M-calpain-mediated cleavage of GAP-43 near Ser41 is negatively regulated by protein kinase C, calmodulin and calpain-inhibiting fragment GAP-43-3

Neuronal protein GAP-43 performs multiple functions in axon guidance, synaptic plasticity and regulation of neuronal death and survival. However, the molecular mechanisms of its action in these processes are poorly understood. We have shown that in axon terminals GAP-43 is a substrate for calcium-activated cysteine protease m-calpain, which participates in repulsion of axonal growth cones and induction of neuronal death. In pre-synaptic terminals in vivo, in synaptosomes, and in vitro, m-calpain cleaved GAP-43 in a small region near Ser41, on either side of this residue. In contrast, micro-calpain cleaved GAP-43 in vitro at several other sites, besides Ser41. Phosphorylation of Ser41 by protein kinase C or GAP-43 binding to calmodulin strongly suppressed GAP-43 proteolysis by m-calpain. A GAP-43 fragment, lacking about forty N-terminal residues (named GAP-43-3), was produced by m-calpain-mediated cleavage of GAP-43 and inhibited m-calpain, but not micro-calpain. This fragment prevented complete cleavage of intact GAP-43 by m-calpain as a negative feedback. GAP-43-3 also blocked m-calpain activity against casein, a model calpain substrate. This implies that GAP-43-3, which is present in axon terminals in high amount, can play important role in regulation of m-calpain activity in neurons. We suggest that GAP-43-3 and another (N-terminal) GAP-43 fragment produced by m-calpain participate in modulation of neuronal response to repulsive and apoptotic signals.     

Palmitoylation alters protein activity: blockade of G(o) stimulation by GAP-43.

The addition of palmitate to cysteine residues enhances the hydrophobicity of proteins, and consequently their membrane association. Here we have investigated whether this type of fatty acylation also regulates protein-protein interactions. GAP-43 is a neuronal protein that increases guanine nucleotide exchange by heterotrimeric G proteins. Two cysteine residues near the N-terminus of GAP-43 are subject to palmitoylation, and are necessary for membrane binding as well as for G(o) activation. N-terminal peptides, which include these cysteines, stimulate G(o). Monopalmitoylation reduces, and dipalmitoylation abolishes the activity of the peptides. The activity of GAP-43 protein purified from brain also is reversibly blocked by palmitoylation. This suggests that palmitoylation controls a cycle of GAP-43 between an acylated, membrane-bound reservoir of inactive GAP-43, and a depalmitoylated, active pool of protein.     

A calcium-43 NMR study of calcium binding to an acidic proline-rich phosphoprotein from human saliva

The 43Ca NMR line width measured for Ca2+ bound to protein A, an acidic proline-rich salivary protein, is 1 order of magnitude narrower than has previously been observed for other proteins of similar molecular weight. The correlation times, quadrupole coupling constants, and chemical shifts estimated for Ca2+ ions bound to the intact protein (Mr approximately 10 000) and its 30 amino acid residue long acidic N-terminal TX peptide were indistinguishable within experimental error. These results--as well as the outcome of 1H NMR relaxation rate measurements--are indicative of extensive motions for the protein residues, which in turn give rise to a high degree of flexibility for the protein-bound Ca2+. Ca2+ titration and pH-dependent measurements on protein A, the TX peptide, and the dephosphorylated TX peptide established the importance of the two phosphoserine residues in the binding of Ca2+. Moreover, a comparison of the 43Ca NMR parameters with those obtained for other Ca2+-binding proteins suggests the presence of Ca2+-binding sites of similar symmetry in all these proteins. No evidence was found for a proposed interaction between the highly acidic N-terminal and the weakly basic C-terminal regions of protein A. In contrast, the high pH inflection that was observed in the pH titration curve for the intact protein was also found for the phospho and dephospho TX peptides, thus suggesting that basic moieties in the N-terminal region rather than those in the C-terminal region may be responsible for this observation.     

Immunohistochemical observation of growth-associated protein 43 (GAP-43) in the developing circumvallate papilla of the rat

The distribution and development of growth-associated protein 43 (GAP-43)-like immunoreactivity (-LI) in the rat circumvallate papilla (CVP) were compared to those of protein gene product 9.5 (PGP 9.5)-LI. In the adult, thick GAP-43-like immunoreactive (-IR) structures gathered densely in the subgemmal region. Some of these further penetrated the apical epithelium and trench wall epithelium. At least two types of GAP-43-IR structures were recognized; taste bud-related and non-gustatory GAP-43-IR neural elements. Immunoelectron microscopy revealed that GAP-43-LI was localized predominantly in the Schwann cells, and a few axons displayed GAP-43-LI in the lamina propria. In the trench epithelium, GAP-43-LI was detected in the cytoplasmic side of the axonal membrane. Some intragemmal GAP-43-IR axons made synaptic-like contacts with taste bud cells. At least four developmental stages were defined on the basis of the changes in distribution of GAP-43-LI. In stage I [embryonic day (E) 16–17] GAP-43-IR structures accumulated at the lamina propria just beneath the newly-formed circumvallate papilla. In stage II (E18–19) GAP-43-IR nerve fibers began to penetrate the apical epithelium. In stage III [E20-postnatal day (P) 0] GAP-43-IR nerve fibers first appeared in the trench wall epithelium. Penetration of GAP-IR nerve fibers occurred in the inner trench wall epithelium first, and then in the outer trench wall epithelium. In stage IV (P1-) the distribution of GAP-43-LI was similar to that observed in the adult; but the density of GAP-43-LI was much higher than in adults. PGP 9.5-LI showed a similar distribution pattern to that of GAP-43-LI, except for round-shaped cells in the apical epithelium at the late embryonic stages, and in taste bud cells and intralingual ganglionic cells which lacked GAP-43-LI. The similarities in distribution patterns of GAP-43-LI and PGP 9.5-LI during the development and mature circumvallate papilla suggest that GAP-43 may be a key neuronal molecule for induction and maintenance of the taste buds.     

A partially structured region of a largely unstructured protein, Plasmodium falciparum merozoite surface protein 2 (MSP2), forms amyloid-like fibrils.

Merozoite surface protein 2 (MSP2) from the human malaria parasite Plasmodium falciparum is expressed as a GPI-anchored protein on the merozoite surface. It has been implicated in the process of erythrocyte invasion and is a leading vaccine candidate. MSP2 is an intrinsically unstructured protein (IUP), and recombinant MSP2 forms amyloid-like fibrils upon storage. We have examined synthetic peptides corresponding to sequences in the conserved N-terminal region of MSP2 for the presence of local structure and the ability to form fibrils related to those formed by full-length MSP2. In a 25-residue peptide corresponding to the entire N-terminal region of mature MSP2, structures calculated from NMR data show the presence of nascent helical and turn-like structures. An 8-residue peptide from the central region of the N-terminal domain (residues 8-15) also formed a turn-like structure. Both peptides formed fibrils that were similar but not identical to the amyloid-like fibrils formed by full-length MSP2. Notably, the fibrils formed by the peptides bound both Congo Red and Thioflavin T, whereas the fibrils formed by full-length MSP2 bound only Congo Red. The propensity of peptides from the N-terminal conserved region of MSP2 to form amyloid-like fibrils makes it likely that this region contributes to fibril formation by the full-length protein. Thus, in contrast to the more common pathway of amyloid formation by structured proteins, which proceeds via partially unfolded intermediates that then undergo beta-aggregation, MSP2 is an example of a largely unstructured protein with at least one small structured region that has an important role in fibril formation.     

Mass spectrometric analysis of GAP-43/neuromodulin reveals the presence of a variety of fatty acylated species

GAP-43 (neuromodulin) is a protein kinase C substrate that is abundant in developing and regenerating neurons. Thioester-linked palmitoylation at two cysteines near the GAP-43 N terminus has been implicated in directing membrane binding. Here, we use mass spectrometry to examine the stoichiometry of palmitoylation and the molecular identity of the fatty acid(s) attached to GAP-43 in vivo. GAP-43 expressed in either PC12 or COS-1 cells was acetylated at the N-terminal methionine. Approximately 35% of the N-terminal GAP-43 peptides were also modified by palmitate and/or stearate on Cys residues. Interestingly, a variety of acylated species was detected, in which one of the Cys residues was acylated by either palmitate or stearate, or both Cys residues were acylated by palmitates or stearates or a combination of palmitate and stearate. Depalmitoylation of membrane-bound GAP-43 did not release the protein from the membrane, implying that additional forces function to maintain membrane binding. Indeed, mutation of four basic residues within the N-terminal domain of GAP-43 dramatically reduced membrane localization of GAP-43 without affecting palmitoylation. These data reveal the heterogeneous nature of S-acylation in vivo and illustrate the power of mass spectrometry for identification of key regulatory protein modifications.     

Identification of regions responsible for heparin-induced amyloidogenesis of human serum amyloid A using its fragment peptides

Human serum amyloid A (SAA) is a precursor protein of amyloid fibrils. Although several studies have been performed, a detailed understanding of the molecular mechanism for SAA fibrillation remains elusive. Glycosaminoglycans such as heparin are suggested to serve as scaffolds in amyloid fibril formation in some cases. In the present study, amyloidogenic properties of synthetic fragment peptides corresponding to the N-terminal (residues 1-27), central (residues 43-63), and C-terminal (residues 77-104) regions of SAA molecule induced by heparin were examined using fluorescence, circular dichroism (CD), and electron microscopy. Fluorescence and CD measurements demonstrated that SAA (1-27) peptide is evidently involved in heparin-induced amyloidogenesis. Correspondingly, relatively minor changes in fluorescence and a quite different pattern in the CD spectrum were observed in SAA (43-63) peptide. In contrast, SAA (77-104) peptide did not show any changes induced by heparin. Transmission electron microscopy indicated that SAA (1-27) peptide forms short and straight fibrils, whereas SAA (43-63) peptide forms much longer and seemingly elastic fibrils. These results suggest that the N-terminal region plays a crucial role as a rigid core and the central region facilitates the elongation of fibrils in heparin-induced amyloidogenesis of SAA molecule.     

Specific Proteolysis of Neuronal Protein GAP-43 by Calpain: Characterization,Regulation, and Physiological Role

The mechanism of specific proteolysis of the neuronal protein GAP-43 in axonal terminals has been investigated. In synaptic terminals in vivo and in synaptosomes in vitro GAP-43 is cleaved only at the single peptide bond formed by Ser41; this is within the main effector domain of GAP-43. Proteolysis at this site involves the cysteine calcium-dependent neutral protease calpain. The following experimental evidences support this conclusion: 1) calcium-dependent proteolysis of GAP-43 in synaptosomes is insensitive to selective inhibitor of micro-calpain (PD151746), but it is completely blocked by micro- and m-calpain inhibitor PD150606; 2) GAP-43 proteolysis in the calcium ionophore A23187-treated synaptosomes is activated by millimolar concentration of calcium ions; 3) the pattern of fragmentation of purified GAP-43 by m-calpain (but not by micro-calpain) is identical to that observed in synaptic terminals in vivo. GAP-43 phosphorylated at Ser41 by protein kinase C (PKC) is resistant to the cleavage by calpain. In addition, calmodulin binding to GAP-43 decreases the rate of calpain-mediated GAP-43 proteolysis. Our results indicate that m-calpain-mediated GAP-43 proteolysis regulated by PKC and calmodulin is of physiological relevance, particularly in axonal growth cone guidance. We suggest that the function of the N-terminal fragment of GAP-43 (residues 1-40) formed during cleavage by m-calpain consists in activation of neuronal heterotrimeric GTP-binding protein G(o); this results in growth cone turning in response to repulsive signals.     

Binding of myotoxin a to sarcoplasmic reticulum Ca(2+)-ATPase: a structural study.

The interaction of myotoxin alpha with intact sarcoplasmic reticulum (SR) components was investigated, and two SR proteins were identified that associated with myotoxin a. One of the proteins has an apparent molecular weight similar to the Ca(2+)-ATPase, the major SR protein responsible for calcium loading. Ca(2+)-ATPase was purified, and its interaction with myotoxin a was studied. Evidence for specific binding of myotoxin a to Ca(2+)-ATPase was established by isolating chemically cross-linked myotoxin a-Ca(2+)-ATPase complexes and further proving their association with anti-myotoxin a antibodies. The binding region of myotoxin a was further delineated by cleaving the protein with cyanogen bromide (CNBr) into two fragments, a larger N-terminal fragment of 28 residues and a smaller C-terminal fragment of 14 residues. Competition experiments with 125I-myotoxin a showed that the C-terminal fragment competed better against 125I-myotoxin a than the N-terminal fragment for SR protein binding. Two overlapping peptides covering the sequence of the N-terminal fragment were synthesized to clarify the interaction of the N-terminal fragment of myotoxin a with SR proteins. A 16-residue peptide corresponding to residues 1-16 competed strongly with 125I-myotoxin a, while a second peptide (residues 13-28) did not.     

Growth-associated protein of 43 kDa (GAP-43) is cleaved nonprocessively by the 20S proteasome.

Purified, nonubiquitinated growth-associated protein of 43 kDa (GAP-43) was attacked by purified reticulocyte 20S proteasome but not by the 26S proteasome. Cleavage yielded 12 N-terminally labelled GAP-43 fragments that could be resolved by SDS/PAGE. Inhibitor experiments suggested that proteasome beta1 activity yielded the resolved bands and that proteasomebeta5 activity generated nonresolvable fragments. Processive degradation, yielding only nonresolvable fragments, therefore did not occur. Most of the resolved fragments co-migrated with fragments formed in the reticulocyte lysate translation mixture used for GAP-43 synthesis, which suggested that the fragments were also produced in the translation mixture by the endogenous reticulocyte lysate proteasome. Consistent with this idea, the addition of proteasome inhibitors to translation mixtures blocked fragment production. Ubiquitinated GAP-43 appeared to be the source of the fragments in the presence of ATP, and nonubiquitinated GAP-43 the source in the absence of ATP. The results therefore suggest that the lack of processing seen with the 20S proteasome is not an artefact arising from the way in which the 20S proteasome was purified. In one purification protocol, the GAP-43 fragments formed in translation mixtures co-purified with full-length GAP-43. These fragments were digested to nonresolvable products upon addition of purified 20S proteasome. Addition of calmodulin or G-actin blocked the consumption of both full-length GAP-43 and the co-purified GAP-43 fragments. This showed that the resolved fragments can re-enter the proteasome and be cleaved to nonresolvable products, indicating that the lack of processivity is not a result of their resistance to further proteasome attack. The difficult step therefore appears to be the transfer of the large fragments within the proteasome from the beta1 to the beta5 activity for further attack.     

Developmental study of the expression of B50/GAP-43 in rat retina

B50/GAP-43 has been implicated in neural plasticity, development, and regeneration. Several studies of axonally transported proteins in the optic nerve have shown that this protein is synthesized by developing and regenerating retinal ganglion cells in mammals, amphibians, and fish. However, previous studies using immunohistochemistry to localize B50/GAP-43 in retina have shown that this protein is found in the inner plexiform layer in adults. Since the inner plexiform layer contains the processes of amacrine cells, ganglion cells, and bipolar cells to determine which cells in the retina express B50/GAP-43, we have now used in situ hybridization to localize the mRNA that codes for this protein in the developing rat retina. We have found that B50/GAP-43 is expressed primarily by cells in the retinal ganglion cell layer as early as embryonic day 15, and until 3 weeks postnatal. Some cells in the inner nuclear layer, possibly a subclass of amacrine cells, also express B50/GAP-43 protein and mRNA; however, the other retinal neurons–bipolar cells, photoreceptors, and horizontal cells express little, if any, B50/GAP-43 at any stage in their development. Early in development, the protein appears in the somata and axons of ganglion cells, while later in development, B50/GAP-43 becomes concentrated in the inner plexiform layer, where it continues to be expressed in adult animals. These results are discussed in terms of previous proposals as to the functions of this molecule. © 1993 John Wiley & Sons, Inc.     

Expression of the GTPase activating domain of the neurofibromatosis type 1 (NF1) gene in Escherichia coli and role of the conserved lysine residue.

A small catalytic domain from the neurofibromatosis type 1 gene, NF1-333, consisting of 333 amino acids between residues 1197 and 1528, including an additional N-terminal methionine, was expressed in Escherichia coli as a soluble protein. Its catalytic activity under non-saturating conditions is similar to the full-length p120-GAP but different from truncated GAP-334. Under saturating conditions its kcat and KM are lower. Lys-1422, which is totally conserved in all GAP proteins, was mutated and the properties of the mutant protein investigated. Lys-1422 seems to be essential for the stability of the proteins and not for its catalytic activity.     

生长相关蛋白GAP-43多肽抗体的制备

目的对生长相关蛋白-43（growth associated proteins-43,GAP-43）进行抗原表位分析并制备兔多克隆抗体。方法通过对GAP-43 cDNA序列及氨基酸序列的结构进行生物信息分析,依据蛋白质的二级结构、亲水性、疏水性、抗原性及理化特性,经过同源性检索后,综合考虑抗体设计的其他因素,选出具有免疫活性的抗原决定簇多肽片段,采用有机固相多肽合成法合成了GAP-43的多肽片段,并与载体蛋白血蓝蛋白（KLH）偶联制备成抗原,免疫新西兰家兔。结果 ELISA测定GAP-43抗体效价为1∶32 000;Western blot检测结果显示：在分子量25kDa出现单一条带;免疫组织化学法检测显示：GAP-43蛋白在小鼠海马神经元中有表达;免疫细胞化学法检测显示：GAP-43蛋白在人神经母细胞瘤株SH-SY5Y中存在表达。结论利用生物信息学软件比较准确地预测GAP-43的抗原决定簇,并成功制备了高效价、高特异性的多肽抗体。     

Identification of two proteins that bind to a pyrimidine-rich sequence in the 3'-untranslated region of GAP-43 mRNA.

GAP-43 is a membrane phosphoprotein that is important for the development and plasticity of neural connections. In undifferentiated PC12 pheochromocytoma cells, GAP-43 mRNA degrades rapidly ( t = 5 h), but becomes stable when cells are treated with nerve growth factor. To identify trans- acting factors that may influence mRNA stability, we combined column chromatography and gel mobility shift assays to isolate GAP-43 mRNA binding proteins from neonatal bovine brain tissue. This resulted in the isolation of two proteins that bind specifically and competitively to a pyrimidine-rich sequence in the 3'-untranslated region of GAP-43 mRNA. Partial amino acid sequencing revealed that one of the RNA binding proteins coincides with FBP (far upstream element binding protein), previously characterized as a protein that resembles hnRNP K and which binds to a single-stranded, pyrimidine-rich DNA sequence upstream of the c -myc gene to activate its expression. The other binding protein shares sequence homology with PTB, a polypyrimidine tract binding protein implicated in RNA splicing and regulation of translation initiation. The two proteins bind to a 26 nt pyrimidine-rich sequence lying 300 nt downstream of the end of the coding region, in an area shown by others to confer instability on a reporter mRNA in transient transfection assays. We therefore propose that FBP and the PTB-like protein may compete for binding at the same site to influence the stability of GAP-43 mRNA.     

The first semi-synthetic serine protease made by native chemical ligation

Selective incorporation of non-natural amino acid residues into proteins is a powerful approach to delineate structure-function relationships. Although many methodologies are available for chemistry-based protein engineering, more facile methods are needed to make this approach suitable for routine laboratory practice. Here, we describe a new strategy and provide a proof of concept for engineering semi-synthetic proteins. We chose a serine protease Streptomyces griseus trypsin (SGT) for this study to show that it is possible to efficiently couple a synthetic peptide containing a catalytically critical residue to a recombinant fragment containing the other active site residues. The 223-residue hybrid SGT molecule was prepared by fusing a chemically synthesized N-terminal peptide to a large C-terminal fragment of recombinant origin using native chemical ligation. This C-terminal polypeptide was produced from full-length SGT by cyanogen bromide cleavage at a genetically engineered Met57 position. This semi-synthetic hybrid trypsin is fully active, showing kinetics identical to the wild-type enzyme. Thus, we believe that it is an ideal model enzyme for studying the catalytic mechanisms of serine proteases by providing a straightforward approach to incorporate non-natural amino acids in the N-terminal region of the protein. In particular, this strategy will allow for replacement of the catalytic His57 residue and the buried N-terminus, which is thought to help align the active site, with synthetic analogs. Our approach relies on readily available recombinant proteins and small synthetic peptides, thus having general applications in chemical engineering of large proteins where the N-terminal region is the focal interest.     

Monoclonal antibodies show that kinase C phosphorylation of GAP-43 during axonogenesis is both spatially and temporally restricted in vivo.

To study the role of kinase C phosphorylation in the distribution and function of GAP-43 we have generated a panel of mAbs that distinguish between GAP-43 that has been phosphorylated by kinase C and forms that have not. One class of antibodies, typified by 2G12/C7, reacts with only the phosphorylated form of GAP-43; it recognizes the peptide IQAS(PO4)FR equivalent to residues 38-43 that includes the single kinase C phosphorylation site at serine. Another, exemplified by 10E8/E7, reacts with both phosphorylated and nonphosphorylated forms. We have used the antibodies to study the distribution of kinase C-phosphorylated GAP-43 during axonogenesis and in the adult nervous system. Two major findings emerge. First, there is a lag between the initiation of axon outgrowth and the phosphorylation of GAP-43 by kinase C. The extent of this lag period varies between the different structures studied. In some cases, e.g., the trigeminal nerve, our result suggest that kinase C phosphorylation may be correlated with proximity of the growing axon to its target. Second, kinase C-phosphorylated GAP-43 is always spatially restricted to the distal axon. It is never seen either proximally or in cell bodies, even those with high levels of GAP-43 protein. This result also implies that GAP-43 is axonally transported in the non-kinase C phosphorylated form. Thus, kinase C phosphorylation of GAP-43 is not required for axon outgrowth or growth cone function per se and may be more related to interactions of the growth cone with its environment.     

Increased expression of GAP-43 in small sensory neurons after stimulation by NGF indicative of neuroregeneration in capsaicin-treated rats.

Intraplantar injections of human recombinant nerve growth factor (rhNGF-beta) into the hind paw of capsaicin-treated adult rats are known to lead to a recovery of depleted peptide transmitter substances, to the immunohistochemical reappearance of peptidergic innervation in the skin and in the dorsal horn of the spinal cord, as well as to a recovery of the function of capsaicin-lesioned neurons. In the present study a marker peptide for neuronal regeneration and outgrowth, growth associated protein 43 (GAP-43), was investigated in lumbar dorsal root ganglia (DRGs) and in the hindpaw skin, in order to differentiate which population of the sensory neurons responds with a neuroregenerative behaviour. In situ hybridization histochemistry (ISH) revealed that at day 8 after the capsaicin treatment GAP-43 expression was significantly increased in small DRG cells as compared to control animals, and treatment with NGF in capsaicinized rats lead to an even more pronounced increase of GAP-43 expression in the small-sized cell population. Intraepidermal labelling of GAP-43 peptide was observed in the skin of control animals, but was markedly reduced in the animals that were treated with capsaicin alone. However, intraepidermal GAP-43 immunoreactive (GAP-43-IR) fibres nearly fully recovered in the capsaicin + NGF-treated group. These results indicate that the population of small DRG cells shows spontaneous regenerative activity after a capsaicin lesion which does not lead to a successful recovery of nerve terminals in the skin. Only after an additional NGF treatment small DRG cells show an even stronger regenerative response which now also involves structural reorganization of neuron membranes and axogenesis in the periphery.     

Self-association of IQGAP1: characterization and functional sequelae

The scaffolding protein IQGAP1 participates in numerous cellular functions by binding to target proteins such as actin, calmodulin, E-cadherin, beta-catenin, Cdc42, Rac1, and CLIP-170. IQGAP1 regulates the cytoskeleton, promotes cell motility, and modulates E-cadherin-mediated cell-cell adhesion. However, how IQGAP1 exerts its functions in vivo is still unclear. In this study we investigate the self-association of IQGAP1 and its role in IQGAP1 function. Endogenous IQGAP1 co-immunoprecipitated from MCF-7 cells with IQGAP1 tagged with enhanced green fluorescent protein, indicating that IQGAP1 self-associates in cells. In vitro assays confirmed that IQGAP1 can self-associate and that this effect is mediated by the N-terminal half of the protein. Gel filtration analysis suggested that full-length IQGAP1 exists as a combination of monomers, dimers, and larger oligomers. Analysis performed with multiple fragments of IQGAP1 narrowed the self-association region to amino acids 763-863. In support of this observation, a peptide comprising residues 763-863 disrupted self-association of full-length IQGAP1 in a dose-dependent manner. Similarly, deleting this sequence from IQGAP1 abolished binding to full-length IQGAP1. In addition, the ability of IQGAP1 to increase the amount of active Cdc42 in cells is abrogated upon removal of this region. Consistent with these findings, transfection into cells of a peptide containing the self-association domain significantly reduced the amount of active Cdc42 in cell lysates. These observations define a sequence of IQGAP1 that is necessary for its oligomerization and demonstrate that self-association is required for the normal cellular function of IQGAP1.