Structural characterization of the pore forming protein TatAd of the twin-arginine translocase in membranes by solid-state N-NMR

The transmembrane protein TatA is the pore forming unit of the twin-arginine translocase (Tat), which has the unique ability of transporting folded proteins across the cell membrane. This ATP-independent protein export pathway is a recently discovered alternative to the general secretory (Sec) system of bacteria. To obtain insight in the translocation mechanism, the structure and alignment in the membrane of the well-folded segments 2-45 of TatA_d from Bacillus subtilis was studied here. Using solid-state NMR in bicelles containing anionic lipids, the topology and orientation of TatA_d was determined in an environment mimicking the bacterial membrane. A wheel-like pattern, characteristic for a tilted transmembrane helix, was observed in ¹⁵N chemical shift /¹⁵N-¹H dipolar coupling correlation NMR spectra. Analysis of this PISA wheel revealed a 14-16 residue long N-terminal membrane-spanning helix which is tilted by 17° with respect to the membrane normal. In addition, comparison of uniformly and selectively ¹⁵N-labeled TatA_2-45 samples allowed determination of the helix polarity angle.     

Cellular uptake of S413-PV peptide occurs upon conformational changes induced by peptide-membrane interactions

In face of accumulated reports demonstrating that uptake of some cell-penetrating peptides occurs through previously described endocytic pathways, or is a consequence of cell fixation artifacts, we conducted a systematic analysis on the mechanism responsible for the cellular uptake of the S4₁₃-PV karyophilic cell-penetrating peptide. The results reviewed here show that the S4₁₃-PV peptide is able to very efficiently accumulate inside live cells in a rapid, non-toxic and dose-dependent manner, through a mechanism distinct from endocytosis. Comparative analysis of peptide uptake by mutant cells lacking heparan sulfate proteoglycans demonstrates that, although not mandatory, their presence at cell surface facilitates the cellular uptake of the S4₁₃-PV peptide. Furthermore, we demonstrate that upon interaction with lipid vesicles, the S4₁₃-PV peptide undergoes significant conformational changes that are consistent with the formation of helical structures. Such conformational changes occur concomitantly with a penetration of the peptide into the lipid bilayer, strongly suggesting that the resulting helical structures are crucial for the non-endocytic cellular uptake of the S4₁₃-PV peptide. Overall, our data support that, rather than endocytosis, the cellular uptake of the S4₁₃-PV cell-penetrating peptide is a consequence of its direct translocation through cell membranes following conformational changes induced by peptide-membrane interactions.     

Biophysical studies of the membrane location of the voltage-gated sensors in the HsapBK and KvAP K channels

The membrane location of two fragments in two different K⁺-channels, the KvAP (from Aeropyrum pernix) and the HsapBK (human) corresponding to the putative “paddle” domains, has been investigated by CD, fluorescence and NMR spectroscopy. Both domains interact with q = 0.5 phospholipid bicelles, DHPC micelles and with POPC vesicles. CD spectra demonstrate that both peptides become largely helical in the presence of phospholipid bicelles. Fluorescence quenching studies using soluble acrylamide or lipid-attached doxyl-groups show that the arginine-rich domains are located within the bilayered region in phospholipid bicelles. Nuclear magnetic relaxation parameters, T₁ and ¹³C-¹H NOE, for DMPC in DMPC/DHPC bicelles and for DHPC in micelles showed that the lipid acyl chains in the bicelles become less flexible in the presence of either of the fragments. An even more pronounced effect is seen on the glycerol carbons. ²H NMR spectra of magnetically aligned bicelles showed that the peptide derived from KvAP had no or little effect on bilayer order, while the peptide derived from HsapBK had the effect of lowering the order of the bilayer. The present study demonstrates that the fragments derived from the full-length proteins interact with the bilayered interior of model membranes, and that they affect both the local mobility and lipid order of model membrane systems.     

The TatAd component of the Bacillus subtilis twin-arginine protein transport system forms homo-multimeric complexes in its cytosolic and membrane embedded localisation

The twin arginine translocation (Tat) system has the capacity to transfer completely folded proteins across the bacterial cytoplasmic membrane and the thylakoid membrane of plant chloroplasts. The most abundant TatA protein of this system has been suggested to form the protein conducting channel. Here, the molecular organisation of soluble and membrane embedded Bacillus subtilis TatA_d was analysed using negative contrast and freeze-fractured electron microscopy. In both compartments, the protein showed homo-oligomerisation. In aqueous solution, TatA_d formed homo-multimeric micelle-like complexes. Freeze-fracture analysis of proteoliposomes revealed self association of membrane-integrated TatA_d independent from TatC_d, the second component of this transport system. Immunogold labelling demonstrated that the substrate prePhoD was co-localised with membrane-integrated TatA_d complexes.     

In vitro Unfolding and Refolding of the Small Multidrug Transporter EmrE

The composition of the lipid bilayer is increasingly being recognised as important for the regulation of integral membrane protein folding and function, both in vivo and in vitro. The folding of only a few membrane proteins, however, has been characterised in different lipid environments. We have refolded the small multidrug transporter EmrE in vitro from a denatured state to a functional protein and monitored the influence of lipids on the folding process. EmrE is part of a multidrug resistance protein family that is highly conserved amongst bacteria and is responsible for bacterial resistance to toxic substances. We find that the secondary structure of EmrE is very stable and only small amounts are denatured even in the presence of unusually high denaturant concentrations involving a combination of 10 M urea and 5% SDS. Substrate binding by EmrE is recovered after refolding this denatured protein into dodecylmaltoside detergent micelles or into lipid vesicles. The yield of refolded EmrE decreases with lipid bilayer compositional changes that increase the lateral chain pressure within the bilayer, whilst conversely, the apparent rate of folding seems to increase. These results add further weight to the hypothesis that an increased lateral chain pressure hinders protein insertion across the bilayer. Once the protein is inserted, however, the greater pressure on the transmembrane helices accelerates correct packing and final folding. This work augments the relatively small number of biophysical folding studies in vitro on helical membrane proteins.     

Structural rearrangement of model membranes by the peptide antibiotic NK-2

We have developed a novel α-helical peptide antibiotic termed NK-2. It efficiently kills bacteria, but not human cells, by membrane destruction. This selectivity could be attributed to the different membrane lipid compositions of the target cells. To understand the mechanisms of selectivity and membrane destruction, we investigated the influence of NK-2 on the supramolecular aggregate structure, the phase transition behavior, the acyl chain fluidity, and the surface charges of phospholipids representative for the bacterial and the human cell cytoplasmic membranes. The cationic NK-2 binds to anionic phosphatidylglycerol liposomes, causing a thinning of the membrane and an increase in the phase transition temperature. However, this interaction is not solely of electrostatic but also of hydrophobic nature, indicated by an overcompensation of the Zeta potential. Whereas NK-2 has no effect on phosphatidylcholine liposomes, it enhances the fluidity of phosphatidylethanolamine acyl chains and lowers the phase transition enthalpy of the gel to liquid cristalline transition. The most dramatic effect, however, was observed for the lamellar/inverted hexagonal transition of phosphatidylethanolamine which was reduced by more than 10 °C. Thus, NK-2 promotes a negative membrane curvature which can lead to the collapse of the phosphatidylethanolamine-rich bacterial cytoplasmic membrane.     

The membrane-induced structure of melittin is correlated with the fluidity of the lipids

The effect of the bee toxin melittin on DMPC dynamics in fast-tumbling bicelles has been investigated. The ¹³C R₁ and ¹³C-¹H NOE relaxation parameters for DMPC were used to monitor the effect of melittin and cholesterol on lipid dynamics. It was found that melittin has the largest effect on the DMPC mobility in DMPC/DHPC bicelles, while less effect was observed in cholesterol-doped bicelles, or in bicelles made with CHAPS, indicating that the rigidity of the membrane affects the melittin-membrane interaction. CD spectra were analysed in terms of cooperativity of the α-helix to random coil transition in melittin, and these results also indicated similar differences between the bicelles. The study shows that bicelles can be used to investigate lipid dynamics by spin relaxation, and in particular of peptide-induced changes in membrane fluidity.     

Solid-state NMR approaches to measure topological equilibria and dynamics of membrane polypeptides

Biological membranes are characterized by a high degree of dynamics. In order to understand the function of membrane proteins and even more of membrane-associated peptides, these motional aspects have to be taken into consideration. Solid-state NMR spectroscopy is a method of choice when characterizing topological equilibria, molecular motions, lateral and rotational diffusion as well as dynamic oligomerization equilibria within fluid phase lipid bilayers. Here we show and review examples where the ¹⁵N chemical shift anisotropy, dipolar interactions and the deuterium quadrupolar splittings have been used to analyze motions of peptides such as peptaibols, antimicrobial sequences, Vpu, phospholamban or other channel domains. In particular, simulations of ¹⁵N and ²H-solid-state NMR spectra are shown of helical domains in uniaxially oriented membranes when rotation around the membrane normal or the helix long axis occurs.     

Effect of lipid composition on the topography of membrane-associated hydrophobic helices: stabilization of transmembrane topography by anionic lipids

To investigate the effect of lipid structure upon the membrane topography of hydrophobic helices, the behavior of hydrophobic peptides was studied in model membrane vesicles. To define topography, fluorescence and fluorescence quenching methods were used to determine the location of a Trp at the center of the hydrophobic sequence. For peptides with cationic residues flanking the hydrophobic sequence, the stability of the transmembrane (TM) configuration (relative to a membrane-bound non-TM state) increased as a function of lipid composition on the order: 1:1 (mol:mol) 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC):1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine ∼ 6:4 POPC:cholesterol < POPC ∼ dioleoylphosphatidylcholine (DOPC) < 1,2-dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] sodium salt (DOPG) ≤ 1,2-dioleoyl-sn-glycero-3-[phospho-l-serine] sodium salt (DOPS), indicating that the anionic lipids DOPG and DOPS most strongly stabilized the TM configuration. TM stabilization was near maximal at 20-30 mol% anionic lipid, which are physiologically relevant values. TM stabilization by anionic lipid was observed for hydrophobic sequences with a diverse set of sequences (including polyAla), diverse lengths (from 12 to 22 residues), and various cationic flanking residues (H, R, or K), but not when the flanking residues were uncharged. TM stabilization by anionic lipid was also dependent on the number of cationic residues flanking the hydrophobic sequence, but was still significant with only one cationic residue flanking each end of the peptide. These observations are consistent with TM-stabilizing effects being electrostatic in origin. However, Trp located more deeply in DOPS vesicles relative to DOPG vesicles, and peptides in DOPS vesicles showed increased helix formation relative to DOPG and all other lipid compositions. These observations fit a model in which DOPS anchors flanking residues near the membrane surface more strongly than does DOPG and/or increases the stability of the TM state to a greater degree than DOPG. We conclude that anionic lipids can have significant and headgroup structure-specific effects upon membrane protein topography.     

Membrane Protein Frustration: Protein Incorporation into Hydrophobic Mismatched Binary Lipid Mixtures

Bacteriophage M13 major coat protein was reconstituted in different nonmatching binary lipid mixtures composed of 14:1PC and 22:1PC lipid bilayers. Challenged by this lose-lose situation of hydrophobic mismatch, the protein-lipid interactions are monitored by CD and site-directed spin-label electron spin resonance spectroscopy of spin-labeled site-specific single cysteine mutants located in the C-terminal protein domain embedded in the hydrophobic core of the membrane (I39C) and at the lipid-water interface (T46C). The CD spectra indicate an overall α-helical conformation irrespective of the composition of the binary lipid mixture. Spin-labeled protein mutant I39C senses the phase transition in 22:1PC, in contrast to spin-labeled protein mutant T46C, which is not affected by the transition. The results of both CD and electron spin resonance spectroscopy clearly indicate that the protein preferentially partitions into the shorter 14:1PC both above and below the gel-to-liquid crystalline phase transition temperature of 22:1PC. This preference is related to the protein tilt angle and energy penalty the protein has to pay in the thicker 22:1PC. Given the fact that in Escherichia coli, which is the host for M13 bacteriophage, it is easier to find shorter 14 carbon acyl chains than longer 22 carbon acyl chains, the choice the M13 coat protein makes seems to be evolutionary justified.     

Influence of the lipid composition on the kinetics of concerted insertion and folding of melittin in bilayers

We have examined the kinetics of the adsorption of melittin, a secondary amphipathic peptide extracted from bee venom, on lipid membranes using three independent and complementary approaches. We probed (i) the change in the polarity of the ¹⁹Trp of the peptide upon binding, (ii) the insertion of this residue in the apolar core of the membrane, measuring the ¹⁹Trp-fluorescence quenching by bromine atoms attached on lipid acyl chains, and (iii) the folding of the peptide, by circular dichroism (CD). We report a tight coupling of the insertion of the peptide with its folding as an α-helix. For all the investigated membrane systems (cholesterol-containing, phosphoglycerol-containing, and pure phosphocholine bilayers), the decrease in the polarity of ¹⁹Trp was found to be significantly faster than the increase in the helical content of melittin. Therefore, from a kinetics point of view, the formation of the α-helix is a consequence of the insertion of melittin. The rate of melittin folding was found to be influenced by the lipid composition of the bilayer and we propose that this was achieved by the modulation of the kinetics of insertion. The study reports a clear example of the coupling existing between protein penetration and folding, an interconnection that must be considered in the general scheme of membrane protein folding.     

Comparison of the membrane interaction mechanism of two antimicrobial RNases: RNase 3/ECP and RNase 7

Eosinophil cationic protein (ECP/RNase 3) and the skin derived ribonuclease 7 (RNase 7) are members of the RNase A superfamily. RNase 3 is mainly expressed in eosinophils whereas RNase 7 is primarily secreted by keratinocytes. Both proteins present a broad-spectrum antimicrobial activity and their bactericidal mechanism is dependent on their membrane destabilizing capacities. Using phospholipid vesicles as membrane models, we have characterized the protein membrane association process. Confocal microscopy experiments using giant unilamellar vesicles illustrate the morphological changes of the liposome population. By labelling both lipid bilayers and proteins we have monitored the kinetic of the process. The differential protein ability to release the liposome aqueous content was evaluated together with the micellation and aggregation processes. A distinct morphology of the protein/lipid aggregates was visualized by transmission electron microscopy and the proteins overall secondary structure in a lipid microenvironment was assessed by FTIR. Interestingly, for both RNases the membrane interaction events take place in a different behaviour and timing: RNase 3 triggers first the vesicle aggregation, while RNase 7 induces leakage well before the aggregation step. Their distinct mechanism of action at the membrane level may reflect different in vivo antipathogen functions.     

Drug screening strategy for human membrane proteins: From NMR protein backbone structure to in silica- and NMR-screened hits

About 8000 genes encode membrane proteins in the human genome. The information about their druggability will be very useful to facilitate drug discovery and development. The main problem, however, consists of limited structural and functional information about these proteins because they are difficult to produce biochemically and to study. In this paper we describe the strategy that combines Cell-free protein expression, NMR spectroscopy, and molecular DYnamics simulation (CNDY) techniques. Results of a pilot CNDY experiment provide us with a guiding light towards expedited identification of the hit compounds against a new uncharacterized membrane protein as a potentially druggable target. These hits can then be further characterized and optimized to develop the initial lead compound quicker. We illustrate such “omics” approach for drug discovery with the CNDY strategy applied to two example proteins: hypoxia-induced genes HIGD1A and HIGD1B.     

Shiga toxin B-subunit sequential binding to its natural receptor in lipid membranes

Shiga toxin B-subunit (STxB), a protein involved in the cell-binding and intracellular trafficking of Shiga holotoxin, binds to a specific glycolipid, the globotriaosyl ceramide (Gb₃). Tryptophan residues of STxB, located at the protein-membrane interface, allow one to study its interaction with model membranes by means of spectroscopic methods with no need for chemical derivatisation with a fluorophore. The protein emits maximally around 346 nm and a blue shift of about 8 nm, as well as the occurrence of changes in the emission fluorescence intensity spectra, is indicative of insertion and partition into the membrane. However, the interaction seems to take place without pentamer dissociation. Acrylamide quenching experiments confirm tryptophan residues become less exposed to solvent when in the presence of vesicles, and the use of lipophilic probes suggests that they are located in a shallow position near the water/membrane interface. Fluorescence intensity and lifetime measurements upon STxB titration with Gb₃-containing vesicles suggest a complex STxB/Gb₃ docking mechanism involving static quenching in the later stages. Based on our observations, a model of the protein-membrane interaction is proposed and the STxB membrane partition and binding constants were calculated.     

HIV Fusion Peptide Penetrates, Disorders, and Softens T-Cell Membrane Mimics

This work investigates the interaction of N-terminal gp41 fusion peptide (FP) of human immunodeficiency virus type 1 (HIV-1) with model membranes in order to elucidate how FP leads to fusion of HIV and T-cell membranes. FP constructs were (i) wild-type FP23 (23 N-terminal amino acids of gp41), (ii) water-soluble monomeric FP that adds six lysines on the C-terminus of FP23 (FPwsm), and (iii) the C-terminus covalently linked trimeric version (FPtri) of FPwsm. Model membranes were (i) LM3 (a T-cell mimic), (ii) 1,2-dioleoyl-sn-glycero-3-phosphocholine, (iii) 1,2-dioleoyl-sn-glycero-3-phosphocholine/30 mol% cholesterol, (iv) 1,2-dierucoyl-sn-glycero-3-phosphocholine, and (v) 1,2-dierucoyl-sn-glycero-3-phosphocholine/30 mol% cholesterol. Diffuse synchrotron low-angle x-ray scattering from fully hydrated samples, supplemented by volumetric data, showed that FP23 and FPtri penetrate into the hydrocarbon region and cause membranes to thin. Depth of penetration appears to depend upon a complex combination of factors including bilayer thickness, presence of cholesterol, and electrostatics. X-ray data showed an increase in curvature in hexagonal phase 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine, which further indicates that FP23 penetrates into the hydrocarbon region rather than residing in the interfacial headgroup region. Low-angle x-ray scattering data also yielded the bending modulus K_C, a measure of membrane stiffness, and wide-angle x-ray scattering yielded the S_xray orientational order parameter. Both FP23 and FPtri decreased K_C and S_xray considerably, while the weak effect of FPwsm suggests that it did not partition strongly into LM3 model membranes. Our results are consistent with the HIV FP disordering and softening the T-cell membrane, thereby lowering the activation energy for viral membrane fusion.     

Quantifying Interactions of β-Synuclein and γ-Synuclein with Model Membranes

The synucleins are a family of proteins involved in numerous neurodegenerative pathologies [α-synuclein and β-synuclein (βS)], as well as in various types of cancers [γ-synuclein (γS)]. While the connection between α-synuclein and Parkinson's disease is well established, recent evidence links point mutants of βS to dementia with Lewy bodies. Overexpression of γS has been associated with enhanced metastasis and cancer drug resistance. Despite their prevalence in such a variety of diseases, the native functions of the synucleins remain unclear. They have a lipid-binding motif in their N-terminal region, which suggests interactions with biological membranes in vivo. In this study, we used fluorescence correlation spectroscopy to monitor the binding properties of βS and γS to model membranes and to determine the free energy of the interactions. Our results show that the interactions are most strongly affected by the presence of both anionic lipids and bilayer curvature, while membrane fluidity plays a very minor role. Quantifying the lipid-binding properties of βS and γS provides additional insights into the underlying factors governing the protein-membrane interactions. Such insights not only are relevant to the native functions of these proteins but also highlight their contributions to pathological conditions that are either mediated or characterized by perturbations of these interactions.     

Differential membrane perturbation caused by the cell penetrating peptide Tp10 depending on attached cargo

The membrane leakage caused by the cell penetrating peptide Tp10, a variant of transportan, was studied in large unilamellar vesicles with the entrapped fluorophore calcein. The vesicles were composed of zwitterionic 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine. A significant decrease in membrane leakage was found when the 55kDa streptavidin protein was attached to Tp10. When a 5.4kDa peptide nucleic acid molecule was attached, the membrane leakage was comparable to that caused by Tp10 alone. The results suggest that direct membrane effects may cause membrane translocation of Tp10 alone and of smaller complexes, whereas these effects do not contribute for larger cargoes.     

Cell-free expression and assembly of ATP synthase

Cell-free (CF) expression technologies have emerged as promising methods for the production of individual membrane proteins of different types and origin. However, many membrane proteins need to be integrated in complex assemblies by interaction with soluble and membrane-integrated subunits in order to adopt stable and functionally folded structures. The production of complete molecular machines by CF expression as advancement of the production of only individual subunits would open a variety of new possibilities to study their assembly mechanisms, function, or composition. We demonstrate the successful CF formation of large molecular complexes consisting of both membrane-integrated and soluble subunits by expression of the atp operon from Caldalkalibacillus thermarum strain TA2.A1 using Escherichia coli extracts. The operon comprises nine open reading frames, and the 542-kDa F₁F_o-ATP synthase complex is composed of 9 soluble and 16 membrane-embedded proteins in the stoichiometry α₃β₃γδ?ab₂c₁₃. Complete assembly into the functional complex was accomplished in all three typically used CF expression modes by (i) solubilizing initial precipitates, (ii) cotranslational insertion into detergent micelles or (iii) cotranslational insertion into preformed liposomes. The presence of all eight subunits, as well as specific enzyme activity and inhibition of the complex, was confirmed by biochemical analyses, freeze-fracture electron microscopy, and immunogold labeling. Further, single-particle analysis demonstrates that the structure and subunit organization of the CF and the reference in vivo expressed ATP synthase complexes are identical. This work establishes the production of highly complex molecular machines in defined environments either as proteomicelles or as proteoliposomes as a new application of CF expression systems.     

The N-terminal fragment of human islet amyloid polypeptide is non-fibrillogenic in the presence of membranes and does not cause leakage of bilayers of physiologically relevant lipid composition

Human islet amyloid polypeptide (hIAPP) forms amyloid fibrils in pancreatic islets of patients with type 2 diabetes mellitus (DM2). The formation of hIAPP fibrils has been shown to cause membrane damage which most likely is responsible for the death of pancreatic islet β-cells during the pathogenesis of DM2. Previous studies have shown that the N-terminal part of hIAPP, hIAPP_1-19, plays a major role in the initial interaction of hIAPP with lipid membranes. However, the exact role of this N-terminal part of hIAPP in causing membrane damage is unknown. Here we investigate the structure and aggregation properties of hIAPP_1-19 in relation to membrane damage in vitro by using membranes of the zwitterionic lipid phosphatidylcholine (PC), the anionic lipid phosphatidylserine (PS) and mixtures of these lipids to mimic membranes of islet cells. Our data reveal that hIAPP_1-19 is weakly fibrillogenic in solution and not fibrillogenic in the presence of membranes, where it adopts a secondary structure that is dependent on lipid composition and stable in time. Furthermore, hIAPP_1-19 is not able to induce leakage in membranes of PC/PS or PC bilayers, indicating that the membrane interaction of the N-terminal fragment by itself is not responsible for membrane leakage under physiologically relevant conditions. In bilayers of the anionic lipid PS, the peptide does induce membrane damage, but this leakage is not correlated to fibril formation, as it is for mature hIAPP. Hence, membrane permeabilization by the N-terminal fragment of hIAPP in anionic lipids is most likely an aspecific process, occurring via a mechanism that is not relevant for hIAPP-induced membrane damage in vivo.     

The effect of membranes on the in vitro fibrillation of an amyloidogenic light-chain variable-domain SMA

Light chain (or AL) amyloidosis is the most common form of systemic amyloidosis, characterized by the pathological deposition of insoluble fibrils of immunoglobulin light-chain fragments in various organs and tissues, especially in the kidney and heart. Both the triggering factors and the mechanisms involved in the abnormal formation of the insoluble fibrillar aggregates from the soluble proteins are poorly understood. For example, although the fibrillar deposits are typically found associated with the extracellular matrix and basement membranes, it is not clear whether fibrils are initially formed intra- or extracellularly, nor it is understood what determines where the deposits will occur; i.e., site tropism. In the present investigation, we studied the interaction of a recombinant amyloidogenic light-chain variable domain, SMA, with lipid vesicles. The nature of the interaction was dependent on the lipid composition and the SMA to lipid ratio. The most pronounced effect was found from vesicles composed of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine/1-palmitoyl-2-oleoyl-sn-glycero-3-phosphate, which dramatically accelerated fibril growth. Interestingly, spectral probes, such as intrinsic fluorescence and far-UV CD spectroscopy did not show significant conformational changes in the presence of the vesicles. The presence of cholesterol or divalent cations, such as Ca²⁺ and Mg²⁺, lead to decreased membrane-induced SMA fibrillation. Thus, membranes may have significant effects on light-chain fibrillation and may contribute to the site selectivity observed in AL amyloidosis.