Remodeling of HIV-1 Nef Structure by Src-Family Kinase Binding

The HIV-1 accessory protein Nef controls multiple aspects of the viral life cycle and host immune response, making it an attractive therapeutic target. Previous X-ray crystal structures of Nef in complex with key host cell binding partners have shed light on protein–protein interactions critical to Nef function. Crystal structures of Nef in complex with either the SH3 or tandem SH3–SH2 domains of Src-family kinases reveal distinct dimer conformations of Nef. However, the existence of these Nef dimer complexes in solution has not been established. Here we used hydrogen exchange mass spectrometry (HX MS) to compare the solution conformation of Nef alone and in complexes with the SH3 or the SH3–SH2 domains of the Src-family kinase Hck. HX MS revealed that interaction with the Hck SH3 or tandem SH3–SH2 domains induces protection of the Nef αB-helix from deuterium uptake, consistent with a role for αB in dimer formation. HX MS analysis of a Nef mutant (position Asp123, a site buried in the Nef:SH3 dimer but surface exposed in the Nef:SH3–SH2 complex), showed a Hck-induced conformational change in Nef relative to wild-type Nef. These results support a model in which Src-family kinase binding induces conformational changes in Nef to expose residues critical for interaction with the μ1 subunit of adaptor protein 1 and the major histocompatibility complex-1 tail, and subsequent major histocompatibility complex-1 downregulation and immune escape of HIV-infected cells required for functional interactions with downstream binding partners.     

A rationally designed monomeric variant of anthranilate phosphoribosyltransferase from Sulfolobus solfataricus is as active as the dimeric wild-type enzyme but less thermostable

The anthranilate phosphoribosyltransferase from Sulfolobus solfataricus (ssAnPRT) forms a homodimer with a hydrophobic subunit interface. To elucidate the role of oligomerisation for catalytic activity and thermal stability of the enzyme, we loosened the dimer by replacing two apolar interface residues with negatively charged residues (mutations I36E and M47D). The purified double mutant I36E+M47D formed a monomer with wild-type catalytic activity but reduced thermal stability. The single mutants I36E and M47D were present in a monomer-dimer equilibrium with dissociation constants of about 1 μM and 20 μM, respectively, which were calculated from the concentration-dependence of their heat inactivation kinetics. The monomeric form of M47D, which is populated at low subunit concentrations, was as thermolabile as monomeric I36E+M47D. Likewise, the dimeric form of I36E, which was populated at high subunit concentrations, was as thermostable as dimeric wild-type ssAnPRT. These findings show that the increased stability of wild-type ssAnPRT compared to the I36E+M47D double mutant is not caused by the amino acid exchanges per se but by the higher intrinsic stability of the dimer compared to the monomer. In accordance with the negligible effect of the mutations on catalytic activity and stability, the X-ray structure of M47D contains only minor local perturbations at the dimer interface. We conclude that the monomeric double mutant resembles the individual wild-type subunits, and that ssAnPRT is a dimer for stability but not for activity reasons.     

Oligomerization is required for HIV-1 Nef-induced activation of the Src family protein-tyrosine kinase, Hck

Hck is a member of the Src protein-tyrosine kinase family and is expressed strongly in macrophages, an important HIV target cell. Previous studies have shown that Nef, an HIV-1 accessory protein essential for AIDS progression, binds and activates Hck through its SH3 domain. Structural analysis suggests that Nef forms oligomers in vivo, which may bring multiple Hck molecules into close proximity and promote autophosphorylation. Using bimolecular GFP fluorescence complementation, we show for the first time that Nef oligomerizes in living cells and that the oligomers localize to the plasma membrane. To test the role of Nef oligomerization in Hck activation, we fused Nef to the hormone-binding domain of the estrogen receptor (Nef-ER), allowing us to control its dimerization with 4-hydroxytamoxifen (4-HT). In Rat-2 fibroblasts co-expressing Nef-ER and Hck, 4-HT treatment induced Nef-ER dimer and tetramer formation, leading to Hck kinase activation and cellular transformation. The number of transformed foci observed with Nef-ER plus Hck in the presence of 4-HT was markedly greater than that observed with wild-type Nef plus Hck, suggesting that enforced oligomerization enhances activation of Hck by Nef in vivo. Enhanced transformation correlated with increased Hck/Nef complex formation at the plasma membrane. In complementary experiments, we observed that a Nef mutant defective for Hck SH3 domain binding (Nef-PA) suppressed Hck kinase activation and transformation by the wild-type Hck/Nef complex. This effect correlated with the formation of a ternary complex between wild-type Nef, Nef-PA, and Hck, suggesting that Nef-PA suppresses Hck activation by blocking wild-type Nef oligomerization. Finally, Nef-ER induced Hck activation in a 4-HT-dependent manner in the macrophage precursor cell line TF-1, suggesting that oligomerization is essential for signaling through Hck in a cell background relevant to HIV infection. Together, these data demonstrate that Nef oligomerization is critical to the activation of Hck in vivo, and suggest that inhibitors of oligomerization may suppress Nef signaling through Hck in HIV-infected macrophages, slowing disease progression.     

Altering CLC stoichiometry by reducing non-polar side-chains at the dimerization interface

CLC-ec1 is a Cl⁻/H⁺ antiporter that forms stable homodimers in lipid bilayers, with a free energy of −10.9 kcal/mol in 2:1 POPE/POPG lipid bilayers. The dimerization interface is formed by four transmembrane helices: H, I, P and Q, that are lined by non-polar side-chains that come in close contact, yet it is unclear as to whether their interactions drive dimerization. To investigate whether non-polar side-chains are required for dimer assembly, we designed a series of constructs where side-chain packing in the dimer state is significantly reduced by making 4–5 alanine substitutions along each helix (H-ala, I-ala, P-ala, Q-ala). All constructs are functional and three purify as stable dimers in detergent micelles despite the removal of significant side-chain interactions. On the other hand, H-ala shows the unique behavior of purifying as a mixture of monomers and dimers, followed by a rapid and complete conversion to monomers. In lipid bilayers, all four constructs are monomeric as examined by single-molecule photobleaching analysis. Further study of the H-helix shows that the single mutation L194A is sufficient to yield monomeric CLC-ec1 in detergent micelles and lipid bilayers. X-ray crystal structures of L194A reveal the protein re-assembles to form dimers, with a structure that is identical to wild-type. Altogether, these results demonstrate that non-polar membrane embedded side-chains play an important role in defining dimer stability, but the stoichiometry is highly contextual to the solvent environment. Furthermore, we discovered that L194 is a molecular hot-spot for defining dimerization of CLC-ec1.     

Structural basis of membrane-induced cardiotoxin A3 oligomerization

Cobra cardiotoxins (CTXs) have previously been shown to induce membrane fusion of vesicles formed by phospholipids such as cardiolipin or sphingomyelin. CTX can also form a pore in membrane bilayers containing a anionic lipid such as phosphatidylserine or phosphatidylglycerol. Herein, we show that the interaction of CTX with negatively charged lipids causes CTX dimerization, an important intermediate for the eventual oligomerization of CTX during the CTX-induced fusion and pore formation process. The structural basis of the lipid-induced oligomerization of CTX A3, a major CTX from Naja atra, is then illustrated by the crystal structure of CTX A3 in complex with SDS; SDS likely mimics anionic lipids of the membrane under micelle conditions at 1.9-A resolution. The crystal packing reveals distinct SDS-free and SDS-rich regions; in the latter two types of interconnecting CTX A3 dimers, D1 and D2, and several SDS molecules can be identified to stabilize D1 and D2 by simultaneously interacting with residues at each dimer interface. When the three CTXSDS complexes in the asymmetric unit are overlaid, the orientation of CTX A3 monomers relative to the SDS molecules in the crystal is strikingly similar to that of the toxin with respect to model membranes as determined by NMR and Fourier transform infrared methods. These results not only illustrate how lipid-induced CTX dimer formation may be transformed into oligomers either as inverted micelles of fusion intermediates or as membrane pore of anionic lipid bilayers but also underscore a potential role for SDS in x-ray diffraction study of protein-membrane interactions in the future.     

The role of quaternary interactions on the stability and activity of ascorbate peroxidase.

Point mutations at the dimer interface of the homodimeric enzyme ascorbate peroxidase (APx) were constructed to assess the role of quaternary interactions in the stability and activity of APx. Analysis of the APx crystal structure shows that Glu112 forms a salt bridge with Lys20 and Arg24 of the opposing subunit near the axis of dyad symmetry between the subunits. Two point mutants, E112A and E112K, were made to determine the effects of a neutral (alanine) and repulsive (lysine) mutation on dimerization, stability, and activity. Gel filtration analysis indicated that the ratio of the monomer to dimer increased as the dimer interface interactions went from attractive to repulsive. Differential scanning calorimetry (DSC) data exhibited a decrease in both the transition temperature (Tm) and enthalpy of unfolding (deltaHc) with Tm = 58.3 +/- 0.5 degrees C, 56.0 +/- 0.8 degrees C, and 53.0 +/- 0.9 degrees C and deltaHc = 245 +/- 29 kcal/mol, 199 +/- 38 kcal/mol, and 170 +/- 25 kcal/mol for wild-type APx, E112A, and E112K, respectively. Similar changes were observed based on thermal melting curves obtained by absorption spectroscopy. No change in enzyme activity was found for the E112A mutant, and only a 25% drop in activity was observed for the E112K mutant which demonstrates that the non-Michaelis Menten kinetics of APx is not due to the APx oligomeric structure. The cryogenic crystal structures of the wild-type and mutant proteins show that mutation induced changes are limited to the dimer interface including an alteration in solvent structure.     

HIV-1 Nef protein: purification, crystallizations, and preliminary X-ray diffraction studies.

Human immunodeficiency virus Nef protein accelerates virulent progression of AIDS by its interaction with specific cellular proteins involved in cellular activation and signal transduction. Here we report the purification and crystallization of the conserved core of HIV-1LAI Nef protein in the unliganded form and in complex with the wild-type SH3 domain of the P59fyn protein-tyrosine kinase. One-dimensional NMR experiments show that full-length protein and truncated fragment corresponding to the product of HIV-1 protease cleavage have a well-folded compact tertiary structure. The ligand-free HIV-1 Nefcore protein forms cubic crystals belonging to space group P23 with unit cell dimensions of a = b = c = 86.4 A. The Nef-Fyn SH3 cocrystals belong to the space group P6(1)22 or its enantiomorph, P6(5)22, with unit cell dimensions of a = b = 108.2 A and c = 223.7 A. Both crystal forms diffract to a resolution limit of 3.0 A resolution using synchrotron radiation, and are thus suitable for X-ray structure determination.     

Engineering the quaternary structure of an enzyme: construction and analysis of a monomeric form of malate dehydrogenase from Escherichia coli.

The citric acid cycle enzyme, malate dehydrogenase (MDH), is a dimer of identical subunits. In the crystal structures of 2 prokaryotic and 2 eukaryotic forms, the subunit interface is conformationally homologous. To determine whether or not the quaternary structure of MDH is linked to the catalytic activity, mutant forms of the enzyme from Escherichia coli have been constructed. Utilizing the high-resolution structure of E. coli MDH, the dimer interface was analyzed critically for side chains that were spatially constricted and needed for electrostatic interactions. Two such residues were found, D45 and S226. At their nearest point in the homodimer, they are in different subunits, hydrogen bond across the interface, and do not interact with any catalytic residues. Each residue was mutated to a tyrosine, which should disrupt the interface because of its large size. All mutants were cloned and purified to homogeneity from an mdh- E. coli strain (BHB111). Gel filtration of the mutants show that D45Y and D45Y/S226Y are both monomers, whereas the S226Y mutant remains a dimer. The monomeric D45Y and D45Y/S226Y mutants have 14,000- and 17,500-fold less specific activity, respectively, than the native enzyme. The dimeric S226Y has only 1.4-fold less specific activity. All forms crystallized, indicating they were not random coils. Data have been collected to 2.8 A resolution for the D45Y mutant. The mutant is not isomorphous with the native protein and work is underway to solve the structure by molecular replacement.     

Variable oligomerization modes in coronavirus non-structural protein 9

Non-structural protein 9 (Nsp9) of coronaviruses is believed to bind single-stranded RNA in the viral replication complex. The crystal structure of Nsp9 of human coronavirus (HCoV) 229E reveals a novel disulfide-linked homodimer, which is very different from the previously reported Nsp9 dimer of SARS coronavirus. In contrast, the structure of the Cys69Ala mutant of HCoV-229E Nsp9 shows the same dimer organization as the SARS-CoV protein. In the crystal, the wild-type HCoV-229E protein forms a trimer of dimers, whereas the mutant and SARS-CoV Nsp9 are organized in rod-like polymers. Chemical cross-linking suggests similar modes of aggregation in solution. In zone-interference gel electrophoresis assays and surface plasmon resonance experiments, the HCoV-229E wild-type protein is found to bind oligonucleotides with relatively high affinity, whereas binding by the Cys69Ala and Cys69Ser mutants is observed only for the longest oligonucleotides. The corresponding mutations in SARS-CoV Nsp9 do not hamper nucleic acid binding. From the crystal structures, a model for single-stranded RNA binding by Nsp9 is deduced. We propose that both forms of the Nsp9 dimer are biologically relevant; the occurrence of the disulfide-bonded form may be correlated with oxidative stress induced in the host cell by the viral infection.     

Structure of the Q237W mutant of HhaI DNA methyltransferase: an insight into protein-protein interactions

We have determined the structure of a mutant (Q237W) of HhaI DNA methyltransferase, complexed with the methyl-donor product AdoHcy. The Q237W mutant proteins were crystallized in the monoclinic space group C2 with two molecules in the crystallographic asymmetric unit. Protein-protein interface calculations in the crystal lattices suggest that the dimer interface has the specific characteristics for homodimer protein-protein interactions, while the two active sites are spatially independent on the outer surface of the dimer. The solution behavior suggests the formation of HhaI dimers as well. The same HhaI dimer interface is also observed in the previously characterized binary (M.HhaI-AdoMet) and ternary (M.HhaI-DNA-AdoHcy) complex structures, crystallized in different space groups. The dimer is characterized either by a non-crystallographic two-fold symmetry or a crystallographic symmetry. The dimer interface involves three segments: the amino-terminal residues 2-8, the carboxy-terminal residues 313-327, and the linker (amino acids 179-184) between the two functional domains--the catalytic methylation domain and the DNA target recognition domain. Both the amino- and carboxy-terminal segments are part of the methylation domain. We also examined protein-protein interactions of other structurally characterized DNA MTases, which are often found as a 2-fold related 'dimer' with the largest dimer interface area for the group-beta MTases. A possible evolutionary link between the Type I and Type II restriction-modification systems is discussed.     

Phosphorylation mutants elucidate the mechanism of annexin IV-mediated membrane aggregation

Site-directed mutagenesis, electron microscopy, and X-ray crystallography were used to probe the structural basis of annexin IV-induced membrane aggregation and the inhibition of this property by protein kinase C phosphorylation. Site-directed mutants that either mimic (Thr6Asp, T6D) or prevent (Thr6Ala, T6A) phosphorylation of threonine 6 were produced for these studies and compared with wild-type annexin IV. In vitro assays showed that unmodified wild-type annexin IV and the T6A mutant, but not PKC-phosphorylated wild-type or the T6D mutant, promote vesicle aggregation. Electron crystallographic data of wild-type and T6D annexin IV revealed that, similar to annexin V, the annexin IV proteins form 2D trimer-based ordered arrays on phospholipid monolayers. Cryo-electron microscopic images of junctions formed between lipid vesicles in the presence of wild-type annexin IV indicated a separation distance corresponding to the thickness of two layers of membrane-bound annexin IV. In this orientation, a single layer of WT annexin IV, attached to the outer leaflet of one vesicle, would undergo face-to-face self-association with the annexin layer of a second vesicle. The 2.0-A resolution crystal structure of the T6D mutant showed that the mutation causes release of the N-terminal tail from the protein core. This change would preclude the face-to-face annexin self-association required to aggregate vesicles. The data suggest that reversible complex formation through phosphorylation and dephosphorylation could occur in vivo and play a role in the regulation of vesicle trafficking following changes in physiological states.     

Dimer structure of an interfacially impaired phosphatidylinositol-specific phospholipase C

The crystal structure of the W47A/W242A mutant of phosphatidylinositol-specific phospholipase C (PI-PLC) from Bacillus thuringiensis has been solved to 1.8A resolution. The W47A/W242A mutant is an interfacially challenged enzyme, and it has been proposed that one or both tryptophan side chains serve as membrane interfacial anchors (Feng, J., Wehbi, H., and Roberts, M. F. (2002) J. Biol. Chem. 277, 19867-19875). The crystal structure supports this hypothesis. Relative to the crystal structure of the closely related (97% identity) wild-type PI-PLC from Bacillus cereus, significant conformational differences occur at the membrane-binding interfacial region rather than the active site. The Trp --> Ala mutations not only remove the membrane-partitioning aromatic side chains but also perturb the conformations of the so-called helix B and rim loop regions, both of which are implicated in interfacial binding. The crystal structure also reveals a homodimer, the first such observation for a bacterial PI-PLC, with pseudo-2-fold symmetry. The symmetric dimer interface is stabilized by hydrophobic and hydrogen-bonding interactions, contributed primarily by a central swath of aromatic residues arranged in a quasiherringbone pattern. Evidence that interfacially active wild-type PI-PLC enzymes may dimerize in the presence of phosphatidylcholine vesicles is provided by fluorescence quenching of PI-PLC mutants with pyrene-labeled cysteine residues. The combined data suggest that wild-type PI-PLC can form similar homodimers, anchored to the interface by the tryptophan and neighboring membrane-partitioning residues.     

An Integrative Approach Combining Noncovalent Mass Spectrometry, Enzyme Kinetics and X-ray Crystallography to Decipher Tgt Protein-Protein and Protein-RNA Interaction

The tRNA-modifying enzyme tRNA-guanine transglycosylase (Tgt) is a putative target for new selective antibiotics against Shigella bacteria. The formation of a Tgt homodimer was suggested on the basis of several crystal structures of Tgt in complex with RNA. In the present study, noncovalent mass spectrometry was used (i) to confirm the dimeric oligomerization state of Tgt in solution and (ii) to evidence the binding stoichiometry of the complex formed between Tgt and its full-length substrate tRNA. To further investigate the importance of Tgt protein-protein interaction, point mutations were introduced into the dimer interface in order to study their influence on the formation of the catalytically active complex. Enzyme kinetics revealed a reduced catalytic activity of these mutated variants, which could be related to a destabilization of the dimer formation as evidenced by both noncovalent mass spectrometry and X-ray crystallography. Finally, the effect of inhibitor binding was investigated by noncovalent mass spectrometry, thus providing the binding stoichiometries of Tgt:inhibitor complexes and showing competitive interactions in the presence of tRNA. Inhibitors that display an influence on the formation of the dimer interface in the crystal structure are promising candidates to alter the protein-protein interaction, which could provide a new way to inhibit Tgt.     

Human immunodeficiency virus type 1 Nef associates with lipid rafts to downmodulate cell surface CD4 and class I major histocompatibility complex expression and to increase viral infectivity

Lipid rafts are membrane microdomains that are functionally distinct from other membrane regions. We have shown that 10% of human immunodeficiency virus type 1 (HIV-1) Nef expressed in SupT1 cells is present in lipid rafts and that this represents virtually all of the membrane-associated Nef. To determine whether raft targeting, rather than simply membrane localization, has functional significance, we created a Nef fusion protein (LAT-Nef) containing the N-terminal 35 amino acids from LAT, a protein that is exclusively localized to rafts. Greater than 90% of the LAT-Nef protein was found in the raft fraction. In contrast, a mutated form, lacking two cysteine palmitoylation sites, showed less than 5% raft localization. Both proteins were equally expressed and targeted nearly exclusively to membranes. The LAT-Nef protein was more efficient than its nonraft mutant counterpart at downmodulating both cell surface CD4 and class I major histocompatibility complex (MHC) expression, as well as in enhancing first-round infectivity and being incorporated into virus particles. This demonstrates that targeting of Nef to lipid rafts is mechanistically important for all of these functions. Compared to wild-type Nef, LAT-Nef downmodulated class I MHC nearly as effectively as the wild-type Nef protein, but was only about 60% as effective for CD4 downmodulation and 30% as effective for infectivity enhancement. Since the LAT-Nef protein was found entirely in rafts while the wild-type Nef protein was distributed 10% in rafts and 90% in the soluble fraction, our results suggest that class I MHC downmodulation by Nef may be performed exclusively by raft-bound Nef. In contrast, CD4 downmodulation and infectivity enhancement may require a non-membrane-bound Nef component as well as the membrane-bound form.     

A novel dimer-tetramer transition captured by the crystal structure of the HIV-1 Nef

HIV-1 Nef modulates disease progression through interactions with over 30 host proteins. Individual chains fold into membrane-interacting N-terminal and C-terminal core (Nef_core) domains respectively. Nef exists as small oligomers near membranes and associates into higher oligomers such as tetramers or hexadecamers in the cytoplasm. Earlier structures of the Nef_core in apo and complexed forms with the Fyn-kinase SH3 domain revealed dimeric association details and the role of the conserved PXXP recognition motif (residues 72–78) of Nef in SH3-domain interactions. The crystal structure of the tetrameric Nef reported here corresponds to the elusive cytoplasmic stage. Comparative analyses show that subunits of Nef_core dimers (open conformation) swing out with a relative displacement of ∼22 Å and rotation of ∼174° to form the ‘closed’ tetrameric structure. The changes to the association are around Asp125, a conserved residue important for viral replication and the important XR motif (residues 107–108). The tetramer associates through C4 symmetry instead of the 222 symmetry expected when two dimers associate together. This novel dimer-tetramer transition agrees with earlier solution studies including small angle X-ray scattering, analytical ultracentrifugation, dynamic laser light scattering and our glutaraldehyde cross-linking experiments. Comparisons with the Nef_core—Fyn-SH3 domain complexes reveal that the PXXP motif that interacts with the SH3-domain in the dimeric form is sterically occluded in the tetramer. However the 151–180 loop that is distal to the PXXP motif and contains several protein interaction motifs remains accessible. The results suggest how changes to the oligomeric state of Nef can help it distinguish between protein partners.     

Structural models of the bovine papillomavirus E5 protein

The bovine papillomavirus E5 protein is thought to be a type II integral membrane protein that exists as a disulfide-linked homodimer in transformed cells. Polarized-infrared measurements show that the E5 dimer in membrane bilayers is largely α-helical and has a transmembrane orientation. Computational searches of helix-helix conformations reveal two possible low-energy dimer structures. Correlation of these results with previous mutagenesis studies on the E5 protein suggests how the E5 dimer may serve as a molecular scaffold for dimerization and ligand-independent activation of the PDGF-β receptor. We propose that on each face of the E5 dimer a PDGF-β receptor molecule interacts directly with Gln17 from one E5 monomer and with Asp33 from the other E5 monomer. This model accounts for the requirement of Gln17 and Asp33 for complex formation and explains genetic results that dimerization of the E5 protein is essential for cell transformation. Proteins 33:601–612, 1998. © 1998 Wiley-Liss, Inc.     

Amino acid substitution at the dimeric interface of human manganese superoxide dismutase

The side chains of His30 and Tyr166 from adjacent subunits in the homotetramer human manganese superoxide dismutase (Mn-SOD) form a hydrogen bond across the dimer interface and participate in a hydrogen-bonded network that extends to the active site. Compared with wild-type Mn-SOD, the site-specific mutants H30N, Y166F, and the corresponding double mutant showed 10-fold decreases in steady-state constants for catalysis measured by pulse radiolysis. The observation of no additional effect upon the second mutation is an example of cooperatively interacting residues. A similar effect was observed in the thermal stability of these enzymes; the double mutant did not reduce the major unfolding transition to an extent greater than either single mutant. The crystal structures of these site-specific mutants each have unique conformational changes, but each has lost the hydrogen bond across the dimer interface, which results in a decrease in catalysis. These same mutations caused an enhancement of the dissociation of the product-inhibited complex. That is, His30 and Tyr166 in wild-type Mn-SOD act to prolong the lifetime of the inhibited complex. This would have a selective advantage in blocking a cellular overproduction of toxic H2O2.     

Thermodynamic stability of the bacteriorhodopsin lattice as measured by lipid dilution

To determine the strength of noncovalent interactions that stabilize a membrane protein complex, we have developed an in vitro method for quantifying the dissociation of the bacteriorhodopsin (BR) lattice, a naturally occurring two-dimensional crystal. A lattice suspension was titrated with a short- and long-chain phosphatidylcholine mixture to dilute BR within the lipid bilayer. The fraction of BR in the lattice form as a function of added lipid was determined by visible circular dichroism spectroscopy and fit with a cooperative self-assembly model to obtain a critical concentration for lattice assembly. Critical concentration values of wild-type and mutant proteins were used to calculate the change in lattice stability upon mutation (DeltaDeltaG). By using this method, a series of mutant proteins was examined in which residues at the BR-BR interface were replaced with smaller amino acids, either Ala or Gly. Most of the mutant lattices were destabilized, with DeltaDeltaG values of 0.2-1.1 kcal/mol at 30 degrees C, consistent with favorable packing of apolar residues in the membrane. One mutant, I45A, was stabilized by approximately 1.0 kcal/mol, possibly due to increased lipid entropy. The DeltaDeltaG values agreed well with previous in vivo measurements, except in the case of I45A. The ability to measure the change in stability of mutant protein complexes in a lipid bilayer may provide a means of determining the contributions of specific protein-protein and protein-lipid interactions to membrane protein structure.