Structural analysis of N-glycans of the planarian Dugesia japonica Shunji Natsuka 1,2 , Yukiko Hirohata 2 , Shin-ichi Nakakita 3 , Wataru Sumiyoshi 3 and Sumihiro Hase 2 1 Department of Biology, Faculty of Science, Niigata University, Japan 2 Department of Chemistry, Graduate School of Science, Osaka University, Japan 3 Life Science Research Center, Kagawa University, Japan Introduction The structures of N-glycans are generally diverse, but the variation in diversity across the phylogeny is still unknown. In particular, structural information on invertebrate glycans is very limited. The nematode Caenorhabditis elegans has various unique N-glycans [1–4], even though it has a very simple body construc- tion. The major N-glycans in the cephalopods squid [5,6] and octopus [7] are similar to the complex glycans of vertebrates. In contrast, gastropods, which, like cephalopods, also belong to the mollusk family, have quite different N-glycans that are modified with methyl residues [8,9]. This suggests that phylogenic closeness and N-glycan similarity are not correlated in a simple way. In this study, we analyzed the N-glycan structures of planarians, which are platyhelminths, in order to verify this proposal. Planarians are lower animals com- pared with mollusks, i.e. squids, octopuses and snails, whose N-glycan structures have already been studied. If N-glycans become more complex along the phylog- eny, the N-glycans of planarians will less complicated than those of higher animals. This prediction arises from the simple logic that an organism with a complex system will have complex and diverse molecules. In contrast, if the planarian has unique and complex N-glycans, the view that the diversity of N-glycans does not correlate simply with the phylogenic process is supported. The species of planarian used in this study, Dugesia japonica, is a free-living platyhelminth. Keywords diversity; methylation; N-glycans; phylogeny; planarian Correspondence S. Natsuka, Department of Biology, Faculty of Science, Niigata University, Niigata 950-2181, Japan Tel ⁄ Fax: +81 25 262 6174 E-mail: natsuka@bio.sc.niigata-u.ac.jp (Received 13 August 2010, revised 15 November 2010, accepted 17 November 2010) doi:10.1111/j.1742-4658.2010.07966.x To investigate the relationship between phylogeny and glycan structures, we analyzed the structure of planarian N-glycans. The planarian Duge- sia japonica, a member of the flatworm family, is a lower metazoan. N-gly- cans were prepared from whole worms by hydrazinolysis, followed by tagging with the fluorophore 2-aminopyridine at their reducing end. The labeled N-glycans were purified, and separated by three HPLC steps. By comparison with standard pyridylaminated N-glycans, it was shown that the N-glycans of planarian include high mannose-type and pauci-mannose- type glycans. However, many of the major N-glycans from planarians have novel structures, as their elution positions did not match those of the stan- dard glycans. The results of mass spectrometry and sugar component anal- yses indicated that these glycans include methyl mannoses, and that the most probable linkage was 3-O-methylation. Furthermore, the methyl resi- dues on the most abundant glycan may be attached to the non-reducing- end mannose, as the glycans were resistant to a-mannosidase digestion. These results indicate that methylated high-mannose-type glycans are the most abundant structure in planarians. Abbreviations PA, pyridylamino. 452 FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS Platyhelminths also include Schistosoma species, which are parasitic flatworms. The N-glycan structures of parasitic flatworms have been reported previously [10] and include multi-fucosylated complex-type structures. However, the glycan structure of parasites may be dif- ferent from their ancestral forms because of high selec- tion pressure due to competing with a host immune system. For that reason, we chose to use a free-living platyhelminth. Results N-glycans were liberated from lyophilized planarians and tagged with 2-aminopyridine at their reducing ter- minus. These pyridylaminated (PA-) N-glycans were separated by anion-exchange chromatography into neutral (N) and acidic (A) fractions (Fig. 1A). Reducing-end analysis indicated that the neutral frac- tion included over 95% of the total amount of PA-N- glycans (Fig. 1B). The neutral fraction was further sep- arated by size-fractionation HPLC (Fig. 2). Thirteen fractions (A–M) with molecular sizes indicating the presence of more than three glucose units were col- lected, and further separated by reversed-phase HPLC (Fig. 3). Peaks that were > 5% in area of the major PA-glycan (pN4) were collected and named pN1– pN18. A two-dimensional glycan map was prepared based on the retention times of the PA-glycans on size- fractionation and reversed-phase HPLC (Fig. 4). To reduce fluctuations in retention time, the retention time on size-fractionation HPLC was normalized against glucose unit, and that on reversed-phase HPLC was converted to the reversed-phase scale, as described in Experimental procedures. The positions of pN1, pN11, pN13, pN15, pN16, pN17 and pN18 on the map coincided with those of the standard PA-glycans M2B, M5A, M6B, M7B, M8A, M9A and G1M9A, respec- tively (Table 1). As the result of MS analysis sup- ported these assignments (Table 2), we concluded that the planarian N-glycans pN1, pN11, pN13, pN15, pN16, pN17 and pN18 did indeed match these standard structures. The results of MS analysis of other PA-glycans suggested multi-methylation of Fig. 1. Anion-exchange HPLC. (A) PA-N-glycans from planarians were separated into neutral (N) and acidic (A) fractions by anion- exchange chromatography. The thick bar indicates the collected fractions. X indicates a peak derived from reagents. The fluores- cence scale in the inset chromatogram is magnified 30-fold. Arrowheads labeled S0–S4 indicate elution positions of standard PA-N-glycans with 0–4 sialic acids, respectively. (B) Contents of PA-glycans that have PA-GlcNAc at their reducing end in the fractions from anion-exchange chromatography. Fig. 2. Size-fractionation HPLC. The neutral PA-glycans were sepa- rated by size-fractionation HPLC. Thirteen major peaks (A–M), the molecular sizes of which were larger than that of isomaltotriose trisaccharide, were isolated. Arrowheads indicate the elution posi- tions of PA-isomalto-oligosaccharides with degrees of polymeriza- tion from 1 to 15. S. Natsuka et al. N-glycans of planarian FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS 453 oligomannosidic N-glycans (Fig. 5 and Table 2). We subsequently analyzed the sugar composition of pN4 (Fig. 6). As methyl residues were not released from sugars under the hydrolyzing conditions of the analy- sis, methylated monosaccharide(s) could be detected. Indeed, a peak that had an elution position that coincided with that of pyridylaminated 3-O-methyl mannose was detected on both reversed-phase and borate-chelating anion-exchange HPLC. Mannose and N-acetylglucosamine were also detected by the HPLC analyses. Based on the results of the MS and sugar component analyses, pN4 was estimated to comprise three methyl mannoses, two mannoses and two N-acetylglucosamines, and the linkage position of the O-methyl residues was tentatively assigned as the third position of the mannoses. Furthermore, the 3-O- methyl mannoses may be located at the non-reducing end, as pN4 was resistant to a-mannosidase digestion (data not shown). We then investigated the core struc- ture of pN4 by partial acid hydrolysis. pN4 was par- tially hydrolyzed to form fragments of various lengths. The reducing-end fragments with pyridylamino resi- dues were analyzed by size-fractionation and reversed- phase HPLC, and plotted as a two-dimensional map (Fig. 7). Fragments with elution positions that coincided with GlcNAcb1-4GlcNAc-PA, Manb1-4Glc- NAcb1-4GlcNAc-PA and Mana1-6Manb1-4GlcNAcb1- 4GlcNAc-PA were formed from pN4. These results indicate that the structure of pN4 was Mana1- 6Manb1-4GlcNAcb1-4GlcNAc-PA, modified with three 3-O-methyl mannoses. Furthermore, based on the results of the MS analysis, pN6 and pN10 may contain fucose, which is a common deoxyhexose in animals. However, further analysis was not performed because the amounts of material available were too small. Discussion The results of this study show that the planarian has unique variation in its N-glycans, which are modified by methyl residues at the 3-position of mannose. It has been reported that certain snails and slugs have methy- lated pauci-mannose-type glycans [8,9]. The linkage position of the methyl residue was also reported to be the 3-position of mannose. However, planarians and snails are not closely related from the phylogenic point of view: planarians are platyhelminths while snails are mollusks. In a previous paper, we described the N-gly- cans of squid, another mollusk [6]. However, the char- acteristic structures of squid N-glycans are quite Fig. 3. Reversed-phase HPLC of fractions (A)–(M). Fractions (A)–(M) obtained on size-fractionation HPLC were further separated by reversed-phase HPLC, and the major PA-glycans isolated were named pN1–pN18. N-glycans of planarian S. Natsuka et al. 454 FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS different from those of snails. These results suggest that phylogenic closeness and N-glycan similarity are not correlated in a simple way. Each metazoan has characteristic N-glycans. For example, C. elegans has highly fucosylated core struc- tures [4], squid has a complex type with type 1 lactosamine [5,6], snails [8,9] and planarians have a 3-O-methylated oligomannosidic type, and vertebrates have a complex type with type 2 lactosamine. Even the protozoan Dictyostelium discoideum has unique N-glycan structures, with intersecting GlcNAc [11]. However, many genes encoding the enzymes that syn- thesize these specific glycans of each organism have not been identified. For example, the existence of an enzyme that catalyzes mannose methylation was sug- gested by the results of this study. Methylated N-gly- cans have been discovered in various invertebrates, for example snails [8,12], slugs [9] and nematodes [1]. However, a gene encoding an enzyme capable of syn- thesizing methylated glycans has not been reported so far. In future studies, research to identify the genes involved in biosynthesis of the methylated glycan will be required. Why does each organism have unique N-glycan structures? One possible answer is that primitive multi- cellular organisms may already have the genetic poten- tial in their primitive genome for synthesizing the vari- ous N-glycan structures. From many possible variations, each organism has ‘chosen’ particular struc- tures during evolution. To verify this hypothesis, it is necessary to compare the genes involved in the synthe- sis of N-glycans in many organisms. The hypothesis would gain strong support if lower metazoans were found to have the genetic potential to synthesize the N-glycans found in higher metazoans. Furthermore, it is possible that particular glycan structures are not inevitable, and that micro-heterogeneity is essential for each organism, i.e. alternative sets of glycans may be able to form similar systems in organisms. The study of glycans from the phylogenic viewpoint remains important in order to be able to answer this question. Experimental procedures Preparation of PA-glycans from planarian The planarians (Dugesia japonica) were purchased from a local animal vendor and kept in water at 14 °C in the dark. They were fed with chicken liver. After starvation for 5 days, the planarians were washed with water and lyophi- lized. N-linked glycans were liberated from the glycopro- teins of lyophilized planarians by hydrazinolysis as described previously [13,14]. Briefly, 2 mg of the sample was heated at 100 °C for 10 h with 0.2 mL of anhydrous hydrazine. After removal of hydrazine by repeated evapora- tions, the glycans were re-N-acetylated with acetic anhy- dride in a saturated sodium bicarbonate solution, and then passed through a Dowex 50Wx2 (H + ) cation exchanger (Dow Chemicals, Midland, MI, USA) to remove sodium ions. The reducing ends of the liberated glycans were tagged with the fluorophore 2-aminopyridine. Lyophilized samples were heated at 90 °C for 60 min with 20 lL of pyr- idylamination reagent, and then heated at 80 °C for 35 min after the addition of 70 lL of reducing reagent, as described previously [14]. Purification of PA-glycans The pyridylamination reaction mixture was diluted with 150 lL of water and extracted twice using 200 lL of phe- nol ⁄ chloroform (1 : 1 v ⁄ v) to remove the excess reagents [15]. The water layer that contained the PA-glycans was purified by gel filtration on a mini-column (1 · 8 cm, TSK- gel Toyopearl HW-40F, Tosoh, Tokyo, Japan). After load- ing the sample, the eluate between 2 and 6 mL was collected as the PA-glycan fraction and lyophilized. The PA-glycans were further purified using a graphite carbon Fig. 4. Two-dimensional HPLC map of the PA-glycans from planar- ian. The retention times on size-fractionation and reversed-phase HPLC were converted to glucose units and the reversed-phase scale, respectively, as described in Experimental procedures. pN1– pN18 were plotted as circles on a two-dimensional map. The diam- eter of the circles is proportional to the cube root of the detected amount of PA-glycans. The positions of standard PA-glycans are indicated by crosses labeled (a)–(g). The standards were: (a) M2B, (b) M5A, (c) M6B, (d) M7B, (e) M8A, (f) M9A and (g) G1M9A. The structures of the standard PA-glycans are shown in Table 1. S. Natsuka et al. N-glycans of planarian FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS 455 Table 1. Estimated structures of PA-N-glycans prepared from planarian. Ratios indicate the relative amounts of the glycans, with the amount of the most abundant glycan, pN4, taken as 100. pN1 Peak # pN2 pN3 pN4 pN5 pN6 pN7 pN8 pN9 pN10 pN11 pN12 pN13 pN14 pN15 pN16 pN17 pN18 Ratio 5.2 14.5 6.0 100 20.2 20.4 22.4 14.3 6.4 5.9 5.6 27.0 19.7 10.6 8.8 17.3 32.9 5.5 Abbreviation M2B Me 3 -M5 M5A M6B M7B M8A M9A G1M9A Estimated Structure Manα1 Manα1 Manα1 Manα1 Manα1 Manα1 Manα1 Manα1 Manα1 Manα1 Manα1 Manα1-2Manα1 Manα1-2Manα1-2Manα1 Manα1-2Manα1-2Manα1 Manα1-2Manα1-2Manα1 Glcα1-3Manα1-2Manα1-2Manα1 Manα1-2Manα1 Manα1-2Manα1 Manα1-2Manα1 Manα1-2Manα1 Manα1-2Manα1 Manα1 Manα 1 Manα1 Manα1 6 6 3 6 6 6 6 6 3 3 3 6 6 6 6 3 3 3 3 3 6 3 3 3 Manα1 (Me-Hex-) 2 Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 2 Hex-Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 3 (Hex-) 2 Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 3 (Hex-) 3 Manβ1-4G1cNAcβ1-4G1cNAc-PA Manβ1-4G1cNAcβ1-4G1cNAc-PA Manβ1-4G1cNAcβ1-4G1cNAc-PA Manβ1-4G1cNAcβ1-4G1cNAc-PA Manβ1-4G1cNAcβ1-4G1cNAc-PA Manβ1-4G1cNAcβ1-4G1cNAc-PA Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 3 (Hex-) 3 Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 3 (Hex-) 4 Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 3 (Hex-) 5 Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-Hex-) 3 Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-) 3 (Hex1-) 3 (dHex1-) Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-) 3 (Hex-) 4 (dHex)(Pen-)Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-) 2 (Hex-) 3 (dHex) 2 Manβ1-4GleNAeβ1-4GleNAe-PA or Manβ1-4G1cNAcβ1-4G1cNAc-PA (Me-3Man1-) 3 Manβ1-4GlcNAcβ1-4GlcNAc-PA 6 6 { N-glycans of planarian S. Natsuka et al. 456 FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS cartridge (GL-Pak Carbograph 300 mg; GL Sciences Ltd, Tokyo, Japan). The glycan mixture was dissolved in 1 mL of 50 mm ammonium acetate, pH 6.0, and loaded onto the cartridge. After washing with 5 mL of 50 mm ammonium acetate, pH 6.0, the glycans were eluted using 5 mL of 60% acetonitrile in the ammonium acetate buffer. High-performance liquid chromatography for PA-glycan separation HPLC was performed using a Waters Alliance HPLC sys- tem with a Waters 2475 fluorescence spectrophotometer (Milford, MA, USA) unless otherwise mentioned. Anion- exchange HPLC was performed using a TSKgel DEAE- 5PW column (0.75 · 7.5 cm; Tosoh) at a flow rate of 1.0 mLÆmin )1 . The column was equilibrated with diluted ammonium solution, pH 9.0. Five minutes after the sample had been injected, the concentration of ammonium acetate, pH 9.0, was increased linearly to 0.2 m in the first 20 min, Fig. 6. Sugar component analysis of pN4. The pyridylaminated monosaccharides prepared from pN4 were analyzed by reversed- phase HPLC (A) and borate-chelating anion-exchange HPLC (B). Arrows (a)–(f) indicate the elution positions of PA-Man, PA-4-O- methyl mannose, PA-GlcNAc, PA-2-O-methyl mannose, PA-6-O- methyl mannose and PA-3-O-methyl mannose, respectively. Fig. 5. MALDI-TOF mass spectrum of pN4. Experimental details are given in Experimental procedures. Table 2. Results of mass analysis. Me, methyl residue; Hex, hexose; dHex, deoxyhexose; Pen, pentose; HexNAc, N-acetylhexosamine; PA, pyridylamino residue. Code Mass number (observed) Estimated composition (Na + adduct) Mass number (calculated) pN1 849.509 Hex 2 HexNAc 2 PA 849.323 pN2 1039.266 Me 2 Hex 3 HexNAc 2 PA 1039.407 pN3 1215.788 Me 3 Hex 4 HexNAc 2 PA 1215.475 pN4 1377.866 Me 3 Hex 5 HexNAc 2 PA 1377.528 pN5 1201.577 Me 2 Hex 4 HexNAc 2 PA 1201.460 pN6 1347.550 Me 2 Hex 4 dHexHexNAc 2 PA 1347.518 pN7 1539.895 Me 3 Hex 6 HexNAc 2 PA 1539.581 pN8 1702.141 Me 3 Hex 7 HexNAc 2 PA 1701.634 pN9 1702.017 Me 3 Hex 7 HexNAc 2 PA 1701.634 pN10 1656.007 Me 3 Hex 5 dHexPenHexNAc 2 PA or Me 2 Hex 5 dHex 2 HexNAc 2 PA 1655.628 pN11 1335.957 Hex 5 HexNAc 2 PA 1335.481 pN12 1864.035 Me 3 Hex 8 HexNAc 2 PA 1863.686 pN13 1497.876 Hex 6 HexNAc 2 PA 1497.534 pN14 2025.914 Me 3 Hex 9 HexNAc 2 PA 2025.739 pN15 1659.802 Hex 7 HexNAc 2 PA 1659.587 pN16 1821.713 Hex 8 HexNAc 2 PA 1821.640 pN17 1983.933 Hex 9 HexNAc 2 PA 1983.692 pN18 2146.133 Hex 10 HexNAc 2 PA 2145.745 S. Natsuka et al. N-glycans of planarian FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS 457 and then to 0.5 m over the next 10 min. The PA-glycans were detected using a fluorescence spectrophotometer with an excitation wavelength of 310 nm and an emission wave- length of 380 nm. Size-fractionation HPLC was performed using a TSK gel Amide 80 column (0.46 · 7.5 cm; Tosoh) at a flow rate of 0.5 mLÆmin )1 . The column was equili- brated with 50 mm ammonium formate, pH 4.4, containing 80% acetonitrile. After the sample had been injected, the acetonitrile concentration was decreased linearly from 80% to 65% over the first 5 min, 65% to 55% over the second 5 min, and then 55% to 30% over the next 25 min. The PA-glycans were detected using a fluorescence spectropho- tometer with an excitation wavelength of 315 nm and an emission wavelength of 400 nm. The molecular size of each PA-glycan is given in terms of glucose units based on the elution times of PA-isomalto-oligosaccharides [14]. Reversed-phase HPLC was performed on a Cosmosil 5C18- P column (0.2 · 25 cm; Nacalai Tesque, Kyoto, Japan) at a flow rate of 0.2 mLÆmin )1 . The column was equilibrated with 100 mm triethylamine acetate, pH 4.0. After injection of the sample, the 1-butanol concentration was increased linearly from 0.075% to 0.5% over 105 min. The PA-gly- cans were detected using an excitation wavelength of 315 nm and an emission wavelength of 400 nm. The reten- tion time of each PA-glycan was converted to the reversed- phase scale as described previously [16]. Reversed-phase HPLC for PA-monosaccharides was performed on an Ul- trasphere-ODS column (1.0 · 25 cm; Beckman Coulter, Brea, CA, USA) at a flow rate of 1.5 mLÆmin )1 . The col- umn was equilibrated with 1% acetonitrile in 0.25 m sodium citrate, pH 4.0. The sample was separated by iso- cratic elution, and detected using a fluorescence spectropho- tometer with an excitation wavelength of 315 nm and an emission wavelength of 400 nm. Borate-chelating anion- exchange HPLC was performed on a TSKgel Sugar AX-I column (0.46 · 15 cm; Tosoh) at a flow rate of 0.3 mLÆmin )1 using a Waters 515 HPLC pump and a Hitachi L-7485 fluorescence spectrophotometer (Tokyo, Japan). The column was equilibrated with 10% acetonitrile in 0.8 m potassium borate, pH 9.0, at a temperature of 74 °Cina column oven (CTO-10ASvp; Shimadzu, Kyoto, Japan). The sample was separated by isocratic elution, and detected using a fluorescence spectrophotometer with an excitation wavelength of 310 nm and an emission wavelength of 380 nm. Reducing-end analysis The reducing-end analysis was performed as reported previ- ously [17]. The PA-glycans were hydrolyzed with 4 m hydrochloric acid at 100 °C for 4 h. After complete evapo- ration of the hydrochloric acid, the liberated monosaccha- rides were re-N-acetylated with acetic anhydride in a saturated sodium bicarbonate solution. PA-monosaccha- rides derived from reducing ends were purified on a small cation-exchange column, as described previously [18]. Briefly, a sample was applied to a Dowex 50Wx2 (H + ) column (0.5 · 3 cm). After the column had been washed with 2 mL of water, PA-sugars were eluted using 3 mL of 2.5% ammonia solution. The eluate was lyophilized to remove ammonia, and then analyzed by HPLC. PA-sugars were separated using a TSKgel Sugar AX-I column (0.46 · 15 cm; Tosoh) equilibrated with 10% acetonitrile in 0.8 m potassium borate, pH 9.0, at a flow rate of 0.3 mLÆmin )1 at 74 °C, and detected using a fluorescence spectrophotometer at an excitation wavelength of 310 nm and an emission wavelength of 380 nm. The molar ratio of GlcNAc detected in the fractions was taken as the molar ratio of N-glycans. Sugar component analysis The PA-glycans were hydrolyzed with 4 m trifluoroacetic acid at 100 °C for 4 h. After complete evaporation of triflu- oroacetic acid, the liberated monosaccharides were re-N- acetylated with acetic anhydride in a mixture of methanol and pyridine (4 : 1 v ⁄ v) for 30 min at room temperature. The solvent was evaporated, and then the monosaccharides Fig. 7. Two-dimensional HPLC map of the partially hydrolyzed frag- ments from pN4. pN4 was partially hydrolyzed, and glycan frag- ments from the pyridylaminated reducing end were separated using two types of HPLC, as described in Experimental procedures. The retention times on size-fractionation and reversed-phase HPLC were converted to glucose units and the reversed-phase scale, respectively. The positions of the standard PA-glycans GlcNAcb1-4GlcNAc-PA, Manb1-4GlcNAcb1-4GlcNAc-PA, Mana1- 3Manb1-4GlcNAcb1-4GlcNAc-PA (M2A) and Mana1-6Manb1-4Glc- NAcb1-4GlcNAc-PA (M2B) are indicated by crosses labeled (a)–(d), respectively. N-glycans of planarian S. Natsuka et al. 458 FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS were tagged with 2-aminopyridine by a pyridylamination method modified for monosaccharides [18]. The excess reagents were removed by repeated evaporation with tolu- ene and methanol. The PA-monosaccharides were analyzed by reversed-phase and borate-chelating anion-exchange HPLC. Standard methylated monosaccharides were prepared as described previously [19]. Enzyme digestion Approximately 10 pmol of PA-glycans were digested with 10 unitsÆmL )1 of jack bean (Canavaria gladiata) a-mannosi- dase (Seikagaku Kogyo, Tokyo, Japan) in 20 lLof50mm sodium citrate buffer, pH 4.5, for 20 h at 37 °C. Mass spectrometric analysis PA-glycans separated by HPLC were lyophilized to remove volatile salts in the buffer, and 1–10 pmol were co-crystal- lized with 1 mgÆmL )1 of 2,5-dihydroxybenzoic acid in ace- tonitrile on an AnchorChip target plate (Bruker Daltonics, Billerica, MA, USA) according to the manufacturer’s protocol. MALDI-TOF mass spectra were recorded using an Autoflex II (Bruker Daltonics) in reflector mode. Partial acid hydrolysis The PA-glycans were hydrolyzed with 1 m trifluoroacetic acid at 100 ° C for 30 min [20]. After complete evaporation of trifluoroacetic acid, the hydrolyzed PA-glycans were re-N-acetylated with acetic anhydride in a saturated sodium bicarbonate solution as described above. The PA-glycans were purified on a small cation-exchange column, as described previously [18], and analyzed by size-fractionation and reversed-phase HPLC to produce a glycan map. References 1 Altmann F, Fabini G, Ahorn H & Wilson IB (2001) Genetic model organisms in the study of N-glycans. Biochimie 83, 703–712. 2 Natsuka S, Adachi J, Kawaguchi M, Nakakita S, Hase S, Ichikawa A & Ikura K (2002) Structural analysis of N-linked glycans in Caenorhabditis elegans. J Biochem 131, 807–813. 3 Cipollo JF, Costello CE & Hirschberg CB (2002) The fine structure of Caenorhabditis elegans N-glycans. J Biol Chem 277, 49143–49157. 4 Hanneman AJ, Rosa JC, Ashline D & Reinhold VN (2006) Isomer and glycomer complexities of core GlcNAcs in Caenorhabditis elegans. Glycobiology 16, 874–890. 5 Takahashi N, Masuda K, Hiraki K, Yoshihara K, Huang HH, Khoo KH & Kato K (2003) N-Glycan structures of squid rhodopsin. Eur J Biochem 270, 2627–2632. 6 Natsuka S, Ishida M, Ichikawa A, Ikura K & Hase S (2006) Comparative biochemical study of N-linked gly- cans from skin of a squid, Todarodes pacifics. J Biochem 140, 87–93. 7 Zhang Y, Iwasa T, Tsuda M, Kobata A & Takasaki S (1997) A novel monoantennary complex-type sugar chain found in octopus rhodopsin: occurrence of the Galb1 fi 4Fuc group linked to the proximal N-acetyl- glucosamine residue of the trimannosyl core. Glycobiolo- gy 7, 1153–1158. 8 van Kuik JA, van Halbeek H, Kamerling JP & Vliegenthart JF (1985) Primary structure of the low-molecular-weight carbohydrate chains of Helix pomatia a-hemocyanin. Xylose as a constituent of N-linked oligosaccharides in an animal glycoprotein. J Biol Chem 260, 13984–13988. 9 Gutternigg M, Ahrer K, Grabher-Meier H, Bu ¨ rgmayr S & Staudacher E (2004) Neutral N-glycans of the gastro- pod Arion lusitanicus. Eur J Biochem 271, 1348–1356. 10 Khoo KH, Chatterjee D, Caulfield JP, Morris HR & Dell A (1997) Structural mapping of the glycans from the egg glycoproteins of Schistosoma mansoni and Shis- tosoma japonicum: identification of novel core structures and terminal sequences. Glycobiology 7, 663–677. 11 Sharkey DJ & Kornfeld R (1989) Identification of an N-acetylglucosaminyltransferase in Dictyostelium dis- coideum that transfers an ‘intersecting’ N-acetylglucos- amine residue to high mannose oligosaccharides. J Biol Chem 264, 10411–10419. 12 Lehr T, Geyer H, Maass K, Doenhoff MJ & Geyer R (2006) Structural characterization of N-glycans from the freshwater snail Biomphalaria glabrata cross-reacting with Schistosoma mansoni glycoconjugates. Glycobiology 17, 82–103. 13 Yosizawa Z, Sato T & Schmid K (1966) Hydrazinolysis of a-1-acid glycoprotein. Biochim Biophys Acta 121, 417–420. 14 Natsuka S & Hase S (1998) Analysis of N- and O-gly- cans by pyridylamination. Methods Mol Biol 76, 101– 113. 15 Yanagida K, Natsuka S & Hase S (1999) A pyridylam- ination method aimed at automatic oligosaccharide analysis of N-linked sugar chains. Anal Biochem 274, 229–234. 16 Yanagida K, Ogawa H, Omichi K & Hase S (1998) Introduction of a new scale into reversed-phase high- performance liquid chromatography of pyridylamino sugar chains for structural assignment. J Chromatogr A 800, 187–198. 17 Makino Y, Kuraya N, Omichi K & Hase S (1996) Clas- sification of sugar chains of glycoproteins by analyzing reducing end oligosaccharides obtained by partial acid hydrolysis. Anal Biochem 238, 54–59. S. Natsuka et al. N-glycans of planarian FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS 459 18 Suzuki J, Kondo A, Kato I, Hase S & Ikenaka T (1991) Analysis by high-performance anion-exchange chromatography of component sugars as their fluores- cent pyridylamino derivatives. Agric Biol Chem 55, 283–284. 19 Mega T, Nishikawa A & Ikenaka T (1983) A conve- nient method for the preparation of all the partially methylated derivatives of methyl a-d-mannopyranoside and a-d-galactopyranoside. Carbohydr Res 112, 313–319. 20 Makino Y, Omichi K & Hase S (1998) Analysis of oli- gosaccharide structures from the reducing end terminal by combining partial acid hydrolysis and a two-dimen- sional sugar map. Anal Biochem 264, 172–179. N-glycans of planarian S. Natsuka et al. 460 FEBS Journal 278 (2011) 452–460 ª 2011 The Authors Journal compilation ª 2011 FEBS . map. The diam- eter of the circles is proportional to the cube root of the detected amount of PA-glycans. The positions of standard PA-glycans are indicated by crosses labeled (a)–(g). The standards. In contrast, if the planarian has unique and complex N-glycans, the view that the diversity of N-glycans does not correlate simply with the phylogenic process is supported. The species of planarian. November 2010) doi:10.1111/j.1742-4658.2010.07966.x To investigate the relationship between phylogeny and glycan structures, we analyzed the structure of planarian N-glycans. The planarian Duge- sia japonica, a member of the flatworm family,