One‐Flask Synthesis of Cyclic Diguanosine Monophosphate (c‐di‐GMP)

The bacterial signaling molecule cyclic diguanosine monophosphate (c‐di‐GMP) plays a key role in controlling biofilm formation and pathogenic virulence, among many other functions. It has widespread consequences for human health, and current research is actively exploring its molecular mechanisms. The convenient one‐flask, gram‐scale synthesis of c‐di‐GMP described here has facilitated these efforts and has been applied to a variety of analogs. © 2023 The Authors. Current Protocols published by Wiley Periodicals LLC.

A number of routes for the synthesis of c-di-GMP have been reported over the years, starting with van Boom's original phosphate triester approach (Ross et al., 1990). Since then, various combinations of amidite and H-phosphonate couplings have been employed (Amiot et al., 2006;Hayakawa et al., 2003;Hyodo & Hayakawa, 2004;Hyodo et al., 2006;Kiburu et al., 2008;Yan & Aguilar, 2007;Zhang et al., 2004). Until now, these methods have given only small amounts of products. In contrast, the method reported here (Gaffney et al., 2010) yields over 1 g of product and can readily be scaled up further. No chromatographic steps are required, and the intermediate tert-butylammonium salt and the final product 5 are purified by simple crystallizations.
NOTE: Reagents and solvents are available from Sigma-Aldrich, and the apparatus from Fisher.  Dry compound 1 1. Place 6.31 g (6.5 mmol) of 1 into a 100-ml pear-shaped flask. Dry the amidite twice by evaporation of 40 ml CH 3 CN on a rotary evaporator, the last time leaving ∼20 ml. Add ten 3-Å molecular sieves, and a stopper with a septum. Set aside until needed.
Perform hydrolysis, β-elimination, and detritylation of second portion of 1 to give 2 2. Place another portion of 1 (4.85 g, 5.0 mmol) into a 500-ml round-bottom flask with a stir bar. Add 1.16 g of pyridinium trifluoroacetate (6.0 mmol, 1.2 equiv) and rinse down the sides of the flask with 25 ml CH 3 CN. Add 0.18 ml H 2 O (10 mmol, 2 equiv) and stir for 1 min. Add 25 ml tert-BuNH 2 and let the solution stir for 10 min. Place the flask on a rotary evaporator and concentrate to a foam. Evaporate twice with 50 ml CH 3 CN, each time to give a foam. 4. After 10 min, quench with 7.0 ml pyridine (87 mmol, 2 equiv relative to DCA). Concentrate the solution to ∼20 ml and then evaporate three times with 40-ml portions of dry CH 3 CN, the last time leaving about 12 ml. Check by HPLC (HPLC conditions are described in the Critical Parameters section; retention times are given in Table  1) before the last concentration to be sure detritylation is complete. (If not, concentrate, evaporate several times with ethyl acetate to remove pyridine, and repeat step 3 with less acid, depending on how much tritylated material was left.) Stopper the flask with a septum.
Upon addition of pyridine, the orange color should immediately disappear.
Perform linear coupling of 1 with 2 and detritylation to give 3 5. Transfer the dried 1 to 2 through a 16-G double-tipped needle by inserting one end of the needle through the septum of the flask containing 1 all the way to the bottom, and the other end through the septum of the flask containing 2 but above the solution. Place a vent needle in the flask with 2. Place a needle-tipped line from a source of dry nitrogen or argon into the flask with 1, thereby causing the solution of 2 to flow into the flask with 1. 6. Immediately inject a 1-ml rinse of dry CH 3 CN into the flask that had contained 1 (from step 1), rotate the flask quickly to rinse the sides of the flask, and transfer this rinse through the double-tipped needle into the flask with 2. Repeat this rinse one more time with 1 ml fresh CH 3 CN. Allow the mixture of 1 and 2 to stir for 2 min, and then take a sample for HPLC analysis (HPLC conditions are described in the Critical Parameters section; retention times are given in Table 1).
7. As soon as the HPLC sample is removed, add 2.73 ml anhydrous 5.5 M tert-butyl hydroperoxide in decane (15 mmol, 3 equiv). Let the solution stir for 30 min and then take an HPLC sample. Once the HPLC sample is removed, place the flask in an ice bath and add 1.25 g NaHSO 3 dissolved in 2.5 ml H 2 O. After 5 min of stirring, concentrate to an oil.
The anhydrous tert-butyl hydroperoxide should be kept in a refrigerator but removed an hour before use. The NaHSO 3 solution takes some time to dissolve, so prepare it 15 min ahead of time, but not earlier.
Use the rest of the 6% dichloroacetic acid prepared in step 3.
9. After 10 min, quench the reaction with 15 ml pyridine. Take an HPLC sample, add 35 ml more pyridine, and concentrate to ∼20 ml. Add 150 ml pyridine and concentrate to 100 ml. Stopper the flask with a septum.
The orange color for this detritylation is often less intense than in step 3.
12. Slowly add 20 g NaHCO 3 (238 mmol) and stir for 5 min. 16. Add 25 ml CH 2 Cl 2 to the foam and immediately swirl vigorously until the solid dissolves. Stopper the flask with a septum, let it sit until crystal deposition has slowed, and then refrigerate overnight.
Crystallization should begin anywhere from 1 min to 1 hr, and should be nearly complete after 30 min.
18. Transfer the above into a 250-ml round-bottom flask. For each 1.00 mmol (1.20 g), add 100 ml CH 3 NH 2 (800 equiv) in anhydrous ethanol (33% by weight). Add a stir bar, cap with a septum, and stir for 90 min. Check completion of the reaction by HPLC. Concentrate to an oil. Add 4 ml pyridine and 2 ml Et 3 N, swirl to mix, and concentrate to an oil. Repeat this procedure two more times.
19. Add 4 ml pyridine to the flask, cap with a septum, and insert a vent needle. Place the flask in an oil bath at 50°C in a hood and stir. For every 1.00 mmol (1.20 g), draw up 14 ml Et 3 N and 8.3 ml Et 3 N·3HF (150 equiv F -) into separate syringes and insert them both through the septum of the flask. Add the two reagents simultaneously over ∼1 min. Stir the mixture for 1 hr at 50°C, occasionally rotating the flask at an angle to dissolve solids on the sides. 20. Remove the flask from the oil bath and start it stirring on a different stir plate. Remove the septum, and for every 1.00 mmol, add 112 ml HPLC-grade acetone in a slow, steady stream over 1 min. Continue to stir the mixture as the product crystallizes. After 10 min, collect the crystals by filtration in a sintered glass funnel and wash them five times using 5-ml portions of acetone. Stir thoroughly between each wash. If the product is sticky, use a spatula to repeatedly smear it against the sides of the funnel, thereby removing trapped impurities. Cover the funnel with filter paper and dry it in a desiccator over KOH overnight. Weigh it and check a sample by HPLC. Analyze 5 by 1 H and 31 P NMR.  159.3 (br), 156.4, 152.1 (br), 139.1, 117.5 (br), 92.8, 83.0, 75.8, 73.0, 65.0, 49.4, 10.8; 31 Take great care when handling the acetone filtrate and washes because they contain excess Et 3 N·3HF.

Critical Parameters
The detritylations in steps 3 and 8 are potentially reversible upon quenching and again upon concentration, in spite of the water added to prevent it. The concentrated oils of detritylated 2 and 3 should be used immediately, and not stored.
Conditions for the initial amidite coupling between 1 and 2 should be as dry as possible.
HPLC is useful for following the course of the synthesis. We used a Waters Atlantis C18 column, 4.6 mm × 50 mm, 3 μm, with gradients of 2% to 100% CH 3 CN and 0.1 M triethylamine acetate (TEAA; pH 6.8) for steps 1 to 17, and 2% to 40% at the end of step 18. However, many steps are time dependent, and the procedure must be continued immediately after taking samples, without waiting to see the results. Note that compounds with a cyanoethyl group may elute as two peaks because there are two diastereomers.
The linear dimer is best cyclized and oxidized as soon as it is made. Therefore, steps 1 to 13 should be performed on the same day.
Because of the danger of using Et 3 N·3HF, HPLC should not be performed to monitor progress of the final desilylation. Once the solids are in solution (there may be two layers), the reaction should go to completion in 1 hr.

Anticipated Results
The yield for crystallization of the intermediate tert-butylammonium salt should be ∼35% from 1, and the yield for crystallization of the final product, 5, should be ∼30% from 1. Both are white powders that are homogeneous by HPLC-mass spectrometry (HPLC-MS).

Time Considerations
The entire sequence of reactions can be done in 4 days, with final NMR analysis taking longer. Steps 1 to 13 should be performed all in 1 day (8 to 10 hr) or the yield will suffer. Steps 14 to 16 take a second day (2 to 3 hr), and step 17 takes only a few minutes of a third day. Steps 18 to 20 take a fourth day (6 to 8 hr).