

The disulfide bond between Cys-110 and Cys-187 in the intradiscal domain is required for correct folding in vivo and function of mammalian rhodopsin. Misfolding in rhodopsin, characterized by the loss of ability to bind 11-cis-retinal, has been shown to be caused by an intradiscal disulfide bond different from the above native disulfide bond. Further, naturally occurring single mutations of the intradiscal cysteines (C110F, C110Y, and C187Y) are associated with retinitis pigmentosa (RP). To elucidate further the role of every one of the three intradiscal cysteines, mutants containing single-cysteine replacements by alanine residues and the above three RP mutants have been studied. We find that C110A, C110F, and C110Y all form a disulfide bond between C185 and C187 and cause loss of retinal binding. C185A allows the formation of a C110-C187 disulfide bond, with wild-type-like rhodopsin phenotype. C187A forms a disulfide bond between C110 and C185 and binds retinal, and the pigment formed has markedly altered bleaching behavior. However, the opsin from the RP mutant C187Y forms no rhodopsin chromophore.

Conformational changes in rhodopsin. Movement of helix f detected by site-specific chemical labeling and fluorescence spectroscopy

A recent proposal for the formation of functionally active rhodopsin has placed critical importance on a movement of one of its transmembrane helices (Farrens, D. L., Altenbach, C., Yang, K., Hubbell, W. L., and Khorana, H. G. (1996) Science 274, 768-770). We investigated this hypothesis using a series of eight rhodopsin mutants containing single reactive cysteine residues in the region (helix F) where movement was previously detected. The cysteine mutants were studied in two ways, by measuring their reactivity to a cysteine-specific reagent (PyMPO-maleimide), and by labeling the cysteines with a fluorescent label (monobromobimane) followed by fluorescence spectroscopic analysis. The chemical reactivity data showed sequence-specific variations in reactivity for the mutants in the dark state, resulting in a pattern suggestive of an alpha helix. Interestingly, only upon photoactivation to the MII form did residues found on the inner "face" of this helix react with the PyMPO-maleimide. The ability of the dark state mutants to react with the fluorescent label monobromobimane followed a similar pattern. Furthermore, fluorescence measurements indicate that a bimane label on the inner face of the helix (at V250C) detects changes in the polarity of its environment and accessibility to a fluorescence quenching agent upon MII formation. Viewed together, the data provide further direct evidence that rhodopsin activation involves a conformational change at helix F.

Rhodopsin: a prototypical G protein-coupled receptor

A variety of spectroscopic and biochemical studies of recombinant site-directed mutants of rhodopsin and related visual pigments have been reported over the past 9 years. These studies have elucidated key structural elements common to visual pigments. In addition, systematic analysis of the chromophore-binding pocket in rhodopsin and cone pigments has led to an improved understanding of the mechanism of the opsin shift, and of particular molecular determinants underlying color vision in humans. Identification of the conformational changes that occur on rhodopsin photoactivation has been of particular recent concern. Assignments of light-dependent molecular alterations to specific regions of the chromophore have also been attempted by studying native opsins regenerated with synthetic retinal analogs. Site-directed mutagenesis of rhodopsin has also provided useful information about the retinal-binding pocket and the molecular mechanism of rhodopsin photoactivation. Individual molecular groups have been identified to undergo structural alterations or environmental changes during photoactivation. Analysis of particular mutant pigments in which specific groups are locked into their respective "off" or "on" states has provided a framework to identify determinants of the active conformation, as well as the minimal number of intramolecular transitions required to switch between inactive and active conformations. A simple model for the active state of rhodopsin can be compared to structural models of its ground state to localize chromophore-protein interactions that may be important in the photoactivation mechanism. This review focuses on the recent functional characterization of site-directed mutants of bovine rhodopsin and some cone pigments. In addition, an attempt is made to reconcile previous key findings and existing structural models with information gained from the analysis of site-directed mutant pigments.

Incorporation of rhodopsin in laterally structured supported membranes: observation of transducin activation with spatially and time-resolved surface plasmon resonance

Rhodopsin-transducin coupling was used as an assay to investigate a laterally patterned membrane reconstituted with a receptor and its G protein. It served as a model system to show the feasibility to immobilize G protein-coupled receptors on solid supports and investigate receptor activation and interaction with G proteins by one-dimensional imaging surface plasmon resonance. Supported membranes were formed by the self-assembly of lipids and rhodopsin from detergent solution onto functionalized gold surfaces. They formed micrometer-sized alternating regions of pure fluid phospholipid bilayers separated by bilayers composed of an outer phospholipid leaflet on a gold-attached inner thiolipid. Rhodopsin was found to incorporate preferentially into the phospholipid bilayer regions, whereas transducin was uniformly distributed over the entire outer surface of the supported patterned membrane. The influence of rhodopsin on the dark binding of transducin to lipid membranes was described quantitatively and compared with previously published data. Coupling reactions with transducin resembled closely the native system, indicating that the native functionality of rhodopsin was preserved in the supported membranes. The spatially varying properties of the membranes resulted in a pattern of rhodopsin activity on the surface. This combination of techniques is very promising for the investigation of the lateral diffusion of transducin, can be extended to include signalling proteins downstream of the G protein, and may be applied to functional screening of other G protein-coupled receptors. In the future, it may also serve as a basis for constructing biosensors based on receptor proteins.

Structure and function in rhodopsin: topology of the C-terminal polypeptide chain in relation to the cytoplasmic loops

Cysteine mutagenesis and site-directed spin labeling in the C-terminal region of rhodopsin have been used to probe the local structure and proximity of that region to the cytoplasmic loops. Each of the native amino acids in the sequence T335-T340 was replaced with Cys, one at a time. The sulfhydryl groups of all mutants reacted rapidly with the sulfhydryl reagent 4,4'-dithiodipyridine, which indicated a high degree of solvent accessibility. Furthermore, to probe the proximity relationships, a series of double Cys mutants was constructed. One Cys in all sets was at position 338 and the other was at a position in the sequence S240-V250 in the EF interhelical loop, at position 65 in the AB interhelical loop, or at position 140 in the CD interhelical loop. In the dark state, no significant disulfide formation was observed between C338 and C65 or C140 under the conditions used, whereas a relatively rapid disulfide formation was observed between C338 and C242 or C245. Spin labels in the double Cys mutants showed the strongest magnetic interactions between the nitroxides attached to C338 and C245 or C246. Light activation of the double mutant T242C/S338C resulted in slower disulfide formation, whereas interactions between nitroxides at C338 and C245 or C246 decreased. These results suggest the proximity of the C-terminal residue C338 to residues located on the outer face of a cytoplasmic helical extension of the F helix with an apparent increase of distance upon photoactivation.

Structure and function in rhodopsin: rhodopsin mutants with a neutral amino acid at E134 have a partially activated conformation in the dark state

The Glu-134-Arg-135 residues in rhodopsin, located near the cytoplasmic end of the C helix, are involved in G protein binding, or activation, or both. Furthermore, the charge-neutralizing mutation Glu-134 to Gln-134 produces hyperactivity in the activated state and produces constitutive activity in opsin. The Glu/Asp-Arg charge pair is highly conserved in equivalent positions in other G protein-coupled receptors. To investigate the structural consequences of charge-neutralizing mutations at Glu-134 and Arg-135 in rhodopsin, single spin-labeled side chains were introduced at sites in the cytoplasmic domains of helices C (140), E (227), F (250), or G (316) to serve as "molecular sensors" of the local helix bundle conformation. In each of the spin-labeled rhodopsins, a Gln substitution was introduced at either Glu-134 or Arg-135, and the electron paramagnetic resonance spectrum of the spin label was used to monitor the structural response of the helix bundle. The results indicate that a Gln substitution at Glu-134 induces a photoactivated conformation around helices C and G even in the dark state, an observation of potential relevance to the hyperactivity and constitutive activity of the mutant. In contrast, little change is induced in helix F, which has been shown to undergo a dominant motion upon photoactivation. This result implies that the multiple helix motions accompanying photoactivation are not strongly coupled and can be induced to take place independently. Gln substitution at Arg-135 produces only minor structural changes in the dark- or light-activated conformation, suggesting that this residue is not a determinant of structure in the regions investigated, although it may be functionally important.

Structure and function in rhodopsin: packing of the helices in the transmembrane domain and folding to a tertiary structure in the intradiscal domain are coupled

A previous study of the retinitis pigmentosa mutation L125R and two designed mutations at this site, L125A and L125F, showed that these mutations cause partial or total misfolding of the opsins expressed in COS cells from the corresponding mutant opsin genes. We now report on expression and characterization of the opsins from the following retinitis pigmentosa mutants in the transmembrane domain of rhodopsin that correspond to six of the seven helices: G51A and G51V (helix A), G89D (helix B), A164V (helix D), H211P (helix E), P267L and P267R (helix F), and T297R (helix G). All the mutations caused partial misfolding of the opsins as observed by the UV/visible absorption characteristics and by separation of the expressed opsins into fractions that bound 11-cis-retinal to form the corresponding mutant rhodopsins and those that did not bind 11-cis-retinal. Further, all the mutant rhodopsins prepared from the above mutants, except for G51A, showed strikingly abnormal bleaching behavior with abnormal metarhodopsin II photointermediates. The results show that retinitis pigmentosa mutations in every one of the transmembrane helices can cause misfolding of the opsin. Therefore, on the basis of these and previous results, we conclude that defects in the packing of the transmembrane helices resulting from these mutations are relayed to the intradiscal domain, where they cause misfolding of the opsin by inducing the formation of a disulfide bond other than the native Cys-110---Cys-187 disulfide bond. Thus, there is coupling between packing of the helices in the transmembrane domain and folding to a tertiary structure in the intradiscal domain.

Light-induced protein conformational changes in the photolysis of octopus rhodopsin

Light-induced protein conformational changes in the photolysis of octopus rhodopsin were measured with a highly sensitive time-resolved transient UV absorption spectrophotometer with nanosecond time resolution. A negative band around 280 nm in the lumirhodopsin minus rhodopsin spectra suggests that alteration of the environment of some of the tryptophan residues has taken place before the formation of lumirhodopsin. A small recovery of the absorbance at 280 nm was observed in the transformation of lumirhodopsin to mesorhodopsin. Kinetic parameters suggest that major conformational changes have taken place in the transformation of mesorhodopsin to acid metarhodopsin. In this transformation, drastic changes of amplitude and a shift of a difference absorption band around 280 nm take place, which suggest that some of the tryptophan residues of rhodopsin become exposed to a hydrophilic environment.

Structure and function in rhodopsin: peptide sequences in the cytoplasmic loops of rhodopsin are intimately involved in interaction with rhodopsin kinase

Phosphorylation of light-activated rhodopsin by the retina-specific enzyme, rhodopsin kinase (RK), is the primary event in the initiation of desensitization in the visual system. RK binds to the cytoplasmic face of rhodopsin, and the binding results in activation of the enzyme which then phosphorylates rhodopsin at several serine and threonine residues near the carboxyl terminus. To map the RK binding sites, we prepared two sets of rhodopsin mutants in the cytoplasmic CD and EF loops. In the first set, peptide sequences in both loops were either deleted or replaced by indifferent sequences. In the second set of mutants, the charged amino acids (E134, R135, R147, E239, K245, E247, K248, and E249) were replaced by neutral amino acids in groups of 1-3 per mutant. The deletion and replacement mutants in the CD loop showed essentially no phosphorylation, and they appeared to be defective in binding of RK. Of the mutants in the EF loop, that with a deletion of 13 amino acids, was also defective in binding to RK while the second mutant containing a replacement sequence bound RK but showed a reduction of about 70% in Vmax for phosphorylation. The mutants containing charged to neutral amino acid replacements in the CD and EF loops were all phosphorylated but to different levels. The charge reversal mutant E134R/R135E showed a 50% reduction in Vmax relative to wild-type rhodopsin. Replacements of charged residues in the EF loop decreased the Km by 5-fold for E239Q and E247Q/K248L/E239Q. In summary, both the CD and EF cytoplasmic loops are intimately involved in binding and interaction of RK with light-activated rhodopsin.

Structure and function in rhodopsin. Cysteines 65 and 316 are in proximity in a rhodopsin mutant as indicated by disulfide formation and interactions between attached spin labels

To probe proximity relationships between different amino acids in the interhelical loops in the cytoplasmic domain of rhodopsin, we are using a general approach in which two cysteine residues are introduced at different locations. Here we report on the characteristics of one such mutant that contains the naturally occurring cysteine 316 near the cytoplasmic end of helix G and a second cysteine at position 65 (H65C), near the cytoplasmic end of helix A. The mutant protein after expression in COS-1 cells and reconstitution with 11-cis-retinal can be bound to anti-rhodopsin antibody 1D4-Sepharose at pH 6 in a form that contains the two cysteines in the free sulfhydryl form. In this form, the mutant protein reacts as expected with N-ethylmaleimide in the dark at room temperature and can be derivatized with nitroxide spin labels. However, under appropriate conditions, the mutant can be isolated with the cysteines in the disulfide form, which has been characterized by analysis of fragments produced on proteolysis with thermolysin. A study of the interactions between nitroxide spin labels attached to the two cysteine residues in the mutant protein indicates that in the dark state they are within about 10 A of each other. On illumination the distance between the spin labels increases. Collectively, the above results show that, upon folding of the mutant opsin in vivo, cysteines 65 and 316, and by inference, helices A and G, are in proximal locations and move further apart upon photoactivation.

Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin

Conformational changes are thought to underlie the activation of heterotrimeric GTP-binding protein (G protein)-coupled receptors. Such changes in rhodopsin were explored by construction of double cysteine mutants, each containing one cysteine at the cytoplasmic end of helix C and one cysteine at various positions in the cytoplasmic end of helix F. Magnetic dipolar interactions between spin labels attached to these residues revealed their proximity, and changes in their interaction upon rhodopsin light activation suggested a rigid body movement of helices relative to one another. Disulfide cross-linking of the helices prevented activation of transducin, which suggests the importance of this movement for activation of rhodopsin.

Structure and function in rhodopsin: high level expression of a synthetic bovine opsin gene and its mutants in stable mammalian cell lines

Stable mammalian cell lines harboring a synthetic bovine opsin gene have been derived from the suspension-adapted HEK293 cell line. The opsin gene is under the control of the immediate-early cytomegalovirus promoter/enhancer in an expression vector that also contains a selectable marker (Neo) governed by a relatively weak promoter. The cell lines expressing the opsin gene at high levels are selected by growth in the presence of high concentrations of the antibiotic geneticin. Under the conditions used for cell growth in suspension, opsin is produced at saturated culture levels of more than 2 mg/liter. After reconstitution with 11-cis-retinal, rhodopsin is purified to homogeneity in a single step by immunoaffinity column chromatography. Rhodopsin thus prepared (> 90% recovery at concentrations of up to 15 microM) is indistinguishable from rhodopsin purified from bovine rod outer segments by the following criteria: (i) UV/Vis absorption spectra in the dark and after photobleaching and the rate of metarhodopsin II decay, (ii) initial rates of transducin activation, and (iii) the rate of phosphorylation by rhodopsin kinase. Although mammalian cell opsin migrates slower than rod outer segment opsin on SDS/polyacrylamide gels, presumably due to a different N-glycosylation pattern, their mobilities after deglycosylation are identical. This method has enabled the preparation of several site-specific mutants of bovine opsin in comparable amounts.

Structure and function in rhodopsin: expression of functional mammalian opsin in Saccharomyces cerevisiae

The yeast Saccharomyces cerevisiae has been investigated for expression of mammalian opsin as an alternative to the currently used expression in COS-1 mammalian cells. The synthetic opsin gene was placed under the control of the inducible promoter GAL1 in the multicopy yeast/ Escherichia coli shuttle vector YEpRF1. Transformation of a GAL+ S. cerevisiae strain with the vector and growth of galactose-induced cultures to saturation showed the production of 2.0 +/- 0.5 mg of opsin from about 10(10) cells by ELISA. The addition of 11-cis-retinal to either cell spheroplasts or lysed cells showed that a fraction (2-4%) of the total expressed opsin reconstituted to rhodopsin. This fraction was purified to homogeneity and was shown to be fully functional and indistinguishable from bovine rhodopsin by the following criteria: (i) UV-visible absorption spectra, (ii) the formation of metarhodopsin II and its rate of decay, and (iii) initial rate of transducin activation as measured by the formation of a complex between transducin (alpha subunit) and guanosine 5'-[gamma-[35S]thio]triphosphate. The purified fraction was homogeneously glycosylated. However, glycosylation was distinct from that of bovine rhodopsin as judged by mobility on SDS/PAGE and endoglycosidase H sensitivity.

Structural features and light-dependent changes in the cytoplasmic interhelical E-F loop region of rhodopsin: a site-directed spin-labeling study

Thirty consecutive single cysteine substitution mutants in the amino acids Q225-I256 of bovine rhodopsin have been prepared and modified with a sulfhydryl specific nitroxide reagent. This sequence includes the E-F interhelical loop, a transducin interaction site. The accessibilities of the attached nitroxides to collisions with hydrophilic and hydrophobic paramagnetic probes in solution were determined, and the electron paramagnetic resonance spectra analyzed in terms of side chain mobility, both in the dark and after photoactivation. Accessibility cata shows that the rhodopsin polypeptide chain crosses an aqueous/ hydrophobic boundary in the range V227-K231 and again in the range V250-V254. In the hydrophobic segments, both the accessibility and mobility data are consistent with helical structures. In the regions of the sequence located within the aqueous phase, periodic variation in both accessibility and mobility of the spin-labeled side chains indicates that the E-F interhelical loop is largely alpha-helical, being formed by regular extensions of the E and F helices by about 1.5 and 3 turns, respectively. Judging from nitroxide mobilities, the putative extension of helix E in the aqueous phase is more dynamic than that of helix F. Changes in the electron paramagnetic resonance characteristics of the spin-labeled rhodopsin upon photoactivation indicate that chromophore isomerization results in patterns of structural changes that can be interpreted in terms of movements of helices that extend into the aqueous loop regions.

Structure and function in rhodopsin. Single cysteine substitution mutants in the cytoplasmic interhelical E-F loop region show position-specific effects in transducin activation

The cytoplasmic interhelical E-F loop in rhodopsin is a part of the region that interacts with the G-protein transducin and rhodopsin kinase during signal transduction. In extending the previous work on systematic single cysteine substitutions of the amino acids in the cytoplasmic C-D loop, we have now replaced, one at a time, the amino acids Q225-I256 in the E-F loop region by cysteines. All the mutants formed the characteristic rhodopsin chromophore with 11-cis-retinal. While most of the mutants bleached normally, L226C, showed abnormal bleaching behavior. A study of the alkylation of the mutants by N-ethylmaleimide in dark showed low reactivity by some mutants, especially L226C. The rates of transducin activation (GT(alpha)-GTP gamma S complex formation) were measured for all the mutants. While these were normal for the bulk of the mutants, some (L226C, T229C, V230C, A233C, A234C, T242C, T243C, and Q244C) showed strikingly reduced transducin activation. The results suggest a specific structure in the E-F loop that interacts with transducin.

Specific tryptophan UV-absorbance changes are probes of the transition of rhodopsin to its active state

The difference of rhodopsin and metarhodopsin II (MII) absorption spectra exhibits a characteristic pattern in the UV wavelength range, consisting of peaks at 278, 286, 294, 302 nm. These difference bands are thought to result from the perturbation of the environments of tryptophan and/or tyrosine residues. We used site-directed mutagenesis to investigate the contribution of tryptophan absorption to these spectral features. Each of the five tryptophan residues in bovine rhodopsin was replaced by either a phenylalanine or a tyrosine. The mutant pigments (W35F, W126F, W161F, W175F, W265F/Y) were prepared and studied by UV-visible photobleaching difference spectroscopy. The difference spectra of the W35F and W175F mutants were identical to that of rhodopsin, whereas in the W161F mutant, the magnitudes of the 294- and 302-nm bands were slightly lowered. The differential absorbance at 294 nm was reduced by over 50% in the W126F and W265F/Y mutants. The difference peak at 302 nm was reduced in the W265F/Y mutants, but was almost completely absent in the W126F mutant. These data indicate that the difference bands at 294 and 302 nm originate from the perturbations of Trp126 and Trp265 environments resulting from a general conformational change concomitant with MII formation and receptor activation. Model studies on tryptophan absorption indicate that the difference peak at 294 nm is due to the differential shift of the Lb absorption of the indole side chain as a result of decreased hydrophobicity or polarizability of the Trp126 and Trp265 environments. The resolution of the 302-nm band, assigned to the differential shift of the indole La absorption, is consistent with hydrogen-bonding interactions of the indole N-H groups of Trp126 and Trp265 becoming weaker in MII. These results suggest that the photoactivation of rhodopsin involves a change in the relative disposition of transmembrane helices 3 and 6, which contain Trp126 and Trp265 respectively, within the alpha-helical bundle of the receptor.

Structure and function in rhodopsin: correct folding and misfolding in point mutants at and in proximity to the site of the retinitis pigmentosa mutation Leu-125-->Arg in the transmembrane helix C

L125R is a mutation in the transmembrane helix C of rhodopsin that is associated with autosomal dominant retinitis pigmentosa. To probe the orientation of the helix and its packing in the transmembrane domain, we have prepared and studied the mutations E122R, I123R, A124R, S127R, L125F, and L125A at, and in proximity to, the above mutation site. Like L125R, the opsin expressed in COS-1 cells from E122R did not bind 11-cis-retinal, whereas those from I123R and S127R formed the rhodopsin chromophore partially. A124R opsin formed the rhodopsin chromophore (lambda max 495 nm) in the dark, but the metarhodopsin II formed on illumination decayed about 6.5 times faster than that of the wild type and was defective in transducin activation. The mutant opsins from L125F and L125A bound 11-cis-retinal only partially, and in both cases, the mixtures of the proteins produced were separated into retinal-binding and non-retinal-binding (misfolded) fractions. The purified mutant rhodopsin from L125F showed lambda max at 500 nm, whereas that from L125A showed lambda max at 503 nm. The mutant rhodopsin L125F showed abnormal bleaching behavior and both mutants on illumination showed destabilized metarhodopsin II species and reduced transducin activation. Because previous results have indicated that misfolding in rhodopsin is due to the formation of a disulfide bond other than the normal disulfide bond between Cys-110 and Cys-187 in the intradiscal domain, we conclude from the misfolding in mutants L125F and L125A that the folding in vivo in the transmembrane domain is coupled to that in the intradiscal domain.