Electron-Activated Tandem Mass Spectrometry Analysis of Glycosaminoglycans

Abstract

Glycosaminoglycans (GAGs) are linear polysaccharides found in a variety of organisms. GAGs contribute to biochemical pathway regulation, cell signaling, and disease progression. GAG sequence information is imperative for determining structure-function relationships. Recent advances in electron-activation techniques paired with high resolution mass spectrometry allow for full sequencing of GAG structures. Electron detachment dissociation (EDD) and negative electron transfer dissociation (NETD) are two electron-activation methods that have been utilized for GAG characterization. Both methods produce an abundance of informative glycosidic and cross-ring fragment ions without producing a high degree of sulfate decomposition. Here we provide detailed protocols for using EDD and NETD to sequence GAG chains. In addition to protocols directly involving performing these MS/MS methods, protocols include sample preparation, method development, internal calibration, and data analysis.

Basic Protocol 1: Preparation of Glycosaminoglycan Samples

Basic Protocol 2: FTICR Method Development

Basic Protocol 3: Internal Calibration with NaTFA

Basic Protocol 4: Electron Detachment Dissociation (EDD) of GAG samples

Basic Protocol 5: Negative Electron Transfer Dissociation (NETD) of GAG samples

Basic Protocol 6: Analysis of MS/MS data

INTRODUCTION

Glycosaminoglycans (GAGs) are linear carbohydrates composed of repeating disaccharide units. GAGs are present in a variety of organisms, including humans, and participate in biological processes. Though linear, GAGs are complex and heterogeneous due to non-template driven enzymatic modifications, including acetylation, sulfation, and epimerization. Analytical techniques such as NMR can determine modification location and type but require purified samples. Mass spectrometry (MS) and tandem mass spectrometry (MS/MS) can be used to characterize GAGs within a mixture and require much lower quantities of samples. Compositions of GAGs, including oligomer length and extent of modifications, can be determined using MS. However, to determine the location of modifications, MS/MS is needed. Vibrational-activation techniques including, collision-induced dissociation (CID) and infrared multiphoton dissociation (IRMPD) yield a high degree of glycosidic cleavages; however, also result in a high extent of sulfate decomposition seen as the loss of SO₃. Electron-activation techniques including electron detachment dissociation (EDD) and negative electron transfer dissociation (NETD) yield high degrees of both glycosidic and cross-ring fragment ions. Additionally, electron-activation methods maintain labile sulfate bonds. Unlike vibrational-activation methods, EDD and NETD do not require highly ionized precursor ions to yield informative fragmentation, as these are nonergodic and dissociate through low energy channels. Since electron-activation methods produce more informative fragment ions than vibrational-activation methods for GAGs, these methods are ideal for determining structural information of GAGs.

EDD irradiates multiply charged precursor anions with moderate energy electrons (~19 eV), resulting in electron detachment and radical species formation (Wolff, Amster, Chi, & Linhardt, 2007; Wolff, Laremore, Busch, Linhardt, & Amster, 2008a). The utility of EDD for sequencing GAG oligomers has been demonstrated for heparin (Hp)/ heparan sulfate (HS) and chondroitin sulfate (CS)/ dermatan sulfate (DS) (Agyekum, Patel, Zong, Boons, & Amster, 2015; Agyekum, Zong, Boons, & Amster, 2017; Huang et al., 2013; Leach et al., 2012; Leach, Wolff, Laremore, Linhardt, & Amster, 2008; Staples & Zaia, 2011; J. J. Wolff et al., 2007; Wolff, Chi, Linhardt, & Amster, 2007; Wolff et al., 2008a; Wolff, Laremore, Busch, Linhardt, & Amster, 2008b). NETD utilizes an electron acceptor, typically radical fluoranthene cation, which reacts with multiply charged precursor anions resulting in electron transfer from the precursor anion to the radical cation (Leach et al., 2011; Wolff et al., 2010). NETD produces similar fragmentation to EDD (Leach, Riley, Westphall, Coon, & Amster, 2017; Leach et al., 2011; Wolff et al., 2010). EDD is a low efficiency method requiring more signal averaging than other MS/MS methods, and utilizes a 1 sec pulse length. These combined result in long experiment times (2–5 mins depending on a number of signal averages). NETD occurs on a faster timescale because of its high efficiency and low signal requirement averaging (0.5–3 mins depending on number of signal averages and other parameters) (Leach et al., 2017). Both methods are available on Fourier transform ion cyclotron resonance (FT ICR) instruments. In these protocols, we discuss the sample preparation necessary for EDD and NETD of GAGs, as well as developing a MS method and calibration. The processes of performing EDD and NETD are also reported, as well as data workup using commercially available software.

BASIC PROTOCOL 1

PREPARATION OF GLYCOSAMINOGLYCAN SAMPLES

Introductory paragraph:

Synthetic or biological glycosaminoglycans (GAGs) are used in this protocol. The nature of these samples is dependent on sample preparation and may require either the addition of Na⁺ via the addition of NaOH, or desalting using spin filters. Small amounts of Na⁺ ions in GAG samples can increase the likelihood of production of highly ionized precursor ions. However, too many Na⁺ ions can add complexity to MS spectra and clog spray needles.

Materials:

Purified glycosaminoglycan (this can be purchased, synthesized (Arungundram et al., 2009; Karst & Linhardt, 2003), or extracted from tissue (Leach et al., 2012; Munoz et al., 2006)

HPLC water (H₂O, Sigma, 270733)

Methanol (MeOH, Fisher, UN1230)

Sodium Hydroxide (NaOH, Sigma, 221465–1KG)

Amicon Ultra 0.5mL 3K Spin filter (Millipore, UFC500396)

Centrifuge (Thermo Scientific, Sorvall Legend Micro 17 Centrifuge)

Eppendorf pipettes (Eppendorf Research, 10 μL, 100 μL, 1000 μL)

Pipet tips (Fisherbrand, SureOne™, Reload Insert Stack, Beveled Pipet Tips, 02–707-420)

5 mL Eppendorf tubes (Fisher Scientific, Eppendorf™ Eppendorf Tubes™ Conical Tubes, 14–282-300)

Vortex (Fisher Scientific, Fisher Vortex Genie 2™, 12–812)

Protocol steps

• 1Dissolve GAG sample in H₂O to create 1 mg/mL stock solution. Quantitation of small samples (0.5–5 μg) can be accomplished using dimethylmethylene blue (DMMB) assay (Coulson-Thomas & Gesteira, 2014).
• 2Vortex solution

To remove sodium salts

Note: Most GAG samples regardless of origin will contain Na⁺ that will produce a heterogenous molecular ion signal through sodium cation exchange with ionizable protons. If abundant sodium satellite peaks are preventing or hindering the analysis of the chosen precursor, salts should be removed with a spin-filter. For precursors with a high degree of sulfate modifications, sodium/proton exchange will enhance the tandem mass spectrometry analysis and introducing additional sodium ions will provide better quality data.

• 3Add 400μL of H₂O to spin filter in accompanying Eppendorf vial
• 4Centrifuge filter (16 000 x g) for 20 mins
• 5Remove filter and dispose of water flow through in Eppendorf vial
• 6Replace filter, add 5 μL of GAG stock solution
• 7Add 400 μL of H₂O to spin filter
• 8Centrifuge filter (16 000 x g) for 30 mins
• 9Remove filter and dispose of flow through
• 10Repeat steps 6–8, 3–10 times depending on extent of salt in sample
• 11Following desalting, dilute GAG stock to 0.1 mg/mL in 50:50 MeOH:H₂O
• 12Vortex solution

To enhance sodium salts

• 13Dissolve NaOH in H₂O to create 0.1 M stock solution
• 14Dilute GAG stock solution to 0.1 mg/mL in 50:50 MeOH:H₂O
• 15Add 2–10 μL of NaOH stock solution to diluted GAG solution (amount depends on total sample amount). Approximately 1mM NaOH should be added for best results.
• 16Vortex solution
• 17Figure 1 shows the benefit of removing Na⁺ to GAG solutions. This figure shows Fondaparinux sodium (Sigma Aldrich, 114870–03-0), which when diluted and analyzed from the bottle without any pretreatment appears as both highly charged and highly sodiated precursor ions. After desalting with centrifuge filters 10 times, lower charged precursor ions are favored, and sodiated precursor ion intensity has decreased.

Figure 1.

Example of the same GAG sample (Fondaparinux, GlcNS6S-GlcA-GlcNS3S6S-IdoA2S-GlcNS6S-Me) when desalted (top) and containing excess Na⁺ ions (bottom). Fully ionized precursors are obtained when Na⁺ ions are added to the sample.

BASIC PROTOCOL 2

FTICR METHOD DEVELOPMENT

Introductory Paragraph:

The development of an MS method for GAGs is discussed in this protocol. Optimization of certain parameters may be necessary based on the length and level of complexity of the GAG being analyzed. Development of this protocol assumes that the FT ICR instrument is properly shimmed, and the quadrupole is calibrated. Monthly bakeouts and glass capillary cleanings may be needed to ensure performance of FT ICR instrument.

Materials:

FTICR MS (Bruker, solariX XR ESI ETD, SN 127250000200)

Protocol steps

1. Open Bruker FTMS Control Software

2. On the Instrument Control Panel (left side of the screen):

   1. Set size to 1 M or 2 M (this can be set lower if resolution isn’t a concern)
   2. Set low m/z to 150 and high m/z to 2000
   3. Set average scans (Avg Scans) to 12
   4. Set accumulation time (Accum) to 0.1 seconds

3. Set instrument to negative mode

4. Under the “Sample Info” tab name your sample

5. Under the “API Source” tab there are four subsections (Syringe Pump, Divert Valve, API Source, Source Gas Tune)

   1. Under Syringe Pump, set the flow rate to 100 μL/h and the syringe type to “Hamilton 250 μL”
   2. Make sure Divert Valve is set to source, not waste
   3. Under API Source, set source to ESI

      1. Capillary voltage may need to be tuned. The ideal voltage should be between 3500 V and 4500 V
      2. End plate offset should be set to −500 V
   4. Source gas tune may need to be tuned, specifically the nebulizer gas

      1. Set nebulizer gas to 2.0 bar, and ensure nebulizer check box is checked
      2. Set dry gas to 4.0 L/min
      3. Set dry temp to 180° C

6. Under the “Ion Transfer” tab there are six subsections (Source Optics, Octopole, Quadrupole, Collision Cell, Transfer Optics and Gas Control)

   1. Under Source Optics, set capillary exit and deflector plate to −220 V, set funnel 1 to −150 V and set funnel RF amplitude to 150 Vpp

      1. Skimmer 1 should be set to −15 V; this can be increased or decreased for in-source collisional fragmentation

   2. Under Octopole, set frequency to 5 MHz and RF amplitude to 350 Vpp

   3. Under Quadrupole, set Q1 mass to the value in your “low m/z” parameter (150 m/z). This works as a low mass filter when not isolating

   4. Under Collision Cell, set DC voltage to −0.8 V, and set collision voltage to 0. If you notice unwanted collisions occurring this can be set to −1.0 V. If you want to see collisions, this can be set to ~ 2 V. (note, this is not the same voltage as the collision induced dissociation voltage, this is a separate, additive value that allows for low energy in-source collisions)

      1. Set RF frequency to 2 MHz and collision RF amplitude to 600 Vpp

   5. Under Transfer Optics, set time of flight to 0.500 ms (this can be increased or decreased depending on the expected m/z range of your sample. For most GAG standards 0.500 ms is deal). Set frequency to 4 MHz and RF amplitude to 350 Vpp

   6. Under Gas Control, set flow to 20 % and ensure the check box for enable is checked

7. Under “Source MS/MS” tab all check boxes under the In Source Fragmentation, Multi CASI, and ETD sub tabs should not be checked. The Quadrupole MS/MS table should also have all check boxes unchecked and the Q1 mass should be set to the “low m/z” value (150 m/z)

8. Under “Analyzer” tab, there are four subsections (Para Cell, Shimming DC Bias, Gated Injection DC Bias, Multiple Cell Accumulations, Gated Trapping)

   1. Under the Para Cell subtab, set transfer exit lens to 20 V, analyzer entrance to 13 V, side kick to 0 V, side kick offset to 1.5 V, front trap plate and back trap plate to −3.000 V, and sweep excitation to 23 %
   2. Shimming DC Bias and Gated Injection DC Bias values should not be adjusted from your current values
   3. Under the Multiple Cell Accumulations subtab, set ICR cell fills to 1
   4. The Gated Trapping subtab should be unchecked

9. The remaining tabs should be left at the default values

10. Save the method

BASIC PROTOCOL 3

INTERNAL CALIBRATION WITH NATFA

Introductory paragraph:

Sodium trifluoroacetate (NaTFA), a common calibrant, is used in the protocol. Proper calibration before performing experimental runs is imperative for ensuring mass accuracy. NaTFA can clog spray lines if not properly rinsed following use.

Materials:

HPLC water (H₂O, Sigma, 270733)

Sodium Trifluoroacetate (NaTFA, Aldrich, 132101–25G)

15 mL Falcon Tube (Fisher Scientific, Falcon™ 15 mL Conical Centrifuge Tube, 14–959-53A)

250 μL glass syringe (Hamilton 1725 Gastight Syringe, 250 μL, Luer-Cemented, 22s G, 2”, Blunt, UX-07940–47)

Vortex (Fisher Scientific, Fisher Vortex Genie 2™, 12–812)

Protocol steps

1. Dissolved NaTFA in H₂O to make 1 mg/mL stock solution

2. Dilute stock solution in H₂O to 0.01 mg/mL

3. Fill glass syringe with diluted NaTFA solution

4. Place syringe in syringe pump

5. On FTMS control, click the syringe pump checkbox in the lower right-hand corner

6. Click the tune button in the upper left-hand corner

7. Hold down the fast forward button (button with 2 forward facing arrows) until you see NaTFA signal

8. Once signal is steady and consistent, click the stop button to stop tunning, and click the acquisition button. Figure 2 shows the expected spectrum when using NaTFA in negative ion mode

9. Once your acquisition has completed, click on the “Calibration” tab

10. In the left corner, click the yellow folder button

    1. This will open the Bruker saved reference lists. When given 2 folder options (ESI and MALDI), double click the ESI folder
    2. This will open up a list of reference files. Find the file labeled “Na TFA” and double click it

11. Once you NaTFA reference file has been loaded, click Automatic

12. Any matched masses will be displayed here. To zoom in on a specific peak, choose the mass within the calibration list. Make sure the central peak is being picked and not a side peak. If the wrong peak is being picked, click on the correct peak. A window will open up confirming you want to replace the reference peak. Click OK

13. Uncheck any reference peaks that are not within 1 ppm of the expected mass

14. Once all peaks remaining are within 1 ppm, select Accept

15. Continue this until all masses within your expected mass range are within 1 ppm error

16. Note: changes to the method will negate the calibration. Only calibrate using the method you plan to use for MS/MS

Figure 2.

Expected results when examining NaTFA in negative ion mode with a time of flight of 0.500 ms.

BASIC PROTOCOL 4

ELECTRON DETACHMENT DISSOCIATION (EDD) OF GAG SAMPLES

Introductory paragraph:

Electron detachment dissociation (EDD) has been widely used for the structural characterization of GAG samples. EDD irradiates multiply charged precursor anions with electrons of moderate energy (~19 eV), inducing electron detachment, resulting in radical formation. The radical anion undergoes more extensive fragmentation than other activation methods. EDD is primarily performed on an FT ICR MS; however, the development of an electromagnetostatic electron capture dissociation (ECD) cell allows for both ECD and EDD to be implemented on a wider variety of instruments (Fort et al., 2018). For this protocol, a Bruker solariX XR FTMS instrument is utilized.

Materials:

FTICR MS (Bruker, solariX XR ESI ETD, SN 127250000200)

Nanospray source

Static Nanospray glass tips (New Objective, Econo-12N)

Microloader pipet tips (Eppendorf, 0.5–20 μL, 100 mm, light gray, 192 pcs., 930001007)

Eppendorf pipette (Eppendorf Research, 10 μL)

250 μL glass syringe (Hamilton 1725 Gastight Syringe, 250 μL, Luer-Cemented, 22s G, 2”, Blunt, UX-07940–47)

Methanol (MeOH, Fisher, UN1230)

Protocol Steps

1. This work can be done with either conventional ESI or nanoESI. If your sample quantities are low, nanoESI is recommended

2. If planning on performing EDD, turn ECD heater on before doing internal calibration. Allow 30 mins for the heater to fully heat up. ECD heater should be set to 1.5 A

3. If using conventional ESI:

   1. load the glass syringe with your diluted sample

   2. under the “API Source” tab, ensure ESI is selected

4. If using nanoESI:

   1. Set a micropipette to 10 μL and put a gel loading tip onto the pipette
   2. Fill the gel loading pipette with MeOH
   3. Insert gel loading pipette tip into nanospray tip
   4. Slowly fill nanospray tip with MeOH, then remove MeOH, repeat 3x
   5. When finished, remove all excess MeOH from nanospray tip
   6. Fill the gel loading pipette with the diluted GAG sample
   7. Insert gel loading pipette tip into nanospray tip
   8. Slowly fill nanospray tip with sample

      1. As filling nanospray tip, very slowly remove gel loading tip
      2. Be cautious to prevent bubbles from forming
   9. Once nanospray tip is full, insert electrode into tip
   10. Carefully position tip into nanospray source and be cautious to not break tip
   11. Under “API Source”, switch the source from ESI to NanoESI on-line
   12. To stop sample flow, set capillary voltage to 500 V. Tune voltage to get best spray, usually between 1000–1700 V

5. Click the tune button. Accumulation time may need to be increased (0.1– 0.4 s)

6. Once sample spray is consistent, click stop, and then acquisition to collect an MS1 spectrum

7. Once MS1 spectrum has been collected, choose a highly charged precursor for activation

8. Under the “Source MS/MS” tab, change the Q1 mass to the m/z value of the selected precursor, set the isolation window to 5 m/z and click the isolation box

9. Click the tune button

10. When isolating, the accumulation time will need to be increased (to 0.5–2 s)

11. If it appears that collisions are occurring, decrease the skimmer 1 and collisional voltages

12. Once isolation appears clean, click stop and then acquisition to collect an isolation spectrum

13. Select the “In-cell MSn” tab

    1. Under Internal MS/MS subtab, click the ECD box
    2. An ECD tab will appear
    3. Set the ECD pulse length to 1.00 sec
    4. Set ECD Bias to 19.0 V
    5. Set ECD Lens to −18 V

       1. This value can be tuned to achieve ideal fragmentation. Typical voltages range from −17.5 - −19.0 V
    6. Figure 3 displayed sample EDD data of a GAG sample

14. Click the tune button

15. Adjust ECD lens to achieve ideal fragmentation

16. Click stop

17. Change the average number of scans to 48

18. Click acquisition

19. Figure 4 shows a comparison of precursor and fragment ion intensity. EDD is a low efficiency fragmentation method, and fragment ion intensity is expected to be low compared to the precursor

20. Repeat these steps as needed to collect multiple spectra of the same precursor, or of different precursors

Figure 3.

EDD spectrum of the [M-5H+Na]⁴⁻ precursor ion of chondroitin sulfate A hexasaccharide (GlcA-GlcNAc4S-GlcA-GlcNAc4S-GlcA-GlcNAc4S) with annotations. Results were obtained with an ECD lens value of −18 V and 48 average scans.

Figure 4.

Expected EDD spectrum of the [M-5H+Na]⁴⁻ precursor ion of chondroitin sulfate A hexasaccharide (GlcA-GlcNAc4S-GlcA-GlcNAc4S-GlcA-GlcNAc4S). Fragment ion intensity compared to precursor intensity is displayed after 48 signal averages.

BASIC PROTOCOL 5

NEGATIVE ELECTRON TRANSFER DISSOCIATION (NETD) OF GAG SAMPLES

Introductory paragraph:

Negative electron transfer dissociation (NETD) is a powerful tool for sequencing GAG samples. Similar to EDD, NETD produces a radical anion which undergoes more extensive fragmentation than other activation methods, specifically threshold activation methods. Unlike EDD, NETD uses a reagent cation which strips an electron from the precursor anion. Multiple reagents have been used for NETD; however, for this protocol fluoranthene is used. NETD can be utilized on a variety of instrumentation; however, for this protocol NETD is performed on a Bruker solariX XR FTMS instrument.

Materials:

FTICR MS (Bruker, solariX XR ESI ETD, SN 127250000200)

Fluoranthene (Sigma Aldrich, 89793–5G)

Nanospray source

Static Nanospray tips (New Objective, Econo-12N)

Microloader tips (Eppendorf, 0.5–20 μL, 100 mm, light gray, 192 pcs., 930001007)

250 μL glass syringe (Hamilton 1725 Gastight Syringe, 250 μL, Luer-Cemented, 22s G, 2”, Blunt, UX-07940–47)

UHP Argon (Ar, AirGas, AR UHP300)

Methanol (MeOH, Fisher, UN1230)

Eppendorf Pipettes (Eppendorf Research, 10 μL)

Protocol steps

1. This work can be done with either conventional ESI or nanoESI. If your sample quantities are low, nanoESI is recommended. For one hour of analysis, 100 μL is needed for conventional ESI, whereas only 10 μL of sample is needed for nanoESI.

2. Use UHP argon instead of the Bruker recommended UHP methane for negative mode ETD

3. Go to “Tools”, “CI source”, the click the flush check box

   1. Allow lines to flush for at least 5 mins (note: if the flushing step is skipped, oxygen will damage the ETD filament)
   2. Once flush is complete, uncheck flush box
4. After flushing lines, turn “CI Source” on by clicking the “CI Source” check box in the Instrument Control Panel

   1. Once the CI source is on and turned green, it is ready to use
5. If using conventional ESI:

   1. load the glass syringe with your diluted sample
   2. under the “API Source” tab, ensure ESI is selected
6. If using nanoESI:

   1. Set a micropipette to 10 μL and put a gel loading tip onto the pipette
   2. Fill the gel loading pipette with MeOH
   3. Insert gel loading pipette tip into nanospray tip
   4. Slowly fill nanospray tip with MeOH, then remove MeOH, repeat 3x
   5. When finished, remove all excess MeOH from nanospray tip
   6. Fill the gel loading pipette with the diluted GAG sample
   7. Insert gel loading pipette tip into nanospray tip
   8. Slowly fill nanospray tip with sample

      1. As filling nanospray tip, very slowly remove gel loading tip
      2. Be cautious to prevent bubbles from forming
   9. Once nanospray tip is full, insert electrode into tip
   10. Carefully position tip into nanospray source and be cautious to not break tip
   11. Under “API Source”, switch the source from ESI to NanoESI on-line
   12. To stop sample flow, set capillary voltage to 500 V. Tune voltage to get best spray, usually between 1000–1700 V

7. Select the tune button. Accumulation time may need to be increased (0.1– 0.4 s)

8. Once sample spray is consistent, click stop, and then acquisition to collect an MS1 spectrum

9. Once MS1 spectrum has been collected, choose a highly charged precursor for activation

10. Under the “Source MS/MS” tab, change the Q1 mass to the m/z value of the selected precursor, set the isolation window to 5 m/z and click the isolation box

11. Click the tune button

12. When isolating, the accumulation time will need to be increased (to 0.5–2 s)

13. If it appears that collisions are occurring, decrease the skimmer 1 and collisional voltages

14. Once isolation appears clean, click stop and then acquisitions to collect an isolation spectrum

15. Under “Source MS/MS” tab, click the ETD on/off check box

16. A new tab will appear called “ETD Reagent”

17. Click the “ETD Reagent” tab

18. Click the reagent tune mode check box

19. Set the reagent mass to 202 m/z (mass of fluoranthene)

20. Under the Reagent Trapping subtab, set the entrance lens to 0 V, reaction voltage to 1.5 V, exit lens to 0 V

21. Under the Reagent Running Quench subtab, set the entrance lens to 2.3 V, CC bias quench to −0.5 V and exit lens to 15 V

22. Under the Reagent Tuning Extract subtab, set entrance lens to 15 V, CC bias to 0.6 V and exit lens to −20 V

23. Click tune

24. If the reagent peak is not seen or in low intensity, you will need to adjust the following instrument parameters:

    1. The time of flight may need to be lowered
    2. In the “Ion Transfer” tab, set octopole frequency to 5 MHz
    3. In the “ETD Reagent tab, adjust the collision RF amplitude and mirror RF amplitude until reagent ion intensity is 10⁷ intensity

25. Once the reagent has been tuned, deselect reagent tune mode check box

26. Under the “Source MS/MS” tab, under the ETD subtab there are three adjustable parameters

    1. Accumulation time (Accum s) will be the same value as what you have set in the instrument control panel
    2. Reaction time should be set to 50 ms
    3. Reagent accumulation time should be set to 50 ms

27. Click the tune button

28. While tuning, increase the reagent accumulation time slowly to achieved desired level of fragmentation (200 ms-1000 ms depending on ionization state of precursor)

29. Be careful to not fully react your precursor ion, try to maintain your precursor no less than half the intensity of your most intense fragment ion

30. Once fragmentation is at the desired level, click stop

31. Figure 5 shows sample NETD data for a GAG sample.

32. Change the average number of scans in the instrument control panel to 24

33. Click the acquisition button

34. Figure 6 shows a comparison of precursor and fragment ion intensity. Unlike EDD, NETD fragment ion intensity is expected to be higher in intensity

35. Repeat these steps as needed to collect multiple spectra of the same precursor, or of different precursors

Figure 5.

Example NETD spectrum of the [M-10H+5Na]⁵⁻ precursor ion of Fondaparinux (GlcNS6S-GlcA-GlcNS3S6S-IdoA2S-GlcNS6S-Me) with annotations. Results were obtained with a 150 ms reagent accumulation time and 24 signal averages.

Figure 6.

Example NETD spectrum of the [M-10H+5Na]⁵⁻ precursor ion of Fondaparinux (GlcNS6S-GlcA-GlcNS3S6S-IdoA2S-GlcNS6S-Me). Fragment ion intensity compared to precursor intensity is displayed after 24 signal averages.

BASIC PROTOCOL 6

ANALYSIS OF MS/MS DATA

Introductory paragraph:

The analysis of MS/MS data is a key component to GAG sequencing. GlycoWorkbench 2.0 is free software available for both Windows and iOS (Ceroni et al., 2008). GlycoWorkBench can be used to assigned glycosidic, cross-ring, and SO₃ loss fragment ions. Assignment of other neutral loss fragments will need to be done manually.

Materials:

PC computer

Bruker Data Analysis

GlycoWorkbench 2.0

Microsoft Excel

Optional: Microsoft PowerPoint and Adobe Illustrator

Protocol steps

1. Note: the following protocol assumes a background knowledge of the Domon and Costello nomenclature for denoting carbohydrate fragment ions (Domon & Costello, 1988)

2. Open data file in Bruker Data analysis software

3. Click the “MassList” tab, then click Parameters

   1. Under peak finder, click FTMS
   2. Set the S/N threshold to 4, relative intensity threshold to 0.01% and absolute intensity threshold to 100
   3. Click OK

4. Click the “MassList” tab again, then click Find

5. If your masslist is not shown, click the “Window” tab, then click Spectrum Data

6. Copy your masslist and paste into an Excel sheet

   1. In Excel, delete all columns besides m/z and intensity
   2. Save masslist as a text file

7. Open Glycoworkbench 2.0

8. Draw your expected structure using the glycan cartoons

   1. Right click on the drawing screen, click add residue to add the appropriate sugar residue
   2. Repeat this for all sugar residues and any substituents (such as sulfate modifications)
   3. Add linkage information
   4. Add location of any known sulfate modifications
9. Right click on the drawn structure, click Mass Options of Selected Structures

   1. Modify the derivatization and reducing end tabs to reflect the expected structure
   2. Check the negative mode checkbox
   3. Under # H ions set the charge state of the precursor ion
   4. Add any sodium exchanges for the precursor to ex. Na ions
   5. Click OK

10. On the right-hand side, click “PeakList”

11. On the bottom, there is a series of symbols. Click the symbol that looks like an open white folder. This will prompt you to upload a document. Click on your peaklist text file that was previously saved

12. Figure 7 shows the expected screen when a structure has been drawn and peaklist has been imported

13. Once your peaklist is imported, at the top of the screen click “Tools”

14. Hover the mouse over Annotation, do not click

15. Move the mouse down to Annotate Peaks with Fragments from Current Structure

16. A “Fragment options” window will open

    1. For fragment types, check B, C, Y, Z, A and X (do not have internal fragments clicked)
    2. Set the max n.o. cleavages to 1
    3. Set the max n.o. cross rings to 1
    4. Make sure the max # H ions and max ex. Na ions are the same values as what you entered for your precursor

       1. To change these, you will need to uncheck the derive options from parent ion check box. Remember to re check this box when finished making changes
    5. Set the accuracy to 10 ppm
    6. Click OK

17. Once the software has finished loading, on the right-hand side you will see an image of your expected structure and a value for the number of assigned peaks (ex. x/xx (x%))

18. Click the “Details” tab- this will display all the assigned fragments

19. Figure 8 shows the expected screen when peaklist masses have been matched with known fragment ion masses

20. Click the ion tab to organize the table by ion

21. Highlight all matches that have an input for ion

    1. Right click and select show only selected annotations. This will eliminate any unassigned fragment peaks

22. Click the intensity tab to organize the table by intensity

23. We will now create a calibration file using high intensity glycosidic fragment ions

24. Toggle back to Bruker Data Analysis

25. Click the “Calibrate” tab, then click Edit Reference Mass Lists

    1. This will open a new “Compass ReferenceMassListEditor” window
    2. The following information will be pulled from your Glycoworkbench results

       1. Under name put the fragment type (ex. B3)
       2. Under z put the charge of the fragment ion
       3. Under m/z put the Ion m/z (note this is not the mass to charge observed, but the expected mass to charge of the assigned fragment)
    3. Repeat this for a total of 8 reference masses (this is the maximum number of reference masses you can input)
    4. Save your reference mass list

26. Click the “Calibrate” tab, then click Internal

    1. This will open a “Internal Mass Spectrum Calibration” window

       1. At the top, select the masslist you created

       2. Any matching masses to your reference masslist will appear

       3. The Error [ppm] column should be under 1 ppm. Delete any reference masses that have a high error; as masses are eliminated the error will change.

       4. Note: you need at least 2 calibration points

       5. When finished, click “recalibrate”

27. Once the internal calibration is complete, the masslist will have slightly changed. Save your new masslist as a text file (see step 5)

28. Toggle back to Glycoworkbench 2.0

29. Repeat steps 10–19 with your new masslist (in step 15e, change the accuracy to 5 ppm)

30. If you would like to save this information into a table, the data can be copied from Glycoworkbench 2.0 and pasted into an excel document

31. If you would like to annotate your spectrum with your matches, you can annotate directly in Bruker Data Analysis by clicking “Annotation” and then clicking Annotate. You can then click on the peak you want to label, and type the label

    1. Your annotated spectrum can then be copied (right click the blue bar at the top of the spectrum and select copy) and pasted into either Microsoft PowerPoint or Adobe Illustrator to create a figure

Figure 7.

Sample GlycoWorkBench screen with structure drawn and peaklist inputted.

Figure 8.

Sample GlycoWorkBench screen displaying annotation results.

Commentary

Background Information

Glycosaminoglycans (GAGs) are challenging analytical targets due to their inherent structural diversity. GAGs are linear carbohydrates that are covalently bound to a core protein, making up proteoglycans. There are four main groups of GAGs: Heparin (Hp)/ heparan sulfate (HS), chondroitin sulfate/ dermatan sulfate (DS), keratan sulfate (KS), and hyaluronic acid (HA) (Ly, Laremore, & Linhardt, 2010; Pepi, Sanderson, Stickney, & Amster, 2020; Varki et al., 2009). Of these groups, Hp/HS and CS/DS are the focus of structural characterization as they are more diverse and complex and participate in a variety of biological processes. Hp/HS and CS/DS are composed of repeating disaccharide units of a N-acetyl amino sugar and a uronic acid (Clegg et al., 2006; Rabenstein, 2002). Initial biosynthesis steps compose an elongated uniform chain. Deacetylases, sulfotransferases, and epimerases produce heterogeneous complex chains (Varki et al., 2009). Hp/HS are strong anticoagulants and prevent vein thrombosis and pulmonary embolism. Hp/HS are the most structurally complex group of GAGs and can contain up to four sulfate modifications per disaccharide unit (Rabenstein, 2002). CS/DS have anti-inflammatory and pain reducing properties and are used as a treatment for osteoarthritis and cataracts (Sugahara & Kitagawa, 2000). CS/DS are less structural complex than Hp/HS and can have up to three sulfate modifications per disaccharide unit, though it is more common to see one- two sulfate modifications. Due to sulfate modifications and carboxyl groups, GAGs are negatively charged. Mass spectrometry analysis in negative ion mode is therefore a useful tool for the analysis of GAGs.

Electron capture dissociation (ECD) is a strong ion activation method for cations. ECD utilizes low-energy electrons (<1 eV) for recombination with multiply charged precursor cations (Zubarev, 2004; Zubarev, Kelleher, & McLafferty, 1998). This results in radical formation. Electron detachment dissociation (EDD) can be considered a negative mode complement to ECD. EDD utilized moderate energy electrons (~19 eV) for irradiation of multiply charged precursor anions (J. J. Wolff et al., 2007; Wolff et al., 2008a). Both ECD and EDD are nonergodic processes, and produce more extensive fragmentation than threshold-based activation methods. In addition to GAGs, EDD can be used to sequence a variety of negatively charged samples. Due to the acidic nature of some post-translational modifications (PTMs), EDD has been utilized to sequence proteins, and fragments along the protein backbone. Additionally, EDD has been used for oligonucleotides and non-GAG oligosaccharides.

Electron transfer dissociation (ETD) is an ion-ion reaction the utilizes a reagent anion and multiply charge analyte cations. The reagent anion delivers an electron to the analyte cation, resulting in electron transfer and radical formation (Syka, Coon, Schroeder, Shabanowitz, & Hunt, 2004). Negative electron transfer dissociation (NETD) is the negative mode complement to ETD. NETD uses a reagent cation that removes an electron from the analyte anion (Coon, Shabanowitz, Hunt, & Syka, 2005; Leach et al., 2017; Riley, Bern, Westphall, & Coon, 2016). Like EDD, NETD has also been utilized on acidic proteins, oligonucleotides and non-GAG oligosaccharides.

In addition to EDD and NETD, there are other ion activation methods that can be used for GAG sequencing. Collision induced dissociation (CID) and infrared multiphoton dissociation (IRMPD) are vibrational excitation methods which produce an abundance of glycosidic fragment ions (Agyekum, Pepi, et al., 2017; Muchena J. Kailemia, Li, Ly, Linhardt, & Amster, 2012; M. J. Kailemia et al., 2015; Wolff, Laremore, Leach, Linhardt, & Amster, 2009). However, vibrational excitation methods also produce an abundance of sulfate decomposition and require highly ionized precursor ions. Electron induced dissociation (EID) can be used on singly-charged precursor ions, and produces similar fragmentation to EDD. Recent work has shown that charge transfer dissociation (CTD) and ultraviolet photodissociation (UVPD) are strong tools for GAG analysis, producing glycosidic and cross-ring fragmentation similar to EDD and NETD (Klein, Leach, Amster, & Brodbelt, 2019; Pepi, Sasiene, Mendis, Jackson, & Amster, 2020).

Critical Parameters

The protocols here have been proven effective for Hp/HS and CS/DS GAGs. When analyzing other GAG groups or glycan classes some variation may occur. The chain length and complexity of the GAG sample may also influence some of the parameters discussed in these protocols. Proper storage of samples is imperative. Long-term storage of GAGs in 4°C can result in degradation. To maintain GAG samples, storing at −20°C is strongly recommended. These protocols are based on a 0.1 mg/mL working concentration of sample. If working with a higher or lower concentration, many of the parameters discussed will need to be optimized. These protocols utilize FT ICR MS and assume users will have access to this instrumentation and accompanying software. NETD can be implemented on a variety of instrumentation; however, high-resolution MS is needed to confidently assign fragment peaks. A windows computer is needed to used Bruker’s Data analysis software. FT ICR data save as folders, and data files can be quite large. Appropriate storage space is needed to store data files, and an external hard drive may be useful.

Troubleshooting

Sulfate modifications of GAGs are extremely labile. Soft source conditions are crucial to maintaining these modifications before activation occurs. The source conditions outlined in these protocols may need to be adjusted if in-source fragmentation occurs. Skimmer 1 and collision voltage are the main causes of in-cell fragmentation and should be lowered if collisions occur. Accumulation time may also need to be lowered or increased to maintain the appropriate number of ions in the cell. The total ion count, or TIC, should be maintained at low 10⁸ intensity (1 ×10⁸ – 3×10⁸). Depending on the size of the GAG being investigated, the time of flight may also need to be adjusted. 0.500 ms is typical for GAGs ranging from 4–10 degrees of polymerization (dp); however, this may need to be increased slightly for longer GAG chains. For EDD, the ECD lens value can be adjusted to achieve optimal fragmentation. Typical values range from −17.5 - −19.0 V. For NETD, the reagent accumulation time may need to be adjusted to achieve optimal fragmentation. Typical values range from 200 – 1000 ms.

Understanding Results

The described protocols have been successfully used for sequencing Hp/HS and CS/DS GAG chains ranging from dp4- dp10 in size. Anticipated results are rich spectra of fragment ions both below and above the precursor ion m/z. For EDD, fragment ion intensity is expected to be low (relative intensity ~ 2–10%). For NETD, fragment ion intensity can range from low to high, occasionally with fragment ions becoming more intense than precursor ions. Both methods should produce both glycosidic and cross-ring fragment ions, as well as neutral loss peaks. Additionally, both methods should produce charge reduced precursor ions. NETD can result in charge reduced precursor ions paired with fluoranthene. Samples with uniform chains and no linker region on the reducing end will result in isobaric fragments.

Time Considerations

The time required for this analysis depends on multiple factors. If using a nanospray source instead of conventional ESI, nanospray tips can become clogged and may need to be replaced. Basic protocol 1 can require as long as 1 working day and as short as 30 mins, dependent on the nature of the sample. Basic protocol 2 may take between 30 mins- 1 hr; however, once the method is saved, the protocol will not need to be repeated for future analysis. Basic protocol 3 typically requires less than 30 mins. Basic protocol 4 is dependent on the number of samples being analyzed and the number of spectra to be collected per sample. 1 spectrum of 1 sample takes about 2.5 min, depending on the number of scans averaged. Basic protocol 5 is also dependent on the number of samples and the number of spectra being collected. 1 spectrum of 1 sample takes about 1 min, depending on number of scans averaged and reagent accumulation time. Basic protocol 6 can be fairly time consuming depending on the extent of data workup desired. To create a table with the information provided from GlycoWorkBench, about an hour is needed. To create a table and create a figure can take up to 3 hrs, depending on the familiarity of the scientist with PowerPoint and Illustrator.