Myosin-induced movement of alphaalpha, alphabeta, and betabeta smooth muscle tropomyosin on actin observed by multisite FRET

The interaction of the alphaalpha, betabeta, and alphabeta smooth muscle tropomyosin (Tm) isoforms with F-actin was systematically studied in the absence and in the presence of myosin subfragment 1 (S1) using multifrequency phase/modulation F?rster resonance energy transfer (FRET). A Gaussian double distance distribution model was adopted to fit FRET data between a 5-(2-iodoacetyl-amino-ethyl-amino)naphthalene-1-sulfonic acid donor at either Cys-36 of the beta-chain or Cys-190 of the alpha-chain and a 4-dimethylaminophenylazophenyl 4'-maleimide acceptor at Cys-374 of F-actin. Experimental data were obtained for singly and doubly labeled alphabeta Tm (donor only at alpha, only at beta, or both) and for doubly labeled alphaalpha or betabeta Tm. Data for singly labeled alphabetaTm were combined in a global analysis with doubly labeled alphabetaTm. In all doubly labeled isoforms, upon S1 binding, one donor-acceptor "apparent" distance increased slightly by 0.5-2 A, whereas the other decreased by 6-9 A. These changes are consistent with a uniform "rolling" motion of Tm over the F-actin surface. The analysis indicates that Tm occupies relatively well-defined positions, with some flexibility, in both the predominantly closed (-S1) and open (+S1) thin-filament states. The results for the alphabetaTm heterodimer indicate that the local twofold symmetry of alphaalpha or betabeta Tm is effectively broken in alphabetaTm bound to F-actin, which implies a difference between the alpha- and beta-chains in terms of their interaction with F-actin.     

CD8 Raft localization is induced by its assembly into CD8alpha beta heterodimers, Not CD8alpha alpha homodimers

The coreceptor CD8 is expressed as a CD8alphabeta heterodimer on major histocompatibility complex class I-restricted TCRalphabeta T cells, and as a CD8alphaalpha homodimer on subsets of memory T cells, intraepithelial lymphocytes, natural killer cells, and dendritic cells. Although the role of CD8alphaalpha is not well understood, it is increasingly clear that this protein is not a functional homologue of CD8alphabeta. On major histocompatibility complex class I-restricted T cells, CD8alphabeta is a more efficient TCR coreceptor than CD8alphaalpha. This property has for the mouse protein been attributed to the recruitment of CD8alphabeta into lipid rafts, which is dependent on CD8beta palmitoylation. Here, these divergent distributions of CD8alphabeta and CD8alphaalpha are demonstrated for the human CD8 proteins as well. However, although palmitoylation of both CD8alpha and CD8beta chains was detected, this modification did not contribute to raft localization. In contrast, arginines in the cytoplasmic domain are crucial for raft localization of CD8betabeta. Most strikingly, the assembly of a non-raft localized CD8beta chain with a non-raft localized CD8alpha chain resulted in raft-localized CD8alphabeta heterodimers. Using chimeric CD8 proteins, this property of the heterodimer was found to be determined by the assembly of CD8alpha and CD8beta extracellular regions. The presence of two CD8alpha extracellular regions, on the other hand, appears to preclude raft localization. Thus, heterodimer formation and raft association are intimately linked for CD8alphabeta. These results emphasize that lipid raft localization is a key feature of human CD8alphabeta that clearly distinguishes it from CD8alphaalpha.     

The X-ray crystal structure of human beta-hexosaminidase B provides new insights into Sandhoff disease

Human lysosomal beta-hexosaminidases are dimeric enzymes composed of alpha and beta-chains, encoded by the genes HEXA and HEXB. They occur in three isoforms, the homodimeric hexosaminidases B (betabeta) and S (alphaalpha), and the heterodimeric hexosaminidase A (alphabeta), where dimerization is required for catalytic activity. Allelic variations in the HEXA and HEXB genes cause the fatal inborn errors of metabolism Tay-Sachs disease and Sandhoff disease, respectively. Here, we present the crystal structure of a complex of human beta-hexosaminidase B with a transition state analogue inhibitor at 2.3A resolution (pdb 1o7a). On the basis of this structure and previous studies on related enzymes, a retaining double-displacement mechanism for glycosyl hydrolysis by beta-hexosaminidase B is proposed. In the dimer structure, which is derived from an analysis of crystal packing, most of the mutations causing late-onset Sandhoff disease reside near the dimer interface and are proposed to interfere with correct dimer formation. The structure reported here is a valid template also for the dimeric structures of beta-hexosaminidase A and S.     

Production of heterogeneous dimer lignostilbenedioxygenase II from lsdA and lsdB in Escherichia coli cells

The lignostilbenedioxygenase isozyme I and III genes, lsdA and lsdB, were coexpressed within the same Escherichia coli cells. The lignostilbenedioxygenase isozymes produced were separated on QAE anion-exchange column chromatography. Three parts of active fractions corresponding to alphaalpha, alphabeta, and betabeta dimers were detected. The purified isozyme from second active fractions corresponded to the native heterogeneous isozyme II.     

Crystal structure of the TCR co-receptor CD8alphaalpha in complex with monoclonal antibody YTS 105.18 Fab fragment at 2.88 A resolution

The CD8 glycoprotein functions as an essential element in the control of T-cell selection, maturation and the TCR-mediated response to peptide antigen. CD8 is expressed as both heterodimeric CD8alphabeta and homodimeric CD8alphaalpha isoforms, which have distinct physiological roles and exhibit tissue-specific expression patterns. CD8alphaalpha has previously been crystallized in complex with class I pMHC and, more recently, with the mouse class Ib thymic leukemia antigen (TL). Here, we present the crystal structure of a soluble form of mouse CD8alphaalpha in complex with rat monoclonal antibody YTS 105.18 Fab fragment at 2.88 A resolution. YTS 105.18, which is commonly used in the blockade of CD8+ T-cell activation in response to peptide antigen, is specific for mouse CD8alpha. The YTS 105.18 Fab is one of only five rat IgG Fab structures to have been reported to date. Analysis of the YTS 105.18 Fab epitope on CD8alpha reveals that this antibody blocks CD8 activity by hydrogen bonding to residues that are critical for interaction with both class I pMHC and TL. Structural comparison of the liganded and unliganded forms of soluble CD8alphaalpha indicates that the mouse CD8alphaalpha immunoglobulin-domain dimer does not undergo significant structural alteration upon interaction either with class I pMHC or TL.     

Interactions of enolase isoforms with tubulin and microtubules during myogenesis

Enolase is a glycolytic enzyme, expressed as cell-type specific isoforms in higher vertebrates. Herein we demonstrated for the first time that enolase isoforms interact with microtubules during muscle satellite cell differentiation. While in undifferentiated myoblasts the ubiquitous alphaalpha enolase isoform, expressed at high level, exhibited extensive co-localization with microtubules, the muscle-specific betabeta isoform, expressed at low level, did not. During differentiation, the level of beta subunit increased significantly; the alpha and beta enolase immunoreactivities were detected both in cytosol and along the microtubules. We identified tubulin from muscle extract as an interacting protein for immobilized betabeta enolase. ELISA and surface plasmon resonance measurements demonstrated the direct binding of enolase isoforms to tubulin with an apparent KD below the micromolar range, and indicated that the presence of 0.8 mM 2-phosphoglycerate abolished the interaction. Our data showed that, at various stages of myogenic differentiation, microtubules were decorated by different enolase isoforms, which was controlled by the abundance of both partners. We suggest that the binding of enolase to microtubules could contribute to the regulation of the dynamism of the cytoskeletal filaments known to occur during the transition from myoblast to myotubes.     

Intramolecular disulfide linked alphabeta and alphaalpha in oxidized tropomyosin: separation, identification, and process of formation

The oxidation of rabbit skeletal tropomyosin (TM) by repeated cycles of freezing and melting in 0.3 mM Na bicarbonate was studied by electrophoresis and column chromatography. The oxidized TM showed two bands at ca. 70,000 daltons on sodium dodecyl sulfate (SDS)-polyacrylamide gels. Each band component was separated into disulfide-linked alphabeta and alphaalpha by carboxymethyl cellulose (CMC) column chromatography in urea. Oxidized TM before fractionation, as well as the alphabeta and alphaalpha components, was found to have a molecular weight of about 80,000 daltons, indicating the disulfide bonds to be primarily intramolecular. Oxidation of dilute TM in 1 M NaCl by exposure to air also produced disulfide-linked alphabeta. Partially oxidized TM was found to separate into beta, alphabeta, alpha, and alphaalpha on CMC chromatography, and these were eluted with a linear gradient of NaCl at molarities of ca. 0.09, 0.11, 0.12, and 0.14 M, respectively. The oxidation process was investigated by CMC chromatography, and a possible mechanism is presented. The alphabeta and alphaalpha components may exist as dominant component in TM in vitro rather than as a random mixture of two subunits. A splitting of the electrophoretic band of the alpha subunit into a doublet was observed.     

One interferon gamma receptor binds one interferon gamma dimer

We investigated the stoichiometry of the interferon gamma and interferon gamma receptor interaction, using recombinant interferon gamma and recombinant soluble interferon gamma receptor, applying chemical cross-linking and chromatographic techniques, and analyzing the resulting products in denaturing polyacrylamide gels. Interferon gamma cross-linked to itself produced a major band of an apparent molecular mass of 34 kDa, which suggests that it exists as a dimer in physiological buffer and which agrees with published data. Soluble interferon gamma receptor cross-linked to itself produced mainly a 28-kDa band, suggesting that the interferon gamma receptor exists as a monomer. Interferon gamma cross-linked to the soluble interferon gamma receptor resulted in the formation of two main products of apparent molecular masses of 60 and 44 kDa. The predominant 60-kDa band resulted from the cross-linking of one interferon gamma dimer (34 kDa) to one interferon gamma receptor molecule (27 kDa). The 44-kDa band was formed by the cross-linking of one interferon gamma molecule to one interferon gamma receptor. Kinetic studies showed that the cross-linking of interferon gamma dimer to the soluble receptor proceeds through the intermediate formed by cross-linking one molecule of the interferon gamma dimer to the receptor. Reducing and dissociating agents inhibited complex formation. When chromatographed on Sephadex G-100, interferon gamma was eluted as a protein of 34-kDa molecular mass, the soluble interferon gamma receptor as a protein of 40 kDa, and their mixture was eluted in one peak corresponding to an apparent molecular mass of 73 kDa. Sodium dodecyl sulfate-polyacrylamide gel analysis of the eluted mixture showed the presence of both interferon gamma and interferon gamma receptor at a ratio of 2:1. The found results suggest that the interferon gamma receptor binds interferon gamma as a dimer.     

Fibre-type distribution and subcellular localisation of alpha and beta enolase in mouse striated muscle

Enolase is a dimeric glycolytic enzyme exhibiting tissue specific isoforms. During ontogenesis, a transition occurs from the embryonic alphaalpha towards the specific alphabeta, and betabeta isoforms in striated muscle. Immunocytochemical analyses on transverse sections of adult mouse gastrocnemius muscle, allowed us to compare the expression of alpha and beta subunits to that of myosin heavy chain (MHC) isoforms. Levels of beta immunoreactivity followed the order IIB > IIX > IIA > I. This gradient parallels the ATPase activity associated to MHC isoforms, indicating that the expression of beta enolase in myofibres is finely regulated as a function of energetic requirements. By contrast, variations in alpha immunolabelling intensity appeared independent of fibre types. Longitudinal muscle sections exhibited a striated pattern of alpha immunoreactivity. Confocal microscopy analyses demonstrated that alpha was localised at the M band. Most beta immunoreactivity was diffuse all over the sarcoplasm. However, some beta immunoreactivity was striated and localized at both Z and M bands. Thus, betabeta enolase could participate to multi-enzyme complexes present at the I band, and involved with local ATP production. Our results support the notion that isozymes differ in their ability to interact with other macromolecules, thus segregating to different subcellular sites where they would respond to specific functional demands.     

A study of the quaternary structure of Escherichia coli RNA polymerase using bis(imido esters).

The quaternary structure of Escherichia coli RNA polymerase has been studied by cross-linking with a periodate-cleavable bis(imido ester), N,N'-bis(2-carboximidoethyl)tartaramide dimethyl ester dihydrochloride (CETD). The cross-linked holoenzyme gives a characteristic five-band pattern after electrophoresis on sodium dodecyl sulfate-polyacrylamide gels. The components of each band have been unambiguously identified by (a) molecular-weight measurements, (b) comparisons of cross-linking patterns of holoenzyme and core enzyme, and (c) periodate cleavage of cross-links followed by a second dimension sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The bands are (1) alphabeta and alphabeta', (2) sigmabeta and sigmabeta', (3) alphasigmabeta', (4) betabeta', and (5) sigmabetabeta'. Bands 2 and 4 are the most prominent at low reagent concentrations (up to 2.5 mM) but band 5 becomes the most prominent at higher concentrations. There are no bands corresponding to alphaalpha and alphasigma; a faint band has been tentatively identified as alphabetabeta'. Shorter bis(imido esters) are much less effective cross-linking reagents than CETD and they do not give rise to any other cross-linked species. On the basis of these observations, a model for the subunit arrangement of RNA polymerase is proposed.     

Cross-linking studies of the protein topography of rat liver microsomes

A cleavable cross-linking reagent, dithiobis(succinimidyl propionate), DSP, was used to study the topography of the proteins in the endoplasmic reticulum membrane of rat liver. Reaction of untreated (control), phenobarbital- or 3-methylcholanthrene-induced microsomes with 0.5 mM DSP for 30 min at 0°C resulted in the cross-linking of a protein with a molecular weight of about 52 000 to form an apparent dimer. In phenobarbital microsomes, a smaller amount of a 52 000-dalton protein also appeared in a dimer in the absence of DSP if N-ethylmaleimide was not included during homogenization. In phenobarbital and 3-methylcholanthrene microsomes, a 48 000-dalton protein was cross-linked by DSP to a protein of about 57 000. In all three types of microsomes, a protein with an M_I of about 52 000 was also cross-linked to a protein of about 79 000. In phenobarbital and control microsomes, cross-linking resulted in an oligomeric protein of approximate molecular weight 180 000 which contained three proteins, two with M_r of about 52 000 and one about 79 000. Under the cross-linking conditions, little or no denaturation of cytochrome P-450 and NADPH-cytochrome c reductase was observed. The aryl hydrocarbon hydroxylase activity was significantly inhibited by the bifunctional cross-linking reagent, DSP, but not by the monofunctional reagent N-succinimidyl-3-(4-hydroxyphenyl) propionate. However, attempts to regenerate the aryl hydrocarbon hydroxylase activity by cleavage of the disulfide linkage with 2-mercapto-ethanol or dithiothreitol were not successful.     

Subunit structure and properties of the glycogen-bound phosphoprotein phosphatase from skeletal muscle

A high molecular weight phosphoprotein phosphatase was purified approximately 11,000-fold from the glycogen-protein complex of rabbit skeletal muscle. Polyacrylamide gel electrophoresis of the preparation in the absence of sodium dodecyl sulfate showed a major protein band which contained the activity of the enzyme. Gel electrophoresis in the presence of sodium dodecyl sulfate also showed a major protein band migrating at 38,000 daltons. The sedimentation coefficient, Stokes radius, and frictional ratio of the enzyme were determined to be 4.4 S, 4.4 nm, and 1.53, respectively. Based on these values the molecular weight of the enzyme was calculated to be 83,000. The high molecular weight phosphatase was dissociated upon chromatography on a reactive red-120 agarose column. The sedimentation coefficient, Stokes radius, and frictional ratio of the dissociated enzyme (termed monomer) were determined to be 4.1 S, 2.4 nm, and 1.05, respectively. The molecular weight of the monomer enzyme was determined to be 38,000 by polyacrylamide gel electrophoresis. Incubation of the high molecular weight phosphatase with a cleavable cross-linking reagent, 3,3'-dithiobis(sulfosuccinimidyl propionate), showed the formation of a cross-linked complex. The molecular weight of the cross-linked complex was determined to be 85,000 and a second dimension gel electrophoresis of the cleaved cross-linked complex showed that the latter contained only 38,000-dalton bands. Limited trypsinization of the enzyme released a approximately 4,000-dalton peptide from the monomers and dissociated the high molecular weight phosphatase into 34,000-dalton monomers. It is proposed that the catalytic activity of the native glycogen-bound phosphatase resides in a dimer of 38,000-dalton subunits.     

MHC class I allele dosage alters CD8 expression by intestinal intraepithelial lymphocytes

The development of TCR alphabeta(+), CD8alphabeta(+) intestinal intraepithelial lymphocytes (IEL) is dependent on MHC class I molecules expressed in the thymus, while some CD8alphaalpha(+) IEL may arise independently of MHC class I. We examined the influence of MHC I allele dosage on the development CD8(+) T cells in RAG 2(-/-) mice expressing the H-2D(b)-restricted transgenic TCR specific for the male, Smcy-derived H-Y Ag (H-Y TCR). IEL in male mice heterozygous for the restricting (H-2D(b)) and nonrestricting (H-2D(d)) MHC class I alleles (MHC F(1)) were composed of a mixture of CD8alphabeta(+) and CD8alphaalpha(+) T cells, while T cells in the spleen were mostly CD8alphabeta(+). This was unlike IEL in male mice homozygous for H-2D(b), which had predominantly CD8alphaalpha(+) IEL and few mostly CD8(-) T cells in the spleen. Our results demonstrate that deletion of CD8alphabeta(+) cells in H-Y TCR male mice is dependent on two copies of H-2D(b), whereas the generation of CD8alphaalpha(+) IEL requires only one copy. The existence of CD8alphabeta(+) and CD8alphaalpha(+) IEL in MHC F(1) mice suggests that their generation is not mutually exclusive in cells with identical TCR. Furthermore, our data imply that the level of the restricting MHC class I allele determines a threshold for conventional CD8alphabeta(+) T cell selection in the thymus of H-Y TCR-transgenic mice, whereas the development of CD8alphaalpha(+) IEL is dependent on, but less sensitive to, this MHC class I allele.     

Cross-linking of transmembrane helices in proton-translocating nicotinamide nucleotide transhydrogenase from Escherichia coli: implications for the structure and function of the membrane domain

Proton-pumping nicotinamide nucleotide transhydrogenase from Escherichia coli contains an alpha and a beta subunit of 54 and 49 kDa, respectively, and is made up of three domains. Domain I (dI) and III (dIII) are hydrophilic and contain the NAD(H)- and NADP(H)-binding sites, respectively, whereas the hydrophobic domain II (dII) contains 13 transmembrane alpha-helices and harbours the proton channel. Using a cysteine-free transhydrogenase, the organization of dII and helix-helix distances were investigated by the introduction of one or two cysteines in helix-helix loops on the periplasmic side. Mutants were subsequently cross-linked in the absence and presence of diamide and the bifunctional maleimide cross-linker o-PDM (6 A), and visualized by SDS-PAGE. In the alpha(2)beta(2) tetramer, alphabeta cross-links were obtained with the alphaG476C-betaS2C, alphaG476C-betaT54C and alphaG476C-betaS183C double mutants. Significant alphaalpha cross-links were obtained with the alphaG476C single mutant in the loop connecting helix 3 and 4, whereas betabeta cross-links were obtained with the betaS2C, betaT54C and betaS183C single mutants in the beginning of helix 6, the loop between helix 7 and 8 and the loop connecting helix 11 and 12, respectively. In a model based on 13 mutants, the interface between the alpha and beta subunits in the dimer is lined along an axis formed by helices 3 and 4 from the alpha subunit and helices 6, 7 and 8 from the beta subunit. In addition, helices 2 and 4 in the alpha subunit together with helices 6 and 12 in the beta subunit interact with their counterparts in the alpha(2)beta(2) tetramer. Each beta subunit in the alpha(2)beta(2) tetramer was concluded to contain a proton channel composed of the highly conserved helices 9, 10, 13 and 14.     

The cryptophycin-tubulin ring structure indicates two points of curvature in the tubulin dimer

Cryptophycin-1 is the parent compound of a group of cyclic peptides with potent antineoplastic activity. Cryptophycins are thought to function by modulating the dynamic instability of spindle microtubules, and in vitro are known to bind in an equimolar ratio to the beta-tubulin subunit and to induce the formation of ring-like complexes. However, the detailed mechanisms whereby the cryptophycins interact with tubulin are not known. We have investigated the origin of the conformational changes in tubulin both biochemically and by electron microscopy and image analysis. Cryptophycin was found to protect both alpha- and beta-tubulin against proteolysis by trypsin, indicating conformational changes in specific regions of both subunits. The ring mass was determined to be approximately 0.81 MDa by sedimentation velocity combined with dynamic light scattering and by STEM, indicating a complex of eight alphabeta dimers. Statistical analysis of rings imaged by cryoelectron microscopy revealed 16-fold symmetry, corresponding to eight dimers. Computational averaging based on this symmetry yielded an image of a 24 nm diameter ring, at 2.6 nm resolution, that clearly distinguishes intradimer contacts from interdimer contacts, and allows discrimination of alpha-subunits from beta-subunits. Fitting of the tubulin dimer crystal structure into this projected density map indicates two points of curvature: a 13 degrees intradimer bend and a 32 degrees interdimer bend. We conclude that drug binding to one subunit (beta) results in two bends per dimer, affecting both subunits.     

IL-15 promotes survival but not effector function differentiation of CD8+ TCRalphabeta+ intestinal intraepithelial lymphocytes

CD8 single-positive cells, including CD8alphaalpha+ and CD8alphabeta+ subsets, constitute the majority of TCRalphabeta+ intestinal intraepithelial lymphocytes (alphabeta iIEL) in mice. CD8+ alphabeta iIEL show significantly weaker responses to TCR stimulation in the presence of exogenous IL-2 than do CD8+ T cells of the central immune system. IL-15 is a T cell growth factor likely expressed in the intestine mucosa. To understand the role of IL-15 in CD8+ alphabeta iIEL biology, we compared the effects of exogenous IL-15 and IL-2 on the survival and primary responses of the two CD8+ alphabeta iIEL subsets in vitro. In contrast to the death of approximately 60% of both CD8alphaalpha+ and CD8alphabeta+ iIEL cultured in IL-2 with or without TCR stimulation, IL-15 promoted survival of the CD8alphaalpha+ subset in the presence of TCR stimulation and promoted survival of both subsets in the absence of TCR stimulation. The higher proliferation level of TCR stimulated CD8alphaalpha+ alphabeta iIEL cultured in IL-15 compared with those cultured in IL-2 is likely due to IL-15's prosurvival effects. In addition, unlike exogenous IL-2, exogenous IL-15 did not support the effector functions of either iIEL subsets, including IFN-gamma production, IL-4-induced Th2 cytokine production, and anti-TCR mAb-redirected cytotoxicity. These findings demonstrate that IL-15 and IL-2 are functionally distinct and suggest that IL-15 plays a unique role in the maintenance of the CD8+ alphabeta iIEL pool in the absence of Ag stimulation and in the survival and expansion of CD8alphaalpha+ alphabeta iIEL upon Ag stimulation.     

The uncoupling protein dimer can form a disulfide cross-link between the mobile C-terminal SH groups

Isolated uncoupling protein (UCP) can be cross-linked, by various disulfide-forming reagents, to dimers. The best cross-linking is achieved with Cu2+-phenanthroline oxidation. Because cross-linking is independent of UCP concentration and prevented by SDS addition, a disulfide bridge must be formed between the two subunits of the native dimer. Cross-linking is prevented by SH reagent and reversed by SH-reducing reagents. In mitochondria, cross-linking of UCP with disulfide-forming agents is even more efficient than in isolated state. It proves that UCP is a dimer in mitochondria, before isolation. Disulfide-bridge formation does not inhibit GTP-binding to UCP. Cross-linked UCP re-incorporated in proteoliposomes either before or after cross-linking fully retains the H1-transport function. Rapid cross-linking by membrane impermeant reagents indicates a surface localization of the C-terminus in soluble UCP and projection to the outer surface in mitochondria. Intermolecular disulfide-bridge formation in a dimer requires juxtaposition of identical cysteines at the twofold symmetry axis. A rigid juxtaposition of cysteines is unlikely, unless intended for a native disulfide bridge. The absence of such a bridge in UCP suggests that juxtaposition of cysteines is generated by high mobility. In order to localize the cysteine involved, cross-linked UCP was cleaved by BrCN. The CB-7 C-terminal peptide, which contains cysteines at positions 287 and 304, disappears. Limited trypsinolytic cleavage, previously shown to occur at Lys-292, removed cross-linking in UCP both in the solubilized and mitochondrially bound state. The cleaved C-terminal peptide of 11 residues contains only cystein-304 which, thus, should be the only one (out of 7 cysteines in UCP) involved in the S-S bridge formation. Obviously, the C-terminal location of the cysteine, because of its high mobility, permits juxtapositioning for cross-linking. This agrees with predictions from hydrophobicity analysis that the last 14 residues in UCP protrude from the membrane.     

The formation of a disulfide cross-link between the two subunits demonstrates the dimeric structure of the mitochondrial oxoglutarate carrier

Isolated oxoglutarate carrier (OGC) can be cross-linked to dimers by disulfide-forming reagents such as Cu²⁺-phenanthroline and diamide. Acetone and other solvents increase the extent of Cu²⁺-phenanthroline-induced cross-linking of OGC. Cross-linked OGC re-incorporated in photeoliposomes fully retains the oxoglutarate transport activity. The amount of cross-linked OGC calculated by densitometry of scanned gels depends on the method of staining, since cross-linked OGC exhibits a higher sensitivity to Coomassie brilliant blue as compared to silver nitrate. Under optimal conditions the formation of cross-linked OGC dimer (stained with Coomassie brilliant blue) amounts to 75% of the total protein. Approximately the same cross-linking efficiency was evaluated from Western blots. Cross-linking of OGC is prevented by SH reagents and reversed by SH-reducing reagents, which shows that it is mediated by disulfide bridge(s). The formation of SS bridge(s) requires the native state of the protein, since it is suppressed by SDS and by heating. Furthermore, the extent of cross-linking is independent of OGC concentration indicating that disulfide bridge(s) must be formed between the two subunits of native dimers. The number and localization of disulfide bridge(s) in the cross-linked OGC were examined by peptide fragmentation and subsequent cleavage of disulfide bond(s) by β-mercaptoethanol. Our experimental results show that cross-linking of OGC is accomplished by a single disulfide bond between the cysteines 184 of the two subunits and suggest that these residues in the putative transmembrane helix four are fairly close to the twofold axis of the native dimer structure.