Kevin R. MacKenzie

Sequence-specific Dimerization of the Transmembrane Domain of the “BH3-only” Protein BNIP3 in Membranes and Detergent

Mitochondria-mediated apoptosis is regulated by proteins of the Bcl-2 superfamily, most of which contain a C-terminal hydrophobic domain that plays a role in membrane targeting. Experiments with BNIP3 have implicated the transmembrane (TM) domain in its proapoptotic function, homodimerization, and interactions with Bcl-2 and Bcl-xL. We show that the BNIP3 TM domain self-associates strongly in Escherichia coli cell membranes and causes reversible dimerization of a soluble protein in the detergent SDS when expressed as an in-frame fusion. Limited mutational analysis identifies specific residues that are critical for BNIP3 TM self-association in membranes, and these residues are also important for dimerization in SDS micelles, suggesting that the self-association observed in membranes is preserved in detergent. The effects of sequence changes at positions Ala176 and Gly180 suggest that the BNIP3 TM domain associates using a variant of the GXXXG motif previously shown to be important in the dimerization of glycophorin A. The importance of residue His173 in BNIP3 TM domain dimerization indicates that polar residues, which have been implicated in self-association of model TM peptides, can act in concert with the AXXXG motif to stabilize TM domain interactions. Our results demonstrate that the hydrophobic C-terminal TM domain of the pro-apoptotic BNIP3 protein dimerizes tightly in lipidic environments, and that this association has a strong sequence dependence but is independent of the identity of flanking regions. Thus, the transmembrane domain represents another region of the Bcl-2 superfamily of proteins that is capable of mediating strong and specific protein-protein interactions.

Association energetics of membrane spanning α-helices

Since Popot and Engelman proposed the ‘two-stage’ thermodynamic framework for dissecting the energetics of helical membrane protein folding, scientists have endeavored to measure the free energies of helix–helix associations to better understand how interactions between helices stabilize and specify native membrane protein folds. Chief among the biophysical tools used to probe these energies are sedimentation equilibrium analytical ultracentrifugation, fluorescence resonance energy transfer, and thiol disulfide interchange experiments. Direct and indirect comparisons of thermodynamic results suggest that differences in helix–helix stabilities between micelles and bilayers may not be as large as previously anticipated. Genetic approaches continue to become more quantitative, and the propensities for helices to interact in bacterial membranes generally correlate well with in vitro measurements.

Structural basis for dimerization of the BNIP3 transmembrane domain.

Mutagenesis data suggest that BNIP3 transmembrane domain dimerization depends critically on hydrogen bonding between His 173 and Ser 172, but a recent structural analysis indicates that these residues adopt multiple conformations and are not always hydrogen bonded. We show that in dodecylphosphocholine micelles the structure of the BNIP3 transmembrane domain is modulated by phospholipids and that appropriate reconstitution and lipid titration yield a single set of peptide resonances. NMR structure determination reveals a symmetric dimer in which all interfacial residues, including His 173 and Ser 172, are well-defined. Small residues Ala 176, Gly 180, and Gly 184 allow close approach of essentially ideal helices in a geometry that supports intermonomer hydrogen bond formation between the side chains of His 173 and Ser 172. Bulky residues Ile 177 and Ile 181 pack against small residues of the opposite monomer, and favorable polar backbone-backbone contacts at the interface likely include noncanonical Calpha-H.O=C hydrogen bonds from Gly 180 to Ile 177. Modeling mutations into the structure shows that most deleterious hydrophobic substitutions eliminate the His-Ser hydrogen bond or introduce an intermonomer clash, indicating critical roles for sterics and hydrogen bonding in the sequence dependence of dimerization. Substitutions at most noninterfacial positions do not alter dimerization, but the disruptive effects of substitutions at Ile 183 cannot be rationalized in terms of peptide-peptide contacts and therefore may indicate a role for peptide-detergent or peptide-lipid interactions at this position.

Controlling Peptide Structure with Coordination Chemistry: Robust and Reversible Peptide–Dirhodium Ligation

Selective bond formation of biomolecule substrates requires synthetic chemistry in an aqueous, functional-grouprich environment to alter and control the structure, function, and aggregation state of biomolecules. Ideally, new methods would be non-denaturing, would utilize chemistry that is orthogonal to existing methods, and would be reversible in response to external stimuli. We here report a reversible method for addressing and bridging glutamate and aspartate carboxylates under mild aqueous conditions. One powerful way to affect polypeptide structure is through methods that link, or bridge, two amino-acid side chains to form a cyclic product, but there exist few methods to selectively and reversibly bridge amino acid side chains. Cysteine is commonly used as a handle for selective bond formation through redox-mediated formation of disulfide linkages or selective alkylation reactions, but working with cysteine-containing peptides can be difficult. Other methods to address naturally occurring amino acids include activated esters of organic dicarboxylic acids for cross-linking lysine residues, but reversibility in this case is limited. Metal ions serve structural roles in metalloproteins, where side chains serve as ligands which are bridged by a metal ion to create a folded metal-binding pocket. Taking a cue from these biological examples, the effects of metal binding on peptide structures is an active area of study. Peptide– metal interactions have been used to understand metalloprotein folding and energetics and to shed light on potential toxicity pathways. Many transition metals can bind to peptides in aqueous solution, most commonly through cysteine or histidine residues. Although metalloproteins can bind extremely tightly to metal ions, smaller designed polypeptides often bind metal ions dynamically in aqueous solution, and systems involving a small number of binding groups often require excesses of metal ion in solution. We set out to develop a robust metal-binding method that would selectively address side chain functional groups in a manner complementary to extant methods. Carboxylates are common, naturally occurring polypeptide side chains, and the dirhodium–carboxylate interaction is sufficiently stable that ligand exchange is not observed under a variety of biologically relevant conditions. Selective binding to polycarboxylate regions has been observed with lanthanides, but well defined bridging of a small number of carboxylate side chains has not been demonstrated. We set out to explore the ability of dirhodium tetracarboxylates to bind chemoselectively and tightly to side chain carboxylates and hoped to develop a reversible metal ligation protocol to link two carboxylate side chains under nondenaturing conditions. In contrast to equilibrium binding exhibited in many peptide-metal binding interactions the robust carboxylate dirhodium bond allows us to enforce structural changes of an isolable peptide metal adduct. For initial studies, we focussed on the bis ACHTUNGTRENNUNG(cysteine) hairpin domain from a typical zinc finger protein, ZIF268. We chose to examine the generality of the zinc-binding domain, determining if it could serve as the basis for new dirhodiumbinding domains through amino acid substitutions to position two Asp residues in place of the zinc-binding Cys. The peptide sequence ZF, derived from ZIF268 P62–A73, contains a number of reactive functional groups, necessitating a uniquely selective complexation method (Scheme 1). In addition, the parent zinc finger sequence folds into the common a-helix,b-sheet motif in the presence of three or four ligands, but exhibits no change from a random coil in the presence of only two ligands. Bridging the dirhodium tetracarboxylate core with traditional ligands has proven challenging. Preparative yields of bridged structures have been largely limited m-phenylene structures, and it has been reported that aliphatic a,w-diacids give product mixtures and low yields of chelate prod[a] A. N. Zaykov, Prof. Z. T. Ball Department of Chemistry MS-60, Rice University Houston, Texas 77005 (USA) Fax: (+1) 713-348-5155 E-mail : zb1@rice.edu [b] Prof. K. R. MacKenzie Department of Biochemistry and Cell Biology, Rice University Houston, Texas 77005 (USA) Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.200901266.

Intermonomer Hydrogen Bonds Enhance GxxxG-Driven Dimerization of the BNIP3 Transmembrane Domain: Roles for Sequence Context in Helix–Helix Association in Membranes

We determined the sequence dependence of human BNIP3 transmembrane domain dimerization using the biological assay TOXCAT. Mutants in which intermonomer hydrogen bonds between Ser172 and His173 are abolished show moderate interaction, indicating that side-chain hydrogen bonds contribute to dimer stability but are not essential to dimerization. Mutants in which a GxxxG motif composed of Gly180 and Gly184 has been abolished show little or no interaction, demonstrating the critical nature of the GxxxG motif to BNIP3 dimerization. These findings show that side-chain hydrogen bonds can enhance the intrinsic dimerization of a GxxxG motif and that sequence context can control how hydrogen bonds influence helix-helix interactions in membranes. The dimer interface mapped by TOXCAT mutagenesis agrees closely with the interfaces observed in the NMR structure and inferred from mutational analysis of dimerization on SDS-PAGE, showing that the native dimer structure is retained in detergents. We show that TOXCAT and SDS-PAGE give complementary and consistent information about BNIP3 transmembrane domain dimerization: TOXCAT is insensitive to mutations that have modest effects on self-association in detergents but readily discriminates among mutations that completely disrupt detergent-resistant dimerization. The close agreement between conclusions reached from TOXCAT and SDS-PAGE data for BNIP3 suggests that accurate estimates of the relative effects of mutations on native-state protein-protein interactions can be obtained even when the detergent environment is strongly disruptive.

Structure of androcam supports specialized interactions with myosin VI

Androcam replaces calmodulin as a tissue-specific myosin VI light chain on the actin cones that mediate D. melanogaster spermatid individualization. We show that the androcam structure and its binding to the myosin VI structural (Insert 2) and regulatory (IQ) light chain sites are distinct from those of calmodulin and provide a basis for specialized myosin VI function. The androcam N lobe noncanonically binds a single Ca2+ and is locked in a “closed” conformation, causing androcam to contact the Insert 2 site with its C lobe only. Androcam replacing calmodulin at Insert 2 will increase myosin VI lever arm flexibility, which may favor the compact monomeric form of myosin VI that functions on the actin cones by facilitating the collapse of the C-terminal region onto the motor domain. The tethered androcam N lobe could stabilize the monomer through contacts with C-terminal portions of the motor or recruit other components to the actin cones. Androcam binds the IQ site at all calcium levels, constitutively mimicking a conformation adopted by calmodulin only at intermediate calcium levels. Thus, androcam replacing calmodulin at IQ will abolish a Ca2+-regulated, calmodulin-mediated myosin VI structural change. We propose that the N lobe prevents androcam from interfering with other calmodulin-mediated Ca2+ signaling events. We discuss how gene duplication and mutations that selectively stabilize one of the many conformations available to calmodulin support the molecular evolution of structurally and functionally distinct calmodulin-like proteins.