Palladium catalysis for the synthesis of hydrophobic C-6 and C-2 aryl 2′-deoxynucleosides. Comparison of C-C versus C-N bond formation as well as C-6 versus C-2 reactivity

DOI: 10.1021/ja0107172

Source and publish data:

Journal of the American Chemical Society p. 7779 - 7787 (2001)

Update date:2022-07-30

Topics:

Authors:

Lakshman

Hilmer Hilmer

Martin Martin

Keeler Keeler

Dinh Dinh

Ngassa Ngassa

Russon Russon

Read Full Text PDF DownLoad Join now for total 90,000,000 free articles

Article abstract of DOI:10.1021/ja0107172

Suzuki-Miyaura cross-coupling of haloaromatic compounds with arylboronic acids provides a simple entry to biaryl systems. Despite its ease, to date, there are no detailed investigations of this procedure for deoxynucleoside modification. As shown in this

Full text of DOI:10.1021/ja0107172

J. Am. Chem. Soc. 2001, 123, 7779-7787

7779

Palladium Catalysis for the Synthesis of Hydrophobic C-6 and C-2

Aryl 2′-Deoxynucleosides. Comparison of C-C versus C-N Bond

Formation as well as C-6 versus C-2 Reactivity

Mahesh K. Lakshman,*^,§,‡John H. Hilmer,^§,‡Jocelyn Q. Martin,^†,‡John C. Keeler,^‡

Yen Q. V. Dinh,^‡Felix N. Ngassa,^‡and Larry M. Russon

Contribution from the Department of Chemistry, City College of CUNY, 138th Street at ConVent AVenue,

New York, New York 10031-9198, Department of Chemistry, UniVersity of North Dakota,

Grand Forks, North Dakota 58202-9024, and Mass Spectrometry SerVices, Department of Chemistry

and Biochemistry, UniVersity of Oklahoma, Norman, Oklahoma 73019-3050

ReceiVed March 19, 2001

Abstract: Suzuki-Miyaura cross-coupling of haloaromatic compounds with arylboronic acids provides a simple

entry to biaryl systems. Despite its ease, to date, there are no detailed investigations of this procedure for

deoxynucleoside modification. As shown in this study, a wide variety of C-6 arylpurine 2′-deoxyriboside (C-6

aryl 2′-deoxynebularine analogues) and C-2 aryl 2′-deoxyinosine analogues can be conveniently prepared via

the Pd-mediated cross-coupling of arylboronic acids with the C-6 halonucleosides, 6-bromo- or 6-chloro-9[2-

deoxy-3,5-bis-O-(tert-butyldimethylsilyl)-â-D-erythro-pentofuranosyl]purine (1 and 2), and the C-2 halonucleo-

side, 2-bromo-O⁶-benzyl-3′,5′-bis-O-(tert-butyldimethylsilyl)-2′-deoxyinosine (3). Although bromonucleoside

1 proved to be a good substrate for the Pd-catalyzed Suzuki-Miyaura cross-couplings, we have noted that for

several C-6 arylations, the chloronucleoside 2 provides superior coupling yields. Also described in this study

is a detailed evaluation of catalytic systems that led to optimal product recoveries. Finally, a comparison of

the C-C and C-N bond-forming reactions of deoxynucleosides is also reported. On the basis of this comparison,

we provide evidence that C-N bond formation at the C-6 position, leading to N-aryl 2′-deoxyadenosine

analogues, is more sensitive to the ligand used, whereas C-C bond-forming reactions at the same position are

not. In contrast to the ligand dependency exhibited in C-N bond formation at the C-6 position, comparable

reactions at the C-2 position of purine deoxynucleosides proceed with less sensitivity to the ligand used.

Introduction

possess π-stacking ability. Thus, DNA containing such nucleo-

sides could exhibit unusual properties. For instance, in unrelated

studies, we have seen that covalent modification of the N⁶amino

group of 2′-deoxyadenosine with carcinogenic metabolites of

the multiple aryl ring-containing benzo[a]pyrene leads to an

increase in thermal stability of a short N-ras duplex in the

presence of an apyrimidinic site opposite the adducted adenine.⁷

Further, in these cases, the pyrene moiety seemed to undergo

intercalative accommodation within the duplex.⁷Similar effects

have also been seen with unusual pyrene nucleosides.⁸In

addition, aryl nucleosides could have potential pharmacological

activity, and they could be used in the development of new

molecular assembly motifs. Therefore, we became interested

in probing questions on the applications of hydrophobic

nucleosides, as well as studies on DNA containing these unusual

nucleosides. As a first step in our studies it is therefore necessary

to delineate a general chemical method for the preparation of

aryl nucleosides.

Synthetic access to unusual nucleosides is becoming markedly

important for a variety of reasons. For example, unusual

deoxyribosides have recently been used to probe enzymatic

activities such as recognition by polymerases, base-pairing

properties, and repair.^1-4In addition, unnatural nucleosides have

been evaluated for fidelity in replicative processes with a view

to expansion of the genetic alphabet.^5,6Purine deoxynucleosides

bearing aryl rings at the C-6 and C-2 positions are interesting

for several reasons. Such nucleosides, while maintaining the

basic purine skeleton that is potentially recognizable by intra-

cellular systems, also bear hydrophobic aromatic moieties that

* Corresponding author. Phone: (212) 650-7835. Fax: (212) 650-6107.

E-mail: lakshman@sci.ccny.cuny.edu.

^§City College of CUNY.

^‡University of North Dakota.

^†Undergraduate research participant.

University of Oklahoma.

Although several methods have been described for C-C bond

formation among purines and to an extent on nucleosides,^9-13

surprisingly, the application of the Suzuki-Miyaura cross-

coupling¹⁴for purine and nucleoside modification has only

recently received attention,¹⁵with the most recent reports

appearing while this work was in progress.^15b-dThis cross-

(1) Morales, J. C.; Kool, E. T. J. Am. Chem. Soc. 1999, 121, 2323-

2324.

(2) Kool, E. T.; Morales, J. C.; Guckian, K. M. Angew. Chem., Int. Ed.

2000, 39, 990-1009.

(3) Morales, J. C.; Kool, E. T. J. Am. Chem. Soc. 2000, 122, 1001-

1007.

(4) Chepanoske, C. L.; Langelier, C. R.; Chmiel, N. H.; David, S. S.

Org. Lett. 2000, 2, 1341-1344.

(5) McMinn, D. L.; Ogawa, A. K.; Wu, Y.; Liu, J.; Schultz, P. G.;

Romesberg, F. E. J. Am. Chem. Soc. 1999, 121, 11585-11586.

(6) Ogawa, A. K.; Wu, Y.; McMinn, D. L.; Liu, J.; Schultz, P. G.;

Romesberg, F. E. J. Am. Chem. Soc. 2000, 122, 3274-3287.

(7) Chaturvedi, S.; Lakshman, M. K. Carcinogenesis 1996, 17, 2747-

2752.

(8) Guckian, K. M.; Schweitzer, B. A.; Ren, R. X.-F.; Sheils, C. J.;

Tahmassebi, D. C.; Kool, E. T. J. Am. Chem. Soc. 2000, 122, 2213-2222.

Published on Web 07/11/2001

7780 J. Am. Chem. Soc., Vol. 123, No. 32, 2001

Lakshman et al.

coupling method offers distinct advantages over others because

of its operational simplicity as well as the broad range of

compounds that can be prepared on the basis of the ready

availability (or easy synthesis) of a wide assortment of arylbo-

ronic acids. Among nucleosides, aryl group introduction at the

C-6 position of purine ribosides via the Suzuki-Miyaura

reaction has received much more attention in the recent

past,^15a,b,dwhereas far fewer examples are available on similar

reactions with the more labile 2′-deoxyribosides.^15cIn addition,

to our knowledge, arylation at the C-2 position of 2′-deoxy-

nucleosides has not been explored to date. Thus, details of the

Suzuki-Miyaura cross-coupling remain largely unstudied for

deoxynucleoside modification particularly with respect to

understanding the catalytic systems necessary as well as the

generality of the approach. On the basis of recent developments

in C-C bond-formation methods,^16-21as well as our own

interests in Pd-catalyzed methods for nucleoside and, conceiv-

ably, DNA modification, we report herein a facile approach to

C-6 aryl 2′-deoxynebularine and C-2 aryl 2′-deoxyinosine

analogues via the Suzuki-Miyaura protocol. We also compare,

for the first time, the utility of C-6 chloro- and C-6 bromopurine

nucleosides in these reactions. With a view to understanding

Pd-catalyzed cross-coupling as a general entry to modified

nucleosides, we have compared C-C and C-N bond formation

at the C-6 position and also C-N bond formation at the C-6

and C-2 positions. Therefore, this study offers the first, relatively

extensive investigation of palladium catalysis for nucleoside

modification.

Figure 1. Structures of C-6 and C-2 halo 2′-deoxynucleosides that

are suitable substrates for Suzuki-Miyaura cross-coupling reactions.

Results and Discussion

The Suzuki-Miyaura protocol for C-C bond formation,

leading to biaryl compounds, is an operationally simple method

involving a Pd-catalyzed cross-coupling of haloaromatics with

arylboronic acids.¹⁴This method does not necessitate the use

of anhydrous conditions, and a variety of protecting groups are

well-tolerated under the reaction conditions. Application of this

procedure for the introduction of aryl groups into nucleosides

would represent a simple and efficient approach to unusual,

hydrophobic nucleosides. Two possible alternatives can be

envisioned for the arylation of nucleosides through such an

approach: (a) coupling of a halonucleoside with arylboronic

acids or (b) coupling of a nucleoside boronic acid with halo

aromatic compounds. Between the two, approach (a) is more

straightforward for two reasons; syntheses of halonucleosides,

such as those shown in Figure 1, are well-established,^22-24and

a wide assortment of arylboronic acids are commercially

available.

(9) For examples on tin chemistry, please see: (a) Gundersen, L.-L.

Tetrahedron Lett. 1994, 35, 3155-3158. (b) Gundersen, L.-L.; Bakkestuen,

A. K.; Aasen, A. J.; Øvera˚s, H.; Rise, F. Tetrahedron 1994, 50, 9743-

9756. (c) Hocek, M.; Masoj´ıdkova´, M.; Holy´, A. Tetrahedron 1997, 53,

2291-2302. (d) Moriarty, R. M.; Epa, W. R.; Awasthi, A. K. Tetrahedron

Lett. 1990, 31, 5877-5880. (e) Sessler, J. L.; Wang, B.; Harriman, A. J.

Am. Chem. Soc. 1993, 115, 10418-10419. (f) Van Aerschot, A. A.; Mamos,

P.; Weyns, N. J.; Ikeda, S.; De Clercq, E.; Herdewijn, P. A. J. Med. Chem.

1993, 36, 2938-2942.

(10) For an example on aluminum chemistry, please see: Hirota, K.;

Kitade, Y.; Kanbe, Y.; Maki, Y. J. Org. Chem. 1992, 57, 5268-5270.

(11) For an example on copper chemistry, please see: Dvorˇa´kova´, H.;

Dvorˇa´k, D.; Holy´, A. Tetrahedron Lett. 1996, 37, 1285-1288.

(12) For examples on zinc chemistry, please see: (a) Stevenson, T. M.;

Prasad, A. S. B.; Citineni, J. R.; Knochel, P. Tetrahedron Lett. 1996, 37,

8375-8378. (b) Prasad, A. S. B.; Stevenson, T. M.; Citineni, J. R.; Nyzam,

V.; Knochel, P. Tetrahedron 1997, 53, 7237-7254.

Suzuki-Miyaura reactions usually involve the use of a base

such as aqueous carbonate;^14,15ahowever, halonucleosides could

potentially face hydrolysis with an aqueous base. Further, on

the basis of our recent experience with Pd-mediated C-N bond

formation,²²we anticipated the use of the tert-butyldimethylsilyl

(TBDMS) protecting group at the 3′,5′-hydroxyls of the nucleo-

side. The use of this protecting group precludes the use of the

commonly employed CsF²⁵and KF.^17,19,20,25aWe have found

the TBDMS group to be stable under reaction conditions when

employing bases such as K₃PO₄and Cs₂CO₃.²²Whereas K₃-

PO₄has been utilized in Suzuki-Miyaura reactions,^16,17,19Cs₂-

CO₃has been used in Heck couplings²⁶as well as Suzuki-

Miyaura reactions,²¹and both seemed worthy choices. However,

we preferred K₃PO₄on the basis of its compatibility with the

systems we that anticipated investigating.

(13) For examples of free-radical-based arylation, please see: (a) Nair,

V.; Richardson, S. G.; Coffman, R. E. J. Org. Chem. 1982, 47, 4520-

4524. (b) Nair, V.; Young, D. A. J. Org. Chem. 1984, 49, 4340-4344. (c)

McKenzie, T. C.; Epstein, J. W. J. Org. Chem. 1982, 47, 4881-4884.

(14) (a) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457-2483. (b)

Suzuki, A. J. Organomet. Chem. 1999, 576, 147-168. (c) Suzuki, A. In

Metal-Catalyzed Cross-Coupling Reactions; Diederich, F., Stang, P. J., Eds.;

Wiley-VCH: New York, 1998; Chapter 2, pp 49-97. (d) For a general

review, please see: Stanforth, S. P. Tetrahedron 1998, 54, 263-303.

(15) (a) Havelkova´, M.; Hocek, M.; Cˇesnek, M.; Dvorˇa´k, D. Synlett 1999,

1145-1147. (b) Hocek, M.; Holy´, A.; Votruba, I.; Dvorˇa´kova´, H. J. Med.

Chem. 2000, 43, 1817-1825. (c) Hocek, M.; Holy´, A.; Votruba, I.;

Dvorˇa´kova´, H. Collect. Czech. Chem. Commun. 2000, 65, 1683-1697. (d)

Hocek, M.; Holy´, A.; Votruba, I.; Dvoˇra´kova´, H. Collect. Czech. Chem.

Commun. 2001, 66, 483-499.

Synthesis of C-6 Aryl 2′-Deoxynebularine Derivatives.

(a) Utility of Bromonucleoside 1 for Suzuki-Miyaura Cross-

Coupling. The halonucleoside presursor, 6-bromo-9[2-deoxy-

3,5-bis-O-(tert-butyldimethylsilyl)-â-D-erythro-pentofuranosyl]-

(22) Lakshman, M. K.; Keeler, J. C.; Hilmer, J. H.; Martin, J. Q. J. Am.

Chem. Soc. 1999, 121, 6090-6091. In the Supporting Information for this

paper, the reported use of CBr₄for the synthesis of the bromonucleoside 1

is an error. The correct reagent for this synthesis is CHBr₃.

(23) (a) Synthesis of 6-chloro-9(2-deoxy-â-^D-erythro-pentofuranosyl)-

purine: Robins, M. J.; Basom, G. L. Can. J. Chem. 1973, 51, 3161-3169.

(b) Synthesis of the corresponding bis TBDMS ether, 6-chloro-9[2-deoxy-

3,5-bis-O-(tert-butyldimethylsilyl)-â-^D-erythro-pentofuranosyl]purine (2):

Maruenda, H.; Chenna, A.; Liem, L.-K.; Singer, B. J. Org. Chem. 1998,

63, 4385-4389. Compound 2 has been reported as an oil, but we have

isolated this material as a colorless solid.

(24) (a) Harwood, E. A.; Sigurdsson, S. T.; Edfeldt, N. B. F.; Reid, B.

R.; Hopkins, P. B. J. Am. Chem. Soc. 1999, 121, 5081-5082. (b) Harwood,

E. A.; Hopkins, P. B.; Sigurdsson, S. T. J. Org. Chem. 2000, 65, 2959-

2964.

(25) (a) The use of CsF for Suzuki-Miyaura reactions was initially

reported by Wright, S. W.; Hageman, D. L.; McClure, L. D. J. Org. Chem.

1994, 59, 6095-6097. (b) For some examples, please see refs 16 and 18.

(26) (a) Littke, A. F.; Fu, G. C. J. Org. Chem. 1999, 64, 10-11. (b)

Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 1999, 38, 2411-2413.

(16) Old, D. W.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem. Soc. 1998,

120, 9722-9723.

(17) Wolfe, J. P.; Singer, R. A.; Yang, B. H.; Buchwald, S. L. J. Am.

Chem. Soc. 1999, 121, 9550-9561.

(18) Bei, X.; Crevier, T.; Guram, A. S.; Jandeleit, B.; Powers, T. S.;

Turner, H. W.; Uno, T.; Weinberg, W. H. Tetrahedron Lett. 1999, 40, 3855-

3858.

(19) Wolfe, J. P.; Buchwald, S. L. Angew. Chem., Int. Ed. 1999, 38,

2413-2416.

(20) Littke, A. F.; Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2000, 122, 4020-

4028.

(21) (a) Zhang, C.; Huang, J.; Trudell, M. L.; Nolan, S. P. J. Org. Chem.

1999, 64, 3804-3805. (b) Zhang, C.; Trudell, M. L. Tetrahedron Lett. 2000,

41, 595-598.

Synthesis of Hydrophobic C-6 and C-2

J. Am. Chem. Soc., Vol. 123, No. 32, 2001 7781

Table 1. Evaluation of Catalytic Systems for the Coupling of Bromonucleoside 1 with Phenylboronic Acid^a,b

entry

catalyst/ligand

solvent

temp, °C

time, h; result

Pd(OAc)₂/L-1

Pd(OAc)₂/L-2

Pd(OAc)₂/L-3

Pd(OAc)₂/L-4

Pd(OAc)₂/L-5

Pd(OAc)₂/L-6

Pd(OAc)₂/L-3

1,4-dioxane

1,2-DME

100

12, 40%

1.75, 73%

6, 67%

1, 91%

reaction ∼50% complete after ∼48 h^c

after ∼12 h, product formation ceased;

degradation after ∼36 h^d

0.5, 80%

Pd₂(dba)₃/L-2

Pd(PPh₃)₄

1,4-dioxane

toluene

100

8, 87%^e

^aReaction conditions (entries 1-8): bromonucleoside (1) 0.15 M in anhydrous 1,4-dioxane (or 1,2-DME for entry 7), 1.5 molar equiv phenylboronic

acid, 2 molar equiv K₃PO₄, 10 mol % Pd(OAc)₂, 15 mol % L, 100 °C (80 °C for entry 7). ^b% yields refer to isolated and purified products.

^cProduct was not isolated, but extent of reaction was estimated by TLC. ^dHeating beyond 12 h seemed to produce byproducts, as visualized on

TLC; product not isolated. ^eReaction conditions: bromonucleoside (1) 0.05 M in anhydrous toluene, 1.5 molar equiv phenylboronic acid, 1.25

molar equiv powdered anhydrous K₂CO₃, 25 mol % Pd(PPh₃)₄.

From these data, several points are worth specific mention.

In every case, the ratio of phosphine:Pd was 1.5:1. Thus,

formation of bis-coordinated, 1:1 Pd(0)-ligand complexes

would be a significant likelihood only in the case of ligands

L-1, L-2, and L-3 through intramolecular P,P-Pd or P,N-Pd

linkages. In contrast, ligands L-4, L-5, and L-6 are mono-

coordinating, and any bis-coordinated Pd(0) intermediate would

arise as a consequence of two phosphine ligands per Pd. Evident

from Table 1 is the fact that the combination of L-4/Pd(OAc)₂/

K₃PO₄(entry 4) was superior to all others tested, resulting in a

91% isolated yield of 4a. The (Ph₃P)₄Pd/anhydrous K₂CO₃

system (entry 9), which is similar to that recently reported in

Suzuki-Miyaura reactions of nucleosides, provided approxi-

mately comparable results, but required a longer reaction

Figure 2. Structures of commercially available or readily synthesized

phosphine ligands that are potentially useful for C-C and C-N bond-

forming reactions of nucleosides.

time.^15a,30Interestingly, L-5, which is similar to but sterically

bulkier than L-4, was relatively ineffective, with only ∼50%

product formation after 48 h (entry 5). Similar results (∼50%

purine (1), can be conveniently synthesized on the multigram

product formation after 48 h) were obtained with (t-Bu)₃P (L-

scale by a diazotization-bromination.²²In the recent literature,

6, entry 6), a ligand that has otherwise found admirable use in

two palladium reagents have been successfully used for Suzuki-

simpler systems.^20,26We have recently shown that ligand L-3,

Miyaura reactions on simple systems; these are Pd(OAc)₂and

which differs from L-4 by only the dimethylamino moiety, was

Pd₂(dba)₃. On the basis of the relative costs of the two reagents

highly effective for C-N bond formation at the C-6 position

as well as the results of a recent study²⁷that suggested the

of deoxynucleosides.²²However, in the present case, L-3 proved

superiority of Pd(OAc)₂over Pd₂(dba)₃, the former was chosen

to be less effective as compared to L-4 in both time for the

for detailed investigation. In addition, the recently described

reaction to attain completion and the ensuing product yield (entry

beneficial influence of the acetate counter-ion in Pd-catalyzed

3). In addition, in the case of L-3, when the higher-boiling 1,4-

Heck reactions as well as other cross-couplings further justified

dioxane was replaced with 1,2-DME (1,2-dimethoxyethane, the

the choice of Pd(OAc)₂.²⁸In contrast to the limited number of

solvent used in C-N bond-forming reactions), the reaction did

Pd reagents, a large assortment of ligands is available (Figure

not progress to completion, and degradation seemed to set in

2).²⁹Therefore, choice of the catalytic system posed a significant

upon prolonged heating. The system L-2/Pd(OAc)₂/K₃PO₄was

number of possibilities. To simplify the identification of a

also quite effective, with a relatively short reaction time, but

generally applicable catalytic system, we chose to perform the

produced a lower product yield (entry 2) than the L-4/Pd(OAc)₂/

initial optimization experiments using 1 and phenylboronic acid.

K₃PO₄system. Use of Pd₂(dba)₃in combination with L-2 and

Table 1 lists the results of these initial experiments.

K₃PO₄provided what appeared to be a fast reaction with a

slightly lower isolated yield of 4a (80%, entry 8).

(27) Wolfe, J. P.; Buchwald, S. L. J. Org. Chem. 2000, 65, 1144-1157.

(28) Amatore, C.; Jutand, A. Acc. Chem. Res. 2000, 33, 314-321.

(29) Ligands L-1 to L-6 are commercially available; however, at the

time this study was conducted, 2-(dicyclohexylphosphino)-2′-(N,N-dimeth-

ylamino)-1,1′-biphenyl (L-3) was not commercially available. This com-

pound was synthesized through a minor modification of the procedure

described in ref 16 (best overall yield, 20% over 4 steps). A shorter synthesis

of L-3 has recently been reported: Tomori, H.; Fox, J. M.; Buchwald, S.

L. J. Org. Chem. 2000, 65, 5334-5341.

On the basis of the optimization experiments carried out using

unsubstituted phenylboronic acid, the next step was the utiliza-

(30) In the studies described in ref 15, anhydrous K₂CO₃has been used

for the cross-coupling of hydrolytically labile compounds. On the other

hand, the reactions were observed to be faster with aqueous K₂CO₃(ref

15a). To avoid complications, such as hydrolysis of the bromonucleoside

or protecting group degradation, we employed anhydrous K₂CO₃.

7782 J. Am. Chem. Soc., Vol. 123, No. 32, 2001

Lakshman et al.

Table 2. Synthesis of C-6 Aryl 2′-Deoxynebularine Derivatives

4a-g Using Bromonucleoside 1^a

Figure 3. Two possible structures of compound 4d with coordinated

Pd.

entry

arylboronic acid

time, h^bproduct %yield^c

phenylboronic acid

1.5

19.5

1.5

8.5

nucleoside complexes of Pd(II). Pd(II), a class B or “soft” metal,

prefers coordination with nitrogen donor atoms (N¹and N⁷) of

the purine nucleobases,^31,32and purine nucleosides in their basic

form have been shown to bind Pd(II) at both sites, with a

preference for N¹-binding.³³Thus, it is plausible that 4d could

compete with the ligand for coordination to Pd utilizing the O

on the aryl moiety and either the N¹or N⁷of the purine (Figure

3). This is similar to the belief that formation of bis indolyl

Pd(II) complexes is responsible for low yields in some C-N

bond-forming reactions of indoles.³⁴Support for our rationale

is also obtained from the fact that 2-tolylboronic acid, in contrast

to 2-ethoxyphenylboronic acid, couples rather smoothly with a

6-chloropurine ribonucleoside precursor.^15bOn the other hand,

synthesis of biphenyls with more than one ortho substituent is

generally problematic,¹⁷and it is conceivable that the prevailing

reasons are also the causes of the anomalous reaction with

2-ethoxyphenylboronic acid.

In contrast to 4d, placement of the alkoxy groups at the para

or meta positions on the aryl ring dispenses with the bis

coordinating ability of the products, because the alkoxy groups

are more remote from the N⁷and N¹sites. Alternatively, any

steric conflicts are also diminished. Correspondingly, reactions

leading to the 4- and 3-methoxyphenyl nucleosides (4b and 4c)

are rapid (1-1.5 h), and good product yields are obtained with

1.5 molar equiv of the respective boronic acids (Table 2, entries

2 and 3).

(b) Utility of Chloronucleoside 2 for Suzuki-Miyaura

Cross-Coupling. With the experiments described above, the

generality of Pd catalysis for the synthesis of C-6 arylated purine

2′-deoxyribosides from a bromonucleoside precursor had been

established, as well as an understanding of the optimal catalytic

system for the process. However, the fact that chloropurine

derivatives were efficient substrates in Suzuki-Miyaura reac-

tions was interesting to us.¹⁵Ni-promoted cross-coupling of

boronic acids with aryl chlorides has been quite useful for C-C

bond formation in relatively simple systems,³⁵and a modestly

yielding Ni-catalyzed synthesis of C-6 phenyl nebularine has

been reported.³⁶On the other hand, highly effective Pd catalysts

for Suzuki-Miyaura reactions involving chloroaromatics that

provide consistent results have only recently been identified.^16-21,37

On the basis of these reasons, an evaluation of the utility of 2

4-methoxyphenylboronic acid

3-methoxyphenylboronic acid

2-ethoxyphenylboronic acid^d

3-nitrophenylboronic acid

4-acetylphenylboronic acid

3-thiopheneboronic acid

^aReaction conditions: bromonucleoside (1) 0.15 M in anhydrous

1,4-dioxane, 1.5 molar equiv arylboronic acid, 2 molar equiv K₃PO₄,

10 mol %, Pd(OAc)₂, 15 mol % L-4, 100 °C. ^bReactions were

monitored by TLC for complete disappearance of 1 and are approximate

values. ^cIsolated and purified products. ^dIn this case, 3 molar equiv

of 2-ethoxyphenylboronic acid was used.

tion of the method for the synthesis of an assortment of

arylnucleosides having different substituents on the aryl moiety.

Thus, a selection of arylboronic acids bearing either electron-

withdrawing or electron-donating substituents was made so as

to evaluate the overall utility of the process for the generation

of the modified nucleosides. Such a detailed investigation, to

our knowledge, has not been reported in the literature for

nucleosides. Table 2 shows the phenylboronic acids used and

the yields of the C-6 arylnucleosides 4a-g.

From the data in Table 2, it is apparent that arylboronic acids

with electron-withdrawing groups seem to return lower product

yields, as compared to the remaining substituted boronic acids.

This type of a trend, though not evident in recent reports of

Suzuki-Miyaura cross-couplings in relatively simpler sys-

tems,^17,20has been observed in the arylation at the C-6 position

of a chloropurine precursor.^15a,bAs described later in this paper,

the low-yield problem can be resolved, and substantial yield

improvements can be attained.

The coupling with 2-ethoxyphenylboronic acid presented an

interesting case. In the reaction of this arylboronic acid, use of

10 mol % Pd(OAc)₂and 15 mol % L-4 resulted in incomplete

reaction even after 67 h, and the product yield from this reaction

was 31%. Replacement of L-4 with L-3 under otherwise

identical conditions also resulted in incomplete reaction after

77 h. In an attempt to achieve complete conversion in a

reasonable period and in better yield, we rationalized increasing

the concentration of the catalytic system. When 30 mol % Pd-

(OAc)₂and 45 mol % L-4 were employed, the reaction was

complete within 20 h; however, the isolated product yield was

only 13%. Since this result was somewhat unexpected, a second

alternative was attempted. Increasing the molar equivalence of

2-ethoxyphenylboronic acid from 1.5 to 3 resulted in a complete

reaction within 20 h with a 62% isolated yield of 4d. Although

we cannot be certain about the reasons for this behavior, it is

conceivable that complexes such as those shown in Figure 3

are responsible for the observed results. Such complexes could

potentially contribute to degradation of the catalytic system,

leading to termination of the catalytic cycle, causing product

degradation by processes such as deglycosylation, or both, etc.,

leading to lowered yields. In the absence of a direct precedent

for the formation of these complexes in the Suzuki-Miyaura

reaction, some understanding can be garnered from studies on

(31) Kazakov, S. A. In Bioorganic Chemistry: Nucleic Acids; Hecht, S.

M., Ed.; Oxford University Press: New York, 1996; Chapter 9, pp 244-

287.

(32) Although N³could also be a potential coordination site, steric

hindrance by the sugar is thought to preclude coordination here.³¹

(33) Martin, R. B. In Metal Ions in Biological Systems; Sigel, A., Sigel,

H., Eds.; Marcel Dekker: New York, 1996; Vol. 32, Chapter 3, pp 61-89.

(34) (a) Paul, F.; Patt, J.; Hartwig, J. F. Organometallics 1995, 14, 3030-

3039. (b) Old, D. W.; Harris, M. C.; Buchwald, S. L. Org. Lett. 2000, 2,

1403-1406.

(35) For examples, please see: (a) Saito, S.; Oh-tani, S.; Miyaura, N. J.

Org. Chem. 1997, 62, 8024-8030. (b) Lipshutz, B. H.; Sclafani, J. A.;

Blomgren, P. A. Tetrahedron 2000, 56, 2139-2144. (c) Galland, J.-C.;

Savignac, M.; Geneˆt, J.-P. Tetrahedron Lett. 1999, 40, 2323-2326.

(36) Bergstrom, D. E.; Reddy, P. A. Tetrahedron Lett. 1982, 23, 4191-

4194.

Synthesis of Hydrophobic C-6 and C-2

J. Am. Chem. Soc., Vol. 123, No. 32, 2001 7783

Table 3. Synthesis of Five C-6 Aryl 2′-Deoxynebularine

Derivatives Using Chloronucleoside 2^a

Scheme 1

entry

arylboronic acid

time, h^bproduct %yield^c

phenylboronic acid

1.5

4-methoxyphenylboronic acid

3-nitrophenylboronic acid

4-acetylphenylboronic acid

3-thiopheneboronic acid

^aReaction conditions: chloronucleoside (2) 0.15 M in anhydrous

1,4-dioxane, 1.5 molar equiv arylboronic acid, 2 molar equiv K₃PO₄,

10 mol % Pd(OAc)₂, 15 mol % L-4, 100 °C. ^bReactions were

monitored by TLC for complete disappearance of 2 and are approximate

values. ^cIsolated and purified products.

gratifying cross-coupling with 2 (84% yield), although a modest

49% yield of 4f was realized from 1, as well. Thus, our results

demonstrate that arylboronic acids with electron-withdrawing

substituents not only couple with both halo precursors 1 and 2,

but they do so rapidly and efficiently when chloronucleoside 2

is utilized.

From the foregoing experiments, the effectiveness of Pd

catalysis for the synthesis of C-6 aryl 2′-deoxynebularines is

clear. In addition, both halonucleosides 1 and 2 undergo

Suzuki-Miyaura cross-coupling reactions, with the caveat that

the chloro precursors may be more suitable in some cases.

Introduction of aryl moieties at the C-2 position of nucleosides

became our next consideration.

(prepared in three steps from 2′-deoxyinosine²³) in the Suzuki-

Miyaura coupling reactions was warranted.

Again, as an initial test, 2 was subjected to cross-coupling

with phenylboronic acid using the optimized catalytic system

[Pd(OAc)₂, L-4, K₃PO₄] under the conditions described above

(1,4-dioxane, 100 °C). However, one change was incorporated

into the procedure, which was premixing the Pd(OAc)₂and L-4.

This was done largely because of its reported beneficial effect

on C-N bond formation,²⁷as well as the possibly lower

reactivity of chloroaromatic 2, as compared to the bromoaro-

matic 1, under Pd-catalyzed conditions (it should be noted that

2 possesses higher reactivity under S_NAr displacement condi-

tions). The reaction of 2 with phenylboronic acid proved to be

very facile, attaining completion within 1.5 h with a 93%

isolated yield of 4a. On the basis of this result, we decided to

investigate the reaction of 2 with other boronic acids, particularly

those that provided lower yields in their reaction with 1. Table

3 shows the results that were obtained in the cross-coupling of

four substituted boronic acids with 2.

Clear from Table 3 is the fact that C-6 aryl nucleosides can

be obtained by Suzuki-Miyaura reactions of the chloronucleo-

side 2 with arylboronic acids. Also indicated by the results, is

that in many cases, the product yields from this precursor may,

in fact, be superior to those obtained by utilizing the bromo-

nucleoside 1. To evaluate whether premixing Pd(OAc)₂and L-4

had any effect on the rate or yield of a reaction involving

bromonucleoside 1, a single cross-coupling reaction of 1 with

3-nitrophenylboronic acid was attempted. With the premixed

catalyst, the reaction was complete within ∼1 h but the isolated

yield of 4e was 56% (which is comparable to the 59% obtained

without premixing), indicating no particular advantage to

premixing the catalyst when 1 was employed for these reactions.

A point that is significant to note is that our results with

chloronucleoside 2 contrast with recent observations, which

report that arylboronic acids with electron-withdrawing groups

do not provide satisfactory yields of the cross-coupled products.^15b

In one study, the yield of 6-(3-nitrophenyl)-9-benzylpurine from

a cross-coupling involving 3-nitrophenyboronic acid has been

reported as 19% (in 48 h under nonaqueous conditions) and

66% (in 7 h using aqueous K₂CO₃).^15aIn our case, the yield of

4e was 84% (reaction complete within 1.5 h under nonaqueous

conditions). Further, 4-acetylphenylboronic acid provided a

Synthesis of C-2 Aryl 2′-Deoxyinosine Derivatives. Al-

though Pd-catalyzed C-C bond-forming reactions at the 2-posi-

tion of an O⁶-protected inosine (a ribonucleoside) have been

carried out using a C-2 iodo precursor, these have been largely

Heck and Stille cross-couplings.³⁸To our knowledge, there are

no studies utilizing the Suzuki-Miyaura reaction, and more

importantly, there are no reports of C-C bond formation using

Pd catalysis on the more labile 2′-deoxyribonucleosides.

A simple reaction protocol for the synthesis of C-2 aryl 2′-

deoxyinosine analogues is shown in Scheme 1. The synthesis

of 3 from 2′-deoxyguanosine has been reported in the literature,²⁴

and it is easily prepared on the multigram scale. As in the case

of C-6 modification, initial experimentation involved the Pd-

catalyzed cross-coupling of 3 with phenylboronic acid using

conditions that had been optimized for the generation of the

C-6 aryl 2′-deoxynebularine derivatives. Under the cross-

coupling conditions, the C-2 phenyl O⁶-benzyl-2′-deoxyinosine

derivative 5a was obtained in 87% yield within 1 h. The cross-

coupling of 3 and phenylboronic acid was briefly investigated

using another catalytic system; Pd(OAc)₂/L-1/Cs₂CO₃(0.1, 0.15,

and 2 molar equiv, respectively). Under these conditions,

although 5a was obtained in 84% yield, the reaction time appears

to be longer (∼10 h, unoptimized). Given the overall superiority

and generality of the Pd(OAc)₂/L-4/K₃PO₄system for C-C

bond formation at both the C-6 and the C-2 positions, further

reactions on 3 were performed with this combination. Because

the final step in the synthesis of 2-aryl 2′-deoxyinosine

derivatives involves a catalytic hydrogenolysis of the benzyl

group, boronic acids in which the substituents would not pose

problems in the reduction step were selected. However, reactions

with boronic acids bearing electron-donating and electron-

withdrawing groups were studied in order to assess the ease

and generality of the reaction. Table 4 shows the yields for the

(38) (a) Nair, V.; Turner, G. A.; Chamberlain, S. D. J. Am. Chem. Soc.

1987, 109, 7223-7224. (b) Nair, V.; Turner, G. A.; Buenger, G. S.;

Chamberlain, S. D. J. Org. Chem. 1988, 53, 3051-3057.

(37) Bei, X.; Turner, H. W.; Weinberg, W. H.; Guram, A. S.; Peterson,

J. L. J. Org. Chem. 1999, 64, 6797-6803.

7784 J. Am. Chem. Soc., Vol. 123, No. 32, 2001

Lakshman et al.

Table 4. Two-Step Synthesis of C-2 Aryl 2′-Deoxinosine

Derivatives 6a-d^a,b,c

Table 5. Results of C-N Bond Formation at the C-6 Position with

Various Catalytic Systems^a

time step 1: compd, step 2: compd,

entry

arylboronic acid

h^d

% yield^e

% yield^e,f

phenylboronic acid

4-methoxyphenylboronic acid 6

3-methoxyphenylboronic acid 4

5a, 87

5b, 67

5c, 78

5d, 80

6a, 84

6b, 86

6c, 93

6d, 70

4-acetylphenylboronic acid

1.25

^aStep 1, cross-coupling. ^bStep 2, removal of the O⁶-benzyl group.

^cReaction conditions for step 1: bromonucleoside (3) 0.13 M in

anhydrous 1,4-dioxane, 1.5 molar equiv arylboronic acid, 2 molar equiv

K₃PO₄, 10 mol % Pd(OAc)₂, 15 mol % L-4, 100 °C. ^dStep 1 reactions

were monitored by TLC for complete disappearance of 3 and are

approximate values. ^eIsolated and purified products. ^fReaction condi-

tions for step 2: products 5a-d were reduced with 5% Pd-C in 1:1

THF-MeOH at room temperature and 1 atm of H₂.

Pd species

(mol %)

ligand

(mol %) base

T, time,

entry

solvent °C h^byield^c

Pd₂(dba)₃(10) L-1 (30) Cs₂CO₃PhMe

Pd₂(dba)₃(10) L-1 (20) Cs₂CO₃PhMe

Pd₂(dba)₃(5) L-1 (10) Cs₂CO₃PhMe

Pd₂(dba)₃(10) L-1 (30) K₃PO₄1,2-DME 80 24

Pd(OAc)₂(5) L-1 (10) Cs₂CO₃PhMe

100 63

100 2.5 54

100

70 12

Pd₂(dba)₃(10) L-2 (30) Cs₂CO₃PhMe

100

cross-coupling (step 1, leading to 5a-d) as well as for the

hydrogenolysis (step 2, leading to 6a-d). In each case, step 1

proceeded smoothly, in good yield, and step 2 yielded the final

products uneventfully.

Pd₂(dba)₃(10) L-2 (30) K₃PO₄1,2-DME 80

Pd₂(dba)₃(10) L-3 (30) K₃PO₄1,2-DME 80

Pd₂(dba)₃(5) L-3 (15) K₃PO₄1,2-DME 70

10 Pd₂(dba)₃(10) L-4 (30) K₃PO₄1,2-DME 80 19

11 Pd₂(dba)₃(10) L-4 (12) K₃PO₄1,2-DME 80 20

12 Pd₂(dba)₃(10) L-5 (30) K₃PO₄1,2-DME 80 20

Inc.

C-C versus C-N Bond Formation at the C-6 Position of

Purine Deoxynucleosides. In this report, we have clearly

demonstrated the facile manner in which Suzuki-Miyaura

reactions can be accomplished at the C-6 and C-2 positions of

purine 2′-deoxynucleosides under anhydrous conditions. In

contrast to C-C bond formation, C-N bond formation at the

C-6 and C-2 positions of purine deoxynucleosides has been

investigated to a greater extent in the recent past by us²²and

others.^24,39It would, therefore, be interesting and instructive to

compare the catalytic systems involved in successful C-C and

C-N bond formations with a view to understanding the

requirements imposed on the catalytic systems in each case.

Quite clearly evidenced in this study, a variety of ligand-Pd

complexes are effective in the cross-coupling of 1 with

phenylboronic acid in yields ranging from 67 to 91%. In contrast

to this, efficient C-N bond formation at the C-6 position appears

to depend much more on the catalytic system that is used.

Significantly, our studies show that reactions involving the bis

coordinating ligands L-1, L-2, and L-3 generally produce faster

reactions, as compared to the mono-coordinating ligands L-4

and L-5 (compare, for example, entries 1-3 as well as 6-8 to

entries 10 and 12 in Table 5). Whereas the mono-coordinating

L-4 led to a complete reaction in 20 h, use of L-5, which has

t-Bu instead of cyclohexyl residues on the phosphorus, resulted

in incomplete conversion within the same duration (entries 10

and 12 in Table 5). Catalytic systems L-1/Pd₂(dba)₃/Cs₂CO₃,

L-2/Pd₂(dba)₃/Cs₂CO₃, and L-3/Pd₂(dba)₃/K₃PO₄provided rea-

sonably good yields (62-69%); however, the use of L-1 and

L-2 resulted in products that were somewhat more darkly

^aReactions were performed essentially as described in ref 22.

^bReactions were monitored by TLC for complete disappearance of 1

and are approximate values. ^cIsolated and purified products.

Scheme 2

These reactions also indicate that bis-coordinating ligands are

generally significantly better in effecting the arylamination of

1, and this is a striking contrast to the Suzuki-Miyaura cross-

coupling reactions involving 1.

C-C versus C-N Bond Formation at the C-2 Position of

Purine Deoxynucleosides. General methods for C-N bond

formation at the C-2 position of purine deoxynucleosides

utilizing 3 for the cross-couplings have been reported in the

literature.^24,39cThe report by Hopkins and co-workers describes

the use of Pd₂(dba)₃/t-BuONa, whereas Johnson and colleagues

have utilized the Pd(OAc)₂/Cs₂CO₃combination;^24,39chowever,

both reports describe reactions involving either (()- or (+)-

L-1 as the ligand. Yields for the arylamination in both studies

are good, although only one arylamination is reported in ref

39c.

colored than that produced with L-3, although the H NMR

A particularly important question that arose was whether a

high ligand dependence would be observed in the arylamination

of 3, as was seen in the case of 1, and for the reasons explained

later, our initial hypothesis was that this would not be the case.

Given the known success of the Pd₂(dba)₃/t-BuONa/L-1 as well

as the Pd(OAc)₂/Cs₂CO₃/L-1 system,^24,39cwe decided to

investigate only the efficiency of the bis-coordinating ligand

L-3 to that of the mono-coordinating ligand L-4, again using

p-toluidine as a model amine for this purpose (Scheme 2).

Therefore, two parallel C-N bond-forming reactions were

conducted using conditions described for the amination of 1,

where one reaction involved use of L-3 and the other, L-4.

Consistent with our expectation, both reactions were complete

quite rapidly (within 3 h), and the isolated yields of 8 were

also comparable (75% using L-3 and 65% using L-4). Thus, it

is quite clear that C-N bond-forming reactions at the C-2

spectra of the products obtained in each case indicated no

impurities to account for the coloration. Use of K₃PO₄in

conjunction with L-1 or L-2 in 1,2-DME prolonged the reaction

time, leading to 7 in only 31-33% yield. Changing the 1:3 ratio

of Pd₂(dba)₃to ligand or lowering catalyst loading in reactions

involving L-1 and L-3 led to complete reaction rapidly, albeit

in lower yield. It is noteworthy, however, that 5 mol % Pd₂-

(dba)₃/15 mol % L-3 resulted in an ∼50% yield of 7, as

compared to the 69% yield obtained with 10 mol % Pd₂(dba)₃/

30 mol % L-3, within 3 h. Thus, in our experience a 10 mol %

Pd₂(dba)₃/30 mol % L-3 combination is optimal for the synthesis

of N⁶-aryl 2′-deoxyadenosine analogues using K₃PO₄as base.

(39) (a) De Riccardis, F.; Bonala, R. R.; Johnson, F. J. Am. Chem. Soc.

1999, 121, 10453-10460. (b) De Riccardis, F.; Johnson, F. Org. Lett. 2000,

2, 293-295. (c) Bonala, R. R.; Shishkina, I. G.; Johnson, F. Tetrahedron

Lett. 2000, 41, 7281-7284.

Synthesis of Hydrophobic C-6 and C-2

J. Am. Chem. Soc., Vol. 123, No. 32, 2001 7785

sides via the Suzuki-Miyaura protocol has been demonstrated.

Whereas the C-6 bromonucleoside 1 is a convenient starting

point in most cases, providing good yields of arylated products,

the C-6 chloronucleoside 2 proved to be superior as compared

with 1 in certain cross-coupling reactions. Thus, both C-6

halonucleosides can be used in a complementary manner for

C-C bond-forming reactions. In reactions involving modifica-

tion at the C-2 position, four boronic acids were studied, and

in each case, good cross-coupling yields were obtained using

the O⁶-benzyl-2-bromo-2′-deoxyinosine derivative 3; however,

the boronic acids used in the Suzuki-Miyaura reactions with 3

were limited to those in which the substituent would not cause

complications during the hydrogenolytic deprotection of the O⁶.

A remedy to this potential problem with some boronic acids is

the use of a different protecting group at the O⁶. Examples

include the 2-cyanoethyl and the 4-nitrophenylethyl groups, both

of which can be readily cleaved under basic conditions.⁴¹

Comparison of the C-C and C-N bond-forming reactions

at C-6 and C-2 of purine deoxynucleosides provides some

interesting results. C-C bond formation at the C-6 position is

relatively independent of the ligand used, and with the exception

of ligands containing tert-butyl groups (L-5 and L-6), both

mono- and bis-coordinating ligands provide complete reactions

with good to high yields. In contrast, C-N reactions at the C-6

position are more ligand-dependent, with bis-coordinating

ligands being effective, among which L-3 is the best in our

experience. In comparison to C-N bond formation at the C-6

position, similar reactions at the C-2 position are less sensitive

to the ligand used, with both the mono-coordinating L-4 and

the bis-coordinating L-3 producing complete reactions in good

yields. Thus, among nucleosides, modification at C-6 can

potentially be more difficult if a coordinating moiety is

introduced at C-6. On the other hand, this may not be the case

for C-2 modification, and reactions at this site may be simpler

to achieve. Remarkably, despite the potential multiple coordina-

tion sites in purine deoxynucleosides, Pd-catalyzed reactions

with these compounds are quite effective in constructing hitherto

unknown and unusual nucleosides for exploring new avenues

on modified DNA. The method described in this report

complements known reactions of nucleosides and enables the

development of products that are not readily attainable other-

wise. Further studies on Pd catalysis within the domain of

nucleoside and DNA modification are currently being explored

in our laboratories.

Figure 4. Panel A: two possible structures of C-6 aminoaryl

nucleosides with coordinated Pd. Panel B: two possible structures of

C-2 aminoaryl nucleosides with coordinated Pd.

position are not quite so dependent on the ability of a ligand to

bis-coordinate, although a slightly better yield was obtained with

the bis-coordinating L-3. The yields in these two reactions also

compare favorably with yields obtained for other arylamination

reactions on 3 using Pd₂(dba)₃/(+)-L-1/t-BuONa but involve

the use of a much milder base.²⁴This factor will be significant

when the use of either t-BuONa or Cs₂CO₃is precluded. An

important point to note is that in contrast to the reported efficient

arylaminations of 3 with t-BuONa as base,²⁴we have observed

that arylamination of 1 with p-toluidine in the presence of

t-BuONa proceeds at room temperature, and modest product

formation is accompanied by significant amounts of uncharac-

terized byproducts.

One possible explanation for the differences observed in

reactions at the C-6 and the C-2 position has to do with the

competition of the product with the ligand for coordination with

the palladium, thereby lowering throughput. Structures of

possible Pd-chelated nucleosides that could be produced in C-N

bond-forming reactions are shown in Figure 4. In each of these

structures, there can be N,N bis-coordination of the product with

Pd. As alluded to earlier, studies on palladium-nucleoside

complexes indicate that the N¹and N⁷are most probable

coordination sites, with N³usually being considered a disfavored

site.^31-33On the basis of these reasons, as well as the fact that

five-membered palladacycles are preferred over the four-

membered ones,⁴⁰the N⁶,N⁷-coordinated complex is a plausible

species that can be produced in C-N bond-forming reactions

at the C-6 position of purines. Ligands that more effectively

sequester the palladium and prevent Pd coordination with the

product are, therefore, more likely to be effective in C-N bond

formation at the C-6 position. Thus, the bis-coordinating ligands,

such as L-1, L-2, and L-3, provide efficient conversion in the

amination reaction at this site. On the basis of the same rationale,

C-N bond formation at the C-2 position should be less ligand-

dependent. Our observations are, in fact, consistent with this,

and both the mono- and the bis-coordinating ligands (L-3 and

L-4) provide comparable results.

Experimental Section

The arylboronic acids, palladium(II)acetate [Pd(OAc)₂], tris(diben-

zylidineacetone)dipalladium [Pd₂(dba)₃], ligands L-1, L-2, L-4, L-5,

L-6, Cs₂CO₃, K₃PO₄, and anhydrous 1,2-dimethoxyethane (1,2-DME),

as well as 5% Pd on C, were purchased from commercial sources and

were used without additional purification. Ligand L-3 was synthesized

as previously reported.¹⁶The catalysts and ligands were stored under

N₂in a desiccator at 0 °C and the bases were stored under N₂in a

desiccator at room temperature. For each cross-coupling, 1,4-dioxane

was distilled from NaBH₄just prior to performing the reaction, whereas

THF used for catalytic reduction was distilled from LiAlH₄. The

halonucleosides 1, 2, and 3 were synthesized as described.^22-24

Reactions were monitored by TLC (silica gel, 250 µ), and column

chromatographic purifications were performed on 200-300 mesh silica

gel. Solvents used for eluting the compounds, as well as TLC conditions

and R_fvalues, are provided under individual compound headings. Proton

Conclusions

In this study, the applicability of palladium catalysis to effect

arylation at the C-6 and C-2 positions of purine deoxynucleo-

(40) (a) Kocˇovsky´, P.; Vyskocˇil, Sˇ.; C´ısaˇrova´, I.; Sejbal, J.; Tisˇlerova´,

I.; Smrcˇina, M.; Lloyd-Jones, G. C.; Stephen, S. C.; Butts, C. P.; Murray,

M.; Langer, V. J. Am. Chem. Soc. 1999, 121, 7714-7715. (b) Newkome,

G. R.; Puckett, W. E.; Gupta, V. K.; Kiefer, G. E. Chem. ReV. 1986, 86,

451-489. (c) Zhang, L.; Zetterberg, K. Organometallics 1991, 10, 3806-

3813.

(41) For removal of the 2-cyanoethyl group, please see: (a) Gaffney, B.

L.; Jones, R. A. Tetrahedron Lett. 1982, 23, 2257-2260. (b) Beaucage, S.

L.; Iyer, R. P. Tetrahedron 1992, 48, 2223-2311. For removal of the

4-nitrophenylethyl group, please see: Trichtinger, T.; Charubala, R.;

Pfleiderer, W. Tetrahedron Lett. 1983, 24, 711-714.

7786 J. Am. Chem. Soc., Vol. 123, No. 32, 2001

Lakshman et al.

NMR spectra were obtained at 500 MHz; chemical shifts (δ) are

reported in ppm, and coupling constants (J) are in Hz.

0.11 (2s, 12H, Si-CH₃). HRMS calcd for C₂₈H₄₃N₅O₅Si₂Na (M⁺

Na), 608.2700. Found, 608.2704.

Typical Procedure for the Coupling of Bromonucleoside 1 with

Arylboronic Acids. In an oven-dried, screw-cap vial equipped with a

stirring bar were placed Pd(OAc)₂(2 mg, 8.9 µmol), L-4 (4.8 mg, 13.6

µmol), the boronic acid (1.5 molar equiv), and K₃PO₄(39 mg, 0.18

mmol). Finally, the bromonucleoside (1, 50 mg, 0.092 mmol) was

added, followed by the addition of freshly distilled, dry 1,4-dioxane

(0.6 mL). The vial was flushed with nitrogen, sealed with a Teflon-

lined cap, and placed in a sand bath that was maintained at 100-102

°C. The reactions were monitored by TLC, and upon completion, the

reaction mixtures were filtered through Celite. In each case, the residue

was washed with CH₂Cl₂and the filtrate was evaporated to dryness.

The products were purified by column chromatography on silica gel

using appropriate solvents (listed under individual compound headings,

vide infra). Fractions that contained the pure products were combined,

evaporated and finally dried under high vacuum to remove traces of

solvent.

6-Phenyl-3′,5′-bis-O-(tert-butyldimethylsilyl)-2′-deoxynebu-

larine (4a). Chromatography, CH₂Cl₂, followed by 2% acetone in CH₂-

Cl₂. Colorless oil, R_f(2% acetone in CH₂Cl₂) ) 0.31. ¹H NMR

(CDCl₃): 9.02 (s, 1H, purine-H), 8.79-8.77 (m, 2H, Ar-H), 8.43 (s,

1H, purine-H), 7.60-7.52 (m, 3H, Ar-H), 6.59 (t, 1H, 1′, J ) 6.5),

4.67 (m, 1H, 3′), 4.07 (app q, 1H, 4′, J ∼ 3.6), 3.90 (dd, 1H, 5′, J )

4.3, 11.2), 3.81 (dd, 1H, 5′, J ) 3.2, 11.2), 2.73 (app sept, 1H, 2′, J )

5.9, 6.9, 13.0), 2.51 (ddd, 1H, 2′, J ) 3.7, 6.1, 13.1), 0.94, 0.92 (2s,

18H, C(CH₃)₃), 0.13, 0.10, 0.098 (3s, 12H, Si-CH₃). HRMS calcd for

C₂₈H₄₄N₄O₃Si₂Na (M⁺+ Na), 563.2850. Found, 563.2864.

6-(4-Methoxyphenyl)-3′,5′-bis-O-(tert-butyldimethylsilyl)-2′-deox-

ynebularine (4b). Chromatography, CH₂Cl₂, followed by 2% acetone

in CH₂Cl₂. Colorless oil, R_f(2% acetone in CH₂Cl₂) ) 0.11. ¹H NMR

(CDCl₃): 8.96 (s, 1H, purine-H), 8.81 (d, 2H, Ar-H, J ) 9.0), 8.40 (s,

1H, purine-H), 7.09 (d, 2H, Ar-H, J ) 9.0), 6.57 (t, 1H, 1′, J ) 6.5),

4.66 (m, 1H, 3′), 4.06 (app q, 1H, 4′, J ∼ 3.6), 3.91 (s, 3H, OCH₃),

3.89 (dd, 1H, 5′, J ) 4.3, 11.2), 3.80 (dd, 1H, 5′, J ) 3.2, 11.2), 2.72

(app sept, 1H, 2′, J ) 6.0, 6.8, 13.0), 2.50 (ddd, 1H, 2′, J ) 3.7, 6.1,

13.0), 0.94, 0.92 (2s, 18H, C(CH₃)₃), 0.13, 0.10, 0.098 (3s, 12H, Si-

CH₃). HRMS calcd for C₂₉H₄₆N₄O₄Si₂Na (M⁺+ Na), 593.2955. Found,

593.2960.

6-(4-Acetylphenyl)-3′,5′-bis-O-(tert-butyldimethylsilyl)-2′-deox-

ynebularine (4f). Chromatography, CH₂Cl₂followed by 2% acetone

in CH₂Cl₂. Colorless oil, R_f(2% acetone in CH₂Cl₂) ) 0.17. ¹H NMR

(CDCl₃): 9.06 (s, 1H, purine-H), 8.91 (d, 2H, Ar-H, J ) 8.5), 8.50 (s,

1H, purine-H), 8.15 (d, 2H, Ar-H, J ) 8.5), 6.60 (t, 1H, 1′, J ) 6.4),

4.67 (m, 1H, 3′), 4.08 (app q, 1H, 4′, J ∼ 3.5), 3.91 (dd, 1H, 5′, J )

4.1, 11.2), 3.81 (dd, 1H, 5′, J ) 3.1, 11.2), 2.72 (app quint, 1H, 2′,

J ∼ 6.1), 2.69 (s, 3H, COCH₃), 2.52 (ddd, 1H, 2′, J ) 3.8, 6.1, 13.1),

0.94, 0.92 (2s, 18H, C(CH₃)₃), 0.13, 0.11 (2s, 12H, Si-CH₃). HRMS

calcd for C₃₀H₄₇N₄O₄Si₂(M⁺+ H), 583.3136. Found, 583.3134.

6-(3-Thiopheno)-3′,5′-bis-O-(tert-butyldimethylsilyl)-2′-deoxyneb-

ularine (4g). Chromatography, CH₂Cl₂, followed by 5% EtOAc in CH₂-

Cl₂. Colorless oil, R_f(5% EtOAc in CH₂Cl₂) ) 0.35. ¹H NMR

(CDCl₃): 8.95 (s, 1H, purine-H), 8.93 (dd, 1H, Ar-H, J ) 1.1, 3.1),

8.43 (s, 1H, purine-H), 8.29 (dd, 1H, Ar-H, J ) 1.1, 5.1), 7.46 (dd,

1H, Ar-H, J ) 3.1, 5.1), 6.57 (t, 1H, 1′, J ) 6.4), 4.66 (m, 1H, 3′),

4.06 (app q, 1H, 4′, J ∼ 3.6), 3.90 (dd, 1H, 5′, J ) 4.2, 11.2), 3.80

(dd, 1H, 5′, J ) 3.2, 11.2), 2.71 (app quint, 1H, 2′, J ∼ 6.7), 2.50

(ddd, 1H, 2′, J ) 3.9, 6.1, 13.1), 0.94, 0.93 (2s, 18H, C(CH₃)₃), 0.13,

0.11 (2s, 12H, Si-CH₃). HRMS calcd for C₂₆H₄₂N₄O₃SSi₂Na (M⁺

Na), 569.2414. Found, 569.2437.

Typical Procedure for the Coupling of Chloronucleoside 2 with

Arylboronic Acids. In an oven-dried, screw-cap vial equipped with a

stirring bar were placed Pd(OAc)₂(2 mg, 8.9 µmol) and L-4 (4.8 mg,

13.6 µmol). Freshly distilled, dry 1,4-dioxane (0.6 mL) was added, the

vial was flushed with nitrogen and sealed with a Teflon-lined cap, and

the mixture was stirred at room temperature for a few minutes. The

boronic acid (1.5 molar equivalents), chloronucleoside (2, 46.0 mg,

0.092 mmol) and K₃PO₄(39 mg, 0.18 mmol) were then added to the

vial, the vial was again flushed with nitrogen, sealed with a Teflon-

lined cap, and placed in a sand bath that was maintained at 100-102

°C. Reactions were monitored by TLC, and upon completion, the

reaction mixtures were filtered through Celite. In each case, the residue

was washed with CH₂Cl₂, and the filtrate was evaporated to dryness.

The products were purified by column chromatography on silica gel

using appropriate solvents (described under the individual compound

headings, vide supra). Fractions that contained the pure products were

combined, evaporated, and finally dried under high vacuum to remove

traces of solvent. The ¹H NMR spectra of compounds 4a, 4b, and 4e-g

were identical to those obtained from the bromonucleoside reactions.

Typical Procedure for the Coupling of Bromonucleoside 3 with

Arylboronic Acids. In an oven-dried, screw-cap vial equipped with a

stirring bar were placed Pd(OAc)₂(1.7 mg, 7.5 µmol), L-4 (4.0 mg,

11.4 µmol), the boronic acid (1.5 molar equivalents), and K₃PO₄(39

mg, 0.15 mmol). Finally, the bromonucleoside (3, 50 mg, 0.077 mmol)

was added, followed by the addition of freshly distilled, dry 1,4-dioxane