Mammalian glycophosphatidylinositol anchor transfer to proteins and posttransfer deacylation.

The glycophosphatidylinositol (GPI) anchors of proteins expressed on human erythrocytes and nucleated cells differ with respect to acylation of an inositol hydroxyl group, a structural feature that modulates their cleavability by PI-specific phospholipase C (PI-PLC). To determine how this GPI anchor modification is regulated, the precursor and protein-associated GPIs in two K562 cell transfectants (ATCC and .48) exhibiting alternatively PI-PLC-sensitive and resistant surface proteins were analyzed and the temporal relationship between GPI protein transfer and acquisition of PI-PLC sensitivity was determined. Nondenaturing PAGE analyses demonstrated that, whereas in .48 transfectants the GPI anchors in decay accelerating factor (DAF) and placental alkaline phosphatase (PLAP) were >95% acylated, in ATCC transfectants, they were 60 and 33% unsubstituted, respectively. In contrast, TLC analyses revealed that putative GPI donors in the two lines were identical and were >/=95% acylated. Studies of de novo DAF biosynthesis in HeLa cells bearing proteins with >90% unacylated anchors showed that within 5 min at 37 degreesC (or at 18 degreesC, which does not permit endoplasmic reticilum exit), >50% of the anchor in nascent 44-kDa proDAF protein exhibited PI-PLC sensitivity. In vitro analyses of the microsomal processing of miniPLAP, a truncated PLAP reporter protein, demonstrated that the anchor donor initially transferred to prominiPLAP was acylated and then progressively was deacylated. These findings indicate that (i) the anchor moiety that initially transfers to nascent proteins is acylated, (ii) inositol acylation in mature surface proteins is regulated via posttransfer deacylation, which in general is cell-specific but also can be protein-dependent, and (iii) deacylation occurs in the endoplasmic reticulum immediately after GPI transfer.

Stepwise transformation of astrocytes by simian virus 40 large T antigen and epidermal growth factor receptor overexpression.

We have investigated the transformed phenotype of neonatal mouse cortical astrocytes immortalized by retrovirus-mediated transfer of the SV40 large T antigen gene. Expression of T antigen was driven by the Moloney murine leukemia virus long terminal repeat. Cell lines were selected based on coexpression of neomycin resistance, which provides a selection method believed to be unbiased for transformation state. Astrocyte cell lines derived in this manner express T antigen over a relatively narrow range (approximately 4-fold), are contact inhibited, are able to enter a quiescent state in the presence of growth factors, and do not readily form colonies in soft agar. Compared to mortal astrocytes, the population growth rate is increased 3-fold, saturation densities are 4-fold higher, and the genome is relatively unstable as measured by the presence of DNA-aneuploid stem lines and by changes in DNA ploidy over time. However, changes in transformation phenotype occur at a low rate, making the cell lines amenable to experimentation. Most often, the growth phenotype remained unchanged during months of culture. Transfection of an epidermal growth factor receptor (EGFR) gene was used to generate a subline that was conditionally transformed (colony formation in soft agar was dependent on transforming growth factor alpha). v-raf transfection was used to generate constitutive transformation. Thus, these cell lines appear to be excellent experimental models for progressive transformation. With them, untested hypotheses of brain tumor progression derived from human genetic studies may be tested experimentally.

Effect of glycoinositolphospholipid anchor lipid groups on functional properties of decay-accelerating factor protein in cells.

The inositol ring in the glycoinositolphospholipid (GPI) anchor of human decay-accelerating factor (DAF) is unmodified in nucleated cells, whereas it is fatty acid acylated in erythrocytes (Ehu). To assess the effect of this and of the glycerol sn-2-associated acyl substituent on the abilities of DAF to cell membrane incorporate and function, 1) endogenous (physiologically anchored) DAF proteins bearing three- and two-"footed" GPI anchors were purified from Ehu and HeLa cells and 2) synthetic DAF variants bearing alternative one- "footed" anchors (retaining either the sn-1 glycerol- or inositol-associated lipid) were prepared by alkaline hydroxylamine treatment and phosphatidylinositol-specific phospholipase D digestion of Ehu DAF, respectively. The different DAF species were added to antibody-sensitized sheep erythrocytes (EshA) and their abilities to insert into the plasma membranes of the cells and control subsequent complement activation on their surfaces were compared. DAF proteins bearing all four GPI anchor structures adhered to the Esh hemolytic intermediates and inhibited expression of C3 convertase (C4b2a) activity. However, mixing of DAF-treated EshA with untreated EshAC142 and stripping of cell-associated DAF proteins with vesicles showed that only the physiologically anchored proteins remained stably associated with the lipid bilayer and functioned intrinsically. Both three- and two-"footed" Ehu and HeLa DAF proteins exhibited comparable ability to incorporate and function in the intermediates as well as to accumulate to levels 1000-fold higher/cell in Schistosoma mansoni schistosomula. These findings indicate that 1) an intact inositolphospholipid-containing GPI anchor is necessary for stable membrane integration and intrinsic function, 2) endogenous GPI anchors (with either unsubstituted and acylated inositol) incorporate and function with comparable efficiency, and 3) the transfer of either endogenous DAF form can account for the previously described circumvented uptake of human C3b by blood stage schistosomula.

The influence of membrane components on regulation of alternative pathway activation by decay-accelerating factor.

Decay-accelerating factor (DAF) is a C regulatory protein which functions in membranes to inhibit autologous C activation on cell surfaces. A liposome model was used to study the mechanism of DAF action and examine the effects of membrane-bound glycophorin and LPS on the regulatory activity of DAF. Liposomes were incubated in MgEGTA-treated human serum and activation of the alternative pathway measured by C3b binding. Liposomes composed of phosphatidylcholine, phosphatidylethanolamine, and cholesterol activated the alternative pathway in proportion to their content of PE. Incorporation of 10(-7) mol/mol phospholipid of either human E or HeLa cell-derived DAF inhibited C activation by liposomes containing 40% phosphatidylethanolamine by 50%, an efficiency comparable to that observed in intact E. HeLa DAF that had been treated with phosphatidylinositol-specific phospholipase C to remove its glycolipid anchor had no effect on C activation by liposomes at concentrations as high as 10(-5) mol/mol phospholipid. Incorporation of DAF into liposomes prepared with bound C3b inhibited the deposition of additional C3b by C3bBbP. However, the incorporated DAF increased the amount of Bb generated from B in the presence of D indicating that accelerated decay of the convertase was the primary effect of DAF. Similarly, treatment of intact human E with anti-DAF decreased the amount of Bb generated by the alternative pathway convertase. To study the effects of other membrane components on DAF activity, liposomes were prepared with purified human glycophorin A or LPS. In glycophorin liposomes the presence of PE was required to activate the alternative pathway and DAF inhibited this activation. In contrast, LPS liposomes bound C3b independently of PE and the incorporation of DAF had no effect. These results demonstrate that within a membrane, DAF's inhibitory activity on the alternative pathway C3 convertase is mediated independently of other membrane proteins, that in this model the major activity of DAF is to accelerate convertase decay, and that the presence of other membrane molecules that may serve as C3 acceptors can circumvent DAF function.

Expression of two molecular forms of the complement decay-accelerating factor in the eye and lacrimal gland.

Complement is present in ocular fluids, but the molecular mechanism(s) restricting its activation to exogenous targets and not to autologous ocular cells are currently unknown. To clarify how this control is achieved, monoclonal antibody (mAb)-based techniques were used to examine the eye, the lacrimal gland, and ocular fluids for the decay-accelerating factor (DAF), a membrane regulatory protein which protects blood cells from autologous complement activation on their surfaces. Immunohistochemical staining of tissue sections revealed DAF antigen on corneal and conjunctival epithelia, corneal endothelium, trabecular meshwork, and retina, as well as on lacrimal gland acinar cells and in adjacent lumens. By flow cytometry, cultures of conjunctival epithelium exhibited the highest DAF levels and levels on corneal epithelium greater than corneal endothelium greater than conjunctival fibroblasts. Biosynthetic labeling of corneal endothelium yielded de novo DAF protein with an apparent molecular weight (Mr) of 75 kD, approximating that of blood cell DAF protein, and digestions of conjunctival epithelium with phosphatidylinositol-specific phospholipase C (PI-PLC), an enzyme which cleaves glycoinositolphospholipid membrane anchors, released approximately 70% of the ocular surface DAF protein similar to leukocyte surface DAF protein. Quantitations of DAF by radioimmunometric assay employing mAbs against two DAF epitopes revealed 325 ng/ml (n = 12), 4.8 ng/ml (n = 10), and 22.0 ng/ml (n = 8) of soluble DAF antigen in tears, aqueous humor, and vitreous humor, respectively. Western blot analyses of the tear DAF antigen revealed two DAF forms, one with an apparent Mr of 72 kD resembling membrane DAF forms in other sites, and a second with an apparent Mr of 100 kD, which is previously undescribed. Since DAF activity is essential physiologically in protecting blood cells from autologous complement attack, the identification of DAF on the ocular surface, intraocularly, in the lacrimal gland, and in tears suggests that DAF-mediated control of complement activation is also required in these locations.

Structural basis for variations in the sensitivity of human decay accelerating factor to phosphatidylinositol-specific phospholipase C cleavage.

Human decay-accelerating factor (DAF) proteins expressed on E and nucleated cells differ in their susceptibility to phosphatidylinositol (PI)-specific phospholipase C (PLC) cleavage/release. To investigate the mechanism of this difference, the glycoinositol-phospholipid anchoring moieties of E DAF, and of HeLa cell, polymorphonuclear cell, and lymphocyte DAF were structurally compared. Labeling of PI-PLC-resistant E DAF with 3-(trifluoromethyl)-3-(m-[125I]-iodophenyl)-diazirine ([125I]TID) and TLC analysis of nitrous acid deamination anchor fragments showed a predominant phospholipid species with less polar migration than the 125I-TID-labeled PI. Gas chromatographic analyses of methanolyzed E protein revealed 2.20 +/- 0.16 mol of fatty acids [16:0, 18:0, 18:1, 20:4, 22:4, and 22:5 (0.76, 0.36, 0.25, 0.15, 0.40, 0.28 mol, respectively)] and 0.86 +/- 0.05 mol of inositol per mol of N-terminal Asp. Gas chromatography-mass spectroscopy demonstrated principally myo-inositol but also variable amounts of the chiro-isomer. Nondenaturing polyacrylamide gel electrophoresis of 14C-radiomethylated E protein revealed that pretreatment with hydroxylamine, a reagent which removes ester-linked lipids, rendered it PI-PLC susceptible. In contrast, parallel analyses of 35S-cys-labeled PI-PLC-sensitive HeLa DAF protein revealed only minor amounts of the hydroxylamine-sensitive phospholipid species. Similar results were obtained with 125I-surface-labeled DAF from polymorphonuclear cells, as well as from unstimulated peripheral blood and anti-CD3-activated lymphocytes. These findings demonstrate that, rather than PI, the E DAF anchor contains an inositol alkylacylglycerol-phospholipid which is heterogeneous with respect to acyl groups and inositol isomers, that an ester-linked substitution in this inositolphospholipid underlies the resistance of E DAF protein to PI-PLC cleavage/release, and that this structural modification is cell-specific.

Heightened complement sensitivity of acquired immunodeficiency syndrome lymphocytes related to diminished expression of decay-accelerating factor.

Although the human immunodeficiency virus can induce cytopathic changes in human lymphocytes in vitro, the mechanism(s) underlying progressive lymphopenia in patients with AIDS and AIDS-related complex has not been elucidated. To investigate this issue, peripheral blood lymphocytes of AIDS and AIDS-related complex patients and healthy control subjects were examined for their ability to resist homologous complement-mediated lysis. Upon sensitization with monoclonal antibodies to major histocompatibility complex class I antigen, as much as 48% lysis of patients' cells was observed in as little as a 1:32 dilution of human serum compared to 18 +/- 8% (mean +/- SD) lysis of controls' cells even in a 1:8 dilution of human serum. To investigate the mechanism of the abnormal complement sensitivity, AIDS and AIDS-related complex cells were analyzed for expression of decay-accelerating factor (DAF), a complement regulatory protein that functions intrinsically in blood cell membranes to prevent complement activation on their surfaces. Flow cytometric assays using anti-DAF monoclonal antibodies demonstrated that patients' lymphocytes and monocytes were DAF-deficient, in contrast to their polymorphonuclear leukocytes, which showed normal DAF levels. Expression of DAF was diminished on CD4+ as well as CD8+ T-lymphocyte subpopulations as opposed to expression of CD3, which was comparable in patients and controls. Incubation of normal lymphocytes with anti-DAF monoclonal antibodies or phosphatidylinositol-specific phospholipase C, an enzyme that cleaves DAF, enhanced lysis. Conversely, reconstitution of patients' cells with exogenous DAF reduced their lysis. The findings of heightened complement sensitivity and DAF deficiency of patients' lymphocytes in vitro suggest the possibility that the DAF deficit may contribute to the progressive lymphopenia of AIDS in vivo.

Association of cytoskeletal re-organization with capping of the complement decay-accelerating factor on T lymphocytes.

Recent studies have identified cell-associated proteins that are membrane anchored by glycosyl-inositol-phospholipid structures but the biologic implications of this mode of membrane attachment are incompletely understood. Among proteins anchored in this way is the decay-accelerating factor (DAF), a complement (C) regulatory factor that functions on blood cell surfaces to prevent autologous C attack. As one approach to investigate the functional consequences of glycosyl-inositol-phospholipid-anchoring of DAF in T lymphocytes, the effects of crosslinking surface DAF molecules were compared to those of crosslinking conventionally by anchored cluster of differentiation (CD) proteins. Upon incubation with anti-DAF mAb and anti-murine IgG, DAF re-distributed to a pole of the cell with a t1/2 at 37 degrees C of 4.4 min as compared to t1/2 of 3.5 to 7 min for CD3, CD4, and CD8. Re-distribution of DAF occurred independently of CD2, CD3, CD4, or CD8. Anti-DAF immunoprecipitates of membrane extracts of cells chemically cross-linked with dithiobis(succinimidylpropionate) contained only monomeric DAF. Immunofluorescent staining demonstrated clustered actin, tubulin, and vimentin beneath the capped DAF protein. Pre-treatment of cells with colchicine or 8-azidoadenosine 3',5'-cyclic phosphate, but not lumicolchicine, resulted in reduction of the t1/2 for DAF to 1 to 2.6 min. Conversely, treatment of cells with cytochalasins B or D completely blocked DAF capping. The results indicate that, upon cross-linking, glycosyl-inositol-phospholipid-anchored DAF molecules undergo capping similar to conventionally anchored CD molecules and that DAF capping is associated with cytoskeletal reorganization.

Glycolipid reanchoring of T-lymphocyte surface antigen CD8 using the 3' end sequence of decay-accelerating factor's mRNA.

Decay-accelerating factor (DAF) is one of a family of cell-associated proteins that undergo posttranslational modifications in which glycolipid anchoring structures are substituted for membrane-spanning sequences. The signals that direct the covalent substitution reaction in these proteins are unknown. Human DAF was expressed in Chinese hamster ovary (CHO) cells and murine BW lymphocytes. In both cases, the xenogeneic DAF in transfectants incorporated a glycolipid anchor. A chimeric CD8-DAF cDNA, encompassing the extra-cellular region of the T-lymphocyte surface antigen CD8 and the 3' end of DAF mRNA (encoding the C-terminal region of mature DAF as well as the hydrophobic extension peptide), was expressed in human leukemia lines after transfection with an Epstein-Barr virus-based episomal vector. The chimeric protein in transfectants demonstrated glycolipid anchoring, whereas unaltered CD8 in control experiments did not. The signals directing glycolipid anchoring in eukaryotic cells are thus evolutionarily conserved and contained in the 3' end of the DAF sequence.

Decay-accelerating factor protects human tumor cells from complement-mediated cytotoxicity in vitro.

The disialoganglioside GD2 is expressed on a wide spectrum of human tumor types, including neuroblastomas and melanomas. Upon binding of 3F8, a murine monoclonal antibody (MAb) specific for GD2, neuroblastomas and some melanomas are sensitive to killing by human complement, whereas some melanomas are not. To investigate the mechanism underlying these differences in complement mediated cytotoxicity, complement-insensitive melanoma cell lines were compared with respect to expression of the decay-accelerating factor (DAF), a membrane regulatory protein that protects blood cells from autologous complement attack. While DAF was undetectable among neuroblastomas, it was present in complement-insensitive melanomas. When the function of DAF was blocked by anti-DAF MAb, C3 uptake and complement-mediated lysis of the insensitive melanoma lines were markedly enhanced. F(ab')2 fragments were as effective in enhancing lysis as intact anti-DAF MAb. The DAF-negative and DAF-positive melanoma cell lines were comparably resistant to passive lysis by cobra venom factor-treated serum. The data suggest that in some tumors, DAF activity accounts for their resistance to complement-mediated killing. The ability to render these cells complement-sensitive by blocking DAF function may have implications for immunotherapy.

Identification of the complement decay-accelerating factor (DAF) on epithelium and glandular cells and in body fluids.

Decay-accelerating factor (DAF) is a 70 kD membrane regulatory protein that prevents the activation of autologous complement on cell surfaces. Using immunohistochemical methods and a radioimmunometric assay based on mAbs to DAF, we found large amounts of membrane-associated DAF antigen on the epithelial surface of cornea, conjunctiva, oral and gastrointestinal mucosa, exocrine glands, renal tubules, ureter and bladder, cervical and uterine mucosa, and pleural, pericardial and synovial serosa. Additionally, we detected soluble DAF antigen in plasma, tears, saliva, and urine, as well as in synovial and cerebrospinal fluids. While plasma, tear, and saliva DAF are larger than erythrocyte (Ehu) membrane DAF by Western blot analysis, urine DAF is slightly smaller (67,000) in Mr. Unlike purified Ehu DAF, however, urine DAF is unable to incorporate into the membrane of red cells. Although its inhibitory activity on the complement enzyme C3-convertase is lower than that of Ehu DAF, it is comparable to that of serum C4 binding protein (C4bp). Biosynthetic studies using cultured foreskin epithelium and Hela cells disclosed DAF levels (approximately 2 X 10(5) molecules/cell) exceeding those on blood cells. In addition, these studies revealed the synthesis of two DAF species, one with apparent Mr corresponding to that of epithelial cell membrane DAF and the other to urine DAF, suggesting that the urine DAF variant arises from adjacent epithelium. The function of DAF in body fluids is unknown, but the observation that urine DAF has C4bp-(or factor H-)like activity shows that it could inhibit the fluid phase activation of the cascade.

Decay accelerating factor of complement is anchored to cells by a C-terminal glycolipid.

Membrane-associated decay accelerating factor (DAF) of human erythrocytes (Ehu) was analyzed for a C-terminal glycolipid anchoring structure. Automated amino acid analysis of DAF following reductive radiomethylation revealed ethanolamine and glucosamine residues in proportions identical with those present in the Ehu acetylcholinesterase (AChE) anchor. Cleavage of radiomethylated 70-kilodalton (kDa) DAF with papain released the labeled ethanolamine and glucosamine and generated 61- and 55-kDa DAF products that retained all labeled Lys and labeled N-terminal Asp. Incubation of intact Ehu with phosphatidylinositol-specific phospholipase C (PI-PLC), which cleaves the anchors in trypanosome membrane form variant surface glycoproteins (mfVSGs) and murine thymocyte Thy-1 antigen, released 15% of the cell-associated DAF antigen. The released 67-kDa PI-PLC DAF derivative retained its ability to decay the classical C3 convertase C4b2a but was unable to membrane-incorporate and displayed physicochemical properties similar to urine DAF, a hydrophilic DAF form that can be isolated from urine. Nitrous acid deamination cleavage of Ehu DAF at glucosamine following labeling with the lipophilic photoreagent 3-(trifluoromethyl)-3-(m-[125I]iodophenyl)diazirine ([125I]TID) released the [125I]TID label in a parallel fashion as from [125I]TID-labeled AChE. Biosynthetic labeling of HeLa cells with [3H]ethanolamine resulted in rapid 3H incorporation into both 48-kDa pro-DAF and 72-kDa mature epithelial cell DAF. Our findings indicate that DAF and AChE are anchored in Ehu by the same or a similar glycolipid structure and that, like VSGs, this structure is incorporated into DAF early in DAF biosynthesis prior to processing of pro-DAF in the Golgi.