Boundary lipids of the nicotinic acetylcholine receptor: spontaneous partitioning via coarse-grained molecular dynamics simulation
BBoundary lipids of the nicotinic acetylcholine receptor: spontaneous partitioningvia coarse-grained molecular dynamics simulation
Liam Sharp a , Reza Salari a,b , Grace Brannigan a,c a Center for Computational and Integrative Biology, Rutgers University-Camden, Camden, NJ b Now at Washington University School of Medicine in St Louis c Department of Physics, Rutgers University-Camden, Camden, NJ
Abstract
Reconstituted nicotinic acetylcholine receptors (nAChRs) exhibit significant gain-of-function upon addition of choles-terol to reconstitution mixtures, and cholesterol a ff ects organization of nAChRs within domain-forming membranes,but whether nAChR partitions to cholesterol-rich liquid-ordered (“raft” or l o ) domains or cholesterol-poor liquid-disordered ( l do ) domains is unknown. We use coarse-grained molecular dynamics simulations to observe spontaneousinteractions of cholesterol, saturated lipids, and polyunsaturated (PUFA) lipids with nAChRs. In binary Dipalmi-toylphosphatidylcholine:Cholesterol (DPPC:CHOL) mixtures, both CHOL and DPPC acyl chains were observedspontaneously entering deep “non-annular” cavities in the nAChR TMD, particularly at the subunit interface andthe β subunit center, facilitated by the low amino acid density in the cryo-EM structure of nAChR in a native mem-brane. Cholesterol was highly enriched in the annulus around the TMD, but this e ff ect extended over (at most) 5-10Å.In domain-forming ternary mixtures containing PUFAs, the presence of a single receptor did not significantly a ff ectthe likelihood of domain formation. nAChR partitioned to any cholesterol-poor l do domain that was present, regard-less of whether the l do or l o domain lipids had PC or PE headgroups. Enrichment of PUFAs among boundary lipidswas positively correlated with their propensity for demixing from cholesterol-rich phases. Long n − n − ff ective at displacing cholesterol from non-annular sites. Keywords:
Polyunsaturated Fatty Acids (PUFA), Docosahexaenoic acid (DHA), cholesterol, nicotinic acetylcholinereceptor, nAChR partitioning, liquid order ( l o ), liquid disorder ( l do ), lipid-protein interactions, lipid rafts
1. Introduction
The nicotinic acetylcholine receptor (nAChR) is anexcitatory pentameric ligand gated ion channel (pLGIC)commonly found in the neuronal post synaptic mem-brane and neuromuscular junction (NMJ) in mammalsas well as the electric organs of the
Torpedo electricray. nAChRs play a fundamental role in rapid exci-tation within the central and peripheral nervous sys-tem, and neuronal nAChRs are also critical for cogni-tion and memory (1,2) . Acetylcholine is the orthostericnAChR ligand, but numerous other exogenous and en-dogenous small molecules modulate nAChRs, includ-ing nicotine, general anesthetics, the tipped-arrow poi-son curare, phospholipids, cholesterol, and cholesterol-derived hormones. (3,4)
The larger pLGIC super familythat includes nAChRs has been shown to play roles in numerous diseases related to inflammation, (5–7) , ad-diction (7) , chronic pain (8) , Alzheimer’s Disease (9–11) ,spinal muscular atrophy (12) , schizophrenia (13) and neu-rological autoimmune diseases (14) .nAChRs are highly sensitive to the surrounding lipidenvironment (15–18) for reasons that remain poorly un-derstood. In the late 1970s it was observed that re-constituted nAChRs only exhibit native conductanceif model phospholipid membranes contained at least10-20% cholesterol (19–21) . Three generations of inves-tigation into the mechanism have followed, with thefirst generation of studies (19–42) aiming to di ff erenti-ate between the role of bulk, annular, and non-annularcholesterol. The second generation (16,18,42–54) of studiesprobed membrane-mediated e ff ects on organization ofmultiple nAChRs, while the third generation (55–58) hasapplied x-ray crystallography and high-resolution cryo Preprint submitted to BBA December 4, 2018 a r X i v : . [ q - b i o . S C ] D ec lectron microscopy to directly observe lipid bindingmodes.Members of the pLGIC family other than nAChR arealso lipid-sensitive, (59–62) and lipids other than choles-terol can also modulate function (15,52,63–67) , but thesemechanisms have not been as extensively studied. Therecent publication of several crystal and cryo-EM struc-tures (55–58) has confirmed that specific lipid-pLGIC in-teractions extend beyond cholesterol and nAChR. Suchinteractions are also well-established in other trans-membrane proteins, including G-protein coupled re-ceptors (GPCRs) and other ion channels, as reviewedin (68–72) .Even in the specific case of cholesterol-nAChR in-teractions, results from di ff erent approaches have sug-gested complex behavior and even contradictory inter-pretations. Results have indicated that both cholesterolenrichment (19–21) and cholesterol depletion (73) causegain of function, that anionic phospholipids are un-necessary for native function (19–21) or must be (65,66) in-cluded in a reconstitution mixture, that cholesterol in-creases nAChR clustering (18,47,48) and directly interactswith nAChR (29,32) , but nAChR does not consistentlypartition into cholesterol-rich domains (74) . We suggesthere that some of these apparent contradictions may beexplained by competition between cholesterol and otherlipids found in native membranes, primarily lipids withpolyunsaturated fatty acyl chains (PUFAs).Interactions of nAChR with PUFAs have not beensystematically investigated experimentally, but a largeamount of circumstantial experimental evidence sug-gests an important role for PUFAs in nAChR function.Clinically, long-chain n − ω −
3) lipids have a neuroprotective role (75) , andnAChR-associated pathologies can arise for patientswith low levels of n − α (13) , and dietary supplementationwith n − ff ectsin some individual cases. (76) In vitro , PUFA-rich asolectin (77,78) is one of the mostrobust additives (20) for obtaining native nAChR func-tion: restoration of native function by cholesterylhemisuccinate (CHS) is observed only over a narrowCHS concentration range in monounsaturated PE / PSmembranes, but a much wider concentration range inasolectin (20) . The specific component(s) of asolectinthat complement cholesterol in improving nAChR func-tion have not been isolated. Long chained n − (79,80) and those of the Torpedo electric or- gan, (36,81) . Both such membranes also have an abun-dance of phosphoethanolamine (PE) headgroups andsaturated glycerophospholipids, and a scarcity of mo-nounsaturated acyl chains and sphingomyelin comparedto thhe
Xenopus oocyte membranes (82) common infunctional studies, or a “generalized” mammalian cellmembrane (83) .Membranes composed of ternary mixtures of satu-rated lipids, unsaturated lipids, and cholesterol tendto demix into separate domains. Saturated lipids andcholesterol constitute a rigid liquid ordered phase ( l o ) inwhich acyl chains remain relatively straight. (84–87) Un-saturated lipids form a more flexible liquid disorderedphase ( l do ) in which the chains remain fully melted. l o domains are often visualized as signaling “platforms”,restricting membrane proteins into high density “rafts”that di ff use within a fluid membrane (88,89) . This concep-tualization requires that l o domains have a much smallerarea than l do domains, and does not well-represent mem-branes that are over 30% cholesterol, such as neuronalmembranes.The first generation of studies into the mechanism un-derlying cholesterol-modulation of nAChR were con-ducted and interpreted in an era preceding the discov-ery of lipid-induced domain formation in membranes.The second generation explicitly considered potentialinteractions of nAChR with lipid domains, in part todetermine the requirements for the extremely high den-sity ( ∼ µ − ) of nAChRs at the neuromuscular junc-tion (90) . Since direct interaction between nAChR andcholesterol had been demonstrated in the first gener-ation of studies, a sensible initial hypothesis was thatnAChR persistently partitioned to l o domains, retaininglittle contact with unsaturated chains. Tests of this hy-pothesis have yielded results that are inconclusive, con-tradictory, or highly sensitive to lipid composition.Barrantes and colleagues (52) found that the additionof nAChRs to a domain-forming lipid mixture increasedthe size of Dipalmitoylphosphatidylcholine / Cholesterol(DPPC / Chol) lipid-ordered domains, which (combinedwith additional FRET data) was interpreted as indicat-ing nAChR was embedded in liquid-ordered domains.Some studies (45,91,92) suggest that nAChRs are associ-ated with microdomains independently of stimulationby other proteins associated with the neuromuscularjunction. Other studies (48,93) suggested that nAChRs re-quire stimulation by a protein such as agrin to partitioninto microdomains. Formation and disassembly of thenAChR-rich microdomains is highly sensitive to choles-terol concentration. (18,44,45,48,94)
These studies suggested a role for cholesterol-induced phase separation, but did not confirm that2AChR partitions to the cholesterol-rich phase. To testfor an intrinsic nAChR domain preference, Barrantesand co-workers checked for enrichment of nAChRsin the detergent resistant membrane (DRM). nAChRswere not enriched in the DRM of a model, domain-forming mixture (1:1:1 Chol: palmitoyloleoylphos-phatidylcholine(POPC): sphingomyelin) (74) but induc-ing compositional asymmetry across leaflets did yieldnAChR enrichment in the DRM fraction (95) . Whilemore precise and robust experimental methods for de-termining partitioning preference and specific boundarylipids such as mass spectrometry have been applied forother transmembrane proteins (96,97) , they have not beenapplied to complex heteromers like nAChR.Fully atomistic molecular dynamics (MD) simula-tions (98–101) have served as a natural complement tothe third-generation structural biology approach, but arelimited in their ability to resolve contradictions betweenfirst and second generation studies, because lipids areunable to di ff use over simulation time scales. (102–106) .E ffi cient lipid di ff usion is a requirement for equilibrat-ing domains or detecting protein-induced lipid sorting.Coarse-grained MD (CG-MD) has been used to greatsuccess in a number of simulations for both lipid-proteinbinding and membrane organization (103–108) . Here weuse CG-MD as a “computational microscope” to ob-serve the equilibrium distribution of lipids local to thenAChR in a range of binary and ternary lipid mix-tures inspired by native membranes. We observe a re-markable enrichment of polyunsaturated lipids amongnAChR boundary lipids. To our knowledge, these arethe first molecular simulations of the nAChR in non-randomly mixed membranes, and the first study to sys-tematically investigate the likelihood of polyunsaturatedlipids as nAChR boundary lipids.
2. Methods
All simulations reported here used the coarse-grainedMARTINI 2.2 (109) topology and forcefield. nAChRcoordinates were based on a cryo-EM structure of the αβγδ muscle-type receptor in native torpedo membrane(PDB 2BG9 (110) ). This is a medium resolution struc-ture (4Å) and was further coarse-grained using the mar-tinize.py script; medium resolution is su ffi cient for usein coarse-grained simulation, and the native lipid envi-ronment of the proteins used to construct 2BG9 is crit-ical for the present study. The secondary, tertiary andquaternary structure in 2BG9 was preserved via soft backbone restraints during simulation as described be-low, so any inaccuracies in local residue-residue inter-actions would not cause instability in the global confor-mation.Coarse-grained membranes were built using the Mar-tini script insane.py, which was also used to embed thecoarse-grained nAChR within the membrane. The in-sane.py script randomly places lipids throughout theinter- and extra-cellular leaflets, and each simulationpresented in this manuscript was built separately. Bi-nary mixed membranes were composed of one saturatedlipid species (Dipalmitoylphosphatidylcholine-DPPCor Dipalmitoylphosphatidylethanolamine-DPPE) andcholesterol (CHOL), while ternary mixed membranesalso included either two n − n − x x and 25 x x
25 nm , with about ∼ ∼ x x
40 nm with about ∼ ∼ ∼ with about ∼ ∼ l do phase didnot reflect finite size e ff ects. Molecular dynamics simulations were carried outusing GROMACS (111) ; small boxes used GROMACS5.0.6 and large and very large boxes used GROMACS5.1.2 or 5.1.4. All systems were run using van der Waals(vdW) and Electrostatics in shifted form with a dielec-tric constant of (cid:15) r =
15. vdW cuto ff lengths were between0.9 and 1.2 nm, with electrostatic cuto ff length at 1.2nm.Energy minimization was performed over 10000 to21000 steps. Molecular dynamics were run using atime step of 25 fs, as recommended by MARTINI,for 2 µ s for small membranes,and 10 µ s for large andvery large membranes. Simulations were conductedin the isothermal-isobaric (NPT) ensemble, by usinga Berendsen thermostat set to 323 K with temperature3oupling constant set to 1 ps, as well as isotropic pres-sure coupling with compressibility set to 3 × − bar − and a pressure coupling constant set to 3.0 ps.Secondary structures restraints consistent with MAR-TINI recommendations were constructed by the mar-tinize.py (109) script and imposed by Gromacs (111) . Pro-tein conformation was maintained in small systems viaharmonic restraint (with a spring constant of 1000 kJ · mol − ) on the position of backbone beads. nAChR con-formation in large systems was preserved via harmonicbonds between backbone beads separated by less than0.5 nm, calculated using the ElNeDyn algorithm (112) associated with MARTINI (109) with a coe ffi cient of900 kJ · mol − . These restraints limited the root-mean-squared-displacement (RMSD) of the backbone to lessthan 2.5 Å throughout the simulation.The minimum equilibration time depended on thesystem size. Small systems typically began domain for-mation by 500 ns, with domains fully formed by 1000ns. Large systems and very large simulations requiredabout 5 µ s of equilibration for stabilization of metricsdescribed below. Extent of domain formation within the membranewas tracked by M A , B ≡ (cid:104) n A , B (cid:105) x B − M A ≡ (cid:104) n A , A (cid:105) x A − n A , B is the number of type B molecules amongthe 6 nearest neighbors for a given type A molecule, theaverage is over time and all molecules of type A , andthe self-association metric is notated M A ≡ M A , A forbrevity. For a random mixture, (cid:104) n A , B (cid:105) = x B , where x B is the fraction of overall bulk lipids that are of typeB. M A = M A > M A < l o or l do do-main was tracked by counting the number b sat of satu-rated annular boundary lipids and comparing with theexpectation for a random mixture, via the order param-eter Q sat : Q sat ≡ x sat (cid:42) b sat b tot (cid:43) − , (2)where b tot is the total number of lipids in the annularboundary region and x sat is the fraction of overall bulklipids that are saturated phospholipids. Q sat < l do phase,while Q sat > l o phase. Each frame, b tot and b sat werecalculated by counting the number of total and saturatedlipids, respectively, for which the phosphate bead fellwithin a distance of 1.0 nm to 3.5 nm from the M2helices, projected onto the membrane plane.Two-dimensional density distribution of the beadswithin a given lipid species B around the protein wascalculated on a polar grid: ρ B ( r i , θ j ) = (cid:68) n B ( r i , θ j ) (cid:69) r i ∆ r ∆ θ (3)where r i = i ∆ r is the projected distance of the bin cen-ter from the protein center, θ j = j ∆ θ is the polar angleassociated with bin j, ∆ r =
10Å and ∆ θ = π radiansare the bin widths in the radial and angular direction re-spectively, and (cid:68) n B ( r i , θ j ) (cid:69) is the time-averaged numberof beads of lipid species B found within the bin cen-tered around radius r i and polar angle θ j . In order todetermine enrichment or depletion, the normalized den-sity ˜ ρ B ( r i , θ j ) is calculated by dividing by the approxi-mate expected density of beads of lipid type B in a ran-dom mixture, x B s B N L / (cid:104) L (cid:105) , where s B is the number ofbeads in one lipid of species B, N L is the total numberof lipids in the system, and (cid:104) L (cid:105) is the average projectedbox area: ˜ ρ B ( r i , θ j ) = ρ B ( r i , θ j ) x B s B N L / (cid:104) L (cid:105) (4)This expression is approximate because it does not cor-rect for the protein footprint or any undulation-induceddeviations of the membrane area. The associated cor-rections are small compared to the membrane area andwould shift the expected density for all species equally,without a ff ecting the comparisons we perform here.
3. Results
Lipid sorting was characterized for nAChRs in binaryDPPC:CHOL membranes (Figure 1A) using severalmetrics. Non-random lipid mixing (including domainformation) was quantified using the self-associationmetric M A as defined in Equation 1. As expected, insimulated binary membranes containing only DPPC and0-40% cholesterol, minimal demixing was observed,with values of M DPPC (Fig 1B) rising slightly for highercholesterol concentrations but remaining persistentlybelow 0.05.4
000 ns X
Chol M D PP C Q A B C ˜ ρ Chol
Figure 1: nAChR boundary lipids in binary mixtures of DPPC and CHOL. A: Representative frame from a simulated trajectory of a single nAChRembedded in a small membrane, colored by subunit ( α :green, β :purple, δ :gray, γ :cyan) in a 4:1 DPPC (blue):Chol (red) mixture. B: Extent ofdemixing ( M DPPC defined in Eq. 1) and depletion of saturated lipids from the boundary ( Q sat defined in Eq.2) in small binary membranes. Inthis binary system, cholesterol depletion / enrichment is directly related to the saturated lipid depletion / enrichment: Q chol = − x sat Q sat / x Chol . Errorbars represent standard error for a blocking average over 50 ns. C: Average normalized density (Eq. 4) of cholesterol for the system in A. Data isequivalent to that in Figure 5: Binary Mixture “Chol” row.
Depletion of saturated lipids among nAChR bound-ary lipids (relative to those expected for a random mix-ture) was quantified using the metric Q sat defined inequation 2. Negative and positive values of Q sat re-flect depletion or enrichment of saturated lipids in thenAChR boundary, respectively. In binary systems con-taining cholesterol and saturated lipids, depletion ofsaturated lipids corresponds directly to enrichment ofcholesterol: Q chol = − Q sat x sat / x Chol .In binary DPPC:CHOL mixtures, Q sat was veryslightly negative for x Chol < (19,20,113) and a phase transition at about 20% cholesterol in binaryDPPC:CHOL model membranes is indicated by di ff er-ential scanning calorimetry. (114) Spontaneous binding of cholesterol to non-annular or“embedded” sites, similar to what we previously pro-posed (98) , was observed in these CG-simulations, andpenetration of the TMD bundle by DPPC acyl chainswas also observed at lower cholesterol concentrations(Fig 1A). Distribution of density for embedded lipids isfurther discussed in Section 3.4.Annular cholesterol (enrichment of cholesterol at theprotein-lipid interface), is visible for the binary systemsvia a ring of high (red) cholesterol density just aroundthe protein in Figure 1C. Enrichment of cholesterol nearthe protein is highly localized with a ring that is lessthan 5Å wide. This is in general agreement with evi- dence for annular cholesterol in randomly-mixed binarymembranes. (49) ff ected by introduction of annAChR In order to test whether nAChR a ff ected domain for-mation in domain-forming membranes, we character-ized M PUFA for systems containing DPPC, Cholesterol,and PE or PC with either n-3 (DHA) or n-6 (LA) acylchains. Addition of phospholipids with unsaturatedacyl chains to systems containing a saturated lipid andcholesterol is well-established to induce domain for-mation, and polyunsaturated phospholipids make thesedomains more well-defined (115) . As expected, we ob-served that addition of PUFAs to DPPC / CHOL bilayersdid induce domain formation over a range of composi-tions, and values for M PUFA are shown as filled symbolsin Figure 2 A.Introducing a single nAChR to these same systemsdid not significantly a ff ect domain formation. M DHA was determined for an isolated nAChR in ternary mixedmembranes with over 40 di ff erent combinations ofDHA, DPPC, and Cholesterol (Figure 2A, shaded con-tours). Its e ff ect on membrane organization is repre-sented by the di ff erence in color of the circular symboland the shaded contour at the same composition. Intro-ducing a single nAChR into the DHA-containing sys-tems does slightly reduce the amount of DHA requiredto obtain a given value of M DHA . This subtle trend mayreflect increased likelihood of DHA-DHA interactions5 QX CH X CH X U n s a t X U n s a t DHA LA BA Figure 2: Quantitative analysis of bulk membrane mixing andnAChR boundary lipid composition across small membranes contain-ing DPPC, Cholesterol, and either dDHA-PE or dLA-PC. Shadedcontours were constructed based on 40 individual simulations withdDHA-PE and 30 with dLA-PC. A: M PUFA , defined in eq 1. Circlesrepresent mixing of systems with the same lipid composition but nonAChR. B: Q sat , defined in Eq 2. due to nucleation of DHA-containing lipids around theprotein (Figure 3).Across ternary mixtures with two long n − M DHA approached 5 (Figure 2A), and were signif-icantly reduced (to less than 0.5) when DHA chainswere replaced with linoleic acyl (LA) chains. This re-sult is consistent with a previously-observed significantincrease in miscibility temperature upon supplementa-tion of plasma membranes with n − (115) Substantial lipid demixing in DHA-containing mix-tures was observed even at low cholesterol concentra-tions. Over the range we tested, M DHA was not sensitiveto cholesterol concentration x Chol , as shown by the hor-izontal contours for DHA in Figure 2A.
For more than 70 lipid compositions tested, nAChRalways partitioned into a PUFA-rich l do phase if sucha phase was present. We never observed nAChR par-titioning to an l o phase. Representative frames fromtrajectories of domain formation in the presence of nAChR are shown in Figure 3. This observation in-cludes all tested concentrations of the ternary mix-tures, regardless of whether the zwitterionic head-group was PC or PE (Figure 4), or whether DPPCwas replaced by dioleoylphosphatidylcholine (DOPC)(di-18:1), Palmitoyloleoylphosphatidylcholine (POPC)(16:0,18:1), or dilauroylphosphatidylcholine (DLPC)(di-14:0), as shown in Figure S1.These results are quantified for nAChR embedded internary membranes containing DPPC, CHOL, and ei-ther dDHA-PE or dLA-PC in Figure 2 B, using the met-ric Q sat defined in equation 2. In all systems studiedhere, Q sat <
0, indicating depletion of saturated lipidsas boundary lipids, consistent with observed partition-ing to the l do domain in Figure 3. Furthermore, de-pletion was much stronger in systems containing DHA( Q DHA sat << Q LA sat ), consistent with the more well-definedDHA domains ( M DHA >> M LA ).The nAChR annulus is highly enriched in DHA:DHA-PE constitute nearly 100% of the local lipids evenin membranes with very low DHA concentrations. Thisstrong signal could indicate multiple high a ffi nity sitesfor DHA chains across the transmembrane protein sur-face. At another extreme, DHA enrichment could bedriven by a very slight preference for DHA in a highlynon-ideal bulk: since DHA is found in well-defined do-mains without protein, even one DHA molecule thatbinds to the protein surface could stabilize the rest ofthe l do domain nearby. Comparing boundary lipid anddomain formation trends can help distinguish betweenthese two scenarios. If boundary lipid enrichment isdetermined purely by how well-defined domains are(the latter scenario), we would expect similar trendsfor M DHA and Q sat in the DHA column of Figure 2.In contrast, Figure 2 shows that while domain forma-tion in DHA-containing systems is only weakly sensi-tive to cholesterol content (horizontal contours), com-position of boundary lipids is highly sensitive to choles-terol content (diagonal contours). These results suggestthat direct interactions between multiple favorable siteson nAChRand DHA-containing lipids dominate the ob-served enrichment of DHA among boundary lipids.The simulations represented in Figure 2 do comparethe e ff ects of two unsaturated lipids that also have dif-ferent headgroups. DHA is far more commonly pairedwith PE in native membranes, while LA is more com-monly found with PC. We found no qualitative dif-ferences in nAChR domain partitioning or significantquantitative e ff ect on Q sat upon switching PC and PEheadgroups on the PUFA lipid. We did observe a quan-titative e ff ect of saturated lipid headgroup on boundarylipid composition: Q sat was reduced by half when satu-6 HALA DHA
AB C
Figure 3: Trajectories of ternary mixtures at ratios of 2:2:1 DPPC:PUFA:Chol. A and B: Trajectories of simulation systems with a single nAChRembedded within small membranes, using lipids containing DHA acyl chains or LA acyl chains. Both simulations were run for 2 µ s. C: Finalsnapshot of 4 µ s trajectory of a system within a large ∼ membrane with the same composition as in A. Subunits are colored: α : green, β : purple, δ : gray, γ : cyan. Lipids are colored: Chol: red, DPPC: blue, dDHA-PE: white, dLA-PC: tan. SatFA (blue)PC PEPCPE P U F A ( S il v e r) Q Sat =-0.56±0.01Q
Sat =-0.44±0.02 Q
Sat =-0.26±0.01Q
Sat =-0.28±0.03
Figure 4: Comparison of nAChR partitioning based on lipidheadgroups (PC and PE). All images represent last frame of2 µ s simulations of small membranes with composition 2:2:1Sat:PUFA:Cholesterol. Rows represent the head-group for the PUFA-containing lipid, while columns represent the head-group of the satu-rate lipid. Each image includes Q sat values related to individual sys-tems with errors across averaging 50 ns blocks. rated PE was used instead of saturated PC. (Figure 4).As shown in Figure 4, nAChR is bordered by l o domainson two opposing faces when saturated PE is used, com-pared to only one face if PC is used. The particular do-main topology shown in Figure 4 is an artifact of the pe-riodic boundary conditions, but still indicates more fa-vorable interactions of nAChR with an l o domain com-posed of DPPE vs DPPC. This may reflect a di ff erence in the lipid shape (wedge-shaped DPPE vs cylindrical-shaped DPPC) and the associated monolayer sponta-neous curvature. For PUFA lipids in flexible l do do-mains, lipid shape is less likely to play a significant rolein determining partitioning. The dramatic di ff erence indomain flexibility is apparent in Figure S2. The nAChR structure used for these simulations wasdetermined in a native membrane with a high fractionof polyunsaturated lipids. While we previously (98) pro-posed that unresolved density in this structure could beembedded cholesterol, the possibility of occupation byphospholipids other than POPC was not investigated.Furthermore, we did not consider possible asymmetryacross subunits in binding previously. Here we do ob-serve penetration of both the intersubunit (“type B”)and the intrasubunit (“type A / C”) sites previously pro-posed (98) , by both phospholipids and cholesterol, butwith a high degree of subunit specificity.Two dimensional density distributions of DPPC, PU-FAs, and cholesterol over short and long length scaleswere measured for two ternary mixtures and one binarymixture (Figure 5). In binary DPPC / cholesterol mem-branes, DPPC was more likely than cholesterol to oc-cupy intrasubunit sites. DPPC binds shallowly in the α subunit and more deeply in the β subunit. Introduc-ing PUFAs resulted in displacement of both cholesterol7 PPCCholdDHA-PEDPPCCholdLA-PCTernary Mixture with short n-6 PUFAs ˜ ρ Ternary Mixture with long n-3 PUFAs DPPCCholBinary Mixture ˜ ρ log ˜ ρ ˜ ρ = 0 Boundary regionMembrane
Figure 5: Lipid density enrichment or depletion around a single cen-tral nAChR. Heatmaps are colored according to the normalized den-sity ˜ ρ a (left, defined in eq 4) or ln ˜ ρ a (right), averaged over the final5 µ s of a 10 µ s simulation. Membrane column (left) depicts densityacross the simulated membrane; ˜ ρ a < ρ a > and DPPC from intrasubunit sites, except for the β intra-subunit site, which became more likely to be occupiedby cholesterol. The interior of the β subunit TMD hasthe largest amount of available volume, could sequestercholesterol (but not DPPC) from the PUFA lipids in theannulus, and filling the interior with a PUFA chain maybe entropically costly. PUFA chains did occupy otherintrasubunit sites, but remained fluid, as shown in Fig-ure 6.Intersubunit sites were rarely occupied by DPPC,with the exception of the β + /α − site in the binary sys-tem (Figure 5). Intersubunit sites were more likely tobind cholesterol, particularly the β + /α − , α + /γ − , and α + /δ − subunit interfaces. Occupation of the α + /δ − in-terface is consistent with cryo-EM observations (116) ofenhanced cholesterol density around the α + /δ − site.Intersubunit sites that were not significantly occupiedby cholesterol ( δ + /β − and γ − /α + ) did show signifi-cant and deep occupation by DHA, which tended to en-ter from the adjacent intrasubunit site rather than fromthe membrane. Even those intersubunit sites with sig-nificant cholesterol occupancy can simultaneously bindpart of a DHA chain, yielding non-vanishing DHA den-sity. We also calculated density distributions of each lipidspecies at distances beyond the “annular” ring, overthe 5-20 nm range. As shown in Figure 5 (left col-umn), observed sorting of lipids within 5-20 nm of thenAChR is dependent on the overall composition of themembrane. For all compositions shown, cholesterol isdepleted within 5-20 nm and enriched even farther fromthe protein. Within the binary systems this e ff ect isminor ( ˜ ρ CHOL ∼ ρ CHOL ∼ .
5) andsubstantial ( ˜ ρ CHOL ∼ .
25) for the highly-segregatedDHA containing systems. A similar pattern is ob-served for DPPC, which suggests that “sorting” overthe 5-20 nm range is primarily driven by intrinsic di ff er-ences in membrane organization that would be observedwithout the receptor. PUFAs are also most highly en-riched at intermediate distances : the deepest red bandis found at about 5 nm in LA-containing systems andabout 8 nm in DHA-containing systems. This would beexpected when nAChR partitions near a curved domainboundary, as in Figure 4.8 nm Figure 6: Embedded lipids in the nAChR. Main image: Represen-tative frame from equilibrated small membrane simulation of nAChRin 2:2:1 DPPC:DHA-PE:CHOL. Backbone beads of the TMD helicesare colored by subunit as in Figure 2; side-chain beads are not shown.Both DHA-PE (white) and cholesterol (red) equilibrate to embed-ded sites in the subunit center and subunit interfaces, although mostcholesterol is found in the l o phase with DPPC (blue). Inset : Cryo-EM density of nAChR from (117) as rendered in (98) ; dark blue indi-cates high density, white is medium density, and red is low density.
4. Discussion
In this work we used coarse-grained simulations topredict the local lipid composition around the nicotinicacetylcholine receptor, in a range of domain formingmembranes. We observed nAChR partitioning to theliquid-disordered phase in all systems for which such aphase was present. This is inconsistent with the modelof lipid rafts as platforms that contain a high densityof nAChRs, and unexpected in light of the establishedcholesterol dependence of nAChR. As shown in thesesimulations, partitioning to the l do phase does not pre-vent nAChR from accessing cholesterol.The simulations presented here involve only one re-ceptor per system. Using the present results only, thesimplest extrapolation to multiple receptors would as-sume that receptors are simply distributed randomlyacross the l do domain. The local receptor area densitywould be the number of receptors divided by the totalarea of l do domains.In the model membranes used here, as well as in na-tive nAChR membranes, the lipid composition wouldbe expected to yield l do phases that were about the samesize as l o phases. The l o “raft” in an l do “sea” analogy isnot representative when over 50% of the membrane is inthe “raft” phase. A more representative analogy wouldbe receptors as boats, floating on an l do lake within an l o rigid land mass. Filling in the lake by adding to the coastline would force any boats in the lake closer to-gether. Similarly, any process that decreased total l do area while keeping the number of receptors constantwould increase the local receptor density. In this model,observing increased nAChR density by adding mem-brane cholesterol (as in (18,44–48,52,53,118) ) would be con-sistent with nAChR partitioning to the cholesterol-poorphase rather than the cholesterol-rich phase.This extrapolation from a single receptor assumesthat introduction of additional receptors does not changepartitioning behavior. We do still find reliable parti-tioning to the l do phase upon adding more receptors,and we will characterize systems with multiple recep-tors in a future publication. Due to receptor dimeriza-tion and trimerization, distribution of individual recep-tors within the l do phase will not be random. This wouldnot change the expected trend of density increasing withadded cholesterol, however.Observed partitioning into the l do phase could be con-sidered inconsistent with interpretations of some exper-iments, (74,95) which suggest minimal nAChR partition-ing preference in symmetric model membranes or anactual preference for an l o phase in asymmetric modelmembranes. These experiments used only monounsatu-rated acyl chains, and may have had less well-defineddomains. They further relied on detergent resistantmembrane (DRM) methods, which are sensitive to thechoice of detergent (120) and could be unable to distin-guish between proteins with no partitioning preferencevs proteins that persistently partition to one side of aboundary.The origin of preferential partitioning observed inthese simulations for the l do domain is still unclear,but may reflect di ff erent elastic properties of the l do and l o domains. In general, proteins embedded inmembranes will introduce a boundary condition onthe membrane shape, such that (1) the thickness ofthe membrane matches the thickness of the transmem-brane domain (121–123) and (2) interfacial lipids are par-allel to the protein surface. (124) . Transmembrane pro-teins with hydrophobic mismatch with the surround-ing membrane may deform the membrane thickness tosatisfy constraint (1), while cone-shaped proteins likepLGICs must also introduce a “tilt” deformation to sat-isfy (2). Each leaflet of the membrane has an elasticresistance to bending away from its spontaneous cur-vature, and satisfying these constraints is energeticallycostly.Continuum theories based on the Helfrich Hamil-tonian have been used to predict shape deforma-tions around protein inclusions in homogeneous mem-branes. (121,123,124) In mixed membranes, minimization9f the protein-deformation free energy may also inducelipid sorting. Two distinct sorting mechanisms couldminimize the bending free energy: sorting that A) re-duces the required bending deformation, by selectingboundary lipids with a specific thickness, leaflet asym-metry, or shape or B) reduces the free energy cost of thebending deformation, by selecting for flexible boundarylipids. Mechanism (B) is the most generally applicableapproach, and would stabilize partitioning to the mostflexible domains, consistent with our observations (Fig-ure S2). In some cases, mechanism (A) may also con-tribute to partitioning or lipid-sorting, and could explainwhy nAChR tends to attract saturated PE over saturatedPC, or how leaflet asymmetry can promote partitioningto more rigid phases as observed in (95) .We previously (98) proposed unresolved density in thecryo-EM structure of nAChR in the
Torpedo mem-brane could be embedded cholesterol, based on gainof function caused by cholesterol in reconstitution mix-tures (113,125–131) , but we did not consider the possibil-ity of occupation by polyunsaturated chains. Here weobserve spontaneous binding of cholesterol to coarse-grained embedded sites, but long-chain PUFA tails dis-place cholesterol in some binding sites. Long acylchains may penetrate far into the TMD bundle with-out requiring the entire head group also be incorporated,and long-chain PUFAs may do so without as substantialan entropic penalty as long saturated chains. Choles-terol (like phosphatidic acid, another lipid known tocause gain of function under some preparations (127–131) )has a much smaller headgroup than PC or PE. It can be-come fully incorporated into the TMD without the TMDneeding to accommodate the bulky headgroup. Thesecomplex associations underlie the challenges of pre-dicting local lipid environment in heterogeneous, highlynon-ideal mixtures.All simulations reported here contain lipids with di-saturated tails or di-PUFA tails. While lipid specieswith two identical acyl chains do exist in the nativemembrane, they are far less common than hybrid lipidswith heterogeneous acyl chains. Including hybrid lipidswould reduce the potential for formation of large do-mains, while increasing the length of the domain in-terface. Incorporation of hybrid lipids would also re-duce the nAChR-local concentration of PUFA chains.Even 5-10% DHA is a saturating concentration fornAChR cavities, however, so we expect occupation ofcavities to be minimally a ff ected by replacement of di-DHA lipids with twice the number of hybrid lipids.
5. Acknowledgment
GB and RS were supported by research grants NSFMCB1330728 and NIH P01GM55876-14A1. GB andLM were also supported through a grant from the Re-search Corporation for Scientific Advancement. Thisproject was supported with computational resourcesfrom the National Science Foundation XSEDE programthrough allocation NSF-MCB110149, a local clusterfunded by nsf-dbi1126052, the Rutgers University Of-fice of Advanced Research Computing (OARC) and theRutgers Discovery Informatics Institute (RDI2), whichis supported by Rutgers and the State of New Jersey. Weare grateful to Dr. J´erˆome H´enin for his helpful sugges-tions throughout this study.
References [1] J. A. Dani, Overview of nicotinic receptors and their roles inthe central nervous system, Biological psychiatry 49 (2001)166–174.[2] J.-P. Changeux, P.-J. Corringer, U. Maskos, The nicotinicacetylcholine receptor: From molecular biology to cognition.,Neuropharmacology 96 (2015) 135–136.[3] C. D. Klaassen, J. B. Watkins, Casarett & Doull’s essentials oftoxicology, McGraw Hill Professional, 2015.[4] A. Taly, P.-J. Corringer, D. Guedin, P. Lestage, J.-P. Changeux,Nicotinic receptors: allosteric transitions and therapeutic tar-gets in the nervous system, Nature Publishing Group 8 (2009).[5] H. Patel, J. Mcintire, S. Ryan, A. Dunah, R. Loring, Anti-inflammatory e ff ects of astroglial α κ B pathway andactivation of the Nrf2 pathway, Journal of Neuroinflammation14 (2017).[6] G. T. Yocum, D. L. Turner, J. Danielsson, M. B. Barajas,Y. Zhang, D. Xu, N. L. Harrison, G. E. Homanics, D. L. Far-ber, C. W. Emala, GABA A receptor α α lzheimer’s disease, Journal of Neurochemistry 73 (1999)1635–1640.[12] A. S. Arnold, M. Gueye, S. Guettier-Sigrist, I. Courdier-Fruh,G. Coupin, P. Poindron, J. P. Gies, Reduced expression ofnicotinic AChRs in myotubes from spinal muscular atrophy Ipatients, Laboratory Investigation 84 (2004) 1271–1278.[13] S. N. Haydar, J. Dunlop, Neuronal nicotinic acetylcholine re-ceptors - targets for the development of drugs to treat cognitiveimpairment associated with schizophrenia and alzheimer’s dis-ease., Current topics in medicinal chemistry 10 (2010) 144–152.[14] V. A. Lennon, L. G. Ermilov, J. H. Szurszewski, S. Vernino,Immunization with neuronal nicotinic acetylcholine receptorinduces neurological autoimmune disease, Journal of ClinicalInvestigation 111 (2003) 907–913.[15] A. K. Hamouda, M. Sanghvi, D. Sauls, T. K. Machu, M. P.Blanton, Assessing the lipid requirements of the torpedo cal-ifornica nicotinic acetylcholine receptor., Biochemistry 45(2006) 4327–4337.[16] J. E. Baenziger, J. A. Domville, J. P. D. Therien, The role ofcholesterol in the activation of nicotinic acetylcholine recep-tors., Current topics in membranes 80 (2017) 95–137.[17] L. F. Padilla-Morales, J. O. Coln-Sez, J. E. Gonzlez-Nieves,O. Quesada-Gonzlez, J. A. Lasalde-Dominicci, Assessment ofthe functionality and stability of detergent purified nachr fromtorpedo using lipidic matrixes and macroscopic electrophysi-ology., Biochimica et biophysica acta 1858 (2016) 47–56.[18] F. J. Barrantes, Cholesterol e ff ects on nicotinic acetylcholinereceptor, Journal of neurochemistry 103 (2007) 72–80.[19] A. W. Dalziel, E. S. Rollins, M. G. McNamee, The e ff ectof cholesterol on agonist-induced flux in reconstituted acetyl-choline receptor vesicles, FEBS Letters 122 (1980) 193–196.[20] M. Criado, H. Eibl, F. Barrantes, E ff ects of lipids on acetyl-choline receptor. essential need of cholesterol for maintenanceof agonist-induced state transitions in lipid vesicles, Biochem-istry 21 (1982) 3622–3629.[21] E. L. Ochoa, A. W. Dalziel, M. G. Mcnamee, Reconstitu-tion of acetylcholine receptor function in lipid vesicles of de-fined composition, Biochimica et Biophysica Acta (BBA)-Biomembranes 727 (1983) 151–162.[22] D. Marsh, F. Barrantes, Immobilized lipid in acetylcholinereceptor-rich membranes from torpedo marmorata, Proceed-ings of the National Academy of Sciences 75 (1978) 4329–4333.[23] D. Marsh, A. Watts, F. Barrantes, Phospholipid chain immobi-lization and steriod rotational immobilization in acetylcholinereceptor-rich membranes from torpedo marmorata, Biochimicaet Biophysica Acta (BBA)-Biomembranes 645 (1981) 97–101.[24] J. Gonzalez-Ros, M. Llanillo, A. Paraschos, M. Martinez-Carrion, Lipid environment of acetylcholine receptor from tor-pedo californica, Biochemistry 21 (1982) 3467–3474.[25] M. Mcnamee, J. Ellena, A. Dalziel, Lipid-protein interactionsin membranes containing the acetylcholine receptor, Biophys-ical journal 37 (1982) 103.[26] J. F. Ellena, M. A. Blazing, M. G. Mcnamee, Lipid-protein in-teractions in reconstituted membranes containing acetylcholinereceptor, Biochemistry 22 (1983) 5523–5535.[27] J. R. Zabrecky, M. A. Raftery, The role of lipids in the functionof the acetylcholine receptor, Journal of Receptors and SignalTransduction 5 (1985) 397–417.[28] D. R. Bristow, I. L. Martin, Solubilisation of the γ -Aminobutyric Acid / Benzodiazepine Receptor from Rat Cere-bellum: Optimal Preservation of the Modulatory Responses byNatural Brain Lipids, Journal of Neurochemistry 49 (1987)1386–1393. [29] W. S. Leibel, L. L. Firestone, D. C. Legler, L. M. Braswell,K. W. Miller, Two pools of cholesterol in acetylcholinereceptor-rich membranes from¡ i¿ torpedo, Biochimica et Bio-physica Acta (BBA)-Biomembranes 897 (1987) 249–260.[30] D. S. Middlemas, M. A. Raftery, Identification of subunitsof acetylcholine receptor that interact with a cholesterol pho-toa ffi nity probe, Biochemistry 26 (1987) 1219–1223.[31] O. T. Jones, J. H. Eubanks, J. P. Earnest, M. G. McNamee, Aminimum number of lipids are required to support the func-tional properties of the nicotinic acetylcholine receptor., Bio-chemistry 27 (1988) 3733–3742.[32] O. T. Jones, M. G. Mcnamee, Annular and nonannular bindingsites for cholesterol associated with the nicotinic acetylcholinereceptor, Biochemistry 27 (1988) 2364–2374.[33] T. M. Fong, M. G. McNamee, Correlation between acetyl-choline receptor function and structural properties of mem-branes, Biochemistry 25 (1986) 830–840.[34] T. M. Fong, M. G. McNamee, Stabilization of acetylcholinereceptor secondary structure by cholesterol and negativelycharged phospholipids in membranes, Biochemistry 26 (1987)3871–3880.[35] M. McNamee, T. Fong, E ff ects of membrane lipids and fluid-ity on acetylcholine receptor function, Lipid Domains and theRelationship to Membrane Function. New York, Liss (1988)43–62.[36] F. J. Barrantes, The lipid environment of the nicotinic acetyl-choline receptor in native and reconstituted membrane, Criticalreviews in biochemistry and molecular biology 24 (1989) 437–478.[37] C. Sunshine, M. G. McNamee, Lipid modulation of nicotinicacetylcholine receptor function: the role of neutral and nega-tively charged lipids, Biochimica Et Biophysica Acta (BBA)-Biomembranes 1108 (1992) 240–246.[38] C. Sunshine, M. G. McNamee, Lipid modulation of nico-tinic acetylcholine receptor function: the role of membranelipid composition and fluidity, Biochimica et Biophysica Acta(BBA)-Biomembranes 1191 (1994) 59–64.[39] V. Narayanaswami, M. G. McNamee, Protein-lipid interac-tions and torpedo californica nicotinic acetylcholine recep-tor function. 2. membrane fluidity and ligand-mediated alter-ation in the accessibility of gamma subunit cysteine residues tocholesterol., Biochemistry 32 (1993) 12420–12427.[40] G. H. Addona, H. S. Jr, M. A. Kloczewiak, S. S. Husain, K. W.Miller, Where does cholesterol act during activation of thenicotinic acetylcholine receptor?, Biochimica et BiophysicaActa (BBA)-Biomembranes 1370 (1998) 299–309.[41] J. Corbin, H. H. Wang, M. P. Blanton, Identifying the choles-terol binding domain in the nicotinic acetylcholine receptorwith [ 125 I] azido-cholesterol, Biochimica et Biophysica Acta(BBA)-Biomembranes 1414 (1998) 65–74.[42] F. J. Barrantes, S. S. Antollini, M. P. Blanton, M. Prieto,Topography of nicotinic acetylcholine receptor membrane-embedded domains, Journal of Biological Chemistry 275(2000) 37333–37339.[43] J. E. Baenziger, C. M. H´enault, J. P. D. Therien, J. Sun, Nico-tinic acetylcholine receptor-lipid interactions: Mechanistic in-sight and biological function, Biochimica et Biophysica Acta- Biomembranes 1848 (2015) 1806–1817.[44] J. L. Brus´es, N. Chauvet, U. Rutishauser, Membrane lipidrafts are necessary for the maintenance of the (alpha)7 nico-tinic acetylcholine receptor in somatic spines of ciliary neu-rons., Journal of Neuroscience 21 (2001) 504–512.[45] S. Marchand, A. Devillers-Thi´ery, S. Pons, J.-P. Changeux,J. Cartaud, Rapsyn escorts the nicotinic acetylcholine recep-tor along the exocytic pathway via association with lipid rafts., he Journal of neuroscience : the o ffi cial journal of the Societyfor Neuroscience 22 (2002) 8891–8901.[46] J. Oshikawa, Y. Toya, T. Fujita, M. Egawa, J. Kawabe,S. Umemura, Y. Ishikawa, Nicotinic acetylcholine receptoralpha 7 regulates cAMP signal within lipid rafts., Americanjournal of physiology: Cell physiology 285 (2003) C567–74.[47] C. Pato, F. Stetzkowski-Marden, K. Gaus, M. Recouvreur,A. Cartaud, J. Cartaud, Role of lipid rafts in agrin-elicitedacetylcholine receptor clustering, Chemico-Biological Inter-actions 175 (2008) 64–67.[48] D. Zhu, W. C. Xiong, L. Mei, Lipid rafts serve as a signalingplatform for nicotinic acetylcholine receptor clustering., TheJournal of neuroscience : the o ffi cial journal of the Society forNeuroscience 26 (2006) 4841–4851.[49] F. J. Barrantes, V. Bermudez, M. V. Borroni, S. S. Antollini,M. F. Pediconi, J. C. Baier, I. Bonini, C. Gallegos, A. M. Roc-camo, A. S. Valles, V. Ayala, C. Kamerbeek, Boundary lipidsin the nicotinic acetylcholine receptor microenvironment, in:Journal of Molecular Neuroscience, volume 40, pp. 87–90.[50] V. Bermudez, S. S. Antollini, G. A. F. Nievas, M. I. Aveldao,F. J. Barrantes, Partition profile of the nicotinic acetylcholinereceptor in lipid domains upon reconstitution, Journal of lipidresearch 51 (2010) 2629–2641.[51] V. L. Perillo, D. A. Pe˜nalva, A. J. Vitale, F. J. Barrantes, S. S.Antollini, Transbilayer asymmetry and sphingomyelin compo-sition modulate the preferential membrane partitioning of thenicotinic acetylcholine receptor in Lo domains, Archives ofBiochemistry and Biophysics 591 (2016) 76–86.[52] J. J. Wenz, F. J. Barrantes, Nicotinic acetylcholine recep-tor induces lateral segregation of phosphatidic acid and phos-phatidylcholine in reconstituted membranes, Biochemistry 44(2005) 398–410.[53] M. V. Borroni, A. S. Valls, F. J. Barrantes, The lipid habitats ofneurotransmitter receptors in brain., Biochimica et biophysicaacta 1858 (2016) 2662–2670.[54] N. Unwin, Segregation of lipids near acetylcholine-receptorchannels imaged by cryo-em., IUCrJ 4 (2017) 393–399.[55] S. Basak, N. Schmandt, Y. Gicheru, S. Chakrapani, Crys-tal structure and dynamics of a lipid-induced potentialdesensitized-state of a pentameric ligand-gated channel, eLife6 (2017).[56] T. Altho ff , R. E. Hibbs, S. Banerjee, E. Gouaux, X-ray struc-tures of GluCl in apo states reveal a gating mechanism of Cys-loop receptors., Nature 512 (2014) 333–337.[57] D. Laverty, P. Thomas, M. Field, O. J. Andersen, M. G.Gold, P. C. Biggin, M. Gielen, T. G. Smart, Crystal struc-tures of a GABA A -receptor chimera reveal new endogenousneurosteroid-binding sites, Nature Structural and MolecularBiology 24 (2017) 977–985.[58] S. Zhu, C. M. Noviello, J. Teng, R. M. Walsh, J. J. Kim, R. E.Hibbs, Structure of a human synaptic gabaa receptor., Nature559 (2018) 67–72.[59] S. M. Dunn, C. R. Martin, M. W. Agey, R. Miyazaki, Func-tional reconstitution of the bovine brain gabaa receptor fromsolubilized components, Biochemistry 28 (1989) 2545–2551.[60] T. Sooksawate, M. A. Simmonds, E ff ects of membrane choles-terol on the sensitivity of the gabaa receptor to gaba in acutelydissociated rat hippocampal neurones, Neuropharmacology 40(2001) 178–184.[61] J. E. Baenziger, P.-J. Corringer, 3D structure and al-losteric modulation of the transmembrane domain of pen-tameric ligand-gated ion channels, Neuropharmacology 60(2011) 116–125.[62] Z. Dostalova, X. Zhou, A. Liu, X. Zhang, Y. Zhang, R. De-sai, S. A. Forman, K. W. Miller, Human α β γ
2l gamma- aminobutyric acid type a receptors: High-level production andpurification in a functional state., Protein science : a publica-tion of the Protein Society 23 (2014) 157–166.[63] A. Bhushan, M. G. McNamee, Correlation of phospholipidstructure with functional e ff ects on the nicotinic acetylcholinereceptor. a modulatory role for phosphatidic acid, Biophysicaljournal 64 (1993) 716–723.[64] M. H. Cheng, L. T. Liu, A. C. Saladino, Y. Xu, P. Tang, Molec-ular dynamics simulations of ternary membrane mixture: phos-phatidylcholine, phosphatidic acid, and cholesterol, The Jour-nal of Physical Chemistry B 111 (2007) 14186–14192.[65] J. Corrie, A. A. Ogrel, E. A. Mccardy, M. P. Blanton, J. E.Baenziger, Lipid-protein interactions at the nicotinic acetyl-choline receptor: A functional coupling between nicotinic re-ceptors and phosphatidic acid-containing lipid bilayers, Jour-nal of Biological Chemistry 277 (2002) 201–208.[66] C. J. B. DaCosta, A. A. Ogrel, E. A. Mccardy, M. P. Blan-ton, J. E. Baenziger, Lipid-Protein Interactions at the NicotinicAcetylcholine Receptor, The Journal of biological chemistry277 (2002) 201–208.[67] S. E. Rankin, G. H. Addona, M. A. Kloczewiak, B. Bugge,K. W. Miller, The cholesterol dependence of activation and fastdesensitization of the nicotinic acetylcholine receptor., Bio-physical journal 73 (1997) 2446–2455.[68] K. Burger, G. Gimpl, F. Fahrenholz, Regulation of receptorfunction by cholesterol., Cell Mol Life Sci 57 (2000) 1577–1592.[69] A. G. Lee, How lipids a ff ect the activities of integral membraneproteins, 2004.[70] T. J. Pucadyil, A. Chattopadhyay, Role of cholesterol inthe function and organization of g-protein coupled receptors.,Progress in lipid research 45 (2006) 295–333.[71] M. Landreh, M. T. Marty, J. Gault, C. V. Robinson, A slidingselectivity scale for lipid binding to membrane proteins, 2016.[72] N. Smithers, J. H. Bolivar, A. G. Lee, J. M. East, Charac-terizing the fatty acid binding site in the cavity of potassiumchannel KcsA, Biochemistry 51 (2012) 7996–8002.[73] J. Santiago, G. R. Guzm´an, L. V. Rojas, R. Marti, G. A. Asmar-Rovira, L. F. Santana, M. McNamee, J. A. Lasalde-Dominicci,Probing the e ff ects of membrane cholesterol in the torpedocalifornica acetylcholine receptor and the novel lipid-exposedmutation α c418w in xenopus oocytes, Journal of BiologicalChemistry 276 (2001) 46523–46532.[74] V. Berm´udez, S. S. Antollini, G. A. F. Nievas, M. I. Avelda˜no,F. J. Barrantes, Partition profile of the nicotinic acetylcholinereceptor in lipid domains upon reconstitution, Journal of LipidResearch 51 (2010) 2629–2641.[75] D. Piomelli, G. Astarita, R. Rapaka, A neuroscientist’s guideto lipidomics., Nat Rev Neurosci 8 (2007) 743–754.[76] G. P. Amminger, M. R. Sch¨afer, K. Papageorgiou, C. M. Klier,S. M. Cotton, S. M. Harrigan, A. Mackinnon, P. D. McGorry,G. E. Berger, Long-chain ω -3 fatty acids for indicated preven-tion of psychotic disorders: a randomized, placebo-controlledtrial, Archives of general psychiatry 67 (2010) 146–154.[77] C. Regost, J. Arzel, J. Robin, G. Rosenlund, S. J. Kaushik,Total replacement of fish oil by soybean or linseed oil with areturn to fish oil in turbot (Psetta maxima) 1. Growth perfor-mance, flesh fatty acid profile, and lipid metabolism, Aquacul-ture 217 (2003) 465–482.[78] R. E. Olsen, B. T. Dragnes, R. Myklebust, E. Ringø, E ff ectof soybean oil and soybean lecithin on intestinal lipid com-position and lipid droplet accumulation of rainbow trout, On-corhynchus mykiss Walbaum, Fish Physiology and Biochem-istry 29 (2003) 181–192.[79] W. C. Breckenridge, I. G. Morgan, J. P. Zanetta, G. Vincendon, dult rat brain synaptic vesicles II. Lipid composition, BBA -General Subjects 320 (1973) 681–686.[80] C. Cotman, M. L. Blank, A. Moehl, F. Snyder, Lipid Compo-sition of Synaptic Plasma Membranes Isolated from Rat Brainby Zonal Centrifugation, Biochemistry 8 (1969) 4606–4612.[81] O. Quesada, C. Gonz´alez -Freire, M. C. Ferrer, J. O. Col´on -S´aez, E. Fern´andez-Garc´ıa, J. Mercado, A. D´avila, R. Morales,J. Lasalde-Dominicci, Uncovering the lipidic basis for thepreparation of functional nicotinic acetylcholine receptor de-tergent complexes for structural studies, Scientific Reports 6(2016) 32766.[82] W. G. Hill, N. M. Southern, M. Bryce, E. Potter, G. Apodaca,C. P. Smith, M. L. Zeidel, Isolation and characterization of thexenopus oocyte plasma membrane: a new method for studyingactivity of water and solute transporters., American Journal ofPhysiology-Renal Physiology 289 (2005) F217–24.[83] H. I. Ing´olfsson, M. N. Melo, F. J. Van Eerden, C. Arnarez,C. A. Lopez, T. A. Wassenaar, X. Periole, A. H. De Vries,D. P. Tieleman, S. J. Marrink, Lipid organization of the plasmamembrane, Journal of the American Chemical Society 136(2014) 14554–14559.[84] S. E. Feller, Acyl chain conformations in phospholipid bilay-ers: a comparative study of docosahexaenoic acid and saturatedfatty acids, Chemistry and Physics of Lipids 153 (2008) 76–80.[85] P. L. Yeagle, Chapter 7 - structures of lipid assemblies, in:P. L. Yeagle (Ed.), The Membranes of Cells (Third Edition),Academic Press, Boston, third edition edition, 2016, pp. 115 –154.[86] P. Cicuta, S. L. Keller, S. L. Veatch, Di ff usion of liquid do-mains in lipid bilayer membranes, Journal of Physical Chem-istry B 111 (2007) 3328–3331.[87] J. V. Bleecker, P. A. Cox, S. L. Keller, Mixing Temperatures ofBilayers Not Simply Related to Thickness Di ff erences betweenLo and Ld Phases, Biophysical Journal 110 (2016) 2305–2308.[88] K. Simons, E. Ikonen, Functional rafts in cell membranes.: EB-SCOhost, 1997.[89] K. Simons, D. Toomre, Lipid rafts and signal transduction.,Nature Reviews Molecular Cell Biology 1 (2000) 31–39.[90] W. C. Breckenrldge, G. Gombos, I. G. Morgan, the Lipid Com-position of Adult Rat Brain Synaptosomal Plasma Membranes,Biochimica et Biophysica Acta 266 (1972) 695–707.[91] F. Stetzkowski-Marden, K. Gaus, M. Recouvreur, A. Cartaud,J. Cartaud, Agrin elicits membrane lipid condensation at sitesof acetylcholine receptor clusters in c2c12 myotubes, Journalof lipid research 47 (2006) 2121–2133.[92] R. Willmann, S. Pun, L. Stallmach, G. Sadasivam, A. F. Santos,P. C. et al., Cholesterol and lipid microdomains stabilize thepostsynapse at the neuromuscular junction, The EMBO journal25 (2006) 4050–4060.[93] J. Campagna, J. Fallon, Lipid rafts are involved in C95 (4,8) agrin fragment-induced acetylcholine receptor clustering,Neuroscience 138 (2006) 123–132.[94] S. Pun, M. Sigrist, A. F. Santos, M. A. Ruegg, J. R. Sanes,T. M. J. et al., An intrinsic distinction in neuromuscular junc-tion assembly and maintenance in di ff erent skeletal muscles,Neuron 34 (2002) 357–370.[95] V. L. Perillo, D. A. Pe˜nalva, A. J. Vitale, F. J. Barrantes, S. S.Antollini, Transbilayer asymmetry and sphingomyelin compo-sition modulate the preferential membrane partitioning of thenicotinic acetylcholine receptor in Lo domains, Archives ofBiochemistry and Biophysics 591 (2016) 76–86.[96] K. Gupta, J. Li, I. Liko, J. Gault, C. Bechara, D. Wu, J. T. S.Hopper, K. Giles, J. L. P. Benesch, C. V. Robinson, Identifyingkey membrane protein lipid interactions using mass spectrom-etry., Nature protocols 13 (2018) 1106–1120. [97] D. S. Chorev, L. A. Baker, D. Wu, V. Beilsten-Edmands, S. L.Rouse, T. Zeev-Ben-Mordehai, C. Jiko, F. Samsudin, C. Gerle,S. Khalid, A. G. Stewart, S. J. Matthews, K. Grnewald, C. V.Robinson, Protein assemblies ejected directly from nativemembranes yield complexes for mass spectrometry., Science(New York, N.Y.) 362 (2018) 829–834.[98] G. Brannigan, J. Henin, R. Law, R. Eckenho ff , M. L. Klein,Embedded cholesterol in the nicotinic acetylcholine receptor,Proceedings of the National Academy of Sciences 105 (2008)14418–14423.[99] M. H. Cheng, Y. Xu, P. Tang, Anionic lipid and cholesterolinteractions with α β ffi ne De-formation of Lipids, Biophysical Journal 115 (2018) 117–128.[108] A. J. Sodt, M. L. Sandar, K. Gawrisch, R. W. Pastor, E. Lyman,The molecular structure of the liquid-ordered phase of lipidbilayers, Journal of the American Chemical Society 136 (2014)725–732.[109] S. J. Marrink, H. J. Risselada, S. Yefimov, D. P. Tieleman,A. H. de Vries, The martini force field:¢ coarse grained modelfor biomolecular simulations, The Journal of Physical Chem-istry B 111 (2007) 7812–7824. PMID: 17569554.[110] N. Unwin, Refined structure of the nicotinic acetylcholine re-ceptor at 4 Å resolution, Journal of Molecular Biology 346(2005) 967–989.[111] H. Berendsen, D. van der Spoel, R. van Drunen, Gromacs: Amessage-passing parallel molecular dynamics implementation,Computer Physics Communications 91 (1995) 43 – 56.[112] X. Periole, M. Cavalli, S. J. Marrink, M. A. Ceruso, Combiningan elastic network with a coarse-grained molecular force field:Structure, dynamics, and intermolecular recognition, Journalof Chemical Theory and Computation 5 (2009) 2531–2543.[113] T. Fong, M. McNamee, Correlation between acetylcholine re-ceptor function and structural properties of membranes, Bio-chemistry 25 (1986) 830–840.[114] D. Marsh, Liquid-ordered phases induced by cholesterol: a ompendium of binary phase diagrams, Biochimica et Bio-physica Acta (BBA)-Biomembranes 1798 (2010) 688–699.[115] K. R. Levental, J. H. Lorent, X. Lin, A. D. Skinkle, M. A.Surma, E. A. Stockenbojer, A. A. Gorfe, I. Levental, Polyun-saturated lipids regulate membrane domain stability by tuningmembrane order, Biophysical Journal 110 (2016) 1800–1810.[116] N. Unwin, Segregation of lipids near acetylcholine-receptorchannels imaged by cryo-EM, IUCrJ 4 (2017) 393–399.[117] A. Miyazawa, Y. Fujiyoshi, N. Unwin, Structure and gatingmechanism of the acetylcholine receptor pore, Nature 423(2003) 949–955.[118] F. J. Barrantes, Cell-surface translational dynamics of nicotinicacetylcholine receptors, 2014.[119] H. A. Rodrigues, R. F. Lima, M. D. C. Fonseca, E. A. Ama-ral, P. M. Martinelli, L. A. Naves, M. V. Gomez, C. Kushm-erick, M. A. Prado, C. Guatimosim, Membrane cholesterolregulates di ff erent modes of synaptic vesicle release and re-trieval at the frog neuromuscular junction, European Journalof Neuroscience 38 (2013) 2978–2987.[120] D. A. Brown, Analysis of raft a ffi nity of membrane proteins bydetergent-insolubility., Methods in molecular biology (Clifton,N.J.) 398 (2007) 9–20.[121] H. Aranda-Espinoza, a. Berman, N. Dan, P. Pincus, S. A.Safran, Interaction between inclusions embedded in mem-branes, Biophysical journal 71 (1996) 648–656.[122] M. Ø. Jensen, O. G. Mouritsen, Lipids do influence pro-tein function-the hydrophobic matching hypothesis revisited.,Biochimica et biophysica acta 1666 (2004) 205–226.[123] G. Brannigan, F. L. H. Brown, A consistent model for thermalfluctuations and protein-induced deformations in lipid bilay-ers., Biophys. J. 90 (2006) 1501–1520.[124] M. Goulian, R. Bruinsma, P. Pincus, Long-range forces in het-erogeneous fluid membranes, EPL (Europhysics Letters) 22(1993) 145.[125] C. Sunshine, M. McNamee, Lipid modulation of nicotinicacetylcholine receptor function: the role of neutral and neg-atively charged lipids., Biochim Biophys Acta 1108 (1992)240–6.[126] A. K. Hamouda, M. Sanghvi, D. Sauls, T. K. Machu, M. P.Blanton, Assessing the lipid requirements of the Torpedocalifornica nicotinic acetylcholine receptor, Biochemistry 45(2006) 4327–4337.[127] D. Butler, M. McNamee, FTIR analysis of nicotinic acetyl-choline receptor secondary structure in reconstituted mem-branes, Biochimica et Biophysica Acta - Biomembranes(1993).[128] A. Bhushan, M. McNamee, Correlation of phospholipid struc-ture with functional e ff ects on the nicotinic acetylcholine re-ceptor. a modulatory role for phosphatidic acid, BiophysicalJournal (1993).[129] T. Fong, M. McNamee, Stabilization of acetylcholine recep-tor secondary structure by cholesterol and negatively chargedphospholipids in membranes, Biochemistry (1987).[130] P. Bednarczyk, A. Szewczyk, K. Dolowy, Transmembrane seg-ment m2 of glycine receptor as a model system for the pore-forming structure of ion channels, Acta Biochimica Polonica(2002).[131] C. J. B. Dacosta, A. A. Ogrel, E. A. McCardy, M. P. Blan-ton, J. E. Baenziger, Lipid-protein interactions at the nicotinicacetylcholine receptor. A functional coupling between nico-tinic receptors and phosphatidic acid-containing lipid bilayers,Journal of Biological Chemistry 277 (2002) 201–208.[132] W. Humphrey, A. Dalke, K. Schulten, VMD – Visual Molecu-lar Dynamics, Journal of Molecular Graphics 14 (1996) 33–38.[133] J. Stone, An E ffi cient Library for Parallel Ray Tracing and An- imation , Master’s thesis, Computer Science Department, Uni-versity of Missouri-Rolla, 1998., Master’s thesis, Computer Science Department, Uni-versity of Missouri-Rolla, 1998.