Ectocytosis caused by sublytic autologous complement attack on human neutrophils. The sorting of endogenous plasma-membrane proteins and lipids into shed vesicles

Fused p47phox and p67phox truncations efficiently reconstitute NADPH oxidase with higher activity and stability than the individual components

Activation of the neutrophil NADPH oxidase occurs via assembly of the cytosolic regulatory proteins p47(phox), p67(phox), and Rac with the membrane-associated flavocytochrome b(558). Following cell-free activation, enzymatic activity is highly labile (Tamura, M., Takeshita, M., Curnutte, J. T., Uhlinger, D. J., and Lambeth, J. D. (1992) J. Biol. Chem. 267, 7529-7538). To try to stabilize the activity and investigate the nature of the complex, fusion proteins between p47N-(1-286) and p67N-(1-210) were constructed. In a cell-free system, a fusion protein, p67N-p47N, had an 8-fold higher efficiency and produced a higher activity than the individual proteins, and also resulted in an 8-fold improved efficiency for Rac and a lowered K(m) for NADPH. O(2) generating activity was remarkably stabilized by using p67N-p47N. The cytosolic proteins fused in the opposite orientation, p47N-p67N, showed similar activity and stability as individual proteins, but with a 4-fold improved efficiency compared with the individual cytosolic factors. In the system efficiency for Rac and affinity for NADPH were also higher than those with the nonfused components. Interestingly, the p67N-p47N showed nearly full activation in the absence of an anionic amphifile in a cell-free system containing cytochrome b(558) relipidated with phosphatidylinositol- or phosphatidylserine-enriched phospholipid mixtures. From the results we consider multiple roles of anionic amphifiles in a cell-free activation, which could be substituted by our system. The fact that a fusion produces a more stable complex indicates that interactions among components determine the longevity of the complex. Based on the findings we propose a model for the topology among p47N, p67N, and cytochrome b(558) in the active complex.

Acquisition of intact allogeneic human leukocyte antigen molecules by human dendritic cells

In an attempt to transduce monocyte-derived dendritic cells (DCs) by a retroviral vector coding for a cell surface marker, we were confronted by the observation of high transfer of the surface molecule in the absence of vector proviral DNA in the treated cells. Indeed, DCs acquired the surface marker by a mechanism independent of the vector machinery, requiring cell-to-cell contact and involving transfer of lipids and a variety of intact membrane proteins. Most important, this property of DCs also includes acquisition of foreign human leukocyte antigen (HLA) molecules. Consequently, DCs become immunological hybrids as they display their own and foreign HLA molecules. The newly acquired HLA is fully functional because it allows recognition by allo-specific T lymphocytes and the binding and presentation of antigen peptides.

Acquisition of intact allogeneic human leukocyte antigen molecules by human dendritic cells

In an attempt to transduce monocyte-derived dendritic cells (DCs) by a retroviral vector coding for a cell surface marker, we were confronted by the observation of high transfer of the surface molecule in the absence of vector proviral DNA in the treated cells. Indeed, DCs acquired the surface marker by a mechanism independent of the vector machinery, requiring cell-to-cell contact and involving transfer of lipids and a variety of intact membrane proteins. Most important, this property of DCs also includes acquisition of foreign human leukocyte antigen (HLA) molecules. Consequently, DCs become immunological hybrids as they display their own and foreign HLA molecules. The newly acquired HLA is fully functional because it allows recognition by allo-specific T lymphocytes and the binding and presentation of antigen peptides.

Secretion of the galectin family of mammalian carbohydrate-binding proteins

Galectins are cytosolic proteins that lack any signal sequence for transport into the endoplasmic reticulum and are not glycosylated, although several galectins contain consensus sites for N-glycosylation, indicating that these proteins do not traverse the ER-Golgi network. However, there is abundant evidence for the extracellular localisation of some galectins at cell surfaces, in the extracellular matrix and in cell secretions consistent with other evidence for extracellular roles of galectins as modulators of cell adhesion and signalling. How then are galectins secreted if not through the classical secretory pathway? Do all galectins share the same secretory pathway? Can a particular galectin utilise more than one secretory pathway? If galectins play important extracellular roles how is their secretion regulated in relation to function? These are still largely unanswered questions but recent studies are beginning to give glimpses into some novel aspects of the secretion of these intriguing proteins.

Ectosomes Released by Human Neutrophils Are Specialized Functional Units

Here we show that human polymorphonuclear leukocytes (PMN) release ectosomes independently of complement attack during their activation both in vitro and at the site of inflammation in vivo. Patterns of biotinylated proteins on the surface of PMN and on PMN-derived ectosomes indicated a specific sorting of cell surface proteins into and out of ectosomes. Ectosomes expressed clusters of complement receptor 1 (CR1), which allowed them to bind efficiently to opsonized bacteria. Myeloperoxidase and human leukocyte elastase, both stored within the azurophilic granules of PMN, were found to colocalize on ectosomes with CR1. Furthermore, myeloperoxidase colocalized with human leukocyte elastase. In contrast, not present on CR1-expressing ectosomes were CD63, a selective marker for the azurophilic granules, and CD14, which is located within the same granules and the secretory vesicles as CR1. Of the other complement regulatory proteins expressed by PMN, only CD59 colocalized with CR1, while CD55 and CD46 were almost absent. Ectosomes released by activated PMN at the site of inflammation may function as a well organized element (ecto-organelle), designed to focus antimicrobial activity onto opsonized surfaces.

Determinants in the N-terminal domains of galectin-3 for secretion by a novel pathway circumventing the endoplasmic reticulum-Golgi complex

Galectin-3 is a beta-galactoside-binding protein that is secreted from many cells although the protein lacks a signal sequence for transfer into the endoplasmic reticulum and Golgi compartments and entry into classical secretory pathways. Previously it was shown that attachment of the first 120 amino acid residues of the N-terminal sequence of hamster galectin-3 to the cytoplasmic protein chloramphenicol acetyltransferase (CAT) supported the rapid secretion of the fusion protein from transiently transfected Cos cells under conditions in which CAT protein was not secreted. Here we report that progressive N-terminal truncation gradually reduced secretion of the fusion proteins, eventually to very low levels compared with the starting product, but did not totally eliminate secretion until a significant majority of the sequence was removed. Mutant CAT fusion proteins containing internal deletions in residues 97-120 of the galectin-3 N-terminal sequence were also secreted to a similar extent to the starting product, but further deletion of residues 89-96 abolished detectable secretion. Proline to alanine mutagenesis of the sequence YP(90)SAP(93)GAY in two secretion-competent CAT fusion proteins greatly reduced or abolished their secretion, whereas similar mutagenesis of proline pairings present elsewhere in the galectin-3 N-terminal segments of these proteins had no effect. The results indicate that this sequence is one essential determinant for secretion of galectin-3-CAT fusion proteins and by inference galectin-3, at least from transfected Cos cells. However, the short sequence of residues 89-96 by itself is insufficient to direct secretion of CAT fusion proteins and appears to be active only in the context of a larger portion of the galectin-3 N-terminal sequence.

Activated T Cells Acquire Endothelial Cell Surface Determinants During Transendothelial Migration

Activated T cells acquire endothelial cell (EC) plasma membrane constituents during transendothelial migration. This was assessed using an in vitro model system in which human peripheral blood CD4+ T cells migrated through confluent monolayers of HUVEC. Flow cytometry of migrated CD4+ T cells demonstrated that activated, but not resting, T cells acquired a variety of endothelial surface determinants, including CD31, CD49d, CD54, CD61, and CD62E. The extracellular domains of these molecules were detected on migrated T cells with mAbs, including those directed to the ligand-binding regions. A number of approaches were employed to document that the acquisition of these molecules was uniquely accomplished by activated T cells and clearly involved transfer from both resting and TNF-α-activated EC. Acquisition of endothelial markers by activated T cells occurred as part of the transfer of membrane components, as migrating T cells acquired EC membranes prelabeled with the lipophilic dye, 3,3′-dihexadecyloxacarbocyanine perchlorate (DiOC-16), along with EC surface proteins. Thus, during transendothelial migration, activated T cells acquire endothelial membrane components, and as a result may deliver them to perivascular sites.

Imaging bioluminescent indicators shows Ca2+ and ATP permeability thresholds in live cells attacked by complement

A series of permeability thresholds to Ca2+ metabolites and macromolecules, occurring at different times when cells are attacked by complement, has been established by imaging HeLa cells transiently expressing a recombinant cytosolic fusion protein of firefly luciferase and aequorin (luciferase-aequorin) to measure changes in ATP and cytosolic free Ca2+. Nuclear fluorescence of propidium was used as a measure of permeability to small molecules, and luciferase activity imaged to assess lysis. The rise in cytosolic free Ca2+ observed after C9 attack preceded by at least 60 s both the increase in propidium fluorescence, measured in single cells, and the decrease in ATP monitored by luciferase light emission. These effects were dependent on the concentration of C9. At concentrations of C9 up to 4 micrograms/ml no loss of luciferase-aequorin protein was detected at the end of the experiment. Thus the membrane integrity of the cells remained intact, even though the cells were permeable to propidium. These results confirmed our earlier observations that propidium permeability in cells attacked by complement was not a reliable measure of cell death. They also show that it is vital to take account of cellular heterogeneity if the mechanisms by which cells respond to membrane pore former attack are to be correctly interpreted.

Plasma membrane targetting, vesicular budding and release of galectin 3 from the cytoplasm of mammalian cells during secretion

Galectin 3, a 30 kDa galactoside-binding protein distributed widely in epithelial and immune cells, contains no signal sequence and is externalized by a mechanism independent of the endoplasmic reticulum (ER)-Golgi complex. We show here that hamster galectin 3 overexpressed in transfected cos-7 cells is secreted at a very low rate. A chimaera of galectin 3 fused to the N-terminal acylation sequence of protein tyrosine kinase p56(lck), Nt-p56(lck)-galectin 3, which is myristoylated and palmitoylated and rapidly transported to plasma membrane domains, is efficiently released from transfected cells indicating that movement of cytoplasmic galectin 3 to plasma membrane domains is a rate limiting step in lectin secretion. N-terminal acylation is not sufficient for protein secretion since p56(lck) and the chimaera Nt-p56(lck)-CAT are not secreted from transfected cells. The amino-terminal half of galectin 3 is sufficient to direct export of a chimaeric CAT protein indicating that part of the signal for plasma membrane translocation lies in the N-terminal domains of the lectin. Immunofluorescence studies show that Nt-p56(lck)-galectin 3 aggregates underneath the plasma membrane and is released by membrane blebbing. Vesicles of low buoyant density isolated from conditioned medium are enriched in galectin 3. The lectin is initially protected from exogenous collagenase but is later released in soluble protease-sensitive form from the lectin-loaded vesicles. Using murine macrophages, which secrete their endogenous galectin 3 at a moderate rate especially in the presence of Ca2+-ionophores, we were also able to trap a galectin 3-loaded vesicular fraction which was released into the culture supernatant.

Mutation of recombinant complement component C9 reveals the significance of the N-terminal region for polymerization

Complement component C9 binds to C5b-8 sites on target cells and polymerizes to form the membrane attack complex (MAC). The aim of the work reported here was to discover which region within C9 was responsible for protecting the globular protein against self-polymerization. Computer prediction modelling highlighted the domain at the N-terminus of C9, which was then investigated by site-directed mutagenesis. The mutated proteins were expressed using insect cells infected with baculovirus. Removal of 16, 20 or 23 amino acids at the N-terminus of C9 resulted in inactivation due to self-polymerization. In contrast, removal of 4, 8 or 12 amino acids resulted in a C9 that did not polymerize spontaneously, had two to threefold enhanced lytic activity on erythrocytes, and had increased binding to C5b-8 sites on rat neutrophils. These results suggest that the domain within the first 16 amino acids at the N-terminus of C9 is crucial in preventing the self-polymerization of the globular protein. We have also found that C9 contains a motif (27WSEWS31) common to a family of cytokine receptors that is similar to a tryptophan-rich motif (WEWWR) of the membrane pore formers, thiol-activated cytolysins. Mutation of this motif in C9 resulted in polymerized protein, consistent with this site keeping the N-terminus in a protected conformation and preventing premature self-polymerization.

CR1, CD35 in synovial fluid from patients with inflammatory joint diseases

OBJECTIVE

To investigate synovial fluid (SF) for the presence of CR1 and to study its relationship to SF leukocytes and to serum levels of soluble CR1 (sCR1) in patients with rheumatic diseases.

METHODS

Synovial fluids were collected from 35 patients with rheumatoid arthritis (RA) and 26 patients with other inflammatory joint diseases. Total CR1 in the SF and serum were measured with a sandwich enzyme-linked immunosorbent assay (ELISA) that recognized both soluble and transmembrane forms of CR1. The characteristics of CR1 in SF were analyzed by ultracentrifugation and by a second ELISA specific for transmembrane CR1.

RESULTS

CR1 was found in all SF samples tested (range 5-281 ng/ml). SF CR1 was higher in patients with RA (mean +/- SD 81 +/- 66 ng/ml) than in those with other inflammatory joint diseases (31.8 +/- 23.8 ng/ml) (P < 0.001). Serum sCR1 was not significantly increased in the patients compared with the normal subjects. There was no correlation between serum sCR1 and SF CR1. In 44% of the patients, the SF CR1 level was higher than the serum sCR1 level. A fraction (30-80%) of SF CR1 was pelleted by ultracentrifugation and, unlike serum sCR1, it reacted in an ELISA specific for transmembrane CR1. Thus, SF contained 2 forms of CR1: a membrane-associated and a soluble form, which was confirmed by sucrose density-gradient ultracentrifugation. SF CR1 levels correlated directly with the number of SF total leukocytes and polymorphonuclear leukocytes (PMN). These 2 forms of CR1 were also found in the supernatant of in vitro-activated PMN from normal subjects. SF CR1 exhibited the capacity to act as a cofactor for the factor I degradation of C3b.

CONCLUSION

CR1 is found in the SF of patients with joint inflammation. The data suggest that SF CR1 originates from the infiltrating leukocytes, which shed both a soluble and a membrane-associated form. Whether SF CR1 participates in the local regulation of complement activation remains to be examined.

Necrosis and apoptosis associated with distinct Ca2+ response patterns in target cells attacked by human natural killer cells

1. During the process of cell death, rises in cytosolic Ca2+ concentration ([Ca2+]i) together with structural changes were investigated in target cells attacked by purified CD3-,CD16+ human natural killer (NK) cells. 2. In the target cell line K562, a rapid [Ca2+]i rise to 1-2 microM occurred a few minutes after NK cell-target cell contact, immediately followed by leakage of the Ca2+ indicator dye fura-2 from the cell. Cells were permeabilized, but their chromatin was not fragmented. The changes were basically consistent with those seen in necrosis induced by activated complement. 3. In the target cell line MOLT-4, which expressed the apoptosis-inducing surface antigen Fas much more strongly than K562, the majority of attacked cells displayed a [Ca2+]i rise to 0.7-1 microM followed by a slow decline, often associated with diminishing [Ca2+]i oscillations. As a whole, [Ca2+]i remained higher than 150 nM for at least 1.5-3 h (approximately 100 nM in control cells). 4. MOLT-4 cells attacked by NK cells became bubble shaped within 20 min of the main [Ca2+]i rise reaching its peak, and then both the cell and chromatin were fragmented into small pieces. These findings were basically consistent with those in apoptosis induced by a monoclonal antibody against the surface antigen Fas. 5. NK cells induced both necrosis and apoptosis in cell lines insensitive to NK cells in the presence of an antibody against the major histocompatibility complex class I (antibody-dependent cell-mediated cytotoxicity, ADCC). The distinct Ca2+ responses patterns described above corresponded to necrosis or apoptosis in different cells stimulated by the common ADCC pathway. 6. Human NK cells were found to be capable of inducing necrosis (membrane damage) or apoptosis (nuclear damage) depending on the target cell types. The characteristic Ca2+ response profile was a good indicator for distinguishing between the modes of cell death induced by the cytotoxicity of the killer cells.

Distinct Ca2+ response patterns in human natural killer cells during induction of necrosis or apoptosis of target cells

Human CD3-, CD16+ natural killer cells (NKs) induced necrosis in the cell line K562 and the minor population of MOLT-4, while induced apoptosis in the majority of MOLT-4, in one-to-one cell interaction. Simultaneous Ca2+ imaging of NK and target cells revealed remarkable differences in changes of cytosolic Ca2+ concentration ([Ca2+]i) during the process of distinct target cell killing. A rapid [Ca2+]i rise in NK was generated within 2 min after contact with any target. For induction of necrosis, the [Ca2+]i rise in NK reached a peak of 580 +/- 195 nM (mean +/- SD, n = 41) and was followed by an abrupt large [Ca2+]i rise in target cells probably due to Ca2+ entry through pores formed by perforin which was released from NK. The interval between onsets of NK and target Ca2+ responses (Tnk-t) was shorter than 3 min in most cases. Target cells were soon permeabilized, and NK Ca2+ responses ceased in 14 +/- 7 min. For induction of apoptosis, NK Ca2+ responses were relatively smaller (405 +/- 125 nM, n = 16), and Tnk-t was widely scattered up to 30 min. NK Ca2+ responses lasted much longer in a form of Ca2+ oscillations (total duration, 66 +/- 33 min) in parallel with target Ca2+ responses, and ceased as apoptotic changes such as cellular and nuclear fragmentation advanced. The difference in Tnk-t suggests that rapid synchronous release of perforin tends to cause necrosis rather than apoptosis. The duration of NK Ca2+ responses suggests that some 'off signals' may be transmitted to NK following recognition of target cell damage.

Annexin 1 is overexpressed and specifically secreted during experimentally induced colitis in rats

Annexin 1 is a protein induced by glucocorticoids endowed with extracellular anti-inflammatory properties. In this study, the local expression and secretion of annexins 1-6, in rat proximal colon, were studied at different times after intracolonic administration of 30 mg trinitrobenzenesulfonic acid in 50% ethanol. Secretion was identified by incubating colonic tissues in a culture medium. The expression of annexins was detected by immunoblotting in tissue homogenates and incubation media. Inflammatory stages were evaluated by measuring myeloperoxidase activity. Annexin 1 expression in colons increased after trinitrobenzenesulfonic acid treatment and was maximal between days 1 to 9, during the cellular stage of the inflammation that corresponded to maximal myeloperoxidase activity. From 12 h to 9 days after trinitrobenzenesulfonic acid/ethanol treatment, annexin 1 was specifically secreted. Annexin 3 was also overexpressed during the cellular stage, but the expression of annexins 2, 4, 5, and 6 was unchanged; none of these annexins were secreted. Annexin 1 was shown to be physiologically secreted because its release was specific, abundant, and not correlated with cellular lysis. Annexin 1 may be considered as a putative candidate in the control of the gut inflammatory processes.

Cranial meninges of goldfish: age-related changes in morphology of meningeal cells and accumulation of surfactant-like multilamellar bodies

In the optic tectum of goldfish, the outer, middle and inner layers of the endomeninx were evident in animals ranging in age from 1 month to several years. The outer layer in young animals consisted of closely overlapping cells with intertwined processes, whereas in the older animals it contained large extracellular spaces. The intermediate layer cells were always arranged in a single continuous layer, but in young animals they overlapped extensively with one another toward their edges whereas in the oldest animals they became extremely flat and non-overlapping. The inner layer included an outer tier of cells with their bases adhering to the intermediate layer, and an inner tier of cells detached from both the intermediate layer and the basal lamina overlying the brain parenchyma. Inner layer cells contained many large vacuoles that were in continuity with the extracellular space. With age, the extracellular space and the vacuolar system expanded, and the inner layer evolved into a meshwork of attenuated cytoplasmic processes embedded in the granular extracellular matrix. Another age-related feature was the accumulation adjacent to the basal lamina of uniform disc-shaped membranous structures, resembling multilamellar bodies of lung surfactant. These "disc bodies" were apparently generated by the coalescence of vesicles formed at the surface of the inner layer cells, possibly as a by-product of protein secretion by these cells.

Efflux of cellular cholesterol and phospholipid to lipid-free apolipoproteins and class A amphipathic peptides

The mechanism(s) by which lipid-free apolipoprotein (apo) AI is able to stimulate efflux of cholesterol and phospholipid from cells in cultures has (have) been examined. This process was found to be enhanced when macrophages were enriched with cholesterol. There were 12- and 4-fold increases in cholesterol and phospholipid efflux, respectively, from cholesterol-enriched mouse macrophages when compared to cells not loaded with cholesterol. This enhancement in cholesterol efflux to lipid-free apo AI from macrophages enriched with cholesterol was found to be controlled by the level of free cholesterol in the cells. When cholesterol-enriched mouse macrophages were exposed to lipid-free apo AI at 20 micrograms/mL (706 nM), there was significant efflux of [14C]cholesterol and [3H]phospholipid (20% +/- 0.5%/24 h and 6% +/- 0.3%/24 h, respectively). In comparison, HDL at equivalent protein concentrations only stimulated 11% and 4% efflux of cholesterol and phospholipid, respectively. Synthetic peptides containing amphipathic helical segments that mimic those present in apo AI were used to examine the structural features of the apoprotein which stimulate lipid efflux. Peptides containing only one (18A) or two (37pA) amphipathic helical segments stimulated as much cholesterol efflux from both mouse macrophages and L-cells as apo AI. The order of efficiency, as assessed by the mass concentration at which half-maximal efflux was reached (EC50), was apo AI > 37pA > 18A, indicating that acceptor efficiency was dependent on the number of amphipathic helical segments per molecule. When the helical content of 18A was increased by neutralizing the charges at the ends of the peptide (Ac-18A-NH2), there was a substantial increase in the efficiency for cholesterol efflux (EC50 18A = 17 micrograms/mL vs Ac-18A-NH2 = 6 micrograms/mL). In contrast, when the amphipathicity of the helix in 18A was decreased by scrambling the amino acid sequence, thereby reducing its lipid affinity, cholesterol and phospholipid efflux were not stimulated. The efficiency with which the peptides stimulated cholesterol efflux was in order of their lipid affinity (37pA > Ac-18A-NH2 > 18A), and this order was similar for phospholipid efflux. The time course of lipid release from mouse macrophages and L-cells indicated that phospholipid appeared in the extracellular medium before cholesterol. These results suggest that the apo AI or peptides first interacted with the cell to form protein/phospholipid complexes, that could then accept cholesterol.

X-linked vacuolated myopathy: complement membrane attack complex on surface membrane of injured muscle fibers

We describe a probable recessive X-linked myopathy characterized by the presence of vacuolated muscle fibers. Four males and their shared maternal grandfather were affected. Clinical characteristics include juvenile onset, very slow progression, and predominant proximal muscle involvement. The clinical picture and the morphological findings are compared with those previously described in a family. By immunofluorescence, all histologically abnormal muscle fibers, in particular those vacuolated, showed a strong deposition of the complement C5b-9 membrane attack complex over the whole muscle fiber surface. Weak immunostaining for membrane attack complex was also found in endomysial capillaries and perimysial vessel walls. Muscle fibers showed sarcolemmal immunolabeling with anti-major histocompatibility complex I, which was also present on the margins of many vacuoles. All vacuoles were stained by antidystrophin antibody, which colocalized in most of them with antilaminin immunostaining. Taken together, these results suggest that the deposition of membrane attack complex on the damaged cell surface membrane could be important in the pathogenesis of this muscle disorder, and that the membrane-bounded vacuoles could be a consequence of sarcolemmal invagination.

Complement component C9-dependent cytosolic free Ca2+ rise and recovery in neutrophils

The cytosolic free Ca2+ concentration in rat neutrophils was determined by ratiometric fluorometry and imaging of Fura-2. Transient elevations of cytosolic free Ca2+ concentration were provoked by addition of C9 to neutrophils pre-coated with C5b-8. The rate of rise of the cytosolic free Ca2+ concentration was dependent upon the concentration of C9. These changes in cytosolic free Ca2+ concentration occurred in the absence of cell lysis, since there was no release of Fura-2, and were the result of increased permeability to extracellular Ca2+. More than 96% of the rise in cytosolic free Ca2+ generation by C9 was dependent upon the presence of extracellular Ca2+, but did not occur via channels which were inhibited by the Ca2+ channel blocker SKF96365. The decrease in the permeability of the membrane to Ca2+ after C9 was not triggered by the rise in cytosolic free Ca2+. After attack by C9, individual neutrophils remained responsive to f-met-leu-phe, or further attack by C9, both producing Ca2+ transients. The recovery of the Ca2+ signal was consistent with the complement membrane attack complex generating a series of permeability thresholds in the plasma membrane. These data have implications for the understanding of the mechanisms underlying the inappropriate responsiveness of neutrophils at inflammatory sites.

Secretory phospholipase A2 generates the novel lipid mediator lysophosphatidic acid in membrane microvesicles shed from activated cells

Nonpancreatic secretory phospholipase A2 (sPLA2) displays proinflammatory properties; however, its physiological substrate is not identified. Although inactive toward intact cells, sPLA2 hydrolyzed phospholipids in membrane microvesicles shed from Ca(2+)-loaded erythrocytes as well as from platelets and from whole blood cells challenged with inflammatory stimuli. sPLA2 was stimulated upon degradation of sphingomyelin (SPH) and produced lysophosphatidic acid (LPA), which induced platelet aggregation. Finally, lysophospholipid-containing vesicles and sPLA2 were detected in inflammatory fluids in relative proportions identical to those used in vitro. We conclude that upon loss of phospholipid asymmetry, cell-derived microvesicles provide a preferential substrate for sPLA2. SPH hydrolysis, which is provoked by various cytokines, regulates sPLA2 activity, and the novel lipid mediator LPA can be generated by this pathway.

Altered glycosylation and selected mutation in recombinant human complement component C9: effects on haemolytic activity

Recombinant wild-type and mutated forms of human complement component C9 have been synthesized in baculovirus-infected insect cells. Wild-type recombinant C9 was indistinguishable from native C9, as judged by haemolytic activity, trypsin and alpha-thrombin digestion, reaction with antibodies to C9, enzymatic deglycosylation to the same core size and polymerization in the presence of Zn2+. Replacement of the native signal peptide with the honey-bee melittin signal peptide, and replacement of Spodoptera frugiperda (Sf9) cells with Trichoplusia ni cells produced yields of 5 micrograms C9/ml supernatant. Three C9 mutants were generated; one mutant, with four acidic residues changed to alanines in a putative calcium-binding site, had the same biological activity as recombinant C9. Another mutant, lacking 23 N-terminal amino acids, previously showing increased polymerization when produced in vitro, polymerized on secretion, rendering it inactive. It was not possible to demonstrate haemolytic activity of the third mutant, cysteines 33 and 36 mutated to alanine, as it was secreted a hundredfold less than the wild-type protein.

Construction, expression and functional analysis of a glycolipid-linked form of CR1

By genetic engineering of human CR1 cDNA and its stable transfection into cells we have produced a cell line which expresses CR1 anchored to the cell surface by a glycolipid anchor. The glycosyl-phosphatidylinositol (GPI)-CR1 protects cells intrinsically from damage mediated by complement activated through the classical pathway. Cell surface GPI-CR1 is more efficient on a molar basis than soluble CR1 in the assay, but extrinsic protection of other cells was not obtained. Soluble CR1-protected cells extrinsically in the assay but was required at nearly ten fold higher amounts than the intrinsic protection conferred by GPI-anchored CR1. Additionally, GPI-CR1 was shown to act as a co-factor to Factor I in the generation of C3c from iC3b. Since GPI-anchored proteins can incorporate spontaneously into the membranes of living cells by virtue of their lipid tails, the isolated GPI-CR1 will be used to introduce CR1 on to the surfaces of many different types of cell so that its role in immunity can be further investigated.

Stress-relaxation of fibroblasts in collagen matrices triggers ectocytosis of plasma membrane vesicles containing actin, annexins II and VI, and beta 1 integrin receptors

To learn about the effects of tension on fibroblast function, we have been studying initial cellular responses to stress-relaxation. Human foreskin fibroblasts were cultured in anchored collagen matrices for 2 days, during which time mechanical stress developed. Subsequently, the matrices were dislodged; thereby allowing stress to dissipate. Within 5 min after initiating stress-relaxation, fibroblasts retracted their pseudopodia. At this time, we observed the disappearance of cellular stress fibers and the formation of actin clusters along the cell margins. The actin was found to be located inside 200 nm diameter vesicles that were budding from the cell surface. Vesicles isolated from the matrix after stress-relaxation contained prominent 24 kDa, 36 kDa (doublet), 45 kDa, and 135 kDa polypeptides. The 45 kDa polypeptide was the major component in the Triton-insoluble vesicle fraction and appeared to be actin. The 36 kDa (doublet) polypeptide, which was found predominantly in the Triton-soluble vesicle fraction, was identified as annexin II. Vesicles also contained annexin VI and beta 1 integrin receptors but not tubulin, vimentin, vinculin or annexin I. The results suggest that stress-relaxation of fibroblasts induces a novel ectocytotic process involving transient budding of intact, plasma membrane vesicles from the cell cortex. On the basis of their morphological and biochemical features, these vesicles may be analogous to the ‘matrix vesicles’ released by chondrocytes and could play a role in extracellular matrix remodeling after wound contraction.

Killing of cells by perforin. Resistance to killing is not due to diminished binding of perforin to the cell membrane

Different cell types vary widely in their susceptibility to killing by the pore-forming cytolytic molecule perforin. In particular, the cells responsible for synthesis of perforin, i.e. cytotoxic T lymphocytes (CTL) and natural killer (NK) cells, are very resistant to cytolysis by this molecule. It has previously been suggested that resistance is due, at least in part, to diminished binding of perforin to these cells. The purpose of the present study was to compare binding of perforin to sensitive and resistant cell types. To this end, perforin was biosynthetically labelled prior to purification. The purified labelled protein was then utilized to obtain a direct measure of the amount of perforin bound to cells during attack. Resistant cells (CTL, neutrophils) bound at least as much perforin as did sensitive cells (K562, HL60 etc.), indicating that resistance to perforin involves mechanisms operating after binding of the lytic molecule.

An acetylation method for the quantification of membrane lipids, including phospholipids, polyphosphoinositides and cholesterol

A method for the quantification of membrane lipids by acetylation with [3H]acetic anhydride is described. A standard lipid, labelled with 14C or 32P, is added to the sample and is simultaneously acetylated. The 3H/14C or 3H/32P ratios obtained in the acetylated lipids are proportional to the initial lipid concentration. The method has been used to quantify cholesterol, phosphatidylethanolamine, phosphatidylinositol, sphingomyelin, phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate in amounts between 2 and 10 nmol.

Incidence of hypothyroidism after 131I therapy for hyperthyroidism. Relation to pretherapy serum levels of T3, T4 and thyroid antibodies

A correlation is reported between serum levels of triiodothyronine (S-T3) and thyroxine (S-T4) before treatment, as well as levels of thyroid antibodies before treatment, and the development of hypothyroidism following 131I therapy in 86 patients with hyperthyroidism. Patients with marked elevation of S-T3 or S-T4 had demonstrable antibodies to thyroid cytoplasmic antigen more often than those with normal or moderately elevated levels, and patients with markedly elevated levels of S-T3 also had a higher incidence of hypothyroidism after treatment. Patients with nodular thyroid glands and with markedly elevated levels of S-T3 required a larger number of 131I doses before no signs of hyperthyroidism persisted in comparison to those with moderately elevated levels.