INTRODUCTION

Sialic acids (Sias)¹ are a family of 9-carbon carboxylated monosaccharides often found as terminal residues of vertebrate oligosaccharides (1). Sialic acid-containing cell surface glycoconjugates (sialoglycoconjugates) are known to influence many biological processes. For example, they may alter the physical properties of the plasma membrane or serve as specific ligands for certain lectins and toxins (1). Sialic acids can also be modified in a variety of ways, giving rise to a family of more than 30 different structures (2, 3) which, in turn, can be presented in a variety of linkages to the underlying sugar chain. These modifications can significantly affect the physicochemical properties of the parent molecule and are therefore predicted to modify their function and/or to create new functions. In mammals, one of the most common modifications of sialic acids is the addition of O-acetyl esters to the hydroxyl group at the C-9 position. These esters are well known to affect recognition of Sias by viral hemagglutinins and bacterial sialidases (2, 3) and by the recently described sialic acid-binding mammalian proteins belonging to the I-type lectin family (4, 5, 6, 7, 8). Indirect evidence also suggests that they may affect tissue morphogenesis during development (9) and modulate the alternative pathway of complement activation (10).

Murine erythroleukemia (MEL) cells are virus-transformed erythroid precursors that proliferate in culture until they are given a signal for terminal differentiation (11). In the preceding paper (12), we showed that expression of 9-O-acetylation on MEL cells is a highly regulated modification, being selectively found on cell surface mucin-type glycoproteins, down-regulated upon cellular differentiation, and affected by the integrity of the microtubular network. We also showed that a recombinant soluble form of the influenza C hemagglutinin esterase (13) allows direct and selective removal of 9-O-acetyl esters from the surface of MEL cells. This work has provided a system wherein the functional consequences of altering the level of cell surface 9-O-acetylation can be directly examined.

Here we have investigated the influence of modulating the levels of MEL cell 9-O-acetylation upon the activation of alternative complement pathway, recognition by some I-type lectins, and on tissue homing in intact mice. We find that in each process, 9-O-acetylation has a substantial influence.

EXPERIMENTAL PROCEDURES

Most of materials used were obtained from Sigma .The following materials were obtained from the sources indicated. Alkaline phosphatase-conjugated goat anti-human IgG antibody was from Bio-Rad; fluorescein isothiocyanate-conjugated goat anti-human IgG and phycoerythrin-conjugated goat anti-human IgG were from CalTag Laboratories (South San Francisco, CA); fluorescein isothiocyanate-conjugated rabbit anti-sheep IgG antibody and Arthrobacter ureafaciens neuraminidase were from Calbiochem; specific anti-factor H antibody (sheep antiserum) was from Accurate Chemicals (Netherlands); diisopropyl fluorophosphate was from Aldrich; dimethyl sulfoxide (Me₂SO) was from Fisher; [³H]thymidine (specific activity 50 mCi/mmol) was from ICN Radiochemicals (Costa Mesa, CA); fetal calf serum was from Hyclone (Logan, UT). Protein assays were determined with the bicinchoninic acid protein assay reagent kit (Pierce) using BSA as a standard. The O-sialoglycoprotease enzyme was a kind gift from Dr. Alan Mellors, University of Guelph, Canada. All other chemicals were of reagent grade or better and were obtained from commercial sources.

Murine erythroleukemia (MEL) cells were obtained from George Palade at University of California San Diego (14, 15) and cultured in RPMI 1640 with 10% heat-inactivated fetal calf serum.

The soluble chimeric protein CHE-Fc, consisting of the extracellular domain of influenza C hemagglutinin esterase fused to the Fc portion of human IgG₁, was generated and characterized as described elsewhere (13). The modified form CHE-FcD was generated by treating CHE-Fc with 1 mM diisopropyl fluorophosphate to inactivate the esterase as described (13). CHE-Fc specifically releases 9-O-acetyl esters from sialic acids, whereas CHE-FcD specifically recognizes and binds to 9-O-acetylated sialic acids (13). The CD22Rg chimera was prepared exactly as described previously (16), by stable expression in wild-type Chinese hamster ovary cells. Chimeric sialoadhesin was purified from a Chinese hamster ovary lec2 cell line that was stably transfected with a plasmid encoding a chimeric fusion protein of the outer three Ig domains of murine sialoadhesin fused to the Fc portion of human IgG. The original construct (Sn-Rg) has been described elsewhere (17) and was kindly provided by Paul Crocker (ICRF, Oxford, UK) and Sorge Kelm (University of Kiel, Germany).

Blood was collected from volunteers and allowed to clot at room temperature for 1 h. Serum was separated from the clot by centrifugation at 2000 rpm for 20 min and stored in aliquots at 70 °C until use.

Washed cell pellets were resuspended into a buffer containing 50 mM Tris (pH 7.5), 150 mM NaCl, 1% Triton X-100, and protease inhibitors (1 mM EDTA, 50 µg/ml leupeptin, and 4 µg/ml pepstatin) and incubated for 10 min at room temperature with occasional mixing. After high speed ultracentrifugation, the supernatant was saved for analysis.

1 × 10⁶ MEL cells were resuspended in 100 µl of PBS, containing 0.02% NaN₃ (control cells), or 100 ml of PBS, containing 0.02% NaN₃ with 10 µg of CHE-Fc. After incubation at 37 °C for 60 min, the cells were washed three times in ice-cold PBS and then stained for flow cytometry analysis as described below.

1 × 10⁶ MEL cells were resuspended in 100 µl of 50 mM NaAc (pH 6.0) with 10 milliunits of A. ureafaciens sialidase at 37 °C for 2 h, and the cells were washed three times in ice-cold PBS to remove the residual enzyme.

Proteins (40-µg aliquots) from MEL cells were incubated with 2.5 milliunits of enzyme in 100 mM HEPES (pH 7.4), in the presence of BSA (stabilizer) for 2 h at 37 °C. For treatment of cell surface molecules, MEL cells (100 µl, 10⁶ cells/ml) in Hanks' balanced saline solution, 1% fetal calf serum, 0.1% NaN₃, and 10 mM Neu2en5Ac (sialidase inhibitor) were incubated for 2 h at 37 °C in the presence or absence of 5 milliunits of O-sialoglycoprotease. The cells were then washed with Hanks' balanced saline solution, 1% fetal calf serum, 0.1% NaN₃ prior to the cell lysis assay.

Proteins were separated by SDS-polyacrylamide gel electrophoresis in 7.5% polyacrylamide gels, transferred to Immobilon-P membranes (Millipore), and the blots incubated overnight at 4 °C with sialoadhesinRg (10 µg/ml) in PBS, 1% BSA. Glycoproteins reacting with the chimera were colorimetrically detected with a goat anti-human IgG antibody conjugated with alkaline phosphatase and developed with p-nitrophenyl phosphate.

Differentiation was induced by growing MEL cells (starting inoculum of 5 × 10⁵ cells/ml) in the presence of 2% Me₂SO in complete medium for 3 days. The extent of spontaneous differentiation of MEL cells was determined using benzidine staining to detect hemoglobin accumulation.

MEL cells were washed three times with 1% BSA in PBS and incubated in the same buffer with a mixture of fluorescein isothiocyanate-conjugated goat anti-human IgG and CHE-FcD or phycoerythrin-conjugated goat anti-human IgG and sialoadhesinRg for 2 h at 4 °C. Control cells were incubated with similar amounts of human IgG₁ in 1% BSA in PBS. Cells were washed once in PBS, fixed in 2% formaldehyde in PBS, and analyzed by flow cytometry using a Becton Dickinson FACScan instrument. Intact cells were gated based on their forward and side scattering characteristics.

Activation of the human alternative pathway was measured by direct lysis of MEL cells in dilutions of normal human serum. The serum was diluted with Gelatin Veronal Buffer containing Mg²⁺ and EGTA (GVB-Mg-EGTA, Sigma) to prevent any contribution by the classical complement pathway (18, 19, 20). MEL cells (1 × 10⁶ in 0.1 ml of GVB-MG-EGTA, buffer) were added to serially diluted serum. The reaction mixtures were incubated for 45 min at 37 °C with agitation. 1 ml of PBS was added to each sample, and the percentage lysis of the MEL cells was determined by trypan blue staining. The reagent blank showed no lysis; 100% lysis was obtained with the addition of 1 ml of distilled H₂O to 1 × 10⁶ MEL cells.

MEL cells (1 × 10⁶ in 100 µl of GVB²⁺ buffer) were incubated with normal human serum (1:1 in PBS) at 37 °C for 15 min. The cells were washed twice with PBS, 1% BSA and resuspended in 100 µl of sheep anti-human factor H antibody (30 µg/ml) in PBS, 1% BSA. Samples were incubated on ice for 1 h, washed with PBS, 1% BSA, and resuspended in 100 µl of fluorescein isothiocyanate-conjugated rabbit anti-sheep IgG antibody (1:50). After further incubation on ice for 1 h, cells were washed twice, resuspended in 500 µl of PBS, 3.7% formaldehyde, and analyzed for bound factor H protein by measuring fluorescence intensity on a FACScan instrument. The effect of de-O-acetylation and de-sialylation on factor H protein binding to MEL cells was studied by prior treatment of the cells with esterase (CHE-Fc) or sialidase, respectively, as described above.

2 × 10⁵ MEL cells were labeled with 100 µCi of [³H]thymidine for 3 days and chased in normal media overnight. The labeled cells were washed with PBS, suspended in 100 µl of PBS, and treated with CHE-Fc/or sham-treated at 37 °C for 1 h. 100 µl of the treated and sham-treated MEL cells were injected into the tail vein of 8-week-old female DBA/2J mice (Jackson Laboratory, Bar Harbor, ME). After 2 h, mice were sacrificed under anesthesia, and different organs were collected for analysis. Femoral bone marrow, spleen, liver, lungs, brain, thymus, lymph nodes, heart, and kidneys were removed and homogenized in PBS containing 1% Triton X-100, using a Polytron (Kinematica, Switzerland). The extracts of different organs were analyzed for presence of ³H radioactivity as an index of presence of MEL cells (0.5 ml of extract/20 ml of scintillation fluid). Under these conditions, counting efficiency was comparable among the different extracts.

RESULTS AND DISCUSSION

As shown in Fig. 1A, MEL cells are sensitive to lysis by normal human serum. Such spontaneous lysis by heterologous serum can be mediated by either the classical (antibody-dependent) or the alternative (antibody-independent) pathways of complement activation. The specific involvement of the alternative pathway in this lysis is indicated by the use of GVB-Mg-EGTA buffer to prevent any contribution of the classical complement pathway (18, 19, 20) and by the striking effects of dilution (Fig. 1A). This is confirmed by the finding that human serum deficient in the alternative pathway factor B does not kill cells (Fig. 1B). Previous studies have shown that lysis of murine erythrocytes by human serum occurs via the alternative pathway (10) and that the extent of lysis roughly correlated with the level of 9-O-acetylation of sialic acids (10). Although not proven directly, it was suggested that the O-acetyl groups were abrogating the normal function of sialic acid in restricting the alternative pathway of complement lysis. This type of restriction is caused by the regulatory factor H (18, 19, 20, 23), a soluble factor in serum which is believed to bind to surfaces via the intact exocyclic (C₇-C₈-C₉) side chain of sialic acids (18, 19, 20, 23). Since MEL cells also have cell surface 9-O-acetylation, we considered the possibility that a similar mechanism might be operating in this case.

In the preceding paper (12), we showed that induced differentiation of MEL cells is accompanied by the loss of cell surface 9-O-acetylation. As shown in Fig. 2, such differentiated MEL cells become resistant to human serum-mediated lysis, even when exposed to a high concentration (1:1 dilution) of serum. This could potentially be a direct result of the marked loss of cell surface 9-O-acetylation which should allow increased binding of factor H. However, differentiation is a complex process, and this loss of sensitivity could also be due to some unrelated molecular changes. To address this concern, we studied the time dependence of changes in sensitivity to lysis during the course of differentiation, in relationship to the density of cell surface 9-O-acetylation, as measured by flow cytometry with the FcD probe. As shown in Fig. 3, the extent of loss of cell surface 9-O-acetylation over time following the induction of differentiation correlates well with the increasing resistance to lysis.

To more directly address the role of cell 9-O-acetylation in complement sensitivity of MEL cells, we used two ways to remove cell surface 9-O-acetyl groups. In the preceding paper (12), we showed that treatment with the enzyme O-sialoglycoprotease selectively removes the cell surface mucins carrying O-acetylated sialic acids. As shown in Fig. 4A, this treatment also gave protection from lysis. An even more selective removal of 9-O-acetyl groups can be achieved by treatment with the recombinant soluble influenza CHE-Fc esterase. As shown in Fig. 4B, treatment with increasing concentrations of this enzyme also resulted in increasing resistance to lysis.

In the accompanying paper (12), we also showed that treatment with nocodazole causes an accumulation of cell surface 9-O-acetylation on MEL cells, apparently by preventing endocytosis and turnover of the cell surface molecules. As shown in Fig. 5, this treatment is accompanied by the increased sensitivity to killing by diluted serum. Taken together with the effects of induced differentiation and enzymatic treatments described above, these data directly demonstrate for the first time that 9-O-acetylation can regulate the alternative pathway of complement. The prior work of others indicates that the exocyclic side chain of sialic acids is required for binding of factor H of the alternative pathway (18, 19, 20, 23), which normally restricts amplification of the C3bBb complex on surfaces (20, 23, 24, 25). Thus, it is reasonable to suggest that substitution of this side chain with a bulky O-acetyl ester abrogates this restrictive function, allowing amplification to proceed.

To test the hypothesis mentioned above, we studied direct binding of human factor H to MEL cells, exploiting different conditions that allowed us to modify cell surface sialylation and acetylation of sialic acids. Factor H binding was studied by incubating cells for 15 min at 37 °C with human serum, which has ~0.5 mg/ml factor H (25). After this incubation, factor H binding was monitored using anti-factor H antibody and an adequate fluoresceinated conjugate by flow cytometry. As shown in Fig. 6, factor H binding increased significantly upon Me₂SO-induced differentiation. As we discussed in the accompanying paper (12), Me₂SO-induced differentiation was followed by increase in wheat germ agglutinin binding on the cell surface and decrease of sialic acid O-acetylation in sialomucins. Factor H binding to MEL cells is dependent on sialic acids, since it could be abrogated by sialidase treatment (Fig. 6, sialidase treatment also caused a loss of wheat germ agglutinin reactivity, indicating the efficiency of the treatment, data not shown). De-O-acetylation of MEL sialic acids was achieved by esterase treatment (CHE-Fc treatment). Factor H binding in esterase-treated cells was significantly increased, suggesting that the O-acetyl groups can mask factor H recognition or binding to sialic acid-rich surfaces as those represented by MEL cells. These results correlate well with the increased sensitivity of both Me₂SO-differentiated and esterase-treated MEL cells to the lytic activity of the alternative pathway of complement. They also illustrate the role of sialic acid modifications in modulating the binding activity of sialic acid binding proteins (in this case the complement factor H) and thus their function.

Recently, several other sialic acid-binding molecules have been found in vertebrate systems. One group is a newly recognized family of mammalian lectins belonging to the immunoglobulin superfamily (I-type lectins), which include sialoadhesin and CD22 (8). Sialoadhesin is an adhesion molecule restricted to macrophages, and its highest expression is found in hematopoietic and lymphoid tissues like bone marrow, spleen, and lymph nodes. Sialoadhesin mediates cell adhesion by binding to cell surface glycoconjugates terminating in Sia2-3Gal1-3GalNAc and/or Sia2-3Gal1-3(4)GlcNAc and has been proposed to play a role in the interactions of macrophages with blood cells (5). Earlier studies have shown that sialoadhesin can bind to MEL cells (26). CD22 is known to bind selectively to the sequence Sia2-6Gal1-4GlcNAc. Since I-type lectins usually require the side chain of sialic acids for recognition (5, 6, 7, 17, 27, 28), we investigated the possibility that the cell surface 9-O-acetylation on these cells might be restricting binding of sialoadhesin. Indeed, as shown in Fig. 7A, the binding of recombinant soluble murine sialoadhesin to MEL cells was increased upon differentiation, shown earlier to be accompanied by loss of cell surface 9-O-acetylation. In keeping with this, sialoadhesin binding to total sialoglycoconjugates on a Western blot is markedly increased after differentiation. Interestingly, the increased binding appears to be confined to the O-sialoglycoprotease-sensitive sialomucins (Fig. 7B). Some binding of CD22Rg was also noted but no increase occurred upon CHE-Fc esterase treatment (data not shown). This is in contrast to some other cell types such as lymphocytes and hepatocytes, in which O-acetylation can mask CD22 ligands (4).

Although differentiation of MEL cells gave a decrease of 9-O-acetylation and increase of sialoadhesin binding, the two phenomena could still be unrelated. To directly address this issue, we treated MEL cells with the CHE-Fc esterase and then studied sialoadhesin binding by flow cytometry. As shown in Fig. 8, this treatment resulted in a >1 log increase in the mean fluorescence intensity of sialoadhesin binding. Taken together, these data show that 9-O-acetylation abrogates recognition of sialomucins by sialoadhesin. They also fit with recent studies by others (7) in which exposure of murine erythrocytes to whole virions of influenza C increased interactions with macrophages carrying sialoadhesin.

Upon injection into syngeneic DBA/2J mice, MEL cells are known to colonize organs like the liver, spleen, and bone marrow, eventually resulting in death from leukemic infiltration (11). However, the adhesive mechanisms involved in the initial homing of injected cells are unknown. Since sialoadhesin is known to be present in lymph nodes, spleen, bone marrow, and liver, we reasoned that specific interactions could occur with injected MEL cells, which in turn could be affected by the level of 9-O-acetylation. We therefore labeled MEL cells in culture with [³H]thymidine, injected them into the tail veins of DBA/2 mice, and studied the short-term homing of the injected cells by harvesting organs after 2 h. As shown in Fig. 9, a major fraction of the injected cells was recovered in the spleen and liver, and this fraction was significantly increased by prior treatment of the cells with the CHE-Fc esterase. Some changes were also noted in recovery in other organs. However, the only other statistically significant change was the decrease in homing to the brain. This in vivo study involves complex variables and is obviously subject to many interpretations. Since induced differentiation of cells is also a complex process, we did not study the effects of this manipulation on homing. Regardless, the effects of the direct enzymatic esterase treatment indicate that the extent of cell surface 9-O-acetylation can modulate the tissue distribution of the erythroleukemia cells within the intact animal.

CONCLUSIONS AND PERSPECTIVES

Many potential functions have been attributed to the 9-O-acetylation of sialic acids (2, 3), but few have been directly proven. Here we have developed a system wherein some of these hypotheses can be tested. Thus, we have shown that 9-O-acetylation can abrogate the normal function of the exocyclic side chain of Sia in preventing activation of the alternative complement pathway (18, 19, 20, 23). Addition of a bulky acetyl group caused a loss of binding of factor H, a protein which normally restricts the amplification loop of alternative pathway activation by dissociation of Bb from C3b and by acting as a cofactor for factor I proteolysis of C3b (18, 19, 20, 23, 24, 25). We have also shown here an example of how an O-acetyl group can block recognition of sialic acid by an I-type lectin, sialoadhesin. In other work, it has been shown that influenza A and B virus hemagglutinin recognition of sialic acids cannot tolerate O-acetyl substitution of the side chain (29). In each case, O-acetylation can thus potentially act as a physiological modulator of binding. On the other hand, some viral and animal hemagglutinins require a 9-O-acetyl group for recognition.

There are many other intriguing observations that suggest that 9-O-acetylation may modulate biological phenomena involving sialic acids. For example, indirect observations suggest the O-acetylation of sialic acids on murine erythrocytes may confer resistance to malarial parasite binding (30). Also, in adult mice subjected to repeated bleeding, the level of 9-O-acetylation in circulating red blood cells appeared to increase (10). Some thymocytes and mature lymphocytes have O-acetylated Sias, whereas others do not (4, 31, 32, 33). 9-O-Acetylation is also selectively expressed at high levels on certain lymphocyte populations (32, 33, 34, 35, 36), and T-cells of patients with various malignancies are said to acquire high levels of O-acetylation (37). O-Acetylated Sias are often found as developmentally regulated components of neuroectodermal gangliosides (38, 39, 40, 41), suggesting a role in morphogenesis. The approaches used in the present study might be adapted to successfully elucidate the significance of some of these phenomena as well.