Huntingtin interacting protein 1 Is a clathrin coat binding protein required for differentiation of late spermatogenic progenitors.

Huntingtin-interacting protein 1 (HIP1) interacts with huntingtin, the protein whose gene is mutated in Huntington's disease. In addition, a fusion between HIP1 and platelet-derived growth factor beta receptor causes chronic myelomonocytic leukemia. The HIP1 proteins, including HIP1 and HIP1-related (HIP1r), have an N-terminal polyphosphoinositide-interacting epsin N-terminal homology, domain, which is found in proteins involved in clathrin-mediated endocytosis. HIP1 and HIP1r also share a central leucine zipper and an actin binding TALIN homology domain. Here we show that HIP1, like HIP1r, colocalizes with clathrin coat components. We also show that HIP1 physically associates with clathrin and AP-2, the major components of the clathrin coat. To further understand the putative biological role(s) of HIP1, we have generated a targeted deletion of murine HIP1. HIP1(-/-) mice developed into adulthood, did not develop overt neurologic symptoms in the first year of life, and had normal peripheral blood counts. However, HIP1-deficient mice exhibited testicular degeneration with increased apoptosis of postmeiotic spermatids. Postmeiotic spermatids are the only cells of the seminiferous tubules that express HIP1. These findings indicate that HIP1 is required for differentiation, proliferation, and/or survival of spermatogenic progenitors. The association of HIP1 with clathrin coats and the requirement of HIP1 for progenitor survival suggest a role for HIP1 in the regulation of endocytosis.

Clathrin- and AP-2-binding sites in HIP1 uncover a general assembly role for endocytic accessory proteins.

Clathrin-mediated endocytosis is a major pathway for the internalization of macromolecules into the cytoplasm of eukaryotic cells. The principle coat components, clathrin and the AP-2 adaptor complex, assemble a polyhedral lattice at plasma membrane bud sites with the aid of several endocytic accessory proteins. Here, we show that huntingtin-interacting protein 1 (HIP1), a binding partner of huntingtin, copurifies with brain clathrin-coated vesicles and associates directly with both AP-2 and clathrin. The discrete interaction sequences within HIP1 that facilitate binding are analogous to motifs present in other accessory proteins, including AP180, amphiphysin, and epsin. Bound to a phosphoinositide-containing membrane surface via an epsin N-terminal homology (ENTH) domain, HIP1 associates with AP-2 to provide coincident clathrin-binding sites that together efficiently recruit clathrin to the bilayer. Our data implicate HIP1 in endocytosis, and the similar modular architecture and function of HIP1, epsin, and AP180 suggest a common role in lipid-regulated clathrin lattice biogenesis.

A soluble form of the urokinase plasminogen activator receptor (suPAR) can bind to hematopoietic cells.

The receptor for urokinase plasminogen activator (uPAR; CD87) is a 50- to 65-kDa glycosylphosphatidylinositol (GPI)-anchored glycoprotein expressed by leukocytes and tumor cells where it facilitates uPA-dependent, plasmin-mediated pericellular proteolysis during cellular invasion. Because uPAR is inducibly shed into culture supernatants and human body fluids, we tested the hypothesis that soluble uPAR (suPAR) can bind to the plasma membrane of hematopoietic cells where it might modulate their invasive phenotype. As measured by flow cytometry, recombinant biotinylated-suPAR (B-suPAR) bound in a specific fashion to THP-1 leukemia cells and blood PMNs and monocytes (but not to lymphocytes). B-suPAR also demonstrated specific binding to a variety of leukemic lines, including cells that are positive or negative for membrane uPAR expression. Binding of B-suPAR to THP-1 cells was enhanced four- to sevenfold by 24-h exposure of cells to PMA or by co-incubation with uPA ligand (but not its isolated catalytic and binding fragments). Conversely, binding of B-suPAR to PMNs was unaffected by brief exposure to fMLP, and was inhibited by co-incubation with uPA. B-suPAR biding to PMA-differentiated THP-1 cells in the presence of uPA was further enhanced by acid washing (removing endogenous uPA) but was partially inhibited by treatment of cells with trypsin. Pretreatment of PMA-differentiated THP-1 cells and unstimulated PMNs with soluble sugars, calcium chelators, and antibodies specific for integrins or extracellular matrix proteins failed to consistently block the binding of B-suPAR. Whereas the binding of suPAR did not measurably affect cell-associated plasmin activation, suPAR did competitively inhibit the binding of exogenous uPA to membrane-associated uPAR. These observations support the hypothesis that suPAR can bind specifically to trypsin-sensitive receptors expressed by certain normal and neoplastic hematopoietic cells where its binding is variably influenced by uPA ligand.

Urokinase-type plasminogen activator receptors associate with beta1 and beta3 integrins of fibrosarcoma cells: dependence on extracellular matrix components.

We have shown previously that the urokinase-type plasminogen activator receptor (uPAR) physically associates with beta2 integrins on human leukocyte membranes. We now report that uPAR associates with certain members of the beta1 and beta3 integrin families expressed by a nonhematopoietic fibrosarcoma cell line (HT1080) when adherent to certain extracellular matrix molecules. Flow cytometry studies indicated that HT1080 cells expressed uPAR and beta1 and beta3 integrins. Double staining immunofluorescence was used to label uPAR and beta1 and beta3 integrins. The staining patterns of uPAR and beta1 integrins were strikingly similar when attached to fibronectin, laminin, or vitronectin but not polylysine-coated substrates. Resonance energy transfer (RET) between uPAR and beta1 integrins was observed, especially at focal adhesion plaques; this indicates that these molecules are within about 7 nm of each other on these cell membranes. uPAR and beta3 integrin coclustering and RET were also observed on tumor cells adherent to vitronectin but not to fibronectin, laminin, or polylysine-coated surfaces. Because N-acetyl-D-glucosamine was found previously to inhibit beta2 integrin-uPAR association, we tested the effect of saccharides on the beta1-uPAR and beta3-uPAR colocalization and RET. Colocalization and RET between uPAR and beta1 or beta3 integrins were effectively inhibited by N-acetyl-D-glucosamine on extracellular matrix-coated surfaces. To better define which members of beta1 and beta3 integrin families associate with uPAR, we studied the association of several alpha subunits with uPAR on tumor cells. We found that: (a) alpha5 colocalizes with uPAR on cells attached to fibronectin-coated surfaces; (b) alpha5 and alpha(v) colocalize with uPAR on cells adherent to vitronectin; and (c) alpha3 and alpha6 associate with uPAR on cells attached to laminin. In further support of physical associations between integrins and uPAR on tumor cells, uPAR was found to coimmunoprecipitate with beta1 integrins in Brij-58 lysates of HT1080 cells (as detected by anti-uPAR Western blotting of material isolated from an anti-beta1 integrin immunoaffinity column). Thus, uPAR may laterally associate with integrins of tumor cells when attached to specific extracellular matrix elements to enable directional proteolysis for tumor cell migration and invasion.

Enzyme-linked immunoabsorbent assay detection of a soluble form of urokinase plasminogen activator receptor in vivo.

The receptor for urokinase plasminogen activator (uPA-R, CD87) is a glycosylphosphatidylinositol (GPI)-anchored 50 to 65 kD glycoprotein that, by regulating membrane-associated plasmin activity, may facilitate the invasion of inflammatory and malignant cells. Certain other GPI-anchored glycoproteins are shed from the cell membrane and exist as soluble products in vitro and in vivo. To determine if uPA-R undergoes a similar phenomenon, we have developed a sensitive enzyme-linked immunoabsorbent assay (ELISA) (using a rabbit antiserum as both capture and detection reagents) to measure the quantity of soluble uPA-R (suPA-R) in tissue culture supernatants and biologic fluids. Using this ELISA, we have detected suPA-R in the culture supernatants of U-937 cells and human monocytes stimulated in vitro by certain soluble inflammatory mediators (Sitrin et al, Blood 84:1268, 1994; Mizukami et al., Clin Res 42:115A, 1994). To determine if suPA-R exists in vivo, we have screened the plasma of 20 normal volunteers (mean +/- SD, 3 +/- 3 ng/mL; median, 2 ng/mL; range, 1 to 11 ng/mL [serum values slightly higher]); the plasma of 13 ICU patients with clinical sepsis syndrome (mean +/- SD, 30 +/- 11 ng/mL; median, 11 ng/mL; range, 4 to 221 ng/mL); and the extravascular fluids (pleural, pericardial, and peritoneal) of 84 individuals with presumed inflammatory or malignant conditions (mean +/- SD, 21 +/- 39 ng/mL; median, 10 ng/mL; range, 2 to 253 ng/mL). Among the latter specimens, most were inflammatory exudates (only six were malignant by positive cytology) with the highest quantities of suPA-R associated with neutrophilic exudates. The solubility of suPA-R contained within these fluids was confirmed by reanalysis after ultracentrifugation to remove particulate material. When tested in a uPA ligand capture ELISA, representative specimens of extravascular body fluids and sepsis plasma contained suPA-R capable of binding uPA ligand (generally representing a small fraction of the immunoreactive material). We conclude from these data that suPA-R is immunologically detectable in vitro and in vivo with high concentrations of receptor found under conditions of inflammatory stimulation. The possibility of suPA-R's biologic activity is suggested by its partial retention of ligand binding capacity.

Human urokinase-type plasminogen activator primes neutrophils for superoxide anion release. Possible roles of complement receptor type 3 and calcium.

Urokinase-type plasminogen activator (uPA), which binds to cells via a specific receptor (uPAR), participates in pericellular proteolysis during leukocyte migration. Previous studies have indicated that uPAR is physically associated with CR3 (CD11b/CD18). To test the functional interactions of CR3 and uPAR, we have examined the ability of uPA to elicit changes in cytosolic calcium levels of normal neutrophils, neutrophils from a leukocyte adhesion deficiency (LAD) patient, and 3T3 transfectants expressing CR3, uPAR, or both. We found that calcium levels of neutrophils increased from 106 +/- 6 nM in untreated cells to 199 +/- 25 nM in the presence of uPA. In contrast, no significant change in calcium was observed when neutrophils from an leukocyte adhesion deficiency patient were examined. The uPA-dependent calcium rise was inhibited by mAb directed against either CR3 or uPAR and required intact uPA. To substantiate further these findings, we prepared transfectants expressing genes encoding uPAR, CR3, and both receptors; only cells expressing both receptors experienced a rise in intracellular calcium. Although uPA's calcium signal is insufficient to trigger superoxide production, FMLP dose-dependent superoxide production was greatly enhanced by incubating neutrophils with intact, but not fragmented, uPA. Flow cytometry experiments utilizing an FMLP analogue exclude the possibilities that urokinase binds to the FMLP receptor or up-regulates its expression. We suggest that calcium is a second messenger of uPA, that this message is mediated in a CR3-dependent fashion, and that this signal primes neutrophils for superoxide production.

Cytokine-specific regulation of urokinase receptor (CD87) expression by U937 mononuclear phagocytes.

Mononuclear phagocytes concentrate urokinase-type plasminogen activator (uPA) at the cell surface by expressing membrane uPA receptors (uPAR). This study examines the ability of exogenous cytokines to alter expression of membrane-associated uPA and uPAR in U937 mononuclear phagocytes. Cells were stimulated with recombinant interferon gamma (IFN gamma) or tumor necrosis factor alpha (TNF alpha), followed by immunolabeling for uPA or uPAR and flow cytometry. IFN gamma increased surface uPA 2.2-fold relative to unstimulated controls (P < .001), whereas TNF alpha had no significant effect. Likewise, maximal uPA binding capacity was increased 2.8-fold by IFN gamma (P < .02), but was not affected by TNF alpha. In unstimulated cells, 50% of receptors were occupied by endogenously generated uPA, and this proportion was not affected by either cytokine. IFN gamma upregulated uPAR 2.1-fold relative to unstimulated controls (P < .001), whereas TNF alpha had no effect. In contrast to effects on surface protein, TNF alpha induced a substantial increase in uPAR mRNA, equaling the effect of IFN gamma. In addition, both cytokines doubled the intracellular uPAR pool (P < .01). By contrast, TNF alpha induced a 2.5-fold increase in the level of uPAR protein released into conditioned medium (compared with unstimulated cells), whereas IFN gamma had no effect. These results indicate that uPAR expression is regulated in a cytokine-specific fashion. Some stimuli, such as TNF alpha, may increase uPAR synthetic activity without a corresponding change in membrane expression, because of enhanced release of uPAR from the cell. Cytokine-specific modulation of uPAR may be important in regulating the function of mononuclear phagocytes in inflammation and tissue repair.

Immunologic detection of the cellular receptor for urokinase plasminogen activator.

The cellular receptor for urokinase plasminogen activator (uPA-R) is a monomeric phosphatidylinositol-linked glycoprotein (gp40-65) that may contribute to the invasive capacity of tumor and inflammatory cells by focusing the activity of urokinase (uPA) in converting plasminogen to plasmin, a serine protease capable of degrading extracellular matrix proteins. The further characterization of uPA-R has been facilitated by our recent development of a monoclonal antibody, anti-Mo3f, specific for uPA-R. This mAb bound to uPA-R expressed by phorbol myristate acetate-stimulated U-937 cells and by NIH-3T3 cells permanently transfected with uPA-R cDNA. In competitive binding assays, anti-Mo3f inhibited the binding of fluorescein-conjugated uPA ligand to uPA-R expressed by U-937 cells and uPA-R transfectants; conversely, preexposure of cells to saturating quantities of exogenous uPA partially blocked the subsequent binding of anti-Mo3f mAb to uPA-R. Anti-Mo3f mAb was employed as the capture reagent in an ELISA for the quantitation of soluble forms of uPA-R (derived from U-937 cells and recombinant uPA-R) which had a sensitivity of approximately 4-12 ng/ml. Anti-Mo3f mAb was also applied as a serologic probe for the detection of uPA-R expressed by human tumor tissues. By immunoperoxidase staining, anti-Mo3f demonstrated positive tumor cell staining in 4 of 16 breast and 7 of 31 prostate carcinomas in formalin-fixed, paraffin-embedded specimens. These data indicate that the anti-Mo3f mAb detects an epitope proximate to or within the ligand binding domain (domain 1) of uPA-R and may be useful as a tool for the serologic detection of uPA-R in soluble form or associated with human tumors.

The urokinase receptor is expressed in invasive breast cancer but not in normal breast tissue.

We have studied expression of the urokinase receptor (u-PAR) in paraffin-embedded breast tissues at various stages of malignant progression. Forty-nine of 59 invasive cancers studied showed varying degrees of reactivity with our polyclonal antibody. The staining pattern was variable from case to case, although strong surface staining of tumor-associated macrophages was evident in most of these sections. In several cases, blood vessels in selected tumor areas were stained, as confirmed by treatment of adjacent sections with an anti-factor VIII antibody. These could represent regions of recent angiogenesis. Staining of tumor cells was observed in 21 of 59 cases and was extensive in 5 cases but confined to a small percentage of cells in the remaining 16 samples. In contrast with the cancer sections, all normal breast tissue (12 cases) was negative, as well as all fibroadenomas (4 cases), papillomas (5 cases), and hyperplasia with atypia (2 cases) studied. Seven carcinomas in situ examined were also negative for u-PAR, with the exception of few macrophages in two cases, suggesting that u-PAR expression may be associated with invasive tumor. The presence of u-PAR in human breast cancer and its absence from nonmalignant breast tissue supports the idea that plasminogen activation plays an important role in the process of cancer invasion. Expression of u-PAR on macrophages, endothelial cells, and cancer cells suggests the existence of complex paracrine interactions between tumor cells and stroma.

cDNA for Mo3, a monocyte activation antigen, encodes the human receptor for urokinase plasminogen activator.

We have cloned the cDNA for Mo3, an activation Ag expressed by human monocytes and myelomonocytic cell lines after stimulation by PMA, LPS, muramyl dipeptide, certain cytokines, and cAMP agonists. We have previously shown that Mo3 expression in vivo is associated predominantly with macrophages in inflammatory sites. Mo3 is a highly glycosylated protein of about 50 kDa in monocytes and U-937 cells and is anchored to the plasma membrane by glycosyl-phosphatidylinositol linkage. We purified Mo3 protein by cleavage from the U-937 cell surface with phosphatidylinositol-specific phospholipase C, followed by affinity chromatography using a mAb. An internal peptide sequence was determined and used to design oligonucleotide probes for screening an expression cDNA library. Nucleotide sequencing indicated that the complete coding sequence encodes 335 amino acids, including a predicted signal peptide of 22 residues and a hydrophobic C-terminal portion that is probably cleaved during formation of the GPI linkage. The resulting mature protein of about 290 amino acids is consistent with the 29-kDa molecular mass of deglycosylated Mo3. A Northern blot of RNA from U-937 cells revealed a 1.5-kb band that was induced by PMA treatment. Mo3 cDNA was transfected into Cos cells and surface expression of Mo3 was detected by ELISA using various anti-Mo3 mAb. We performed a computer search of the National Biomedical Research Foundation database and found that Mo3 is identical to the human receptor for the urokinase plasminogen activator (uPA-R). Purified soluble Mo3, as well as anti-Mo3 antibodies, were able to block uPA binding to its receptor on U-937 cells, indicating that Mo3 is indeed uPA-R. The use of these anti-Mo3 antibodies may be helpful in assessing the role of uPA-R in processes such as inflammation and tumor invasion.

Purification, biochemical composition, and biosynthesis of the Mo3 activation antigen expressed on the plasma membrane of human mononuclear phagocytes.

Mo3 is an activation Ag expressed on the surface of human mononuclear phagocytes stimulated in vitro or in vivo by various activating factors. Mo3 is obtained by immunoprecipitation with anti-Mo3 mAb from lysates of PMA-stimulated U-937 cells. The Ag is a heterogeneous glycoprotein with a molecular mass range of 42 to 66 kDa (nonreducing conditions) containing N-linked carbohydrate chains. When the cells are treated with phosphatidylinositol-specific phospholipase C, greater than 60% of total precipitable gp42-66 Ag is released in the supernatant. This phosphatidylinositol-specific phospholipase C-sensitive linkage to the plasma membrane has provided a means for the one-step purification of Mo3 by immunoaffinity chromatography. The eluted soluble Mo3 (sMo3) was greater than 90% pure as documented by the appearance of a single major protein peak on reverse phase HPLC and SDS-PAGE. The average yield was 12.1 micrograms/10(8) cells. Sufficient quantities of sMo3 have been purified to permit the determination of amino acid and carbohydrate composition. Complex N-linked carbohydrates make up nearly 50% of the glycoprotein content and contribute to its heterogeneity. An anti-Mo3 polyclonal antiserum generated from sMo3 was used to immunoprecipitate Mo3 and its precursor from biosynthetically labeled, PMA-stimulated U-937 cells or LPS-stimulated monocytes. These 35S-methionine "pulse-chase" experiments demonstrated the existence of a 40- to 42-kDa endo-beta-N-acetylglucosaminidase-sensitive precursor, which over a period of 4 to 5 h gave rise to an endo-beta-N-acetylglucosaminidase-resistant, but N-glycanase-sensitive 42- to 66-kDa mature form.

A structural characterization of the Mo3 activation antigen expressed on the plasma membrane of human mononuclear phagocytes.

Mo3 is an activation Ag expressed by human monocytic cells after stimulation in vitro by PMA, LPS, certain cytokines, and muramyl dipeptide. The structural characterization of Mo3 has been made possible by the development of a mAb (anti-Mo3f) that immunoprecipitates Mo3 from Nonidet P-40 lysates of radiolabeled PMA-stimulated U-937 cells and LPS-activated monocytes. On SDS-PAGE (nonreducing conditions) of anti-Mo3f immunoprecipitates, U-937 Mo3 is a single broad band of 39 to 66 kDa, whereas monocyte Mo3 is smaller with an apparent molecular mass of 32 to 56 kDa. Under reducing conditions, there is an increase in the m.w. of both species of Mo3 suggesting the existence of internal disulfide bonds. Mo3 is a glycoprotein with carbohydrate of the N-linked complex type as evidence by a reduction in m.w. by 40 to 50% after treatment with endoglycosidase F or N-glycanase; neuraminidase treatment produces a 3-kDa reduction in m.w. Deglycosylated Mo3 isolated from U-937 and monocytes have similar m.w. suggesting that the molecular heterogeneity of the native Mo3 may be due to differences in glycosylation. Mo3 is sensitive to phosphatidylinositol-specific phospholipase C with the release of native Mo3 from the surface of PMA-stimulated U-937 cells. These results indicate that Mo3 is a member of the glycosylphosphatidylinositol-linked family of surface glycoproteins.

Human mononuclear phagocyte activation antigens.

Activation of mononuclear phagocytes causes changes in plasma membrane composition that include the expression of surface antigens and receptors. Monoclonal antibody technology has made it possible to identify and characterize newly expressed surface antigens. Among these "activation antigens" is a glycoprotein, Mo3, which (among hematopoietic cells) is selectively expressed by human mononuclear phagocytes that have been exposed to inflammatory factors in vitro and in vivo. Progress toward a functional and structural analysis of Mo3 is described.

The effect of erythrocyte membrane on the birefringence formation of sickle cell hemoglobin.

The birefringence formation of sickle cell hemoglobin (HbS) in a thin liquid layer was observed while its environment was deoxygenated at different rates, and the effect of membrane was examined. Under slow rate of deoxygenation at 37 degrees C, at pH 7.4, the birefringence of purified HbS appeared at a concentration higher than 24% and its relative magnitude increased as the concentration was increased. Similarly, the partial pressure of oxygen, at which the birefringence formation was evident, increased from 0 to 27 torr as the concentration of HbS was increased from 24 to 28%, but it remained the same above this protein concentration. In all the samples tested relative birefringence was largest at the slow rate of deoxygenation (30 torrO2/min) and the magnitude decreased as the rate of deoxygenation was increased. The samples showed different sensitivity to the rate of deoxygenation. For example, while the total untreated hemolysate made by freeze-thawing of packed sickle cells was most resistant to the increased rates of deoxygenation, purified HbS was not. Washed open ghosts partially restored the birefringence formation pattern of purified HbS. The results indicate that the inner surface of the membranes of erythrocytes could behave as a template for large HbS polymer formation at relatively higher rates of deoxygenation.

Association of calmodulin inhibition, erythrocyte membrane stabilization and pharmacological effects of drugs.

The present study was designed to determine whether there is an association of drug-induced inhibition of calmodulin functions, drug-induced membrane stabilization (protection against osmotic lysis), and pharmacological effects of drugs. First, data on drugs which have been studied for both calmodulin inhibition and membrane antihemolysis were collected from the literature and an association of the two properties was established. Second, ten additional drugs were selected for study of all three properties. Four drugs, with known antihemolytic effects, were studied for calmodulin inhibition. One drug, which was a known calmodulin inhibitor, was studied for antihemolysis. Our results show that membrane-stabilizing drugs are usually calmodulin inhibitors, and vice versa; that drugs in certain therapeutic classes inhibit calmodulin-activated functions and protect against osmotic lysis; and finally, that there is a significant correlation (P less than 0.01) in terms of potency between these two actions of drugs. Data from the literature which bear on these mechanisms of drug actions suggest that the interactions between drugs and calmodulin, and drugs and the membrane, appear to be hydrophobic in nature. At this point, we do not know whether there is some causal relationship between calmodulin inhibition and the antihemolytic effect of drugs, or whether the two are simply a result of hydrophobic properties of drugs. Similarly, the roles of calmodulin inhibition and/or membrane antihemolysis in producing therapeutic efficacy are unknown.