Phase separation of integral membrane proteins in Triton X-114 solution

A solution of the nonionic detergent Triton X-114 is homogeneous at 0 degrees C but separates in an aqueous phase and a detergent phase above 20 degrees C. The extent of this detergent phase separation increases with the temperature and is sensitive to the presence of other surfactants. The partition of proteins during phase separation in solutions of Triton X-114 is investigated. Hydrophilic proteins are found exclusively in the aqueous phase, and integral membrane proteins with an amphiphilic nature are recovered in the detergent phase. Triton X-114 is used to solubilize membranes and whole cells, and the soluble material is submitted to phase separation. Integral membrane proteins can thus be separated from hydrophilic proteins and identified as such in crude membrane or cellular detergent extracts.     

A Mild Purification Method for Polysaccharide Binding Membrane Proteins: Phase Separation of Digitonin Extracts to Isolate the Hyaluronate Synthase fromStreptococcussp. in Active Form

A new method was developed to purify the streptococcal hyaluronate synthase in active form to electrophoretic homogeneity. The method is based on the extraction of protoplast membranes with digitonin and a phase separation into an aqueous and a detergent phase induced by addition of polyethylene glycol 6000 at 0°C. Proteins bound to hyaluronate were enriched in the aqueous phase, whereas other membrane proteins resided in the detergent phase. Final purification of the hyaluronate synthase was achieved by ion exchange chromatography.     

Phase separation in the isolation and purification of membrane proteins

Phase separation is a simple, efficient, and cheap method to purify and concentrate detergent-solubilized membrane proteins. In spite of this, phase separation is not widely used or even known among membrane protein scientists, and ready-to-use protocols are available for only relatively few detergent/membrane protein combinations. Here, we summarize the physical and chemical parameters that influence the phase separation behavior of detergents commonly used for membrane protein studies. Examples for the successful purification of membrane proteins using this method with different classes of detergents are provided. As the choice of the detergent is critical in many downstream applications (e.g., membrane protein crystallization or functional assays), we discuss how new phase separation protocols can be developed for a given detergent buffer system.     

Aminoacylase I is not a glycolipid-anchored ectoenzyme in pig kidney.

Subcellular fractionation of pig kidney cortex revealed that aminoacylase I (EC 3.5.1.14, N-acyl-L-amino-acid aminohydrolase) is predominantly a soluble enzyme with only 0.5% of the total activity being recovered in the membrane fraction. The aminoacylase I activity associated with the membrane preparations displayed neither rapid release following incubation with phosphatidylinositol-specific phospholipase C from Bacillus thuringiensis nor the distinctive differential pattern of detergent solubilization which was seen with glycosyl-phosphatidylinositol-anchored proteins (renal dipeptidase, alkaline phosphatase). When fractionated by phase separation in Triton X-114, integral membrane proteins of kidney microvillar membranes partitioned predominantly (greater than 90%) into the detergent-rich phase. In contrast, only 3.7% of aminoacylase I activity associated with microvillar membranes partitioned into the detergent-rich phase. Aminoacylase I activity of pig kidney would therefore appear to be a hydrophilic protein in nature and is not, as suggested previously, a G-PI-anchored integral membrane protein.     

Mechanisms of phase behaviour and protein partitioning in detergent/polymer aqueous two-phase systems for purification of integral membrane proteins

Detergent/polymer aqueous two-phase systems are studied as a fast, mild and efficient general separation method for isolation of labile integral membrane proteins. Mechanisms for phase behaviour and protein partitioning of both membrane-bound and hydrophilic proteins have been examined in a large number of detergent/polymer aqueous two-phase systems. Non-ionic detergents such as the Triton series (polyoxyethylene alkyl phenols), alkyl polyoxyethylene ethers (C(m)EO(n)), Tween series (polyoxyethylene sorbitol esters) and alkylglucosides form aqueous two-phase systems in mixtures with hydrophilic polymers, such as PEG or dextran, at low and moderate temperatures. Phase diagrams for these mixtures are shown and phase behaviour is discussed from a thermodynamic model. Membrane proteins, such as bacteriorhodopsin and cholesterol oxidase, were partitioned strongly to the micelle phase, while hydrophilic proteins, BSA and lysozyme, were partitioned to the polymer phase. The partitioning of membrane protein is mainly determined by non-specific hydrophobic interactions between detergent and membrane protein. An increased partitioning of membrane proteins to the micelle phase was found with an increased detergent concentration difference between the phases, lower polymer molecular weight and increased micelle size. Partitioning of hydrophilic proteins is mainly related to excluded volume effects, i.e. increased phase component size made the hydrophilic proteins partition more to the opposite phase. Addition of ionic detergent to the system changed the partitioning of membrane proteins slightly, but had a strong effect on hydrophilic proteins, and can be used for enhanced separation between hydrophilic proteins and membrane protein.     

Vaccinia Virus Virion Membrane Biogenesis Protein A11 Associates with Viral Membranes in a Manner That Requires the Expression of Another Membrane Biogenesis Protein,A6

A group of vaccinia virus (VACV) proteins, including A11, L2, and A6, are required for biogenesis of the primary envelope of VACV, specifically, for the acquisition of viral membrane precursors. However, the interconnection among these proteins is unknown and, with the exception of L2, the connection of these proteins with membranes is also unknown. In this study, prompted by the findings that A6 coprecipitated A11 and that the cellular distribution of A11 was dramatically altered by repression of A6 expression, we studied the localization of A11 in cells by using immunofluorescence and cell fractionation analysis. A11 was found to associate with membranes and colocalize with virion membrane proteins in viral replication factories during normal VACV replication. A11 partitioned almost equally between the detergent and aqueous phases upon Triton X-114 phase separation, demonstrating an intrinsic affinity with lipids. However, in the absence of infection or VACV late protein synthesis, A11 did not associate with cellular membranes. Furthermore, when A6 expression was repressed, A11 did not colocalize with any viral membrane proteins or associate with membranes. In contrast, when virion envelope formation was blocked at a later step by repression of A14 expression or by rifampin treatment, A11 colocalized with virion membrane proteins in the factories. Altogether, our data showed that A11 associates with viral membranes during VACV replication, and this association requires A6 expression. This study provides a physical connection between A11 and viral membranes and suggests that A6 regulates A11 membrane association.     

A new method for the isolation of renal basement membranes.

A method is described for the isolation of basement membranes from rabbit renal cortex in which the detergent N-lauroyl sarcosine is used as the disruptive agent. The isolated membranes have been compared with membranes prepared using ultrasonication and they were comparable both in terms of purity and gross chemical composition. Glomerular and tubular basement membranes were isolated by first separating glomeruli from tubules by density gradient centrifugation followed by detergent treatment of the separated tissues. The detergent method has the advantage that the basement membranes retained their native structure to a large degree, whereas sonicated membranes were severely fragmented. Collagen fibres were a significant contaminant in both preparations and were revealed more clearly by negative staining than by examination of thin sections. Studies with the detergent-treated membrane revealed that a few proteins, which seemed to be membrane components, were extracted with 1 M NaCl and that these proteins were lost from the basement membranes during sonication used in the conventional isolation procedure.     

Cholesterol sensitivity of detergent resistance: a rapid flow cytometric test for detecting constitutive or induced raft association of membrane proteins.

BACKGROUND: Lipid rafts are cholesterol- and glycosphingolipid-rich microdomains in the cellular plasma membranes that play critical roles in compartmentalization (concentration, coupling, and isolation) of receptors and signal molecules. Therefore, detecting constitutive or induced raft associations of such proteins is of central interest in cell biology. This has mostly been done with time- and cell-consuming immunobiochemical techniques affected by several sources of artifacts. A flow cytometric analysis of immunocytochemical staining under differential circumstances of detergent treatment offers a new alternative to this method. METHODS: Membrane microdomains are resistant to nonionic detergents due to extensive, strong interactions between their molecular constituents. We used this feature to develop a rapid flow cytometric assay of differential detergent resistance based on immunocytochemical labeling of extracellular domain epitopes in membrane proteins. Data evaluation is based on comparative detection of their detergent solubility without and with cholesterol depletion of cell membranes, resolved by moderate concentrations of nonionic detergents. RESULTS: Nonionic detergents Triton X-100 and Nonidet-40 (0.05-0.1%) in cold or Brij-98 (0.1-0.5%) at 37 degrees C efficiently resolved detergent solubility or resistance of many lymphocyte cell surface proteins. Kinetic data revealed that a short (5-10 min) detergent treatment is sufficient for this assay. Comparison of detergent solubility in untreated and cholesterol-depleted cells differentiated membrane proteins associated with or excluded from raft microdomains, respectively. Confocal microscopy showed that this mild detergent treatment leaves the cytoskeleton of the cells intact, with a detectable expression of raft marker detergent-resistant proteins attached to it. An induced association with rafts of immunoglobulin E receptors upon antigen cross-linking was also easily detectable in rat mast cells by this approach. CONCLUSIONS: A protocol is proposed for a rapid (5-10 min) test of detergent resistance of membrane proteins in cells. The approach requires only a small amount of cells (10(4)/sample) and offers a good resolution of detergent solubility or resistance of membrane proteins, also in terms of the underlying mechanisms, with an advantage of applicability for all conventional bench-top flow cytometers.     

Tyrosine hydroxylase in secretory granules from bovine adrenal medulla. Evidence for an integral membrane form

Intact secretory granules isolated from bovine adrenal medulla express tyrosine hydroxylase (TH) activity. Granule-associated TH sediments on continuous sucrose gradients with dopamine beta-hydroxylase, a marker for granule membranes, indicating that TH is associated with chromaffin granules. Membranes prepared from lysed granules retain TH, whereas granule contents are free of the enzyme. TH immunoreactivity was detected in granule membranes by immunoblot analysis using a polyclonal antiserum against TH. TH immunoreactivity cannot be removed from membranes by washes in high ionic strength buffers and is only partially removed from membranes by treatment with either urea or Na2CO3. TH can be removed from granule membranes by the detergents Nonidet P-40, Triton X-100, and 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate. Treatment of membranes with a phosphatidylinositol-specific phospholipase C did not remove TH, ruling out the possibility of a glycosyl phosphatidyl anchor. Fractionation of granule membranes by temperature-induced phase separation in Triton X-114 revealed that TH is recovered in phases in which integral (detergent phase) and hydrophobic (phospholipid phase) membrane proteins are typically found. By contrast, TH from adrenal cytosol fractionated exclusively into the aqueous phase along with other soluble proteins. Digestion of granules with various protease enzymes revealed that TH is resistant to degradation, suggesting that the enzyme is embedded within membranes. TH becomes phosphorylated when intact granules are exposed to the catalytic subunit of the cAMP-dependent protein kinase, indicating that at least the N-terminal region of TH is exposed on the cytoplasmic surface of granules. These results establish that a fraction of TH is an integral component of bovine granule membranes. The association of TH with granule membranes may play a role in coordinating TH activity and catecholamine release.     

Functional reconstitution of the HIV receptors CCR5 and CD4 in liposomes.

Reconstitution of membrane proteins allows their study in a membrane environment that can be manipulated at will. Because membrane proteins have diverse biophysical properties, reconstitution methods have so far been developed for individual proteins on an ad hoc basis. We developed a postinsertion reconstitution method for CCR5, a G protein coupled receptor, with seven transmembrane alpha helices and small ecto- and endodomains. A His6-tagged version of CCR5 was expressed in mammalian cells, purified using the detergent N-dodecyl-beta-d-maltoside (DDM) and reconstituted into preformed liposomal membranes saturated with DDM, removing the detergent with hydrophobic polystyrene beads. We then attempted to incorporate CD4, a protein with a single transmembrane helix and a large hydrophilic ectodomain into liposomal membranes, together with CCR5. Surprisingly, reconstitution of this protein was also achieved by the method. Both proteins were found to be present together in individual liposomes. The reconstituted CCR5 was recognized by several monoclonal antibodies, recognized its natural ligand, and CD4 bound a soluble form of gp120, a subunit of the HIV fusion protein that uses CD4 as a receptor. Moreover, cells expressing the entire fusion protein of HIV bound to the liposomes, indicating that the proteins were intact and that most of them were oriented right side out. Thus, functional coreconstitution of two widely different proteins can be achieved by this method, suggesting that it might be useful for other proteins.     

Transfer of cytochrome b 5 and NADPH cytochrome c reductase between membranes

NADPH-cytochrome c reductase also reduces cytochrome b 5. The reduction is very slow when the proteins are in solution or bound to different membranes. Only when both proteins share a common membrane, is cytochrome b 5 reduced rapidly by NADPH. The difference in reaction rates indicates recombination on a common membrane of cytochrome b 5 and NADPH reductase originally bound to different vesicles. The recombination of the two proteins occurs with a variety of biological membranes (previously enriched with either reductase or cytochrome b 5) as well as with liposomes. We explain this process as protein transfer rather than vesicle fusion for several reasons: 1. The vesicles do not alter shape or size during incubation. 2. The rate of this process corresponds to the rate of incorporation of the single proteins into liposomes carrying the 'complementary' protein. 3. The exchange of proteins between biological membranes and liposomes occupied by protein does not change the density of either membrane. Protein transfer between membranes appears to be limited to those proteins which had spontaneously recombined with a preformed membrane. In contrast, proteins incorporated into liposomes by means of a detergent were not transferred, nor were endogenous cytochrome b 5 and NADPH-cytochrome c reductase transferred from microsomes to Golgi membranes or lipid vesicles. We conclude that the endogenous proteins and proteins incorporated in the presence of a detergent are linked to the membrane in another manner than the same proteins which had been inserted into a preformed membrane.     

L-DOPA decarboxylase association with membranes in mouse brain

This work presents evidence on the association of active DDC molecules with membranes in mammalian brain. L-DOPA decarboxylase (DDC) is generally considered to be a cytosolic enzyme. Membrane-associated DDC was detected by immunoblotting and enzymatic assay experiments. DDC activity and immunoreactivity could be partially extracted from mammalian brain membranes by detergent. Fractionation of membranes by temperature-induced phase separation in Triton X-114, resulted in the recovery of membrane-associated DDC in separation phases where integral and hydrophobic membrane proteins separate. Treatment of membranes with phosphatidylinositol-specific phospholipase C or proteinase K, did not elute membrane-associated DDC activity, suggesting that a population of DDC molecules exist embedded within membranes. The elucidation of the functional significance of the enzyme's association with membranes could provide us with new information leading to the better understanding of the biological pathways that DDC is involved in.     

Phase separation of rat intestinal brush border membrane proteins using Triton X-114

Rat intestinal microvillus membrane contains at least 24 polypeptides, of which 18 can be solubilized using Triton X-114 at 4 degrees C. Upon phase separation at 32 degrees C, 11 proteins separated nearly completely into the detergent-rich phase, while 9 proteins were found exclusively in the aqueous phase. Enzymes which were uniquely included in the detergent phase were alkaline phosphatase, leucine aminopeptidase, gamma-glutamyl transpeptidase, and Ca2+-Mg2+ ATPase. The proteins which were excluded from the detergent phase and found exclusively in the aqueous phase included the disaccharidases (glucoamylase, sucrase-isomaltase, trehalase, lactase) and the ileal receptor for the intrinsic factor-cobalamin complex. Integral membrane proteins can thus be separated during solubilization into two groups prior to further purification or characterization.     

Comparative lipid analysis and structure of detergent-resistant membrane raft fractions isolated from human and ruminant erythrocytes

The lipid composition and structure of detergent-resistant membrane rafts from human, goat, and sheep erythrocytes is investigated. While the sphingomyelin:cholesterol ratio varied from about 1:5 in human to 1:1 in sheep erythrocytes a ratio of 1:1 was found in all raft preparations insoluble in Triton X-100 at 4 degrees C. Excess cholesterol is excluded from rafts and saturated molecular species of sphingomyelin assayed by gas chromatography-mass spectrometry determines the solubility of cholesterol in the detergent. Freeze-fracture electron microscopy shows that vesicles and multilamellar structures formed by membrane rafts have undergone considerable rearrangement from the original membrane. No membrane-associated particles are observed. Synchrotron X-ray diffraction studies showed that d spacings of vesicle preparations of rafts cannot be distinguished from ghost membranes from which they are derived. Dispersions of total polar lipid extracts of sheep rafts show phase separation of inverted hexagonal structure upon heating and this phase coexists with multilamellar structures at 37 degrees C.     

Determination and application of empirically derived detergent phase boundaries to effectively crystallize membrane proteins

Elucidating the structures of membrane proteins is essential to our understanding of disease states and a critical component in the rational design of drugs. Structural characterization of a membrane protein begins with its detergent solubilization from the lipid bilayer and its purification within a functionally stable protein‐detergent complex (PDC). Crystallization of the PDC typically occurs by changing the solution environment to decrease solubility and promote interactions between exposed hydrophilic surface residues. As membrane proteins have been observed to form crystals close to the phase separation boundaries of the detergent used to form the PDC, knowledge of these boundaries under different chemical conditions provides a foundation to rationally design crystallization screens. We have carried out dye‐based detergent phase partitioning studies using different combinations of 10 polyethylene glycols (PEG), 11 salts, and 11 detergents to generate a significant amount of chemically diverse phase boundary data. The resulting curves were used to guide the formulation of a 1536‐cocktail crystallization screen for membrane proteins. We are making both the experimentally derived phase boundary data and the 1536 membrane screen available through the high‐throughput crystallization facility located at the Hauptman‐Woodward Institute. The phase boundary data have been packaged into an interactive Excel spreadsheet that allows investigators to formulate grid screens near a given phase boundary for a particular detergent. The 1536 membrane screen has been applied to 12 membrane proteins of unknown structures supplied by the structural genomics and structural biology communities, with crystallization leads for 10/12 samples and verification of one crystal using X‐ray diffraction.     

Separation of membrane proteins according to their hydropathy by serial phase partitioning with Triton X-114

The detergent Triton X-114, because of its convenient cloud point temperature (22 °C), has been used extensively to extract membrane proteins and to separate them in two phases according to their hydropathy. The upper detergent-poor phase contains mostly hydrophilic proteins, whereas hydrophobic ones are found mainly in the lower detergent-rich phase. In this work, we developed a method to fractionate membrane proteins and estimate their hydropathy based on a series of cloud point partitions with Triton X-114. With this method, beetroot plasma membrane proteins were separated in different fractions according to their hydropathy, following the binomial distribution law as expected. This method revealed the presence of both hydrophilic and hydrophobic Ca²⁺-dependent protein kinases in those membranes. At least five distinct Ca²⁺-dependent kinases were observed in in-gel kinase activity assays. This separation procedure was also used as the first step in the purification of a hydrophobic 60-kDa kinase.     

Sorting of Lens Aquaporins and Connexins into Raft and Nonraft Bilayers: Role of Protein Homo-Oligomerization

Two classes of channel-forming proteins in the eye lens, the water channel aquaporin-0 (AQP-0) and the connexins Cx46 and Cx50, are preferentially located in different regions of lens plasma membranes ( [1] and [2]). Because these membranes contain high concentrations of cholesterol and sphingomyelin, as well as phospholipids such as phosphatidylcholine with unsaturated hydrocarbon chains, microdomains (rafts) form in these membranes. Here we test the hypothesis that sorting into lipid microdomains can play a role in the disposition of AQP-0 and the connexins in the plane of the membrane. For both crude membrane fractions and proteoliposomes composed of lens proteins in phosphatidylcholine/sphingomyelin/cholesterol lipid bilayers, detergent extraction experiments showed that the connexins were located primarily in detergent soluble membrane (DSM) fractions, whereas AQP-0 was found in both detergent resistant membrane and DSM fractions. Analysis of purified AQP-0 reconstituted in raft-containing bilayers showed that the microdomain location of AQP-0 depended on protein/lipid ratio. AQP-0 was located almost exclusively in DSMs at a 1:1200 AQP-0/lipid ratio, whereas ∼50% of the protein was sequestered into detergent resistant membranes at a 1:100 ratio, where freeze-fracture experiments show that AQP-0 oligomerizes (3). Consistent with these detergent extraction results, confocal microscopy images showed that AQP-0 was sequestered into raft microdomains in the 1:100 protein/lipid membranes. Taken together these results indicate that AQP-0 and connexins can be segregated in the membrane by protein-lipid interactions as modified by AQP-0 homo-oligomerization.     

Ciliary membrane tubulin and associated proteins: a complex stable to Triton X-114 dissociation

When either membranes from scallop gill cilia or reconstituted membranes from the same source are solubilized with Triton X-114 and the detergent is condensed by warming, no significant fraction of any major membrane protein partitions into the micellar detergent. Rather, most of the membrane lipids condense with the detergent phase, forming mixed micelles from which nearly pure lipid vesicles may be produced by adsorption of detergent with polystyrene beads. One minor membrane protein, with a molecular weight of about 20 000, is associated consistently with these vesicles. The aqueous phase contains a fairly homogeneous protein-Triton X-114 micelle sedimenting at 2.6 S in the analytical ultracentrifuge. Sucrose gradient velocity analysis in a detergent-free gradient indicates moderate size polydispersity but constant polypeptide composition throughout the sedimenting protein zone. Sucrose gradient equilibrium analysis (also in a detergent-free gradient) results in a protein-detergent complex banding at a density of 1.245 g/cm3. Sedimentation of the protein-detergent complex in the ultracentrifuge, followed by fixation and normal processing for electron microscopy, reveals a fine, reticular material consisting of 5-10-nm granules. These data are consistent with previous evidence that membrane tubulin and most other membrane proteins exist together as a discrete lipid-protein complex in molluscan gill ciliary membranes.     

GUV preparation and imaging: Minimizing artifacts

The components of biological membranes are present in a physical mixture. The nonrandom ways that the molecules of lipids and proteins mix together can strongly influence the association of proteins with each other, and the chemical reactions that occur in the membrane, or that are mediated by the membrane. A particular type of nonrandom mixing is the separation of compositionally distinct phases. Any such phase separation would result in preferential partition of some proteins and lipids between the coexisting phases, and thus would influence which proteins could be in contact, and whether a protein could find its target. Phase separation in a plasma membrane would also influence the binding of molecules from outside the cell to the membrane, including recognition proteins on viruses, bacteria, and other cells. The concept of these and other events associated with membrane phase separation are sometimes grouped together as the “raft model” of biological membranes. Several types of experiments are aimed at detecting and characterizing membrane phase separation. Visualizing phase separation has special value, both because the immiscibility is so decisively determined, and also because the type of phase can often be identified. The fluorescence microscope has proven uniquely useful for yielding images of separated phases, both in certain cell preparations, and especially in models of cell membranes. Here we discuss ways to prepare useful model membranes for image studies, and how to avoid some of the artifacts that can plague these studies.     

Purification of proctolin-binding proteins from the foregut of the insect Blaberus craniifer.

A membrane protein that specifically binds the insect neuropeptide proctolin was purified using standard chromatography from cockroach foregut membranes. Proctolin-binding sites were efficiently solubilized with either the nonionic detergent digitonin or the zwitterionic detergent Chaps, as indicated by the specific binding of 3H-proctolin to solubilized samples. A solubilized sample obtained from 1600 foregut membranes was subjected to a five-step chromatographic purification including chromatofocusing, anion-exchange and size-exclusion chromatographies. The final size-exclusion separation resulted in the isolation of approximately 100 pmol of purified proctolin-binding proteins, eluting as a single peak at approximately 74 kDa. Analysis of the purified sample using SDS/PAGE and silver staining showed two bands at 80 kDa and 76 kDa. Densitometric analysis of the gel indicated that each band contained approximately 7-8 microg of protein, suggesting that one band corresponds to the proctolin-binding activity. Proctolin-binding proteins were thus purified 1800-fold using standard chromatography.