No abstract

Chemokines constitute a large family of small proteins that regulate leukocyte trafficking to the site of inflammation by binding to specific cell-surface receptors belonging to the GPCR superfamily. The interactions between N-terminal (Nt-) peptides of these GPCRs and chemokines have been studied extensively using NMR spectroscopy. However, due to lower affinities of peptides representing the three extracellular loops (ECLs) of chemokine receptors to their respective chemokine ligands, information concerning these interactions is scarce. To overcome the low affinity of ECL peptides to chemokines, we linked two or three CCR5 extracellular domains by either biosynthesis in Escherichia coli or by chemical synthesis. Using such chimeras, CCR5 binding to RANTES was followed using 1H-15N-HSQC spectra to monitor titration of the chemokine with peptides corresponding to the extracellular surface of the receptor. Nt-CCR5 and ECL2 were found to be the major contributors to CCR5 binding to RANTES, creating a nearly closed ring around this protein by interacting with opposing faces of the chemokine. A RANTES positively charged surface involved in Nt-CCR5 binding resembles the positively charged surface in HIV-1 gp120 formed by the C4 and the base of the V3. The opposing surface on RANTES, composed primarily of β2-β3 hairpin residues, binds ECL2 and was found to be analogous to a surface in the crown of the gp120 V3. The chemical and biosynthetic approaches for linking GPCR surface regions discussed herein should be widely applicable to investigation of interactions of extracellular segments of chemokine receptors with their respective ligands.

The large, catalytic subunits (LSUs; ilvB, ilvG and ilvI, respectively) of enterobacterial acetohydroxyacid synthases isozymes (AHAS I, II and III) have molecular weights approximately 60 kDa and are paralogous with a family of other thiamin diphosphate dependent enzymes. The small, regulatory subunits (SSUs) of AHAS I and AHAS III (ilvN and ilvH) are required for valine inhibition, but ilvN and ilvH can only confer valine sensitivity on their own LSUs. AHAS II is valine resistant. The LSUs have only approximately 15, <<1 and approximately 3%, respectively, of the activity of their respective holoenzymes, but the holoenzymes can be reconstituted with complete recovery of activity. We have examined the activation of each of the LSUs by SSUs from different isozymes and ask to what extent such activation is specific; that is, is effective nonspecific interaction possible between LSUs and SSUs of different isozymes? To our surprise, the AHAS II SSU ilvM is able to activate the LSUs of all three of the isozymes, and the truncated AHAS III SSUs ilvH-Delta80, ilvH-Delta86 and ilvH-Delta89 are able to activate the LSUs of both AHAS I and AHAS III. However, none of the heterologously activated enzymes have any feedback sensitivity. Our results imply the existence of a common region in all three LSUs to which regulatory subunits may bind, as well as a similarity between the surfaces of ilvM and the other SSUs. This surface must be included within the N-terminal betaalphabetabetaalphabeta-domain of the SSUs, probably on the helical face of this domain. We suggest hypotheses for the mechanism of valine inhibition, and reject one involving induced dissociation of subunits.

The envelope spike of HIV-1, which consists of three external gp120 and three transmembrane gp41 glycoproteins, recognizes its target cells by successively binding to its primary CD4 receptor and a coreceptor molecule. Until recently, atomic-resolution structures were available primarily for monomeric HIV-1 gp120, in which the V1, V2 and V3 variable loops were omitted (gp120 core ), in complex with soluble CD4 (sCD4). Differences between the structure of HIV gp120 core in complex with sCD4 and the structure of unliganded simian immunodeficiency virus gp120 core led to the hypothesis that gp120 undergoes a major conformational change upon sCD4 binding. To investigate the conformational flexibility of gp120, we generated two forms of mutated gp120 amenable for NMR studies: one with V1, V2 and V3 omitted ( mut gp120 core ) and the other containing the V3 region N]Ile-labeled unliganded mut gp120 core showed many fewer crosspeaks than the expected number, and also many fewer crosspeaks in comparison with the labeled mut gp120 core bound to the CD4-mimic peptide, CD4M33. This finding suggests that in the unliganded form, mut gp120 core shows considerable flexibility and motions on the millisecond time scale. In contrast, most of the expected crosspeaks were observed for the unliganded mut gp120 core (+V3), and only a few changes in chemical shift were observed upon CD4M33 binding. These results indicate that mut gp120 core (+V3) does not show any significant conformational flexibility in its unliganded form and does not undergo any significant conformational change upon CD4M33 binding, underlining the importance of V3 in stabilizing the gp120 core conformation.