Structural analysis of the N-glycans of the major cysteine proteinase of Trypanosoma cruzi Identification of sulfated high-mannose type oligosaccharides Mariana Barboza1, Vilma G Duschak2, Yuko Fukuyama3,*, Hiroshi Nonami3, Rosa Erra-Balsells4, Juan J Cazzulo1 and Alicia S Couto4 ´ Instituto de Investigaciones Biotecnologicas-INTECH, Universidad Nacional de Gral San Martin, Buenos Aires, Argentina ´ ´ Instituto Nacional de Parasitologıa ‘Dr Mario Fatala Chaben’, ANLIS, Ministerio de Salud y Ambiente, Buenos Aires, Argentina College of Agriculture, Ehime University, Matsuyama, Japan ´nica, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, ´ CIHIDECAR (CONICET) Departamento de Quımica Orga Argentina Keywords Cruzipain; nor-harmane; sulfated oligosaccharides; Trypanosoma cruzi; UVMALDI-TOF MS Correspondence A S Couto, CIHIDECAR (CONICET) ´nica, ´ Departamento de Quımica Orga Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Buenos Aires, CP 1428, Argentina Fax: +54 11 45763346 Tel: +54 11 45763346 E-mail: acouto@qo.fcen.uba.ar *Present address Koichi Tanaka Mass Spectrometry Research Laboratory, Shimadzu Corporation, Nishinokyo-Kuwabaracho, Nakagyo-ku, Kyoto 604-8511, Japan (Received April 2005, accepted 23 May 2005) doi:10.1111/j.1742-4658.2005.04787.x Trypanosoma cruzi, the parasitic protozoan that causes Chagas disease, contains a major cysteine proteinase, cruzipain This lysosomal enzyme bears an unusual C-terminal extension that contains a number of posttranslational modifications, and most antibodies in natural and experimental infections are directed against it In this report we took advantage of UV-MALDI-TOF mass spectrometry in conjunction with peptide N-glycosidase F deglycosylation and high performance anion exchange chromatography analysis to address the structure of the N-linked oligosaccharides present in this domain The UV-MALDI-TOF MS analysis in the negativeion mode, using nor-harmane as matrix, allowed us to determine a new striking feature in cruzipain: sulfated high-mannose type oligosaccharides Sulfated GlcNAc2Man3 to GlcNAc2Man9 species were identified In accordance, after chemical or enzymatic desulfation, the corresponding signals disappeared In addition, by UV-MALDI-TOF MS analysis (a) a main population of high-mannose type oligosaccharides was shown in the positive-ion mode, (b) lactosaminic glycans were also identified, among them, structures corresponding to monosialylated species were detected, and (c) as an interesting fact a fucosylated oligosaccharide was also detected The presence of the deoxy sugar was further confirmed by high performance anion exchange chromatography In conclusion, the total number of oligosaccharides occurring in cruzipain was shown to be much higher than previous estimates This constitutes the first report on the presence of sulfated glycoproteins in Trypanosomatids Trypanosoma cruzi, the parasitic protozoan that causes the American Trypanosomiasis or Chagas disease, contains a major cysteine proteinase (CP), cruzipain This enzyme is present in the epimastigote, amastigote, metacyclic and tissue culture trypomastigote forms [1] It has been reported to be placed in the lysosomal compartment [2–4], but it seems to be also located at the cell surface Accordingly, plasma membrane-bound isoform(s) of CPs have been shown in the different developmental stages of T cruzi [5] Cruzipain is encoded by numerous genes which contain no introns, encoding a signal peptide, a propeptide and a mature enzyme As for all Type I CPs from trypanosomatids, the protein presents a catalytic Abbreviations CP, cysteine proteinase; HPAEC-PAD, high pH anion exchange chromatography with pulsed amperometric detection; PNGase F, peptide N-glycosidase F FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS 3803 Structure of the N-glycans in cruzipain moiety and a characteristic C-terminal domain [6], the latter being the most characteristic structural feature of this protein [7] As in the CPs of Trypanosoma rangeli and Crithidia fasciculata but in contrast to similar CPs from Leishmania mexicana and Trypanosoma brucei, cruzipain C-terminal is retained in the natural mature form of the enzyme [8] This extension is 130 aminoacid residues long [9], and most antibodies in natural and experimental infections are directed against it [10] Cruzipain is purified from epimastigotes as a complex mixture of isoforms [8] This microheterogeneity is probably due to the presence of a number of posttranslational modifications [11] including carbohydrate heterogeneity [12] as well as some point mutations leading to amino-acid replacements, most if not all present in the C-terminal domain [9,11] It is known that the single N-glycosylation site in the C-terminal domain (Asn255) as well as the first potential N-glycosylation site in the catalytic moiety (Asn33) are glycosylated in vivo [13]; the latter bears only highmannose type oligosaccharides It had been suggested that the C-terminal domain of cruzipain presents either a high-mannose oligosaccharide, a hybrid monoantennary or a complex biantennary oligosaccharide chain [12] In vivo labeling of the parasites with 32Pi discounted the presence of Pi in the mature enzyme [11] Recently, two types of post-translational modifications involving carbohydrates have been described: a complex N-glycosidic oligosaccharide bearing sialic acid and single N-acetyl-glucosamine residues with an O-glycosidic linkage [14] The fact that cruzipain is a complex mixture of isoforms with a great diversity in the N-linked structures made the separation of species very difficult We took advantage of UV-MALDI-TOF mass spectrometry in combination with enzymatic digestions and complemented with high pH anion exchange chromatography (HPAEC) analysis to characterize the N-linked glycans present in the protein In this paper we report, for the first time, the presence of sulfate in the N-linked oligosaccharides of cruzipain This structural feature was confirmed in the unique N-glycosidic site present in the C-terminal domain; a major population of high-mannose type oligosaccharides, as well as lactosaminic, fucosylated and sialylated complex type glycans in a minor extent, were shown The diversity of structures present in the C-terminal domain might account for the microheterogeneities found in natural cruzipain Results In an attempt to perform a structural study of the oligosaccharide chains localized in the C-terminal 3804 M Barboza et al 66 45 29 kDa 10 11 Fig Analysis by SDS ⁄ PAGE followed by silver stainning of C-terminal domain purification Lane 1, natural cruzipain; lane 2, self-proteolysed cruzipain; lanes 3–11 are fractions corresponding to the Bio Gel P-30 columm Lanes 6–11 correspond to purified C-terminal Molecular mass markers are indicated (in kDa) at the left side of the figure domain of cruzipain, the protein was subjected to selfproteolysis and the C-terminal was further purified via Bio Gel P-30 column chromatography [14] Fractions containing purified C-terminal (Fig 1, lanes 8–11) were joined, freeze-dried and subjected to peptide N-glycosidase F (PNGase F) treatment The released oligosaccharides were separated from the polypeptide by Ultrafree McFilters (MW 5000) A fraction of these oligosaccharides was reduced with NaB3H4 and desalted by Biogel P-2 as already described [23] Labeled oligosaccharides included in the column corresponded to neutral N-linked oligosaccharides as shown by HPAEC (Fig 2A) The acidic glycans already reported [14] were recovered in the excluded fraction Under conditions where acidic glycans are resolved, the HPAEC profile of the excluded fraction showed four major peaks (Rt ¼ 8, 13, 25 and 32.5 min) (Fig 2B) Mild acid hydrolysis of this fraction to release sialic acid and further analysis showed the absence of the peak with Rt ¼ 25 (Fig 2C) When this desialylated fraction was analysed under conditions where neutral glycans are resolved, a peak coincident with a standard of a biantennary complex-type oligosaccharide was obtained (Fig 2E) The fact that not all the acidic glycans were sensitive to the mild acid hydrolysis strongly suggested that another acidic group could be present Thus, a digestion with sulfatase was performed and the profile obtained by HPAEC (Fig 2D) showed the disappearance of the peak with Rt ¼ 32.5 More evidence of the presence of sulfated species in cruzipain was achieved using anion-exchange chromatography [24] A sample of the labeled oligosacFEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS M Barboza et al Fig HPAEC analysis of the oligosaccharides released by PNGase F treatment from the C-terminal domain of cruzipain (A) Total neutral fraction; (B) total acidic fraction; (C) acidic fraction treated with mild acid to release sialic acid; (D) acidic fraction treated with sulfatase; (E) same as (C) (A) and (E), analysis under ‘conditions a’ for neutral glycan; (B), (C) and (D), analysis under ‘conditions b’ for acidic glycans Standards: 1, Man3GlcNAc2OH; 2, Man5GlcNAc2OH; 3, Man6GlcNAc2OH; 4, Gal2GlcNAc2Man3GlcNAc2OH; 5, Man9GlcNAc2OH; 6, monosialylated; 7, disialylated; 8, trisialylated oligosaccharides charides (12 000 cpm) was applied to a QAE-Sephadex column equilibrated with mm Tris-base Although a major fraction (9000 cpm) passed through, 25% (3000 cpm) of the label bound to the resin and was eluted with mm Tris ⁄ 20 mm NaCl correlating with the presence of another acidic group in addition to sialic acid To determine the structural identity of the N-glycans present in cruzipain, a UV-MALDI-TOF MS analysis was performed Figure 3A shows the positive UV-MALDI spectrum of the whole oligosaccharide fraction released from cruzipain by PNGase F digestion Figure 3B shows the spectrum corresponding to the analogous fraction obtained from the purified C-terminal domain The m ⁄ z-value, composition and structure of the detected glycans are listed in Tables and Molecular ions were determined as monosodium adducts [M + Na]+ Although the quality of the spectra (signal ⁄ noise ratio) obtained for each sample was different, both spectra showed major peaks at m ⁄ z-values: 933.2 (933.7), 1094.8 (1096.1), 1259.1 (1258.3), 1420.4 (1420.4), 1581.7 (1582.7), 1744.5 (1744.8) and 1906.6 (1907.6) (Table 1, Fig 3) These signals correspond to high-mannose glycans containing from to mannose residues (GlcNAc2Man3 to GlcNAc2Man9) (Table 1) Interestingly, signals at m ⁄ zvalues 2029.5 (2029.9), 2069.1 (2069.3) and 2395.4 (2395.1) compatible with lactosaminic type oligosaccharides, not reported so far as components of cruzipain, were found in both spectra (Table 1, Fig 3) The C-terminal spectrum also showed signals with m ⁄ z-values 1013.9; 1176.4; 1338.7; 1500.6; 1663.6 and 1825.1 (Fig 3B) compatible with sulfated high-mannose species (Table 2) Some of these peaks were also observed in the cruzipain spectrum In addition, a signal at m ⁄ z 1809.6 was compatible with a fucosylated oligosaccharide (Fig 3B, Table 1) As this deoxy-sugar had not been previously identified as component of cruzipain, its presence was investigated by HPAECPAD analysis The sugar was released from the C-terminal domain by specific a-l-fucosidase treatment, separated through Ultrafree McFilters (MW 5000), labeled by reduction with NaB3H4 and analysed under FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS Structure of the N-glycans in cruzipain A 400 23 CPM 300 200 100 B 500 CPM 400 300 200 100 C 500 CPM 400 300 200 100 D 500 CPM 400 300 200 100 E 425 CPM 325 225 125 10 15 20 25 Retention time (min) 30 35 40 3805 Structure of the N-glycans in cruzipain M Barboza et al A 100 1906.6 1744.5 90 80 1581.7 Relative abundance 70 933.2 1013.9 1420.4 60 1094.8 1541.6 1259.1 50 1217.0 1379.2 1923.2 1703.5 1338.7 40 1663.5 1843.1 2069.1 2029.5 30 2395.4 20 10 602 976 B 100 1350 2099 1725 2474 1258.3 90 1096.1 80 1420.4 Relative abundance 933.7 2029.9 70 1968.6 1977.8 2069.3 2395.1 60 1582.7 50 40 8X 1744.8 * 30 1500.6 1338.7 20 1054.8 1013.9 1176.4 1217.0 1200.0 1809.6 1663.6 1379.7 10 1825.1 881 1199 1517 1907.6 1815.1 1835 1968.6 2029.9 2069.3 2395.1 2153 2471 m/z Fig UV-MALDI-TOF MS analysis of the released oligosaccharides in the lineal positive mode using GA as matrix (A) oligosaccharides obtained from cruzipain, m ⁄ z range: 600–2474 Da (B) oligosaccharides obtained from C-terminal, m ⁄ z range: 881–2471 Da Inset corresponds to expanded m ⁄ z range: 1950–2460 Da Structures are detailed in Tables and conditions c (Fig 4) The same peak was obtained when fucose was released by acid hydrolysis (not shown) Also minor signals at m ⁄ z 1815.1 and 1977.8 corresponding to biantennary sialylated oligosaccharides were detected in the C-terminal spectrum (Fig 3B, Table 1) 3806 b-Carbolines have proven to be effective matrices for the detection of sulfated carbohydrates in the negative ion mode [20–22] That is the reason why we carried out a UV-MALDI-TOF MS analysis using nor-harmane as matrix to confirm the presence of sulfated species The spectrum of the whole oligosaccharide fraction obtained FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS M Barboza et al Structure of the N-glycans in cruzipain Table m ⁄ z-Value, composition and structure of the high mannose and complex type glycans of cruzipain and C-terminal domain h, N-Acetylglucosamine; d, mannose; , galactose; b, sialic acid; ,, fucose Calculated m ⁄ z a [M+Na]+ Measured m ⁄ z b [M+Na]+ Proposed composition 933.31 933.78 1095.37 1095.92 1257.42 1258.06 933.2 933.7 1094.8 1096.1 1259.1 1258.3 HexNAc2Man3 1419.47 1420.21 1420.4 1420.4 HexNAc2Man6 1581.53 1582.35 1581.7 1582.7 HexNAc2Man7 1743.58 1744.49 1744.5 1744.8 HexNAc2Man8 1905.63 1906.63 1906.6 1907.6 HexNAc2Man9 1809.66 1810.56 – 1809.6 HexNAc4Man3Gal2Fuc 1815.61 1816.53 – 1815.1 HexNAc4Man3Gal1SA1 1977.67 1978.67 – 1977.8 HexNAc4Man3Gal2SA1 2028.71 2029.73 2029.5 2029.9 HexNAc5Man3Gal3 2069.74 2070.78 2069.1 2069.3 HexNAc6Man3Gal2 2393.85 2395.06 2395.4 2395.1 HexNAc6Man3Gal4 Structure HexNAc2Man4 HexNAc2Man5 a Upper number indicates the average mass; lower number indicates the monoisotropic mass b Upper number indicates the m ⁄ z value obtained for oligosaccharides from cruzipain; lower number indicates the m ⁄ z value obtained for oligosaccharides from the purified C-T domain from the C-terminal domain in the negative ion mode is shown in Fig 5A Signals with m ⁄ z 990.0, 1152.8, 1314.1, 1476.6, 1639.4, 1801.5 and 1963.7 were assigned to sulfated high-mannose oligosaccharides as [M-H]– ions (Table 2) No signals corresponding to neutral oligosaccharides were detected Noticeably, the major population corresponded to signals with m ⁄ z 1007.6, 1170.5, 1332.4, 1494.7, 1656.9, 1819.1 and 1981.5 compatible with [M + H2O-H]– ions of the same glycans (Table 2; Fig 5A) Similarly, cationic adducts attaching FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS water in the positive ion mode were detected among C-terminal oligosaccharides using GA as matrix (Fig 3B, peaks at: m ⁄ z 1054.8, 1217.0 and 1379.7; Table 2) In order to discard the complex nature of the acidic oligosaccharides, treatment of another sample of the C-terminal oligosaccharides with endo-b-galactosidase was performed The spectra obtained (Fig 5B), showed a pattern of signals similar to that of the untreated sample (Fig 5A), assuring the high-mannose nature of the sulfated glycans 3807 3808 1176.4 1338.7 1500.6 1663.6 1825.1 1175.9 1338.1 1500.2 1662.4 1824.5 a 1980.7 1865.5 1703.4 1541.2 1379.1 1216.9 1054.8 Calculated m ⁄ z a [M+H2O+Na]+ Sulfates were present as the sodium salt – 1013.9 1013.8 1961.5 Measured m ⁄ z [M+Na]+ Calculated m ⁄ z [M+Na]+ – – 1703.5 1541.6 1379.7 1217.0 1054.8 Measured m ⁄ z a [M+H2O+Na]+ 1800.5 HexNAc2Man8 + SO4 1962.7 1638.4 HexNAc2Man7 + SO4 HexNAc2Man9 + SO4 1476.2 HexNAc2Man6 + SO4 1314.1 HexNAc2Man5 + SO4 989.8 Calculated m ⁄ z [M–H]– 1151.9 Structure HexNAc2Man4 + SO4 HexNAc2Man3 + SO4 Proposed composition 1963.7 1801.5 1639.4 1476.6 1314.1 1152.8 990.0 Measured m ⁄ z [M–H]– 1980.7 1818.5 1656.4 1494.2 1332.1 1170.0 1007.0 Calculated m ⁄ z [M+H2O–H]– Table m ⁄ z-Value, composition and structure of sulfated oligosaccharides of cruzipain and C-terminal domain h, N-Acetylglucosamine; d, mannose ; q -SO4 1981.5 1819.1 1656.9 1494.7 1332.4 1170.5 1007.6 Measured m ⁄ z [M+H2O–H]– Structure of the N-glycans in cruzipain M Barboza et al FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS M Barboza et al Structure of the N-glycans in cruzipain A 25000 20000 15000 Radioactivity (CPM) 10000 5000 12 16 B 25000 20000 15000 m ⁄ z 1656.2 assigned to sulfated HexNAc2Man7 in relation with signal at m ⁄ z 1310.1 decreased more than 60% in the solvolysis treated sample and completely disappeared after the enzymatic treatment (Fig 6B) It is interesting to point out that although the background signals shown in Fig 6B were not assigned, they were used as internal reference to express relative abundance of the 1493.9, 1656.2, 1818.5 and 1980.3 peaks In addition, the released sulfate ion was identified by ion chromatography using conductivity detection (Fig 6D, as an inset in Fig 6B) Altogether, these data confirm the presence of sulfated high-mannose type oligosaccharides in cruzipain and in its C-terminal domain This is the first report of the use of nor-harmane as matrix for structural characterization of N-linked oligosaccharides of glycoproteins 10000 Discussion 5000 12 16 Time (min) Fig HPAEC analysis of the a-L-fucose released from the C-terminal domain of cruzipain (A) Purified C-terminal domain was digested with a-L-fucosidase The released sugar was labeled by reduction with NaB3H4 desalted and analysed by HPAEC under conditions c (B) Same as (A) without enzyme Standards: 1, fucitol; 2, sorbitol In another experiment, oligosaccharides obtained from cruzipain were analysed using nor-harmane as matrix in the negative ion mode (Fig 6) Accordingly, signals corresponding to sulfated glycans compatible with [M + H2O-H]– ions were shown (m ⁄ z 1170.6, 1331.8, 1493.9, 1656.2, 1818.5 and 1980.3) (Table 2, Fig 6A) The corresponding adducts in the positive ion mode, [M + H2O + Na]+, were also detected when cruzipain oligosaccharides were analysed using GA as matrix (Fig 3A, peaks at m ⁄ z 1217.0, 1379.2, 1541.6 and 1703.5; Table 2) In order to confirm the nature of the substitution present in glycans obtained from cruzipain, desulfation was carried out using sulfatase The UV-MALDI-TOF spectra of the treated sample, obtained in the negative ion mode using nor-harmane as matrix, showed the complete disappearance of the aforementioned signals (Fig 6B) Furthermore, when desulfation of the released oligosaccharides was performed by solvolysis, although treatment was not complete a significant reduction of those signals was observed (Fig 6C) It should be noted that the relative abundance of the major peak at FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS In this study we have examined the structures of the N-linked oligosaccharides present in cruzipain HPAEC analysis of the total neutral oligosaccharide fraction obtained from the C-terminal domain showed a mixture of different types of carbohydrate chains as previously reported [12,14] However, when the acidic glycans were analyzed the results obtained pointed to the presence of sulfated oligosaccharides in addition to the sialylated glycans previously reported [14] The fact that part of the labeled oligosaccharides were bound to a QAE-Sephadex column and eluted with mm Tris-base ⁄ 20 mm NaCl also agreed with the presence of sulfate groups in this fraction [24] UV-MALDI-TOF MS analysis demonstrated that the overall glycosylation pattern of cruzipain is characterized by a remarkable structural diversity There is no doubt that oligosaccharides of cruzipain are mainly of the high-mannose type However, it is interesting to note that lactosaminic glycans bearing from to lactosamine units are also present, a feature that had not been detected before as component of T cruzi glycoproteins The fact that these signals were found in the C-terminal UV-MALDI-TOF mass spectra confirms their localization in this domain as previously suggested [12] Taking into account that cruzipain bears sialic acid in N-linked structures [14] which can only be acquired at the cell surface through the action of trans-sialidase [25], the finding that polylactosaminic units are also present in this enzyme triggers the possibility that cruzipain would be transported via an endocytic recycling and ⁄ or lysosomal transport pathway as proposed for T brucei [26,27] 3809 Structure of the N-glycans in cruzipain A 100 M Barboza et al 1494.7 1656.9 Relative abundance (%) 1332.4 1819.1 1476.6 1170.5 1639.4 1314.1 50 1801.5 1007.6 1152.8 990.0 1981.5 1963.7 Relative abundance (%) B 100 1494.3 1332.1 50 1152.1 1312.1 1172.1 1656.8 1476.2 1818.3 1639.3 1800.9 1981.1 1962.1 800 1200 1600 2000 Mass (m/z) In addition, a signal compatible with a fucosylated structure was also found in the C-terminal domain The existence of the a-l-fucose unit in C-terminal was further supported by HPAEC analysis Up to now, two surface glycoproteins of T cruzi have been described containing a-l-fucose in their structure: Gp-72, isolated from epimastigote forms [28,29] and the trypomastigote stage specific glycoprotein belonging to the Tc-85 family [30] However, in T cruzi, no fucosyl transferase has been reported so far Although their low abundances, signals corresponding to monosialylated oligosaccharides could be detected in the positive-ion mode without derivatization using GA as matrix compatible with [M + Na]+ adducts The main proposed structure was assigned to a biantennary complex oligosaccharide, in accordance with the results obtained by HPAEC However, relative abundances of these peaks showed variations from batch to batch It is known that a serious limitation in the study of sulfated oligosaccharides are the few reliable analytical methods of structural characterization [31] For that reason, in the last years, the development of new matrices have made UV-MALDI-TOF MS a suitable tool for 3810 2400 2800 Fig UV-MALDI-TOF MS analysis of the oligosaccharides released from C-terminal domain by PNGase F treatment in the linear negative-ion mode using nor-harmane as matrix (A) Analysis of the total oligosaccharide fraction, m ⁄ z range: 800–2800 Da (B) Same as (A) after endo-b-galactosidase treatment Structures are detailed in Table their analysis [20,32,33] Nor-harmane was optimal for the analysis of sulfated N-linked oligosaccharides in negative ion mode because it provided not only good detection sensitivity, but also no interference with the matrix adduct ions The conditions used allowed the production of intense signals without any desulfation Signals corresponding to sulfated GlcNAc2Man3 to GlcNAc2Man9 glycans were identified Interestingly, the major signals were attributed to [M + H2O-H]– adducts The water adducts (M + H2O) were also found in the positive ion mode cationized by sodium (M + H2O + Na)+ using either GA or nor-harmane as matrix, in different samples (oligosaccharides from C-terminal domain or from the whole protein) and using different equipment Therefore, the retention of water can be explained taking into account its strong interaction with the negative charge site of the sulfated analytes [33,34] The fact that the resulting signals were totally resistant to endo-b-galactosidase digestion allowed us to discard the complex structure of the sulfated species On the other hand, considering that when the core glycan structure is substituted, the modifications are present on the NAc-glucosamine unit, the detection of the signals at m ⁄ z-value 990 and 1007.6 in the C-terminal spectrum FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS M Barboza et al Structure of the N-glycans in cruzipain 1656.2 A 100 90 Relative abundance 80 1818.5 70 60 50 1493.9 40 30 1331.8 1170.6 1310.8 20 1980.3 1598.6 1893.6 10 1000 1200 1400 1600 1800 2000 2200 2400 B 100 90 D 60 50 40 30 1310.1 1195.8 20 10 1000 1200 1400 1531.0 1622.0 1750.3 1600 1800 SO4 4.18 70 PO4 Conductivity(µs) Relative abundance 80 Time (min) 2000 2200 2400 2200 2400 C 100 90 Fig UV-MALDI-TOF MS analysis of the oligosaccharides released from cruzipain domain by PNGase F treatment in the linear negative-ion mode using nor-harmane as matrix (A) Analysis of the total oligosaccharide fraction, m ⁄ z range: 900–2500 Da (B) Same as (A) after sulfatase treatment (C) Same as (A) after solvolysis (D) Ion chromatography analysis of sulfate released from oligosaccharides obtained from cruzipain Relative abundance 80 70 60 1656.9 50 40 1310.1 1195.8 30 1818.5 1893.6 20 10 1000 and the growing sulfated high-mannose series suggest that the sulfate group should be located on the chitobiosyl core (Fig 5A) The presence of sulfate groups in N-linked oligosaccharides has been reported in virus [35] and especially in mammalian cells [36–40] However, these reports are mostly based on the results of radioisotope labeling and there are only a few reports on the detailed structure of these sulfated glycans Such oligosaccharides usually FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS 1493.9 1598.6 1200 1400 1600 1800 m/z 2000 sulfated on galactose, mannose, N-acetyl galactosamine, N-acetyl glucosamine or glucuronic acid residues have been implicated in several specific molecular recognition processes [41,42] In T cruzi, sulfated structures have been described as part of glycolipids [43,44] The present study constitutes the first report on the presence of sulfated oligosaccharides in glycoproteins of T cruzi Sulfated high-mannose type glycans have only been described as component of glycoproteins from 3811 Structure of the N-glycans in cruzipain Dictyostelium discoideum [24] Likewise, these glycoproteins are localized in lysosome, however, Man6-SO4 accounts for the majority of the sulfated sugar In conclusion, the results obtained provide evidence of the nature of the glycans present in the unique N-glycosylation site present in the C-terminal domain of cruzipain (Asn255) This site is mainly occupied by neutral or sulfated high-mannose type oligosaccharides To a minor extent, biantennary lactosaminic chains, some of them bearing sialic acid, or fucose are also present The finding of sulfated glycans indicates the activity of a sulfotransferase which has not been described in T cruzi to date In the present study, the precise location of the sulfate group and its biological significance remain to be established Studies to address these questions are currently in progress in our laboratory Experimental procedures Materials All solvents used were of analytical or HPLC grade Ultra free-MC centrifugal filter units Amicon Bio separations were from Millipore Corporation (Bedford, MA, USA) Radioactivity was determined in a 1214 Rackbeta Wallac liquid scintillation counter using Optiphase’Hisafe scintillation cocktail (LKB) Lowry’s method [15] was used to quantify protein content Neutral glycan standards were obtained from Oxford Glyco System (Abingdon, UK) and sialylated oligostandards obtained from fetuin were from Dionex Corporation (Dionex Corporation, Sunnyvale, CA, USA) All chemicals used in UV-MALDI-TOF MS analysis were ACS grade or higher Purification of the C-terminal domain Highly purified cruzipain was obtained by a procedure using chromatography on Con-A Sepharose and Mono Q [8]; fractions strongly bound to the anionic resin, eluted with 0.25–0.50 m NaCl were used The C-terminal domain was obtained by self-proteolysis of cruzipain in sodium acetate buffer pH 6.0 at 40 °C for 48 h The C-terminal domain was purified by gel filtration in a Bio Gel P-30 column (1.5 · 100 cm) eluted with Tris ⁄ HCl buffer pH 7.6 containing 50 mm NaCl Fractions of mL were collected and monitored by measuring UV absorption at 280 ⁄ 230 nm A sample of each fraction was analysed by SDS ⁄ PAGE followed by silver staining or electroblotting, developing with anticruzipain polyclonal antibody [14] Mild acid hydrolysis Sialic acid was hydrolysed with 0.01 m HCl for 20 at 100 °C and freeze-dried For fucose analysis hydrolysis was 3812 M Barboza et al performed with 0.1 m HCl for h at 100 °C Samples were freeze dried, dissolved in water (0.5 mL), the solution was adjusted to pH and labeled with NaB3H4 (0.12 mCi) for h at room temperature Reduction was completed with NaBH4 for h more and the reaction was stopped by addition of acetic acid to pH Solvolysis Samples were passed over 0.5 mL of AG50W-X8 resin (H+) and the column was washed with water (2 mL) Pyridine (0.015 mL) was added to the sample, which was then lyophilized, dissolved in dimethylsulfoxide ⁄ methanol (9 : 1, v ⁄ v; 0.2 mL), adjusted to pH with dilute HCl, heated at 100 °C for h and freeze-dried [16] High pH anion exchange chromatography (HPAEC) For HPAEC analysis the released monosaccharides or oligosaccharides were labelled by reduction with NaB3H4 Boric acid was removed by repeated coevaporations with methanol and the labelled oligosaccharides were desalted by passage through a Bio-Gel P-2 column [14] A DX-300 Dionex BioLC system (Dionex Corporation) with a pulse amperometric detector was used The following columns and conditions were employed: (a) Carbopack PA-100 column equipped with a PA-100 precolumn; gradient elution with 50 mm NaOH and 0–50 mm sodium acetate during 40 The flow rate was 0.6 mLỈmin)1; (b) Carbopack PA-100 column equipped with a PA-100 precolumn; isocratic elution with 100 mm NaOH ⁄ 50 mm sodium acetate for min, followed by a gradient elution with 100 mm NaOH and 50–170 mm sodium acetate during 60 The flow rate was mLỈmin)1; and (c) Carbopack MA-1 column equipped with a MA-1 precolumn and an isocratic elution with 25% solution A (NaOH 200 mm), 75% solution B (water) The flow rate was 0.4 mLỈmin)1 Ion chromatography analysis was performed on a Dionex AS4A column using 1.8 mm Na2CO3 ⁄ 1.7 mm NaHCO3 as eluent, with postcolumn in-line anion micromembrane suppression and conductivity detection The flow rate was mLỈmin)1 [17] Enzymatic digestions PNGase F digestion was performed in 10 mm Tris ⁄ HCl buffer pH 8.3, containing PNGase F (New England Biolabs Inc., Beverly, MA, USA) (15 mU) The oligosaccharides were separated from the protein by Ultrafree McFilters (MW 5000) Endo-b-d-galactosidase digestion was performed in 50 mm sodium acetate pH 6.0 with mU of B fragilis FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS M Barboza et al endo-b-d-galactosidase (Oxford Glyco System, Abingdon, UK) containing 250 lgỈmL)1 bovine serum albumine and mm NaCl a-l-Fucosidase digestion was performed in 30 mm sodium phosphate pH 5.6 containing 14 mU of b-l-fucosidase from bovine kidney (Sigma-Aldrich Co., St Louis, MO, USA) Sulfatase digestion was performed in 50 mm sodium acetate pH 5.0 with sulfatase from Abalone entrailis (Type VII Sigma-Aldrich Co., St Louis, MO, USA) (25 mU) All digestions were performed for 18 h at 37 °C Before UV MALDI-TOF MS analysis, the oligosaccharides were treated with 150 mm acetic acid for h at room temperature [18] and desalted through a microcolumnn containing AG-50 (H+) [19] UV-MALDI-TOF mass spectrometry Measurements were performed with an Applied Biosystems Voyager DE-STR (Applied Biosystems, Foster City, CA, USA) laser-desorption time-of-flight mass spectrometer equipped with UV-nitrogen laser (337 nm) and an AXIMA-CFR plus (Shimadzu Biotech, Manchester, UK) All spectra were acquired in the linear and reflectron modes, in both positive- and negative-ion modes at an accelerating voltage of 20 kV, grid voltage 95%, guide wire 0.05 and delay time 400 ns; 30–150 shots were averaged for each spectrum Gentisic acid (2,5-dihydroxybenzoic acid, GA), nor-harmane (9H-pirido-[3,4]-b-indole) [20–22] were used as matrices The best results were obtained in positive-ion mode with GA and in negative-ion mode with nor-harmane Matrix solutions were made by dissolving mg of the selected compound in mL of MeOH ⁄ H2O (1 : 1, v ⁄ v) Analyte solutions were freshly prepared by dissolving the carbohydrates in water The analyte-matrix deposit was prepared following the thin-film layer method; 0.5 lL of the matrix solution was placed on the sample probe tip and the solvent removed by blowing air at room temperature Subsequently, 0.5 lL of the analyte solution was placed on the same probe tip covering the matrix and partially dissolving it, and the solvent was removed by blowing air Then, two additional portions (0.5 lL) of the matrix solution were deposited on the same sample probe tip producing a partial dissolution of the previously deposited thin-film matrix and analyte layers The matrix to analyte ratio was : (v ⁄ v) and the matrix and analyte solution loading sequence was matrix, analyte, matrix, matrix External calibration was performed with: insulin (10 pmolỈlL)1 in H2O 0.1% trifluoroacetic acid) and ribonuclease [10 pmolỈlL)1 in H2O 0.1% (v ⁄ v) trifluoroacetic acid] using sinapinic acid (1 mg per 0.04 mL MeCN plus 0.06 mL H2O 0.1% (v ⁄ v) trifluoroacetic acid as matrix]; b-cyclodextrine (10 pmolỈlL)1 in H2O) using nor-harmane as matrix [2 mgỈmL)1 MeOH ⁄ H2O (1 : 1, v ⁄ v)]; bradykinin [10 pmolỈlL)1 in H2O 0.1% (v ⁄ v) trifluoroacetic acid] and FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS Structure of the N-glycans in cruzipain angiotensin II [10 pmolỈlL)1 in H2O 0.1% (v ⁄ v) trifluoroacetic acid] using CHCA [10 mgỈmL)1 in 50% (v ⁄ v) MeCN ⁄ H2O 0.1% (v ⁄ v) trifluoroacetic acid], in positive and negative ion modes The following glycans were used as standards: Man6GlcNAc2 (Oxford Glyco System) and oligosaccharides obtained from urine follicle-stimulating hormone (uFSH) (Massone S.A., Argentina) Acknowledgements The authors are indebted to the National Research Council of Argentina (CONICET, PIP ⁄ 904 and PIP ⁄ 447), the University of Buenos Aires (UBA, 022 ´ and 218), Agencia Nacional de Promocion Cientı´ fica y ´ Tecnologica, Argentina (ANPCyT, BID 1201 ⁄ OC-AR Pict 12312 and 06–06545) and UNDP ⁄ World Bank ⁄ WHO Special Programme for Research and Training in Tropical Diseases (TDR; Project ID 970629) for partial financial support J J Cazzulo, A S Couto, V G Duschak and R Erra-Balsells are Research Members of CONICET M Barboza has a fellowship from the same Institution UV-MALDI-TOF MS experiments were performed: (a) as part of the Academic Agreement between REB (FCEyN-UBA, Argentina) and HN (CA-EU, Japan) with the facilities of the High Resolution Liquid Chromatography-integrated Mass Spectrometry System of the United Graduated School of Agricultured Sciences (EU, Japan) and (b) thanks to the Shimadzu Corporation with the facilities of the Koichi Tanaka Mass Spectrometry Research Laboratory (Kyoto, Japan) References Franke de Cazzulo BM, Martinez J, North MJ, Coombs GH & Cazzulo JJ (1994) Effects of proteinase inhibitors on growth and differentiation of Trypanosoma cruzi FEMS Microbiol Lett 124, 81–86 Bontempi E, Martinez J & Cazzulo JJ (1989) Subcellular localization of a cystein proteinase from Trypanosoma cruzi Mol Biochem Parasitol 33, 43–47 Murta ACM, Persechini PM, de Souto-Padron T, De Souza W, Guimaraes JA & Scharfstein J (1990) Structural and functional identification of GP57 ⁄ 51 antigen of Trypanosoma cruzi as a cystein proteinase Mol Biochem Parasitol 43, 27–38 ´ Souto-Padron T, Campetella O, Cazzulo JJ & de Souza W (1990) Cystein proteinase in Trypanosoma cruzi: immunocytochemical localization and involvement in parasite–host cell interaction J Cell Sci 96, 485–490 Parussini F, Duschak VG & Cazzulo JJ (1998) Membrane-bound cysteine proteinase isoforms in different developmental stages of Trypanosoma cruzi Cell Mol Biol 44, 513–519 3813 Structure of the N-glycans in cruzipain Coombs GH & Mottram JC (1997) Trypanosomiasis and Leishmaniasis (Hide G & Mottram JC, Coombs GH & Holmes PH, eds.), pp 177–197 CAB International, Oxford, UK Cazzulo JJ, Stoka V & Turk V (1997) Cruzipain, the major cystein proteinase from the protozoan parasite Trypanosoma cruzi Biol Chem 378, 1–10 Cazzulo JJ, Labriola C, Parussini F, Duschak VG, Martinez J & Franke de Cazzulo BM (1995) Cystein proteinases in Trypanosoma cruzi and other Trypanosomatid parasites Acta Chimica Slovenica 42, 409–418 Campetella O, Henriksson J, Aslund L, Frasch ACC, Pettersson U & Cazzulo JJ (1992) The major cystein proteinase (cruzipain) from Trypanosoma cruzi is encoded by multiple polymorphic tandemly organized genes located on different chromosomes Mol Biochem Parasitol 50, 225–234 10 Martinez J, Campetella O, Frasch AC & Cazzulo JJ (1993) The reactivity of sera from chagasic patients against different fragments of cruzipain, the major cystein proteinase of Trypanosoma cruzi, suggests the presence of defined antigenic and catalytic domains Immunol Lett 35, 191–196 11 Cazzulo JJ, Martı´ nez J, Parodi AJ, Wernstedt C & Hellman U (1992) On the post-translational modifications at the C-terminal domain of the major cystein proteinase (cruzipain) from Trypanosoma cruzi FEMS Microbiol Lett 100, 411–416 12 Parodi AJ, Labriola C & Cazzulo JJ (1995) The presence of complex- type oligosaccharides at the C-terminal domain glycosylation site of some molecules of cruzipain Mol Biochem Parasitol 69, 247–255 13 Metzner SI, Souza MC, Hellman U, Cazzulo JJ & Parodi AJ (1996) The use of UDP-Glc: glycoprotein glucosyltransferase for radiolabelling protein-linked high mannose-type oligosaccharides Cell Mol Biol 42, 631–635 14 Barboza M, Duschak VG, Cazzulo JJ, Lederkremer RM & Couto AS (2003) Presence of sialic acid in N-linked oligosaccharide chains and O-linked N-acetylglucosamine in cruzipain, the major cysteine proteinase of Trypanosoma cruzi Mol Biochem Parasitol 126, 293–296 15 Lowry OH, Rosebrough NJ, Farr AL & Randall RJ (1951) Protein measurement with Folin phenol reagent J Biol Chem 193, 265–275 16 Freeze HH, Yeh R, Miller AL & Kornfeld S (1983) Structural analysis of the Asparagine-linked oligosaccharides from free lysosomal enzymes of Dictyostelium discoideum J Biol Chem 258, 14874–14879 17 Bousfield GR, Baker VL, Gotschall RR, Butnev VY & Butnev VY (2000) Carbohydrate analysis of glycoprotein hormones Methods 21, 15–39 18 Papac DI, Briggs JB, Chin ET & Jones AJS (1998) A high throughput microscale method to release N-linked oligosaccharides from glycoproteins for matrix-assisted- 3814 M Barboza et al 19 20 21 22 23 24 25 26 27 28 29 30 laser desorption ⁄ ionization time of flight mass spectrometric analysis Glycobiology 8, 445–454 Kuster B, Wheeler SF, Hunter AP, Dwek RA & Harvey DJ (1997) Sequencing of N-linked oligosaccharides directly from protein gels: in-gel deglycosylation followed by matrix-assisted laser desorption ⁄ ionization mass spectrometry and normal-phase high-performance liquid chromatography Anal Biochem 250, 82–101 Nonami H, Fukui S & Erra-Balsells R (1997) b-carboline alkaloids as matrices for matrix-assisted ultraviolet laser desorption time-of-flight mass spectrometry of proteins and sulfated oligosaccharides: a comparative study using phenylcarbonyl compounds, carbazoles and classical matrices J Mass Spectrom 32, 287–296 Nonami H, Tanaka K, Fukuyama Y & Erra-Balsells R (1998) B-carboline alkaloids as matrices for UV-matrixassisted laser desorption ⁄ ionization time of flight mass spectrometry in positive and negative ion modes Analysis of proteins of high molecular mass, and of cyclic and acyclic oligosaccharides Rapid Commun Mass Spectrom 12, 285–296 Erra-Balsells R & Nonami H (2002) Nor-Harmane (9H-pyrido [3,4-b]indole) as outstanding matrix for UVmatrix-assisted laser desorption ⁄ ionization time-of-flight mass spectrometry analysis of synthetic and biopolymers Environ Control Biol 40, 55–73 Duschak VG, Barboza M & Couto AS (2003) Trypanosoma cruzi: partial characterization of minor cruzipain isoforms non-adsorbed to concanavalin A-Sepharose Exp Parasitol 104, 122–130 Freeze HH & Wolgast D (1986) Structural analysis of the N-linked oligosaccharides glycoproteins secreted by Disctyostelium discoideum Identification of mannose-6sulfate J Biol Chem 261, 127–134 Frasch ACC (2000) Functional diversity in the transsialidase and mucin families in Trypanosoma cruzi Parasitol Today 16, 282–286 McConville MJ, Mullin KA, Ilgout SC & Teasdale RD (2002) Secretory pathway of trypanosomatid parasites Microbiol Mol Biol Rev 66, 122–154 Nolan DP, Geuskens M & Pays E (1999) N-linked glycans containing linear polyN-acetyllactosamine as sorting signals in endocytocis in Trypanosoma brucei Curr Biol 9, 1169–1172 Snary D, Ferguson MAJ, Scott MT & Allen AK (1981) Cell surface antigens of Trypanosoma cruzi: use of monoclonal antibodies to identify and isolate an epimastigote specific glycoprotein Mol Biochem Parasitol 3, 343–358 Ferguson MAJ, Allen AK & Snary D (1983) Studies on the structure of a phosphoglycoprotein from the parasitic protozoan Trypanosoma cruzi Biochem J 213, 313–320 Couto AS, Colli W, Goncalves MF & Lederkremer ¸ RM (1990) The N- linked carbohydrate chain of the FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS M Barboza et al 31 32 33 34 35 36 37 Tc-85 glycoprotein from Trypanosoma cruzi trypomastigotes contains sialyl, fucosyl and galactosyl a-(1–3) galactose units Mol Biochem Parasitol 39, 101–108 Dai Y, Whittal RM, Bridges CA, Isogai Y, Hindsgaul O & Li L (1997) Matrix-assisted laser desorption ionization mass spectrometry for the analysis of monosulfated oligosaccharides Carbohydr Res 304, 1–9 Wheeler SF & Harvey DJ (2001) Extension of the in-gel release method for structural analysis of neutral and sialylated N-linked glycans to the analysis of sulfated glycans: application to the glycans from bovine thyroid-stimulating hormone Anal Biochem 296, 92–100 Fukuyama Y, Ciancia M, Nonami H, Cerezo A, Erra Balsells R & Matulewicz MC (2002) Matrix-assisted ultraviolet-laser desorption ⁄ ionization and electrospray ion ionization time-of-flight mass spectrometry of sulfated neocarrabiose oligosaccharides Carbohydr Res 337, 1553–1562 Karas M, Gluckmann M & Schafer J (2000) Ionization in matrix-assisted laser desorption ⁄ ionization: singly charged molecular ions are the lucky survivors J Mass Spectrom 35, 1–12 Bernstein HB & Compans RW (1992) Sulfation of the human immunodeficiency virus envelope glycoprotein J Virol 66, 6953–6959 Van Rooijen JJ, Kamerling JP & Vliegenthart JF (1998) Sulfated di-, tri- and tetraantennary N-glycans in human Tamm-Horsfall glycoprotein Eur J Biochem 256, 471–487 Kawasaki N, Haishima Y, Ohta M, Itoh S, Hyuga M, Hyuga S & Hayakawa T (2001) Structural analysis of FEBS Journal 272 (2005) 3803–3815 ª 2005 FEBS Structure of the N-glycans in cruzipain 38 39 40 41 42 43 44 sulfated N-linked oligosaccharides in erythropoietin Glycobiology 11, 1043–1049 Green ED, van Halbeek H, Boime I & Baenziger JU (1985) Structural elucidation of the disulfated oligosaccharide from bovine lutropin J Biol Chem 260, 15623– 15630 Noguchi N & Nakano M (1992) Structure of the acidic N-linked carbohydrate chains of the 55-kDa glycoprotein family (PZP3) from porcine zona pellucida Eur J Biochem 213, 39–56 Pierce M & Arango J (1986) Rous sarcoma virus-transformed baby hamster kidney cells express higher levels of asparagine-linked tri- and tetraantennary glycopeptides containing [GlcNAc-beta(1,6)Man-alpha(1,6)Man] and poly–N–acetyllactosamine sequences than baby hamster kidney cells J Biol Chem 261, 10772–10777 Kawasaki N, Ohta M, Hyuga S, Hashimoto O & Hayakawa T (2000) Application of liquid chromatography ⁄ mass spectrometry and liquid chromatography with tandem mass spectrometry to the analysis of the sitespecific carbohydrate heterogeneity in erythropoietin Anal Biochem 285, 82–91 Honke K & Taniguchi N (2002) Sulfotransferases and sulfated oligosaccharides Med Res Rev 22, 637–654 Petry K, Nudelman E, Eisen H & Hakomori S (1988) Sulfated lipids represent common antigens on the surface of Trypanosoma cruzi and mammalian tissues Mol Biochem Parasitol 30, 113–121 Uhrig ML, Couto AS, Zingales B, Colli W & Lederkremer RM (1992) Metabolic labelling and partial characterization of a sulfoglycolipid in Trypanosoma cruzi trypomastigotes Carbohydr Res 231, 329–334 3815 ... Altogether, these data confirm the presence of sulfated high-mannose type oligosaccharides in cruzipain and in its C-terminal domain This is the first report of the use of nor-harmane as matrix for structural. .. accounts for the majority of the sulfated sugar In conclusion, the results obtained provide evidence of the nature of the glycans present in the unique N-glycosylation site present in the C-terminal... Cazzulo JJ (1993) The reactivity of sera from chagasic patients against different fragments of cruzipain, the major cystein proteinase of Trypanosoma cruzi, suggests the presence of defined antigenic