Acetylcholinesterase H and T Dimers Are Associated through the Same Contact
2001; Elsevier BV; Volume: 276; Issue: 40 Linguagem: Inglês
10.1074/jbc.m103192200
ISSN1083-351X
AutoresN. Costedoat‐Chalumeau, Jacqueline Leroy, Annick Ayon, Jean Massoulié, Suzanne Bon,
Tópico(s)Phosphodiesterase function and regulation
ResumoAcetylcholinesterase (AChE) exists as AChEH and AChET subunits, which differ by their C-terminal H or T peptides, generating glycophosphatidylinositol-anchored dimers and various oligomers, respectively. We introduced mutations in the four-helix bundle interface of glycophosphatidylinositol-anchored dimers, and analyzed their effect on the production and oligomerization of AChEH, of AChET, and of truncated subunits, AChEC (without H or T peptide). Dimerization was reduced for all types of subunits, showing that they interact through the same contact zone; the formation of amphiphilic tetramers (Torpedo AChET) and 13.5 S oligomers (rat AChET) was also suppressed. Oligomerization appeared totally blocked by introduction of an N-linked glycan on the surface of helix α7,8. Other point mutations did not affect the synthesis or the catalytic properties of AChE but reduced or blocked the secretion of AChET subunits. Secretion of AChET was partially restored by co-expression with QN, a secretable protein containing a proline-rich attachment domain (PRAD); QN organized PRAD-linked tetramers, except for the N-glycosylated mutants. Thus, the simultaneous presence of an abnormal four-helix bundle zone and an exposed T peptide targeted the enzyme toward degradation, indicating a cross-talk between the catalytic and tetramerization domains. Acetylcholinesterase (AChE) exists as AChEH and AChET subunits, which differ by their C-terminal H or T peptides, generating glycophosphatidylinositol-anchored dimers and various oligomers, respectively. We introduced mutations in the four-helix bundle interface of glycophosphatidylinositol-anchored dimers, and analyzed their effect on the production and oligomerization of AChEH, of AChET, and of truncated subunits, AChEC (without H or T peptide). Dimerization was reduced for all types of subunits, showing that they interact through the same contact zone; the formation of amphiphilic tetramers (Torpedo AChET) and 13.5 S oligomers (rat AChET) was also suppressed. Oligomerization appeared totally blocked by introduction of an N-linked glycan on the surface of helix α7,8. Other point mutations did not affect the synthesis or the catalytic properties of AChE but reduced or blocked the secretion of AChET subunits. Secretion of AChET was partially restored by co-expression with QN, a secretable protein containing a proline-rich attachment domain (PRAD); QN organized PRAD-linked tetramers, except for the N-glycosylated mutants. Thus, the simultaneous presence of an abnormal four-helix bundle zone and an exposed T peptide targeted the enzyme toward degradation, indicating a cross-talk between the catalytic and tetramerization domains. acetylcholinesterase four-helix bundle phosphatidylinositol-specific phospholipase C glycophosphatidylinositol proline-rich attachment domain tryptophan amphiphilic tetramerization Acetylcholinesterase (AChE)1 is an essential component of the cholinergic synapse. This enzyme allows a precise control of cholinergic transmission by rapidly hydrolyzing the neurotransmitter, acetylcholine. In vertebrates, AChE is encoded by a single gene. In Torpedo and mammals, alternative splicing produces several types of subunits, mainly AChEH and AChET, which possess the same catalytic domain but distinct C-terminal peptides (Table I). The H and T (tryptophan amphiphilic tetramerization; WAT) peptides determine different post-transcriptional modifications and quaternary associations (1Massoulié J. Pezzementi L. Bon S. Krejci E. Vallette F.M. Prog. Neurobiol. 1993; 41: 31-91Crossref PubMed Scopus (1058) Google Scholar).Table IDifferent types of C-terminal peptides in cholinesterasesSAT represents the last three residues of the common catalytic domain. The underlined regions of H1 and H2 are removed from the mature GPI-anchored proteins; the last residue of the mature protein (ω) is shown in boldface type, and the C-terminal region is shown in italics. Conserved aromatic residues in the T peptide are shown in boldface letters; in H and T peptides, cysteines that may establish intercatenary disulfide bonds are shaded. In the C2-T construct, six residues from the H peptide (doubly underlined), including the proximal cysteine C6 were inserted upstream of the T peptide. The mutated residue in T1 is underlined. Open table in a new tab SAT represents the last three residues of the common catalytic domain. The underlined regions of H1 and H2 are removed from the mature GPI-anchored proteins; the last residue of the mature protein (ω) is shown in boldface type, and the C-terminal region is shown in italics. Conserved aromatic residues in the T peptide are shown in boldface letters; in H and T peptides, cysteines that may establish intercatenary disulfide bonds are shaded. In the C2-T construct, six residues from the H peptide (doubly underlined), including the proximal cysteine C6 were inserted upstream of the T peptide. The mutated residue in T1 is underlined. The H peptide contains one or two cysteines that are close to the catalytic domain and may form intercatenary disulfide bonds and also contains a signal for cleavage and the addition of a glycophosphatidylinositol (GPI). The AChEH subunits thus generate mature GPI-anchored amphiphilic dimers, which retain only a few residues beyond the catalytic domain. X-ray crystallography studies of Torpedo dimers showed that the contact zone between the two monomers is a “four-helix bundle” (FHB), formed by the α7,8 and α10 helices from each catalytic domain (Fig. 1); hydrophobic residues occupy a large proportion of this interface (2Sussman J.L. Harel M. Frolow F. Oefner C. Goldman A. Toker L. Silman I. Science. 1991; 253: 872-879Crossref PubMed Scopus (2440) Google Scholar). The same contact was observed in crystals of mouse AChE from which the C-terminal T peptide had been deleted and which was monomeric in solution (3Bourne Y. Taylor P. Marchot P. Cell. 1995; 83: 503-512Abstract Full Text PDF PubMed Scopus (319) Google Scholar). The T (WAT) peptide possesses a free cysteine, which is located near its C terminus, at a distance of 36 residues from the catalytic domain (Table I); this peptide forms an amphiphilic α helix, and its presence allows a variety of quaternary associations (1Massoulié J. Pezzementi L. Bon S. Krejci E. Vallette F.M. Prog. Neurobiol. 1993; 41: 31-91Crossref PubMed Scopus (1058) Google Scholar). The AChET subunits generate monomers, dimers, and tetramers. Tetramers of AChET subunits also assemble around structural proteins; the hydrophobic P protein anchors the enzyme in cell membranes of the central nervous system (4Gennari K. Brunner J. Brodbeck U. J. Neurochem. 1987; 49: 12-18Crossref PubMed Scopus (68) Google Scholar, 5Inestrosa N.C. Roberts W.L. Marshall T.L. Rosenberry T.L. J. Biol. Chem. 1987; 262: 4441-4444Abstract Full Text PDF PubMed Google Scholar), and collagen ColQ attaches it to the basal lamina of neuromuscular junctions (6Krejci E. Thomine S. Boschetti N. Legay C. Sketelj J. Massoulié J. J. Biol. Chem. 1997; 272: 22840-22847Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar, 7Feng G. Krejci E. Molgo J. Cunningham J.M. Massoulié J. Sanes J.R. J. Cell Biol. 1999; 144: 1349-1360Crossref PubMed Scopus (150) Google Scholar). Isolated T peptides can associate with a short “proline-rich attachment domain” (PRAD) of the N-terminal region of ColQ (8Bon S. Massoulié J. J. Biol. Chem. 1997; 272: 3007-3015Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, 9Bon S. Coussen F. Massoulié J. J. Biol. Chem. 1997; 272: 3016-3021Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar); the C-terminal T peptide therefore behaves as an autonomous interaction domain and has been named the WAT domain (10Simon S. Krejci E. Massoulié J. EMBO J. 1998; 17: 6178-6187Crossref PubMed Scopus (76) Google Scholar). This suggests that the catalytic domain of AChE does not play a major part in the assembly of an AChET tetramer around the PRAD. X-ray crystallography of Electrophorus AChETtetramers, derived from collagen-tailed molecules, showed that the subunits are organized in a skewed quasi-square arrangement (11Bourne Y. Grassi J. Bougis P.E. Marchot P. J. Biol. Chem. 1999; 274: 30370-30376Abstract Full Text Full Text PDF PubMed Scopus (147) Google Scholar,24Raves M. Structure-Function Studies on Acetylcholinesterase in Space and Time.Ph.D. thesis. Weizmann Institute, Rehovot, Israel1998Google Scholar). However, the structure has not been solved at a sufficient resolution to identify the contact zones between subunits. In a tentative model, K. Giles proposed to call these contacts the weak hydrophobic interaction, corresponding to the FHB that exists in AChEH dimers, and the strong hydrophobic interaction, based on the T (WAT) peptides (12Giles K. Protein Eng. 1997; 10: 677-685Crossref PubMed Scopus (28) Google Scholar). If these denominations really reflect the relative strength of the quaternary interactions, AChETdimers should be associated through strong hydrophobic interaction rather than weak hydrophobic interaction. AChEH and AChET dimers would therefore be organized differently, and this appears consistent with the fact that the intercatenary disulfide bonds are not located at the same distance from the catalytic domains in the two types of subunits. We introduced mutations in the α7,8 and α10helices of Torpedo and rat AChEs to modify the FHB interaction by steric hindrance, reduction of hydrophobic interactions, or electrostatic repulsion. We analyzed their effects on the biogenesis of AChE and on its molecular forms. In Torpedo, we studied AChEH and AChET subunits; in rat, we also analyzed truncated molecules, AChEC, which were derived from AChEH subunits by deletion of the GPI addition signal and produced secreted enzyme. To assess the importance of intercatenary disulfide bonds, we mutated the cysteines of H and T peptides; in the case of AChET subunits, we also introduced a proximal cysteine, as in the H peptide. The resulting C-terminal peptides are shown in Table I. AChET subunits were expressed alone or with an N-terminal fragment of ColQ, called QN, that contains the PRAD and consequently induces the organization of soluble PRAD-linked tetramers, which are readily secreted (8Bon S. Massoulié J. J. Biol. Chem. 1997; 272: 3007-3015Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, 9Bon S. Coussen F. Massoulié J. J. Biol. Chem. 1997; 272: 3016-3021Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar). This strategy was expected to reveal whether AChET dimers are based on the same interactions as AChEH or on another type of quaternary contact and also to provide some information on the respective roles of the catalytic domain and of the T peptide (WAT) in the formation of AChET dimers and of tetramers, both homomeric and associated with a PRAD. Our results demonstrate that AChET dimers are linked through the FHB, like AChEH dimers. In addition, we found that FHB mutations could direct AChE toward degradation rather than secretion; this was much more marked for AChET than for AChEH or AChEC. In contrast with wild type AChETsubunits, the mutant subunits were not secreted to any significant extent, except when they were engaged in PRAD-linked tetramers. This suggests a cross-talk between two independent functional domains of AChE, the catalytic domain (FHB mutations), and the tetramerization domain (WAT peptide). Torpedo and rat AChEH and AChET were expressed in the pEF-BOS vector (13Legay C. Bon S. Vernier P. Coussen F. Massoulié J. J. Neurochem. 1993; 60: 337-346Crossref PubMed Scopus (99) Google Scholar). Mutagenesis was performed by the method of Kunkel et al.(14Kunkel T.A. Roberts J.D. Zakour R.A. Methods Enzymol. 1987; 154: 367-382Crossref PubMed Scopus (4560) Google Scholar). Truncated rat AChEC1 and AChEC2 subunits were obtained by introducing stop codons immediately after the catalytic domain or at the position of the second cysteine of AChEH, respectively. The C2-T construct was obtained by inserting the first six residues of H (ATEVPC) upstream of the rat T peptide. COS cells were transfected by the DEAE-dextran method, as previously described (8Bon S. Massoulié J. J. Biol. Chem. 1997; 272: 3007-3015Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar), using 5 μg of DNA encoding the AChE catalytic subunit per 100-mm dish. Co-expression of AChETwith QN/stop 551 (8Bon S. Massoulié J. J. Biol. Chem. 1997; 272: 3007-3015Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar) was obtained by adding 3 μg of DNA encoding this protein. After transfection, COS cells were incubated for 1 day at 37 °C, followed by 2 days at 27 °C in the case ofTorpedo AChE, or for 2–3 days at 37 °C in the case of rat AChE, in a medium containing 10% Nuserum (Inotech, Dottikon, Switzerland), which had been pretreated with 10−5m soman to inactivate serum cholinesterases. The cells were extracted with TMg buffer (1% Triton X-100, 50 mmTris-HCl, pH 7.5, 10 mm MgCl2) at 20 °C to solubilize the GPI-anchored molecules. For PI-PLC digestion, samples of detergent extracts (50 μl) were incubated in TMg buffer for 1 h at 30 °C with a 1% volume (0.025 IU) of PI-PLC from Bacillus thuringiensis (Immunotech (Marseille, France or Glyko Europe, Upper Heyford, UK)). Electrophoretic analyses in nondenaturing conditions and evaluation of PI-PLC-sensitive and -resistant components were performed as previously described (15Coussen F. Ayon A. Le Goff A. Leroy J. Massoulié J. Bon S. J. Biol. Chem. 2001; 276: 27881-27894Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). AChE activity was determined by the colorimetric method of Ellman et al. (16Ellman G.L. Courtney K.D. Andres V. Featherstone R.M. Biochem. Pharmacol. 1961; 7: 88-95Crossref PubMed Scopus (21721) Google Scholar) at room temperature (20 °C). Enzyme samples (10 μl) were added to 0.2 ml of Ellman assay medium in microplates, and the reaction was monitored at 414 nm with a Labsystems Multiskan RC automatic plate reader (Helsinki, Finland); the optical density was recorded at 20-s intervals over a period of 5 min. Two days after transfection, COS cells were preincubated for 45 min in Dulbecco's modified Eagle's medium without cysteine and methionine and labeled for 30 min with 50–100 μCi/ml [35S]methionine/cysteine (Amersham Pharmacia Biotech). After labeling, the cells were rinsed with phosphate-buffered saline and chased in medium containing Nuserum; at various times, the medium was collected, and the cells were extracted in a buffer containing 20 mm sodium borate, pH 9, 1m NaCl, 1 mm EDTA, 1% Triton X-100, 0.1 mg/ml bovine serum albumin, and a protease inhibitor mixture (Sigma; reference number P8340); the cell extracts were centrifuged at 13,000 rpm for 30 min. The cell extracts and the media were incubated overnight at 8 °C with a 1:300 dilution of the polyclonal anti-rat AChE rabbit antiserum A63 under rotatory agitation (17Marsh D. Grassi J. Vigny M. Massoulié J. J. Neurochem. 1984; 43: 204-213Crossref PubMed Scopus (113) Google Scholar) and then mixed with 20 μl of a suspension of protein A-Sepharose 4B beads (Sigma) and incubated for 3 h at 8 °C. The beads were rinsed three times with extraction buffer, and the immunoprecipitated AChE sample was analyzed by electrophoresis under reducing and denaturing conditions (SDS-polyacrylamide gel electrophoresis), using a Fuji image analyzer (BAS 1000). Centrifugation was performed in 5–20% sucrose gradients (50 mm Tris-HCl, pH 7.5, 50 mm MgCl2, either in the presence of 1% Brij or in the presence of 0.2% Triton X-100) in a Beckman SW41 rotor, at 36,000 rpm, for 18 h at 6 °C. The gradients containedEscherichia coli β-galactosidase (16 S) and alkaline phosphatase (6.1 S) as internal sedimentation standards. Electrophoresis in nondenaturing polyacrylamide gels were performed as described previously (18Bon S. Toutant J.P. Méflah K. Massoulié J. J. Neurochem. 1988; 51: 786-794Crossref PubMed Scopus (57) Google Scholar), and AChE activity was revealed by the histochemical method of Karnovsky and Roots (19Karnovsky M.J. Roots L. J. Histochem. Cytochem. 1964; 12: 219-222Crossref PubMed Scopus (2999) Google Scholar). 100 μl of medium containing secreted AChE was incubated at 47 °C. At appropriate time intervals, 10-μl aliquots were withdrawn and stored on ice. Their residual activity was assayed at 20 °C by the Ellman method within 1 h. Several dishes of transfected cells, expressing AChE at the steady state, were incubated for 30 min in the presence of 2 × 10−6 msoman, a membrane-permeant, irreversible inhibitor of cholinesterases. They were then extensively washed and incubated in fresh medium. The cellular activities were determined from individual dishes over a period of 3 h. Fig. 1 was prepared with the GRASP software (20Nicholls A. Sharp K.A. Honig B. Proteins Struct. Funct. Genet. 1991; 11: 281-296Crossref PubMed Scopus (5318) Google Scholar). The presence of anN-glycan in the zone of contact between two subunits was expected to weaken or prevent the formation of dimers of AChEH subunits. We introduced a potentialN-glycosylation site at the surface of the α10helix, in the dimeric contact zone, with the double mutation F527N/P529S. The mutant protein was efficiently synthesized by COS cells, as judged by intracellular immunofluorescence of permeabilized cells with an anti-Torpedo AChE antiserum, but the protein remained inactive, and it was not externalized at the plasma membrane. This suggests that the presence of anN-glycosylation site on the α10 helix interfered with the folding of the enzyme. We also added a potential N-glycosylation site at position370, on the α7,8 helix, by mutation A370N (a single mutation was sufficient, because residue 372 is a threonine). Mutations A370S and A370Q were used as controls. Mutation A370N did not abolish the production of active AChEH or AChET subunits but reduced it to about 1% of the wild type; mutations A370S and A370Q produced 100 and 50%, respectively, of the wild type level for AChEH and 15 and 5% for AChET. Like the wild type, the A370S mutant AChEHsubunits produced GPI-anchored dimers (G2a) and a low level of secreted nonamphiphilic dimers; in contrast, the A370Q mutant produced mainly GPI-anchored monomers. The proportion of dimers, determined from sedimentation profiles of cell extracts, decreased in the same order as the level of activity: wild type > A370S > A370Q ≫ A370N. The fact that the levels of activity and dimerization were much more affected when alanine 370 was replaced by an asparagine (A370N) than by a more bulky residue (A370Q) suggests that the N-glycosylation site was actually used. In the case of AChET subunits, the wild type and mutant A370S produced mostly dimers, but mutant A370Q produced essentially monomers (Fig. 2). Both mutations suppressed amphiphilic tetramers; in contrast, nonamphiphilic tetramers were only slightly reduced by mutation A370Q. Co-expression of the wild type and mutant AChET subunits with the QN fragment of the ColQ collagen, which contains the PRAD (9Bon S. Coussen F. Massoulié J. J. Biol. Chem. 1997; 272: 3016-3021Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar), induced the assembly and the secretion of PRAD-associated tetramers. In addition, it increased the total production of activity for mutant A370S and to a lesser extent for mutant A370Q, showing that the mutated subunits retained the capacity to associate with the PRAD and were partially rescued by this association. There was no increase for A370N, showing that N-glycosylation prevented the tetramerization of AChET subunits. The three mutations of Ala370 had stronger effects on AChET subunits than on AChEH subunits, reducing dimerization, further oligomerization, and also the levels of activity and secretion. We wondered whether this was related to the fact that the intercatenary disulfide bond in AChET dimers and other oligomers is more distal than in AChEH dimers. We therefore introduced the double mutation I538C/C572S, to replace the distal cysteine by a proximal one. In the absence of other mutations, this increased the level of activity by about 30%. In the case of the A370Q mutant, the total activity and the proportion of secreted AChE remained low but were approximately doubled. The levels of cellular and secreted activity were not increased for the A370N mutant. Thus, the fact that FHB mutations affected AChET subunits more dramatically than AChEH subunits was not simply due to the presence of a distal or proximal cysteine, respectively. We then examined whether these results could be generalized to mammalian AChE, in particular because it is more abundantly secreted. To introduce anN-glycosylation site in the α7,8 helix of rat AChE, we constructed the double mutant A370N/V372T (in this paper, we use theTorpedo numbering (21Massoulié J. Sussman J.L. Doctor B.P. Soreq H. Velan B. Cygler M. Rotundo R.L. Shafferman A. Silman I. Taylor P. Shafferman A. Velan B. Multidisciplinary Approaches to Cholinesterase Functions. Plenum Press, New York1992: 285-288Google Scholar)). Mutants A370N and A370Q/V372T were made as nonglycosylable controls. To assess the role of hydrophobic interactions in the FHB, we replaced phenylalanine 527 in the α10 helix by a leucine, an alanine, or a glutamic acid. As shown in Table I, we analyzed the impact of mutations in the FHB contact zone on several types of subunits, characterized by the presence of the C-terminal H peptide (subunits H1 and H2), of a truncated H peptide (subunits C1 and C2), and of the T peptide (subunits T1 and T2); in subunits H1, C1, and T1, the cysteines that may be involved in intercatenary disulfide bonds were mutated or deleted. In the case of T subunits, we replaced cysteine 572 by a cysteine in a proximal position, by point mutations (L538C/C572S) in T1L538C subunits. We also constructed C2-T subunits, in which a fragment of the H peptide, containing the first cysteine, was inserted upstream of the T peptide (Table I). In the case of wild type C2 subunits, the sedimentation profiles obtained for both the cells and the medium showed a predominant peak corresponding to nonamphiphilic dimers, with a small proportion of monomers (Fig. 3,upper row). In the mutants, the cellular active enzyme was almost exclusively monomeric, indicating that the mutations indeed compromised dimerization. The secreted enzyme was partly dimeric, and the proportion of dimers decreased in the following order: wild type > F527L > A370N, F527E, F527A > A370Q/V372T ≫ A370N/V372T. Clearly, the presence of anN-glycosylation site had a much stronger effect than the other mutations on the production of dimers. This analysis was confirmed by electrophoresis in nondenaturing conditions (not shown); the wild type C2 AChE presented a major dimeric component, while mutants showed variable proportions of a faster migrating band, corresponding to monomeric AChE. This species represented the low level of activity produced by the A370N/V372T mutant. H1 and H2 subunits produced GPI-anchored AChE forms, which were characterized by their sensitivity to PI-PLC. The cellular extracts and the medium also contained amphiphilic uncleaved molecules and nonamphiphilic lytic molecules (15Coussen F. Ayon A. Le Goff A. Leroy J. Massoulié J. Bon S. J. Biol. Chem. 2001; 276: 27881-27894Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). The level of activity was considerably lower in the case of the A370N/V372T mutant, as discussed below, and this mutant produced an unusual nonamphiphilic species. For all other mutants, quantitative scanning of electrophoretic patterns showed that the cells contained the same proportion of GPI-anchored enzyme as for the wild type, in agreement with the fact that they possessed the same GPI addition signal. Sedimentation analyses (not shown) and electrophoretic patterns (Fig.4) showed that the wild type H2 subunits produced mostly GPI-anchored dimers, while the mutants produced variable proportions of GPI-anchored monomers and dimers (not shown). In the medium, the proportions of secreted monomers and dimers were similar to those obtained with the corresponding C2 subunits. These results confirm the involvement of the two α helices in the dimerization of glypiated H subunits. Sedimentation and electrophoretic analyses showed that, surprisingly, the wild type H1 subunits did not behave as simple monomers like the C1 subunits, although they could not form disulfide-linked dimers; instead, they appeared to dimerize reversibly, in a detergent-dependent manner. The sedimentation profiles of cell extracts showed a single peak, corresponding to monomers, in gradients containing Triton X-100 (Fig.5A) but two peaks in the presence of Brij-96, suggesting that the enzyme was partially dimerized. The proportion of the faster sedimenting component increased with the concentration of enzyme in the sample that was layered on the gradient (Fig. 5, A and C). When gradient fractions corresponding to each of these components were diluted and analyzed in a second gradient, they showed again a similar bimodal distribution (Fig. 5D). The bimodal sedimentation pattern therefore reflects a reversible association-dissociation equilibrium between the two species. In contrast, fractions corresponding to monomers and dimers produced by mutant H2 subunits sedimented as homogenous, distinct species when analyzed in a second gradient (not shown). The existence of a reversible equilibrium between two components, in the case of wild type H1 subunits, was confirmed by nondenaturing electrophoresis of gradient fractions; all fractions produced the same broad distribution, overlapping the bands obtained for the monomers and dimers, produced by mutant H2 AChE (not shown). The lytic molecules obtained from GPI-anchored H1 subunits after treatment with PI-PLC sedimented as homogenous monomers, similar to those produced by C1 subunits; in both cases, nonamphiphilic G1na forms, produced either directly by C1 or after PI-PLC digestion of H1, did not generate reversible dimers under our experimental conditions (Fig. 5B). Together with the fact that reversible dimerization was observed in the presence of Brij-96, but not of Triton X-100, this suggests that the noncovalent interaction of GPI-anchored H1 subunits was dependent on their interaction with detergent micelles. The GPI-anchored H1 mutants did not show any significant dimerization in sedimentation and electrophoretic analyses, under similar experimental conditions, indicating that the FHB interaction is required for this reversible association. Sedimentation analyses (Fig. 3, lower row) and nondenaturing electrophoresis (Fig.6) showed that in the case of the wild type rat T2 subunits, the cell extracts contained mainly monomers, together with smaller proportions of dimers, nonamphiphilic tetramers, and 13.5 S aggregates, whereas in the case of Torpedo they contained mostly dimers and amphiphilic tetramers. COS cells expressing the rat T2 subunits secreted similar amounts of monomers and dimers, together with some nonamphiphilic tetramers. All mutations, including F527L, suppressed the 13.5 S component, which seems to be intrinsically unstable, since it readily dissociates in the presence of Triton X-100 (8Bon S. Massoulié J. J. Biol. Chem. 1997; 272: 3007-3015Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar). The F527L mutation had little effect on dimerization. The other mutations suppressed the formation of dimers, as shown by electrophoretic patterns (Fig. 6). However, the sedimentation profiles still showed a broad 4.5 S component, sedimenting like the wild type G2aform, in the presence of Brij-96 (Fig. 3). An analysis of the corresponding fractions by nondenaturing electrophoresis confirmed the identity of the 3 S amphiphilic monomers (G1a) in the wild type and in the mutants and of the 4.5 S amphiphilic dimers (G2a) in the wild type. In the mutants, the 4.5 S component was found to contain nonamphiphilic molecules, as illustrated in the case of F527A (Fig.7). These molecules did not interact with detergent micelles and were not assembled into tetramers by co-expression with QN (Fig. 6), but they differed in their electrophoretic mobility from the nonamphiphilic monomers produced by C1 subunits. All FHB mutants, except A370N/V372T, produced some G4na tetramers, even when the formation of dimers and 13.5 S oligomers was blocked; different quaternary interactions are obviously responsible for the assembly and stability of dimers and of nonamphiphilic tetramers, as already suggested in the case of Torpedo. To assess the importance of an intercatenary disulfide bond and of its position for dimerization, we analyzed T1, T1L538C, and C2-T subunits. We found that the T1 subunits produced mainly monomers, together with a small proportion of tetramers, but no dimers, in agreement with previous findings on human AChE (22Velan B. Grosfeld H. Kronman C. Leitner M. Gozes Y. Lazar A. Flashner Y. Marcus D. Cohen S. Shafferman A. J. Biol. Chem. 1991; 266: 23977-23984Abstract Full Text PDF PubMed Google Scholar); in contrast, the T1L538C and C2-T subunits produced a higher proportion of dimers than the T2 subunits, particularly for the wild type (Fig. 8), showing that a proximal cysteine can more efficiently form an intercatenary disulfide bond. We examined whether T1 monomers could reversibly associate into covalent dimers, because they interact with detergent micelles, although their electrophoretic migration differs from that of GPI-anchored monomers produced from H1 subunits. We could not detect any sign of T1 dimers, either in sedimentation or in nondenaturing electrophoresis, under the same experimental conditions
Referência(s)