Optimization of Phosphorylation and Other Reaction Conditions using 2’-

Compound 17 in Scheme 6.3 was prepared in 65% yield by tritylation of 2’-

deoxyadenosine using DMTrCl, TEA and 4-DMAP (as a catalyst) in pyridine.4 This relatively low yield was due to the conversion of the remainder of 2’-deoxyadenosine to a ditritylated product as identified by 1H NMR. This latter product was separated first during a silica gel column using 3% MeOH in CH2Cl2 as eluent; then 17 was collected by increasing the polarity of

N N N N NH₂ O OH DMTrO N N N N NH₂ O OPO(OBn)2 DMTrO N N N N NH2 O OPO(OBn)₂ HO N N N N NH2 O OPO(OBn)O- HO N N N N NH₂ O OH HO a b c d 17 18 19 20 65% 60-89% 86% 80%

Scheme 6.3. Compounds used to optimize reaction conditions. Reagents and conditions: (a) DMTrCl, 4-DMAP, TEA, Py, rt, 2 h; (b) i- (BnO)2PN(iPr)2, Me-tetrazol, THF, rt, 1 h; ii- m-

CPBA in CH2Cl2, -78 °C, 15 min; (c) 30% DCA in CH2Cl2, rt, 30 min; (d) DABCO, 1,4-

dioxane, reflux, 2 h.

5’-O-Dimethoxytrityl-2’-deoxyadenosine 3’-dibenzyl phosphate (18) was prepared from 17 using the earlier described phosphoramidite chemistry. Similar phosphitylation reactions in the literature used a range of equivalents of the phosphoramidite reagent (2.5-1.1).8-10 To test reacting the 3’-OH of 17 while avoiding N6-phosphorylation and to optimize the yield of 18, the reaction was run three times in THF using 2.5, 1.5, and 1.1 equivalents of (BnO)2PN(iPr)2 (see

Table 6.1). The amount of Me-tetrazol was equimolar to (BnO)2PN(iPr)2, and (BnO)2PN(iPr)2

was added dropwise at 0 °C to increase the reaction selectivity; finally, the reaction mixture was slowly warmed to room temperature as stirred for 1 h. Reacting 2.5 equivalents of the reagent led to complete consumption of 17 and showed a single spot on TLC. Later, 1H NMR of the

final oxidized phosphorylation products showed that this single TLC spot was actually due to phosphorylation at 3’-OH as well as at 6-NH2. Subsequent treatment with m-CPBA at -78 °C for

1 h produced 2 spots, one from each phosphorylation product. Table 6.1 shows that the yield of 18 in this case was 72% after chromatographic purification. Reacting 1.5 equivalents of (BnO)2PN(iPr)2 with 17 using the same procedures produced 18 in 89% yield after

chromatographic separation also, less N6-phosphorylated product was produced than with 2.5 equivalents of reagent. Decreasing the equivalents of (BnO)2PN(iPr)2 to 1.1, however, dropped

the yield of 18 to about 60% (as observed via TLC before oxidation) and 40% of 17 remained unreacted.

Table 6.1. Yield of 18 Versus the Number of Equivalents of (BnO)2PN(iPr)2.

Entry No. of equiv. %Yield

1 2.5 72a

2 1.5 89a

3 1.1 60b

a _{Yield after separation of final oxidized product 18.} b _{Yield estimated based on TLC spot size before the}

m-CBPA oxidation step.

As shown in Scheme 6.3, detritylation of 18 produced 2’-deoxyadenosine 3’-dibenzyl phosphate (19). This was accomplished by treatment of 18 with 40 equivalents of 30% DCA in dichloromethane at room temperature for 30 min. Lower concentrations or equivalents of the acid caused incomplete detritylation. This result was surprising as DMTr was readily cleaved at 5’-O in less than 1 min either using 3% DCA,11 or 1% TFA12 during solid-phase oligodeoxynucleotide synthesis. Here use of TFA was avoided, because it might lead to

depurination.11 After the treatment of 18 with DCA, the acidic reaction mixture was neutralized with TEA or saturated aqueous sodium bicarbonate, and 19 was extracted from the aqueous phase with dichloromethane. Purification of the crude product by silica gel chromatography gave 19 in 85% yield. This result showed that dibenzyl protection of the phosphate group was stable under DMTr-deprotection conditions; additionally, no depurination occurred for either the nucleotides 18 or 19.

Monodeprotection of the dibenzyl phosphate 19 formed 20 in 80% yield. Deprotection of benzyl groups by hydrogenolysis could not be used in this work, because it would reduce the ethynyl linker in the synthesis of nucleotide 1. Therefore, we tested two other literature procedures for deprotection of dibenzyl phosphates. The first procedure used 5 equivalents of TMSBr in anhydrous CH2Cl2, with quenching of the reaction under slightly basic conditions

using a 1 M aqueous solution of ammonium bicarbonate to yield a dideprotected product.13 However, in our reaction with 19 a dark color was produced after evaporation of the reaction solvent, and 1H NMR showed a mixture of unidentified products. These products could have been due to side reactions with HBr generated by silylation of 5’-OH, as a red color developed immediately upon addition of TMSBr. A second reported procedure used stoichiometric amounts of DABCO in refluxing toluene for 2 h to monodeprotect benzylphosphonic acid dibenzyl ester.14 In an attempt to cleave both benzyl groups, 19 was treated with 2.2 equivalents of DABCO in refluxing 1,4-dioxane/toluene (1:1). 1,4-Dioxane was used because 19 was not sufficiently soluble in toluene. 1H NMR for the crude product showed that only monodeprotection of 19 occurred to give 20 as an anion with Bn-quaternarized DABCO as the counter cation. A remaining trace amount of 19 was removed by dissolving the crude product in water and washing it with dichloromethane. This did not remove, however, the excess DABCO.

No side products were observed. Thus, the phosphate group of 20 was shown to be stable to reflux at 101-112 °C. Since 20 was water soluble nucleotide, we applied the reaction steps summarized in Scheme 6.3 to the synthesis of the AQ-dA nucleotide conjugate 1 (in Chapter 2). The conjugate 1 was also water soluble.