Respiratory syncytial virus infection of airway epithelial cells, in vivo and in vitro, supports pulmonary antibody responses by inducing expression of the B cell differentiation factor BAFF.

BACKGROUND: The mechanisms regulating antibody expression within the human lung during airway infection are largely unknown. In this study, our objectives were to determine if infection with respiratory syncytial virus (RSV) upregulates expression of the B cell differentiation factors A proliferation inducing ligand (APRIL) and B cell activating factor of the TNF family (BAFF), if this is a common feature of viral airway infection, and how this is regulated in human airway epithelial cells. METHODS: We measured BAFF and APRIL protein expression in bronchoalveolar lavage (BAL) fluid from infants with severe RSV disease, and healthy control children, and in nasopharyngeal aspirates from preschool children with other single respiratory viral infections. We also measured mRNA expression in bronchial brushings from RSV-infected infants, and in RSV-infected paediatric primary airway epithelial cell cultures (pAEC). Beas-2B cell cultures were used to examine mechanisms regulating BAFF expression. RESULTS: BAFF protein and mRNA were elevated (in marked contrast with APRIL) in BAL and bronchial brushings, respectively, from RSV-infected infants. BAFF protein was also found in upper airway secretions from children with human metapneumovirus, H1N1, bocavirus, rhinovirus, RSV and Mycoplasma pneumoniae infection. BAFF mRNA and protein were expressed following in vitro RSV infection of both pAEC and Beas-2B cultures, with mRNA expression peaking 12-h postinfection. BAFF induction was blocked by addition of a neutralising anti-interferon-ß antibody or palivizumab. CONCLUSIONS: BAFF, produced through an interferon-ß-dependent process, is a consistent feature of airway infection, and suggests a role for the airway epithelia in supporting protective antibody and B cell responses in the lung.

Glucocorticoid resistance in chronic lymphocytic leukaemia is associated with a failure of upregulated Bim/Bcl-2 complexes to activate Bax and Bak.

Glucocorticoids (GCs) represent an important component of modern treatment regimens for fludarabine-refractory or TP53-defective chronic lymphocytic leukemia (CLL). However, GC therapy is not effective in all patients. The molecular mechanisms responsible for GC-induced apoptosis and resistance were therefore investigated in primary malignant cells obtained from a cohort of 46 patients with CLL. Dexamethasone-induced apoptosis was unaffected by p53 dysfunction and more pronounced in cases with unmutated IGHV genes. Cross-resistance was observed between dexamethasone and other GCs but not fludarabine, indicating non-identical resistance mechanisms. GC treatment resulted in the upregulation of Bim mRNA and protein, but to comparable levels in both GC-resistant and sensitive cells. Pre-incubation with Bim siRNAs reduced GC-induced upregulation of Bim protein and conferred resistance to GC-induced apoptosis in previously GC-sensitive cells. GC-induced upregulation of Bim was associated with the activation of Bax and Bak in GC-sensitive but not -resistant CLL samples. Co-immunoprecipitation experiments showed that Bim does not interact directly with Bax or Bak, but is almost exclusively bound to Bcl-2 regardless of GC treatment. Taken together, these findings suggest that the GC-induced killing of CLL cells results from the indirect activation of Bax and Bak by upregulated Bim/Bcl-2 complexes, and that GC resistance results from the failure of such activation to occur.

Active Ras induces heterodimerization of cRaf and BRaf.

Growth factor-induced signalling leads to activation of members of the Ras family and subsequent stimulation of different Raf isoforms. Within the mechanism of Raf activation, two isoforms of Raf, cRaf and BRaf, may cooperate. We investigated the relationship between cRaf and BRaf and found that active Ras induced heterodimerization of cRaf and BRaf, an effect that was dependent on the serine residue at position 621 of cRAF: Moreover, we also found that cRaf COOH-terminus constitutively associated with BRaf, whereas the NH(2) terminus did not, even in the presence of active RAS: These data suggest that Ras induces the cRaf-BRaf complex formation through the exposure of 14-3-3 binding sites in the COOH-terminus of cRAF: Thus, Ras-induced cRaf-Braf heterodimerization may explain the observed cooperativity of cRaf and BRaf in cells responding to growth factor signals.

Acetyl-boswellic acids are novel catalytic inhibitors of human topoisomerases I and IIalpha.

Acetyl-boswellic acids (acetyl-BA) are pentacyclic triterpenes derived from the gum resin of frankincense. We have previously shown that these compounds are effective cytotoxic agents, acting through a mechanism that appears to involve the inhibition of topoisomerase activity. We have now investigated the mechanism of action of acetyl-BA and show that these compounds are more potent inhibitors of human topoisomerases I and IIalpha than camptothecin, and amsacrine or etoposide, respectively. Our data demonstrate that acetyl-BA and, to a lesser extent, some other pentacyclic triterpenes, such as betulinic acid, ursolic acid, and oleanolic acid, inhibit topoisomerases I and IIalpha through a mechanism that does not involve stabilization of the cleavable complex or the intercalation of DNA. Surface plasmon resonance analysis revealed that topoisomerases I and IIalpha bind directly to an immobilized derivative of acetyl-BA. This acetyl-BA derivative interacts with human topoisomerases through high-affinity binding sites yielding K(D) values of 70.6 nM for topoisomerase I and 7.6 nM for topoisomerase IIalpha. Based on our data, we propose that acetyl-BA inhibit topoisomerases I and IIalpha through competition with DNA for binding to the enzyme. Thus, acetyl-BA are a unique class of dual catalytic inhibitors of human topoisomerases I and IIalpha.

Mitogenic signaling of Ras is regulated by differential interaction with Raf isozymes.

In the mitogenic signaling cascade interaction of Ras with Raf represents a critical step for the regulation of cell growth and differentiation. The major effector of Ras, the serine/threonine kinase Raf exists as three isoforms with different tissue distributions. We demonstrate that transient transfection of oncogenic Ha-Ras leads to a preferential activation of endogenous c-Raf-1 in HEK 293 cells as opposed to A-Raf. In vitro binding studies using purified Ras binding domains of Raf as well as in vivo bindings tests with full length molecules reveals significantly lower binding affinities of A-Raf to Ha-Ras as compared to other Raf isoforms. The Ras-binding interface of c-Raf differs from A-Raf by a conservative Arg to Lys exchange at residue 59 or 22 respectively. Mutational analysis reveals that this residue represents a point of isozyme discrimination: c-Raf-R59K binds Ha-Ras weaker than the wildtype, likewise A-Raf-K22R increases its affinity to Ha-Ras in vivo and in vitro. Differential binding affinities are reflected in downstream signaling. Immunecomplex kinase assays reveal that Ha-Ras mediated Raf activation is decreased for c-Raf-R59K and increased for A-Raf-K22R when compared to the respective wildtype forms. Thus our observations introduce a new level of isoform discrimination in Ras/Raf signaling as a functional consequence of a conservative amino acid exchange in the Ras binding domains.

Binding of Gbetagamma subunits to cRaf1 downregulates G-protein-coupled receptor signalling.

Receptors of the seven transmembrane domain family are coupled to heterotrimeric G proteins [1]. Binding of ligand to these receptors induces dissociation of the heterotrimeric complex into free GTP-Galpha and Gbetagamma subunits, which then interact with their respective effector molecules to stimulate specific cellular responses. In some cases, these cellular responses involve mitogenic signalling [2]. The mitogen-activated protein (MAP) kinase cascade is initiated by the protein kinase cRaf1 and links growth factor receptor signalling to cell growth and differentiation [3]. The main activator of cRaf1 is the small GTP-binding protein Ras [4], and the binding of cRaf1 to GTP-Ras translocates cRaf1 to the plasma membrane, where it is activated [5]. It has been reported that cRaf1 associates directly with the beta subunit of heterotrimeric G proteins in vitro, and with the betagamma subunit complex in vivo [6], but the role of this association is not yet understood. Here, we show that cRaf1 associates with Gbeta1gamma2, and that this association in mammalian cells is significantly enhanced when active p21(Ras) is present or when cRaf1 is otherwise targeted to the membrane. Association with Gbeta1gamma2 has no effect on the kinase activity of cRaf1, but cRaf1 can affect Gbetagamma-mediated signalling events. Thus, membrane-localised cRaf1 inhibits G-protein-coupled receptor (GPCR)-stimulated activation of phospholipase Cbeta (PLCbeta) by sequestration of Gbetagamma subunits, an effect also observed with endogenous levels of cRaf1. Our data suggest that cRaf1 may be an important regulator of signalling by Gbetagamma, particularly in those GPCR systems that stimulate the MAP kinase cascade through the activation of p21(Ras).

Activated platelets induce tissue factor expression on human umbilical vein endothelial cells by ligation of CD40.

CD40 is a type I member of the tumour necrosis factor (TNF) receptor superfamily of proteins, and is present on a wide variety of cells including vascular endothelial cells. Ligation of this receptor on endothelial cells is known to increase expression of inflammatory adhesion molecules. We have recently demonstrated that platelets express the ligand of CD40 (CD154) within seconds of exposure to agonist, and interact with endothelial cells to participate directly in the induction of an inflammatory response. Here we show that activated platelets induce tissue factor (TF) expression on endothelial cells in a CD40/CD154-dependent manner, and that the magnitude of this response can equal that induced by TNFe. Moreover, CD40 ligation on endothelial cells downregulates the expression of thrombomodulin. We also show that CD40-mediated TF expression is less sensitive to inhibition with the oxidative radical scavenger pyrrolidine dithiocarbamate than is that mediated by TNFalpha, indicating that CD40 has a distinct signalling pathway. Tissue factor is a cell membrane protein which functions as the main trigger of the extrinsic pathway of blood coagulation, and its expression on endothelial cells is implicated in wound healing and angiogenesis. Since platelets are among the first cells involved in haemostasis following tissue injury, our data showing that ligation of CD40 by CD154 induces a procoagulant phenotype on vascular endothelial cells suggests that platelets may play an important role in the induction of wound healing.

CD40 ligand on activated platelets triggers an inflammatory reaction of endothelial cells.

CD40 ligand (CD40L, CD154), a transmembrane protein structurally related to the cytokine TNF-alpha, was originally identified on stimulated CD4+ T cells, and later on stimulated mast cells and basophils. Interaction of CD40L on T cells with CD40 on B cells is of paramount importance for the development and function of the humoral immune system. CD40 is not only constitutively present on B cells, but it is also found on monocytes, macrophages and endothelial cells, suggesting that CD40L has a broader function in vivo. We now report that platelets express CD40L within seconds of activation in vitro and in the process of thrombus formation in vivo. Like TNF-alpha and interleukin-1, CD40L on platelets induces endothelial cells to secrete chemokines and to express adhesion molecules, thereby generating signals for the recruitment and extravasation of leukocytes at the site of injury. Our results indicate that platelets are not only involved in haemostasis but that they also directly initiate an inflammatory response of the vessel wall.

Analysis of CD9, CD32 and p67 signalling: use of degranulated platelets indicates direct involvement of CD9 and p67 in integrin activation.

The use of agonist monoclonal antibodies (mAbs) to probe the signalling function of platelet membrane proteins is severely limited by the dependence of the mAb effect on Fc-FcgammaRII interaction. Furthermore, in addition to its anchoring role, the FcgammaRII receptor itself generates a stimulation signal resulting in granule secretion. Platelet stimulation by the released granule contents can then further obscure the original activation signal. Here we demonstrate that these problems are largely overcome by the use of platelets which had been degranulated with thrombin prior to stimulation with mAbs. We found that, like intact cells, degranulated platelets could also be activated and induced to aggregate by mAbs against a 67 kD membrane protein (known as PTA1) and CD9, and by crosslinked CD32 (FcgammaRII). However, the signal generated by crosslinked FcgammaRII was weak compared with that induced by the other monoclonal antibodies. Thus, by diminishing the FcgammaRII signal contribution, we have succeeded for the first time to clearly dissect the target antigen signal from that generated by FcgammaRII. In addition to differences in the degree of aggregation, analysis of the signals generated by each mAb showed differences in Ca2+ fluxes and protein phosphorylation. Moreover, the signals generated by CD9 and PTA1 antigens differed significantly in their sensitivity to PKC inhibition or ADP-ribosylation of the small GTP-binding protein rhoA. Despite these differences, the signals initiated by all three antigens converged to a common signalling pathway which included activation of tyrosine kinase(s). The pattern of protein phosphorylation strongly resembled that induced by gpIIb/IIIa-mediated platelet interaction with macromolecular ligands and by mutual cell contact. The multiple intercellular links formed by mAb would have a similar effect since the Fc-receptor anchorage required for antigen stimulation is already known to be provided by adjacent cells. The present findings suggest that the function of both CD9 and PTA1 antigens is closely associated with gpIIb/IIIa activation.

The platelet antigens CD9, CD42 and integrin alpha IIb beta IIIa can be topographically associated and transduce functionally similar signals.

Investigation of the specific effects of different mAb known to stimulate platelets (agonist mAb) is complicated by interaction of the Fc portion of these mAb with the platelet Fc gamma RII. This has led to the conclusion that nearly all agonist-mAb-induced activation of platelets is mediated by this receptor. However, the target antigen-mediated signal can be analysed provided that the effects of Fc gamma RII engagement can either be reduced or eliminated. We have therefore blocked platelet Fc gamma RII with IV.3 Fab fragments (an anti-Fc gamma RII mAb), and stimulated the platelets by cross-linking intact agonist mAb with F(ab')2 fragments of an Fc-specific anti-mouse antibody. By analysing functional platelet responses and protein-tyrosine phosphorylation, we found that such non-Fc gamma RII-mediated cross-linking of CD9, CD42 and glycoprotein (gp) IIb/IIIa generates closely similar signals. Since this may indicate molecular associations, we analyzed the surface topography of platelets using the chemical cross-linking agent dithiobis(sulfosuccinimidyl propionate). We found that a proportion of CD9, gpIIb/IIIa and CD42 molecules associate with each other on the platelet surface membrane. Thus, our results suggest that these antigens are able to form a larger molecular complex and induce similar signals. Furthermore, cross-linking of CD9 and CD42 stimulated thrombasthenic platelets completely lacking gpIIb/IIIa. These data therefore indicate that CD9 and CD42 can signal independently of gpIIb/IIIa, and that signals generated by all these molecules may converge on a common pathway.

GTPgammaS-stimulated phospholipase D activation in human neutrophils occurs by protein kinase C-dependent and -independent pathways but not tyrosine kinases.

Addition of GTPgammaS to saponin-permeabilised human neutrophils activated both the NADPH oxidase and phospholipase D (PLD). This PLD activation was hardly affected by staurosporine or Ro31-8220 (at concentrations which inhibited PMA stimulated PLD activity), indicating that it was largely independent of protein kinase C (PKC). This GTPgammaS stimulated PLD activity was enhanced by 1 mM ATP, but this ATP-enhanced activity was blocked by inhibitors of PKC. Addition of GTPgammaS resulted in very low levels of phosphorylation on tyrosine residues, but higher levels of phosphorylation on serine/threonine residues. Addition of pervanadate hydroperoxides stimulated phosphorylation on tyrosine residues and activated PLD which was blocked by addition of inhibitors of tyrosine kinases. Thus, GTPgammaS can stimulate PKC-dependent and -independent pathways of PLD activation. Whilst phosphorylation on tyrosine residues can result in activation of PLD, this is regulated independently of activation via G-proteins.

Properties of fibrinogen cleaved by Jararhagin, a metalloproteinase from the venom of Bothrops jararaca.

Haemorrhagic metalloproteinases from Bothrops jararaca and other venoms degrade vessel-wall and plasma proteins involved in platelet plug and fibrin clot formation. These enzymes also cause proteolytic digestion of fibrinogen which has been suggested to cause defective platelet function. Fibrinogen degradation by jararhagin, a metalloproteinase from B. jararaca, and the effect of jararhagin fibrinogenolysis on both platelet aggregation and fibrin clot formation were investigated. Jararhagin was found to cleave human fibrinogen in the C-terminal region of the A alpha-chain giving rise to a 285-290 kDa fibrinogen molecule lacking the A alpha-chain RGD 572-574 platelet-binding site. Platelet binding and aggregation of ADP-activated platelets is unaffected by this modification. This indicates that the lost site is not essential for platelet aggregation, and that the remaining platelet binding sites located in the N-terminal portion of A alpha chains (RGD 95-97) and the C-terminal of gamma chains (dodecapeptide 400-411) are unaffected by jararhagin-digestion of fibrinogen. Fibrin clot formation with thrombin of this remnant fibrinogen molecule was defective, with poor polymerization of fibrin monomers but normal release of FPA. The abnormal polymerization could be explained by the loss of one of the two complementary polymerization sites required for side-by-side association of fibrin protofibrils. Jararhagin-induced inhibition of platelet function, an important cause of haemorrhage in envenomed patients, is not caused by proteolysis of fibrinogen, as had been thought, and the mechanism remains to be elucidated.

The peanut-agglutinin (PNA)-binding surface components of malignant plasma cells.

Plasma cells within bone marrow aspirates from multiple myeloma patients have been shown to be reactive with the lectin peanut agglutinin (PNA). This has been recently exploited by using PNA for purging bone marrow of malignant cells in autotransplantation therapy of the disease. The purpose of this investigation was to isolate and characterize the PNA-binding proteins of myeloma cells. We used the malignant plasma cell-derived line Karpas-620 (K620) as a model, and showed by affinity chromatography, SDS-PAGE, and immunoprecipitation that, among several PNA-binding proteins, a major one is an incompletely sialylated form of CD44. CD44 is a well-known homing receptor protein which is rich in carbohydrate and usually completely sialylated so that it does not react with PNA. We have then examined the PNA reactivity of myeloma cells from different patients and showed a clear difference in the profile of PNA-binding proteins from case to case. Moreover, in contrast to K620 cells, some of the patient plasma cells tested did not have a PNA-binding form of CD44. In conclusion, therefore, we have shown that a number of different proteins participate in PNA binding by malignant plasma cells. Moreover, we have demonstrated a novel, incompletely sialylated form of CD44 on a myeloma cell line. It is known that the level of glycosylation of CD44 and other proteins may affect their function, but how this relates to the malignant behaviour of plasma cells remains to be determined.

Role of Fc gamma RII in platelet activation by monoclonal antibodies.

mAb are being widely used to probe the function of cell-surface proteins. The cell stimulation that may be produced is often dependent on mAb interaction with both the target Ag and FcR. However, it remains unclear whether these interactions take place on the same cell or between adjacent cells and whether the FcR plays an anchoring or signaling role. Using the model of platelet activation, we demonstrate that two different Fc-dependent mAb, LeoA1 and ALB6, both activate the cell by forming intercellular links between Ag on one cell and FcR on the opposing platelet. We also show that the mAb differ with respect to the relative roles of target Ag vs FcR in provision of the stimulation signal. Thus Fc gamma RII played a mainly anchorage role in LeoA1 stimulation, whereas its role in ALB6 stimulation was mainly signaling. Therefore the precise contribution of each of these roles to the overall effect of a stimulatory antibody should be determined before the antibody is used as a specific functional probe.

Myristic acid is incorporated into the two acylatable domains of the functional glycoprotein CD9 in ester, but not in amide bonds.

CD9 is a signal-initiating glycoprotein of uncertain membrane insertion which contains more than one locus of acylation and is distinguished by being the major acylatable platelet protein. The N-terminus of CD9 is blocked to Edman degradation. We investigated whether [3H]myristic acid could be incorporated into CD9, whether that incorporation occurred via an amide linkage, and whether myristate and palmitate were differentially incorporated into the two domains. Pulse-labeling studies, performed on the human osteogenic sarcoma cell line SKOSC which expresses 22 and 24 kDa variants of CD9 demonstrated that the respective precursors of 20.5 and 23 kDa were not radiolabeled by either [3H]myristic acid or [3H]palmitic acid, but that both fatty acids could be ligated to CD9 during the later stages of protein maturation. The failure to incorporate myristic acid cotranslationally suggest that CD9 does not contain amino-terminal amide-bonded myristic acid. Incorporation of radiolabel from both fatty acids proceeded very rapidly and could be visualized after a 10 s pulse. Although myristic acid was partially metabolized into palmitic acid, incorporation of authentic [3H]myristate into CD9 could be demonstrated. The myristic acid bonds were shown to be as sensitive to hydroxylamine treatment as those linking palmitate. Both fatty acids were also incorporated into CD9 in hydroxylamine-sensitive bonds in the presence of cycloheximide, reaching 30-40% of the levels in untreated controls. The sensitivity of myristate ligands to hydroxylamine demonstrates that this fatty acid is not linked via amide, but rather via ester bonds. The sensitivity of [3H]myristate and [3H]palmitate bonds to 2-mercaptoethanol further suggests that either fatty acid is linked via thioester rather than hydroxyester bonds to each domain on CD9. Limited proteolysis analysis with Staphylococcus aureus V8 proteinase of CD9, labeled in the absence or presence of cycloheximide, showed that [3H]myristic acid and [3H]palmitic acid labeled identical peptides, and to the same extent, suggesting that myristate is an alternative substrate for the transacylase(s) involved.

Anti-CD9 monoclonal antibodies induce homotypic adhesion of pre-B cell lines by a novel mechanism.

Anti-CD9 mAb are known agonists of platelet aggregation, but have not been implicated in cell-cell adhesion. We show here in an experimental system that the anti-CD9 mAb 50H.19, ALB6, and BA-2 can induce rapid, and irreversible, homotypic aggregation of the CD9-positive pre-B lymphoblastoid cell lines NALM-6 and HOON, but not of the CD9-negative B cell line Raji. The specificity of the response is indicated by the failure to effect aggregation with mAb directed to CD24, or to HLA class I Ag. The initiation of strong homotypic aggregates of lymphoid cells is a property ascribed to lymphocyte function-associated Ag-1 (LFA-1), a member of the beta 2 subfamily of leukocyte integrins. We show that CD9-induced aggregation is an active process which proceeds at 37 degrees C, but not at 4 degrees C, requires the expenditure of metabolic energy, and a functioning cytoskeleton, and is not inhibited by Arg-Gly-Asp-Ser peptide. These are properties described for LFA-1-mediated aggregation. However, because beta 2-integrins are not expressed on NALM-6 or HOON cells, they are not the mediators of CD9-induced aggregation. In contrast to LFA-1-mediated adhesion which is Mg2+ dependent, CD9-induced adhesion has an absolute requirement for Ca2+, but not Mg2+, indicating that a Ca2(+)-dependent event is sufficient for adhesion. However, Mg2+ enhances adhesion even at optimal concentrations of Ca2+, implicating an additional Mg2(+)-dependent event which requires Ca2+ to be effective. These findings suggest that CD9 Ag regulates a novel mechanism for promoting tight cell-cell adhesion which requires both Ca2+ and Mg2+ for optimal expression.

Evidence that monoclonal antibodies against CD9 antigen induce specific association between CD9 and the platelet glycoprotein IIb-IIIa complex.

Monoclonal antibodies to the CD9 antigen are powerful platelet agonists. We report here the novel finding that the anti-CD9 monoclonal antibodies 50H.19 and ALB6 promote physical association between CD9 antigen and the glycoprotein IIb-IIIa complex (GPIIb-IIIa) component of the platelet fibrinogen receptor. The monoclonal antibodies do not consistently immunoprecipitate proteins other than CD9 from 125I-labeled human platelets even if the platelets are first treated with the homobifunctional cross-linking reagent dithiobis(succinimidyl propionate), indicating that CD9 antigen is not physically associated with other membrane proteins in the resting state. However, the addition of agonistic concentrations of either monoclonal antibody before cross-linking results in the coprecipitation of proteins corresponding in mobility and peptide composition to GPIIb, and GPIIIa. The association of CD9 with the GPIIb-IIIa complex is unaffected by a combination of aspirin and ADP scavengers sufficient to abrogate anti-CD9 monoclonal antibody-induced platelet aggregation, and is therefore not dependent upon thromboxane- and ADP-mediated pathways of intracellular signalling. The specificity of the association is demonstrated by the lack of other coprecipitating major proteins, by the requirement for induction by anti-CD9 monoclonal antibodies, and by the failure to promote reciprocal association with either of the anti-GPIIb-IIIa complex monoclonal antibodies P2 or HuP1-m1a.

The functional glycoprotein CD9 is variably acylated: localization of the variably acylated region to a membrane-associated peptide containing the binding site for the agonistic monoclonal antibody 50H.19.

Recent studies have shown that [3H]palmitic acid strongly labels both glycosylated forms (gp22 and gp24) of the signal-initiating cell surface glycoprotein CD9. We performed a two-dimensional limited proteolysis analysis with Staphylococcus aureus V8 proteinase in order to localize the palmitylation sites to final peptides on both glycosylated forms of CD9. Analysis of [3H]leucine- and [3H]amino acid mixture-labeled gp22 delineated 4 final peptides of 11, 8, 7 and 4 kDa. gp24 produced a similar pattern with the exception that the 11 kDa peptide was replaced by an N-glycosylated 13 kDa peptide. Since all four final peptides (total molecular mass of 30/32 kDa) could not be accommodated by a parent molecule of 22/24 kDa, it is likely that one of the final peptide coexists in two differently modified states. Palmitic acid labeled the 11 kDa/13 kDa final peptides, and the 7 kDa final peptide, with equal intensity, but was not incorporated into the 4 kDa final peptide, demonstrating that fatty acid is ligated in two distinct regions of the molecule. The 8 kDa final peptide was strongly labeled by [3H]palmitic acid, but only weakly by [3H]leucine. We present evidence that this peptide is derived by further acylation of the region defined by the 7 kDa peptide, and that this occurs in only 15% of the molecules. Palmitic acid is turned over faster at these additional sites, indicating that they may be more accessible to membrane transacylases. Proteolysis of CD9 on the intact cell with papain enabled the highly acylated region to be localized to a membrane-associated fragment which contains the binding site for the agonistic monoclonal antibody 50H.19. The co-localization of a functional domain with a region of variable acylation suggests that acylation events may play a role in the transduction of the signal initiated by interaction of the antibody with CD9.

The functional cell surface glycoprotein CD9 is distinguished by being the major fatty acid acylated and a major iodinated cell-surface component of the human platelet.

We showed that a 22 kDa protein (which comigrated with the leukocyte differentiation antigen CD9 as determined by immunoblotting with the platelet-activating mAb 50H.19) is a major iodinated component of the platelet surface. The iodinated protein was identified as CD9 by limited proteolysis analysis. The major acylated protein in platelets incubated with [3H]palmitic acid also had a mobility of 22 kDa. The radiolabelled fatty acid in CD9 appears to be ester bonded, as it is removed by treatment with hydroxylamine. Non-enzymatic ligation of the fatty acid is not involved. Since platelets lack protein synthetic capacity, the palmitolation of a surface protein indicates the existence of a plasma-membrane located transacylase which functions independently of protein synthesis. Limited proteolysis analysis of the palmitylated protein obtained by immunoprecipitation with mAb 50H.19 confirmed its identity as CD9. An additional novel minor component of 27 kDa was detected in platelets by immunoprecipitation of 125I-surface-labelled, or [3H]palmitic acid-labelled protein, and by immunoblotting with mAb 50H.19. The analogous cleavage patterns obtained by the limited proteolysis analysis of the 22, 24 and 27 kDa glycoproteins suggest that they may be differently modified variants of a single polypeptide.

Characterization of cardiac calsequestrin.

Calsequestrin, a calcium-binding protein found in the sarcoplasmic reticulum of muscle cells, was purified from rabbit and canine cardiac and skeletal muscle tissue. The amino acid compositions and amino-terminal sequences of skeletal and cardiac calsequestrin from rabbit and dog were determined. The amino acid composition of the cardiac form was very similar to the skeletal form. The amino-terminal sequence of the cardiac form was homologous to, but not identical with, the amino-terminal sequence of the skeletal form of the protein. Few species differences in the amino-terminal sequences were observed. The calcium-binding capacity of the cardiac form was half the capacity of the skeletal form although the affinities of the two forms of calsequestrin for Ca2+ were similar (Kd = 1 mM). Calcium binding to the cardiac form induced structural changes in the protein as determined by circular dichroism and intrinsic fluorescence spectroscopy. The alpha-helical content of cardiac calsequestrin increased from 3.5% to 10.9% upon binding calcium, while the intrinsic fluorescence of the protein increased 14%. Potassium ions also affected the conformation of cardiac calsequestrin.