In its active form, the GTP-binding protein rab8 interacts with a stress-activated protein kinase.

Rab8 is a small GTP-binding protein that plays a role in vesicular transport from the trans-Golgi network to the basolateral plasma membrane in polarized epithelial cells (MDCK), and to the dendritic surface in hippocampal neurons. As is the case for most other rab proteins, the precise molecular interactions by which rab8 carries out its function remain to be elucidated. Here we report the identification and the complete cDNA-derived amino acid sequence of a murine rab8-interacting protein (rab8ip) that specifically interacts with rab8 in a GTP-dependent manner. Rab8ip displays 93% identity with the GC kinase, a serine/threonine protein kinase recently identified in human lymphoid tissue that is activated in the stress response. Like the GC kinase, rab8ip has protein kinase activity manifested by autophosphorylation and phosphorylation of the classical serine/threonine protein kinase substrates, myelin basic protein and casein. When coexpressed in transfected 293T cells, rab8 and the rab8ip/GC kinase formed a complex that could be recovered by immunoprecipitation with antibodies to rab8. Cell fractionation and immunofluorescence analyses indicate that in MDCK cells endogenous rab8ip is present both in the cytosol and as a peripheral membrane protein concentrated in the Golgi region and basolateral plasma membrane domains, sites where rab8 itself is also located. In light of recent evidence that rab proteins may act by promoting the stabilization of SNARE complexes, the specific GTP-dependent association of rab8 with the rab8ip/GC kinase raises the possibility that rab-regulated protein phosphorylation is important for vesicle targeting or fusion. Moreover, the rab8ip/GC kinase may serve to modulate secretion in response to stress stimuli.

Homophilic adhesion mediated by the neural cell adhesion molecule involves multiple immunoglobulin domains.

The neural cell adhesion molecule (N-CAM) mediates homophilic binding between a variety of cell types including neurons, neurons and glia, and neurons and muscle cells. The mechanism by which N-CAM on one cell interacts with N-CAM on another, however, is unknown. Attempts to identify which of the five immunoglobulin-like domains (Ig I-V) and the two fibronectin type III repeats (FnIII 1-2) in the extracellular region of N-CAM are involved in this process have led to ambiguous results. We have generated soluble recombinant proteins corresponding to each of the individual immunoglobulin domains and the combined FnIII 1-2 and prepared polyclonal antibodies specific for each. The purified proteins and antibodies were used in aggregation experiments with fluorescent microspheres and chicken embryo brain cells to determine possible contributions of each domain to homophilic adhesion. The recombinant domains were tested for their ability to bind to purified native N-CAM, to bind to each other, and to inhibit the aggregation of N-CAM on microspheres and the aggregation of neuronal cells. Each of the immunoglobulin domains bound to N-CAM, and in solution all of the immunoglobulin domains inhibited the aggregation of N-CAM-coated microspheres. Soluble Ig II, Ig III, and Ig IV inhibited neuronal aggregation; antibodies against whole N-CAM, the Ig III domain, and the Ig I domain all strongly inhibited neuronal aggregation, as well as the aggregation of N-CAM-coated microspheres. Of all the domains, the third immunoglobulin domain alone demonstrated the ability to self-aggregate, whereas Ig I bound to Ig V and Ig II bound to Ig IV. The combined FnIII 1-2 exhibited a slight ability to self-aggregate but did not bind to any of the immunoglobulin-like domains. These results suggest that N-CAM-N-CAM binding involves all five immunoglobulin domains and prompt the hypothesis that in homophilic cell-cell binding mediated by N-CAM these domains may interact pairwise in an antiparallel orientation.

Human semaphorins A(V) and IV reside in the 3p21.3 small cell lung cancer deletion region and demonstrate distinct expression patterns.

Semaphorins and collapsins make up a family of conserved genes that encode nerve growth cone guidance signals. We have identified two additional members of the human semaphorin family [human semaphorin A(V) and human semaphorin IV] in chromosome region 3p21.3, where several small cell lung cancer (SCLC) cell lines exhibit homozygous deletions indicative of a tumor suppressor gene. Human semaphorin A(V) has 86% amino acid homology with murine semaphorin A, whereas semaphorin IV is most closely related to murine semaphorin E, with 50% homology. These semaphorin genes are approximately 70 kb apart flanking two GTP-binding protein genes, GNAI-2 and GNAT-1. In contrast, other human semaphorin gene sequences (human semaphorin III and homologues of murine semaphorins B and C) are not located on chromosome 3. Human semaphorin A(V) is translated in vitro into a 90-kDa protein, which accumulates at the endoplasmic reticulum. The human semaphorin A(V) (3.4-kb mRNA) and IV (3.9- and 2.9-kb mRNAs) genes are expressed abundantly but differentially in a variety of human neural and nonneural tissues. Human semaphorin A(V) was expressed in only 1 out of 23 SCLCs and 7 out of 16 non-SCLCs, whereas semaphorin IV was expressed in 19 out of 23 SCLCs and 13 out of 16 non-SCLCs. Mutational analysis in semaphorin A(V) revealed mutations (germ line in one case) in 3 of 40 lung cancers. Our data suggest the need to determine the function of human semaphorins A(V) and IV in nonneural tissues and their role in the pathogenesis of lung cancer.

Modulation of promoter occupancy by cooperative DNA binding and activation-domain function is a major determinant of transcriptional regulation by activators in vivo.

Binding of transcriptional activators to a promoter is a prerequisite process in transcriptional activation. It is well established that the efficiency of activator binding to a promoter is determined by the affinity of direct interactions between the DNA-binding domain of an activator and its specific target sequences. However, I describe here that activator binding to a promoter is augmented in vivo by the effects of two other determinants that have not been generally appreciated: (i) the number of activator binding sites present in a promoter and (ii) the potency of activation domains of activators. Multiple sites within a promoter can cooperatively recruit cognate factors regardless of whether they contain an effective activation domain. This cooperativity can result in the synergistic activation of transcription. The second effect is the enhancement of activator binding to a promoter by the presence of activation domains. In this case, activation domains are not simply tethered to the promoter by the DNA-binding domain but instead assist the DNA-binding domain being tethered onto the promoter. This effect of activation domains on DNA binding is instrumental in determining how potent activators can induce steep transcriptional increases at low concentrations.

Evidence for distinct signaling mechanisms in two mammalian olfactory sense organs.

In mammals, olfactory stimuli are detected by sensory neurons at two distinct sites: the olfactory epithelium (OE) of the nasal cavity and the neuroepithelium of the vomeronasal organ (VNO). While the OE can detect volatile chemicals released from numerous sources, the VNO appears to be specialized to detect pheromones that are emitted by other animals and that convey information of behavioral or physiological importance. The mechanisms underlying sensory transduction in the OE have been well studied and a number of components of the transduction cascade have been cloned. Here, we investigated sensory transduction in the VNO by asking whether VNO neurons express molecules that have been implicated in sensory transduction in the OE. Using in situ hybridization and Northern blot analyses, we found that most of the olfactory transduction components examined, including the guanine nucleotide binding protein alpha subunit (G-alpha-olf), adenylyl cyclase type III, and an olfactory cyclic nucleotide-gated (CNG) channel subunit (oCNC1), are not expressed by VNO sensory neurons. In contrast, VNO neurons do express a second olfactory CNG channel subunit (oCNC2). These results indicate that VNO sensory transduction is distinct from that in the OE but raise the possibility that, like OE sensory transduction, sensory transduction in the VNO might involve cyclic nucleotide-gated ion channels.

A role for the Msx-1 homeodomain in transcriptional regulation: residues in the N-terminal arm mediate TATA binding protein interaction and transcriptional repression.

In a previous study we showed that the murine homeodomain protein Msx-1 is a potent transcriptional repressor and that this activity is independent of its DNA binding function. The implication of these findings is that repression by Msx-1 is mediated through its association with certain protein factors rather than through its interaction with DNA recognition sites, which prompted investigation of the relevant protein factors. Here we show that Msx-1 interacts directly with the TATA binding protein (TBP) but not with several other general transcription factors. This interaction is mediated by the Msx-1 homeodomain, specifically through residues in the N-terminal arm. These same N-terminal arm residues are required for repression by Msx-1, suggesting a functional relationship between TBP association and transcriptional repression. This is further supported by the observation that addition of excess TBP blocks the repressor action of Msx-1 in in vitro transcription assays. Finally, DNA binding activity is separable from both TBP interaction and repression, which further shows that these other activities of the Msx-1 homeodomain are distinct. Therefore, these findings define a role for the Msx-1 homeodomain, particularly the N-terminal arm residues in protein-protein interaction and transcriptional repression, and implicate a more complex role overall for homeodomains in transcriptional regulation.

Detection of sub-8-nm movements of kinesin by high-resolution optical-trap microscopy.

Kinesin is a molecular motor that transports organelles along microtubules. This enzyme has two identical 7-nm-long motor domains, which it uses to move between consecutive tubulin binding sites spaced 8 nm apart along a microtubular protofilament. The molecular mechanism of this movement, which remains to be elucidated, may be common to all families of motor proteins. In this study, a high-resolution optical-trap microscope was used to measure directly the magnitude of abrupt displacements produced by a single kinesin molecule transporting a microscopic bead. The distribution of magnitudes reveals that kinesin not only undergoes discrete 8-nm movements, in agreement with previous work [Svoboda, K., Schmidt, C. F., Schnapp, B. J. & Block, S.M. (1993) Nature (London) 365, 721-727], but also frequently exhibits smaller movements of about 5 nm. A possible explanation for these unexpected smaller movements is that kinesin's movement from one dimer to the next along a protofilament involves at least two distinct events in the mechanical cycle.

ARD1, a 64-kDa bifunctional protein containing an 18-kDa GTP-binding ADP-ribosylation factor domain and a 46-kDa GTPase-activating domain.

The alpha subunits of the heterotrimeric guanine nucleotide-binding proteins (G proteins) hydrolyze GTP at a rate significantly higher than do most members of the Ras family of approximatelly 20-kDa GTP-binding proteins, which depend on a GTPase-activating protein (GAP) for acceleration of GTP hydrolysis. It has been demonstrated that an inserted domain in the G-protein alpha subunit, not present in the much smaller Ras-like proteins, is responsible for this difference [Markby, D. W., Onrust, R. & Bourne, H. R. (1993) Science 262, 1895-1900]. We report here that ARD1, a 64-kDa protein with an 18-kDa carboxyl-terminal ADP-ribosylation factor (ARF) domain, exhibited significant GTPase activity, whereas the ARF domain, expressed as a recombinant protein in Escherichia coli, did not. Addition of the 46-kDa amino-terminal extension (similarly synthesized in E. coli) to the GTP-binding ARF-domain of ARD1 enhanced GTPase activity and inhibited GDP dissociation. The kinetic properties of mixtures of the ARF and non-ARF domains were similar to those of an intact recombinant ARD1. Physical association of the two proteins was demonstrated directly by gel filtration and by using the immobilized non-ARF domain. Thus, like the alpha subunits of heterotrimeric G proteins, ARD1 appears to consist of two domains that interact to regulate the biological activity of the protein.

Cloning, expression, and properties of the microtubule-stabilizing protein STOP.

Nerve cells contain abundant subpopulations of cold-stable microtubules. We have previously isolated a calmodulin-regulated brain protein, STOP (stable tubule-only polypeptide), which reconstitutes microtubule cold stability when added to cold-labile microtubules in vitro. We have now cloned cDNA encoding STOP. We find that STOP is a 100.5-kDa protein with no homology to known proteins. The primary structure of STOP includes two distinct domains of repeated motifs. The central region of STOP contains 5 tandem repeats of 46 amino acids, 4 with 98% homology to the consensus sequence. The STOP C terminus contains 28 imperfect repeats of an 11-amino acid motif. STOP also contains a putative SH3-binding motif close to its N terminus. In vitro translated STOP binds to both microtubules and Ca2+-calmodulin. When STOP cDNA is expressed in cells that lack cold-stable microtubules, STOP associates with microtubules at 37 degrees C, and stabilizes microtubule networks, inducing cold stability, nocodazole resistance, and tubulin detyrosination on microtubules in transfected cells. We conclude that STOP must play an important role in the generation of microtubule cold stability and in the control of microtubule dynamics in brain.

Factor XII-induced mitogenesis is mediated via a distinct signal transduction pathway that activates a mitogen-activated protein kinase.

Clotting factor XII (Hageman factor) contains epidermal growth factor (EGF)-homologous domains and is reported to be a potent mitogen for human hepatoma (HepG2) cells. In this study, we tested whether factor XII exhibits growth factor activity on several other EGF-sensitive target cells, including fetal hepatocytes, endothelial cells, alveolar type II cells, and aortic smooth muscle cells. We found that factor XII significantly enhanced [3H]thymidine incorporation in aortic smooth muscle cells (SMCs) and all other cells tested. Tyrphostin, a growth factor receptor/tyrosine kinase antagonist, inhibited both EGF- and factor XII-induced responses. However, differences in the levels of magnitude of DNA synthesis, the observed synergism between EGF and factor XII, and the differential sensitivity to tyrphostin suggest that the EGF receptor and the factor XII receptor may be nonidentical. The factor XII-induced mitogenic response was achieved at concentrations that were 1/10th the physiologic range for the circulating factor and was reduced by popcorn inhibitor, a specific factor XII protease inhibitor. Treatment of aortic SMCs with factor XII, as well as activated factor XII, resulted in a rapid and transient activation of a mitogen-activated/extracellular signal-regulated protein kinase with peak activity/tyrosine phosphorylation observed at 5 to 10 min of exposure. Taken together, these data (i) confirm that clotting factor XII functions as a mitogenic growth factor and (ii) demonstrate that factor XII activates a signal transduction pathway, which includes a mitogen-activated protein kinase.

Differential functions of the two Src homology 2 domains in protein tyrosine phosphatase SH-PTP1.

SH-PTP1 (also known as PTP1C, HCP, and SHP) is a non-transmembrane protein tyrosine phosphatase (PTPase) containing two tandem Src homology 2 (SH2) domains. We show here that the two SH2 (N-SH2 and C-SH2) domains in SH-PTP1 have different functions in regulation of the PTPase domain and thereby signal transduction. While the N-terminal SH2 domain is both necessary and sufficient for autoinhibition through an intramolecular association with the PTPase domain, truncation of the C-SH2 domain [SH-PTP1 (delta CSH2) construct] has little effect on SH-PTP1 activity. A synthetic phosphotyrosine residue (pY) peptide derived from the erythropoietin receptor (EpoR pY429) binds to the N-SH2 domain and activates both wild-type SH-PTP1 and SH-PTP1 (delta CSH2) 60- to 80-fold. Another pY peptide corresponding to a phosphorylation site on the IgG Fc receptor (Fc gamma RIIB1 pY309) associates with both the C-SH2 domain (Kd = 2.8 microM and the N-SH2 domain (Kd = 15.0 microM) and also activates SH-PTP1 12-fold. By analysis of the effect of the Fc gamma RIIB1 pY309 peptide on SH-PTP1 (delta CSH2), SH-PTP1 (R30K/R33E), SH-PTP1 (R30K/R136K), and SH-PTP1 (R136K) mutants in which the function of either the N- or C-SH2 domain has been impaired, we have determined that both synthetic pY peptides stimulate SH-PTP1 by binding to its N-SH2 domain; binding of pY ligand to the C-SH2 domain has no effect on SH-PTP1 activity. We propose that the N-terminal SH2 domain serves both as a regulatory domain and as a recruiting unit, whereas the C-terminal SH2 domain acts merely as a recruiting unit.

Binding in the growth hormone receptor complex.

Binding reactions between human growth hormone (hGH) and its receptor provide a detailed account of how a polypeptide hormone activates its receptor and more generally how proteins interact. Through high-resolution structural and functional studies it is seen that hGH uses two different sites (site 1 and site 2) to bind two identical receptor molecules. This sequential dimerization reaction activates the receptor, presumably by bringing the intracellular domains into close proximity so they may activate cytosolic components. As a consequence of this mechanism it is possible to build antagonists to the receptor by introducing mutations in hGH that block binding at site 2 and to build even more potent antagonists by combining these with mutants that enhance binding at site 1. Alanine-scanning mutagenesis of all contact residues at the site 1 interface shows that only a small and complementary set of side chains clustered near the center of the interface affects binding. The most important contacts are hydrophobic, and these are surrounded by polar and charged interactions of lesser importance. Kinetic analysis shows for the most part that the important side chains function to maintain the complex, not to guide the hormone to the receptor. Hormone-induced homodimerization or heterodimerization reactions are turning out to be pervasive mechanisms for signal transduction. Moreover, the molecular recognition principles seen in the hGH-receptor complex are likely to generalize to other protein-protein complexes.

A role of amphiphysin in synaptic vesicle endocytosis suggested by its binding to dynamin in nerve terminals.

Amphiphysin, a major autoantigen in paraneoplastic Stiff-Man syndrome, is an SH3 domain-containing neuronal protein, concentrated in nerve terminals. Here, we demonstrate a specific, SH3 domain-mediated, interaction between amphiphysin and dynamin by gel overlay and affinity chromatography. In addition, we show that the two proteins are colocalized in nerve terminals and are coprecipitated from brain extracts consistent with their interactions in situ. We also report that a region of amphiphysin distinct from its SH3 domain mediates its binding to the alpha c subunit of AP2 adaptin, which is also concentrated in nerve terminals. These findings support a role of amphiphysin in synaptic vesicle endocytosis.

Domains with transcriptional regulatory activity within the ALL1 and AF4 proteins involved in acute leukemia.

The ALLI gene, located at chromosome band 11q23, is involved in acute leukemia through a series of chromosome translocations and fusion to a variety of genes, most frequently to A4 and AF9. The fused genes encode chimeric proteins proteins. Because the Drosophila homologue of ALL1, trithorax, is a positive regulator of homeotic genes and acts at the level of transcription, it is conceivable that alterations in ALL1 transcriptional activity may underlie its action in malignant transformation. To begin studying this, we examined the All1, AF4, AF9, and AF17 proteins for the presence of potential transcriptional regulatory domains. This was done by fusing regions of the proteins to the yeast GAL4 DNA binding domain and assaying their effect on transcription of a reporter gene. A domain of 55 residues positioned at amino acids 2829-2883 of ALL1 was identified as a very strong activator. Further analysis of this domain by in vitro mutagenesis pointed to a core of hydrophobic and acidic residues as critical for the activity. An ALL1 domain that repressed transcription of the reporter gene coincided with the sequence homologous to a segment of DNA methyltransferase. An AF4 polypeptide containing residues 480-560 showed strong activation potential. The C-terminal segment of AF9 spanning amino acids 478-568 transactivated transcription of the reporter gene in HeLa but not in NIH 3T3 cells. These results suggest that ALL1, AF4, and probably AF9 interact with the transcriptional machinery of the cell.

Alternative translation initiation site usage results in two functionally distinct forms of the GATA-1 transcription factor.

GATA-1 is a zinc-finger transcription factor that plays a critical role in the normal development of hematopoietic cell lineages. In human and murine erythroid cells a previously undescribed 40-kDa protein is detected with GATA-1-specific antibodies. We show that the 40-kDa GATA-1 (GATA-1s) is produced by the use of an internal AUG initiation codon in the GATA-1 transcript. The GATA-1 proteins share identical binding activity and form heterodimers in erythroleukemic cells but differ in their transactivation potential and in their expression in developing mouse embryos.