Skip Navigation


Glycobiology Advance Access originally published online on March 29, 2007
Glycobiology 2007 17(7):754-766; doi:10.1093/glycob/cwm035
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
17/7/754    most recent
cwm035v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (2)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Arigi, E.
Right arrow Articles by Levery, S. B
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Arigi, E.
Right arrow Articles by Levery, S. B
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org

Characterization of neutral and acidic glycosphingolipids from the lectin-producing mushroom, Polyporus squamosus

Emma Arigi2, Suddham Singh2, Ardalan H Kahlili3, Harry C Winter4, Irwin J Goldstein4 and Steven B Levery1,2

2 Department of Chemistry, University of New Hampshire, Durham, NH 03824-3598
3 Department of Biochemistry and Molecular Biology, The Complex Carbohydrate Research Center, University of Georgia, Athens, GA 30602-7229
4 Department of Biological Chemistry, School of Medicine, University of Michigan, Ann Arbor, MI 48109-0606

1 To whom correspondence should be addressed; Tel: +1-603-862-2529; Fax: +1-603-862-4278; E-mail: slevery{at}cisunix.unh.edu

Received on December 30, 2006; revised on March 15, 2007; accepted on March 18, 2007


   Abstract
Top
 Abstract
Introduction
 Results
 Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
The polypore mushroom Polyporus squamosus is the source of a lectin that exhibits a general affinity for terminal ß-galactosides, but appears to have an extended carbohydrate-binding site with high affinity and strict specificity for the nonreducing terminal trisaccharide sequence NeuAc{alpha}2 -> 6Galß1 -> 4Glc/GlcNAc. In considering the possibility that the lectin's in vivo function could involve interaction with an endogenous glycoconjugate, it would clearly be helpful to identify candidate ligands among various classes of carbohydrate-containing materials expressed by P. squamosus. Since evidence has been accumulating that glycosphingolipids (GSLs) may serve as key ligands for some endogenous lectins in animal species, possible similar roles for fungal GSLs could be considered. For this study, total lipids were extracted from mature fruiting body of P. squamosus. Multistep fractionation yielded a major monohexosylceramide (CMH) component and three major glycosylinositol phosphorylceramides (GIPCs) from the neutral and acidic lipids, respectively. These were characterized by a variety of techniques as required, including one- and two-dimensional 1H- and 13C-nuclear magnetic resonance (NMR) spectroscopy; electrospray ionization–mass spectrometry (ESI-MS, tandem-MS/collision-induced decay-MS, and ion trap-MSn); and component and methylation linkage analysis by gas chromatography–mass spectrometry. The CMH was determined to be glucosylceramide having a typical ceramide consisting of 2-hydroxy fatty-N-acylated (4E,8E)-9-methyl-sphinga-4,8-dienine. The GIPCs were identified as Man{alpha}1 -> 2Ins1-P-1Cer (Ps-1), Galß1 -> 6Man{alpha}1 -> 2Ins1-P-1Cer (Ps-2), and Man{alpha}1 -> 3Fuc{alpha}1 -> 2Gal{alpha}1 -> 6Galß1 -> 6Man{alpha}1 ->2Ins1-P-1Cer (Ps-5), respectively (where Ins = myo-inositol, P = phosphodiester, and Cer = ceramide consisting mainly of long-chain 2-hydroxy and 2,3-dihydroxy fatty-N-acylated 4-hydroxy-sphinganines). Of these GSLs, Ps-2 could potentially interact with P. squamosus lectin, and further investigations will focus on determining the binding affinity, if any, of the lectin for the GIPCs isolated from this fungus.

Key words: basidiomycete / electrospray ionization / glycolipid / mass spectrometry / NMR spectroscopy


   Introduction
Top
 Abstract
 Introduction
Results
 Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
The polypore mushroom, Polyporus squamosus, is the source of a lectin that exhibits a general affinity for terminal ß-galactosides, but appears to have an extended carbohydrate-binding site with high affinity and strict specificity for the nonreducing terminal trisaccharide sequence NeuAc{alpha}2 -> 6Galß1 -> 4Glc/GlcNAc (Mo et al. 2000Go). Negligible binding was observed with oligosaccharides bearing the isomeric terminal trisaccharides NeuAc{alpha}2 -> 3Galß1 -> 4Glc/GlcNAc or Galß1 -> 3(NeuAc{alpha}2 -> 6)GlcNAc (Zhang et al. 2001Go). The specificity of the lectin toward NeuAc{alpha}2 -> 6Galß1 -> terminated glycans suggests considerable potential for its use in glycobiological studies, but its in vivo function is unknown. In considering the possibility that its function could involve interaction with an endogenous glycoconjugate, it would clearly be helpful to identify candidate ligands among various classes of carbohydrate-containing materials expressed by P. squamosus. Since evidence has been accumulating that glycosphingolipids (GSLs) may serve as key ligands for some endogenous lectins in animal species (Vyas et al. 2002Go), possible similar roles for fungal GSLs could be considered.

Two subclasses of GSLs appear to be widely expressed in the fungal kingdom: (i) monohexosylceramides (CMH) such as glucosylceramide (GlcCer) and galactosylceramide (GalCer) and (ii) glycosylinositol phosphorylceramides (GIPCs), which comprise a structurally diverse class of anionic and zwitterionic fungal (and plant) biosynthetic products (Steiner et al. 1969Go; Smith and Lester 1974Go; Barr and Lester 1984Go; Barr et al. 1984Go; Levery et al. 1998Go, 2000Go; Jennemann et al. 1999Go, 2001Go; Loureiro y Penha et al. 2001Go; Toledo, Levery, Glushka, et al. 2001Go; Toledo, Levery, Straus, et al. 2001Go; Heise et al. 2002Go; Bennion et al. 2003Go; Aoki et al. 2004Go), capable of bearing nonreducing terminal Galpß1 -> residues (Barr et al. 1984Go; Jennemann et al. 2001Go). Interestingly, the ceramides found in CMHs of most fungi are structurally distinct from those found in GIPCs, suggesting that the corresponding lipid acceptor substrates are [with few exceptions (Barreto-Bergter et al. 2004Go)] synthesized in divergent pathways adapted for different physiological functional roles (Toledo et al. 1999Go; Toledo et al. 2000Go; Toledo, Levery, Suzuki, et al. 2001Go; Leipelt et al. 2001Go; Levery et al. 2002Go; Warnecke and Heinz 2003Go; compare ceramide structural features in Schemes 1 and 2). Mushrooms and other basidiomycetes that have so far been examined appear to express GlcCer, and not GalCer, with typical cerebroside ceramide features widespread in the fungal kingdom. In addition, GIPCs of most basidiomycetes have been found to possess complex glycans which are highly diversified in structure, but based on a characteristic core motif, Galß1 -> 6Man{alpha}1 -> 2Ins-P-Cer (Ps-2) (where Ins = myo-inositol, P = phosphodiester, Cer = ceramide; Scheme 1, Ba-2). This core has not been found in yeast such as Saccharomyces cerevisiae or Candida albicans, nor in filamentous fungi (e.g. euascomycete species); so far the most widely distributed GIPC core in euascomycetes has been found to be Man{alpha}1 -> 3Man{alpha}1 -> 2Ins-P-Cer (Scheme 1, Eu-2), along with less frequently observed alternate core structures based on Man{alpha}1 -> 6Ins-P-Cer and GlcN{alpha}1 -> 2Ins-P-Cer (Levery et al. 2000Go; Loureiro y Penha et al. 2001Go; Toledo, Levery, Glushka, et al. 2001Go; Toledo, Levery, Straus, et al. 2001Go; Aoki et al. 2004Go).


Figure 7
View larger version (22K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Scheme 1. Biosynthesis of fungal GIPCs starting with transfer of myo-inositol-1-O-phosphate (IP) from the diacylglycerol moiety of phosphatidylinositol to ceramide, catalyzed by-AUR1 gene-encoded IPC synthase. Common intermediate Man{alpha}1 -> 2InsPCer (MIPC) is next synthesized by action of Man{alpha}2-T (encoded by SUR1/CSG1, CSG2, or both). In S. cerevisae, IPT1-encoded protein transfers a second mole of IP from phosphatidylinositol to MIPC to make PIM{alpha}2IPC (Lester and Dickson 1993Go; Dickson and Lester 2002Go). In many filamentous fungi, an as yet unknown Man{alpha}3-T (X) transfers a second {alpha}-Man residue to MIPC to make a common intermediate M{alpha}3M{alpha}2IPC (right arrow). In most basidiomycetes, including mushrooms and Cryptococcus neoformans, an as yet unknown Galß6-T (Y) transfers a ß-Gal residue to MIPC to make a common intermediate Galß6M{alpha}2IPC (left arrow).

 

Figure 8
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Scheme 2. Structure of prototypical fungal GlcCer, (4E,8E)-N-2''-hydroxyhexadecanoyl-1-ß-D-glucopyranosyl-9-methyl-4,8-sphingadienine, along with fragmentation nomenclature of Costello and Vath (Costello and Vath 1990Go) as modified by Adams and Ann (1993)Go.

 
In this work, we extracted and characterized the major cerebroside and GIPC components of mature fruiting body of P. squamosus using one- and two-dimensional nuclear magnetic resonance (NMR) spectroscopy; electrospray ionization–mass spectrometry (ESI-MS, -MS/CID-MS, and -pseudo-[CID-MS]2) in a hybrid quadrupole/time-of-flight (ESI-Qq/oa-TOF-MS) instrument; multistep fragmentation in an ion trap electrospray-ionization ion trap–mass spectrometry (ESI-IT-MSn) instrument; and gas chromatography–mass spectrometry (GC–MS).


   Results
Top
 Abstract
 Introduction
 Results
Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
HPTLC profile and fractionation of P. squamosus GIPCs
Reproduced in Figure 1 is an high-performance thin-layer chromatography (HPTLC) showing crude (lane C) and high-performance liquid chromatography (HPLC) fractionated (lanes 2–7) acidic lipids obtained from P. squamosus, stained with orcinol for detection of hexose-containing material. Three abundant components [designated Man{alpha}1 -> 2Ins1-P-1Cer (Ps-1), Galß1 -> 6Man{alpha}1 -> 2Ins1-P-1Cer (Ps-2), and Man{alpha}1 -> 3Fuc{alpha}1 -> 2Gal{alpha}1 -> 6Galß1 -> 6Man{alpha}1 -> 2Ins1-P-1Cer (Ps-5)] were observed in the crude preparation (lane C) along with one or more less abundant components. The retention factor (Rf) value of the fastest migrating component was found to be identical to that of an authentic basidiomycete-derived Man{alpha}2IPC [MIPC; from Agaricus blazei (Toledo, Levery, Glushka, et al. 2001Go); data not shown]. HPLC purification yielded two major components appearing as single orcinol-stained bands (lanes 2 and 5), and partial purification of the third major component (lane 3). The upper band in lane 3 had an Rf identical to that in lane 2, but was of lower abundance (approximately 20–30%). The differences in Rf values were consistent with the increasing numbers of sugar residues in each component. The major components in the fractions corresponding to the lanes marked 2, 3, and 5 were subjected to characterization by GC–MS, MS, and NMR methods as described below.


Figure 1
View larger version (60K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. HPTLC analysis (orcinol stain) of acidic lipids (GIPCs) from P. squamosus. Mobile phase, chloroform/methanol/water (50:47:14 v/v/v, containing 0.038% w/v CaCl2). Lane C, crude acidic fraction from P. squamosus. Lanes 2–7, fractions eluted from HPLC of crude P. squamosus acidic lipids (40 x 2 mL fractions collected at 0.5 mL/min). Lane 2; Frs 27–30; 3, Fr31; 4, Fr 32; 5, Frs 33–34; 6, Fr 35; and 7, Fr 36.

 
Monosaccharide, inositol, fatty acid, and sphingoid component analysis of P. squamosus GIPCs
Following methanolysis of the three putative GIPC fractions of P. squamosus, i.e. Ps-1, Ps-2, and Ps-5, GC–MS analysis of monosaccharides as their trimethylsilyl (TMS) methyl glycosides (not shown) showed Ps-1 to consist of mannose as a dominant sugar component, consistent with an MIPC. Ps-2 yielded mannose and galactose in an approximately 2:1 ratio, consistent with a mixture of MIPC and an MIPC derivative with a galactose residue added (GalMIPC) in an approximately 1:2 ratio. Ps-5 yielded fucose, galactose, and mannose in an approximately 1:2:2.6 ratio, suggesting a GIPC with 1 fucose, 2 galactose, and 2–3 mannose residues, consistent with the substantially decreased Rf value for this fraction. Glucose was detected in all the GIPC fractions, however, based on subsequent MS and NMR analysis, its source was considered to be a non-GIPC impurity. Myo-inositol was also detected in all the GIPC fractions. The GC–MS analysis of the fatty acid methyl esters (FAMEs) and N-acetyl-sphingoid-O-TMS derivatives exhibited abundant peaks for 2-hydroxy fatty acids and phytosphingosines. Ceramides of Ps-1, Ps-2, and Ps-5 were found to be composed primarily of (t18:0) 4-hydroxy-sphinganine N-acylated with h22:0, h23:0, and h24:0 in all the fractions. The GC retention times and electron ionization (EI)-mass spectra corresponding to each of the peaks were identical to those of authentic standards. The EI spectrum of the N-acetyl-1,3,4-tri-O-trimethyl-(t18:0)-4-hydroxy derivative had peaks corresponding to [M-15]+, [M-15–90]+, [M-59 – 90]+, [M-174]+, and [M-174 – 90]+, ions observed at m/z 560, 470, 426, 401, and 311, respectively (Thorpe and Sweeley 1967Go; Levery et al. 1998Go). The long-chain monohydroxy fatty acids had EI spectra with peaks corresponding to [M]+, [M-15]+, and [M-59]+ ions similar to previously reported data (Laine et al. 1974Go). Also detected in Ps-1 was a trace of dihydroxy fatty acid (dh24:0). The presence of dh24:0 was confirmed using FAME derivatives of the threo and erythro forms of 2,3-dihydroxy palmitic acid (dh16:0) as standards. Although we were unable to confirm the precise stereochemistry based on retention times, the EI spectrum of dh24:0 qualitatively resembled that of dh16:0, but with the addition of 112 atomic mass units to the major ions observed. The spectrum was distinctly different from that of the monohydroxy fatty acids, with ions corresponding to [M-15]+, [M-59]+, [M-15 – 90]+, and [M-161]+ observed at m/z 543, 499, 453, and 397, respectively [compare results of Heise et al. (2002)Go and Mayberry (1981)Go].

Glycosyl linkage analysis by GC–MS
The glycosyl linkage structures of Ps-1, Ps-2, and Ps-5 were determined by GC–MS of partially methylated alditol acetate derivatives (PMAAs) after permethylation of the ammonia released glycosylinositols, followed by depolymerization, reduction and peracetylation. The structure of Ps-1 was confirmed by the detection of a derivative for a terminal mannose (t-Manp) residue (1,5-di-O-acetyl-1-deuterio-2,3,4,6-tetra-O-methyl-mannitol), consistent with an MIPC. The linkage analysis of Ps-2 yielded derivatives for both terminal galactose (t-Galp) and t-Manp residues (1,5-di-O-acetyl-1-deuterio-2,3,4,6-tetra-O-methyl-galactitol and 1,5-di-O-acetyl-1-deuterio-2,3,4,6-tetra-O-methyl-mannitol, respectively), as well as a derivative for a 6-linked Manp residue (1,5,6-tri-O-acetyl-1-deuterio-2,3,4-tri-O-methyl-mannitol). This is consistent with a mixture of MIPC and Gal -> 6MIPC. The linkage analysis of Ps-5 was characterized by the detection of a 6-linked Manp residue, a t-Manp, a 2-linked Galp (1,2,5-tri-O-acetyl-1-deuterio-3,4,6-tri-O-methyl-galactitol), and a 6-linked Galp (1,5,6-tri-O-acetyl-1-deuterio-2,3,4,-tri-O-methyl-galactitol). Interestingly, a 3-linked fucose derivative (1,3,5-tri-O-acetyl-1-deuterio-6-deoxy-2,4-di-O-methyl-galactitol) was also detected. The absence of any branching residue derivatives suggested that Ps-5 has an unbranched structure, wherein the 6-linked Manp is assumed to originate from an MIPC core, while the t-Manp originates from a mannose residue at the nonreducing terminal. The presence of an internal sequence comprised of two galactoses and a fucose residue was also indicated, consistent with the monosaccharide analysis.

+ESI-MS and -MS/CID-MS analysis of P. squamosus GIPCs
Ps-1
Major molecular adduct species [M(Li) + Li]+ in the ESI-Q/oa-TOF-MS profile of Ps-1 were observed at m/z 1100, 1086, and 1072, consistent with monohexosyl-IPCs having ceramides containing t18:0 4-hydroxysphinganines with h24:0, h23:0, and h22:0 fatty N-acylation, respectively (spectrum not shown). The +ESI-MS/CID spectra acquired from the dilithiated molecular adduct at m/z 1072 and 1100, respectively, were similar to previous results obtained from a mushroom MIPC using a triple quadrupole instrument (Bennion et al. 2003Go), exhibiting pairs of nonphosphorylated fragments representing the nonreducing end of the glycosylinositol observed at m/z 169 and 187 [Hex + Li]+ ([B1 + Li]+ and [C1 + Li]+, respectively), and at m/z 331 and 349 [HexbulletIns + Li]+ ([B2 + Li]+ and [C2 + Li]+, respectively). A pair of glycosylinositol phosphate fragments [HexbulletInsbulletP + Li2]+ were observed at m/z 417 and 435 ([B2PO3(Li) + Li]+ and [C2PO3(Li) + Li]+, respectively). Ceramide-related product ions from the molecular ion m/z 1072 included m/z 644 ([Z0 + Li]+), 662 ([Y0 + Li]+), 730 ([Z0PO3(Li) + Li]+), and 748 ([Y0PO3(Li) + Li]+), consistent with a t18:0/h22:0 ceramide component. A uniform increment of m/z 28 was observed for the ceramide-related products of the m/z 1100 molecular ion, i.e. 672 ([Z0 + Li]+), 690 ([Y0 + Li]+), 758 ([Z0PO3 (Li) + Li]+), and 776 ([Y0PO3(Li) + Li]+), consistent with a t18:0/h24:0 ceramide component. These cleavages are summarized in Scheme 3 and Tables I and II.


Figure 9
View larger version (27K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Scheme 3. Characteristic fragmentations of glycosylated IPCs, represented by MIPC (Ps-1) under conditions of positive ion ESI-Qq/oa-TOF-MS. Panel A, nomenclature of Adams and Ann (1993)Go for fragmentation of the ceramide moiety, and of Costello and Vath (1990)Go and Singh et al. (1991)Go for glycosylinositol phosphoryl group; Panel B, sphingoid the d3b ion, proposed by Hsu and Turk (2001)Go; Panel C, hydrated analog of the d3b ion, proposed in this study as a product of t18:0 or t20:0 phytosphingosine-containing ceramides; Panel D, sphingoid c1b ion, proposed by Hsu and Turk (2001)Go; Panel E, hydrated analog of c1b ion, proposed in this study as a product of t18:0 or t20:0 phytosphingosine-containing ceramides. The adduct designation "+Li" and the charge form have been omitted from the labels for clarity.

 

View this table:
[in this window]
[in a new window]

  		Table I. Glycosylinositol- and phosphorylglycosylinositol-related product ions formed in low-energy ESI-MS/CID-TOF-MS spectra of mono-, di-, and pentaglycosylinositol phosphorylceramides from fruiting body of P. squamosus (Ps). Base peak distinguished by bold font. Fragment designations as in Costello and Vath (1990)Go, as expanded for GIPCs in Singh et al. (1991)Go, illustrated in Schemes 3–5

 

View this table:
[in this window]
[in a new window]

  		Table II. Molecular adduct ions [M(Li) + Li]+ observed in ESI-Q/-TOF-MS profile and significant ceramide-containing product ions (m/z) formed in Q-TOF ESI-MS/CID-TOF-MS spectra of mono-, di-, and pentaglycosylinositol phosphorylceramides from fruiting body of P. squamosus (Ps). Fragment designations as in Costello and Vath (1990)Go, as expanded for GIPCs in Singh et al. (1991)Go, with modifications of Adams and Ann (1993)Go for ceramide-derived ions; illustrated in Schemes 3–5

 
Ps-2
Two sets of molecular adducts ions were observed in the [M(Li) + Li]+ profile of Ps-2 obtained by ESI-Q/oa-TOF-MS. In addition to a group of ions at m/z 1100, 1086, and 1072, consistent with the same set of monohexosyl-IPC species observed in Ps-1 (data not shown), an abundant second group of [M(Li) + Li]+ species could be observed at m/z 1262, 1248, and 1234 (Figure 2A). These values correspond to addition of a uniform m/z 162 increment to the first group of ions, consistent with addition of a second hexose residue to Ps-1. The +ESI-MS/CID spectra from the dilithiated molecular adduct at m/z 1262, reproduced in Figure 2B, exhibit an abundant pair of glycosylinositol phosphate fragments at m/z 579 and 597, corresponding to [B3PO3(Li) + Li]+ and [C3PO3(Li) + Li]+, respectively. Pairs of ions at m/z 417 and 435, and at m/z 255 and 273, are consistent with additional glycosidic cleavages within the glycosylinositol phosphate ion, corresponding to [Y2/B3PO3(Li) + Li]+ and [Y2/C3PO3 (Li) + Li]+, and to [Y1/B3PO3(Li) + Li]+ and [Y1/C3PO3(Li) + Li]+, respectively (see Scheme 4). Nonphosphorylated glycosylinositol fragments include m/z 493 and 511, corresponding to [B3 + Li]+ and [C3 + Li]+ ions, respectively. Other fragments include m/z 331 ([Y2/B2 + Li]+) and 349 ([Y2/C2 + Li]+), and ceramide-related ions 672 ([Z0 + Li]+), 690 ([Y0 + Li]+), 758 ([Z0PO3 (Li) + Li]+), and 776 ([Y0PO3(Li) + Li]+), again consistent with a t18:0/h24:0 ceramide component. The data are summarized in Tables I and II, Schemes 3 and 4. Together with the glycosyl composition and linkage analysis, these data are consistent with a Gal -> 6MIPC structure for Ps-2, with the fraction also containing a significant amount of MIPC (Ps-1), and with both having similar ceramide component profiles.


Figure 2
View larger version (18K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Positive ion ESI-Qq/oa-TOF-MS of Fr 31 (Ps-2). (A) Molecular ion profile as Li adducts. (B) MS/CID-MS of [M(Li) + Li]+ at m/z 1262 (low m/z region).

 

Figure 10
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Scheme 4. Fragmentation of GIPC Ps-2 in positive ion ESI-Qq/oa-TOF-MS; ion designations according to Costello and Vath (1990)Go and Singh et al. (1991)Go. The adduct designation "+Li" and the charge form have been omitted from the labels for clarity.

 
Ps-5
The major molecular adduct species [M(Li) + Li]+ in the ESI-Q/oa-TOF-MS profile of Ps-5 (Figure 3A) are observed at m/z 1732, 1718, and 1704, corresponding to the set of Ps-2 molecular adduct ions uniformly incremented by two hexose and one deoxyhexose residues (2 x m/z 162 + m/z 146 = m/z 570). Reproduced in Figure 4B and C, are +ESI-MS/CID spectra of the dilithiated molecular adduct at m/z 1704, acquired at low- and high-collision energy, respectively. In these spectra, the fragment at m/z 1067 is consistent with an intact glycosylinositol phosphate moiety having the expected composition [deoxyHexbulletHex4bulletInsbulletP + Li2]+ (corresponding to [C6PO3(Li) + Li]+). Glycosidic cleavages within the glycosylinositol phosphate fragment are represented by loss of a hexose residue (m/z 162 decrement), yielding an ion m/z 905 ([Y5/C6PO3(Li) + Li]+); loss of a hexose and a deoxyhexose residue (m/z 162 + m/z 146 = m/z 308 decrement), yielding a highly abundant ion m/z 759 ([Y4/C6PO3(Li) + Li]+); and further sequential losses of residues along the chain, yielding ions at m/z 597 ([Y3/C6PO3(Li) + Li]+), m/z 435 ([Y2/C6PO3(Li) + Li]+), and m/z 273 ([Y1/C6PO3(Li) + Li]+) (see Scheme 5). The corresponding [B6PO3(Li) + Li]+ fragment m/z 1049 undergoes analogous sequence-specific losses of hexose or deoxyhexose residues to yield ions at m/z 741 ([Y4/B6PO3(Li) + Li]+), m/z 579 ([Y3/B6PO3(Li) + Li]+), and m/z 255 ([Y1/B6PO3(Li) + Li]+). Other significant ions include unphosphorylated fragments observed at m/z 1542 ([Y5 + Li]+), m/z 1396 ([Y4 + Li]+), m/z 1234 ([Y3 + Li]+), m/z 963 ([B6 + Li]+), and m/z 981 ([C6 + Li]+). Summarized in Tables I and II, and Schemes 3 and 5, these data are consistent with a pentaglycosyl-IPC structure for Ps-5, which must have the linear sequence Hex-O-deoxyHex-O-Hex-O-Hex-O-Hex-O-InsPCer to be consistent with the results of the glycosyl linkage analysis. The terminal hexose must be a mannose residue (t-Man) based on linkage analysis. The abundant fragments at m/z 1396, 759, and 741, corresponding to combined losses of a hexose and a deoxyhexose residue (m/z 308 decrements from the molecular adduct ion and the two glycosylinositol phosphate ions) imply that the ->3Fuc residue identified in the linkage analysis represents the penultimate residue to which the t-Man is attached. Given the likely assumption that the other mannose residue identified in the linkage analysis (->6Man) is directly linked to IPC to form a ManIPC core, the next two internal hexoses in the sequence must be the ->2Gal and ->6Gal residues observed in the linkage analysis, one of which must be linked 1 -> 6 to the ManIPC core. The order of the ->2Gal and ->6Gal residues is not further specified by any of the data so far. Overall, the data are consistent with the partial sequence Man -> 3Fuc -> 2/6Gal -> 6/2Gal -> 6ManIPC. The remaining features of the sequence were supplied by further ESI-MSn analysis of the permethylated glycosylinositol derived from Ps-5, and acquisition of NMR spectroscopic data for all three GIPC fractions.


Figure 3
View larger version (26K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. Positive ion ESI-Qq/oa-TOF-MS of Fr 33-34 (Ps-5). (A) Molecular ion profile as Li adducts. (B) MS/CID-MS of [M(Li) + Li]+at m/z 1704, low-collision energy (57 eV), and (C) MS/CID-MS of [M(Li) + Li]+at m/z 1704, high-collision energy (78 eV).

 

Figure 4
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. Positive ion ESI-MSn analysis of the partially depolymerized and permethylated glycosylinositol from Ps-5. (A) MS1 profile of sodiated adduct ions in the mixture of depolymerized products, exhibiting molecular species resulting from progressive loss of monosaccharide residues from the nonreducing end of the glycosylinositol moiety. Inset: glycosylinositol residue compositions corresponding to sodiated adducts. (B) MS2 product ion spectrum of products from precursor adduct m/z 899, which contains the triglycosyl inositol sequence Gal-O-Gal-O-Man-O-Ins. Inset: principle glycosidic cleavage products observed in the MS2 spectrum. (C) MS3 product ion spectrum of the nonreducing terminal disaccharide fragment m/z 445 (m/z 899 -> 445 ->). Inset: scheme showing origin of principle fragments correlating with a 1 -> 6 linked dihexoside.

 

Figure 11
View larger version (14K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Scheme 5. Fragmentation of GIPC Ps-5 in positive ion ESI-Qq/oa-TOF-MS; ion designations according to Costello and Vath (1990)Go and Singh et al. (1991)Go. The adduct designation " + Li" and the charge form have been omitted from the labels for clarity.

 
Positive ion mode ESI-IT-MSn of permethylated glycosylinositols
The glycosylinositol moiety was released from Ps-5 by high temperature ammonolysis, permethylated, and analyzed by ion trap ESI-MSn (Ashline et al. 2005Go). An MS1 profile of the sample is reproduced in Figure 4A. Interestingly, a pattern of sodiated molecular ions was obtained which would normally be consistent with a hydrolytic "sequence ladder" obtained by brief acid treatment before permethylation, although no acidic pretreatment was actually applied. This tends to contradict earlier reports that ammonolysis results in the release of intact glycosylinositols without significant effects on glycosidic linkages (Barr et al. 1984Go; Levery et al. 1998Go; Heise et al. 2002Go); regardless of the source of degradation, however, the result was quite fortuitous in this case, as it allowed isolation and fragmentation of each pair of sugar residues in the overall sequence as a permethylated nonreducing terminal disaccharide. In this profile, an abundant molecular adduct at m/z 1277 represents the intact pentaglycosylinositol, while sequentially ordered hydrolysis products, each with the reducing end inositol residue retained, and one or more nonreducing end residues lost, are observed at m/z 1073, 899, 695, and 491. Other hydrolysis products, such as segments that include loss of the inositol residue, are apparent in the profile, but are not useful for further sequence analysis, since it is not possible to distinguish unambiguously between the reducing and nonreducing ends in subsequent MSn steps. On the other hand, it is possible to tell which end of the chain bears the inositol after permethylation because its residue mass is no longer identical to that of a hexose (as it would be in the underivatized state).

MSn analysis of the sodiated molecular ions at m/z 899 and 1073 proved to the most useful for determining the unspecified linkage positions, since these clearly included the relevant residues as terminal disaccharides (analysis of internal linkages can be performed in principle, but in some cases the relevant internal disaccharide ion is not generated in sufficient abundance). The molecular adduct at m/z 899, which represents the sequence Hex-O-Hex-O-Hex-O-Ins, must include the two Gal residues that were linked to each either internally in Ps-5, but are now in the nonreducing terminal disaccharide. It remains only to determine whether this linkage is Gal -> 2Gal or Gal -> 6Gal to complete the sequence except for the monosaccharide anomeric configurations. In an MS2 spectrum of m/z 899 (Figure 4B), the product ion m/z 445 represents the terminal disaccharide-1-ene fragment (Figure 4B, Inset). The MS3 spectrum of products from the m/z 445 intermediate (Figure 4C), exhibits a pattern of fragments that could only arise from a Hex -> 6Hex disaccharide (Figure 4C, Inset) (Ashline et al. 2005Go), specifying the origin of the ->6Gal derivative in the linkage analysis. An essentially identical pattern was obtained by an MS4 pathway proceeding initially via the trisaccharide-1-ene fragment at m/z 649, i.e. m/z 899 -> 649 -> 445-> (spectrum not shown). From this result it can be deduced that the ->2Gal observed in the linkage analysis is derived from the other unspecified linkage, which must be Fuc -> 2Gal. A similar MSn analysis of the molecular ion at m/z 1073 confirmed the Fuc -> 2Gal sequence for its terminal disaccharide (results not shown). The monosaccharide linkage sequence of Ps-5 can therefore be proposed as Man -> 3Fuc -> 2Gal -> 6Gal -> 6ManIPC.

1H-NMR spectroscopic analysis of P. squamosus GIPCs
Ps-1
A downfield expansion of the one-dimensional 1H-NMR spectrum of Ps-1 is reproduced in Figure 5 A. Key salient features are the lone H-1 resonance, typical for {alpha}-Man of a fungal GIPC, observed at 5.041 ppm; a typical sphingoid H-1b resonance at 4.052 ppm, a characteristic chemical shift for this signal in a GIPC (the combined influences of phytosphingosine, 2-hydroxy fatty N-acylation, and 1-O-phosphorylation); and a set of ring proton resonances, with characteristic chemical shift and splitting patterns, that are often observable for the myo-Ins moiety of GIPCs, at 3.956 (H-2), 3.221 (H-3), and 2.948 ppm (H-5). Taken together, these features were strongly similar to those previously observed for MIPCs of basidiomycetes and euascomycetes (Jennemann et al. 1999Go, 2001Go; Toledo, Levery, Glushka, et al. 2001Go). This superficial analysis was confirmed by complete assignment of all 1H and 13C resonances using homo- and heteronuclear two-dimensional NMR spectroscopic analysis [1H–1H-gradient-selected correlation spectroscopy (gCOSY), 1H–1H-total correlation spectroscopy (TOCSY), and 1H-detected 1H-13C gradient-selected heteronuclear single-quantum coherence (gHSQC), spectra not shown; see Table III for list of assignments]. Analysis of the approximate relative magnitudes of three-bond 1H–1H coupling constants around the monosaccharide and inositol spin systems confirmed their stereochemical identities (Koerner et al. 1987Go). Particularly compared with our own data previously acquired under identical conditions (Toledo, Levery, Glushka, et al. 2001Go), all chemical shifts were essentially identical to those observed for MIPC, and not from Man{alpha}6IPC. Key diagnostic evidence for the linkage position of the {alpha}-Man residue is provided by the significant downfield shift of myo-Ins C-2 (Toledo, Levery, Glushka, et al. 2001Go).


Figure 5
View larger version (35K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5. Downfield anomeric proton regions of one-dimensional 1H-NMR spectra (800 MHz; DMSO-d6/2% D2O; 35 ºC) of HPLC purified GIPC fractions isolated from P. squamosus corresponding to lanes 2, 3, and 5 of Fig. 1. (A): Ps-1 (lane 2); (B) Ps-2 (lane 3); *resonances from residual Ps-1; (C) Ps-5 (lane 5).

 

View this table:
[in this window]
[in a new window]

  		Table III. 1H and 13C chemical shifts (ppm) for monosaccharide, inositol, ceramide sphingoid and fatty acyl (in parentheses) residues of Galß6Man{alpha}2IPC (Ps-2) and MIPC (Ps-1) from P. squamosus, in DMSO-d6/2% D2O at 308 °K (35 °C). 3J1,2 (Hz) for monosaccharide residues are given in parentheses. 13C chemical shifts for signals at site of glycosidic linkage given in bold italic font. 13C chemical shifts ± 0.1 ppm unless otherwise indicateda

 
Ps-2
Analysis of the downfield expansion of the one-dimensional 1H–NMR spectrum of the Ps-2 fraction (Figure 5B) supported the conclusion above that it is a mixture composed of MIPC and another component derived from it by addition of a second glycosyl residue, in an approximately 1:2 ratio. In addition to a new H-1 resonance at 4.265 ppm, with a splitting constant indicative of a ß-configuration (3J1,2 = 7.6 Hz), resonances with chemical shifts and splitting patterns identical to those in MIPC can be observed in the spectrum (asterisks), accompanied by identically split resonances at somewhat different chemical shifts, e.g. {alpha}-Man H-1 at 3.974 ppm ({Delta}{delta} = –1.067 ppm), myo-Ins H-2 at 4.002 ppm ({Delta}{delta} = +0.037 ppm), and myo-Ins H-3 at 3.228 ppm ({Delta}{delta} = +0.007 ppm). Since the glycosyl composition and linkage analysis indicate the additional monosaccharide in the derivative to be a nonreducing terminal Gal residue, and the Man residue to be ->6Man, the structure of Ps-2 can be tentatively proposed as Galß6Man{alpha}2IPC. This was confirmed unambiguously, using a two-dimensional NMR strategy as above for Ps-1, by complete assignment of all 1H and 13C resonances in the spin systems of both the residual Ps-1 and the major Ps-2 component (see Table III for list of assignments), followed by determination of interglycosidic connectivities in Ps-2 by two-dimensional 1H-1H-nuclear overhauser enhancement spectroscopy and 1H-detected 1H-13C gradient-selected heteronuclear multiple bond coherence (gHMBC) (spectra not shown). Despite severe overlap in some cases, all of the resonances crucial to the structure elucidation were readily assignable. Key additional features confirming the Galß -> 6Man linkage in Ps-2 are the major downfield shift changes for Man H-6, H-6', H-5, and C-6 ({Delta}{delta} = +0.061, +0.356, +0.289, and 7.0 ppm, respectively); and observation of three bond 1H-13C J-couplings through the glycosidic bond correlating Gal H-1/Man C-6 and Gal C-1/Man H-6/H-6'. Taken together, these data confirm the structure of Ps-2 as Galß6Man{alpha}2IPC, a compound previously observed in the majority of basidiomycetes examined so far, either unmodified or serving as the core structure for more complex GIPCs (Jennemann et al. 1999Go, 2001Go; Heise et al. 2002Go).

Ps-5
The downfield expansion of the one-dimensional 1H-NMR spectrum of the Ps-5 fraction (Figure 5C) indicated that it consists essentially of a single GIPC component, and the data appeared superficially consistent with a glycosylinositol moiety composed of five monosaccharide residues attached to myo-Ins. Observation of a signal for a ß-configured H-1 at 4.283 ppm (3J1,2 = 7.1 Hz) suggests inclusion of a Galß6Man{alpha}2IPC core sequence, although there appear to be two candidate signals for the Man{alpha}2 H-1 observable at 4.983 and 4.965 ppm. Two additional H-1 signals with splitting constants indicative of {alpha}-configured glycosyl residues are apparent at 4.821 and 4.723 ppm. Taken together, the ESI-MS data, the glycosyl composition and linkage analysis, and the anomeric configurations indicated by the one-dimensional NMR spectrum, suggest a complete structure for Ps-5, Man{alpha}3Fuc{alpha}2Gal{alpha}6Galß6Man{alpha}2IPC.

Anomeric proton assignments and further confirmation of the sequence were provided by assignment of nearly all proton resonances by two-dimensional homonuclear NMR spectroscopy (spectra not shown), as above for Ps-1 and Ps-2 was readily accomplished (Table IV); unfortunately, the amount of material was insufficient to obtain a high-quality 1H–13C gHSQC spectrum sufficient for assigning all 13C resonances in the complex glycan (partial assignments are listed in Table IV), nor could a useful 1H–13C gHMBC spectrum be obtained. Nevertheless, the data were sufficient to confirm the identity and anomeric configurations of the monosaccharide residues, as well as most of the linkage sites.


View this table:
[in this window]
[in a new window]

  		Table IV. 1H and 13C chemical shifts (ppm) for monosaccharide, inositol, ceramide sphingoid and fatty-N-acyl (in parentheses) residues of Man{alpha}3Fuc{alpha}2Gal{alpha}6Galß6Man{alpha}2IPC (Ps-5) from P. squamosus in DMSO-d6/2% D2O at 35 °C. 3J1,2 (Hz) for monosaccharide residues are given in parentheses. ND = not determined

 
A discrete seven-proton spin system was observed to connect the H-1 resonance at 4.965 ppm with a set of downfield shifted H-5, H-6, and H-6' resonances at 4.101, 3.540, and 3.925 ppm, respectively; this is consistent with assignment of the spin system to the internal ->6Man{alpha} residue, while the spin system starting with the H-1 at 4.983 ppm could be assigned to the nonreducing terminal Man{alpha} residue. The H-1 resonance at 4.273 ppm was found to be part of an eight proton {alpha}-Fuc spin system that included a characteristic H-5 broadened quartet at 3.961 ppm coupled to an H-6 methyl group doublet at 1.074 ppm. The difficulty in following the connectivity from the H-1 resonance at 4.283 ppm further than H-4 is typical for ß-Gal residues, due to the vanishingly small coupling between H-4 and H-5 in this system. The remaining H-1 resonance at 4.821 ppm can be assigned to the {alpha}-Gal residue, by the process of elimination, but also based on analysis of the magnitudes of three-bond 1H–1H coupling constants between all ring protons in the spin system.

Linkage sites for most of these residues could be confirmed by the relative downfield shifts of 13C resonances assigned from the 1H–13C gHSQC spectrum (denoted by bold italics in Table IV, for ->3Fuc{alpha} C-3, ->2Gal{alpha} C-2, ->6Man{alpha} C-6, and ->2Ins C-2); the sole exception was the ->6Galß linkage, due to the lack of unambiguous connectivities to the H-5 and H-6 resonances of this residue, and by extension to find the expected downfield shifted C-6 resonance. Fortunately, this linkage is supported by the process of exclusion (only the Galß residue can account for the appearance of ->6Gal in the permethylation linkage analysis), as well as by the +ESI-MSn analysis of the permethylated Gal-O-Gal-O-Man-O-Ins fragment. Thus a complete monosaccharide sequence for Ps-5, Man{alpha}3Fuc{alpha}2Gal{alpha}6Galß6Man{alpha}2IPC, is consistent with all data.

1H-NMR spectroscopic analysis of P. squamosus CMH
The downfield expansion of the one-dimensional 1H-NMR spectrum of the CMH fraction (spectrum not shown) exhibited a set of resonances with chemical shifts and coupling patterns characteristic of the seven-proton ß-glucopyranosyl spin system, observed at 4.125 (H-1), 2.958 (H-2), 3.143 (H-3), 3.039 (H-4), 3.091 (H-5), 3.433 (H-6a), and 3.662 ppm (H6b). Additional resonances were consistent with a (4E, 8E)-9-methyl-4,8-sphingadienine structure, with a characteristic pair of vinyl resonances (5.403 and 5.597 ppm) for the (E)-{Delta}4-unsaturation, and the single vinyl resonance (5.097 ppm) for the (E)-{Delta}8-unsaturation where C-9 is substituted with a methyl group (1.542 ppm, s, 3H). Resonances characteristic for (E)-{Delta}3-unsaturation of the 2-hydroxy fatty-N-acyl group were not observed. Altogether, these spectral results were essentially identical to those previously reported for GlcCer from the yeast forms of P. brasiliensis (Toledo et al. 1999Go) and H. capsulatum (Toledo, Levery, Suzuki, et al. 2001Go), and from a mutant strain of Neurospora crassa (Park et al. 2005Go), which express almost exclusively a saturated 2-hydroxy fatty-N-acyl group.

+ESI-MSn analysis of P. squamosus CMH
The major molecular ion species [M + Li]+ in the ESI-MS1 profile of the CMH fraction (Figure 6A, Inset) observed at m/z 734 is consistent with a lipoform having a h16:0/d19:2 fatty acyl/sphingoid combination. The MS2 spectrum acquired from this adduct reproduced in Figure 6A, is analogous to previously reported findings using low-energy tandem (Levery et al. 2000Go; Park et al. 2005Go) and ion trap (Levery 2005Go) instruments, with characteristic ions at m/z 572 and 480 representing the loss of a hexose residue (Y0 fragment) and the h16:0 fatty acid residue (O fragment), respectively (see Scheme 2). The h16:0 fatty-N-acyl group is included in the characteristic ion at m/z 304 (T {equiv} Z0/G fragment), although this fragment appears at much greater abundance in low-energy tandem MS/CID-MS spectra (Levery et al. 2000Go; Park et al. 2005Go) than in ion trap MSn spectra (Levery 2005Go). Further dissection of these primary precursors via MS3 confirmed their presumptive origins and structural significance. Key MS3 products of m/z 572 (Figure 6B) included the characteristic d19:2 sphingoid N ({equiv} Y0/G) ion m/z 318, and the h16:0 fatty-N-acyl W ion m/z 233. Key MS3 products of m/z 480 (Figure 6C) also included the characteristic d19:2 sphingoid N ion m/z 318, as well as hexose ions m/z 169 (B1) and 187 (C1), and a pair of fragments from losses of NH3 (m/z 463) and H2O (m/z 462) (the O fragment and the latter two products correspond to ions designated g1, g2, and g3, respectively, by Hsu and Turk (Hsu and Turk 2001Go).


Figure 6
View larger version (18K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6. Positive ion ESI-MSn analysis of purified GlcCer from P. squamosus. (A) MS1 lithium adduct ion profile showing a predominant adduct at m/z 734 (inset) and MS2 product ion spectrum of the selected precursor at m/z 734. (B) MS3 product ion spectrum of the selected Y0 fragment precursor at m/z 572 (m/z 734 -> 572 ->). (C) MS3 product ion spectrum of the selected O fragment precursor at m/z 480 (m/z 734 -> 480 ->).

 

   Discussion
Top
 Abstract
 Introduction
 Results
 Discussion
Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
In this work, both types of fungal GSLs, CMH, and GIPCs were extracted from the fruiting body of P. squamosus and characterized. The P. squamosus CMH was found to be GlcCer exclusively, as previously observed for basidiomycete species, with a typical basidiomycete structure characterized by the absence of fatty-N-acyl {Delta}3-unsaturation, the predominant ceramide having an h16:0/d19:2 fatty-N-acyl/sphingoid composition [for extended reviews of fungal GlcCer structure, biosynthesis, and function, see Warnecke and Heinz (2003)Go and Barreto-Bergter et al. (2004)Go]. The structures of three major GIPCs, Ps-1, Ps-2, and Ps-5, were elucidated, with the latter found to have a previously unreported pentaglycosylinositol headgroup structure. The ceramide moieties of the GIPCs had typical (phytoceramide) structures for these components found in other fungi; these are distinct from those found in most fungal GlcCer, although some exceptions have been found (Warnecke and Heinz 2003Go; Barreto-Bergter et al. 2004Go). The predominant GIPC sphingoid was found to be t18:0 4-hydroxysphinganine (phytosphingosine), in combination with h22:0 and h24:0 as the predominant long-chain fatty-N-acyl moieties. In Ps-1, dihydroxy fatty-N-acylation (dh24:0) was also detected as a minor component.

The GIPC headgroup structures appear consistent with a transparent sequential biosynthetic relationship. Ps-1 represents a common MIPC core structure found in many fungi (Barr et al. 1984Go; Levery et al. 1998Go, 2000Go; Jennemann et al. 1999Go, 2001Go; Heise et al. 2002Go; Bennion et al. 2003Go); Ps-2 represents modification of MIPC by addition of a Galß6 residue to form Galß6Man{alpha}2IPC (Ba-2; see Scheme 1), a core structure typical of almost all basidiomycete species so far characterized (Jennemann et al. 1999Go, 2001Go; Heise et al. 2002Go), while not being observed in other types of fungi, such as saccharomycetales (e.g. Saccharomyces and Candida spp.) or euascomycetes (filamentous or thermally dimorphic species, e.g. Aspergillus spp., Histoplasma capsulatum, Paracoccidioides brasiliensis, and Sporothrix schenckii). Ps-5 was found to have the structure Man{alpha}3Fuc{alpha}2Gal{alpha}6Galß6Man{alpha}2IPC, consistent with sequential elongation of Ps-2/Ba-2 with three additional monosaccharide residues. All three of these residues, Gal{alpha}6, Fuc{alpha}2, and Man{alpha}3, have been observed previously in mushroom GIPCs, but so far not in a strictly linear sequence as found for Ps-5.

Understanding fungal GSL functions, particularly in terms of their interactions with endogenous carbohydrate- or lipid-binding proteins at the molecular level, is at a very early stage. The recent discovery, cloning, structural and functional analysis of an authentic galectin homolog in the model basidiomycete Coprinopsis cinerea raised the possibility that GIPCs could be endogenous binding partners for galectins or other lectins during early development in mushrooms (Cooper et al. 1997Go; Boulianne et al. 2000Go; Walser et al. 2004Go). Mushroom GIPCs are rich in some of the same monosaccharide residues that are known to interact with galectins and mannose-binding proteins, such as {alpha}- and ß-Gal, {alpha}-Fuc, and {alpha}-Man (Jennemann et al. 1999Go, 2001Go). Interestingly, cryptic human blood group H and B determinants (Fuc{alpha}2Galß/{alpha}- and Fuc{alpha}2(Gal{alpha}3)Galß/{alpha}-, respectively) can be observed on GIPCs from some of these mushrooms. In general, the variety of glycosylinositol headgroup structures found so far in basidiomycetes suggests the potential operation of a "Red Queen Effect" in diversifying GIPC expression (Gagneux and Varki 1999Go). So far, however, little direct evidence has emerged for functional interactions between fungal carbohydrate-binding proteins and GIPCs, but only a limited number of investigations have been carried out, and further studies may reveal more about the functions not only of fungal GIPCs but of a plethora of known fungal lectins which so far have been more of interest as in vitro reagents than as endogenous effector molecules.

The lectin from P. squamosus strongly recognizes NeuAc{alpha}2 -> 6Galß1 -> terminated glycans; the lectin also recognizes, albeit more weakly, nonsialylated terminal Gal residues (Mo et al. 2000Go; Zhang et al. 2001Go). Ps-2/Ba-2, which makes up a significant fraction of P. squamosus GIPCs, does possess a terminal ß-Gal residue, suggesting it could potentially interact with P. squamosus or other fungal lectins. In the near future, we plan to study the possible interactions of fungal GIPCs with lectins from P. squamosus and other mushrooms. Such further investigations could provide insights on the role of GIPCs in the fungal life cycle and developmental biology.


   Materials and methods
Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
Conflict of interest statement
 Acknowledgments
 References
 
Fungal isolate and growth conditions
Carpophores of P. squamosus (Huds.) Fr. were collected in late summer 1998 from a decaying Ulmus stump in Ann Arbor, Michigan (Mo et al. 2000Go). Freeze-dried samples of the fruiting body were used as the source of GSLs for this study.

High-performance thin-layer chromatography
Analytical HPTLC was performed on silica gel 60 plates (E. Merck, Darmstadt, Germany) using chloroform/methanol/water (50:47:14 v/v/v, containing 0.038% w/v CaCl2; solvent A) as mobile phase for acidic lipids, and chloroform/methanol/water (60:40:9 v/v/v, containing 0.025% w/v CaCl2; solvent B) for neutral lipids. Acidic lipid samples were solubilized in isopropanol/hexane/water (55:25:20, v/v/v, upper phase discarded) and applied by streaking from 5 µL Micro-caps (Drummond, Broomall, PA). Neutral lipid samples were solubilized in chloroform/methanol (1:1, v/v) and applied similarly. Detection was made by Bial's orcinol reagent [orcinol 0.55% (w/v) and H2SO4 5.5% (v/v) in ethanol/water 9:1 (v/v); the plate is sprayed and heated briefly to approximately 200–250 °C].

Extraction and purification of GSLs
Extraction, fractionation into neutral and acidic lipids on diethylamino ethanol (DEAE)-Sephadex ion exchange chromatography, and purification of acidic GIPCs from the DEAE-Sephadex-retained fraction by preparative-scale HPLC, were carried out as described previously (Bennion et al. 2003Go). The identity and purity of each acidic fraction from HPLC was assessed by analytical HPTLC using solvent A as above, and those containing putative GIPCs were further assessed by one-dimensional 1H-NMR spectroscopy, before complete characterization by the full range of NMR and MS techniques described below. The glycolipid content of the neutral fraction (DEAE-Sephadex A-25 flow-through) was assessed by analytical HPTLC using solvent B; a band co-migrating with an authentic fungal GlcCer was isolated by preparative HPTLC using the same solvent as described (Toledo et al. 1999Go; Toledo et al. 2000Go; Toledo, Levery, Suzuki, et al. 2001Go; Levery et al. 2002Go; Park et al. 2005Go).

One-dimensional 1H and two-dimensional 1H–1H and 1H–13C NMR spectroscopy
Samples of underivatized GIPC (approximately 0.5 – 1.0 mg) were deuterium exchanged by repeated lyophilization from D2O; samples of GlcCer (approximately 0.5 – 1.0 mg) were deuterium exchanged by repeated evaporation from deuterated solvent B (CD3Cl–CD3OD 1:1 v/v). Samples were dissolved in 0.5 mL of dimethylsulfoxide (DMSO)-d6/2% D2O (Dabrowski et al. 1980Go) for NMR analysis as described for GIPCs (Levery et al. 1998Go; Toledo, Levery, Glushka, et al. 2001Go; Toledo, Levery, Straus, et al. 2001Go; Bennion et al. 2003Go) and GlcCer (Toledo et al. 1999Go; Toledo et al. 2000Go; Toledo, Levery, Suzuki, et al. 2001Go; Park et al. 2005Go). One-dimensional 1H-NMR; two-dimensional 1H–1H-gCOSY, 1H–1H-TOCSY, and proton-detected 1H–13C-gHSQC experiments were performed at 35 °C on Varian Unity Inova 500 MHz (Department of Chemistry, University of New Hampshire), 600 and 800 MHz (Complex Carbohydrate Research Center, University of Georgia) spectrometers, using standard acquisition software available in the Varian VNMR software package. Proton chemical shifts are referenced to internal tetramethylsilane ({delta} = 0.000 ppm). Carbon chemical shifts are referenced to solvent DMSO ({delta} = 40.0 ppm); since these were not obtained from directly detected one-dimensional spectra, but measured in F1 of the gHSQC experiment, values are given to one decimal place only.

Positive ion mode ESI-MS of underivatized GIPCs and GlcCer
MS of GIPCs was performed in the positive ion mode on a Micromass (Manchester, UK) hybrid ESI-Qq/oa-TOF-MS instrument, with GIPC sample introduction via direct infusion in 100% MeOH (approximately 100 ng/µL). A nano-spray capillary tip was employed, from which the flow rate was estimated to be approximately 200 nL/min. MS of GlcCer was also performed via positive ion mode ESI, but on a linear ion trap (LTQ, ThermoFinnigan San Jose, CA). Lithium iodide (10 mM in MeOH) was added to the analyte solution t