Carbohydrate Research 337 (2002)

Transcription

1 Carbohydrate Researh 337 (2002) Determination of the omposition of the oligosaharide phosphate fration of Pihia (Hansenula) holstii NRRL Y-2448 phosphomannan by apillary eletrophoresis and HPLC Vito Ferro, a, * Caiping Li, a Kym Fewings, a Maria C. Palermo, a Robert J. Linhardt, b Toshihiko Toida a Department of Researh & De elopment, Progen Industries Ltd, PO Box 28, Rihlands BC, Qld 4077, Australia b Department of Chemistry, Di ision of Mediinal and Natural Produts Chemistry, Department of Chemial and Biohemial Engineering, Uni ersity of Iowa, Iowa City, IA 52242, USA Shool of Pharmaeutial Sienes, Chiba Uni ersity, 1-33 Yayoi, Inage, Chiba , Japan Reeived 3 August 2001; aepted 6 November 2001 Abstrat The promising new antianer agent, PI-88, is prepared by the sulfonation of the oligosaharide phosphate fration of the extraellular phosphomannan produed by the yeast Pihia (Hansenula) holstii NRRL Y The omposition of the oligosaharide phosphate fration was determined by apillary eletrophoresis (CE) with indiret UV detetion using 6 mm potassium sorbate at ph 10.3 as the bakground eletrolyte. Further onfirmation of the omposition was obtained by HPLC analysis of a sample dephosphorylated by treatment with alkaline phosphatase. The struture of the hexasaharide omponent has been determined by isolation and NMR spetrosopi analysis of its dephosphorylated derivative. Additionally, the struture of a seond, previously undeteted tetrasaharide omponent (a hexosamine) has been determined by isolation and NMR spetrosopi analysis of the aetate of its dephosphorylated derivative. It is demonstrated that CE is an ideal method for the quality ontrol of the oligosaharide phosphate fration for use in the prodution of PI Elsevier Siene Ltd. All rights reserved. Keywords: Pihia (Hansenula) holstii; Phosphomannan; Oligosaharide phosphate fration; Capillary eletrophoresis; PI Introdution The novel sulfated oligosaharide agent known as PI-88 has reently been identified as a promising inhibitor of tumour growth and metastasis 1 and is urrently undergoing linial evaluation in aner patients. PI-88 is an inhibitor of angiogenesis by virtue of its ability to blok heparan sulfate binding of angiogenesis-induing growth fators. PI-88 also bloks metastasis by inhibiting heparanase, 2 a key enzyme involved in the degradation of the extraellular matrix surrounding tumour ells, thus preventing their spread to other sites * Corresponding author. Tel.: ; fax: address: vito.ferro@progen.om.au (V. Ferro). via entry into blood vessels and lymphatis. In addition to its antianer ativities, PI-88 shows promise as a potential antioagulant/antithromboti agent with a novel mode of ation. 3 5 It is a speifi ligand for heparin ofator II, enhaning its ability to inhibit thrombin (fator IIa). However, it does not interat with antithrombin-iii and thus shows no anti-xa or AT-III mediated ant-iia ativity. PI-88 is prepared 6 by the sulfonation of the oligosaharide phosphate fration (OPF) of the extraellular phosphomannan produed by the yeast Pihia (Hansenula) holstii NRRL Y ,8 Early strutural studies of the OPF found that the prinipal omponent is the pentasaharide phosphate 1 9 and that smaller amounts of tetrasaharide 9,10 and hexasaharide 10 were also present. More reently, NMR and mass spetral studies have demonstrated that it ontains /02/$ - see front matter 2002 Elsevier Siene Ltd. All rights reserved. PII: S (01)

2 140 V. Ferro et al. / Carbohydrate Researh 337 (2002) substantial amounts of tetrasaharide phosphate 3, as well as minor amounts of hexa-, tri- and disaharide phosphates. 7 The presene of the phosphate group in these oligosaharides prevents their suessful separation by size-exlusion hromatography (SEC); 7 thus, the amounts of eah omponent present were not quantified. An important aspet of the prodution of PI-88 for linial trials is the availability of analytial proedures for the haraterisation and quality ontrol of the intermediate OPF. We now report on the determination of the omposition of the OPF by apillary eletrophoresis (CE) with indiret UV detetion ombined with HPLC analysis of dephosphosphorylated OPF. 2. Results and disussion Reently, a large-sale preparation of the OPF was developed 8 that gave 1 H and 13 C NMR and mass spetral data in aord with the literature. 7 The OPF obtained from suh a preparation 8 was subjeted to detailed strutural haraterisation by hemial and spetrosopi means. Methylation/fragmentation followed by GC MS analyses of the alditol aetates (data not shown) revealed three major omponents orresponding to (a) 1,2,5-tri-O-aetyl-3,4,6-tri-O-methyl-Dmannitol (from 1 2 Man); (b) 1,3,5-tri-O-aetyl- 2,4,6-tri-O-methyl-D-mannitol (from 1 3 Man); and () 1,5,6- tri- O- aetyl- 2,3,4- tri- O- methyl- D- mannitol (from nonreduing end Man6P). Ion hromatographi analysis demonstrated that the OPF ontained % of phosphate, whih is onsistent with a single phosphate group in eah oligosaharide omponent. A ombination of 1D and 2D (DQF COSY, HMBC, HMQC, NOESY) NMR spetrosopy (data not shown) then onfirmed the struture of the major omponent as the pentasaharide phosphate 1. 7 Capillary eletrophoresis (CE) was next investigated to determine the proportions of the omponents of the mixture. CE has been developed into a powerful and highly effiient tehnique to analyse a wide variety of moleules, allowing rapid, repetitive, high-resolution analysis with little, if any, sample preparation. The use of CE for arbohydrate analysis has seen a rapid inrease in reent years, despite the inherent lak of hromophores for onventional UV detetion. This apparent drawbak has been overome hiefly by the use of derivatisation strategies that typially introdue a hromophore or fluorophore by redutive amination, thus allowing detetion by UV absorbane or fluoresene. Derivatisation reations, however, may not always be possible or suitable, partiularly for quality ontrol purposes beause of variable yields. An alternative, ommonly used strategy that allows diret sample analysis is that of indiret UV detetion, whih requires the use of a suitable UV-absorbing bakground eletrolyte (BGE). 15 A number of BGEs have been tested for the separation of low-moleular weight arbohydrates with indiret UV detetion, e.g., sorbi aid, 15,16 riboflavin, 17 5-sulfosaliyli aid, 18 tryptophan 19 and 1- naphthylaeti aid. 20 In some works, poor peak resolutions and poor baseline noise have been reported. The BGE hosen for the analysis of OPF was 6 mm potassium sorbate sine it has been used previously for the analysis of aidi monosaharides. 21,22 Sine the omponent phosphorylated oligosaharides are naturally ionised, a relatively low ph of was used in previous studies. 21,22 However, with OPF, suh a low ph gave poor resolution and peak shape, as well as an unaeptably long run time. The previously reported highly alkaline onditions (ph 12) 15,16 required for neutral saharides also gave an unstable baseline. The optimised onditions were found to be ph 10.3, apillary temperature of 20 C, and operating voltage of 18 kv. Although the UV absorbane maximum of sorbi aid is 254 nm, the use of 214 nm as the detetion wavelength was preferred as this gave a less noisy baseline. These onditions resulted in a stable baseline, good resolution and good reproduibility, as depited in Fig. 1. Fig. 1 surprisingly shows that the OPF is made up of at least seven omponents. The two major omponents (Peaks B and C) are presumably due to the penta- and tetrasaharide phosphates 1 and 3, respetively, and make up approximately 59 and 28.5%, respetively, of the total arbohydrate ontent. The amount of 3 is slightly higher than the previous estimate based on ESIMS data. 7 As expeted, the eletropherogram also indiates that there is a small amount of hexasaharide phosphate (Peak A, 1%) and disaharide phosphate 7 (Peak F, 3%). However, the presene of monosaharide phosphate (Peak G, 3%) and an extra peak in addition to trisaharide phosphate 5 (Peak D or E, 5.5% total) were not antiipated (the seond peak is presumably a previously unidentified tri- or tetrasaharide phosphate). Peak G was onfirmed as mannose 6-phosphate by spiking experiments with pure mannose 6-phosphate standard. Samples originating from phosphomannan bathes from several different fermentation runs were similarly analysed. Some samples ontained no mannose 6-phosphate, indiating that it is unlikely to be an atual minor omponent of the phosphorylated side hains of the phosphomannan, but rather a byprodut of the hydrolysis of the larger oligosaharide phosphates. The exhaustive, aidatalysed hydrolysis of the phosphomannan has, in fat, been shown to be a good soure of mannose 6-phosphate. 23 The amounts of the other minor omponents did not hange appreiably, but some slight variation was seen in the ratio of pentasaharide

3 V. Ferro et al. / Carbohydrate Researh 337 (2002) Fig. 1. Capillary eletropherogram of the oligosaharide phosphate fration. phosphate 1 (Peak B) to tetrasaharide phosphate 3 (Peak C) ranging from 2:1 to 3:2. The amount of Peak D relative to Peak E was also variable. Only one trisaharide (6) and one tetrasaharide (4) have previously been isolated from phosphomannan hydrolysates, either from the neutral oligosaharides released 8 or from speifi dephosphorylation of the OPF by treatment with aqueous HF. 7 Other than the ommerially available mannose 6- phosphate, pure samples of the oligosaharide phosphates were not available (beause they are not amenable to hromatographi separation). Other means were thus sought to onfirm the identity of the peaks. Diret HPLC analysis of the OPF was not suessful, giving a single peak with either an SEC olumn or a arbohydrate (ation-exhange) olumn. However, the orresponding neutral oligosaharides are amenable to HPLC analysis and an be isolated by preparative SEC on Bio-Gel P-2. 7,8 A sample of the OPF was thus dephosphorylated by treatment with alkaline phosphatase. The resulting neutral oligosaharides were then analysed by HPLC using a Rezex RNO-Oligosaharide olumn (Fig. 2). The various peaks were readily identified by omparison with the pure oligosaharides 2, 4, 6 and 8, previously isolated as byproduts of the hydrolysis. 8 The results mirrored those seen in the CE analysis, with the exeption that only one peak was learly disernible in the trisaharide region (for 6). The isolation and determination of the struture of the (neutral) hexasaharide, whih had not been done previously beause of its low abundane, were next pursued. The neutral oligosaharide fration from a large-sale hydrolysis of the phosphomannan, 8 represented by the unbound effluent from the ion-exhange hromatography step, was olleted, onentrated by reverse osmosis and freeze dried. Analysis of this sample by HPLC revealed a similar distribution of oligosaharides as seen in the alkaline phosphatase experiment, with the exeption of the amount of disaharide 8. The inreased amount of 8 is due to it being the major non-phosphorylated apping unit of the phosphomannan side hains. 7,8 The sample was then subjeted to repeated SEC on a olumn of Bio-Gel P-2, resulting in the isolation of 13 mg of material, orresponding to the hexasaharide (Peak A, Fig. 2) that was of suffiient purity ( 80% by HPLC) for NMR spetrosopi analysis. This material was onfirmed as being predominantly hexasaharide by ESIMS (positive-ion mode, m/z [M+Na] + ). Its struture was then firmly established as the hexasaharide 10, in whih an extra D-mannose unit is (1 3)-linked at the nonreduing end, by a ombination of 1D ( 1 H and 13 C) and 2D (gcosy, ghsqc) NMR spetrosopy and omparison with the literature 7,24 (NMR data are presented in Table 1). The orresponding hexasaharide phosphate is thus presumably 9, with the phosphate group at the terminal 6-OH of the nonreduing end.

4 142 V. Ferro et al. / Carbohydrate Researh 337 (2002) Fig. 2. HPLC hromatogram of the neutral oligosaharides obtained upon dephosphorylation of the oligosaharide phosphate fration by alkaline phosphatase. During the ourse of the repeated SEC, it beame lear by HPLC analysis of the frations that there was indeed an extra omponent present in the mixture eluting lose to the trisaharide (t R =22.3 min vs. t R =22.9 min for trisaharide 6, data not shown). When analysing the bulk sample, this peak was buried under the trisaharide peak (Peak D) and thus not visible. It eluted towards the end of the tetrasaharide fration (Peak C, t R =19.7 min) and the beginning of the trisaharide fration (Peak D). Unfortunately, all attempts to isolate it diretly failed and only mixed frations were obtained. One suh fration whih ontained only 4 (95%) and the unidentified ompound (5%) was examined by ESIMS (positive-ion mode). This showed a moleular-ion peak orresponding to tetrasaharide (m/z [M+Na] + ), but no peaks of the expeted intensity were disernible for a trisaharide. A seond fration whih ontained the unidentified ompound (6.5%), 4 (85%) and 2 (8.5%) was then peraetylated under standard onditions, and the minor omponent orresponding to the peraetate of the un- Table 1 13 Cand 1 H NMR hemial shift data for hexasaharide 10 (major anomeri omponent) Residue C-1/H-1 C-2/H-2 C-3/H-3 C-4/H-4 C-5/H-5 C-6/H-6a,b I b (1.6) a II b (1.9) d III e b (3.0, 9.4) 3.76 d IV e b (2.8, 9.4) 3.76 d V b d VI (1.6) 4.07 (1.8, 3.4) (9.8, 9.8) 3.79 a Seleted oupling onstants in Hz. b Linkage arbons. Chemial shifts for H-6 s: and d Chemial shifts for H-5 s: e Values may have to be interhanged.

5 Table 2 13 Cand 1 H NMR hemial shift data for ompound 11 V. Ferro et al. / Carbohydrate Researh 337 (2002) Residue C-1/H-1 C-2/H-2 C-3/H-3 C-4/H-4 C-5/H-5 C-6/H-6a,b I a (2.6) b 5.03 (10.2) 5.19 (10.0) 3.71 (3.4, 3.8) 4.08, 4.08 II a (2.3) 5.17 (3.4) 4.20 (9.8) 5.18 (10.3) 4.12 (6.7, 2.4) 4.00, 4.36 (12.3) III a (1.9) 4.99 (3.7) 3.96 (9.9) 5.34 (9.9) 3.86 (3.6, 2.9) 4.03, 4.17 (13.3) IV (1.9) 5.01 (3.3) 5.16 (10.0) 5.25 (10.0) 3.90 (4.7, 2.4) 4.02, 4.25 (12.3) Other 13 C NMR data: (13 C, OCOCH 3 ), 23.4 (NHCOCH 3 ), (14 C, C O). Other 1 H NMR data: (13s, COCH 3 ), 2.02 (s, NHCOCH 3 ), 7.23 (d, J 1,NH 9.7 Hz, NH). a Linkage arbons. b Seleted oupling onstants in Hz. Values may have to be interhanged. known ompound was isolated by repeated reversedphase HPLC. The produt, whih was 84% pure by HPLC, was identified as the tetramannosylamine peraetate 11 by ESIMS (m/z [M+Na] + ) and by a ombination of 1D ( 1 H and 13 C) and 2D (gcosy, ghsqc) NMR spetrosopy (NMR data are presented in Table 2). In the 1 H NMR spetrum of 11, the presene of the amide group is onfirmed by the doublet at 7.23 (J 1,NH 9.7 Hz) that is oupled to the reduing end anomeri proton at 5.35 (d, J 1,2 0, J 1,NH 9.7 Hz). Further onfirmation is provided by the 13 C NMR spetrum that shows that the orresponding anomeri arbon atom is shifted upfield to 76.2, and that there is a single N-aetyl signal at The unidentified ompound is thus the tetramannosylamine 12, whih suggests that one of either Peak D or Peak E in the eletropherogram of the OPF (Fig. 1) orresponds to the tetramannosylamine phosphate 13, whilst the other orresponds to trisaharide phosphate 5. There is no evidene to indiate the presene of other glyosylamines in the OPF, and its origin and possible role in the intat phosphomannan polysaharide is a matter for speulation. Its presene may have impliations for the biosynthesis and release of the phosphomannan from the yeast ells. In onlusion, the omposition of the OPF from P. (Hansenula) holstii NRRL Y-2448 phosphomannan has been determined by CE and HPLC, and the hexasaharide omponent has been isolated and haraterised by NMR spetrosopy. In addition, the presene of a tetramannosylamine omponent has been deteted, and its struture has been haraterised by NMR spetrosopy of its peraetate derivative. In so doing, further light has been shed on the struture of the phosphomannan, and a CE method has been developed that is suitable for quality ontrol of the OPF for use in the manufature of PI-88. It has been demonstrated that CE with indiret UV detetion is an ideal method for the analysis of losely related phosphorylated oligosaharides as it allows rapid separations with high resolution and minimal sample preparation. 3. Experimental General. Mass spetra were obtained by eletrospray-ionisation (ESIMS) on a Fisons VG Quattro II mass spetrometer. Analytial HPLC was performed on a Waters Alliane 2690 Separations Module using a Rezex RNO-Oligosaharide 12- m ( mm) olumn (mobile phase: water, flow rate: 0.3 ml/min, olumn temperature: 80 C). Detetion was with a Waters 2410 refrative index detetor. Reversed-phase HPLC was performed using a Waters Nova-Pak C m ( mm) olumn with detetion by UV absorbane at 230 nm (Waters 2487 UV detetor). Thin-layer hromatography (TLC) was performed on E. Merk Kieselgel 60 F 254 aluminium-baked sheets with speified eluents. Compounds were deteted by harring with 10% aq H 2 SO 4. Reverse osmosis was performed on a Millipore Prosale system (operating temperature C, feed pressure 1200 kpa, inlet flow rate 8 L/min) equipped with a Helion RO4 artridge with Nanomax 50 membrane (membrane area 0.37 m 2 ). All reagents were analytial grade and were used without further purifiation. All water was purified in house to USP purified water standard. DEAE-Spherilose, Bio-Gel P-2, alkaline phosphatase and mannose 6-phosphate were from ISCO, BioRad, ICN and Sigma, respetively. The oligosaharide phosphate fration (OPF) from P. (Hansenula) holstii

6 144 V. Ferro et al. / Carbohydrate Researh 337 (2002) NRRL Y-2448 phosphomannan was prepared as previously desribed. 8 Capillary eletrophoresis. Capillary eletrophoresis was performed on a Bekman P/ACE 5000 System equipped with a P/ACE UV Absorbane Detetor. Bekman ecap fused silia apillaries (50 m i.d., 375 m o.d., length 107 m, 100 m to detetor) were used in these analyses. The bakground eletrolyte (BGE) was 6 mm potassium sorbate, ph 10.3 (adjusted with KOH), apillary temperature was 20 C and operating voltage 18 kv. Samples were prepared by dissolution in water at a onentration of 1 mg/ml and were injeted by pressure injetion for 10 s. Detetion was by indiret UV absorbane at 214 nm. The apillary was rinsed between eah run for 10 min with BGE solution. BGE solutions were replaed after approximately 12 analyses. NMR spetrosopy. One-dimensional (1D) 1 H and 13 C NMR and two-dimensional (2D) gcosy and gh- SQC experiments on ompounds 10 and 11 were performed on a Varian Unity 400 spetrometer with standard Varian software at 25 C. Spetra were referened relative to internal aetone ( for 1 Hand 31.0 for 13 C) for solutions in D 2 O(10), or to the solvent peaks ( 7.23 for 1 Hand 77.0 for 13 C) for solutions in CDCl 3 (11). 1D 1 H and 13 C NMR and 2D DQF

7 V. Ferro et al. / Carbohydrate Researh 337 (2002) COSY, NOESY, HMQC and HMBC experiments on the OPF were performed on a JEOL 400 MHz spetrometer with standard JEOL software at 30 C. Methanolysis of oligosaharides. Oligosaharide samples were thoroughly dried in vauo over P 2 O 5 and dissolved in 0.5 ml of dry methanoli 1 M HCl using srew-apped tubes with Teflon-lined septa in the srew ap. Nitrogen gas was bubbled through the solution for 15 s, and then the tubes were apped. After methanolysis for 24 h at 80 C, the aidi solution was neutralised by the addition of 0.15 ml of pyridine, and amino sugars were N-reaetylated by the addition of 0.1 ml of A 2 O. The sample solution was evaporated with nitrogen gas flow at 35 C. The residue was dried for 16 h in vauo over P 2 O 5. Finally, the sample was trimethylsilylated with 30 L of 2:1 N,O-bis(trimethylsilyl)- trifluoroaetamide (BSTFA) pyridine at 80 C for 1 h. Gas hromatography with flame-ionisation detetion was arried out on a Hitahi G-3000 gas hromatograph equipped with DB-1 fused silia apillary olumn (375 m i.d. 25 m). GC MS analyses were performed on a Hewlett Pakard gas hromatograph model 5890 series II with a DB-1 fused silia apillary olumn (375 m i.d. 25 m), followed by detetion with a mass-seletive detetor model 5971 equipped with a Chemistation data system. All analyses were performed in the eletron-impat ionisation mode, and an ionising voltage of 70 ev was used. The temperatures of the injetion port and the detetor were set at 280 C. The olumn temperature was programmed from 120 to 270 C at the rate of 20 C/min and then held isothermally at 270 C. Methylation analysis aetylation and GC MS analysis. The sample was permethylated by the Hakomori method, 25 and permethylation was arried out repeatedly to obtain permethylated oligosaharide. The produt was subjeted to aetolysis in 80% AOH ontaining 1 M HCl at 80 C for 24 h, and the resultant partially O-methylated monosaharides were redued by 5% NaBH 4 in 10 mm NaOH. After removal of borate by evaporation as methylborate, partially O- methylated alditols were aetylated. 26 The partially methylated alditol aetates were analysed by GC MS under the same onditions as desribed above. Ion-hromatography. A sample of the OPF was hydrolysed with 2.5 M trifluoroaeti aid at 100 C for 6 h under nitrogen. After entrifugation at 1800g for 10 min, the sample was dried at 50 C under a stream of nitrogen and dissolved in water. Inorgani phosphate was analysed by ion-hromatography using a TSK gel IC-anion-PW (4.6 mm i.d. 50 mm) suppressor olumn and a Dowex 50W-X (H+) (5.0 mm i.d. 200 mm) olumn. The temperature of the separation was ontrolled at 30 C, and the mobile phase was 1.42 mm NaHCO 3 and 1.5 mm Na 2 CO 3 at a flow rate of 1.0 ml/min. Detetion was by suppressed ondutivity using a Tosoh model CM-8 (Tosoh Co., Tokyo). Dephosphorylation of the OPF with alkaline phosphatase. A sample of the OPF (100 mg) was dissolved in buffer (1 mm MgCl 2, 0.1 mm ZnCl 2,10mMdiethanolamine, 1 ml), ph 9.5, and treated with alkaline phosphatase (520 units, 0.4 mg) at 26 C for 24 h, by whih time TLC (7:3 EtOH 1 M ammonium aetate) indiated omplete onsumption of the starting material. The mixture was then passed through a olumn of DEAE-Spherilose ( m) equilibrated with 0.01 MNH 4 HCO 3. The frations ontaining produt were then ombined, desalted by passage through a olumn of Sephadex G-25M (Pharmaia PD-10) and analysed by HPLC. Isolation of hexasaharide 10. A sample of the unbound effluent from the ion-exhange hromatography step from a large-sale hydrolysis of phosphomannan 8 (50 L) was onentrated by reverse osmosis and then lyophilised. The residue was subjeted to SEC on a olumn of Bio-Gel P-2 (5 100 m) using 0.1MNH 4 HCO 3 as eluent. The frations were monitored by HPLC. The hexasaharide-enrihed frations were ombined, onentrated and subjeted to repeated SEC on olumns of Bio-Gel P-2 (firstly m and then m) using water as eluent. The hexasaharide 10 was isolated as a white, amorphous solid (13 mg) after lyophilisation. The purity was 80% by HPLC (t R =16.2 min). ESIMS: m/z [M+Na] +. 1 H and 13 C NMR data are presented in Table 1. Also isolated was a fration ontaining ompounds 2 (8.5%, t R =17.6 min), 4 (85%, t R =19.7 min) and 12 (6.5%, t R =22.3 min). Isolation of peraetate 11. The above fration ontaining ompounds 2 (8.5%), 4 (85%) and 12 (6.5%) was lyophilised and then aetylated by treatment with exess A 2 O pyridine (25 C, 24 h). Workup (CH 2 Cl 2 ) gave a residue (846 mg) that was subjeted to repeated reversed-phase HPLC employing a linear gradient of 30 50% MeCN water over 35 min (flow rate: 2 ml/ min). This resulted in the isolation of the peraetate 11 as a white, amorphous solid (11 mg, t R =11.6 min). The purity was 84.5% by reversed-phase HPLC. ESIMS: m/z [M+Na] +. 1 H and 13 C NMR data are presented in Table 2. Aknowledgements We thank the Griffith University Magneti Resonane Faility for the use of their NMR failities. We also thank Mr Jeff Rood for tehnial assistane, and Drs Robert Don and Denis Podger for useful disussions.

8 146 V. Ferro et al. / Carbohydrate Researh 337 (2002) Referenes 1. Parish, C. R.; Freeman, C.; Brown, K. J.; Franis, D. J.; Cowden, W. B. Caner Res. 1999, 59, Parish, C. R.; Freeman, C.; Hulett, M. D. Biohim. Biophys. Ata 2001, 1471, M99 M Wall, D.; Douglas, S.; Ferro, V.; Cowden, W.; Parish, C. Thromb. Res. 2001, 103, Demir, M.; Iqbal, O.; Hoppensteadt, D.; Piolo, P.; Ahmad, S.; Shultz, C. L.; Linhardt, R.; Fareed, J. Clin. Appl. Thromb. Hemost. 2001, 7, Piolo, P.; Iqbal, O.; Demir, M.; Ma, Q.; Gerbutaviius, R.; Fareed, J. Clin. Appl. Thromb. Hemost. 2001, 7, Yu, G.; Gunay, N. S.; Linhardt, R. J.; Toida, T.; Fareed, J.; Hoppensteadt, D. A.; Shadid, H.; Ferro, V.; Li, C.; Fewings, K.; Palermo, M. C.; Podger, D., submitted. 7. Parolis, L. A.; Parolis, H.; Kenne, L.; Meldal, M.; Bok, K. Carbohydr. Res. 1998, 309, Ferro, V.; Fewings, K.; Palermo, M. C.; Li, C. Carbohydr. Res. 2001, 332, Bretthauer, R. K.; Kazorowski, G. J.; Weise, M. J. Biohemistry 1973, 12, Slodki, M. E.; Ward, R. M.; Boundy, J. A. Biohim. Biophys. Ata 1973, 304, El Rassi, Z. Eletrophoresis 1999, 20, Suzuki, S.; Honda, S. Eletrophoresis 1998, 19, Olehno, J. D.; Nolan, J. A. In Handbook of Capillary Eletrophoresis; Landers, J. P., Ed. Carbohydrate Analysis by Capillary Eletrophoresis, 2nd edition; CRC: Boa Raton, FL, 1997; pp Koketsu, M.; Linhardt, R. J. Anal. Biohem. 2000, 283, Vorndran, A. E.; Oefner, P. J.; Sherz, H.; Bonn, G. K. Chromatographia 1992, 33, Klokow, A.; Paulus, A.; Figueiredo, V.; Amadò, R.; Widmer, H. M. J. Chromatogr. A 1994, 680, Xu, X.; Kok, W. T.; Poppe, H. J. Chromatogr. A 1995, 716, Damm, J. B.; Overklift, G. T. J. Chromatogr. A 1994, 678, Lu, B.; Westerlund, D. Eletrophoresis 1996, 17, Lee, Y. H.; Lin, T. I. J. Chromatogr. B 1996, 681, Bergholdt, A.; Overgaard, J.; Colding, A.; Frederiksen, R. B. J. Chromatogr. 1993, 644, Ciringh, Y.; Lindsey, J. S. J. Chromatogr. A 1998, 816, Slodki, M. E. US Patent , 1961; Chem. Abstr. 1962, 56, Ogawa, T.; Sasajima, K. Carbohydr. Res. 1981, 97, Hakomori, S. J. Biohem. (Tokyo) 1964, 55, Anumula, K. R.; Taylor, P. B. Anal. Biohem. 1992, 203,