Sequencing The Decorin Glycosaminoglycan Chains On The Basis Of Glycomics

Glycomics represents one of the last frontiers and most challenging in omic analysis.The recent and rapid progress in genomics and proteomics has not been matched in glycomics,one reason for the slow progress in establishing glycomes of various organisms is the complexities associated in the structural characterization and sequencing of complex carbohydrates.Glycosylation occurs in the endoplasmic reticulum and the Golgi organelle and its control is neither well understood nor predictable based on proteomic or genomic analysis.One of the most structurally complex classes of glycoconjugates is the proteoglycans(PGs)and their glycosaminoglycan(GAG)side chains.The first and the only one sequence identified proteoglycan was bikunin.The current study examines the second GAG structure from porcine skin decorin.Decorin PGs was extracted from porcine skin by sodium citrate,GAG was obtained through thoroughly actinase E digestion and amino acids at the reducing ends were removed by P-elimination.Decorin GAG structure characterization using PAGE,GPC,NMR,LC-MS/MS and anticoagulant activity were undertaken.The initial decorin GAG chains had an average molecular mass of 22 kDa,a chain length ranging from dp14-dp290 and an average sulfation density of 0.98 sulfo groups/disaccharide repeating unit.Disaccharide compositional analysis,performed by chondroitin lyase catalyzed depolymerization of the decorin GAG to unsaturated disaccharides(with a non-reducing terminal AUA,deoxy-?-L-threo-hex-4-enopyranosiduronic acid)and HPLC-MS analysis,afforded a composition of 94.4 mol%?UA(1?3)GalNAc4S,4.3 mol%?UA2S(1?3)GalNAc4S,0.1 mol%?UA(1?3)GalNAc6S,0.5 mol%?UA(1?3)GalNAc.Next,we examined the AC/B domain structure of the decorin GAG chain using NMR spectroscopy,which showed an IdoA/IdoA2S:GlcA ratio of 3:1.The anticoagulant activity assay showed it had a very low activity compared the commercial anticoagulant drug heparin.Domain mapping based on B domain contains IdoA or IdoA2S residues and is susceptible to treatment with endolytic chondroitin B lyase,while the AC domain is susceptible to treatment with endolytic chondroitin AC lyase.Thus,exhaustive treatment of decorin GAG with chondroitin B lyase and chondroitin AC lyase and recovery of intact chains of reduced size afforded AC and B domains,respectively.This analysis is consistent with the presence of 2-O-sulfo groups present only on IdoA residue and only in B domain.HILIC-LC-Orbitrap-MS analyzed the AC and B domains chains,as expected,the B domains were often>dp10 and were on average?3-times longer than the AC domains which were most frequently dp6.Exhaustive treatment of decorin GAG with chondroitin layse ABC afforded tetrasaccharide linkage region,bottom up method HILIC-LC-MS analysis confirmed the presence of the two major reduced linkage region structures,?4)GlcA(1?3)Gal(±4S)(1?3)Gal(1?4)Xylitol comprising>90 mol%of the decorin chains.Disaccharides analysis shows that the AC domains exclusively contain?UA(1?3)GalNAc4S,?UA(1?3)GalNAc6S,?UA(1?3)GalNAc at 93.2 mol%,6.0 mol%and 0.8 mol%,respectively,and B domain exclusively contains only AUA(1->3)GalNAc4S at 95.3 mol%and AUA2S(1?3)GalNAc4S at 4.7 mol%.The first two chapters focus on general structure motif,not about a single GAG chain.Our experience in MS sequencing the much less complex bikunin GAG chain suggested that while a sulfation level of 0.98 O-sulfo groups/disaccharide repeating unit of the decorin GAG posed no problem,the large chain length,up to dp290,was well beyond our MS capabilities.Thus,we set out to fractionate the decorin GAG chains to obtain relatively homogenous fractions of chains of average sulfation level but of small sizes.Decorin GAG(?1 g)of molecular weight(avg)22 kDa was first fractionated by size exclusion chromatography(SEC)fast performance liquid chromatography(FPLC)and approximately the last third of the eluting sample,corresponding to Mw(avg)19 kDa was collected.Next,250 mg of this fraction was applied to strong anion exchange(SAX)-FPLC and the center third of the peak was collected and subjected to SEC-HPLC to again enrich 5 mg of the small chains.Finally,preparative PAGE was applied to obtain 135 ftactions that were analyzed by analytical PAGE of microgram quantitites of fractions ranging in estimated size from 3 kDa to 32 kDa.Orbitrap FTMS analysis was then undertaken on selected PAGE fractions ranging from#38(Mw?4.7 kDa)to#60(Mw?8.7 kDa)to determine the intact mass of fraction components and to assess the suitability of these and neighboring fractions for MS/MS sequencing.Accurate masses were obtained for 57 molecular ions ranging from dp20 to dp44 and carring 8-20 sulfo groups with only a small amount of sodium adduction observed.Fourier transform ion cyclotron resonance mass spectrometry(FT-ICR MS)was applied to neighboring PAGE fraction#39.Initial MS analysis provided molecular compositions which were then selected for sequencing by collisional dissociation.The dp range observed in fraction#39(dp14-26)by FT-ICR MS was slightly wider than had been observed in fraction#38(dp 16-24)by Orbitrap FTMS.An example annotated spectrum of decorin MS/MS where CID-FT-ICR MS/MS of the molecular ion m/z 616.8147,corresponding to dp20-7S,resulted in a spectrum rich in glycosidic bond cleavages suitable for sequence determination.Additional sequence analysis of other tandem MS from fraction#39 and complete composition analysis of fractions#35 and#51 are also provided.Sequencing analysis shows homogeneity in the non-reducing end of porcine decorin GAG.The HexA-GalNAc-S disaccharide motif is consistently repeated at the non-reducing end with no evidence of alternative structures.Although information to determine C-5 uronic sugar stereochemistry is lacking,intense series of B fragments and/or C fragments are presented in all decorin tandem MS for the HexA-GalNAc-S repeating unit.The overall percentage of HexA-GalNAc-4S disaccharide as determined by disaccharide analysis increases with respect to chain length.These findings are reflected in the tandem MS results with the total number of observed sequential B-ions that favor the HexA-GalNAc-S pattern increasing when chains are longer.Observable variations in sulfation patterns of different sequences occurs primarily at the reducing end.GlcA-GalNAc6S,IdoA2S-GalNAc4S and GlcA-GalNAc disaccharide variants occur infrequently as suggested by disaccharide analysis and are minor components in the overall sequence.Sequences derived from tandem MS data show that thses laternative occur 0 to 2 times per chain.Sequential series of Y fragments suggest variability in the region closest to the reducing end after the GlcA-GalNAc4S-GlcA-Gal-Gal-Xyt linker region.The IdoA2S-GalNAc4S modification exists in a similar region.Cross-ring fragments that validate the presence of 6-O-sulfo modification on the GalNAc(hence a GlcA-GalNAc6S unit)are only observed near the reducing end.Sequential Y-ions near the reducing end exist from 2-30%relative ion intensity but exhibit no common intensity-dependent pattern.The reducing end is the region where all variations to the HexA-GalNAc-S pattern exist but are observed in no specific order.The overall sequence motif between dp 14 to dp40 is:[f.e.d.c.b.a.],from non-reducing end[f.GalNAc4S,e.GlcA-GalNAc4S/IdoA-GalNAc4S(25%/75%),d.IdoA2S-GalNAc4S,c.GlcA-GalNAc6S,b.GlcA-GalNAc,a.Hex-GalNAc4S]linker to linkage region(GlcA-GalNAc4S-GlcA-Gal-Gal-Xyt),[f=0-1,e=3-20,d=0-1,c=0-1,b=0-1,a=0-3].