Synthesis of aberrant decay-accelerating factor proteins by affected paroxysmal nocturnal hemoglobinuria leukocytes

Altered lipid raft composition and defective cell death signal transduction in glycosylphosphatidylinositol anchor-deficient PIG-A mutant cells

Paroxysmal nocturnal haemoglobinuria (PNH) is a clonal disorder of haematopoietic stem cells caused by somatic PIGA mutations, resulting in a deficiency in glycosylphosphatidylinositol-anchored proteins (GPI-AP). Because GPI-AP associate with lipid rafts (LR), lack of GPI-AP on PNH cells may result in alterations in LR-dependent signalling. Conversely, PNH cells are a suitable model for investigating LR biology. LR from paired, wild-type GPI(+), and mutant GPI(-) cell lines (K562 and TF1) were isolated and analysed; GPI(-) LR contained important anti-apoptotic proteins, not found in LR from GPI(+) cells. When methyl-beta-cyclodextrin (MbetaCD) was utilized to probe for functional differences between normal and GPI(-) LR, increased levels of phospho-p38 mitogen-activated protein kinase (MAPK), and phospho-p65 nuclear factor NF-kappaB were found in control and GPI(-) cells respectively. Subsequent experiments addressing the inhibition of phosphoinositide-3-kinase (PI3K) suggest that the PI3K/AKT pathway may be responsible for the resistance of K562 GPI(-)cells to negative effects of MbetaCD. In addition, transduction of tumour necrosis factor-alpha (TNF-alpha) signals in a LR-dependent fashion increased induction of p38 MAPK in GPI(+) and increased pro-survival NF-kappaB levels in K562 GPI(-) cells. Therefore, we suggest that the altered LR-dependent signalling in PNH-like cells may induce different responses to pro-inflammatory cytokines from those observed in cells with intact GPI-AP.

Transfer of glycosylphosphatidylinositol-anchored proteins to deficient cells after erythrocyte transfusion in paroxysmal nocturnal hemoglobinuria

In paroxysmal nocturnal hemoglobinuria (PNH), an acquired mutation of the PIGA gene results in the absence of glycosylphosphatidylinositol (GPI)-anchored cell surface membrane proteins in affected hematopoietic cells. Absence of GPI-anchored proteins on erythrocytes is responsible for their increased sensitivity to complement-mediated lysis, resulting in hemolytic anemia. Cell-to-cell transfer of CD55 and CD59, 2 GPI-anchored proteins, by red cell microvesicles has been demonstrated in vitro, with retention of their function. Because red cell units stored for transfusion contain many erythrocyte microvesicles, transfused blood could potentially serve as a source of CD55 and CD59. We examined whether GPI-anchored proteins could be transferred in vivo to deficient cells following transfusions given to 6 patients with PNH. All patients were group A(1) blood type. Each was given transfusions of 3 U of compatible, washed group O blood. Patient group A(1) cells were distinguished from the transfused group O cells by flow cytometry and staining with a labeled lectin, Dolichos biflorus, which specifically binds to group A(1) erythrocytes. Increased surface CD59 was measured on recipient red cells and granulocytes 1, 3, and 7 days following transfusion in all 6 patients. Our data suggest a potential therapeutic role for GPI-anchored protein transfer for severe PNH.

Mechanism of intravascular hemolysis in paroxysmal nocturnal hemoglobinuria (PNH)

Paroxysmal nocturnal hemoglobinuria (PNH) hemolysis requires both intravascular complement activation and affected erythrocytes susceptible to complement. This susceptibility is explained by a deficiency in complement regulatory membrane proteins that are attached to the membrane by a glycosylphosphatidylinositol (GPI) anchor. Affected cells lack a series of GPI-anchored membrane proteins with various functions. The lack is caused by a synthetic defect of the anchor due to an impaired transfer of N-acetylglucosamine to phosphatidylinositol which is an early metabolic precursor in the anchor synthesis. Moreover, PIG-A gene responsible for the membrane defect was recently cloned. Further, a possible mechanism of complement activation has been proposed, especially for an infection-induced hemolytic precipitation which is clinically crucial. Thus, the molecular events, leading to intravascular hemolysis characteristic of PNH, has been virtually clarified. Next major concern is the nature of PIG-A: How does PIG-A explain the complex pathophysiology of PNH which exhibits various clinical manifestations?

Mammalian glycosylphosphatidylinositol-anchored proteins and intracellular precursors

Glycosylphosphatidylinositol-anchored proteins can be specifically identified by several methods. PI-PLC digestion analyses, the most widely used technique, can be performed more reliably when conducted with purified protein and phase partitioning to exclude steric effects and when combined with alkaline hydrolysis to control for inositol acylation. Reductive radiomethylation not only can definitively identify a candidate protein as being GPI anchored, but also can provide information on the number of amine components (GlcN, ethanolamine) in the anchor structure. Biosynthetic labeling with anchor precursors is relatively specific when performed with [3H]ethanolamine or [3H]inositol. Incorporation of the precursors additionally can be used to (1) document anchor transfer to primary translation products, (2) identify soluble derivatives of GPI-anchored proteins that have been released from cell surfaces, and (3) localize the site of GPI anchor attachment within a GPI-anchored protein. A pathway for mammalian GP anchor assembly is depicted in Fig. 12. Initially GlcNAc is transferred to PI. The resulting GlcNAc-PI is then deacetylated to yield GlcN-PI. After that step, several points of divergence are identifiable between the mammalian and T. brucei pathways: (1) all mammalian Man-containing intermediates are built on acylated inositol phospholipids; (2) a proximal phosphoethanolamine is found in mammalian GPI anchor intermediates and is added to Man 1 prior to incorporation of Man 2 and Man 3; (3) no Gal branching substituent is added to the mammalian core glycan; and (4) the most polar mammalian GPI contains a third phosphoethanolamine substituent linked to the 6 position of Man 2.

Activation of human monocytes and granulocytes by monoclonal antibodies to glycosylphosphatidylinositol-anchored antigens

The present study investigated possible receptor-like characteristics of glycosylphosphatidylinositol (GPI)-linked antigens on human monocytes and granulocytes by measuring cytoplasmic calcium fluxes and the oxidative burst in cells following cross-linking of GPI-linked antigens. Cross-linking of cell-bound anti-CD14, -CDw52 and -CD55 induced cytoplasmic calcium fluxes and oxidative bursts in unprimed human monocytes similar to those observed following Fc gamma R cross-linking. In granulocytes primed with 200 mM N-formyl-Met-Leu-Phe (FMLP), cross-linking of cell-bound anti-CD16, -CD24, -CD59 and -CD67 led to calcium fluxes and activation of the oxidative burst. The oxidative bursts mediated by GPI-linked antigens were stronger than those induced by 200 nM FMLP, even though FMLP induced a larger increase in cytoplasmic calcium concentration. The responses were likely to be independent of Fc gamma R interactions as F(ab')2 fragments of IgG or IgM antibodies were used in the experiments. Activating effects of monoclonal antibody to GPI-linked antigens were not observed in cells from patients with paroxysmal nocturnal hemoglobinuria, which are deficient in GPI-linked antigens. In addition, treatment with GPI-specific phospholipase C led to inhibition of cell activation through GPI-linked antigens but not through transmembrane receptors. Cross-linking of a number of non-GPI-linked antigens (CD11a, CD18, CD31, CD35, CD43, and CD45) neither induced calcium fluxes, nor activated the oxidative burst. The results indicate that most, if not all, GPI-linked surface glycoproteins on myeloid cells are capable of mediating cell activation and suggest that the GPI anchor is a structure facilitating signal transduction.

Membrane defence against complement lysis: the structure and biological properties of CD59

The complement system is an important branch of the innate immune response, constituting a first line of defence against invading microorganisms which activate complement via both antibody-dependent and -independent mechanisms. Activation of complement leads to (a) a direct attack upon the activating cell surface by assembly of the pore-forming membrane attack complex (MAC), and (b) the generation of inflammatory mediators which target and recruit other branches of the immune system. However, uncontrolled complement activation can lead to widespread tissue damage in the host, since certain of the activation products, notably the fragment C3b and the C5b-7 complex, can bind nonspecifically to any nearby cell membranes. Therefore it is important that complement activation is tightly regulated. Our own cells express a number of membrane-bound control proteins which limit complement activation at the cell surface and prevent accidental complement-mediated damage. These include decay-accelerating factor, complement receptor 1 and membrane cofactor protein, all of which are active at the level of C3/C5 convertase formation. Until recently, cell surface control of MAC assembly had been attributed to a single 65-kD membrane protein called homologous restriction factor (alternatively named C8-binding protein and MAC-inhibiting protein). However a second MAC-inhibiting protein has since been discovered and it is now clear that this protein plays a major role in the control of membrane attack. This review charts the rapid progress made in elucidating the protein and gene structure, and the mechanism of action of this most recently discovered complement inhibitor, CD59.

Induction of Fc gamma R-III (CD16) expression on neutrophils affected by paroxysmal nocturnal haemoglobinuria by administration of granulocyte colony-stimulating factor

The inducibility of glycosyl-phosphatidylinositol (GPI)-anchored proteins on affected paroxysmal nocturnal haemoglobinuria (PNH) neutrophils (PMN) after both in vitro and in vivo stimulation was investigated. Fc gamma R-III (CD16), decay-accelerating factor (DAF/CD55) and 20 kD homologous restriction factor (HRF20/CD59) were demonstrated to be concurrently deficient on unstimulated defective PNH PMN. Upon in vitro stimulation with either N-formyl-methionyl-leucyl-phenylalanine (fMLP), zymosan-activated serum (ZAS), or recombinant human granulocyte colony-stimulation factor (G-CSF), neither CD16 nor CD55 expression was induced on defective PNH PMN. G-CSF was administered to two patients with PNH when their conditions were complicated by bacterial infections, or to prevent infections associated with the extraction of teeth or cataract surgery. CD16 expression was induced on the defective PNH PMN in both cases during the administration of G-CSF, but the expression of CD55 and CD59 was not. CD16, induced on the defective PNH PMN during the administration of G-CSF, was phosphatidylinositol-specific phospholipase C (PIPLC)-sensitive, implying that it had GPI-linkage to the membranes. The patients treated with G-CSF recovered from infection or evaded infection. These observations suggest that a deficiency of GPI-anchored proteins is not always seen in defective PNH blood cells, at least under certain stimulation conditions.

Defective and normal haematopoietic stem cells in paroxysmal nocturnal haemoglobinuria

The expression of decay-accelerating factor (DAF or CD55) and CD59 during haematopoietic cell development in bone marrow aspirates of two patients with paroxysmal nocturnal haemoglobinuria (PNH) was compared with that in normal bone marrow by five-dimensional flow cytometry. In contrast to early uncommitted haematopoietic progenitor cells (CD34+, CD38-) in normal bone marrow which uniformly express DAF and CD59, the majority of CD34+, CD38- cells in both patients' marrow exhibited the absence of the two proteins. In both specimens, however, subpopulations of CD34+, CD38- cells expressing DAF and CD59 were detectable, indicative of the presence of two lines of haematopoiesis, one abnormal and the other normal. Concurrent abnormal and normal haematopoietic development was further evident by the presence of subpopulations of DAF-, CD59- and DAF+, CD59+ cells along the differentiation and maturation pathways of the myeloid (CD33+, CD15(-)-->CD33+-->++, CD15+), the erythroid (CD45dim, CD71dim-->CD45-, CD71++), and the B-lymphoid cell lineages (CD10++, CD20(-)-->CD10-, CD20++). While the majority of cells differentiating into and maturing along each cell lineage lacked DAF and CD59, the majority of mature B (CD20++, CD10-) and T-lymphocytes lymphocytes (CD3+) expressed both proteins suggestive of the presence of lymphocytes with a long life span which were generated from normal haematopoietic progenitors before the onset of the disease. The detection of distinct sets of CD34+, CD38(-)--> + progenitor cells which are DAF+, CD59+ or DAF-, CD59- in marrow of PNH patients has relevance for the treatment of PNH. Cells with the phenotype CD34+, CD38-, DAF+, CD59+ are capable of self renewal and represent potential candidates for autologous bone marrow transplantation following depletion of CD34+, CD38-, DAF-, CD59- cells.

A soluble form of the glycolipid-anchored receptor for urokinase-type plasminogen activator is secreted from peripheral blood leukocytes from patients with paroxysmal nocturnal hemoglobinuria

The cellular urokinase-type plasminogen-activator (uPA) receptor (uPAR) is a glycolipid-anchored membrane protein thought to be involved in pericellular proteolysis during cell migration and tumor invasion. In the present study, we have identified and characterized two soluble forms of uPAR which have retained their ligand-binding capability. One variant was generated in vitro by treatment of intact normal cells with either a phosphatidylinositol-specific phospholipase C (PLC) or endoproteinase Asp-N. The other soluble uPAR variant was secreted in vivo from peripheral blood leukocytes affected by the stem-cell disorder paroxysmal nocturnal hemoglobinuria (PNH), and was found in the plasma from these PNH patients as well as in the conditioned medium from cultured PNH leukocytes. Under normal conditions, we find no evidence for any shedding or secretion of a soluble uPA-binding counterpart to human uPAR in plasma. Unlike normal leukocytes, the PNH-affected cells do not express uPAR on the cell surface, although they do contain apparently normal levels of uPAR-specific mRNA. The secreted uPAR derived from PNH cells has a mobility in SDS/PAGE that is slightly higher than that of uPAR solubilized by PtdIns-specific PLC or detergent, but resembles that of a truncated, recombinant uPAR variant, which has its C-terminus close to the proposed glycolipid-attachment site, suggesting that the secreted protein has been proteolytically processed for glycolipid attachment. The presence in plasma from PNH patients of such a secreted, hydrophilic form of uPAR lends support to the hypothesis that the lesion underlying the PNH disorder resides either in glycolipid biosynthesis or in the function of an as-yet-unidentified transamidating enzyme assumed to cleave and assemble the truncated uPAR with the preformed glycolipid moiety.

Synthesis of mannosylglucosaminylinositol phospholipids in normal but not paroxysmal nocturnal hemoglobinuria cells

To identify mannosyl (Man)-containing intermediates of the human glycoinositol phospholipid (GPI) anchor pathway and examine their expression in paroxysmal nocturnal hemoglobinuria (PNH), mannolipid products deriving from in vitro guanosine diphosphate [3H]Man labeling of HeLa cell microsomes were characterized. The defined GPI species were correlated with products deriving from in vivo [3H]Man labeling of normal and (GPI-anchor defective) affected leukocytes. In vitro analyses in HeLa cells showed dolichol-phosphoryl (Dol-P)-[3H]Man and a spectrum of [3H]Man lipids exhibiting TLC mobilities approximating those of Trypanosoma brucei (Tryp) GPI precursors. Iatrobead HPLC separations and partial characterizations of the major isolated [3H]Man species (designated H1-H8) showed that all but H1 (Dol-P-Man) were sensitive to HNO2 deamination and serum GPI-specific phospholipase D digestion but were resistant to phosphatidylinositol-specific phospholipase C digestion unless previously deacylated with mild alkali. [3H]Man label in H3, H4, and H6 but not in H5 or H7 was efficiently released into the aqueous phase by jack bean alpha-mannosidase digestion. BioGel P-4 and AX-5 sizing of the dephosphorylated core glycan fragments of H6 and H7 gave values that coincided precisely with the corresponding glycan fragments from the fully assembled Tryp anchor donor A' (P2). Affected leukocytes from four patients with PNH supported formation of GlcNAc- and GlcN-PI but all failed to express H6 and H7 as well as H8 and two showed complete absence of earlier Man-containing intermediates. These findings argue that human intracellular GPI mannolipids are built on acylated inositol phospholipids, that H6 and H7 contain differentially phosphoethanolamine-substituted Man3-GlcN-inositol cores, and that PNH cells are defective in conversion of GlcN-PI into these more mature mannolipid structures.

Structural properties of the glycoplasmanylinositol anchor phospholipid of the complement membrane attack complex inhibitor CD59

CD59, the membrane regulator of autologous C5b-9 channel formation, exhibits variable sensitivity to cleavage by phosphatidylinositol-specific phospholipase C (PI-PLC), an enzyme that releases glyco-inositolphospholipid (GPI)-anchored proteins from cell surfaces. To determine whether the GPI-anchor phospholipid of CD59 is similar to that of decay-accelerating factor (DAF) and whether variation in its structure underlies its variable enzyme susceptibility, the GPI anchors of the two proteins expressed on erythrocytes, polymorphonuclear and mononuclear leucocytes were compared in situ and after purification. Flow cytometric analyses of PI-PLC-treated cells showed parallel cell type specific release of both proteins as a function of enzyme concentration. Non-denaturing PAGE analyses of alkaline/hydroxylamine-treated proteins (affinity-purified from [125I]-surface-labelled cells) provided evidence for (i) comparable proportions of GPI-anchor acylation, and (ii) alkali-resistant rather than alkali-sensitive lipid substituents in erythrocytes. These findings argue that the differential C5b-9 sensitivity that distinguishes paroxysmal nocturnal haemoglobinuria II and III erythrocytes does not derive from expression of CD59 molecules with alternative GPI-anchor phospholipid structures.

Phosphatidylinositol-glycan linked proteins of the erythrocyte membrane

The human erythrocyte bears a number of proteins anchored to the outer membrane surface via a phosphatidylinositol-glycan linkage. This class of proteins includes several complement regulatory proteins (including decay-accelerating factor, CD59 antigen (protectin), and C8 binding protein) as well as several enzymes and at least one protein important in cell-cell interaction. In addition, a number of blood group antigens have been identified to reside on proteins with phosphatidylinositol anchors. One blood group (Cromer) resides on DAF. Study of variants in this blood group system has led to interesting information about the function and expression of this protein. Several other blood groups, such as JMH and Holley/Gregory, appear to reside on as yet unidentified phosphatidylinositol-linked proteins. In paroxysmal nocturnal haemoglobinuria, a variable proportion of red cells fail to express or express weakly all phosphatidylinositol-linked proteins. The origin of this deficiency is now being worked out. In addition, individuals with inherited deficiency of DAF or CD59 (protectin) have been identified. Only the latter deficiency leads to a PNH-like syndrome.

The erythrocytes in paroxysmal nocturnal haemoglobinuria of intermediate sensitivity to complement lysis

The sensitivity to lysis by complement of the erythrocytes of 56 patients with paroxysmal nocturnal haemoglobinuria (PNH) was compared to the membrane expression of decay accelerating factor (DAF, CD55), membrane inhibitor of reactive lysis (MIRL, CD59) and acetylcholinesterase (AChE). Most patients (36/50 72% in whom the analysis could be made) appeared to have erythrocytes of intermediate sensitivity to complement in the blood. These cells appeared as a discrete population of cells (PNH II cells), as a 'tail' of cells slightly less sensitive than the predominant PNH III cells (previously called PNH IIIb cells), or as a continuous spectrum of cells sensitive to complement. The PNH III cells totally lacked all three proteins (DAF, MIRL, AChE) by flow cytometric analysis whereas PNH I cells appeared to have normal or nearly normal amounts of each. The cells of intermediate sensitivity (PNH II) had coordinately decreased expression of all three proteins; the level of expression of DAF and MIRL paralleled the sensitivity of the cells to the haemolytic action of complement.

Assembly and deacetylation of N-acetylglucosaminyl-plasmanylinositol in normal and affected paroxysmal nocturnal hemoglobinuria cells

Decay-accelerating factor (DAF) is anchored in cell membranes by a glycosyl-plasmanylinositol (GPI) moiety that is transferred to it en bloc in the rough endoplasmic reticulum. To analyze the biochemical reactions involved in preassembly of this structure, a human hematopoietic cell-free system was employed. Incubation of cell extracts with UDP-[3H]GlcNAc and butanol partitioning of reaction mixtures yielded two products similar in TLC mobility to intermediates described in Trypanosoma brucei. Both species were sensitive to Bacillus thuringiensis phosphatidylinositol-specific phospholipase C, indicative of association of [3H]GlcNAc label with a plasmanylinositol-containing acceptor. In contrast to trypanosome intermediates, which contain phosphatidylinositol (1,2-diacylglycerophosphoinositol), however, alkali treatment and phospholipase A2 digestion generated butanol-phase products characteristic of glycosylated plasmanylinositol (1-alkyl-2-acylglycerophosphoinositol). Kinetic and pulse-chase experiments indicated that the slower-migrating species was a product of the faster and that it, but not the faster, was sensitive to both GPI-specific phospholipase D and nitrous acid deamination, consistent with conversion of GlcNAc- to GlcN-plasmanylinositol. Accordingly, acetic anhydride acetylation retransformed the slower species back to the faster. Further incubation with cell extracts converted the slower species into more polar products. Lysates of normal and of affected blood leukocytes from two paroxysmal nocturnal hemoglobinuria (PNH) patients supported assembly of the two intermediates within 1 min. Thus, the initial enzymes mediating human GPI-anchor assembly are GlcNAc-plasmanylinositol transferase and GlcNAc-plasmanylinositol deacetylase, their substrates contain plasmanylinositols, and the products of their activities are normal in affected PNH cells.

Preferential expression of human Fc gamma RIIIPMN (CD16) in paroxysmal nocturnal hemoglobinuria. Discordant expression of glycosyl phosphatidylinositol-linked proteins

The isoform of Fc gamma RIII (CD16) expressed on PMN has a GPI membrane anchor, and in paroxysmal nocturnal hemoglobinuria (PNH) there is a deficiency in Fc gamma RIII expression on PMN. Contrary to expectation, however, CD16 expression is preserved (albeit at reduced levels) in all affected PNH PMN that completely lack the GPI-anchored proteins DAF (CD55) and CD59. Fc gamma RIII negative PMN are not observed in any of the six PNH patients examined in this study. Analysis of the molecular weight of both glycosylated and deglycosylated Fc gamma RIII from PMN with reduced Fc gamma RIII expression indicates no variations in size relative to normal donor Fc gamma RIIIPMN. Indeed, the Fc gamma RIII expressed at intermediate levels is phosphatidylinositol-specific phospholipase C (PI-PLC)-sensitive. Thus, there is no evidence suggestive of expression of a transmembrane isoform and all data indicate that Fc gamma RIIIPMN on affected cells in PNH is a GPI-linked isoform. With Fc gamma RIIIPMN expression preserved at reduced levels on affected cells in PNH, PMN from PNH patients retain the capacity to internalize the Fc gamma RIIIPMN-specific probe E-ConA (at reduced levels) as well as IgG-opsonized erythrocytes. Reduced expression of GPI-anchored molecules on PNH PMN is not restricted to Fc gamma RIIIPMN since intermediate levels of CD59 were observed in the PNH PMN that were decay-accelerating factor (DAF)-negative and Fc gamma RIIIPMN intermediate. In addition, discordant expression of GPI-linked molecules in individual cells is not restricted to PMN since DAF+/CD14- monocytes were observed in one PNH patient. These data suggest that, when analyzed on an individual cell level, the GPI anchor defect in PNH is not absolute and must involve either a hierarchy of access of different protein molecules to available GPI anchors, distinct anchor biochemistries for the different proteins, or differential regulation of protein-anchor assembly.

Decay-accelerating factor in the cardiomyocytes of normal individuals and patients with myocardial infarction

The presence of decay-accelerating factor (DAF) was clearly demonstrated on the surface of normal cardiomyocytes. In patients who had died of myocardial infarction (MI) cardiomyocytes displayed different appearances: outside the ischaemically damaged region the myocytes showed no significant variations in DAF expression when compared with controls without MI. Within myocardial zones damaged by ischaemia, however, apparently normal myocytes showed large gaps in surface staining of DAF or formed clusters which were entirely devoid of reactivity with anti-DAF antibodies. The number of DAF-deficient myocytes increased with the extent of necrosis and also with the number of days between onset of MI and death. Even though injury to myocytes is to a large extent related to anoxia and to the presence of free oxygen radicals, the complement system also appears to be involved; DAF may have protective functions against complement-mediated injury. We speculate that phospholipase may be involved in the removal of DAF from the cardiomyocyte surface.