TY - JOUR
AU - Olafson, Robert, W.
AB - Abstract The secreted acid phosphatase (SAcP) of L.donovani is a heterogeneous glycoprotein that displays a wide array of N- and O-linked glycosylations. The O-linked sugars are of particular interest due to their similarity to the phosphoglycan structures of the major lipophosphoglycan surface antigen and released phosphoglycan (Turco et al., 1987; Greis et al., 1992). This study describes a structural analysis of the SAcP O-linked glycosylations using mass spectroscopy, amino acid sequencing, and enzymatic carbohydrate sequencing. Analysis of glycan chain lengths and peptide glycosylation site distribution was performed, revealing that the average O-linked structure was ∼32 repeat units in length. Amino acid sequence analysis of glycosylated peptides showed that phosphoglycosylations did not occur randomly but were localized to specific serine residues within an array of degenerate serine/threonine-rich repeat sequences localized in the C-terminus. No evidence was obtained for modification of threonine residues. The observed pattern suggested that a consensus sequence may exist for localization of phosphoglycan structures. consensus sequence, in vitro glycosylation, phosphodiester, phosphoglycan, structure Introduction With the exception of L.major, most species of Leishmania secrete large quantities of an acid phosphatase enzyme into the growth medium. Initial interest in this enzyme related to the fact that it was by far the most abundant of ∼40 different proteins secreted by the parasite (Bates et al., 1988) leading to the speculation that any enzyme produced in such large quantities should play a significant role in parasite survival. However, recent studies of this enzyme have focused more upon a unique post-translational modification. Bates et al. (1987), Jaffe et al. (1990) and Ilg et al. (1993) discovered that MAbs specific to the lipophosphoglycan (LPG) strongly cross-reacted with the secreted acid phosphatase (SAcP), suggesting that the SAcP shared an antigenic determinant in common with this major surface glycolipid. Subsequent work by Ilg et al. (1994) demonstrated that the common structure was a repeating chain of phosphodisaccharides with a subunit structure of (-6Gal-β-1-4Man-a-1-PO4-)n. The number of repeats present in LPG ranged from 16 to 32, depending upon species and lifecycle stage (McConville et al., 1992). The SAcP in L.mexicana has been shown to have fairly short glycan chains ranging from 1–3 repeat subunits in length which are connected to the protein through phosphodiester linkages to serine residues (Ilg et al., 1994). Phosphoglycosylation of proteins is a novel post-translational modification first described in a cysteine proteinase of Dictyostelium discoideum which was shown to have serines modified by phosphodiester linked GlcNAc residues (Gustafson and Milner, 1980). Phosphoglycosylations have also been described for a Trypanosoma cruzi surface antigen (Haynes et al., 1996). Unfortunately, the phosphoglycan chains produced by Leishmania are acid labile making purification and analysis of phosphodiester-linked glycans difficult. These novel structures have been localized to a serine rich region of the L.mexicana SAcP (Weise et al., 1995), but not to specific residues within this region. The present work focuses upon the SAcP of L.donovani, addressing the question of the phosphoglycan structures, the number of repeats per chain, and the density of glycosylation, as well as identifying the apparent consensus protein sequence recognized in these novel O-linked modifications. Results The O-linked glycans of the SAcP The L.donovani SAcP is a structurally complex molecule which runs on SDS-PAGE as a broad band with a minimum Mr of 100 kDa. In L.mexicana, the polydisperse nature of this protein has been shown to arise from extensive posttranslational modifications involving phosphorylations and glycosylations (Ilg et al. 1994). However, the precise location of these modifications has not been reported from any Leishmania species and the structures have not been elucidated for the old world L.donovani species. In order to investigate the phosphoglycans, we have taken advantage of the acid lability of the phosphodiester linkages, where mild acid treatment results in specific cleavage at hexose-1-phosphate linkages (Bates et al., 1990; McConville et al., 1990). This treatment caused the elimination of the broad SDS-PAGE band and the appearance of a pair of bands at 95 and 107 kDa (Figure 6, lanes A and B), presumably representing the two L.donovani SAcP gene products (Shakarian et al., 1997). Electrospray ionization mass spectral analysis of mild acid released glycans Following the mild acid procedure, the main glycan component released from the SAcP was found to have a mass of 421 Da (Figure 1), consistent with a single phosphodisaccharide. A number of other peaks, of lesser abundance, were also observed, which had masses consistent with incompletely cleaved fragments of the phosphoglycan polymers modified with di-or trisaccharide neutral hexoses. The latter were assumed to be the terminal structures of the phosphoglycan chains as reported by Thomas et al. (1992) in Leishmania donovani LPG (Table I). The observed charge state for each species in the MS data was dependent on the number of phosphate moieties present in the respective structures. The presence of phosphate in these structures was confirmed by collision induced dissociation (CID) of the 421 Da peak which gave rise to CID fragments with atomic masses of 79 and 96 Da, representing PO3- and PO4-, respectively (Figure 2). Fig. 1. Open in new tabDownload slide Negative ion mass spectrum of the glycans released from the SAcP by mild acid treatment. The ion at m/z 421 represents a phosphodisaccharide consistent with the expected PO4-Gal-Man repeat structure. Fig. 1. Open in new tabDownload slide Negative ion mass spectrum of the glycans released from the SAcP by mild acid treatment. The ion at m/z 421 represents a phosphodisaccharide consistent with the expected PO4-Gal-Man repeat structure. Fig. 2. Open in new tabDownload slide Collision-induced dissociation (CID) mass spectrum of the 421 molecular ion from Figure 1. The ions at m/z 78.8 and 97.0 represent phosphate that is being released from the molecule. A number of lower abundance ions were also present representing various fragments of the carbohydrate portion of the molecule. Fig. 2. Open in new tabDownload slide Collision-induced dissociation (CID) mass spectrum of the 421 molecular ion from Figure 1. The ions at m/z 78.8 and 97.0 represent phosphate that is being released from the molecule. A number of lower abundance ions were also present representing various fragments of the carbohydrate portion of the molecule. Table I. Open in new tabDownload slide Low abundance phosphoglycan fragments observed by mass spectrometry Table I. Open in new tabDownload slide Low abundance phosphoglycan fragments observed by mass spectrometry Fig. 3. Open in new tabDownload slide Enzymatic sequencing and Biogel P4 analysis of the major oligosaccharide released by anhydrous hydrazine treatment. The scale at the top of each panel indicates the elution profile of dextran standards. (A) represents the initial structure released by hydrazinolysis which is consistent with the size of a phosphodisaccharide. (B) shows the elution profile after HF dephosphorylation and is consistent with a disaccharide. (C) shows the elution profile after subsequent treatment with β-galactosidase. Fig. 3. Open in new tabDownload slide Enzymatic sequencing and Biogel P4 analysis of the major oligosaccharide released by anhydrous hydrazine treatment. The scale at the top of each panel indicates the elution profile of dextran standards. (A) represents the initial structure released by hydrazinolysis which is consistent with the size of a phosphodisaccharide. (B) shows the elution profile after HF dephosphorylation and is consistent with a disaccharide. (C) shows the elution profile after subsequent treatment with β-galactosidase. Hydrazinolysis and glycan sequencing BioGel P4 chromatographic separation of the SAcP glycans subjected to hydrazinolysis, [3H]-NaBH4 reduction and mild acid hydrolysis showed that more than 95% of the radiolabel was contained in one species. The radiolabeled sugar appeared in the void volume of a BioGel P4 column developed with water (data not shown), but eluted at a position equivalent to that of a dextran trisaccharide standard when run in the presence of salt (Figure 3a). These findings indicated that the structure was charged, and the size was consistent with the presence of a phosphorylated disaccharide. In order to verify the presence of phosphate, the glycan was treated with anhydrous HF and applied to a P4 column where the radiolabeled sugar was found to elute at a volume consistent with a neutral disaccharide (Figure 3b). After treatment with β-galactosidase, the radioactive product was found to elute as a monosaccharide (Figure 3c). Dionex anion exchange chromatography of the acid hydrolyzed glycan indicated that the sample contained only galactose and mannitol (data not shown). These data, together with the fact that β-galactosidase treatment was ineffective prior to HF treatment, indicated that the structure of the main component released from the SAcP by hydrazinolysis/mild acid was PO4-Gal-Man, confirming the structural similarity of these glycans with the LPG repeating structure and the New World Leishmania mexicana SAcP (Turco et al., 1987; McConville et al., 1990; Ilg, 1994). Fig. 4. Open in new tabDownload slide Negative ion mass spectrum of glycans released by partial mild acid hydrolysis of the SAcP. The reaction required an hour to reach completion and this data represented a 12 min time point. A series of ions can be seen representing phosphoglycan polymers of increasing length. The ions in the series are labeled in terms of the number of phosphoglycan units or phosphates present in each ion. Thus, the ion at 412 represents a phosphoglycan dimer, with two phosphates and a mass of 826, which is consistent with (PO4-Gal-Man)2. Fig. 4. Open in new tabDownload slide Negative ion mass spectrum of glycans released by partial mild acid hydrolysis of the SAcP. The reaction required an hour to reach completion and this data represented a 12 min time point. A series of ions can be seen representing phosphoglycan polymers of increasing length. The ions in the series are labeled in terms of the number of phosphoglycan units or phosphates present in each ion. Thus, the ion at 412 represents a phosphoglycan dimer, with two phosphates and a mass of 826, which is consistent with (PO4-Gal-Man)2. Elucidation of the length of the O-linked glycans To ascertain the length of the O-linked sugars present on the SAcP of L.donovani, further mass spectral analysis was performed on the sugars released by mild acid treatment. Several aliquots were removed at different time points during the hour-long reaction and analyzed by ESI-MS. Since all the glycan phosphodiester bonds should have similar susceptibility to acid treatment, the largest phosphoglycan fragments should be apparent early in the hydrolysis reaction. These data confirmed this prediction. The MS results in Figure 4 revealed glycan chains of up to six repeat units derived from L.donovani SAcP. Larger fragments were not readily resolved by this method. Because the preceding data was not quantitative, a more precise determination of the average length of the O-linked glycan chains on the SAcP was undertaken by first determining both the monosaccharide composition and the protein concentration of a sample and finally comparing these data, on a molar basis, with the number of available glycosylation sites. Amino acid sequencing and MS data (discussed below) indicated that there was a limited number of O-linked sites present in the protein and that only the first serine in each of the C-terminal repeat structures was phosphoglycosylated. Using this information, in conjunction with the published gene sequence for the L.donovani SAcP (Shakarian et al., 1997), it was possible to determine that the maximum number of modification sites per protein molecule was 23. The protein concentration in the sample was then determined by amino acid analysis which, upon multiplication by the calculated number of modification sites/mole of protein, provided an estimate of the number of available modification sites within that sample. This value was then compared to the monosaccharide composition data obtained by high performance anion exchange chromatography. Both mannose and galactose were present in equal quantities, and both were used to perform the calculation. These calculations revealed that the O-linked glycan chains of the SAcP contained an average of 32 repeat units per phosphoglycosylation site. This was significantly longer than those found on the SAcP of L.mexicana (Ilg et al., 1994) but was in good agreement with the average size observed for LPG phosphoglycan chains from late log promastigotes (McConville et al., 1990; Sacks and Turco, 1990). Thus the biosynthetic machinery necessary to elaborate such extended structures is known to be present in Leishmania promastigotes. Fig. 5. Open in new tabDownload slide Thin layer chromatography of acid hydrolyzed SAcP. Prior to hydrolysis the protein was radiolabeled with [32P]-PO4 and the O-linked phosphoglycan chains were depolymerized with mild acid leaving single phosphate residues bound to the protein. The data shows that only phosphoserine was present. Overexposure of the plate failed to reveal any other phosphorylated amino acids. Fig. 5. Open in new tabDownload slide Thin layer chromatography of acid hydrolyzed SAcP. Prior to hydrolysis the protein was radiolabeled with [32P]-PO4 and the O-linked phosphoglycan chains were depolymerized with mild acid leaving single phosphate residues bound to the protein. The data shows that only phosphoserine was present. Overexposure of the plate failed to reveal any other phosphorylated amino acids. Identification of the glycosylation sites and modified amino acid residues in SAcP Identifying the sites of phosphoglycosylation in the SAcP structure was complicated by varying glycan lengths coupled with the consequent poor yields upon protein digestion. Attempts to isolate glycopeptides consistently led to complex and unresolvable mixtures. In light of this, several indirect methods were undertaken to identify these modification sites. During [32P]-PO4 radiolabeling experiments, the SAcP was found to retain a significant portion of the radiolabel after mild acid treatment while most of the carbohydrate was removed from the protein. These preparations were subsequently used to identify the amino acid residues retaining the phosphate label. This was accomplished by subjecting the protein to total amino acid hydrolysis followed by separation of the amino acid mixture by thin layer chromatography using a solvent system designed to resolve phosphorylated amino acids. The result, shown in Figure 5, clearly demonstrated the presence of phosphoserine as the only phosphorylated amino acid, along with free phosphate and a small amount of undigested peptide. Although the length of the hydrolysis was varied to allow detection of other phosphoamino acids (Martensen et al., 1984), phosphoserine was the only radiolabeled amino acid observed. Fig. 6. Open in new tabDownload slide SDS-PAGE of various stages during the preparation of radiolabeled SAcP peptides for sequence analysis. Lane A, reduced and alkylated SAcP. Lane B, mild acid treated SAcP. Lane D, CNBr and mild acid treated SAcP. Gels as shown were stained with Stains-All; however, identical results were achieved with [32P]-PO4 autoradiography. Lane C contains a set of molecular weight standards. Fig. 6. Open in new tabDownload slide SDS-PAGE of various stages during the preparation of radiolabeled SAcP peptides for sequence analysis. Lane A, reduced and alkylated SAcP. Lane B, mild acid treated SAcP. Lane D, CNBr and mild acid treated SAcP. Gels as shown were stained with Stains-All; however, identical results were achieved with [32P]-PO4 autoradiography. Lane C contains a set of molecular weight standards. In a recent publication, we showed that the L.donovani SAcP had a serine and threonine rich C-terminal region (Shakarian et al., 1997), similar to that reported for L.mexicana SAcP (Weise et al. 1995). In order to investigate the position of the occupied sites within the enzyme, an attempt was made to isolate and sequence the Ser/Thr rich region, using peptide fragments generated by cyanogen bromide (CnBr) and endoproteinase Asp-N cleavage. Based upon the known amino acid sequence, the CnBr reaction was expected to generate a set of small peptides and one large fragment representing the phosphorylated Ser/Thr rich region, which is devoid of Met (see Appendix). In this experiment, [32P]-labeled SAcP was deglycosylated using the mild acid procedure and then subjected to CnBr cleavage. The phosphorylated peptides were separated from nonphosphorylated peptides using hydroxylapatite chromatography and the SDS-PAGE analysis of the resulting fractions is shown in Figure 6. Identical detection patterns were observed with both autoradiography and phosphopeptide sensitive Stains-All staining (Sigma Chemical Co., St. Louis, MO). The progression from a polydisperse, high molecular weight broad band, to bands at 95 and 107 kDa after mild acid treatment, and finally to bands at 40 and 49 kDa after CnBr digestion was readily apparent. Because the Ser/Thr rich region has internal redundancy within its sequence and acidic residues were present at every eighth position, Asp-N digestion was expected to produce peptides with common sequences, amplifying the yield of sequence in the final sample by up to seven fold. The peptides produced by this protocol were purified using Mono-Q anion exchange chromatography, monitoring both 32PO4 and ultraviolet absorption at 230nm (Figure 7). Radioactive peptides were desalted and subjected to amino acid sequence analysis. The results revealed a clearly determinable sequence with the exception of the residue at position six, corresponding to the first serine in the Asp-N C-terminal segment (Table II). After this cycle, the sequence recovered, and although weaker, was readily determined through the next seven cycles, whereupon it terminated. The reduction in yield after cycle six is not uncommonly observed with modified residues and could be attributed to Schiff base formation of alditols from residual labile sugar with open N-terminals. The observed sequence, however, was conclusive. In agreement with the DNA sequence, serine residues always occurred in groups of three, and only the first of these was modified. These data suggested the presence of a consensus sequence for phosphoglycan modification within the repeating protein structure. However, the above data only accounted for one-third of the putative glycosylation sites within the ser/thr rich region, and it remained possible that other sites within the region could also be phosphoglycosylated. Therefore, the preceding experiment was repeated using tryptic digestion in place of CnBr cleavage as the first protein cleavage method. Results of this experiment were similar, in that the ser/thr rich region contained no basic amino acids, thus providing a mixture of small peptides and one large peptide comprising the ser/thr rich region. As shown by capillary electrophoresis, after 5 h of digestion many peaks were present including one broad peak (Figure 8). Because the protein was not deglycosylated prior to digestion, the C-terminal portion would be expected as a broad heterogeneous assemblage of similar large peptides. The separation of the latter group of very similar peptides from the remainder of the digest mixture was performed by Sephadex G-50 gel permeation chromatography where the ser/thr rich peptides appeared in the column void volume along with a minor contamination from a large tryptic fragment residing outside the ser/thr rich region (data not shown). Peptides from the ser/thr rich region were then exposed to a mild acid treatment followed by endoproteinase Glu-C, to cleave the peptides at the repeating acidic residues. After hydroxylapatite chromatography, the peptides were subjected to sequence analysis with the results shown in Table II. As in the previous experiment, the sequence was strong until the first serine was reached, whereupon no residue was observed. This was followed by partial recovery in the next cycle, pointing once again to the first of three serines as the target site for modification. While the possibility remains that other serine residues in this sequence are occasionally modified, they would be at a much lower frequency than the primary site. These findings were corroborated by ESI-MS data of both native protein peptides and synthetic peptides phosphoglycosylated in vitro which also showed that the frequency of phosphoglycosylation never exceeded one modification per repeat unit (discussed below). It should be noted that the degene Fig. 7. Open in new tabDownload slide MonoQ separation (B) of radiolabeled SAcP peptides after CNBr and Asp-N digestion. Collected fractions were monitored for radioactivity (A) and are displayed in counts per minute (c.p.m.). All of the radioactivity in the sample eluted in one peak. The peak width indicates that the radioactive peptides were heterogeneous with respect to extent of deglycosylation and peptide length. Fig. 7. Open in new tabDownload slide MonoQ separation (B) of radiolabeled SAcP peptides after CNBr and Asp-N digestion. Collected fractions were monitored for radioactivity (A) and are displayed in counts per minute (c.p.m.). All of the radioactivity in the sample eluted in one peak. The peak width indicates that the radioactive peptides were heterogeneous with respect to extent of deglycosylation and peptide length. Fig. 8. Open in new tabDownload slide Capillary electrophoresis monitoring of a trypsin digest of the SAcP. Successive panels represent the SAcP at varying stages of digestion. (A) 0 min. (B) 5 min. (C) 2 h. (D) 5 h. (E) represents the completed digest after Sephadex G-50 separation of the large Ser/Thr rich segment from the rest of the digest. The smaller contaminating peptide was removed during subsequent purification, and was not radiolabeled. Fig. 8. Open in new tabDownload slide Capillary electrophoresis monitoring of a trypsin digest of the SAcP. Successive panels represent the SAcP at varying stages of digestion. (A) 0 min. (B) 5 min. (C) 2 h. (D) 5 h. (E) represents the completed digest after Sephadex G-50 separation of the large Ser/Thr rich segment from the rest of the digest. The smaller contaminating peptide was removed during subsequent purification, and was not radiolabeled. Analysis of nonphosphoglycosylated SAcP peptides Finally, an attempt was made to analyze the remaining fragments of the SAcP as generated by tryptic digestion. The peptide mixture was subjected to MS analysis using a MALDI-TOF instrument. The data tabulated in Table III showed the presence of most of the expected peptides. The only peptides unaccounted for were those small peptides obscured by the matrix (<1% of the sequence). All peptides containing a consensus sequence for N-linked glycosylation were observed to be glycosylated, and the masses obtained indicated that the peptides were modified with structures consistent with high mannose type glycosylations. Aside from a glucose terminal hybrid structure, high mannose structures are the only type of N-linked oligosaccharides reported for Leishmania proteins (Olafson et al., 1990; Funk et al., 1997). This data served to confirm that the phosphoglycan modifications did not occur outside the ser/thr rich region, as none of the peptides from this region of the protein were radiolabeled with phosphate and all but a very few of the SAcP peptides were observed by mass spectroscopy. Generation of glycosylated peptides in vitro Studies of the structurally related LPG have shown that it is possible to synthesize phosphoglycan structures in vitro (Carver, 1991, 1992). Consequently, we attempted to employ a synthetic peptide substrate based on the SAcP sequence in an attempt to synthesize phosphoglycosylated peptides. In the first experiment, a peptide was constructed that spanned two full sections of the repeating amino acid sequence found in the ser/thr rich region of the SAcP. This peptide represented the minimum length that would incorporate all of the possible candidates for a phosphoglycosylation consensus sequence. Mass spectral analysis of the peptides generated by this procedure demonstrated the presence of both unmodified peptide (m/z of 858.3 for the doubly charged species) and a single phosphomannose modified peptide (m/z of 879.4) (Figure 9). More abundant modifications were not observed. These data demonstrated the feasibility of generating modified peptides in vitro and, as indicated above, corroborated the previous sequencing data which showed that only a single serine in each serine triplet was modified. Fig. 9. Open in new tabDownload slide Negative ion mass spectrum of an in vitro phophoglycosylated synthetic peptide. The peptide had the sequence SSSEGTTASSSEGTTASSS and the product was a doubly charged species with a m/z value of 979.4 as expected for a molecule modified with a single phosphomannose. Fig. 9. Open in new tabDownload slide Negative ion mass spectrum of an in vitro phophoglycosylated synthetic peptide. The peptide had the sequence SSSEGTTASSSEGTTASSS and the product was a doubly charged species with a m/z value of 979.4 as expected for a molecule modified with a single phosphomannose. Table II. Open in new tabDownload slide Sequencing results for the [32P]-PO4 modified SAcP peptides Table II. Open in new tabDownload slide Sequencing results for the [32P]-PO4 modified SAcP peptides In a second experiment, a smaller peptide was used containing a single eight-residue repeat, including only one potential modification site from the ser/thr rich region. A tryptophan was also included at the N-terminus to increase retention of the peptide on a reverse phase column facilitating subsequent separation. After glycosylation, the products were purified by RP-HPLC and subjected to MS analysis. Again, a small proportion of the material was found to be glycosylated. Unexpectedly, the phosphoglycosylated peptide was found to be one in which the tryptophan residue had been removed during the glycosylation procedure, presumably enzymatically. The reaction product found in greatest abundance was modified with two Gal-Man-PO4 units (Figure 10) as indicated by a doubly charged species with m/z 858.6 (equivalent to a mass of 1719.2 Da). These data indicated that sufficient structural information was available in a small peptide to allow identification and phosphoglycosylation and that higher order structure was unlikely to play a role in selection of modification sites. It was only a coincidence that the above mass matched that of the unmodified peptide used in the previous experiment which was conducted at a different time. Table III. Open in new tabDownload slide MS analysis of a tryptic digest of the SAP Table III. Open in new tabDownload slide MS analysis of a tryptic digest of the SAP Fig. 10. Open in new tabDownload slide Negative ion mass spectrum of an in vitro phosphoglycosylated synthetic peptide. The peptide used had the sequence SSEGTTASSS and the product was a doubly charged species with a m/z value of 858.6 as expected for a molecule modified with two PO4-Man-Gal repeating units. The ion at m/z 1011 was a reagent baseline contaminant. Fig. 10. Open in new tabDownload slide Negative ion mass spectrum of an in vitro phosphoglycosylated synthetic peptide. The peptide used had the sequence SSEGTTASSS and the product was a doubly charged species with a m/z value of 858.6 as expected for a molecule modified with two PO4-Man-Gal repeating units. The ion at m/z 1011 was a reagent baseline contaminant. Discussion The results presented here show that the SAcP of L.donovani is a heavily glycosylated molecule containing more than 700 moles of hexose per mole of protein—largely due to the abundant C-terminal phosphoglycans. We have confirmed the structural similarity between the glycans of the L.donovani SAcP and LPG which was previously implied by anti-LPG monoclonal antibodies and chemical analyses in several Leishmania sp. (Bates et al., 1990; Jaffe et al., 1990; Ilg et al., 1994). The present investigation demonstrated that the glycans of the L.donovani SAcP were phosphodiester linked to serine residues, a linkage that was first described by Gustafson and Milner (1990) on a cysteine proteinase from Dictyostelium. With respect to phosphoglycan length, however, we observed an interesting difference with that reported for L.mexicana (Ilg et al., 1994). The structures on the L.donovani SAcP had an average length of 32 Gal-Man-PO4 repeating units, which was much larger than the average of 1–3 that was reported for L.mexicana (Ilg, 1994). This is perhaps not surprising considering that the L.mexicana enzyme is synthesized as a very high molecular weight filamentous polymer not observed in other species. Differences in the carbohydrate structures of the phosphoglycan repeats may be important in the formation of the unique L.mexicana polymer. It would appear that the phosphoglycan chain structures are consistent among the modified lipids and proteins of each species, which supports the developing evidence that the two structures share a common biosynthetic pathway. Nothing is known about the enzymatic machinery responsible for protein phosphoglycosylation nor is there any information describing the recognition sequence(s) which specify the site of modification. Such a structure must exist, however, because phosphoglycosylation is severely restricted within the 40 or more secreted proteins of L.donovani (Bates et al., 1988), and only specific serine residues are modified in the SAcP structure. Although the possibility exists that the same phosphomannosyl transferase responsible for phosphoglycosylation during LPG synthesis is utilized to modify proteins, it is more likely that separate enzymes function in both cases. Evidence presented here shows that phosphoglycosylation occurs in a regular pattern of modification at very specific sites within the ser/thr rich region of the L.donovani SAcP, suggesting the presence of a consensus identity sequence. These serine residues occur in groups of three, and only the first of these undergoes phosphoglycosylation. There was no evidence of modified threonines in the SAcP. Further studies are presently underway to determine the fine structural details of the primary structure required for transferase recognition. The nature of the biosynthetic apparatus responsible for protein phosphoglycosylation is of particular interest in that a common structure is found on both a protein and a glycolipid, suggesting a common biosynthetic pathway. Carver and Turco (1991, 1992) demonstrated that LPG is synthesized through the alternate addition of mannose-1-phosphate and galactose to the growing glycan chain. The structures on the SAcP could be produced in the same manner. However, addition of the first mannose-1-phosphate must involve a unique substrate recognition process if the reaction is to proceed with both proteins and LPG. Two likely possibilities exist as to how this process might occur. The protein could mimic the LPG core glycan in order to use the LPG phosphomannosyl transferase or, alternatively, a separate enzyme may exist which catalyzes addition of the first residue to proteins. We believe that the latter situation occurs, and experiments are underway to test this hypothesis. Materials and methods Purification of the L.donovani SAcP Leishmania donovani (MHOM/SD/00/1S-2D/LD3) promastigotes were grown to a density of 1.0 × 107 in M199 medium as previously described (Jardim et al., 1995). Spent culture supernatant was passed over a 1.5 × 30 cm octyl Sepharose column followed by a 2.5 × 45 cm DEAE cellulose column equilibrated in 20 mM Tris/HCl, pH 8.0. Unlike most of the contaminating proteins, the SAcP did not bind to the octyl Sepharose column but was retained by the DEAE column which was developed using a two step NaCl gradient consisting of 0.2 M and 0.4 M NaCl in Tris/HCl with the SAcP eluting in the 0.4 M NaCl eluant. The enzymatically active fraction was concentrated using an Amicon ultrafiltration cell fitted with a 30,000 molecular weight cut-off membrane, washed three times with 10 ml of 0.15 M NaCl in Tris/HCl, pH 8.0, and reduced in volume to 0.5 ml. In order to separate the enzyme from the lower Mr released phosphoglycan contaminant, the preparation was finally passed over a 1 × 30 cm Superdex 200 column (Pharmacia LKB, Uppsala, Sweden) equilibrated with 0.15 M NaCl in Tris/HCl, pH 8.0 where the SAcP lyophilized. Radiolabeling of the SAcP Mid-log phase cells were resuspended in 25 ml of phosphate deficient M199 medium at a concentration of 1.0 × 106/ml; 2 mCi of [32P]-NaH2PO4 was added, and the medium was supplemented with sufficient cold phosphate to bring the concentration to 10% of that found in normal M199. This was found to be the minimum level of phosphate required to allow normal cell growth while still providing maximum incorporation of the radiolabel. The 32P-labeled enzyme was purified as described above. Mild acid deglycosylation of the SAcP The SAcP was deglycosylated using the method of Bates et al. (1990). Briefly, 0.5 ml of 20 mM HCl was added to lyophilized SAcP followed by incubation at 60°C for 60 min. Released sugars were separated from the protein by filtration with a Centricon microconcentrator with a 10,000 Mr cut off. Amino acid and monosaccharide analyses Lyophilized protein samples were hydrolyzed using gaseous HCl at 160°C for 1 h. Free amino acids were subsequently converted to their PTC derivatives in a model 420A derivatizer and analyzed on an Applied Biosystems model 130A amino acid analyzer (Applied Biosystems, Foster City, CA). Sugars, released by the mild acid treatment outlined above, were further hydrolyzed in 2 N HCl at 100°C for 3 h to release monosaccharides. The HCl was removed at reduced pressure and the hydrolyzed sugars were treated with alkaline phosphatase (Sigma, St. Louis, MO) at 37°C in 25 mM Tris/HCl (pH 9.5) with 1 mM MgCl2. This digest was applied to a mixed bed column of Ag-50 (H+) and Ag-3 (OH−) ion exchange resins (1 ml of each resin). The columns were washed with 3 bed volumes (6 ml) of H2O and the effluent was pooled and dried on an Eyela Rotary Evaporator (Tokyo Rikakikai Co. Ltd., Japan). Sugars were analyzed on a CarboPac PA-1 anion exchange column (4 × 250 mm) developed isocratically with 16 mM NaOH and monitored with a pulsed amperometric detector (Dionex Corp., Mississauga, Ontario). Standard curves were prepared using stock solutions of mannose and galactose. Mass spectral analysis of peptides and mild acid released glycans Samples were analyzed on a VG Quattro triple quadrupole mass spectrometer (Micromass, UK) interfaced with an electrospray ionization (ESI) source. The spectrometer was operated in negative ion mode to facilitate the detection of phosphorylated species. Samples were dissolved in 1% NH4OH/50% 2-propanol, and 20 µl aliquots were introduced into a solvent stream of 50% 2-propanol and sprayed at a flow rate of 10 µl/min. The capillary voltage was routinely set at 3 kV, the counterelectrode was set to ∼300 V and the declustering potential varied from 20–50 V depending on the experiment. Horse heart myoglobin (Mr of 16,951) at a concentration of 20 pmol/µl and NaI were used to calibrate the instrument. During MS/MS experiments, the gas in the collision cell was maintained at a level of 5.5 × 10-4 mBar. Release and reduction of SAcP glycans Hydrazinolysis was performed using the method of Takasaki et al. (1982). SAcP (2mg) was dried for 3 days in vacuo over liquid nitrogen, suspended in 2.0 ml of anhydrous hydrazine under dry argon, and heated in a sealed tube at 80°C for 10 h. The hydrazine was removed by evaporation under reduced pressure and the released oligosaccharides were, applied in water to Whatman 3 MM paper, and washed for 2 days with 4:1:1 n-butanol:ethanol:water. The first 6 cm of paper down from the origin was removed, and the oligosaccharides were recovered by elution with water. The sample was then treated with 0.5 ml of 0.1 mM cupric acetate at 27°C for 30 min to remove hydrazide adducts from the reducing termini and passed through a mixed bed column of Chelex 100 (Na+) over Ag-50 (H+) washed with 5 bed volumes of water. The effluent was brought to dryness and the residue dissolved in 100 µl of 50 mM sodium borate pH 11.0. NaB3H4 (2 mCi in 50 µl of 0.05 N NaOH) was then added and the reaction was maintained at 30°C for 4 h. After acidifying with 1 N acetic acid, the reaction mixture was taken to dryness with methanol and deionized by treatment with Dowex 50 (H+). The residue was dissolved in a small amount of water, spotted onto Whatman No. 3 paper, and subjected once again to 4:1:1 paper chromatography as described above. The radioactivity remaining at the origin was eluted with water. Sequencing of the released SAcP glycan Gel permeation Chromatography was performed on a 1 × 100 cm column of Bio-Gel P-4 (200–400 mesh, Bio-Rad, Richmond CA) maintained at 50°C and equilibrated in H2O (or 50 mM ammonium acetate, pH 5.0, for negatively charged oligosaccharides) at a flow rate of 0.2 ml/min. The column was calibrated with a mixture of dextran standards and monitored by differential refractometry. Radioactive samples were applied to the column and 1.0 ml fractions were collected. Five microliters of each fraction was removed for scintillation counting. Dephosphorylation was performed by addition of 50 µl of 50% aqueous HF at −20°C to lyophilized sugar followed by incubation at 0°C for 36 h. The majority of the acid was neutralized using LiOH, and the resulting LiF pellet was removed by centrifugation. Solid NaHCO3 was added to neutralize residual HF. The reaction mixture was then applied to a mixed-bed column of Ag-50 (H+) over Ag-3 (OH−) over QAE (OH−) to remove excess Na+ and any phosphorylated oligosaccharides still remaining. Samples were then treated with E.coli β-galactosidase (Boehringer Mannheim), 250 units/ml in 0.1 M phosphate, pH 7.0, for 48 h. Following both treatments, samples were chromatographed on the P-4 column as indicated above. Thin layer chromatographic separation of phosphorylated amino acids Protein samples were radiolabeled with [32P]-PO4 as indicated above and ∼10,000 c.p.m. of radiolabeled protein was hydrolyzed in the gas phase at 100°C using 6 N HCl for 16 h. HCl was removed under vacuum and the samples were dissolved in H2O and spotted onto precoated cellulose TLC plates (20 × 20 cm) with phosphoserine, phosphothreonine, and phosphotyrosine standards. Plates were developed using 7:2:1 ethyl acetate:formic acid:H2O. Detection of standards was accomplished using ninhydrin spray while 32P labeled amino acids were detected by autoradiography. Standards were marked with [32P]-NaPO4 so that they would also appear on the autoradiogram. Chemical modification of purified SAcP and peptide preparation Reduction of lyophilized SAcP (2 mg) was carried out under N2 in 0.5 ml of 6 M guanidinium HCl, 0.2 M Tris, pH 8.0, containing 0.1 M DTT. The reaction was allowed to proceed for 4 h at room temperature, whereupon 75 µmol of iodoacetic acid was added and reaction was allowed to continue for a further 20 min. The reduced and alkylated SAcP was chromatographed on a 1 × 30 cm Superdex 200 (Pharmacia LKB, Uppsala, Sweden) gel permeation column equilibrated in 20 mM Tris, pH 8.0, 0.15 M NaCl, and finally dialyzed and lyophilized. CNBr digestion of the SAcP was carried out by dissolving 2 mg of reduced and alkylated SAcP in 500 µl of 70% formic acid to which was added a 50-fold molar excess of CNBr in the presence of 35 mg of glycine added to scavenge free sugar aldehydes. The reaction was sparged with nitrogen and allowed to proceed for 16 h at room temperature whereupon it was stopped by dilution and lyophilized. Proteolytic digestion of SAcP was carried out using trypsin, endoproteinase Glu-C and endoproteinase Asp-N individually. In each case, lyophilized SAcP was dissolved in a minimal quantity of appropriate digest buffer (50–100 µl): either 0.1 M Tris, pH 8.0 (trypsin); 0.1 M NaPO4, pH 7.8 (Glu-C); or 0.1 M Tris, pH 8.5 (Asp-N). Enzymes were added to each reaction at a ratio of 50:1 SAcP:protease by weight and digests were allowed to proceed for 16–20 h at 37°C, monitored by either RP-HPLC or capillary electrophoresis. SDS-PAGE and detection of phosphorylated proteins and peptides SDS-PAGE was performed according to Laemmli (1970) using 10% gels. Gels were fixed at room temperature with three 20 min washes of 50% 2-propanol which removed traces of SDS prior to Stains-All (Sigma Chemical Co., St. Louis, MO) staining. Phosphorylated species were visible after five min of staining in this reagent, appearing as blue bands against a pink background. Capillary electrophoresis (CE) CE was performed on a model 270A-HT CE System (Applied Biosystems, Foster City, CA) using a 49 cm capillary. Samples were injected hydrodynamically for 1–3 s with 5 in of pressure and electrophoresed in 50 mM NaPO4, pH 6.8, toward the cathode at 30 kV for 20 min. Protein/peptide elution was monitored at 210 nm. Chromatographic separation of phosphorylated peptides Two milliliters of Macroprep ceramic hydroxylapatite resin (Bio-Rad, Richmond, CA) was packed into a 5 × 50 mm column and equilibrated in 10 mM NaPO4, pH 6.8. Samples were eluted from the column with a gradient of 10–500 mM NaPO4 over 10 min at a flow rate of 1 ml/min using a Beckman HPLC system while collecting 0.5 ml fractions of which 5 µl aliquots were subjected to scintillation counting. Radiolabeled phosphopeptides were subsequently desalted on a Sephadex G-10 column and dried in a Speed Vac concentrator (Savant/EC Instruments Inc., Holbrook, NY). Radioactive peptides were also purified using a MonoQ HR 5/5 anion exchange column (Pharmacia LKB, Uppsala, Sweden) equilibrated with 0.02 M Tris/HCl, pH 8.0, and operated at a flow rate of 1 ml/min. The column was developed with a gradient from 0–500 mM NaCl over 50 min and monitored at 229 nm. Scintillation counting was carried out as above. Protein sequence analysis Protein samples were desalted prior to lyophilization and stored dry at −20°C until required for analysis. N-Terminal sequencing was performed on a model 473A gas phase protein sequencer (Applied Biosystems Inc., Foster City, CA) using standard Edman chemistry. n vitro glycosylation of synthetic peptides Synthetic peptides were prepared with the sequences acetyl-WSSEGTTASSS-amide and NH2-SSSEGTTASSSEGTTASSSamide using a model 430A peptide synthesizer (Applied Biosystems Inc., Foster City, CA) and optimized FastMoc chemistry. The in vitro glycosylation of this molecule was performed as described by Carver et al. (1991, 1992). Briefly, 1 × 1010 L.donovani LD3 promasigotes were suspended in lysis buffer (100 mM HEPES, pH 7.4, 50 mM KCl, 1 mM MnCl2, 1 mM MgCl2, 1 mM TLCK, 1 µg/ml leupeptin) and lysed via N2 cavitation (2400 psi, 20 min). Cell membranes and debris were removed by centrifugation at 1500 × g, and the supernatant was centrifuged at 100,000 × g to collect the microsomal membranes. The pellet was resuspended in 0.5 ml of lysis buffer for use in subsequent assays. The glycosylation reaction was performed by adding 100 µl of the membrane preparation to 250 µl of the above lysis buffer containing 30 mmol GDP-Man, 30 mmol UDP-Gal,, 5 mmol ATP, 1 mmol DTT and 1 mmol of the synthetic peptide substrate. This reaction mixture was allowed to incubate for 16 h at 26°C. After centrifugation at 1500 × g, the supernatant was applied to a C8 reverse phase column (Brownlee, Foster City, CA) equilibrated in 0.1% TFA and developed with a linear isopropanol ramp (0–60% in 40 min). The column effluent was monitored at 230 nm, and peaks were collected manually. These samples were dried in a Speed-Vac concentrator and stored at −20°C until analyzed by MS. Acknowledgments We acknowledge the technical assistance of Sandra Kielland and Darryl Hardie of the Tripartite Microanalytical Center of the University of Victoria and the valuable MS technical advice of Michael Ikonomou at the Institute of Ocean Sciences, Patricia Bay, B°C. We particularly thank Nadja Spitzer for carrying out all the extensive parasite culture associated with this work. This study was supported by the National Sciences and Engineering Research Council of Canada. Appendix Cyanogen bromide fragment containing the Ser/Thr rich region of the L.donovani SAcP. Tryptic cleavage sites are indicated in bold. Taken from Shakarian et al. (1997). GCPRTIADNKPVPSRCWIYRYACPSKACPVTYILSAADH-QCYPGAGIPNSSSSSDVTTASSSEGTTTSSSEGTTASSSDV-TTASSSEGTTASSSEGTTASSSEGTTASSSDVTTASSSEGT-TTSSSEGTTASSSDVTTASSSEGTTASSSEGTTASSSDVTT-ASSSEGITASSSEGTTASSSDVTTASSSEGTTTSSSEGTTAS-SSDVTTASSSEGTTASSSEGSSSSDVSFFKEPASWM Abbreviations Abbreviations SAcP secreted acid phosphatase MAb monoclonal antibody LPG lipophosphoglycan GlcNAc N-acetyl glucosamine DEAE diethylaminoethyl PTC phenylthiocarbamyl MS mass spectrometry QAE quarternary anion exchange TLC thin layer chromatography DTT dithiothreitol CNBr cyanogen bromide RP-HPLC reverse phase high performance liquid chromatography SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis CE capillary electrophoresis CID collision induced dissociation ESI-MS electrospray ionization mass spectrometry MALDI-TOF matrix assisted laser desorption ionization-time of flight References Bates P.A. , Kurtz M.K. , Gottlieb M. , Dwyer D.M. . Leishmania donovani: generation of monospecific antibody reagents to soluble acid phosphatase , Exp. Parasitol. , 1987 , vol. 64 (pg. 157 - 164 ) Google Scholar Crossref Search ADS PubMed WorldCat Bates P.A. , Gottlieb M. , Dwyer D.M. . Leishmania donovani: identification of glycoproteins released by promastigotes during growth in vitro , Exp. Parasitol. , 1988 , vol. 67 (pg. 199 - 209 ) Google Scholar Crossref Search ADS PubMed WorldCat Bates P.A. , Hermes I. , Dwyer D.M. . Golgi-mediated post-translational processing of secretory acid phosphatase by Leishmania donovani promastigotes , Mol. Biochem. Parasitol. , 1990 , vol. 39 (pg. 247 - 255 ) Google Scholar Crossref Search ADS PubMed WorldCat Carver M.A. , Turco S.J. . Cell-free biosynthesis of lipophosphoglycan from Leishmania donovani , J. Biol. Chem. , 1991 , vol. 266 (pg. 10974 - 10981 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Carver M.A. , Turco S.J. . Biosynthesis of lipophosphoglycan from Leishmania donovani: characterization of mannosylphosphate transfer in vitro , Arch. Biochem. Biophys. , 1992 , vol. 295 (pg. 309 - 317 ) Google Scholar Crossref Search ADS PubMed WorldCat Funk V.A. , Thomas-Oates J.E. , Kielland S.L. , Bates P.A. , Olafson R.W. . A unique, terminally glucosylated oligosaccharide is a common feature on Leishmania cell surfaces , Mol. Biochem. Parasitol. , 1997 , vol. 84 (pg. 33 - 48 ) Google Scholar Crossref Search ADS PubMed WorldCat Greis K.D. , Turco S.J. , Thomas J.R. , McConville M.J. , Homans S.W. , Ferguson M.A.J. . Purification and characterization of an extracellular phosphoglycan from Leishmania donovani , J. Biol. Chem. , 1992 , vol. 267 (pg. 5876 - 5881 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Gustafson G.L. , Milner L.A. . Occurrence of N-acetylglucosamine-1-phosphate in proteinase I from Dictyostelium discoideum , J. Biol. Chem. , 1980 , vol. 255 (pg. 7208 - 7210 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Haynes P.A. , Ferguson M.J. , Cross G.A.M. . Structural characterization of novel oligosaccharides of cell-surface glycoproteins of Trypanosoma cruzi , Glycobiology , 1996 , vol. 6 (pg. 869 - 878 ) Google Scholar Crossref Search ADS PubMed WorldCat Ilg T. , Harbecke D. , Wiese M. , Overath P. . Monoclonal antibodies directed against Leishmania secreted acid phosphatase and lipophosphoglycan. Partial characterization of private and public epitopes , Eur. J. Biochem. , 1993 , vol. 217(2) (pg. 603 - 615 ) Google Scholar Crossref Search ADS WorldCat Ilg T. , Overath P. , Ferguson M.A.J. , Rutherford T. , Campbell D.G. , McConville M. J. . O-and N-glycosylation of the Leishmania mexicana secreted acid phosphatase , J. Biol. Chem. , 1994 , vol. 269 (pg. 24073 - 24081 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Jaffe C.L. , Perez M.L. , Schnur L.F. . Lipophosphoglycan and secreted acid phosphatase of Leishmania tropica share species-specific epitopes , Mol. Biochem. Parasitol. , 1990 , vol. 41 (pg. 233 - 240 ) Google Scholar Crossref Search ADS PubMed WorldCat Jardim A. , Funk V. , Caprioli R.M. , Olafson R.W. . Isolation and structural characterization of the Leishmania donovani kinetoplastid membrane protein-11, a major immunoreactive membrane glycoprotein , Biochem. J. , 1995 , vol. 305 (pg. 307 - 313 ) Google Scholar Crossref Search ADS PubMed WorldCat Laemmli U.K. . Cleavage of structural proteins during the assembly of the head of bacteriophage T4 , Nature , 1970 , vol. 227(259) (pg. 680 - 685 ) Google Scholar Crossref Search ADS WorldCat Martensen T.M. . Chemical properties, isolation, and analysis of O-phosphates in proteins , Methods Enzymol. , 1984 , vol. 107 (pg. 3 - 23 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat McConville M.J. , Thomas-Oates J.E. , Ferguson M.A.J. , Homans S.W. . Structure of the lipophosphoglycan from Leishmania major , J. Biol. Chem. , 1990 , vol. 265 (pg. 19611 - 19623 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat McConville M.J. , Turco S.J. , Ferguson M.A.J. , Sacks D.L. . Developmental modification of lipophosphoglycan during the differentiation of Leishmania major promastigotes to an infectious stage , EMBO J. , 1992 , vol. 11 (pg. 3593 - 3600 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Olafson R.W. , Thomas J.R. , Ferguson M.A. , Dwek R.A. , Chaudhuri M. , Chang K.P. , Rademacher T.W. . Structures of the N-linked oligosaccharides of GP63, the major surface glycoprotein, from Leishmania mexicana amazonensis , J. Biol. Chem. , 1990 , vol. 265(21) (pg. 12240 - 12247 ) OpenURL Placeholder Text WorldCat Sacks D.L. , Turco S.J. . Developmental modification of lipophosphoglycan from Leishmania major promastigotes during metacyclogenesis , Mol. Biochem. Parasitol. , 1990 , vol. 42 (pg. 225 - 234 ) Google Scholar Crossref Search ADS PubMed WorldCat Shakarian A.M. , Ellis S.L. , Mallinson D.J. , Olafson R.W. , Dwyer D.M. . Two tandemly arrayed genes encode the (histidine) secretory acid phosphatases of Leishmania donovani , Gene. , 1997 , vol. 196 (1-2) (pg. 127 - 137 ) Google Scholar Crossref Search ADS PubMed WorldCat Takasaki S. , Mizuochi T. , Kobata A. . Hydrazinolysis of asparaginelinked sugar chains to produce free oligosaccharides , Methods Enzymol. , 1982 , vol. 83 (pg. 263 - 268 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Thomas J.R. , McConville M.J. , Thomas-Oates J.E. , Homans S.W. , Fergu-son M.A.J. , Gorin P. , Greis K.D. , Turco S.J. . Refined structure of the lipophosphoglycan of Leishmania donovani , J. Biol. Chem. , 1992 , vol. 267 (pg. 6829 - 6833 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Turco S.J. , Hull S.R. , Orlandi P.A.^Jr. , Shepherd S.D. , Homans S.W. , Dwek R.A. , Rademacher T.W. . Structure of the major carbohydrate fragment of the Leishmania donovani lipophosphoglycan , Biochemistry , 1987 , vol. 26(19) (pg. 6233 - 6238 ) Google Scholar Crossref Search ADS WorldCat Wiese M. , Ilg T. , Lottspeich F. , Overath P. . Ser/Thr-rich repetetive motifs as targets for phosphoglycan modifications in Leishmania mexicana secreted acid phosphatase , EMBO J. , 1995 , vol. 14 (pg. 1067 - 1074 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat © 1999 Oxford University Press
TI - Phosphoglycosylation of a secreted acid phosphatase from Leishmania donovani
JO - Glycobiology
DO - 10.1093/glycob/9.6.627
DA - 1999-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/phosphoglycosylation-of-a-secreted-acid-phosphatase-from-leishmania-eTWZLapSba
SP - 627
EP - 636
VL - 9
IS - 6
DP - DeepDyve
ER -