Ion Mobility Mass Spectrometry of Complex Carbohydrates Collision Cross Sections of Sodiated N- linked Glycans. - Supporting Information -

Transcription

1 Ion Mobility Mass Spectrometry of Complex Carbohydrates Collision Cross Sections of Sodiated N- linked Glycans Kevin Pagel, 1 *, David J. Harvey 2, 3 - Supporting Information - 1 Fritz Haber Institute of the Max Planck Society, Department of Molecular Physics, Faradayweg 4-6, Berlin, Germany, 2 Department of Biological Sciences, University of Warwick, Coventry, CV4 7AL, UK, 3 University of Oxford, Department of Biochemistry, South Parks Road, Oxford OX1 3QU UK S1

2 Experimental Details Glycan Release N-linked glycans were released with hydrazine 1,2 from the well-characterized glycoproteins ribonuclease B 3, porcine thyroglobulin, 4,5 chicken ovalbumin, 6,7 and bovine fetuin 8 obtained from Sigma Chemical Co. Ltd. (Poole, Dorset, UK) and re-acetylated. Alternatively to hydrazinolysis, which produces relatively large amounts of sample, glycans can also be released on a smaller scale with the enzyme protein N-glycosidase F (PNGase F) using well-established methods. 9,10 After cleavage, the samples were stored at -20 C until required. Sialic acids were removed from the thyroglobulin and fetuin samples by heating with 1% acetic acid for 1 hour at 70 C. For electrospray analysis, samples were dissolved in water:methanol (1:1, v:v) at about 1 mg/ml. CCS Measurements Measurement of absolute collision cross sections (CCS) were performed on a modified Synapt HDMS 11 (Waters, Manchester, UK) quadrupole-im-ms instrument with a linear (not travelling wave) drift tube which was described in detail previously. 12,13 Briefly, in this instrument, glycan ions and their fragments are generated in a nano electrospray (nano-esi) ion source using capillaries prepared in-house using an established protocol. 14 In order to ensure sufficient fragmentation, ions where subjected to in-source collisions using high cone and skimmer voltages. Subsequently, the ion beam was focussed via a stacked-ring ion guide and transferred into a quadrupole region where ions with m/z below 32,000 could be pre-selected. Afterwards, the ions enter an array of three stacked-ring assemblies of which the second one was modified to serve as linear, RF-confined drift tube with a length of 18cm. IMS separation was followed by mass analysis using an orthogonal ToF mass analyzer. Typical settings for the analyzed glycans were: capillary voltage, kv; sample cone 150 V, extractor cone 10 V; cone gas, 40 L/h; trap collision voltage, 10 V; trap DC bias, 25 V; IMS drift voltage, V; ion transfer stage pressure, 2-5 mbar; trap pressure, mbar (He), mbar (N 2 ); ion mobility gas, He / N 2 ; ion mobility cell pressure, 3.2 mbar (He), 1.7 mbar (N 2 ); time-of-flight analyzer pressure, mbar. All mass spectra were calibrated externally using an aqueous solution of caesium iodide (100 mg/ml) and were processed with the MassLynx software (Version 4.1, Waters, Manchester, UK). Spectra are shown with minimal S2

3 smoothing and without background subtraction. CCS values were determined from the slopes of drift time versus reciprocal drift voltage plots as described previously. 12,15 CCSs were calculated as reported previously. At constant temperature and pressure, the velocity of the ions in the IMS cell (v) is directly proportional to their mobility (K) and the applied electric field (E) (see eq. 1). v = KE (1) The drift time t D that is needed to traverse a cell of length L is proportional to the inverse mobility (1/K) as well as the inverse field (1/E) (see eq. 2). Consequently, the mobility of a given ion is typically determined by plotting the drift time t D vs. the inverse drift voltage and subsequent linear regression. The intercept of the fit t 0 corresponds to the time required to transport the ions from the end of the drift region into the mass analyzer. t D = L KE + t 0 (2) In practice, each sample was measured at eight different drift voltages ranging from 50 to 150 V. Drift times where extracted from the raw data by fitting a Gaussian distribution to the ATD of a given ion. A representative plot of the reciprocal drift voltages versus the centroid drift time in helium is shown for the singly sodiated ribonuclease B carbohydrates in Figure S1. The correlation coefficients of the fit were typically above 0,9996. From the obtained mobilities, absolute CCSs where calculated using the Mason-Schamp equation: 16 CCS = 3e 16N 2π µk B T 1 K (3) where N is the drift gas number density, µ the reduced mass of the ion and drift gas, k B the Bolzmann constant and T the temperature. The reported CCSs represent averages of three (He) or two (N 2 ) replicates acquired in independent measurements. S3

4 TW IMS Measurements and CCS Estimation High-resolution TW IMS measurements were performed on an unmodified Synapt HDMS G2 S 8 kda quadrupole-im-ms instrument (Waters, Manchester, UK). Typically 5 µl of sample was electrosprayed from platinum-palladium-coated borosilicate capillaries prepared inhouse. 17 Typical instrument settings were: capillary voltage, kv; sample cone 150 V; extractor cone 150 V; cone gas, 60 L/h; trap and transfer collision voltage, 2-5 V; wave velocity, 550, 600, 650, 700 and 750 m/s; wave height, 40 V; ion transfer stage pressure (stepwave), 8 mbar; trap pressure, mbar; ion mobility gas, N 2 ; ion mobility cell pressure, 3 mbar; time-of-flight analyzer pressure, mbar. All spectra were calibrated externally using a solution of caesium iodide (100 mg/ml) and were processed with MassLynx 4.1 software (Waters, Manchester, UK). CCS estimation was carried out as reported previously. 12,18 Briefly, the measured drift times were corrected for m/z dependent delay time in the instrument with c being an empirically determined constant (c = x EDC delay coefficient). t ' D = t D c m / z (4) Absolute glycan CCSs (CCS DT ) were corrected for charge and reduced mass. DT CCS CCS'= z (5) m Ion m Gas The slope of the plot of ln(t D ) versus ln(ccs ) yields an exponential factor A which is used to calculate final corrected drift times (t D ). Finally, estimated CCSs (CCS TW ) are determined by plotting absolute CCS (CCS DT ) as a function of t D. t D ''=(t D ') A z 1 m Ion + 1 m Gas (6) Absolute CCSs determined in He (CCS DT He ) served as calibrants to estimate He TW IMS CCSs (CCS* TW He ) from TW IMS measurements in N 2. The CCSs estimated from the calibration were surprisingly insensitive to the applied wave velocity (550, 600, 650, 700 and 750 m/s) and where, therefore, averaged. S4

5 Data Analysis During the measurements, ion mobility as well as mass spectrometric information is acquired simultaneously. The raw data containing drift time t D and m/z information were analyzed using an in-house developed Python code and an extraction tool provided by Waters. Briefly, the software extracts arrival time distributions (ATDs) for a given mass window. Often the obtained ATDs contained more than one, but well separated features, which is predominantly caused by overlapping clusters with identical m/z but different charge state. In these cases, the two most intense peaks where fitted by Gaussian distributions and the extracted drift times where used to calculate/estimate the corresponding CCSs. Figure S1. Reciprocal drift voltage versus drift time t D plots of singly sodiated ribonuclease B glycans and their CID fragments. Drift times where extracted from raw data by fitting a single Gaussian to the arrival time distribution of the particular ion. The slope of the linear fit is directly proportional to the reciprocal mobility K. Correlation coefficients where typically higher than S5

6 Figure S2. ATDs of A) Man 6 GlcNAc 2 measured on the G1 instrument and B) Man 6 GlcNAc 2 measured on a G2S instrument as native glycan (bottom), fragment of Man 7 GlcNAc 2 (middle), and fragment of Man 9 GlcNAc 2 (top). The broad and not fully resolved peak in B indicates that at least two structurally different fragments with m/z are formed via CID of larger highmannose glycans. S6

7 REFERENCES (1) Patel, T.; Bruce, J.; Merry, A.; Bigge, C.; Wormald, M.; Parekh, R.; Jaques, A. Biochemistry 1993, 32, 679. (2) Wing, D. R.; Rademacher, T. W.; Field, M. C.; Dwek, R. A.; Schmitz, B.; Thor, G.; Schachner, M. Glycoconj. J. 1992, 9, 293. (3) Fu, D.; Chen, L.; O'Neill, R. A. Carbohydr. Res. 1994, 261, 173. (4) de Waard, P.; Koorevaar, A.; Kamerling, J. P.; Vliegenthart, J. F. J. Biol. Chem. 1991, 266, (5) Kamerling, J. P.; Rijkse, I.; Maas, A. A. M.; van Kuik, J. A.; Vliegenthart, J. F. G. FEBS Lett. 1988, 241, 246. (6) Da Silva, M. L. C.; Stubbs, H. J.; Tamura, T.; Rice, K. G. Arch. Biochem. Biophys. 1995, 318, 465. (7) Harvey, D. J.; Wing, D. R.; Küster, B.; Wilson, I. B. H. J. Am. Soc. Mass Spectrom. 2000, 11, 564. (8) Green, E. D.; Adelt, G.; Baenziger, J. U.; Wilson, S.; Van Halbeek, H. J. Biol. Chem. 1988, 263, (9) Morelle, W.; Faid, V.; Chirat, F.; Michalski, J.-C. In Glycomics; Packer, N., Karlsson, N., Eds.; Humana Press: 2009; Vol. 534, p 3. (10) Morelle, W.; Michalski, J.-C. Nat. Protocols 2007, 2, (11) Pringle, S. D.; Giles, K.; Wildgoose, J. L.; Williams, J. P.; Slade, S. E.; Thalassinos, K.; Bateman, R. H.; Bowers, M. T.; Scrivens, J. H. Int. J. Mass Spectrom. 2007, 261, 1. (12) Bush, M. F.; Hall, Z.; Giles, K.; Hoyes, J.; Robinson, C. V.; Ruotolo, B. T. Anal. Chem. 2010, 82, (13) Pagel, K.; Natan, E.; Hall, Z.; Fersht, A. R.; Robinson, C. V. Angew. Chem. Int. Ed. 2013, 52, 361. (14) Hernandez, H.; Robinson, C. V. Nat. Protocols 2007, 2, 715. (15) Kemper, P. R.; Dupuis, N. F.; Bowers, M. T. Int. J. Mass Spectrom. 2009, 287, 46. (16) Revercomb, H. E.; Mason, E. A. Anal. Chem. 1975, 47, 970. (17) Hernández, H.; Robinson, C. V. Nat. Protocols 2007, 2, 715. (18) Ruotolo, B. T.; Benesch, J. L. P.; Sandercock, A. M.; Hyung, S.-J.; Robinson, C. V. Nat. Protocols 2008, 3, S7

8 S8