Medical SciencesSer-752 -> Pro mutation in the cytoplasmic domain of integrin .83 subunit and defective activation of platelet integrin aIlbP3 (glycoprotein lib-Illa) in a variant of Glanzmann thrombasthenia (molcular genetlcs/flbnnogen receptor/Arg-Gly-Asp/poyerase chan reacdon) YI-PING CHEN*, ISABELLE DJAFFAR*, DOMINIQUE PIDARDt, BEAT (8,9), as well as for aLI32 (10). At least in the case of amb43 activation, modulation of affinity for its ligands is coordinated with conformational changes of the integrin (7,11). What remains obscure is the exact link between intracellular signals and aIIbP3 conformation modulation.Glanzmann thrombasthenia is a hereditary bleeding disorder characterized by the absence of platelet aggregation due to either an absence or a functional alteration of the integrin aII03 (12). In several cases, the genetic basis for the quantitative defect was found to be associated with alterations within the genes of ajib,83 subunits-i.e., either aib (glycoprotein lIb) (13,14) or 83 (glycoprotein IIa) (14, 15). In addition, the qualitative defects in aIlbm3 found in variants of Glanzmann thrombasthenia have also become the focus of investigations, and two mutations in Glanzmann variants that disrupted ligand binding sites of (3 have been characterized (16-18). Here we report a variant of Glanzmann thrombasthenia whose biochemical and genetic analyses are consistent with the idea that the functional defect is due to replacement of serine by proline in position 752 in the cytoplasmic domain of P3 and that this mutation impairs the coupling between cellular activation and up-regulation of arnb,83 affinity for fibrinogen. To our knowledge this is the first point mutation reported affecting integrin activation. METHODSPatient's Case. Patient P.'s case of variant Glanzmann thrombasthenia has been previously reported (19). Briefly, patient P. exhibited a life-long bleeding tendency with failure of platelets to aggregate and to bind fibrinogen in response to agonists such as ADP, collagen, or thrombin. The patient's platelets could support clot retraction. The presence of alb/3 in the patient's platelets was established by immunocytochemistry.Protein Analysis and Monoclonal Antibody Binding. Blood was collected in the presence of 5 mM EDTA, and platelets were prepared by conventional methods (20). Platelet proteins were solubilized in 2% NaDodSO4 and two-dimensional nonreduced/reduced NaDodSO4/PAGE was performed as described (21). Total platelet proteins (200 ,ug) were solubilized in 2% NaDodSO4 and separated in a first dimension in a 6% polyacrylamide tube gel. Disulfide bonds were reduced by soaking the gel in 10%6 (vol/vol) 2-mercaptoethanol, and the proteins were then separated again in a perpendicular direction in a 7-12% polyacrylamide slab gel. Proteins were detected by staining with Coomassie blue-R.The monoclonal antibody AP-2, which is specific for glycoprotein Ilb-IIla (22), the a11b13 integrin (a generous gift Abbreviations: RGDS, Arg-Gly-Asp-Ser; nt, nucleotide(s). lTo whom reprint requests...

Hematogenous metastasis involves adhesive interactions between blood-borne tumor cells and the vessel wall. By the use of in vitro assays, the adhesion of human melanoma, osteosarcoma, and kidney carcinoma (but not colon carcinoma) cell lines was shown to involve the cytokine-inducible endothelial cell surface protein inducible cell adhesion molecule 110 (INCAM-110) and the alpha 4 beta 1 integrin, molecules normally involved in endothelial-leukocyte interactions. Tumor adhesion to human endothelial cell monolayers was increased 1.9- to 8.2-fold by endothelial activation with the cytokine tumor necrosis factor (TNF) and inhibited by the anti-INCAM-110 monoclonal antibody (mAb) E1/6. Each of these tumor cells expressed members of the beta 1 integrin family of adhesion molecules, and antibodies to the alpha 4 and beta 1 integrin subunits inhibited tumor-endothelial adhesion (48-87% inhibition). A cDNA encompassing the three N-terminal Ig-like domains of vascular cell adhesion molecule 1 (VCAM-1) encoded a protein recognized by the anti-INCAM-110 mAb E1/6 and, when captured onto plastic, supported melanoma cell adhesion by an alpha 4 integrin-dependent mechanism. In contrast to mAb E1/6, a second anti-INCAM-110 mAb Hu8/4 neither inhibited adhesion to activated endothelium nor bound the first three Ig-like domains of INCAM-110/VCAM-1. These data indicate that the adherence of several human tumors to activated endothelium is mediated by an interaction of alpha 4 beta 1 integrin and the N-terminal Ig-like domains of endothelial INCAM-110/VCAM-1. Tumor acquisition of the alpha 4 integrin subunit and endothelial expression of INCAM-110 may affect the frequency and distribution of metastasis.

SummaryThe human platelet antigen (HPA) 3 system is expressed on GPIIb, one subunit of GPIIb-IIIa, the platelet fibrinogen receptor. It was recently shown that HPA-3 was associated with an Ile843/Ser polymorphism. To investigate further HPA-3 determinant structure, we localized an HPA-3a determinant, recognized by the alloantiserum Leka, within the last 29 amino acids of GPIIbα. This region encompasses the polymorphic Ile843, which, as expected, is substituted into Ser in Leka-negative individuals, as shown by DNA sequence after polymerase chain reaction on platelet RNA. In addition, contribution of glycosylation to the determinant structure was demonstrated since the Leka antigenicity was strongly decreased after specifically removing nonterminal O-linked sugars, but not terminal sialic acids. We have thus refined the localization of an HPA-3a determinant within the last 29 amino acids, including Ile843, of GPIIb heavy chain, and shown that the Leka HPA-3a determinant is dependent, in part, upon the serine-linked carbohydrates adjacent to Ile/Ser843.

Previous biochemical studies suggested that the human platelet Ca2+ATPase system may be cell-specific. To test this hypothesis, we first undertook the molecular cloning of Ca2+ATPase from human erythroleukaemia (HEL) cells, because this human cell line exhibits megakaryocytic features and expresses a Ca2+ATPase that cross-reacts with platelet Ca(2+)-ATPase. For this cloning, an HEL-cell cDNA library was screened with a rat cardiac Ca2+ATPase cDNA probe. The insert of the longest clone isolated was 3.9 kb and its sequence displayed a 100% identity with that of the non-muscle human Ca2+ATPase 2-b isoform, termed SERCA2-b (sarco-endoplasmic-reticulum Ca2+ATPase). The 3.9 kb cDNA covered a subtotal coding region and part of the 3' non-coding end of the SERCA2-b mRNA. It cross-hybridized with the 4 kb transcript species of cardiac SERCA2-a and with non-muscle SERCA2-b mRNAs, but not with fast-skeletal-muscle SERCA1 mRNA. We next confirmed that SERCA2-b was a component of the platelet Ca2+ATPase system because (1) the platelet clones isolated from a platelet cDNA library exhibited a 100% homology with HEL-cell cDNA; (2) SERCA2-b mRNA was amplified by PCR on total platelet RNA and (3) platelet Ca2+ATPase cross-reacted with a polyclonal SERCA2-b-specific antiserum. Platelets therefore contain a Ca2+ATPase definitely identified as the SERCA2-b isoform of Ca2+ATPase, thus eliminating the possibility that they only contain a single specific Ca2+ATPase.

No abstract

A cDNA for integrin beta 3 isolated from a human erythroleukaemia (HEL) cell library contained a 340 bp insert at position 1281. This mRNA, termed beta 3c, results from the use of a cryptic AG donor splice site in intron 8 of the beta 3 gene, and is different from a previously described alternative beta 3 mRNA. The predicted open reading frame of beta 3C stops at a TAG stop codon 69 bp downstream from position 1281. It starts with the signal peptide and the 404 N-terminal extracellular residues of beta 3, encompassing the ligand binding sites, followed by 23 C-terminal intron-derived residues, corresponding to a truncated form of beta 3 lacking the cysteine-rich, transmembrane and cytoplasmic domains. Expression of beta 3C mRNA was demonstrated in human platelets, megakaryocytes, endothelial cells and HEL cells by reverse transcriptase/PCR. The beta 3C transcript was also demonstrated in the mouse, suggesting its conservation through evolution. Finally, a 60 kDa polypeptide corresponding to the beta 3C alternative transcript was demonstrated in platelets by Western blotting using a polyclonal antibody raised against a synthetic peptide designed from the beta 3C intronic sequence. Taken together, these results suggest a biological role for beta 3C, the first alternative transcript showing an altered extracellular domain of a beta integrin.

No abstract

No abstract