Introduction

The emerging field of glycomics aims to expand our understanding of the function of glycans in cellular communication, protein structure and function, and human disease. Among the prerequisites for the development of glycomics is a robust and general method for synthesizing complex oligosaccharides. This can be compared to the fields of genomics and proteomics, which were pre-dated and made possible by the invention of automated methods for synthesis of polynucleotides, peptides and proteins. Although the structural and stereochemical complexity of oligosaccharides makes the development of synthetic methods more difficult, significant progress has been made toward the goal of making synthetic oligosaccharides readily available to the research community. Here, we describe a typical protocol for programmed reactivity-based one-pot synthesis of a complex glycan, using fucosyl GM1 hexasaccharide (1) as an example. Fucosyl GM1 is specifically associated with small-cell lung cancer and is therefore a target for vaccine development. This protocol serves to illustrate the methodology of programmable reactivity-based one-pot oligosaccharide synthesis1, which we have used to prepare a wide variety of oligosaccharides2,3,4,5,6,7.

The approach is based on the observation that the protecting groups attached to a particular glycosyl donor greatly affect its rate of reactivity, with relative rates ranging over several orders of magnitude1,8. By compiling a database of relative reactivity values (RRVs) for a large library of glycosyl donors and developing an algorithm to rank the suitability of specific donor molecules for the synthesis of a given oligosaccharide target, we have made it possible to program a synthetic protocol that can be performed in a single pot, without the need to isolate intermediates or to unmask protecting groups. Although the complete database and computer program are proprietary, it is relatively straightforward to determine RRVs of glycosyl donors experimentally, as described in Box 1. The relative rates are determined by HPLC analysis of a competition assay between the glycosyl donor of interest and a reference donor of known RRV. In addition, general trends in reactivity have emerged to guide the design of an effective synthetic strategy. For instance, electron-donating protecting groups such as the benzyl ethers in 3 increase donor reactivity, whereas electron-withdrawing groups such as the acetate esters in 4 decrease reactivity. In fact, the chemical shift of the donor's anomeric proton, which is sensitive to local electron density, correlates very well with its RRV1. The identity of the sugar is also predictive, with the general RRV pattern being fucose > galactose > glucose > mannose.

Several types of glycosyl donors have been developed over the years. Thioglycosides, first used by Garegg9, are among the most versatile because they are stable and can be activated by a wide variety of different reagents. Therefore, our relative reactivity database was generated using thioglycoside donors1. We also chose tolyl thioglycosides as building blocks because the leaving group tolyl thiol is chromogenic and the RRV can easily be determined by HPLC, measuring UV absorbance. Several thioglycoside donor activation reagents have been investigated, including N-iodosuccinimide-triflic acid (NIS-TfOH)10, N-(phenylthio)-ε-caprolactam-triflic anhydride11 and 1-benzenesulfinylpiperidine-triflic anhydride (BSP-Tf2O)12. The effort to develop new activation reagents is driven by the difficulty in finding a single general reagent that provides high yield and good α/β-selectivity for all substrates. The activation method employed in this protocol is BSP-Tf2O, which usually works very well in our hands. The scheme for preparation of the thioglycoside building blocks 3 and 4 is shown in Figure 1, and detailed protocols for synthesis of these building blocks are described in Box 2. The scheme for one-pot assembly of 1 is shown in Figure 2. For the sake of brevity, the reader is referred to the original publication4 and the supporting information included therein for preparation of trisaccharide acceptor 5. Similar to the case of peptide synthesis, some oligosaccharides are more difficult to assemble in high yield than others—for instance, in cases of steric congestion. In these difficult cases, one-pot synthesis may not be optimal because of the difficulty of purifying the final product obtained in lower yield from the abundance of byproducts and unreacted starting material also present in the reaction mixture. In such cases, traditional stepwise methods, although more time-consuming, may give better results because they allow for purification of intermediates at each step.

Figure 1
figure 1

Scheme for synthesis of glycosyl donor building blocks 3 and 4.

Full size image
Figure 2
figure 2

Scheme for reactivity-based one-pot synthesis of 2 and deprotection to fucosyl GM1 oligosaccharide 1.

Full size image

The protecting groups on the disaccharide glycosyl acceptor 4 were selected such that the rate of the undesired self-reaction of two molecules of 4 would be approximately 25 times slower than the rate of reaction with donor 3 (RRV 2,839 versus 72,000). After completion of the first glycosylation, the trisaccharide intermediate is reacted with trisaccharide acceptor 5 in the same pot, without purification of the intermediate, to produce fully elaborated hexasaccharide 2. Trisaccharide 5 does not have a thioglycoside function on the reducing end and therefore cannot react as a glycosyl donor under these conditions (RRV = 0). A series of standard deprotection steps produces the final product 1. First, zinc (Zn) dust in the presence of acetic anhydride converts the trichloroethyl carbamate (Troc) into an N-acetyl group on glucosamine, then the acetate and sialic acid methyl esters are deprotected with sodium methoxide and sodium hydroxide, respectively. Finally, the benzyl groups and side-chain benzyl carbamate are deprotected by palladium (Pd)-catalyzed hydrogenolysis, and silica gel chromatography provides 1 in 21% overall yield from compounds 3 and 4.

Materials

REAGENTS

Equipment

  • Coffee mill (Bunn)

  • Vacuum pump (Welch)

  • Vacuum manifold with argon bubbler (Chemglass)

  • Propane torch (Bernz-O-Matic)

  • Dewar flask (Chemglass)

  • Teflon-coated magnetic stir bars (Sigma-Aldrich)

  • Glass polypropylene syringes (Hamilton) or disposable polypropylene syringes (VWR)

  • Reusable hypodermic syringe needles (Hamilton) or disposable syringe needles (VWR)

  • Hand-held UV lamp (UVP)

  • Rotary evaporator (Büchi)

  • Pyrex chromatography column (Chemglass)

  • NMR tubes (Kontes)

  • NMR spectrometer (Bruker DRX-500 or DRX-600)

  • Filter paper (Fisher, cat. no. 09-801C)

  • Balloons (Sigma-Aldrich)

  • Silica gel 60 F254 thin-layer chromatography(TLC) plates (EMD Chemicals, Inc., cat. no. M5715-7)

  • pH indicator strips (Fisher)

  • HPLC (Hitachi LaChrom Elite) fitted with an analytical reverse phase column (Vydac)

Procedure

Synthesis of protected Fuc-GM1 (2)

  1. 1

    Remove BSP from refrigerator (4 °C) and allow it to warm to room temperature.

  2. 2

    Grind molecular sieve AW-300 pellets into powder using a coffee mill.

  3. 3

    Weigh 0.75 g powdered molecular sieves AW-300 into a 25 ml round-bottomed flask containing a Teflon-coated magnetic stir bar.

  4. 4

    Put the flask under vacuum (approximately 0.3 mm Hg) and flame-dry it using a propane torch, then allow the flask to cool to room temperature while still under vacuum. It may be left under vacuum overnight if desired.

  5. 5

    Fill the flask with argon gas, remove from vacuum manifold and cap with a rubber septum.

  6. 6

    Weigh 92 mg (0.17 mmol, 1.06 equiv.) fucosyl donor 3, 160 mg (0.16 mmol, 1.0 equiv.) disaccharide building block 4 (Box 3) and 18 mg (0.088 mmol, 0.55 equiv.) BSP into the flask; re-cap with the septum, attached to an argon inlet with a syringe needle through the septum. The flow of argon into the flask is monitored by attaching a bubbler to the argon line through the vacuum manifold. Maintain positive argon flow throughout Steps 6–13. Turn the magnetic stirrer on.

  7. 7

    Transfer 2 ml dry, freshly distilled dichloromethane into the flask with a dry glass syringe fitted with a 20-gauge hypodermic needle.

  8. 8

    Pre-dry the reaction mixture at room temperature under positive argon flow for 1 h, as described in Step 6, and then cool the flask to −70 °C in a dry ice/isopropyl alcohol bath contained in the Dewar flask, while maintaining argon flow into the flask.

  9. 9

    Add 16 μl (0.096 mmol, 0.6 equiv.) Tf2O into the reaction flask using a syringe and allow the temperature to increase gradually from −70 °C to −10 °C over 1.5 h by allowing the dry ice to evaporate.

    Critical Step

    The success of this step depends on the temperature of the reaction. To prevent over-reaction, the temperature should not be allowed to increase above −5 °C. If the temperature rises too quickly, add more dry ice to the Dewar.

  10. 10

    After donor 3 is consumed (as monitored by TLC, developing with hexane/ethyl acetate (2.5/1) and visualizing by UV absorbance at 254 nm), cool the reaction flask to −70 °C again. Samples for TLC analysis can be obtained without disrupting the reaction by inserting a glass capillary through a disposable syringe needle passed through the septum and dipping the capillary into the solution. TLC analysis can be performed every 30 min to monitor the progress of the reaction.

  11. 11

    Weigh 0.34 g trisaccharide acceptor 5 (0.22 mmol, 1.4 equiv.) and 17 mg (0.08 mmol, 0.5 equiv.) BSP into the flask; re-cap with the septum, under a positive argon flow.

  12. 12

    Dispense 15 μl (0.088 mmol, 0.55 equiv.) Tf2O into the flask using a syringe, then increase the temperature gradually from −70 °C to 0 °C as in Step 9. Then continue stirring at 0 °C for an additional 4 h in an ice/water bath.

  13. 13

    Transfer 0.2 ml triethylamine into the flask to quench the reaction, and remove the flask from the cooling bath.

  14. 14

    Add 3 ml dichloromethane (does not have to be distilled) into the flask to dilute the reaction, and remove the molecular sieves AW-300 through a filter funnel attached to a 25 ml round-bottomed flask.

  15. 15

    Transfer the filtrate into a separatory funnel and wash the dichloromethane solution with saturated aqueous sodium thiosulfate (1 × 5 ml), saturated aqueous sodium hydrogen carbonate (1 × 5 ml), water (1 × 5 ml) and saturated aqueous sodium chloride (brine) (1 × 5 ml) sequentially.

  16. 16

    Dry the dichloromethane solution by adding approximately 1 g anhydrous sodium sulfate, stir for 5 min, filter the mixture under gravity through a fluted filter paper on a funnel to remove sodium sulfate and collect the filtrate in a round-bottomed flask.

  17. 17

    Evaporate the dichloromethane using a rotary evaporator at room temperature under aspirator vacuum.

    Pause point

    The crude product can be stored at 4 °C overnight.

  18. 18

    Pack a chromatography column (1.5 cm i.d. × 30 cm length) with silica gel using a 1:2 (vol/vol) mixture of toluene and ethyl acetate.

  19. 19

    Load the crude mixture from Step 17 on the top of the silica bed and then cover the top of the column further with a layer of sand (approximately 0.5 cm thick).

  20. 20

    Elute the products with a 1:2 (vol/vol) mixture of toluene and ethyl acetate, and collect fractions of 5–10 ml.

  21. 21

    Identify fractions containing protected Fuc-GM1 2 by silica gel TLC, developing with a 1:2 (vol/vol) mixture of toluene and ethyl acetate (Rf = 0.32). Products are visualized by UV absorbance at 254 nm.

  22. 22

    Collect the fractions and evaporate the solvent using a rotary evaporator. Dry the residue under reduced pressure to give 0.21 g of the protected Fuc-GM1 2 (47%) as a white glassy solid.

    Pause point

    Compound 2 can be stored indefinitely at 4 °C before carrying on to the deprotection steps described below.

  23. 23

    Characterize the protected Fuc-GM1 2 from Step 22 by NMR spectroscopy using deuterated chloroform as solvent, referring to spectral data provided under ANTICIPATED RESULTS.

Synthesis of deprotected Fuc-GM1 (1)

  1. 24

    Activate the Zn dust by washing sequentially with 1 M aqueous HCl, H2O, methanol and diethyl ether through a filter funnel attached to a 100 ml round-bottomed flask.

  2. 25

    Transfer the freshly activated Zn dust into a vial and dry it under vacuum.

  3. 26

    Weigh 80 mg (29 μmol) protected Fuc-GM1 into a flask containing a Teflon-coated magnetic stir bar and cap with a rubber septum attached to the argon inlet. Turn the magnetic stirrer on.

  4. 27

    Transfer 1 ml freshly distilled dichloromethane and 1 ml acetic anhydride into the flask with a glass syringe fitted with a 20-gauge hypodermic needle.

  5. 28

    Add 1 g freshly activated Zn dust into the flask and stir the reaction at room temperature for 5 h.

    Pause point

    The reaction mixture can be left at room temperature overnight.

  6. 29

    Remove the Zn dust by filtration through a filter funnel attached to a 25 ml round-bottomed flask.

  7. 30

    Evaporate the solvent using a rotary evaporator under aspirator vacuum.

  8. 31

    Dilute the residue with 5 ml dichloromethane and transfer the resulting solution into a separatory funnel.

  9. 32

    Wash the dichloromethane solution with saturated aqueous sodium hydrogen carbonate (1 × 5 ml) and brine (1 × 5 ml).

  10. 33

    Dry the dichloromethane solution by adding anhydrous sodium sulfate as in Step 16, filter the mixture under gravity through a fluted filter paper on a funnel and collect the filtrate in a flask.

  11. 34

    Evaporate the dichloromethane using a rotary evaporator at room temperature under aspirator vacuum.

  12. 35

    Pack a chromatography column (1.5 cm i.d. × 30 cm length) with silica gel using a 3:1 (vol/vol) mixture of toluene and acetone.

  13. 36

    Load the crude mixture from Step 34 on the top of the silica bed and then cover the top of the column further with a layer of sand (approximately 0.5 cm thick).

  14. 37

    Elute the column with a 2:1 (vol/vol) mixture of toluene and acetone.

  15. 38

    Identify fractions containing desired N-acetamido product by TLC by developing plates with 2:1 (vol/vol) mixture of toluene and acetone (Rf = 0.12).

  16. 39

    Collect the fractions and evaporate the solvent using a rotary evaporator. Concentrate the residue under reduced pressure.

  17. 40

    Add a Teflon-coated magnetic stir bar into the flask containing the residue from Step 39 and cap with a rubber septum attached to the argon inlet, as in Step 6. The presence of water will interfere with the reaction. Turn the magnetic stirrer on.

  18. 41

    Transfer 2 ml 1:1 (vol/vol) mixture of anhydrous methanol and freshly distilled dichloromethane into the flask using a glass syringe fitted with a 20-gauge hypodermic needle.

  19. 42

    Dispense 50 μl 25% solution of sodium methoxide in methanol into the reaction flask at room temperature using a syringe. Stir at room temperature for 10 h.

    Pause point

    Can be left overnight at room temperature.

  20. 43

    Add Amberlite IRC-50 acidic resin into the flask to neutralize the reaction until the pH is approximately 7, as determined with a pH indicator strip.

  21. 44

    Remove Amberlite IRC-50 acidic resin through a filter funnel attached to a 25 ml round-bottomed flask and rinse with methanol.

  22. 45

    Evaporate the solvent using a rotary evaporator under aspirator vacuum to give the deacylated residue.

  23. 46

    Add a Teflon-coated magnetic stir bar into the flask containing the residue from Step 45. Turn the magnetic stirrer on.

  24. 47

    Transfer 3 ml 2:1 (vol/vol) mixture of methanol and tetrahydrofuran into the flask using a plastic syringe fitted with a 20-gauge hypodermic needle.

  25. 48

    Add 0.5 ml 2 M aqueous sodium hydroxide solution into the reaction flask at room temperature using a plastic syringe fitted with a 20-gauge hypodermic needle. Continue stirring for 3 h at room temperature.

  26. 49

    Carefully add concentrated HCl (approximately 80 μl) into the flask to neutralize the reaction, until the pH is approximately 7, as determined with a pH indicator strip.

  27. 50

    Evaporate the solvent using a rotary evaporator under aspirator vacuum.

  28. 51

    Add a Teflon-coated magnetic stir bar into the flask containing the residue from Step 50. Turn the magnetic stirrer on.

  29. 52

    Transfer 3 ml solution of methanol with 10% (vol/vol) formic acid into the flask.

  30. 53

    Add 70 mg Pd-black into the reaction flask and displace the air in the flask with a balloon containing argon; then displace the argon with a balloon containing hydrogen.

  31. 54

    Stir the reaction under 1 atmospheric hydrogen (balloon full of hydrogen gas) at room temperature for 18 h.

    Pause point

    Can be left overnight at room temperature.

  32. 55

    Remove the balloon and add a 5 M solution of NaOH in H2O into the flask to neutralize the reaction to pH 7.

  33. 56

    Remove the Pd-black by filtration through a filter funnel containing a layer of celite (approximately 3 cm) attached to a 25 ml round-bottomed flask.

  34. 57

    Evaporate the solvent using a rotary evaporator under aspirator vacuum.

  35. 58

    Pack a chromatography column (1.5 cm i.d. × 15 cm length) with Iatro beads using a 3:4:1 (vol/vol/vol) mixture of dichloromethane, methanol and H2O.

  36. 59

    Load the crude mixture from Step 57 on the top of the Iatro bead bed and then cover the top of the column further with a layer of sand (approximately 0.5 cm thick).

  37. 60

    Elute the column with a 3:4:1 (vol/vol/vol) mixture of dichloromethane, methanol and H2O.

  38. 61

    Identify fractions containing Fuc-GM1 1 by TLC by developing plates with a 3:4:1 (vol/vol/vol) mixture of dichloromethane, methanol and H2O (Rf = 0.1).

  39. 62

    Collect the fractions and evaporate the solvent using a rotary evaporator.

  40. 63

    Dissolve the residue from Step 62 in 5 ml H2O and transfer the solution into a 20 ml vial.

  41. 64

    Freeze the solution with liquid nitrogen and lyophilize under reduced pressure to give 15 mg Fuc-GM1 1 (44% over four steps) as a white solid.

    Troubleshooting

Troubleshooting

Troubleshooting advice can be found in Table 1.

Table 1 Troubleshooting table.
Full size table

Timing

Protected Fuc-GM1 (2) Step 1: 30 min

Step 2: 5 min

Steps 3 and 4: 30 min

Steps 5–7: 10 min

Step 8: 1 h

Step 9: 1.5 h

Steps 10 and 11: 15 min

Step 12: 6 h

Steps 13 and 14: 15 min

Step 15: 20 min

Steps 16 and 17: 30 min

Step 18: 15 min

Step 19: 5 min

Steps 20–22: 3 h

Step 23: 30 min.

Total for Steps 1–17: 11 h, should be performed in a single day.

Steps 18–23: 4 h.

Fuc-GM1 (1) Steps 24 and 25: 30 min

Step 26: 5 min

Step 27: 5 min

Step 28: 5 h

Steps 29 and 30: 30 min

Steps 31 and 32: 15 min

Step 33: 15 min

Step 34: 15 min

Step 35: 15 min

Step 36: 5 min

Steps 37–39: 2 h

Steps 40 and 41: 5 min

Step 42: 10 h

Steps 43–45: 30 min

Steps 46 and 47: 5 min

Step 48: 3 h

Step 49: 5 min

Step 50: 30 min

Steps 51 and 52: 5 min

Step 53: 10 min

Step 54: 18 h

Step 55: 5 min

Steps 56 and 57: 30 min

Step 58: 15 min

Step 59: 5 min

Steps 60–62: 3 h

Steps 63 and 64: 24 h.

Total for Steps 24–64: 3 d.

Anticipated results

A typical yield for the one-pot glycosylation operations (Steps 1–23) is 47%. A typical yield for the deprotection steps (Steps 24–64) is 44%, for a total yield of 21% of Fuc-GM1 1 from protected building blocks 3, 4 and 5.

Analytical data

Protected Fuc-GM1 (2)

1H NMR (500 MHz, CDCl3) δ 7.52 (d, J = 7.4 Hz, 2H), 7.36–7.07 (m, 63H), 5.96 (d, J = 8.8 Hz, 1H), 5.52–5.47 (m, 2H), 5.26-5.09 (m, 6H), 4.96–4.92 (m, 2H), 4.84–4.72 (m, 6H), 4.68–4.38 (m, 19H), 4.33–4.29 (m, 3H), 4.22–4.15 (m, 4H), 4.10–3.94 (m, 11H), 3.89–3.86 (m, 2H), 3.81 (s, 3H), 3.76–3.73 (m, 2H), 3.67–3.62 (m, 5H), 3.56–3.44 (m, 10H), 3.35–3.26 (m, 2H), 2.34–2.30 (m, 1H), 2.15 (s, 3H), 1.99 (s, 3H), 1.97 (s, 3H), 1.96 (s, 3H), 1.92 (s, 3H), 1.90 (s, 3H), 1.88 (s, 3H), 1.03 (s, 3H);

13C-Apt NMR (150 MHz, CDCl3) δ 170.65, 170.36, 170.31, 169.81, 169.60, 169.27, 168.58, 156.42, 156.16, 154.10, 139.15, 138.91, 138.70, 138.61, 138.50, 138.29, 138.12, 137.83, 137.78, 136.61, 136.47, 129.56, 128.97, 128.43, 128.40, 128.33, 128.24, 128.20, 128.12, 128.08, 127.97, 127.90, 127.88, 127.81, 127.76, 127.68, 127.64, 127.61, 127.56, 127.52, 127.42, 127.28, 127.26, 127.23, 127.14, 127.07, 127.03, 126.86, 126.75, 126.67, 126.44, 103.50, 102.76, 101.69, 99.12, 97.49, 95.78, 83.46, 82.54, 81.66, 79.04, 78.92, 77.84, 76.46, 76.17, 76.10, 75.93, 75.87, 74.96, 74.87, 74.81, 74.71, 74.52, 74.22, 74.15, 73.97, 73.52, 73.39, 73.22, 73.12, 72.50, 72.45, 72.31, 71.92, 71.62, 70.72, 70.09, 69.09, 69.06, 68.34, 68.20, 68.18, 68.00, 67.87, 67.22, 67.16, 66.44, 62.48, 61.66, 54.91, 53.15, 51.36, 49.34, 47.03, 45.96, 35.59, 29.63, 23.15, 21.22, 20.90, 20.78, 20.58, 20.55, 20.45, 16.82; HRMS (m/z) [M+Na+] calcd for C151H168Cl3N3O42Na, 2823.0060; found 2823.0114.

Fuc-GM1 (1)

1H NMR (600 MHz, D2O)™ 5.07 (d, J = 3.5 Hz, 1H), 4.51 (d, J = 7.5 Hz, 1H), 4.44 (d, J = 6.6 Hz, 1H), 4.37–4.35 (m, 2H), 4.06–4.02 (m, 1H), 3.95 (dd, J = 10.1, 2.2 Hz, 1H), 3.92–3.90 (m, 3H), 3.81 (d, J = 11.4 Hz, 1H), 3.75–3.41 (m, 29 H), 3.20–3.17 (m, 2H), 2.99 (s, 2H), 2.50 (dd, J = 12.3, 4.4 Hz, 1H), 1.85 (s, 6H), 1.73 (t, J = 12.3 Hz 1H), 1.03 (d, J = 6.6 Hz, 3H); 13C-Apt NMR (150 MHz, D2O)™ 174.56, 173.80, 173.54, 102.76, 102.14, 101.68, 101.58, 100.89, 98.70, 75.99, 75.39, 74.44, 74.34, 73.65, 72.57, 72.29, 71.83, 71.39, 69.52, 69.10, 68.69, 68.23, 67.99, 67.61, 66.89, 66.29, 62.39, 60.67, 60.45, 60.04, 59.54, 51.16, 51.07, 39.17, 36.86, 22.27, 21.60, 14.92; HRMS (m/z) [M-H]− calcd for C45H76N3O33 1186.4366; found 1186.4619.