Search

Recently, a two-dimensional crystal structure of NhaA, the Na+/H+ antiporter of Escherichia coli has been obtained [Williams, K. A., Kaufer, U. G., Padan, E., Schuldiner, S. and Kühlbrandt, W. (1999) EMBO J., 18, 3558-3563]. In these crystals NhaA exists as a dimer. Using biochemical and genetic approaches here we show that NhaA exists in the native membrane as a homooligomer. Functional complementation between the polypeptides of NhaA was demonstrated by coexpression of pairs of conditional lethal (at high pH in the presence of Na+) mutant alleles of nhaA in EP432, a strain lacking antiporters. Physical interaction in the membrane was shown between the His-tagged NhaA polypeptide which is readily affinity purified from DM-solubilized membranes with a Ni2+-NTA column and another which is not; only when coexpressed did both copurify on the column. The organization of the oligomer in the membrane was studied in situ by site-directed cross-linking experiments. Cysteine residues were introduced--one per NhaA--into certain loops of Cys-less NhaA, so that only intermolecular cross-linking could take place. Different linker-size cross-linkers were applied to the membranes, and the amount of the cross-linked protein was analyzed by mobility shift on SDS-PAGE. The results are consistent with homooligomeric NhaA and the location of residue 254 in the interface between monomers. Intermolecular cross-linking of V254C caused an acidic shift in the pH profile of NhaA.

pH and Na(+) homeostasis in all cells requires Na(+)/H(+) antiporters. The crystal structure of NhaA, the main antiporter of Escherichia coli, has provided general insights into antiporter mechanisms and their pH regulation. Functional studies of NhaA in the membrane have yielded valuable information regarding its functionality in situ at physiological pH. Here, we Cys-scanned the discontinuous transmembrane segment (TM) IV (helices IVp and IVc connected by an extended chain) of NhaA to explore its functionality at physiological pH. We then tested the accessibility of the Cys replacements to the positively charged SH reagent [2-(trimethylammonium)ethyl] methanethiosulfonate bromide (MTSET) and the negatively charged 2-sulfonatoethyl methanethiosulfonate (MTSES) in intact cells at pH 8.5 and 6.5 and in parallel tested their accessibility to MTSET in high-pressure membranes at both pH values. We found that the outer membrane of E. coli TA16 acts as a partially permeable barrier to MTSET. Overcoming this technical problem, we revealed that (a) Cys replacement of the most conserved residues of TM IV strongly increases the apparent K(m) of NhaA to both Na(+) and Li(+), (b) the cationic passage of NhaA at physiological pH is lined by the most conserved and functionally important residues of TM IV, and (c) a pH shift from 6.5 to 8.5 induces conformational changes in helix IVp and in the extended chain at physiological pH.

The unique trypsin cleavable site of NhaA, the Na + /H + antiporter of Escherichia coli, was exploited to detect a change in mobility of cross-linked products of NhaA by polyacrylamide gel electrophoresis. Double-Cys replacements were introduced into loops, one on each side of the trypsin cleavage site (Lys 249). The proximity of paired Cys residues was assessed by disulfide cross-linking of the two tryptic fragments, using three homobifunctional cross-linking agents: 1,6-bis(maleimido)hexane (BMH), N,N'-o-phenylenedimaleimide (o-PDM), and N,N'-p-phenylenedimaleimide (p-PDM). The interloop cross-linking was found to be very specific, indicating that the loops are not merely random coils that interact randomly. In the periplasmic side of NhaA, two patterns of cross-linking are observed: (a) all three cross-linking reagents cross-link very efficiently between the double-Cys replacements A118C/ S286C, N177C/S352C, and H225C/S352C; (b) only BMH cross-links the double-Cys replacements A118C/ S352C, N177C/S286C, and H225C/S286C, In the cytoplasmic side of NhaA, three patterns of cross-linking are observed: (a) all three cross-linking reagents cross-link very efficiently the pairs of Cys replacements L4C/E252C, S146C/L316C, S146C/R383C, and E241C/E252C; (b) BMH and p-PDM cross-link efficiently the pairs of Cys replacements S87C/E252C, S87C/L316C, and S146C/E252C; (c) none of the reagents cross-links the double-Cys replacements L4C/L316C, L4C/R383C, S87C/R383C, A202C/ E252C, A202C/L316C, A202C/R383C, E241C/L316C, and E241C/R383C. The data reveal that the N-terminus and loop VIII-IX that have previously been shown to change conformation with pH are in close proximity within the NhaA protein. The data also suggest close proximity between N-terminal and C-terminal helices at both the cytoplasmic and the periplasmic face of NhaA.

We have previously shown that the activity of NhaA is regulated by pH and found mutations that affect dramatically the pH dependence of the rate but not the K(m) (for Na(+) and Li(+)) of NhaA. In the present work, we found that helix IV is involved both in ion translocation as well as in pH regulation of NhaA. Two novel types of NhaA mutants were found clustered in trans membrane segment (TMS) IV: One type (D133C, T132C, and P129L) affects the apparent K(m) of NhaA to the cations with no significant effect on the pH profile of the antiporter; no shift of the pH profile was found when the activity of these mutants was measured at saturating Na(+) concentration. In contrast, the other type of mutations (A127V and A127T) was found to affect both the K(m) and the pH dependence of the rate of NhaA whether tested at saturating Na(+) concentration or not. These results imply that residues involved in the ion translocation of NhaA may (A127) or may not (D133, T132, and P129) overlap with those affecting the pH response of the antiporter. All mutants cluster in the N-terminal half of the putative alpha-helix IV, one type on one face, the other on the opposite. Cys accessibility test demonstrated that although D133C is located in the middle of TMS IV, it is inhibited by N-ethylmaleimide and is exposed to the cytoplasm.

We have isolated from nitrous acid cleavage products of heparin two major octasaccharide fragments which bind with high affinity to human antithrombin. Octasaccharide S, with the predominant structure iduronic acid----N-acetylglucosamine 6-O-sulfate----glucuronic acid-----N-sulfated glucosamine 3,6-di-O-sulfate----iduronic acid 2-O-sulfate----N-sulfated glucosamine 6-O-sulfate----iduronic acid 2-O-sulfate----anhydromannitol 6-O-sulfate, is sensitive to cleavage by Flavobacterium heparinase as well as platelet heparitinase and binds to antithrombin with a dissociation constant of (5-15) X 10(-8) M. Octasaccharide R, with the predominant structure iduronic acid 2-O-sulfate----N-sulfated glucosamine 6-O-sulfate----iduronic acid----N-acetylglucosamine 6-O-sulfate----glucuronic acid----N-sulfated glucosamine 3,6-di-O-sulfate----iduronic acid 2-O-sulfate----anhydromannitol 6-O-sulfate, is resistant to degradation by both enzymes and binds antithrombin with a dissociation constant of (4-18) X 10(-7) M. The occurrence of a 15-17% replacement of N-sulfated glucosamine 3,6-di-O-sulfate with N-sulfated glucosamine 3-O-sulfate and a 10-12% replacement of iduronic acid with glucuronic acid in both octasaccharides indicates that these substitutions have little or no effect on the binding of the oligosaccharides to the protease inhibitor. When bound to antithrombin, both octasaccharides produce a 40% enhancement in the intrinsic fluorescence of the protease inhibitor and a rate of human factor Xa inhibition of 5 X 10(5) M-1 s-1 as monitored by stopped-flow fluorometry. This suggests that the conformation of antithrombin in the region of the factor Xa binding site is similar when the protease inhibitor is complexed with either octasaccharide.

The individual temperature dependencies of the process which control the activity of turkey erythrocyte adenylate cyclase have been determined. The temperature dependence of the fraction of activable cyclase units experiences a thermal transition at 24 degrees C for all three modes of enzyme activation: l-epinephrine, adenosine, and NaF. This thermal transition probably reflects the phase transition in the inner monolayer of the membrane which influences the behavior of the GTP regulatory unit which is involved in all three modes of enzyme activation. The "rate constant" of enzyme activation by adenosine reflects two thermal transitions, at 24 and at 35 degrees C; the apparent rate constant of cyclase activation by NaF activation experiences a transition only at 24 degrees C whereas the rate constant of the beta-receptor-bound agonist decreases monotonously with no "breaks" on the Arrhenium plot. Following the temperature dependence of the fluorescence intensity of dansylphosphatidylethanolamine embedded in both sides of the membrane and exclusively in the outer monolayer, one can assign the thermal transition of 24 degrees C to the inner monolayer and the other two transitions to the outer monolayer (10 and 35 degrees C). We interpret these results as follows. (a) The monomolecular rate constant characterizing the activation of cyclase by the precoupled adenosine receptor experiences both the transition at 24 and 35 degrees C, indicating that the latter may span the bilayer. (b) The bata receptor activates the cyclase units only in fluid areas since it can diffuse exclusively in the fluid areas of the membrane and is unable to interact with cyclase units in "frozen" areas. the linear dependence of the logarithm of the rate constant on 1/T for the bata receptor reflects the change of membrane fluidity as a function of temperature.