Overview

1
Disaccharide Isomers Differentiated with Infrared
Multiple Photon Dissociation Sarah E. Stefan,
John R. Eyler Department of Chemistry,
University of Florida, Gainesville, FL USA
Results
Results Continued
Overview
A
B

• The purpose of this research was to
  differentiate deprotonated glucose-containing
  disaccharides through wavelength-dependent
  fragmentation using a narrow-band, line-tunable
  laser.
• A tunable CO2 laser coupled to an electrospray
  ionization Fourier transform ion cyclotron
  resonance (FTICR) mass spectrometer was used to
  study the fragmentation patterns of disaccharides
  with various linkages and anomeric conformations.
  
• Irradiation over the wavelength range of 9.2-9.7
  µm yielded unique plots of disaccharide
  fragmentation as a function of both mass and
  laser wavelength.

Glcß1-3Glc
Glca1-3Glc
Ratio
Introduction

• Polysaccharides, the most common carbohydrates,
  are joined by a glycosidic bond to lipids
  (glycolipids) or proteins (glycoproteins) and
  play an essential role in numerous activities in
  the body.
• Disaccharides are the smallest unit containing a
  glycosidic linkage. Previous research on
  lithiated disaccharides in the positive ion mode
  showed that fragmentation gives unique 2D plots
  that are wavelength dependent. 2
• The ability to distinguish isomers of
  disaccharides gives insight into their bonding
  and interactions. It can also help tp determine
  the structure and location of these smaller units
  within larger saccharides.
• Use of FTICR-MS allows for unparalleled mass
  resolving power and mass accuracy, along with the
  ability for selective ion mass isolation.3
• Infrared multiple photon dissociation (IRMPD)
  can produce more extensive fragmentation than
  collision induced dissociation for some
  oligosaccharides.4
• Use of a tunable CO2 laser allows for selection
  of wavelengths between 9.2 and 10.8 µm for
  fragmentation.

Figure 2. Comparison of m/z 161/179 for 1-3 and
1-6 linked disaccharides.
Error bars represent the 95 confidence interval.
Conclusions

• Fragmentation of deprotonated disaccharides with
  various linkages yields different
  wavelength-dependent fragmentation patterns
  (Figure 1).
• The appearance of m/z 281 is unique for the 1-6
  linked disaccharides.
• Fragmentation of 1-3 linked disaccharides yields
  an abundance of approximately 11 for m/z 161m/z
  179.
• Fragmentation of 1-4 linked disaccharides
  results in an average abundance of approximately
  61 for m/z 161m/z 179.
• The ratios of m/z 161 to m/z 179 cannot solely
  be used to identify the anomeric conformation of
  disaccharides with the same linkages.
• Higher laser power may produce lower mass
  fragments whose ratios may be
• used for distinguishing the disaccharides
• Day-to-day variation makes this a poor
  quantitative method, but reproducibility of
  fragmentation patterns makes it a good
  qualitative method.
• An average error of 3 for the percent abundance
  of the precursor ion was seen at 9.588 µm for
  isomaltose. The overall average error for all
  wavelengths was 6, with the highest error
  (19) seen at 9.201 µm.
• ESI and laser power fluctuation may be the cause
  of the variances seen.

C
D
Glcß1-4Glc
Glca1-4Glc
Method

• Analysis was done on a Bruker Bio-Apex II FTICR
  mass spectrometer with a 4.7T superconducting
  magnet and Infinity cell in the negative ion
  mode.
• A pneumatically-assisted Apollo external
  electrospray ionization (ESI) source with a flow
  rate of 3-7 µL/min was used for ionization. Flow
  rate was determined based on ion signal
  intensity.
• A Lasy-20G tunable CO2 laser producing infrared
  irradiation with variable wavelengths from 9.2 to
  10.8 µm was passed through a KBr window into the
  FTICR cell. The typical laser power used in these
  studies was less than 1 watt.
• 10-4 M solutions of glucose-containing
  disaccharides with various anomeric conformations
  (a-and ß-anomers) and linkages (1-3, 1-4 and 1-6)
  with 10-3 M NaOH in 8020 methanolwater were
  used.
• A daily power calibration was done by adjusting
  the laser power to give a ratio of m/z 179 to 341
  of 1.19 0.17 at 9.588 µm for isomaltose
  (Glca1-6Glc) when irradiated for 1 second.
• The signal intensity, irradiation time and laser
  power were kept constant throughout the day.
• For each wavelength, three sets of ten scans of
  128K datasets were collected and averaged. To
  test reproducibility, fragmentation spectra of
  isomaltose at various wavelengths were taken on
  two separate days.

F
E
Glcß1-6Glc
Glca1-6Glc
Acknowledgements

• We would like to thank the University of Florida
  and the National Science Foundation (NSF grant
  No. CHE-0718007) for funding, Dr. David Powell
  for use of the FTICR mass spectrometer in his
  laboratory and Dr. Brad Bendiak of the University
  of Colorado Health Sciences Center for providing
  samples.

References

1. Varki, A. Glycobiology 1993, 3, 97-130.
2. Polfer, N. C. Valle, J. J. Moore, D. T.
   Oomens, J. Eyler, J. R. Bendiak, B. Anal. Chem.
   2006, 78, 670-679.
3. Marshall, A. G. Hendrickson, C. L. Jackson, G.
   S. Mass Spectrom Rev 1998, 17, 1-35.
4. Xie, Y. Lebrilla, C. B. Anal. Chem. 2003, 75,
   1590-1598.

Figure 1. Wavelength-dependent fragmentation
patterns for A) nigerose B) laminaribiose C)
maltose D) cellobiose E) isomaltose and
F) gentiobiose over the wavelength
range from 9.2 to 9.7 µm.