Structure determination of large isomeric ... - ACS Publications

Subscriber access provided by TRINITY COLL

Article

Structure determination of large isomeric oligosaccharides of natural origin through multi-pass and multi-stage cyclic traveling wave ion mobility mass spectrometry. David Ropartz, Mathieu Fanuel, Jakub Ujma, Martin Palmer, Kevin Giles, and Hélène Rogniaux Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b03036 • Publication Date (Web): 26 Aug 2019 Downloaded from pubs.acs.org on August 27, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Structure determination of large isomeric oligosaccharides of natural origin through multi-pass and multi-stage cyclic traveling wave ion mobility mass spectrometry. David Ropartz ◊, Mathieu Fanuel ◊, Jakub Ujma ‡, Martin Palmer, ‡ Kevin Giles ‡, Hélène Rogniaux ◊* ◊ INRA, UR1268 Biopolymers Interactions Assemblies, La Géraudière B.P. 71627, F-44316 Nantes, France. ‡ Waters Corporation, Stamford Avenue, Altrincham Road, Wilmslow, SK9 4AX, UK. * Corresponding Author: [email protected] ABSTRACT: Carbohydrate isomers with identical atomic composition cannot be distinguished by mass spectrometry. By separating the ions according to their conformation in the gas phase, ion mobility (IM) coupled to mass spectrometry is an attractive approach to overcome this issue and extend the limits of mass spectrometry in structural glycosciences. Recent technological developments have significantly increased the resolving power of ion mobility separators. One such instrument features a cyclic traveling-wave IM separator integrated in a quadrupole/time-of-flight mass spectrometer. This system allows for multi-pass ion separations and for pre-, intra- and post-IM fragmentation. In the present study we utilise this system to explore a complex mixture of oligoporphyrans derived from the enzymatic digestion of the cell wall of the red alga P. umbilicalis. We are able to deduce their complete structure using IM arrival times and the m/z of specific fragments. This approach was successfully applied for sequencing of oligoporphyrans of up to 1500 Da and included the positioning of the methyl ether and sulphate groups. The structures defined in this study by IM-MS/MS agree with those found in the past; but using much more time consuming analytical approaches. This study also revealed some so far undescribed structures, present at very low abundance. In addition, the results made it possible to compare the abundance of the different isomers released by the enzyme and to draw further conclusions on the specificity of β-porphyranase and more particularly on its accommodation tolerance of anhydro-bridges in sub-sites. Finally, a separation of two isomers with very similar mobility was obtained after 58 passes around the cIM, with an estimated resolving power of 920 for these triply charged species, confirming the structures attributed to these two isomers.

Natural polysaccharides, as the main components of plant cell walls, represent major marine and terrestrial bioresources. Their structural diversity presents a significant challenge for analytical sciences and many of these structures remain unknown. Incomplete knowledge of polysaccharide structures obviously limits the understanding of their biological role as well as the full exploitation of their functional properties. In this context, the use of specific degradation enzymes – when available – provides a first level of valuable structural information. This approach is particularly effective if the degradation is controlled to maintain intermediate degradation products, i.e. larger oligosaccharides structures, from which the structure of the native polysaccharide can be deduced. In addition, intermediate products are more likely to reveal the presence of minor motifs with potentially interesting functional properties, which are otherwise masked by the predominance of final products of degradation. To infer the structure of the native polysaccharide, it is then necessary to resolve the glycosidic sequence of the degradation products (released as a complex mixture)

which is complicated due to presence of configurational isomers and diastereoisomers. The former arises from connectivity and/or the positioning of modifications and lateral branching. The latter is due to the differences in configuration of the hydroxyl function of asymmetric carbons (epimers), including the specific cases of anomers (α/β). Mass spectrometry, owing to its outstanding sensitivity, high information content, and relative tolerance to sample complexity, is a leading method for deciphering polysaccharides structures. In the last decade specifically, the combination of ion mobility and mass spectrometry (IM-MS) has proved to be an attractive and effective approach in structural applications (for a recent review, see Gabelica et al1). Ions are separated according to their mobility in a buffer gas, before the m/z measurement. IM-MS has been shown to resolve different classes of molecules in complex samples2-4 and can also be used to obtain – directly (linear field IM) or after calibration (nonlinear field IM) - the rotationally averaged collision cross sections (CCS).1,5 CCS is an inherent property of analyte ion and buffer gas molecules and it can be calculated from

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

theoretical molecular structure and thus, can help to identify molecules. Improvements in the IM resolving power (R) have been achieved through extension of separation path length. Clemmer and co-workers introduced a cyclotron geometry drift tube,6 which was able to store ions for approximatively 100 cycles (path length of ~180 m) and reached a R of about 1000.7 In 2004 Giles and co-workers introduced the concept of non-linear field IM separation based on travelling waves (TWIM).8 This concept was extended further to construct a cyclic travelling wave enabled IM (cIM) separator allowing multipass separations.9 A second-generation instrument was used to demonstrate R of ~750 after 100 passes around the cIM and allowed multistage IMS (IMSn) experiments on precursor and/or product ions.10,11 This instrument was reported recently as particularly attractive for the structural characterization of carbohydrates, including determination of isomers, thanks to the pre and post-IM fragmentation as well as the IMSn capabilities.12 TWIM separators have also been constructed using structures for lossless ion manipulation (SLIM), developed by the Smith group.13 Their multi-pass setup allowed ultra-long ion path lengths to be accessed (~1 km). The authors reported R of ~1860 after 40 passes (540m) and achieved ions trajectory of up to 1-km over 81 consecutive passes.14

The natural polysaccharides investigated in this study are porphyrans extracted from the cell walls of the red

Until now, application of ion mobility to carbohydrates has often been limited to pure or synthesized standards and/or to structures with reduced degrees of polymerization (DPs, usually lower than 6).15,16 Difficulty of separation of larger oligo/polysaccharides arises from backbone folding in the gas-phase, thus structures with minor structural differences (such as methylation) may have common backbone folds, leading to very similar CCSs. As recently reported, this limitation can be overcome through fragmentation and IM separation of the fragments, where CCS difference is more pronounced.16,17 Deduced CCSs can then be compared to reference values, if available, or to the CCS estimated from a theoretical model structure. Reference CCSs are typically determined from linear field IM separators (drift tubes, DTIM) using well-known structures and/or their synthetic equivalents.

Figure 1. Structure of the dimers in porphyrans of the cell walls of the red algae P. umbilicalis. Structural constraints due to the activity of the β-porphyranase A are indicated, according to the conclusions of a previous study,22 based on a combination of IP-RP-UPLC, 18O labelling of the reducing end and XUVPD MS/MS.

The resolving power of DTIM is rarely >100 while the experimental uncertainty in DTIM-derived CCSs is typically ~0.5-2%.3 Notably, Stow et al. performed the most precise DTIM measurements reported to date (relative standard uncertainty in CCS of 0.27%).18 CCS can also be determined with TWIM separators after calibration, but uncertainties associated with the available reference values limits the applicability of the approach in high-resolution IM separations In addition, there are very few standards that can be easily synthesized or purified that reflect the structural complexity and diversity of natural polysaccharides. Therefore, no or very little CCS reference values are available in that area. Lastly, the degrees of freedom are high for larger oligosaccharides structures (DPs above 4 or 5) which makes it difficult to evaluate the theoretical CCSs.

For the purpose of this work, porphyrans were enzymatically degraded with β-porphyranase A into a mixture of oligoporphyrans.21 The structures produced in the digestion mixture have been explored in depth by our group in the past.22-25 Several SEC fractions were analysed using a fairly complex analytical setup that combined ionpairing reversed-phase UPLC (IP-RP-UPLC), 18O-labelling of the reducing end and high-energy activation tandem MS. High-energy activation methods include activation by UV photons of extreme energy, leading to a dissociative photo-ionization,22,23 or activation by high-energy helium cations, producing charge transfer dissociation.24,25 Indeed, we have shown that, unlike the generic fragmentation method in tandem MS (low-energy CID, LE-CID) which provides limited structural information, these two approaches are able to resolve the complete structures of natural oligosaccharides, including isomeric forms, through the significant production of specific cross-ring

algae Porphyra umbilicalis. Porphyrans form an interesting family of sulphated polysaccharides with a wide range of bioactivities.19,20 Schematically, porphyrans can be described as the concatenation of two different disaccharide subunits (Figure 1). The first moiety, named Por for porphyran, is composed of (4-linked α-L-Galp-6sulfate (1- 3) β-D-Galp abbreviated as L6S-G). The second moiety is annotated Aga, for agarose, and contains one anhydro bridge (4-linked 3,6-anhydro-α-L-Galp(1-3) β-DGalp abbreviated as LA-G).


Page 2 of 9

Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry fragments (A and X types). The results revealed that porphyrans of P. umbilicalis were methylated on the C6 position of the β-D-Galp of the Por subunits. We also established unequivocally that methylation occurred on the C6 position of Aga moieties. We showed that enzymatic treatment produced oligosaccharides with sulphated moieties (i.e. Por units) on each extremity. We proved that β-porphyranase A could cope with a methylation of the Por moiety at the non-reducing end, but not at the reducing end. Finally, we showed that oligoporphyrans with several consecutive Aga moieties could be released. This structural knowledge is summarized in Figure 1 and will form the basis of the present study. In this work, the complete SEC fractions of oligoporphyrans have been subjected to ion mobility separation on the cIM mass spectrometer. We show that although no CCS reference values are available, the structure of large isomeric carbohydrates produced from natural porphyrans could be completely resolved thanks to the high IM resolving power, multi-pass experiments and MS/MS capabilities of the cIM system. EXPERIMENTAL SECTION Oligosaccharides production and MS operation. Complex mixtures of oligosaccharides were produced by the enzymatic degradation of the cell walls of the red algae P. umbilicalis as described in Correc et al.,21 with the exception of the use of β-agarase as a pre-process. Samples were dissolved to 10 µg/ml in MeOH/H2O (1:1, v/v) and infused at a flow rate of 5 μL.min−1 into the ESI source of the MS in negative polarity mode. The ToF was operated in V-mode with a m/z range of 100-1000. IM-MS measurements on Synapt G2-Si HDMS. IM-MS experiments were performed on a Synapt G2-Si HDMS mass spectrometer (Waters Corp., Manchester, UK). The TWIM cell parameters were as follows: 180 mL.min−1 of He in the helium cell, 90 mL.min−1 of N2 in the mobility cell; the optimal value for the IM traveling-wave height was 40 V and velocity was 1000 m s−1. The IM resolving power (R; measured as CCS/ΔCCS) of the Synapt instrument has been shown to be ~32 for singly charged species.26 The experimental methodology employed in the present study is shown schematically in Figure S1. IM-MS measurements on the cIM instrument. IMSnMS/MS experiments were performed on a prototype Select Series cyclic IMS instrument (Waters Corp., Wilmslow, UK), which has a Q-cIM-Tof geometry. The instrument is described in detail by Giles et al. 11 Briefly, ions are transferred from the ion source through a quadrupole mass filter and subsequently accumulated in a trap cell, which also allows the collision induced dissociation (CID) of precursor ions. Ions enter into the high pressure mobility chamber through a helium cell (2 mm diameter entrance and exit aperture) to prevent ion scattering and fragmentation.26 Then ions enter a multi-function region where they can be subjected to multi-stage IM separation. More specifically, ions can be directed into an orthogonal cIM device (length ≈ 98 cm); and undergo one or more

passes around the cIM. Number of passes is limited by a so-called “wrap-around” effect, where the fastest ions overtake the slowest ones. In such situation, the “IM isolation” capability can be used to reduce the mobility range of species in the cIM. Essentially, the species outside the desired range are ejected out, while the ones remaining in the cIM can continue the separation process.11 When satisfactory IM separation is achieved, ions are ejected from the multi-function region. Ions then travel through a segmented quadrupole transfer cell, which also allows CID to be performed, and into the TOF MS.TW height and velocity were set to 30-35V and 375 m/s respectively; the cIM separator was operated at a N2 pressure of 2 mbar. The IM resolving power of the cIM instrument was determined at ~70 (CCS/ΔCCS) for a single pass using a pair of isomeric, reverse sequence peptides (GRGDS1+, SDGRG1+) with known CCS values (208.5 Å², 205.3 Å² respectively). R for multi-pass separations increases as 70(nz)0.5, where n is the number of passes around the device and z is the charge state of an ion.11 The experimental methodologies employed in the present study are shown schematically in Figures S2-3. Instrument Control and Data Analysis. The Synapt G2 instrument was controlled using Masslynx (v4.2) and the cIM instrument using a flexible web-based GUI incorporating the IMSn functionality.11 Data from both instruments were analysed using MassLynx software (v4.2) and Driftscope (v2.9) (Waters Corp., Wilmslow, UK). RESULTS AND DISCUSSION Ion mobility of the intact DP8 oligoporphyran. Figure 2 compares the arrival time distributions (ATDs) of the DP8 oligoporphyran species (3.Por, 1.Aga) with different methylation status. ATDs obtained for the [M-3H]3- ion at m/z 511.21, 515.89 and 520.57 are shown respectively in black, red and blue for the 0, 1 and 2 methylated species (respectively noted as DP8.0Me, DP8.1Me and DP8.2Me). The top panel (Figure 2A) corresponds to the ATD measured on the Synapt G2-Si HDMS instrument. Panels B and C show the ATDs obtained after one pass (98 cm) and four passes (392 cm) respectively on the cIM instrument. Both experiments are illustrated in Figures S1 and S2. The resolving power of the Synapt G2-Si HDMS (R ~ 55 for z=3 ions) was clearly not sufficient for the separation of the DP8 oligoporphyrans (Figure 2A). However, for the singly methylated DP8 species (red trace), two components are partially separated (arrival times of 8.3 ms and 8.6 ms). The asymmetric ATD profile obtained for the doubly methylated DP8 species (blue trace) suggests that different components are present and could be separated with an enhanced R. The corresponding ATDs obtained after a single pass around the cIM (Figure 2B) show an improved separation for all DP8 forms, with an expected R of ~120. Two components were clearly separated for the singly methylated DP8 (red trace, 22.1 and 22.6 ms) and the doubly methylated DP8 (blue trace, arrival times of 22.6 and 23.2 ms). After four passes around the cIM (Figure 2C, R ~ 240), two components were resolved for the non-



methylated species (black trace, 50.8 and 52.5 ms). Considering the known specificity of the β-porphyranase A, which leaves two Por moieties on both ends of the oligoporphyran, the two DP8.0Me isoforms could consist of Por-Por-Aga-Por and Por-Aga-Por-Por isomers. For the DP8.1Me oligoporphyran, four conformations were separated (red trace, 51.7, 52.5, 54.0 and 54.5 ms) while DP8.2Me exhibited three

Figure 2. ATDs of the DP8 oligoporphyran composed of 3.Por and 1.Aga moieties and 0, 1 and 2 methyl ether groups. Ions were isolated as [M-3H]3- in negative ionization mode. Black trace: DP8.0Me (m/z 511.21). Red trace: DP8.1Me (m/z 515.89). Blue trace: DP8.2Me (m/z 520.57). ATDs recorded on a Synapt G2-Si HDMS (A), or on the cIM instrument after one pass (B) and four passes (C). Structural elucidation of the isomers through MS/MS following IM separation. The cIM instrument used in this study allows pre- and post-mobility fragmentation experiments to be performed since the mobility separator is located between two gas-filled cells where ions can be subjected to CID (so called “trap” and “transfer” cells). In the following experiments, the differentially methylated DP8 precursor ions (DP8.0Me, DP8.1Me and DP8.2Me) were m/z selected in the quadrupole. The isomers of each DP8 species were then subjected to four passes around the cIM device and subsequent CID in the transfer cell (see Supplemental figure S3).

precursor remaining in the MS/MS spectra. As in the previous experiment (Figure 2), two conformations are separated with ATD maxima at 50.8 and 52.4 ms. These two forms exhibit a 50-fold difference in intensity, which is comparable to the data presented in Figure 2C, showing that the two isomers undergo dissociation with a similar efficiency. The three upper traces in Figure 3 show the ATDs of three product ions. Importantly, product ions generated post-mobility in the transfer have ATDs that correlate with their precursors (since no further IM separation takes place). Thus, product ions originating from different precursors (i.e. black trace) are readily distinguished. The complete product ion spectra are provided in Figure S4. In this figure and throughout the manuscript, the product ions are labeled according to the nomenclature introduced by Domon and Costello.27 The structural difference between the two isomers arises from the Aga moiety position within the backbone. We identified a unique diagnostic product ion for each structure and can thus unambiguously assign the two isomers to features in the precursor ATD. The signal detected at m/z 322.01 (green trace) is unique to the PorPor-Aga-Por oligosaccharide and corresponds to the B32product ion. This ion (plotted in green) displays an ATD profile that aligns perfectly with the intense feature of the precursor ATD (black trace, 52.4 ms). This confirms that the Por-Por-Aga-Por is the most abundant structure in the DP8.0Me oligoporphyran. The species detected at m/z 493.19 can be attributed to the Y52- product ion, specific of the Por-Aga-Por-Por oligoporphyran. Upon closer inspection, this signal appears to be due to overlapping of singly and doubly charged species. The purple trace in the Figure 3 corresponds to the ATD recorded for the first isotopic peak (m/z 493.20); these species appear to be generated from both precursors, albeit with significantly different efficiency. The second isotopic peak of doubly charged species (m/z 493.70, blue trace) appears to be generated exclusively from the faster precursor (50.8 ms). Thus, we confirm that doubly charged species correspond to Y52- ions (Por-Aga-Por-Por specific) while singly charged species are the non-diagnostic 1.4A3- or 1,4X2- products of Por-Por-Aga- Por isomer. Mass spectra extracted from both ATD features

conformations (blue trace, ATDs at 53.1, 54.2 and 56.2 ms). The isomers for DP8.1Me and DP8.2Me likely arise from the position of the Aga moiety, as for the non-methylated DP8, and additionally from the localisation of the methyl ether group(s). The trend observed with more isoforms for the methylated species is thus consistent with expectations. Unmethylated oligoporphyran. Figure 3 shows the ATD of the DP8.0Me species (m/z 511.21) after four passes (black trace). The trace corresponds to the non-fragmented


Page 4 of 9

Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry Figure 3. ATDs obtained after four passes around the cIM instrument for the non-methylated DP8 oligoporphyran (DP8.0Me, composed of 3.Por and 1.Aga moieties, isolated as [M-3H]3- species in negative ionization mode) and dissociation in the transfer cell. Black trace: non-dissociated precursor ion of the DP8.0Me (m/z 511.21) remaining in the

MS/MS spectra. Green trace: B32- fragment ion of the Por-PorAga-Por isomer (m/z 322.01). Purple trace: Y52- fragment ion of the Por-Aga-Por-Por isomer (m/z 493.20), and 1.4A3- or 1,4X2fragments of the Por-Por-Aga-Por isomer. Blue trace: second isotope of the Y52- fragment ion of the Por-Aga-Por-Por isomer (m/z 493.70). TW: 30 V, 375 m/s.

Figure 4. ATDs of monomethylated oligoporphyran species (DP8.1Me, 3.Por, 1.Aga moieties and one methyl ether group, isolated

as [M-3H]3- species in negative ionization mode) obtained after four passes around the cIM device and their CID products ions

generated in the transfer cell. Top trace (black): ATD non-dissociated precursor isomers (m/z 515.89) remaining in the MS/MS spectra. Traces below (orange, purple and green): ATDs of the B5, B3, Y5 and Y3 product ions. Masses of expected B5, B3, Y5 and Y3 product ions are given in Figure S5. Right panel: schematic representation of the identified precursor isomer structures (A to D). Please refer to Figure 1 for full structural details. TW: 30 V, 375 m/s.

(inset) show the isotopic profiles of singly and doubly charged species, further confirming the above assignment. The cIM-MS/MS functionality of this instrument enabled complete structural elucidation of the possible isomers of the DP8.0Me oligoporphyran. Importantly, the results revealed for the first time that the two forms are not present in the equal abundance, the Por-Por-Aga-Por being predominantly produced by the β-porphyranase A. This suggests that the enzyme has better tolerance for anhydro bridges at sub-binding site -4 than at sub-binding site +3 (sites numbered according to the cleavage point).22 The structures of these isomers were characterized in a previous study by our group, using a combination of highenergy photon activation tandem MS (XUV-DPI) and 18O labelling.22 However, they were not separated by the 23 min IP-RP-UPLC gradient and no information was retrieved on their relative abundance. In the present work and in spite of the relatively limited information obtained with low energy CID, high resolution cIM-MS/MS experiments revealed more comprehensive information on the specificity constraints of the enzyme within a 30 s experiment

Methylated oligoporphyrans. To further the structural elucidation of the oligoporphyrans products, the methylated forms (DP8.1Me and DP8.2Me) were investigated following a similar approach as above. However, the DP8.1Me and DP8.2Me species have a low intensity in the MS relative to the DP8.0Me (17% for the DP8.1Me and 2.5% for the DP8.2Me). Thus, the signal intensity of product ions was sometimes very low. A careful examination of the isotopic patterns was performed to validate “real” product ions from the background. The ATD of the background signal with m/z similar to that of a “real” product ion was plotted as a dashed line in Figures 4 and 5. Figure 4 shows the ATD of the remaining precursor ion at m/z 515.89, corresponding to DP8.1Me. After four passes around the cIM, four features are separated; annotated A to D in the upper trace of Figure 4. The bottom traces correspond to the ATDs of B5, B3, Y5 and Y3 product ions, respectively. As shown in Figure S5, isomers with varying methylation site will produce unique combinations of these four product ions; here we use the high resolution IM separation and post-IM CID, in order to distinguish the relevant products and assign methylation site in the precursor DP8.1Me species.



As an example, the precursor feature A (Figure 4, black trace, 51.4 ms) generates a B52- product ion at m/z 482.19, a B32- fragment at m/z 329.05, a Y5- fragment at m/z 903.70 and a Y3- fragment at m/z 583.30. Considering that the enzyme leaves two Por moieties at both ends of the oligosaccharide and that the Por moiety at the nonreducing end cannot be methylated, this combination is diagnostic for the Por-Por-Aga-Por isomer, methylated at

the second Por moiety (Figure 4, right panel, structure A). Another example is the precursor feature C with arrival time measured at 53.8 ms (Figure 4, black trace). This precursor feature generates B3 and Y5 product ions with the same m/z as those produced from feature A. Yet, products B52- and Y3- are measured with m/z of 475.17 and 597.33, respectively,

Figure 5. ATDs obtained after four passes around the cIM instrument for the DP8.2Me oligoporphyran species (3.Por, 1.Aga moieties and two methyl ether groups, isolated as [M-3H]3- species in negative ionization mode) subjected to CID in the transfer cell. Top trace (black): non-dissociated precursor ion of the DP8.2Me (m/z 520.57) remaining in the MS/MS spectra. Traces below: ATDs of the B5, B3, Y5 and Y3 products. Masses of expected B5, B3, Y5 and Y3 products are provided in Figure S6. Right panel: schematic representation of the expected precursor isomer structures (A, B1, B2, C1 and C2). Please refer to Figure 1 for full structural details. TW: 30 V, 375 m/s.

which leads unambiguously to the isomer with the same backbone as isomer A, but methylated at Aga moiety (Figure 4, right panel, structure C). Note that the singly charged ions at m/z 547.10 or m/z 561.11 were not considered in the above deduction because these species can correspond to B3- product ions, as well as to a sulfate loss from the B32- ions at m/z 322.01 and 329.02. Altogether, the cIM-MS/MS results presented in Figure 4 led to the identification of four isomers of DP8.1Me (Figure 4, right panel, structures A-D). In our previous work, we employed the extreme UV photo-dissociation (XUVPD) tandem MS for structural elucidation of these isomers.22 The above four structures were identified in two, closely eluting IPRP-UPLC chromatographic peaks. A fifth structure of DP8.1Me was also determined but was not detected in the present study. Yet, the latter structure is questionable as it was detected only in positive ion mode. In conclusion, a broad picture of the structural isomers of DP8.1Me was

revealed from the by combination of high-resolution IM separation and MS/MS. Figure 5 shows the ATDs of the doubly methylated oligoporphyran species (DP8.2Me) separated after four passes around the cIM and subjected to CID in the transfer cell. The top trace corresponds to the non-dissociated precursor isomers; the traces below correspond to the B5, B3, Y5 and Y3 products. Figure S6 summarizes the expected product ion masses of the possible isomers, considering the known specificity of the enzyme (see discussion regarding DP8.1Me above). Three features are separated with arrival times of 52.7, 53.9 and 56.0 ms (labelled as A, B and C). The CID of precursor feature A reveals a B52product at m/z 482.19, a Y52- product at m/z 507.22 and a Y3- product at m/z 597.33. This combination of product ions is diagnostic for the Por-(6-O-Me)Aga-(6-O-Me)Por-Por structure (Figure S6). The assignment is more complicated for the remaining four possible isomers, which co-elute in ATD features B and C. For feature B, the combination of


Page 6 of 9

Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry arrival time information and m/z of the observed fragments indicated the unambiguous presence of the B1 and B2 isomers (Figure 5, right panel). These two structures share a common backbone (Por-(6-O-Me)PorAga-Por), with one methyl group positioned on the same Por moiety, but they differ in the positioning of the second methyl group. This leads to clearly discriminant masses for their B5, B3, Y5 and Y3 products. This leaves us with another two possible isomers, co-eluting as feature C (56.0 ms). Investigation of product ions generated from feature C reveals only one Y3- product (m/z 597.33). This, combined with the unique B52- product observed (m/z 482.19), indicates that the isomers have a common (6-O-Me)GL6S-G motif on their reducing end. Furthermore, feature C simultaneously generates singly and doubly charged Y5 ions (m/z 903.70 and m/z 500.21 respectively). This means that one of the co-eluting isomers has an Aga motif on the reducing end part (i.e. the Y5

Figure 6. ATDs obtained after 58 passes around the cIM instrument for the co-eluting isomers C1 and C2 of the DP8.2Me oligoporphyran species (3.Por, 1.Aga moieties and two methyl ether groups, isolated as [M-3H]3- species in negative ionization mode, see Figure 5 for details) and their CID products generated in the transfer cell. Top trace: nondissociated precursor ion of the DP8.2Me (m/z 520.57) remaining in the MS/MS spectra. Traces below: ATDs for the B5, B3, Y5 and Y3 products generated from co-eluting isomers C1 and C2. TW: 30 V, 375 m/s.

corresponds to G-LA-(6-O-Me)G-L6S-G), while the other has it on the non-reducing end (i.e. the Y5 is G-L6S-(6-OMe)G-L6S-G). Based on these product ions, we postulate structures C1 and C2 for the isomers co-eluting as feature C (Figure 5, right panel). In the previous work by XUVPD tandem MS,22 four structures corresponding to DP8.2Me were characterized. Three of them correspond to isomers

A, B1 and C2, while B2 and C1 were not previously observed. The structure that we did not observe in the present study corresponds to (L6S-(6-O-Me)G-LA-(6-O-Me)G-L6S-GL6S-G), assigned in our previous study22 using a single diagnostic product ion of low abundance. Using the present methodology, the presence of this isomer could be alluded to by detection of a Y52- product ion at m/z 500.21 and a B52- at m/z 489.2, generated from the same ATD feature. The Y52- at m/z 500.21 is clearly generated from feature C (56 ms) but this may arise from the C2 species (Figure 5); a B52- at m/z 489.20 with an ATD of 56.0 ms is detected too, although at low abundance. However, this latter species was not considered (i.e. it is plotted as a dashed line) because a clear isotopic pattern could not be distinguished. In order to verify presence of isomers C1 and C2 we further exploited the multi-function capabilities of the cIM instrument. We isolated the feature C inside the cIM separator while features A and B were removed after four passes (for detailed descriptions of the advanced functionality of the cIM platform please refer to Giles et al11). Feature C was then subjected to 58 passes around the cIM, which was followed by CID in the transfer cell. The estimated resolving power after 58 passes is ~920 (Methods) for the triply charged species. The ATD of feature C after 58 passes appear somewhat asymmetric, although no clear structure is seen. The ATDs corresponding to five product ions originating from feature C are plotted in Figure 6. The ions detected at m/z 482.19, m/z 597.33 and m/z 500.21 correspond to B52-, Y3- and Y52product ions, respectively. Their ATDs resemble that of the precursors (top trace), with a hint of shift to the shorter arrival time (relative to the precursor ATD, top trace). On the other hand, the ATDs of m/z 903.70 (Y5-) and m/z 329.02 (B32-) appear shifted to longer arrival time. It appears that isomer C2, which has a specific Y52- product at m/z 500.21, is faster than isomer C1, which is identified from the Y5- at m/z 903.70 and the B32- at m/z 329.02. Based on the measured difference in product ion ATD maxima, their FWHM and knowledge of the expected resolving power, we can estimate that the CCSs of the C1 and C2 precursors differ by around ~0.05%. Thus, although they were not fully resolved by the cIM device, the C1 and C2 isomers were assigned using their diagnostic product ions. CONCLUSION The structural characterization of large oligosaccharides is essential for better understanding their biological function and functional properties. In the present study, we demonstrated that determination of these molecular structures could be achieved using a recently developed QcIM-ToF instrument allowing high-resolution IM separations and MS/MS using LE-CID. Even though product ions generated through LE-CID (Y and B types), are in general insufficient to decipher the structure of coexisting isomers, the LE-CID information combined with high resolution IM separation of precursors proved to be a powerful approach for oligosaccharide characterization. We have demonstrated the complete structural assignment of oligosaccharides using data



obtained in a few minutes, without the need for UPLC separation or 18O labelling and using minimal sample volumes. In addition, some new structures, present in very low abundance, were revealed for the first time. Studies such as this one would typically require very high resolution IM separation which, until recently,11,28,29 was available primarily only on prototype instrumentation.7,9,16 In this study, we used only some of the features offered by the cIM instrument: the high-resolution IM separation and post-IMS MS/MS capabilities. Other features of the cIM instrument (such as “IMSn“) can provide additional level of information,11,12 although it was not used in the present work. The next step in improving the application of mass spectrometry in structural glycosciences would be to leverage the synergistic benefits of high-resolution IM separation, IMSn and structurally informative tandem MS. The ideal platform, yet to be developed, would combine the above capabilities and ion activation through highenergy photons or cations.

ASSOCIATED CONTENT Supporting Information Materials and methods, Supplemental Figures and Scheme. The Supporting Information is available free of charge on the ACS Publications website. Figures S1 to S3 are scheme of the experimental processes used for the different IM acquisitions. Figure S4 is the IM-MS/MS spectra of the DP8.0Me species. Figures S5 and S6 summarize the putative product ions of interest of the different structures discussed.

AUTHOR INFORMATION Corresponding Author [email protected], +33 (0) 240 67 50 34 INRA UR1268 BIA Rue de la Géraudière. B.P. 71627, F-44316 Nantes cedex 3. France

Author Contributions The manuscript was written through contributions of all authors. / All authors have given approval to the final version of the manuscript.

ACKNOWLEDGMENT The authors thank Dr. M. Czjzek and Dr. C. Hervé (CNRSUPMC UMR 8227, Station Biologique, Roscoff, France) for providing the oligosaccharides.

REFERENCES 1. Gabelica, V. r.; Shvartsburg, A. A.; Afonso, C.; Barran, P.; Benesch, J. L. P.; Bleiholder, C.; Bowers, M. T.; Bilbao, A.; Bush, M. F.; Campbell, J. L.; Campuzano, I. D. G.; Causon, T.; Clowers, B. H.; Creaser, C. S.; De Pauw, E.; Far, J.; Fernandez-Lima, F.; Fjeldsted, J. C.; Giles, K.; Groessl, M.; Hogan Jr, C. J.; Hann, S.; Kim, H. I.; Kurulugama, R. T.; May, J. C.; McLean, J. A.; Pagel, K.; Richardson, K.; Ridgeway, M. E.; Rosu, F. d. r.; Sobott, F.; Thalassinos, K.; Valentine, S. J.; Wyttenbach, T. Mass Spectrometry Reviews 2019, 38, 291-320. DOI: 10.1002/mas.21585

2. Lanucara, F.; Holman, S. W.; Gray, C. J.; Eyers, C. E. Nat. Chem. 2014, 6, 281-294. DOI: 10.1038/nchem.1889 3. May, J. C.; Goodwin, C. R.; Lareau, N. M.; Leaptrot, K. L.; Morris, C. B.; Kurulugama, R. T.; Mordehai, A.; Klein, C.; Barry, W.; Darland, E.; Overney, G.; Imatani, K.; Stafford, G. C.; Fjeldsted, J. C.; McLean, J. A. Anal. Chem. 2014, 86, 2107-2116. DOI: 10.1021/ac4038448 4. Fanuel, M.; Garajova, S.; Ropartz, D.; McGregor, N.; Brumer, H.; Rogniaux, H.; Berrin, J. G. Biotech. for Biofuels 2017, 10, 10.DOI: 10.1186/s13068-017-0749-55. May, J. C.; McLean, J. A. Anal. Chem. 2015, 87, 1422-1436. DOI: 10.1021/ac504720m 6. Merenbloom, S. I.; Glaskin, R. S.; Henson, Z. B.; Clemmer, D. E. Anal. Chem. 2009, 81, 1482-1487. DOI: 10.1021/ac801880a 7. Glaskin, R. S.; Ewing, M. A.; Clemmer, D. E. Anal. Chem. 2013, 85, 7003-7008. DOI: 10.1021/ac4015066 8. Giles, K.; Pringle, S. D.; Worthington, K. R.; Little, D.; Wildgoose, J. L.; Bateman, R. H. Rapid Commun. Mass Spectrom. 2004, 18, 2401-2414. DOI: 10.1002/rcm.1641 9. Giles, K.; Wildgoose, J.; Pringle, S.; Garside, J.; Carney, P.; Nixon, P.; Langridge, D. Design and Utility of a Multi-Pass Cyclic Ion Mobility Separator. In 62nd ASMS Conference on Mass Spectrometry and Allied Topics, Baltimore, MD, June 15−19, 2014. 10. Giles, K.; Ujma, J.; Green, M.; Richardson, K.; Langridge, D.; Tomczyk, N. Design and Performance of a Second-Generation Cyclic Ion Mobility Enabled Q-ToF. In 65th ASMS Conference on Mass Spectrometry and Allied Topics, Indianapolis, IN, June 4−8, 2017. 11. Giles, K.; Ujma, J.; Wildgoose, J.; Pringle, S.; Richardson, K.; Langridge, D.; Green, M. Anal. Chem. 2019, 91, 8564-8573. DOI: 10.1021/acs.analchem.9b01838 12. Ujma, J.; Ropartz, D.; Giles, K.; Richardson, K.; Langridge, D.; Wildgoose, J.; Green, M.; Pringle, S. J. Am. Soc. Mass Spectrom. 2019, 30, 1028-1037. DOI : 10.1007/s13361-019-02168-9. 13. Hamid, A. M.; Ibrahim, Y. M.; Garimella, S. V. B.; Webb, I. K.; Deng, L. L.; Chen, T. C.; Anderson, G. A.; Prost, S. A.; Norheim, R. V.; Tolmachev, A. V.; Smith, R. D. Anal. Chem. 2015, 87, 1130111308. DOI: 10.1021/acs.analchem.5b02481 14. Deng, L. L.; Webb, I. K.; Garimella, S. V. B.; Hamid, A. M.; Zheng, X. Y.; Norheim, R. V.; Prost, S. A.; Anderson, G. A.; Sandoval, J. A.; Baker, E. S.; Ibrahim, Y. M.; Smith, R. D. Anal. Chem. 2017, 89, 4628-4634. DOI: 10.1021/acs.analchem.7b00185 15. Hofmann, J.; Hahm, H. S.; Seeberger, P. H.; Pagel, K. Nature 2015, 526, 241-+. DOI: 10.1038/nature15388 16. Hofmann, J.; Pagel, K. Angew. Chem.-Int. Edit. 2017, 56, 83428349. DOI: 10.1002/anie.201701309 17. Harvey, D. J.; Watanabe, Y.; Allen, J. D.; Rudd, P.; Pagel, K.; Crispin, M.; Struwe, W. B. J. Am. Soc. Mass Spectrom. 2018, 29, 1250-1261. DOI: 10.1007/s13361-018-1930-1 18. Stow, S. M.; Causon, T. J.; Zheng, X. Y.; Kurulugama, R. T.; Mairinger, T.; May, J. C.; Rennie, E. E.; Baker, E. S.; Smith, R. D.; McLean, J. A.; Hann, S.; Fjeldsted, J. C. Anal. Chem. 2017, 89, 90489055. DOI: 10.1021/acs.analchem.7b01729 19. Venkatraman, K. L.; Mehta, A. Plant Food Hum. Nutr. 2019, 74, 10-17. DOI: 10.1007/s11130-018-0707-9 20. Fernando, I. P. S.; Kim, K. N.; Kim, D.; Jeon, Y. J. Crit. Rev. Biotechnol. 2019, 39, 99-113. DOI: 10.1080/07388551.2018.1503995 21. Correc, G.; Hehemann, J. H.; Czjzek, M.; Helbert, W. Carbohydr. Polym. 2011, 83, 277-283. DOI: 10.1016/j.carbpol.2010.07.060 22. Ropartz, D.; Giuliani, A.; Hervé, C.; Geairon, A.; Jam, M.; Czjzek, M.; Rogniaux, H. Anal. Chem. 2015, 87, 1042-1049. DOI: 10.1021/ac5036007 23. Ropartz, D.; Giuliani, A.; Fanuel, M.; Herve, C.; Czjzek, M.; Rogniaux, H. Analytica Chimica Acta 2016, 933, 1-9. DOI: 10.1016/j.aca.2016.05.036 24. Ropartz, D.; Li, P.; Fanuel, M.; Giuliani, A.; Rogniaux, H.;


Page 8 of 9

Page 9 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry Jackson, G. P. J. Am. Soc. Mass Spectrom. 2016, 27, 1614-1619. DOI: 10.1007/s13361-016-1453-6 25. Ropartz, D.; Li, P.; Jackson, G. P.; Rogniaux, H. Anal. Chem. 2017, 89, 3824-3828. DOI: 10.1021/acs.analchem.7b0047326. Giles, K.; Williams, J. P.; Campuzano, I. Rapid Commun. Mass Spectrom. 2011, 25, 1559-1566. DOI: 10.1002/rcm.5013 27. Domon, B.; Costello, C. E. Glycoconjugate J. 1988, 5, 397-409. DOI: 10.1007/BF01049915

28. Ridgeway, M. E.; Lubeck, M.; Jordens, J.; Mann, M.; Park, M. A. Int. J. Mass Spectrom. 2018, 425, 22-35. DOI: 10.1016/j.ijms.2018.01.006 29. Fouque, K. J. D.; Ramirez, C. E.; Lewis, R. L.; Koelmel, J. P.; Garrett, T. J.; Yost, R. A.; Fernandez-Lima, F. Anal. Chem. 2019, 91, 5021-5027. DOI: 10.1021/acs.analchem.8b04979


Structure determination of large isomeric ... - ACS Publications

Recommend Documents