Originally published In Press as doi:10.1194/jlr.M300184-JLR200 on August 16, 2003
Journal of Lipid Research, Vol. 44, 2073-2088, November 2003
Copyright © 2003 by American Society for Biochemistry and Molecular Biology
Glycosphingolipids of the model fungus Aspergillus nidulans
:
characterization of GIPCs with oligo-
-mannose-type glycans
Beau Bennion*,
Chaeho Park
,
Matthew Fuller1,,
Rebecca Lindsey
,
Michelle Momany,
Richard Jennemann** and
Steven B. Levery2,*
* Department of Chemistry, University of New Hampshire, Durham, NH 03824-3598
The Complex Carbohydrate Research Center and Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602-7229
Department of Plant Biology, University of Georgia, Athens, GA 30602-7229
** Abteilung für Zelluläre und Molekuläre Pathologie, Deutsches Krebsforschungszentrum-Heidelberg, Im Neuenheimer Feld 280, D69120, Heidelberg, Germany
Published, JLR Papers in Press, August 16, 2003. DOI 10.1194/jlr.M300184-JLR200
1 Present address of M. Fuller: School of Biology, Georgia Institute of Technology, Atlanta, GA 30332. 
1 Although Barr, Laine, and Lester (65) did not specify precisely the linkage between Man and Ins in their compounds, but characterized it as Man1
2/6Ins, NMR spectroscopy of H. capsulatum GIPCs shows that the linkage is Man12Ins1-P-1Cer (S. B. Levery, M. S. Toledo, A. H. Straus, and H. K. Takahashi, unpublished observations), as reported for GIPCs of the closely related fungus, P. brasiliensis (45).
2 To whom correspondence should be addressed. e-mail: slevery{at}cisunix.unh.edu
 |
ABSTRACT
|
|---|
Aspergillus nidulans is a well-established nonpathogenic laboratory model for the opportunistic mycopathogen, A. fumigatus. Some recent studies have focused on possible functional roles of glycosphingolipids (GSLs) in these fungi. It has been demonstrated that biosynthesis of glycosylinositol phosphorylceramides (GIPCs) is required for normal cell cycle progression and polarized growth in A. nidulans (Cheng, J., T.-S. Park, A. S. Fischl, and X. S. Ye. 2001. Mol. Cell Biol. 21: 6198–6209); however, the structures of A. nidulans GIPCs were not addressed in that study, nor were the functional significance of individual structural variants and the downstream steps in their biosynthesis. To initiate such studies, acidic GSL components (designated An-2, -3, and -5) were isolated from A. nidulans and subjected to structural characterization by a combination of one-dimensional (1-D) and 2-D NMR spectroscopy, electrospray ionization-mass spectrometry (ESI-MS), ESI-MS/collision-induced decomposition-MS (MS/CID-MS), ESI-pseudo-[CID-MS]2, and gas chromatography-MS methods.
All three were determined to be GIPCs, with mannose as the only monosaccharide present in the headgroup glycans; An-2 and An-3 were identified as di- and trimannosyl inositol phosphorylceramides (IPCs) with the structures Man13Man12Ins1-P-1Cer and Man13(Man16)Man12Ins1-P-1Cer, respectively (where Ins = myo-inositol, P = phosphodiester, and Cer = ceramide). An-5 was partially characterized, and is proposed to be a pentamannosyl IPC, based on the trimannosyl core structure of An-3.
Abbreviations: Cat, cation; CID, collision-induced dissociation; COSY, correlation spectroscopy; ESI, electrospray ionization; GC, gas chromatography; GIPC, glycosylinositol phosphorylceramide; GSL, glycosphingolipid; HPTLC, high-performance thin-layer chromatography; MS, mass spectrometry; TOCSY, total correlation spectroscopy; TOF, time-of-flight
Supplementary key words glycolipid • mass spectrometry • tandem mass spectrometry • collision-induced dissociation • electrospray ionization • NMR spectroscopy • Aspergillus fumigatus
|
INTRODUCTION
|
|---|
A dramatic increase in the frequency of severe mycotic infections during the past two decades has accompanied the growth in populations of immunosuppressed or immunocompromised individuals, including those with acquired immune deficiency due to human immunodeficiency virus infection, recipients of organ and tissue transplants, and patients with leukemias or other cancers (1–5). As with a variety of other microbial infections, reports of resistance to available therapeutic agents have become a major concern for public health clinicians (6–9). The most commonly occurring mycotic infections are those with Aspergillus (usually A. fumigatus and A. flavus), Candida (chiefly C. albicans), and Cryptococcus (chiefly C. neoformans); a wide variety of other species, including some that until recently were not thought of as pathogenic, have also been detected with increasing frequency in disseminated mycoses (10–12).
These circumstances have created an urgent need for new antifungal therapeutic and diagnostic agents, and, by extension, an imperative to identify components that are not only critical to the fungal life cycle and infectivity but that also contain structural elements distinct from those of the host. Investigations of glycosphingolipid (GSL) biosynthesis and function in Saccharomyces cerevisiae and other fungi have established that GSLs are essential components of the fungal cell membrane (13–15). These studies have focused on glycosylinositol phosphorylceramides (GIPCs), a family of compounds widely distributed among fungi but not found in mammalian cells or tissues (15) (see Scheme 1)
. Fungal components or metabolites that interact strongly with the host immune system are of particular interest, and there is ample evidence that GIPCs can interact with the mammalian immune system, in particular where they present oligosaccharide epitopes not shared with mammalian glycoconjugates (16–21).
Information about GIPC biosynthesis and function has been emerging at an increasing rate, due to generation of a variety of Saccharomyces strains mutated with respect to sphingolipid synthesis, the completion of the yeast genome sequence, and the availability of inhibitors of various steps in the GIPC biosynthetic pathway (13–15, 22). Synthesis of inositol phosphorylceramide (IPC), an obligate intermediate in GIPC catabolism, appears to be required for normal growth of S. cerevisiae (23). The IPC synthases are believed to be coded by orthologous AUR1/IPC1 genes discovered in S. cerevisiae, Schizosaccharomyces pombe, Candida spp., Aspergillus spp., and C. neoformans (24–29), and a number of inhibitors of this reaction have been discovered (30–32). These appear to be very useful tools for studying GIPC biosynthesis and function, as well as offering promise as antifungal agents, insofar as they are highly toxic to many fungi but exhibit low toxicity in mammals (33). Disruption of the AUR1 gene in S. cerevisiae and S. pombe results in morphological changes, disappearance of microtubules, abnormal deposition of chitin, and cessation of growth (25, 29). More recently, Luberto et al. (34) observed that downregulation of the IPC1-encoded IPC synthase in C. neoformans significantly lowered the expression of certain virulence traits, as well as impairing intracellular growth in an in vitro murine macrophage model and pathogenicity in an established rabbit in vivo model. Using both AUR1 gene disruption and IPC synthase inhibition, Cheng et al. (35) demonstrated convincingly that GIPC biosynthesis is required for normal cell cycle progression and polarized growth in Aspergillus nidulans (Emericella nidulans). We have also investigated GSL biosynthesis and function in A. nidulans (36), because it is an established nonpathogenic model for the opportunistic mycopathogen, A. fumigatus (37–41). However, the structures of the GIPCs in A. nidulans have not been determined, nor have the possible functional significance of individual GIPC variants and the downstream steps in their biosynthesis been addressed. In this work, we characterized the major GIPC components of A. nidulans using one-dimensional (1-D) and 2-D NMR spectroscopy, electrospray ionization-mass spectrometry (ESI-MS); ESI-MS/collision-induced dissociation (CID)-MS; and ESI-pseudo-[CID-MS]2 in a hybrid ESI-quadrupole/time-of-flight-MS (ESI-Qq/oa-TOF-MS) instrument; and gas chromatography-mass spectrometry (GC-MS). The proposed GIPC structures are all oligo--mannose ("high-mannose") GIPCs based on the core Man3Man2Ins1P1Cer (Ins = myo-inositol; P = phosphodiester; Cer = ceramide).
|
EXPERIMENTAL PROCEDURES
|
|---|
Fungal isolate and growth conditions
A. nidulans FGSC strain A28 was used for all preparations. Approximately 5 x 109 spores were inoculated to 200 ml or 1 l complete medium (1% glucose, 0.2% peptone, 0.1% yeast extract, 0.1% casamino acids, nitrate salts, trace elements, and 0.01% vitamins, pH 6.5; trace elements, vitamins, nitrate salts, and amino acid supplements are described in the appendix to Kafer (42), incubated at 30 or 37°C with shaking (200 rpm) for 24 h, filtered, and processed as described. Yields were 6–8 g wet weight per 200 ml medium, or 25–35 g wet weight per 1 l medium.
Solvents for extraction and anion exchange chromatography
Solvent A, isopropanol-hexane-water (55:25:20; v/v/v, upper phase discarded); solvent B, chloroform-methanol (1:1; v/v); solvent C, chloroform-methanol-water (30:60:8; v/v/v).
High-performance thin-layer chromatography
Analytical high-performance thin-layer chromatography (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 D) as mobile phase. Lipid samples were dissolved in solvent A and applied by streaking from 5 µl Micro-caps (Drummond, Broomall, PA). 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 was sprayed and heated briefly to 
200–250°C.
Extraction and purification of GSLs
Extraction and purification of GSLs were carried out as described previously (19, 36, 43), but with some modifications and additional steps introduced for the purpose of reducing earlier in the protocol the amounts of irrelevant substances to be dealt with. Briefly, GSLs were extracted by homogenizing mycelia (25–35 g wet weight) in an Omni-mixer (Sorvall Inc., Wilmington, DE) three times with 200 ml of solvent A and three times with 200 ml of solvent B. The six extracts were pooled, dried on a rotary evaporator, dialysed against water, and lyophilized. The dried residue was partitioned between water and 1-butanol presaturated with water (200 ml each) with vigorous shaking in a separatory funnel. The lower (water) layer was removed and similarly extracted three more times with equal volumes of water-saturated 1-butanol. The four 1-butanol extracts were combined in a round-bottom flask, evaporated to dryness, resuspended in a minimal volume of solvent C, and applied to a column of DEAE-Sephadex A-25 (Ac- form). Neutral GSLs were eluted with 5 vol of solvent C. Acidic GSLs were eluted with 5 vol of 0.5 M sodium acetate in MeOH. The acidic fraction was dried, dialyzed exhaustively against deionized water, redried, and treated with 20 ml methanol-water-1-butanol (4:3:1; v/v/v) containing 25–30% methylamine at 55°C for 4 h (flask tightly stoppered), with occasional agitation. After removal of the reagent solution by rotary evaporation, the acidic lipids were further fractionated by repetitive preparative-scale HPLC, employing columns of either 60 cm x 4.6 mm Iatrobeads (6RS-8010; Iatron Chemical Co., Tokyo, Japan) or 50 cm x 4.6 mm Sphereclone (Phenomenex, Torrance, CA) 10 µm porous spherical silica. The mobile phase was a 2-propanol-hexane-water gradient programmed from 55:40:5 to 55:25:20 (v/v/v) over 120 min, followed by isocratic elution for 40 min; flow rate 0.5 ml/min. Generally, 40 x 2 ml fractions were collected for first-stage purifications, and 80 x 1 ml fractions were collected for second-stage purifications, where required. The identity and purity of each fraction were assessed by analytical HPTLC using solvent D, as above, and by 1-D 1H-NMR spectroscopy before further characterization by the full range of NMR and MS techniques described below.
1-D 1H and 2-D 1H-1H and 1H-13C nuclear magnetic resonance spectroscopy
Samples of underivatized GIPC (0.5-1.0 mg) were deuterium exchanged by repeated lyophilization from D2O, and then dissolved in 0.5 ml DMSO-d6/2% D2O (44, 45) for NMR analysis. 1-D 1H-NMR, 2-D 1H-1H-gradient-enhanced correlation spectroscopy (gCOSY), 1H-1H total correlation spectroscopy (TOCSY), and proton-detected 1H-13C-gradient-enhanced heteronuclear single quantum coherence (gHSQC) experiments were performed at 35°C on Varian Unity Inova, 500 MHz (Department of Chemistry, University of New Hampshire), 600 MHz, 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 (
= 0.000 ppm). Carbon chemical shifts are referenced to solvent DMSO ( = 40.0 ppm); because these were not obtained from directly detected 1-D spectra but were measured in F1 of the gHSQC experiment, values are given to single decimal place only.
Positive ion mode ESI-MS
Mass spectrometry was performed in the positive-ion mode on a Micromass Global (Manchester, UK) hybrid ESI-Qq/oa-TOF-MS instrument, with GIPC sample introduction via direct infusion in 100% MeOH (100 ng/µl). The flow rate was usually 0.5 µl/min, but for analysis of An-5, which was available in limited amounts, a nano-spray capillary tip was employed, from which the flow rate was estimated to be 200 nl/min. For generation of [M(Li) + Li]+ adducts of GIPC molecular species, LiI (10 mM) in MeOH was added to the analyte solution until the observed ratio of [M(Li) + Li]+ adducts to mixed Na+/Li+ adducts in MS profile mode was >5:1; the necessary LiI concentration was generally in the range of 3 to 5 mM. In general, spectra represent summations of 50–350 scans (30 scans/min) for MS [M(Li) + Li]+ profiles and MS/CID-MS experiments. Resolution was generally 4,000 (5% valley) for MS profile spectra and 3,000 for MS/CID-MS experiments. Extraction cone voltage (analogous to orifice-to-skimmer potential in Sciex API series instruments) was 35 V for MS profile spectra and MS/CID-MS experiments, and 80 V for pseudo-+ESI-(CID-MS)2 experiments. Nominal, monoisotopic m/z values are used in the labeling and description of +ESI-MS results. Interpretation of spectra derived from [M(Li) + Li]+ adducts of GIPC molecular species was essentially as described previously (46).
Monosaccharide, inositol, fatty acid, and sphingoid component analysis by GC-MS
Monosaccharides were analyzed as their per-O-trimethylsilyl methyl glycosides. A re-N-acetylation step was used after methanolytic depolymerization to ensure that hexosamines, if present, would be detected as their acetamido derivatives. Ceramide-derived sphingoid bases and 2-hydroxy fatty acids were detected as their N-acetyl-per-O-trimethylsilyl and 2-O-trimethylsilyl methyl ester derivatives, respectively. All derivatives were prepared and analyzed by GC-MS according to protocols described previously (45). The instrument used was a GCQ (Finnigan MAT, San Jose, CA) operated in electron ionization (EI) mode. Myo-inositol was detected in the monosaccharide analysis as its per-O-trimethylsilyl derivative, although optimal conditions for its release (16) were not employed.
|
RESULTS
|
|---|
HPTLC profile and fractionation of A. nidulans GIPCs
Reproduced in Fig. 1A
is an HPTLC of the crude acidic lipid fraction from A. nidulans A28 (37°C culture), stained with orcinol for detection of hexose-containing material (lane C), compared with a reference standard containing Man3Man2IPC and Galf ß6(Man3)Man2IPC (lane S). In general, such orcinol-stained bands in the acidic fraction of fungal lipids correspond to GIPCs. (IPC itself, having no carbohydrate content, is not detected by this method; on the other hand, hexose-containing impurities, including acidic oligosaccharides, may appear.) A similar profile, showing essentially three major (along with some minor) GIPC bands was obtained by Cheng et al. (35) from a different wild-type A. nidulans strain (A773) via [3H]myo-inositol or [3H]dihydrosphingosine metabolic precursor labeling and autoradiography. Note that the component with the highest Rf value comigrated with an authentic standard of Man3Man2IPC (M2IPC) previously characterized from Paracoccidioides brasiliensis (45), but the most abundant component (middle band) had a significantly lower Rf than did the other major GIPC from P. brasiliensis, Galf ß6(Man3)Man2IPC (Gf M2IPC). A third, less-abundant component migrated with a considerably lower Rf , consistent with attachment of one or two additional monosaccharide units. The crude lipids were fractionated by preparative-scale HPLC (Fig. 1B), with each of the three major components obtained in at least one fraction as a single orcinol-stained band (lanes 2, 4, and 6); these components, designated An-2, -3, and -5, respectively, were subjected to characterization by 1H-NMR, MS, and GC-MS techniques as described below. An amount of An-3 sufficient for extensive 2-D NMR analysis was accumulated from three HPLC fractionations. An additional very minor component (An-4) could be detected (migrating between An-3 and An-5 in Fig. 1B, lane 5), but could not be purified sufficiently for individual characterization.
Monosaccharide, inositol, fatty acid, and sphingoid component analysis of A. nidulans GIPCs
Mannose was essentially the only monosaccharide identified by GC-MS analysis of the trimethylsilylated methyl glycosides produced following methanolysis of An-2, -3, and -5. Myo-inositol was also detected in all three fractions. The only long-chain fatty acid detected in significant amounts in all fractions was h24:0, identified by its characteristic retention time and fragments observed in its EI mass spectrum [as reviewed in ref. (47)]. A trace (<1%) of h25:0 fatty acid was detected in An-3. Traces (<2%) of nonhydroxy 16:0, 18:0, and 24:0 fatty acids were also detected in all fractions. No dihydroxy fatty acids were detected. The major sphingoid components detected were t18:0 and t20:0 4-hydroxysphinganines (phytosphingosines); the observed ratios were 3:2 for An-2, 2:3 for An-3, and 3:2 for An-5. Under conditions of ion trap detection, EI spectra of the 4-hydroxysphinganine-derived N-acetyl-1,3,4-tri-O-trimethylsilyl-4-hydroxysphinganines were qualitatively similar in most respects to those published previously by Thorpe and Sweeley (48), but with more abundant representation of high-m/z ions, including some additional molecular mass-related ions as observed by Levery et al. (45). Thus, [M - 15]+, [M - 15 - 90]+, [M - 59 - 90]+, [M - 174]+, and [M - 174 - 90]+ ions were observed at m/z 560, 470, 426, 401, and 311 for the t18:0 base derivative, and at m/z 588, 498, 454, 429, and 329 for the later-eluting t20:0 base derivative [data not shown; for detailed interpretations, including identifications of lower-m/z ions, see refs (48–50)].
1H-NMR spectroscopic analysis of A. nidulans GIPCs
An-2
Parts of the 1H-NMR spectrum of An-2 were obscured by resonances from a major orcinol-negative component. However, a pair of proton resonances was clearly observed in the anomeric region (Fig. 2A)
at 4.896 ppm (3J1,2 = 1.5 Hz) and 5.059 ppm (3J1,2 < 2.0 Hz); these chemical shift values are essentially identical to those found previously for H-1 of the Man3 and Man2 residues, respectively, of Man3Man2IPC from P. brasiliensis (45). This GIPC has been obtained from a variety of fungi (45) (S. B. Levery, M. S. Toledo, A. H. Straus, and H. K. Takahashi, unpublished observations) and is easily recognizable by its anomeric proton signature and HPTLC Rf identical to that of a fully characterized Man3Man2IPC. Therefore, this fraction was considered to be identified by adequate criteria and was not purified or characterized further.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2. Downfield expansions (5.20–4.50 ppm) of 1-D 1H-NMR spectra (800 MHz; DMSO-d6/2% D2O; 35°C) of GIPC fractions isolated from A. nidulans; spectra were acquired on fractions corresponding to lanes 2, 4, and 6 shown in Fig. 1B. A: An-2 (lane 2); B, An-3 (lane 4); C: An-5 (lane 6).
|
|
An-3
The 1-D 1H-NMR spectrum of fraction An-3 (Fig. 3A)
showed it to be a nearly homogeneous GIPC component, displaying only a few minor impurity peaks. As shown in detail in Fig. 2B, three anomeric proton resonances are clearly observed in the spectrum at 4.666 ppm (3J1,2 = 1.4 Hz), 4.877 ppm (3J1,2 = 1.5 Hz), and 5.013 ppm (3J1,2 = 1.5 Hz). All three of the 3J1,2 values are consistent with -Manp, although the resonance at 4.666 ppm appears rather upfield (under these conditions) compared with the H-1 signals of other GIPC -Man residues we have encountered. However, ample precedent for this large upfield decrement can be found in NMR spectroscopy (in D2O) of high-mannose glycoprotein N-glycans, where H-1 of nonreducing terminal Man16 residues are observed at 0.18–0.22 ppm upfield from H-1 of terminal Man13 (51, 52). Almost all monosaccharide, inositol, and ceramide resonances were unambiguously assignable from 2-D 1H-1H gCOSY and TOCSY experiments (Fig. 3B, C and Table 1). Analysis of approximate 3Ji,j proton coupling constants around the three monosaccharide spin systems confirmed their identities as -Manp residues. Analysis of approximate 3Ji,j coupling constants around the cyclic Ins spin system confirmed it as myo-inositol (all hydroxyl groups equatorial except that at C-2), while the chemical shift pattern, compared with previously acquired data, was consistent with glycosylation of Ins at O-2 (45, 53–55). In the nuclear Overhauser effect (NOE) spectroscopy spectrum (Fig. 3D), a strong NOE cross-peak is observed between the -Manp H-1 at 5.013 ppm and Ins H-2 at 3.954 ppm (resolvable despite partial overlap with the expected intraresidue -Manp H-1/H-2 cross-peak at 3.939 ppm), thus correlating the partners in a Man12Ins core linkage. The linkage of the other two -Manp residues to the first at O-3 and O-6 is supported by strong NOE cross-peaks correlating the H-1 resonating at 4.877 ppm with H-3 of Man2 at 3.620 ppm (in addition to weaker correlations with H-2 and H-4 of that residue) and the H-1 resonating at 4.666 ppm with H-6a of Man2 at 3.520 ppm (in addition to weaker correlations with H-6b and H-5 of that residue). Although severe spectral overlap renders the interglycosidic Man6 H-1/Man2 H-6a correlation somewhat ambiguous by itself (other protons resonate at or near 3.520 ppm), the additional Man6 H-1/Man2 H-6b and Man6 H-1/Man2 H-5 correlations, though weak, are observed at nondegenerate monosaccharide proton frequencies (3.703 ppm and 4.060 ppm, respectively), and merely incidental appearance of these without that relating Man6 H-1 with Man2 H-6a is highly unlikely. Interestingly, the extreme downfield shift of H-5 was observed previously in a GIPC having the Man2 residue substituted at O-3 and O-6, i.e., in the case of Galf ß6(Man3)Man2IPC from P. brasiliensis (45).
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 3. Downfield expansions (5.15–2.80 ppm) of one-dimensional (1-D) and 2-D 1H-NMR spectra (800 MHz; DMSO-d6/2% D2O; 35°C) of A. nidulans GIPC fraction An-3 (combined yield from 3 HPLC purifications). A: 1-D spectrum. B: Upper left section of 2-D 1H-1H g correlation spectroscopy spectrum. C: Lower right section of 2-D 1H-1H total correlation spectroscopy spectrum (200 msec mixing time). D: 2-D 1H-1H nuclear Overhauser effect spectroscopy spectrum (100 msec mixing time; F1, 5.1–4.5 ppm). In D, inter-residue nuclear Overhauser effect correlations across glycosidic linkages are boxed; weaker nonglycosidic inter-residue correlations are marked by dashed boxes; intraresidue correlations have been left unmarked.
|
|
Finally, 13C resonance assignments (Table 1), made from a 2-D 1H-detected, 13C-1H gHSQC spectrum (not shown), were also consistent with the linkage of two of the -Manp residues to the core Man2 residue at O-3 and O-6; this is supported by a pattern of substantial downfield shift increments (-effects) for Man2 C-3 and C-6, along with upfield shift decrements (ß-effects) (56) for Man2 C-2, C-4, and C-5, compared with 13C spectral data for the parent compound, Man2IPC (54). All other 13C chemical shifts are consistent with the proposed Man2IPC core structure, including those characteristic of the myo-Ins residue substituted at O-2 (54). On the basis of all of these data, the structure of An-3 was proposed to be Man6(Man3)Man2IPC.
Further confirmation of the structure of An-3 was obtained by comparison of its NMR spectrum with that of an authentic sample of Man6(Man3)Man2IPC ("Cc-2") previously found in a Basidiomycete species, Cantharellus cibarius (Chantarelle) (55). As shown in Fig. 4
, the NMR spectrum of C. cibarius Cc-2 is almost identical to that of An-3 except for some doubling of the Ins-5 resonance, a spectral feature that appears to be associated in general with the presence of a second ceramide type containing dihydroxy-fatty-N-acylation (S. B. Levery, unpublished observations), already known to be present in the C. cibarius GIPC preparation (55). In particular, the three monosaccharide H-1 signals are observed at identical chemical shifts, each with the same 3J1,2 coupling constant as found for An-3. Because the branched Man6(Man3)Man structure was confirmed unambiguously for Cc-2 by methylation linkage analysis (55), the structure of An-3 is thereby confirmed as well.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 4. Downfield expansion (5.15–2.80 ppm) of 1-D 1H-NMR spectrum (500 MHz; DMSO-d6/2% D2O; 35°C) of C. cibarius GIPC fraction Cc-2.
|
|
An-5
The third GIPC fraction, An-5, yielded a very complex 1H-NMR spectrum, and too little material was available for extensive application of 2-D NMR analysis; however, five signals with similar intensity and line width (3J1,2 <2.0 Hz) are clearly observable in the H-1 region of the 1-D spectrum at 4.994, 4.878, 4.864, 4.821, and 4.657 ppm (Fig. 2C), and each of their correlated H-2 partners was assignable (Table 2) from correlations observed in an otherwise poor-quality TOCSY spectrum (not shown). Three of the H-1 signals occur at chemical shifts near those of the anomeric protons in An-3, suggesting a structure for An-5 based on elaboration of the Man6(Man3)Man2IPC core, with two additional -Man residues. Even provisional assignment for the linkages of the two additional residues based on these data is problematic; however, a number of arguments can be made to narrow the field of likely structures for An-5.
View this table:
[in this window]
[in a new window]
|
TABLE 2. Proposed structure and 1H chemical shifts (ppm) for H-1 and H-2 of monosaccharide residues of An-5 from A. nidulans in DMSO-d6/2% D2O at 35°C
|
|
First, assignment of the H-1 resonance at 4.994 ppm to the core branching Man2 residue is supported by observation of its correlated H-2 signal at 3.953 ppm, close to the value observed for this proton in An-3 (3.939 ppm). This is a significant point, inasmuch as this relatively downfield-shifted value for -Man H-2 is characteristic for glycosylation of this residue at O-3, e.g., by a second -Man residue, compared with unglycosylated (terminal) -Man; by contrast, the effects of glycosylation of -Man at O-6, and even O-2, on this resonance are relatively negligible (51, 52). Analysis of NMR data on GIPCs acquired under similar conditions (45, 53, 54), including the data for An-3 in this work, shows that H-2 of the penultimate residue in a Man3Man disaccharide can be found in the range 3.88–4.08 ppm (5 cases), compared with 3.66–3.77 ppm for a terminal -Man residue (7 cases). Thus, barring unexpected conformational effects, the lack of significant downfield shifts for H-2 of the four remaining -Man residues of An-5 (all in the range of 3.67–3.77 ppm) (Table 2) tends to preclude glycosylation at O-3; i.e., neither of the two Man residues added onto the An-3 core is in an 13 linkage. Second, we propose that glycosylations at O-4 can be excluded (at least provisionally), based on the lack of reports of structures containing a Man4Man disaccharide, or of -mannosyltransferases able to make this linkage. The two additional Man residues are therefore likely to be linked either 12 or 16.
The field of possible structures can be further narrowed by consideration of the cumulative effects of adding Man residues, in either 12 or 16 linkage, on the H-1 subspectrum of An-3, taking into account what is already known of these effects from "structure reporter group" analysis of many closely related oligo--mannose-type glycans (51, 52). For example, all structures one could propose involving addition of both residues in 16 linkage are excluded, because one would then expect to find three H-1 resonances in the upfield region around 3.66–4.67 ppm. One possible structure for An-5, consistent with the available NMR data, would be addition of a Man2Man6 disaccharide to the Man6 residue of An-3. Considering the effects one at a time, addition of the second Man6 residue could be responsible for a slight upfield shift of H-1 of the first (from 4.666 ppm in An-3 to 4.657 ppm observed in An-5), but would itself be expected to give rise to a second H-1 resonance in the same region of the spectrum. However, subsequent addition of a Man2 residue would then be expected to cause a significant downfield shift for H-1 of this second Man6 residue (observed at 4.821 ppm in An-5). Another possibility, consistent with the same reasoning, is addition of the Man2 residue to the inner Man6 to create a doubly branched Man6(Man2)Man6(Man3)Man structure. However, mass spectrometric analysis (see subsequent section) appears more consistent with a single linear branch as proposed above, and tended to exclude as well structures formed by extension of both the Man6 and Man3 residues of An-3 by one additional monosaccharide. Other "soft" criteria may allow one at least to make some provisional judgments. For example, addition of the same sugars in reverse order (Man6Man2) is not categorically excluded by the NMR data but seems less likely, based on analogy to known oligo--mannose structures. Finally, the possibility that the disaccharide extension has been added to the Man3 residue of An-3 can be considered less likely, because the resonance appearing in the spectrum of An-3 at 4.877 ppm, assigned as H-1 of the Man3 residue, also appears in the spectrum of An-5 virtually unshifted at 4.878 ppm. However, it is also possible that this Man3 H-1 resonance has indeed been shifted upfield (e.g., to 4.864 ppm by Man6 glycosylation), while the H-1 of one of the additional residues coincidentally resonates at the same chemical shift in An-5; therefore, this alternative structure is not definitively excluded.
+ESI-MS and MS/CID-MS analysis of A. nidulans GIPCs
An-3
A molecular profile of GIPC fraction An-3 was acquired via +ESI-Q/oa-TOF-MS (Fig. 5A)
; a pair of [M(Na) + Na]+ salt adduct ions observed at m/z 1,456 and 1,484 is consistent with an M3IPC having ceramides consisting of t18:0/t20:0 4-hydroxysphinganines with h24:0 fatty-N-acylation. Less-abundant ions at m/z 1,472 and 1,500 represent the same molecular species with single substitutions of K+ for Na+ (see below). Addition of LiI shifted the profile to predominant [M(Li) + Li]+ adducts at m/z 1,424 and 1,452, with residual mixed Na+/Li+ adducts present at lower abundance at m/z 1,440 and 1,468 (Fig. 5B). The latter ions could in principle represent molecular components containing dihydroxy-fatty-N-acylation, because this modification would also increment by 16 the m/z values of the predominant components. However, dihydroxy fatty acids were not detected in the component analysis.
View larger version (41K):
[in this window]
[in a new window]
|
Fig. 5. +ESI-hybrid quadrupole/time-of-flight (Qq/oa-TOF) spectra of An-3. A: Profile of molecular ions as [M(Na) + Na]+ adducts. B: Profile of molecular ions as [M(Li) + Li]+ adducts. C: Mass spectrometry/collision-induced dissociation-mass spectrometry (MS/CID-MS) of [M(Li) + Li]+ at m/z 1,452, low m/z region. D: MS/CID-MS of [M(Li) + Li]+ at m/z 1,452, high m/z region. The Y-axis expansion in D relative to C is 27x. Ion designations correspond to Scheme 3. The designations "+Li" and "(Li) + Li" and the charge form have been omitted from the fragment labels for clarity.
|
|
In general, +ESI-MS/CID-MS spectra acquired [via +ESI-Q /q(CID)-oa-TOF-MS] from either disodiated or dilithiated molecular adducts exhibited abundant pairs of fragment ions representing the complete glycosylinositol {[B3 + cation (Cat)]+ and (C3 + Cat)+} and glycosylinositol phosphate {[B3PO3(Cat) + Cat]+ and [C3PO3(Cat) + Cat]+} moieties, along with ceramide ([Y0 + Cat]+ and [Z0 + Cat]+) and ceramide phosphate {[Y0PO3(Cat) + Cat]+ and [Z0PO3(Cat) + Cat]+} ions (see Scheme 2)
| |
Scheme 2. Characteristic fragmentations of glycosylated IPCs. A: Nomenclature of Adams and Ann (73) for fragmentation of the ceramide moiety, and of Singh, Costello, and Beach (74) for the myo-inositol phosphoryl group. B: Sphingoid d3b ion, proposed by Hsu and Turk (57). C: Hydrated analog of d3b ion, proposed in this study as a product of t18:0 or t20:0 phytosphingosine-containing ceramides. D: Sphingoid c1b ion, proposed by Hsu and Turk (57). E: Hydrated analog of c1b ion, proposed in this study as a product of t18:0 or t20:0 phytosphingosine-containing ceramides.
|
|
. Ions in parallel series, representing glycosidic fragmentations occurring singly or in combination with other cleavages (the latter denoted by one or more "/"), were also observed {[Cn + Cat]+, [Bn + Cat]+, [Yn + Cat]+, [Zn + Cat]+, [Ym/Cn + Cat]+, [Ym/Bn + Cat]+, [Ym/CnPO3(Cat) + Cat]+, and [Ym/BnPO3(Cat) + Cat]+}. In many cases, [Zm/BnPO3 (Cat) + Cat]+ and [Zm/Bn + Cat]+ ions were also observed. Of these fragments, the [Ym/CnPO3(Cat) + Cat]+ and [Ym/BnPO3(Cat) + Cat]+ series were the most useful for deducing glycosidic sequence, because the presence of the phosphate moiety clearly labels the inositol-containing reducing end of the glycan. +ESI-MS/CID-MS of [M(Li) + Li]+ adducts yielded better overall S/N than [M(Na) + Na]+ adducts, although the differences were not as pronounced as previously reported with the use of a triple quadrupole for +ESI-MS/CID-MS analysis of GIPCs (46).
An +ESI-MS/CID-MS spectrum acquired from the dilithiated molecular adduct at m/z 1,452 is reproduced in Fig. 5C, D, showing the predominant [B3PO3(Li) + Li]+/[C3PO3(Li) + Li]+ pair (m/z 741/759) and other abundant fragments (summarized in Table 3). Consistent with previous findings (45, 46), a somewhat lower abundance is observed for glycosylinositol phosphate fragment ions requiring at least two glycosidic cleavages for their appearance {[Y2/Y2ß/B3PO3(Li) + Li]+ and [Y2/Y2ß/C3PO3(Li) + Li]+, at m/z 417 and 435, respectively}, compared with other ions in the [Ym/BnPO3(Li) + Li]+ and [Ym/CnPO3(Li) + Li]+ series, indicating the presence of a branch point in the glycan (see Scheme 3)
| |
Scheme 3. Fragmentation of GIPC An-3 in positive-ion electrospray ionization-quadrupole/time-of-flight-mass spectrometry (ESI-Qq/oa-TOF-MS); ion designations according to Costello and Vath (75), and Singh, Costello, and Beach (74). The adduct designation "+Li" and the charge form have been omitted from the labels for clarity.
|
|
. In particular, the ratios of the relative abundances of m/z 435 versus m/z 273 (0.22), and m/z 435 versus m/z 597 (0.16), are attenuated by a factor of 3 in An-3 when compared with the same ratios in an isomeric unbranched triglycosyl GIPC structure (0.70 and 0.52, respectively) (B. Bennion, S. B. Levery, M. S. Toledo, A. H. Straus, and H. K. Takahashi, unpublished observations). Such differences in abundance ratios appear to be typical under these conditions.
View this table:
[in this window]
[in a new window]
|
TABLE 3. Glycosylinositol- and phosphorylglycosylinositol-related product ions formed in low-energy ESI-MS/CID-TOF-MS spectra of tri- and pentamannosylinositol phosphorylceramides from mycelia of A. nidulansa
|
|
In addition to glycosidic fragment series derived from the glycosylinositol phosphate and glycosylinositol moieties, fragments from sequential loss of monosaccharide residues from [M(Li) + Li]+ {[Ym(Li) + Li]+}, which were not observed significantly in previous +ESI-MS/CID-MS of GIPCs acquired on a triple-quadrupole instrument (45, 46), appear at lower abundance in the +ESI-Q/q(CID)-oa-TOF-MS (Fig. 5D, Table 4). In the same region of the spectra and at similar abundance, ions from loss of the acyl chain and loss of acyl C2-C
{[O(Li) + Li]+, [J(Li) + Li]+, and [J(Li) + Li - H2O]+} are also observed. Within this group, the latter ion appears to be predominant. Similar to previous results (46), the ceramide and ceramide phosphate ions {[Y0 + Li]+, [Z0 + Li]+, [Y0PO3(Li) + Li]+, and [Z0PO3(Li) + Li]+} appear at intermediate abundances; in addition, ions corresponding to [O/Z0PO3(Li) + Li]+ and a hydrated d3b analog (see Scheme 2C) are observed at lower abundances (m/z 420 and 319, respectively, denoted by asterisks in Fig. 5C). An essentially identical spectrum, except for those ions containing all or part of the ceramide moiety, was obtained via +ESI-MS/CID-MS of the molecular adduct at m/z 1,424 (not shown); in this case all ions containing the ceramide or sphingoid were decremented by m/z 28 (see Table 4). The [O(Li) + Li]+, [J(Li) + Li]+, [J(Li) + Li - H2O)]+, and [O/Z0PO3(Li) + Li]+ ions all provide information about the carbon number of the sphingoid and, by difference, the N-acyl chain, all or part of which is lost in the process of their formation. In this case, in agreement with the sphingoid and fatty acid analysis, they indicate that the m/z 28 increment between the two predominant molecular species is primarily due to a corresponding C2H4 difference in the sphingoid rather than the N-acyl chain. Thus the [M(Li) + Li]+ adducts at m/z 1,424 and 1,452 correspond to molecular species containing t18:0 and t20:0 4-hydroxysphinganines with h24:0 fatty-N-acylation.
View this table:
[in this window]
[in a new window]
|
TABLE 4. Molecular adduct ions [M(Li) + Li]+ observed in ESI-Q/TOF-MS profile (A), and significant ceramide-containing (B), sphingoid-containing and sphingoid-derived (C, D), and fatty-N-acyl-derived (E) product ions (m/z) formed in low-energy ESI-MS/CID-TOF-MS and ESI-CID-MS/CID-TOF-MS spectra of tri- and pentamannosylinositol phosphorylceramides from mycelia of A. nidulans
|
|
This was further confirmed by pseudo-+ESI-(CID-MS)2 experiments, using a high extraction cone voltage (analogous to high orifice-to-skimmer potential in Sciex APITM series instruments) to generate in-source CID primary fragments [+ESI-(CID)], which are then selected by the first analyzer (Q) for subsequent generation of /q(CID)-oa-TOF product ion spectra (altogether, +ESI-[CID]Q /q[CID]oa-TOF-MS). We and others (46, 53, 57) previously applied this technique to analysis of underivatized GSLs on a triple-quadrupole instrument. The most abundant fragments in a high extraction cone voltage +ESI-(CID)-Q /oa-TOF-MS profile of An-3 (not shown) were ceramide ions [Y0 + Li]+ (m/z 718 and 690) and the pair of glycosylinositol phosphate ions [B3PO3(Li) + Li]+/[C3PO3(Li) + Li]+ (m/z 759 and 741, respectively). Product ion spectra generated subsequently via +ESI-(CID)-Q /q(CID)-oa-TOF-MS from the primary ions at m/z 718 and 690 are reproduced in Fig. 6A
and Fig. 6B, respectively. These display abundant fragments corresponding to loss of the acyl chain ([N + Li]+ 
[O/Y0 + Li]+) and a sphingoid rearrangement including additional loss of the amino group (d3b; see Scheme 2C). The [N + Li]+ and d3b ions are observed at m/z 352 and 319, respectively, in the spectrum of products from m/z 718, and at m/z 324 and 291, respectively, in the spectrum from m/z 690, confirming that the m/z 28 differences in the ceramide moieties reside entirely in the sphingoid chain lengths, while the N-acyl group is h24:0 in both ceramide species. Minor products included ions derived from these by dehydration or further fragmentation (Table 4). Ions derived from loss of all or most of the sphingoid alkyl chain, while retaining the fatty-N-acyl chain (e.g., [S + Li]+ [G/Y0 + Li - H2]+, [T + Li]+ [G/Z0 + Li]+, [U + Li]+ [T + Li - C2H2]+) as well as an aldehyde ion derived from the fatty-N-acyl C2-C ([W + Li]+), are observed at the same m/z in both spectra (see Table 4 and Scheme 2). The results are similar to those previously obtained with a triple-quadrupole spectrometer (46), but these spectra are of much higher quality.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 6. High extraction cone voltage +ESI-Qq/oa-TOF spectra of An-3. Expansions (m/z 200–500) of tandem ESI-CID-MS/CID-MS product ion spectra from [Y0 + Li]+ of An-3. A: Products of [Y0 + Li]+ at m/z 718. B: Products of [Y0 + Li]+ at m/z 690. The adduct designation "+Li" and the charge form have been omitted from the fragment labels for clarity.
|
|
An-5
A molecular adduct profile of GIPC An-5 (Fig. 7A)
was consistent with a pentahexosyl-IPC structure as proposed, i.e., An-3 elaborated with two additional Man residues. In this case, the +ESI-Q /oa-TOF-MS profile consisted of a pair of [M(K) + K]+ adduct ions (m/z 1,812 and 1,840), accompanied by a slightly more abundant second pair of ions m/z 16 lower (m/z 1,796 and 1,824), representing the same molecular species with single substitutions of Na+ for K+, and a third pair m/z 32 lower (m/z 1,780 and 1,808), representing [M(Na) + Na]+ adducts. Addition of LiI effectively consolidated the molecular profile into a pair of [M(Li) + Li]+ adducts at m/z 1,748 and 1,776 (Fig. 7B). +ESI-MS/CID-MS spectra acquired via +ESI-Q /q(CID)-oa-TOF-MS on these [M(Li) + Li]+ adducts both exhibited abundant fragment ion pairs at m/z 1,065 and 1,083 {designated [B5PO3(Li) + Li]+/[C5PO3(Li) + Li]+ or [B6PO3(Li) + Li]+/[C6PO3(Li) + Li]+, depending on whether the structure contains a branch point}, consistent with pentahexosylinositolphosphate as a dilithiated salt adduct. As shown in detail for the m/z 1,748 adduct (Fig. 7C, D), complete series of glycosidic cleavage ions {[Ym/CnPO3(Li) + Li]+, [Ym/BnPO3(Li) + Li]+, [Cn + Li]+, [Bn + Li]+, [Ym/Cn + Li]+, and [Ym/Bn + Li]+} were observed, along with [Zm/BnPO3(Li) + Li]+ and [Zm/Bn + Li]+ ions (see Table 3). Similar to the observation with An-3, a much lower abundance was observed for fragment ions at m/z 417 and 435, compared with other ions in the glycosylinositol phosphate series. In particular, the ratios of the relative abundances of m/z 435 versus m/z 273 (0.27), and m/z 435 versus m/z 597 (0.17), were consistent with the presence of a single branch point in the glycan at the same monosaccharide residue as An-3 (see Scheme 4
| |
Scheme 4. Fragmentation of GIPC An-5 in positive-ion ESI-Qq/oa-TOF-MS; ion designations according to Costello and Vath (75), and Singh, Costello, and Beach (74). The adduct designation "+Li" and the charge form have been omitted from the labels for clarity.
|
|
, where m/z 417 and 435 represent ions requiring at least two glycosidic cleavages for their appearance, [Y2/Y2ß/B5PO3(Li) + Li]+ and [Y2/Y2ß/C5PO3(Li) + Li]+, respectively). The location of this branch point is consistent with extension of one of the nonreducing terminal Man residues of An-3 with two additional residues in a continuous chain. Different branching structures (e.g., symmetric extension of each of the nonreducing terminal Man residues of An-3 with one additional residue, or the presence of a second branch point in the structure), are contraindicated by the observation of all remaining [Ym/CnPO3(Li) + Li]+ and [Ym/BnPO3(Li) + Li]+ series ions in high abundance.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 7. +ESI-Qq/oa-TOF spectra of An-5. A: Profile of molecular ions as mixed [M(Na) + Na]+, [M(K) + K]+, and [M(Na) + K]+ {or [M(K) + Na]+} adducts. B: Simplified profile of molecular ions as [M(Li) + Li]+ adducts. C: MS/CID-MS of [M(Li) + Li]+ at m/z 1,748, low m/z region. D: MS/CID-MS of [M(Li) + Li]+ at m/z 1,748, high m/z region. The Y-axis expansion in D relative to C is 9x. Ion designations correspond to Scheme 4. The designations "+Li" and "(Li) + Li" and the charge form have been omitted from the fragment labels for clarity.
|
|
As observed with An-3, [Ym(Li) + Li]+, fragments derived from sequential loss of monosaccharide residues from [M(Li) + Li]+ appeared at low abundance in the +ESI-Q/q(CID)-oa-TOF-MS of An-5 (Fig. 7C, D, Table 4). Ceramide and ceramide phosphate ions {[Y0 + Li]+, [Z0 + Li]+, [Y0PO3(Li) + Li]+, and [Z0PO3(Li) + Li]+} appeared as well in the product ion spectra of An-5 at intermediate abundances. As with An-3, ions lacking all or part of the acyl chain {[O(Li) + Li]+, [J(Li) + Li]+, and [J(Li) + Li - H2O)]+} indicated that the m/z 28 increment between the two predominant molecular species is primarily due to a corresponding C2H4 difference in the sphingoid rather than the N-acyl chain. This was again confirmed by product ion spectra generated via +ESI-(CID)-Q/q(CID)-oa-TOF-MS from the primary [Y0 + Li]+ ions at m/z 690 and 718 observed in the high extraction cone voltage profile of An-5. These pseudo-+ESI-(CID-MS)2 spectra (not shown) were essentially identical to those acquired from the same ions of An-3. Thus, major [N + Li]+ and d3b fragments are again observed at m/z 324 and 291 in the spectrum of products from m/z 690, and at m/z 352 and 319 in the spectrum from m/z 718. All of these results are summarized in Tables 3 and 4.
Quantitation of A. nidulans GIPCs
Having established that the major orcinol-positive bands, visible in HPTLC of the crude base-stable acidic lipids (Fig. 1A, lane C), are GIPCs incorporating an increasing number of Man residues, we made a crude estimate of their molar proportions by densitometric analysis of these bands, whose staining could now be "normalized" by the relative hexose content of each component (for this purpose, An-4 was assumed to be a tetrahexosyl IPC). This analysis yielded approximate molar proportions of 1:5:0.1:0.4 (estimated error ±20%) for An-2, An-3, An-4, and An-5, respectively. Note that the usefulness of these values should be judged against our observations that the overall yield and proportion of GIPCs from A. nidulans and other fungi can vary between strains and depend considerably as well on culture medium, temperature, and stage of growth. On the basis of these caveats, as well as the fact that extractions were performed on fresh, wet mycelium, no attempt was made to calculate the yield of GIPCs relative to cellular mass for this study, as such values could not be considered reproducible.
|
DISCUSSION
|
|---|
Recent investigations of GSL biosynthesis and function in fungi have focused on Aspergillus nidulans, a nonpathogenic laboratory model for the opportunistic mycopathogen, A. fumigatus. As has been observed with many other fungi, GSL biosynthesis in A. nidulans is partitioned into two major pathways, one for production of GIPCs, and another leading to accumulation of cerebrosides [ß-galactosylceramide (GalCer) and/or ß-glucosylceramide (GlcCer)]. In a previous study (36), we characterized a GlcCer from A. nidulans and observed that strong inhibitors of GlcCer synthase interfered with germination, cell cycle, and hyphal growth in both A. nidulans and A. fumigatus, suggesting that GlcCer biosynthesis is essential in these fungi and that design of inhibitors more specific for the fungal enzyme than for the mammalian one might offer a potential route to a new class of antifungal agents. On the other hand, it is already well appreciated that disruption of the GIPC pathway, by either inhibition or deletion of IPC synthase activity, is fatal for fungi (24–27, 29, 30), except for some yeast strains with compensating mutations (58–60). Some fungal strains selected for resistance to the IPC synthase inhibitor Aureobasidin A were found to have a mutation in the AUR1 gene that affects the interaction of the AUR1 protein with the drug but not catalytic activity (25–27, 29). Overexpression of the AUR1 gene in S. pombe conferred significant resistance to Aureobasidin A (25); in addition, wild-type strains of some fungal species, such as the mycopathogen A. fumigatus, are relatively resistant by virtue of highly efficient transporter-mediated drug efflux (61). Exploitation of IPC synthase inhibitors as antifungal drugs is already being investigated, but generation of new therapeutic candidates will doubtless have to address issues of fungal resistance (33).
Recent detailed studies have demonstrated that GIPC biosynthesis is essential for normal cell cycle progression and polarized growth in A. nidulans (35). To dissect further the possible functional role(s) of fungal GIPC biosynthesis, and perhaps exploit glycosyltransferases specific to this pathway for the development of other antifungal agents, we have sought to characterize GIPC expression in Aspergillus species in more detail, beginning with studies to establish the structures of GIPCs in the model fungus A. nidulans. From the current study, the GIPC profile of A. nidulans appears to be relatively simple, with expression of only three significant components, all bearing oligo--mannose-type glycans. These are:
 |
The presence of some minor components could be detected, but sufficient material was not obtained for their analysis. The possibility exists that these minor components could possess very different glycans, perhaps bearing sugar residues other than mannose; moreover, other components could be expressed by A. nidulans, perhaps under different conditions or by different strains. However, a similar profile, showing essentially three major (along with some minor) GIPC bands, was also obtained from a different wild-type A. nidulans strain (A773), although no detailed structural characterization was reported in that work (35). The structures proposed herein for the three major A. nidulans GIPC components suggest a single biosynthetic pathway connecting them, as depicted in Scheme 5
.
In general, knowledge about structures of fungal GIPCs is still at an early stage, with only a handful of species having been examined so far. The present study is the first to report detailed characterization of GIPCs from an Aspergillus species. Among the Aspergilli, early attempts at characterization of GIPC structure and biosynthesis were initiated with A. niger (62–64); compositional analysis indicated that the structures were complex, but characterizations beyond this point were not published. The first detailed characterizations of complex GIPCs (containing more than one monosaccharide residue) from a Euascomycete fungal species were carried out in the 1980s by Robert Lester and his colleagues (16, 65). In these key early studies, the major GIPCs from both the yeast and mycelial forms of the dimorphic mycopathogen Histoplasma capsulatum were profiled, isolated, and analyzed by MS, GC-MS, and degradative methods. The primary structures of three GIPCs were established completely except for the precise linkage of the first sugar residue to the myo-inositol ring, which was proposed to be either Man12Ins or Man16Ins.1 A decade and a half later, the first unambiguous determination of this core linkage was published for GIPCs of a related dimorphic mycopathogen, Paracoccidioides brasiliensis, where it was shown to be Man12Ins (45). Subsequent studies have shown this to be an important detail, because GIPC series with both Man12Ins or Man16Ins core linkages are now known (54, 66). More recently, a major component of the yeast form of a third dimorphic mycopathogen, S. schenckii, was found to possess an additional core linkage, GlcNH212Ins (53), so that three distinct monosaccharide linkages to myo-Ins have now been identified for GIPCs of Ascomycete fungal species. In the present study, only one core structure, Man12Ins, was observed for the GIPCs of A. nidulans.
Biosynthesis of the M3IPC (An-3) from IPC presumably requires the action of at least three distinct -mannosyltransferase activities. In S. cerevisiae, two genes, SUR1 (CSG1/BCL21) and CSG2 (CLS2), are apparently required for production of MIPC (most likely Man2IPC) (67, 68). Of these, the latter does not resemble a glycosyltransferase ("glycosyl-T") gene (67, 69), and its product has not been found in the Golgi but in the endoplasmic reticulum (70); the CSG2 protein is thought to be an integral membrane protein with a role in regulation of Ca2+ levels in the endoplasmic reticulum (67, 69), although its true function has not been verified. The SUR1 protein, on the other hand, does incorporate a domain with substantial identity to a known 1,6-Man-T (Och1p) (68), and is therefore considered a likely candidate for a true MIPC synthase. Nothing is known about the presumed 1,3-Man-T responsible for synthesis of the common intermediate M2IPC (An-2) or about glycosyltransferases involved in any of the subsequent steps in synthesis of more complex GIPCs (such as the 1,6-Man-T required to convert An-2 to An-3 in A. nidulans).
As a number of confirmed or putative genes for 1,2-, 1,3-, and 1,6-Man-Ts are known, an interesting point to determine is whether the downstream glycosylation steps in the biosynthesis of oligo--mannosyl IPCs such as An-3 are carried out by -man-Ts also used in other pathways, or by enzymes specific to the GIPC pathway, and therefore among candidates with as yet unconfirmed glycosyltransferase functions. Examples of both overlapping and nonoverlapping pathway specificities are known from studies of mammalian glycosyltransferase families such as the Galß1,3- and Galß1,4-Ts (71). Another key question is whether biosynthesis of MIPC or more highly elaborated GIPCs is essential for the survival of fungi. It has been established that biosynthesis of IPC is required for vegetative growth of S. cerevisiae (24) but that conversion to MIPC is not (68); however, this finding doesn't exclude essential roles for MIPC, M2IPC, and further GIPC products in A. nidulans or other higher fungi. If they have essential roles, then glycosyltransferases downstream of IPC synthase could be targets of inhibitors with potential as antifungal agents, provided they are specific to GIPC biosynthesis.
The identity of the glycosylinositol structure of the A. nidulans M3IPC with that of C. cibarius is somewhat surprising, insofar as these fungi are members of evolutionarily distant phyla (Ascomycota and Basidiomycota, respectively). However, the C. cibarius GIPC profile also appears to be distinct from those of all seven other Basidiomycete species examined so far, in that the core structures of the latter all diverge at the point of adding the second monosaccharide residue (Galß16Man12Ins) (55, 72). In any case, GIPCs more polar than M3IPC were not
TOP
ABSTRACT
INTRODUCTION
