Factors That Influence the Homo- and Hetero-Oligomerization of the Leukocyte Integrin Transmembrane Domains

Abstract Integrins are a large family of transmembrane α/β heterodimers that subserve various cell-cell, cell-matrix, and cell-pathogen interactions. Integrins are present on cell surfaces in a dynamic equilibrium between inactive and active conformations that can be regulated by interactions involving their transmembrane (TM) domains. Thus, in their inactive conformation, the α and β subunit TM domains interact, whereas the TM domains separate when integrins assume their active conformations. The role of TM domains in regulating integrin function has been extensively studied using the platelet integrin αIIbβ3 and has revealed that the αIIb and β3 subunit TM domains interact in both a heteromeric and homomeric fashion. These observations have led to a “push-pull” hypothesis for regulating integrin function: events that disrupt the heteromeric association of integrin TM domains stabilize the active integrin conformation and push the equilibrium toward the activated state, whereas interactions that require separation, or are more favorable when the TMs are separated, pull the equilibrium toward the activated state. Although the factors that influence the homo- and hetero-association of the platelet αIIb and β3 TM domains have been well studied, the “push-pull” hypothesis for regulating integrin function has yet to be applied to other integrins, specifically to the leukocyte integrins αMβ2 (MAC-1) and αLβ2 (LFA-1). To determine the importance of TM interactions in αMβ2 and αLβ2, we used modified ToxR-based assays, ToxRed and Dominant-Negative (DN)-ToxRed, to measure the specificity and strength of the homomeric and heteromeric associations of the αM, αL, and β2 TM domains in a bacterial membrane. ToxR is a single-pass TM transcriptional factor from V. cholera that activates the cholera toxin (ctx) promoter when it undergoes TM-driven dimerization in the inner membrane of E. coli. By co-expressing ToxR containing a β2 integrin TM domain with either wild-type ToxR or a R96K ToxR mutant that can dimerize but is unable to activate the ctx promoter, we can measure both the homomeric and heteromeric interaction of each TM domain. In contrast to the TM domains of β3-containing integrins that preferentially form homooligomers, the tendency of the αM, αL, and β2 TM domains to form homo- and heterooligomers is nearly the same. To identify residues that are important for the homo- and heterooligomerization of these TM domains, each was scanned with leucine replacements. A small residue motif that lies along one face of the αM TM helix, similar to that identified in the αIIb TM domain, was found to mediate αM TM domain interactions. Based on these observations, we designed CHAMP (Computed Helical Anti-Membrane Peptide) peptides to specifically target the αM TM interface. In the CHAMP method, the backbone conformation for an αM-binding peptide is selected based on known structural preferences for the motifs identified in the αM helix. The sequence for the binding peptide is then designed computationally using a side-chain repacking algorithm. Ten anti-αM CHAMP peptides were designed and tested for their ability to bind to the αM TM domain in the DN-ToxRed assay. All ten designed sequences interacted with the αM TM domain to an equivalent or better extent than the natural αM and β2 TM domain sequences. Thus, our studies indicate that like the TM domains of β3-containing integrins, the TM domains of the β2 integrins undergo both homomeric and heteromeric interactions, making use of canonical TM helix interaction motifs. Further, we have demonstrated the successful application of computational methods to design peptides that selectively target the αM TM domain, providing a novel method to address mechanistic questions about β2 integrin structure and function.