High-Sensitivity Analysis of Neutral Underivatized Oligosaccharides

rides to larger polymers with molecular masses up to 6. kDa, sample ..... initial (and possibly later generation) droplets will have a chance to pass ...
0 downloads 0 Views 884KB Size
Anal. Chem. 1997, 69, 4530-4535

High-Sensitivity Analysis of Neutral Underivatized Oligosaccharides by Nanoelectrospray Mass Spectrometry Ute Bahr,* Anja Pfenninger, and Michael Karas

Division for Instrumental Analytical Chemistry, JW Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany Bernd Stahl

Milupa Research, Milupa GmbH&Co KG, 61381 Friedrichsdorf, Germany

Nanoscale electrospray ionization (nano-ESI) overcomes the sensitivity problems found and reported for neutral oligosaccharides in conventional microscale forced-flow ESI. For a series of compounds ranging from trisaccharides to larger polymers with molecular masses up to 6 kDa, sample concentrations of 10-5 M, i.e., 10 pmol total sample load, yielded very intense singly or multiply cationized molecule ions in an ion-trap mass spectrometer. In a dilution series, it is exemplified that, at the 10-8 M level, molecule ion signals can be clearly registered with a S/N ratio of about 7. Only 100 amol of sample has been consumed in this experiment. Investigation of an oligosaccharide-peptide mixture revealed that the oligosaccharide is suppressed in conventional ESI, whereas in nano-ESI both analytes are detected at comparable and high intensities. Mechanistic implications are discussed, emphasizing the influence of surface activity for the two ESI techniques. The very low flow rates inherent to nanoESI of about 30 nL/min, together with the high signal intensity, make it possible to fully employ the MSn capabilities of an ion-trap mass spectrometer for structural analysis. From less than 1 µL of sample solution, it is possible to make consecutive fragmentation experiments up to MS7 to obtain valuable information about the structure of complex oligosaccharides. Within the growing field of glycobiology, mass spectrometry is gaining increasing importance as an analytical technique for carbohydrate characterization. Molecular mass determination, often from heterogeneous mixtures, and compositional analysis usually form the first step. The more challenging problem, however, is the structural characterization of a carbohydrate sample, because, in contrast to the linear protein and nucleic acid biopolymers, carbohydrates exhibit a wide structural variety due to different branching of the carbohydrate core and linkage of the monosaccharide subunits as well as derivatization of monomers by, e.g., sulfate or phosphate. Whereas, however, matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI) have truly revolutionized the mass spectrometry of peptides and proteins, applications to the study of nucleic acids and especially of oligosaccharides still are comparably limited in number. As a proof for this observation, it may serve that, within the 2-year Fundamental 4530 Analytical Chemistry, Vol. 69, No. 22, November 15, 1997

Reviews appearing in Analytical Chemistry, the 19941 review dedicated a full chapter to carbohydrates and glycoconjugates, whereas in 19962 only glycosylation of peptides and proteins was accounted for. Moreover, most of the work on neutral carbohydrates is performed with MALDI-MS rather than with ESI. The major reason for this appears to be that, until now, underivatized carbohydrates could not be measured by ESI with the same high level of sensitivity as peptides and proteins. Thus, Burlingame et al. state, “unlike the usual approach to proteins, protocols dealing with...carbohydrate moieties involve chemical derivatization in their strategies to effectively cleanup..., counteract hydrophilicity of the oligosaccharide, and enhance sensitivity of mass spectral detection...”.1 In agreement with that, Reinhold et al. state in their compilation of structure elucidation of carbohydrates3 that “due to the weak acidity or basicity...free oligosaccharides do not efficiently charge by protonation and deprotonation” and use derivatization techniques except for acidic carbohydrates or some glycoconjugates. They conclude that “derivatization melds carbohydrate structural analysis for two fundamental reasons. First, ES when using more volatile organic solvents shows significant increases in detection sensitivity, and second, methylation brings unique insight to structure...”. In this paper, we want to challenge these statements and the rationalization of limited sensitivity for carbohydrates in ESI. We will show for a selection of neutral carbohydrates widely varying in their composition and ranging from trisaccharides to about 6 kDa in mass that, when using nanoscale electrospray ionization (nano-ESI),4 carbohydrates can be measured with sensitivity levels comparable to those for peptides and that this even holds for mixtures of carbohydrates and peptides. As we also observe low sensitivity for carbohydrates in mixtures with peptides in conventional microscale forced-flow ESI, we come to the conclusion that the problems in carbohydrate analysis do not arise from the fact that carbohydrates are refractory to desolvation and ionization but are rather due to their hydrophility and lack of surface activity. In addition, it will be shown that nano-ESI together with the MSn capability of a quadrupole ion trap instrument is a valuable tool (1) Burlingame, A. L.; Boyd, R. K.; Gaskell, S. J. Anal. Chem. 1994, 66, 634R683R. (2) Burlingame, A. L.; Boyd, R. K.; Gaskell, S. J. Anal. Chem. 1996, 68, 599R651R. (3) Reinhold, V. N.; Reinhold, B. B.; Costello, C. E. Anal. Chem. 1995, 67, 1772-1784. (4) Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1-8. S0003-2700(97)00624-0 CCC: $14.00

© 1997 American Chemical Society

Table 1. Oligosaccharides Analyzed with ESI-MS oligosacharide

molecular mass (Da)

melibiose 1-kestose melezitose fructo-OS DP4 nystose maltopentaose cellupentaose cycloheptaamylose dextran DP 11 dextran 1000 maltodextrin MD 25 pullulan 5800 neutral OS (human milk) chito-OS DP4

342 504 504 666 666 828 828 1134 1800 m m m m 830

structure R-Gal-(1-6)-Glc β-Fru-(2-6)-β-Fru(2-1)-R-Glc R-Glc-(1-3)-β-Fru-(2-1)-R-Glc (β-Fru-(2-1))4 (β-Fru-(2-1))3-R-Glc (R-Glc-(1-4))5 (β-Glc-(1-4))5 cyclo-(R-Glc-(1-4))7 (R-Glc-(1-6))11 (R-Glc-(1-6))n (R-Glc-(1-4))n ((R-Glc-(1-4)3)(1-6)n (β-Glc-NAc-(1-4))4

origina Sigma SZ Sigma MRL/GPC SZ Merck Sigma Sigma MRL/GPC Fluka MRL PSS GPC/MRL SZ

a SZ, Su ¨ dzucker AG, ZAFES, Obrigheim, Germany; MRL, Milupa Research Laboratory; PSS, Polymer Standard Service, Mainz, Germany; GPC, fractionized or cleaned by GPC; m, mixture.

for structure elucidation of small sample amounts of underivatized oligosaccharides. EXPERIMENTAL SECTION Materials and Sample Preparation. Table 1 lists the analyzed oligoaccharides together with their molecular masses and structures. For ESI analysis, samples were dissolved in 50: 50 water/methanol (v/v) (HPLC grade) at a concentration of 10-5 M for single-compound samples; for mixtures, an average of 10-5 M per component was adjusted. These samples were investigated by nano-ESI or conventional microscale forced-flow ESI without desalting or addition of salt. For the sensitivity experiments, the starting solution of 10-5 M was desalted by NH4+-loaded cation exchange beads5 and then diluted to yield 10-6, 10-7, and 10-8 M solutions; NaCl was added to all solutions to a final concentration of 10-5 M. Insulin was dissolved to 10-5 M in a water/methanol/ acetic acid (49:49:2 v/v/v) mixture. For mixture analysis, aliquots of insulin and maltopentaose (10-5 M) solutions were mixed together. Electrospray Sources and Mass Spectrometry. Mass spectrometry was performed with an LCQ ion-trap mass spectrometer (Finnigan MAT, San Jose, CA). Two different electrospray sources were used: (1) the standard ESI source of the LCQ instrument, run with the syringe pump filled with 30-50 µL of solution at a flow rate of 1 µL/min with nebulizer gas, and (2) a commercial nanospray ESI source (The Protein Analysis Co., Odense, Denmark) using home-pulled glass capillaries with an orifice diameter of 1-2 µm, which were positioned directly at a distance of