Elsevier

Tetrahedron

Volume 64, Issue 26, 23 June 2008, Pages 6125-6137
Tetrahedron

Site-selective direct arylation of unprotected adenine nucleosides mediated by palladium and copper: insights into the reaction mechanism

https://doi.org/10.1016/j.tet.2008.01.062Get rights and content

Abstract

Reaction conditions facilitating the site-selective direct aryl functionalisation at the C-8 position of adenine nucleosides have been identified. Many different aromatic components may be effectively cross-coupled to provide a diverse array of arylated adenine nucleoside products without the need for ribose or adenine protecting groups. The optimal palladium catalyst loading lies between 0.5 and 5 mol %. Addition of excess mercury to the reaction had a negligible affect on catalysis, suggesting the involvement of a homogeneous catalytic species. A study by transmission electron microscopy (TEM) shows that metal containing nanoparticles, ca. 3 nm with good uniformity, are formed during the latter stages of the reaction. Stabilised PVP palladium colloids (PVP=N-polyvinylpyrrolidone) are catalytically active in the direct arylation process, releasing homogenous palladium into solution. The effect of various substituted 2-pyridine ligand additives has been investigated. A mechanism for the site-selective arylation of adenosine is proposed.

Introduction

C-Modified nucleosides are widely used as pH sensors, fluorescent markers, therapeutic agents and supramolecular building blocks.1 Studies that probe biomolecular structure and biochemical mechanisms2, 2(a), 2(b), 2(c), 2(d), 2(e) are of fundamental importance, driving the need for the development of efficient synthetic methods to access structurally diverse non-natural nucleosides, particularly purines.

There are several ways to functionalise purine compounds. Sonogashira, Suzuki and Stille cross-coupling (hereafter coupling) processes, catalysed by palladium(0),3 are routinely used to access arylated purine nucleosides (Scheme 1).4, 4(a), 4(b), 4(c) Historically the protection of both the sugar hydroxyl groups and the reactive heteroaromatic substituents was required for successful couplings. However, recent studies have demonstrated that unprotected halogenated nucleosides may be effectively cross-coupled with various nucleophilic components.5, 5(a), 5(b), 5(c) Arguably more remarkable are the findings that Suzuki–Miyaura cross-couplings are possible on halogenated triphosphate purine bases,6, 7, 7(a), 7(b) this despite the thermal and base sensitivity of the phosphoester bonds in the triphosphate moiety.

Toxic reagents, e.g., organostannanes, limit the use of Stille couplings in industrial scale-up processes, and whilst organoboronic acids are versatile coupling components for Suzuki–Miyaura couplings, in some cases their stability (protodeborylation) can be problematic, particularly in certain heteroaromatic components. Furthermore, these reactions cannot generally be viewed as atom economic transformations. For example, only the phenyl group (77 mu) is transferred from PhSnBu3 and the rest is waste (290 mu), a problem derived from the need for prefunctionalisation in the ‘nucleophilic’ component. To combat these key issues, greener and more efficient methods for aryl functionalisation continue to be identified. Generally, protocols for direct arylation of organic compounds,8 by C–H functionalisation, including the site-selective modification9 of electron-rich heteroaromatics,10 have reached a point where they may be applied to more challenging molecular architecture.11, 11(a), 11(b), 11(c), 11(d)

With the primary aim of adding to the contemporary portfolio of modified molecular probes for practical use in bioapplications, we wished to investigate whether catalytic direct arylation methods can be employed for the synthesis of unprotected purine nucleosides. In unison with our studies, Hocek and co-workers recently disclosed that protected purines can be arylated at the 8-position using catalytic palladium in the presence of excess CuI in DMF at 160 °C for 60 h.12 However, no sugar variants were given. Our studies on the site-selective direct arylation of purine nucleosides are reported in this paper.13

On consideration of the structures of the nucleosides a problem that emerges is the potential instability of the glycosyl bond at high temperatures, e.g., 160 °C (for adenosine the t1/2>8000 h at 90–100 °C in water or formamide).14 Site selectivity9, 15 further needs to be considered; C-arylation can potentially occur at C2 or C8 (Fig. 1). Whilst accurate experimental pKa values for the C–H bonds in adenosine are not known, based on the theoretical pKa values determined for 9-methyl-9H-purine (in DMSO, H2=40.3; H8=29.3)16 we predict that preferential C–H activation should occur at site 1. N-arylation is further possible—a recent report details that a Pd–Xantphos catalytic system can promote selective N-arylation of protected 2′-deoxyadenosines and 2′-deoxyguanosines with aryl bromides.17

Initially, reaction conditions reported in the literature for direct arylation of various heteroaromatic compounds were screened to assess whether reactions of adenosine with aryl iodides were possible. Using standard arylation conditions {Pd(OAc)2, P(t-Bu)3·HBF4, PhCl or PhBr, K3PO4, DMA, 100 °C, 24 h}, or {Pd(OAc)2, PhBr, NaOAc, 120 °C, 20 h} negligible direct arylation was seen. More success was had using the conditions described by Bellina and Rossi,18, 18(a), 18(b), 18(c) which are similar to those reported by Hocek and co-workers.12 However, deglycosylation was found to be an issue at high temperatures over prolonged reaction times. On monitoring a reaction at 160 °C by HPLC it was revealed that direct arylation had occurred within minutes! On consumption and through extensive deglycosylation of adenosine, any remaining PhI rapidly homocoupled to give biphenyl. Deglycosylation of the cross-coupled product then occurred steadily over the course of several hours. Running the reaction at 120 °C improved the reaction efficacy considerably; the product profile is shown in Scheme 2. Note: homocoupling of adenosine was not observed under these conditions.19

Rapid colour changes are observed in reaction mixtures of adenosine and iodobenzene (Fig. 2). After 5 min of heating a yellow/orange colour was evident (plate A), which changed to a green/brown colour after 25 min (∼30% conversion, plate B). After 1.5 h the formation of metal containing nanoparticles was observed in the reaction mixture (∼65% conversion, plate C).

The pH upon work-up should be ∼6.5, which allows the product to be extracted with an ethyl acetate/isopropanol solvent mixture (9:1). Subjecting the crude product to chromatography on silica gel gave the direct arylation product in 65% yield (Table 1, entry 1). A similar yield was obtained using N-methylpyrrolidone (NMP) as the solvent (63%). The reproducibility of these reactions is ±5%. In the presence of catalytic quantities of CuI negligible turnover was observed (entry 2). Omission of Pd(OAc)2 from the reaction resulted in poor turnover (4%), significant decomposition and deglycosylation (entry 3). The Pd(0) source, Pd2(dba)3 (dba=E,E-dibenzylidene acetone) acted as a catalyst (29% yield, entry 4), but was not as effective as Pd(OAc)2. The more activated Pd2(dm-dba)3·dm-dba20 complex gave an improved yield (59%, entry 5), showing that the dibenzylidene acetone ligand is non-innocent in this direct arylation reaction, in accord with other cross-coupling processes.21, 21(a), 21(b), 21(c), 21(d), 21(e), 21(f), 21(g), 21(h) Finally, microwave heating is ineffective for direct arylation of adenosine due to significant decomposition (mainly deglycosylation).

Having identified the key reaction/work-up parameters a library of aryl halides were evaluated for direct arylation of adenosine (Table 2). A series of para-substituted aryl iodides afforded the 8-arylated purine products in good yields (entries 1–4). Remarkably, ionisable substituents, e.g., OH and NH2 groups are also tolerated (entries 5 and 6, respectively). Substrates possessing electron-deficient ortho-substituents are not accepted for C-arylation (entries 7 and 10). Catellani and co-workers reported that the ortho-trifluoromethyl group can interact with Pd(II), reducing its subsequent reactivity, in norbornene-assisted arylation processes.22 On switching this substituent to the meta position the coupled product was formed in good yield (entry 8). In the presence of a meta-nitro substituent, the cross-coupled product was obtained in modest yield (entry 11). The p-nitro substituent gave the best yield from the most strongly ‘electron-deficient substituent series’ (entry 12). Iodonapthalene couples in near quantitative yield (entry 13). Quite remarkably the chemoselectivity was reversed when an ortho-nitro substituent is present—N-arylation emerged as the exclusive reaction pathway (Fig. 3).

We have established that 1,4-diiodobenzene can also be used as an arylation substrate (Scheme 3). Despite using fewer equivalents of 1,4-diiodobenzene (0.5 rather than 2 equiv with respect to the adenosine), both 1,4-di-(8′-adenosinyl)benzene and 8-(p-iodophenyl)adenosine were formed in this reaction, albeit in low yield (non-optimised). These compounds could be useful in supramolecular assembly or ligand design.

2′-Deoxyadenosine can also be arylated under the standard conditions by slight modification to the reaction conditions; at 120 °C substantial deglycosylation was observed, giving 8-phenyladenine (Scheme 4), reflecting the lower stability of 2′-deoxyribose. However, arylation worked well at 80 °C giving the coupled product in 84% yield.

Section snippets

Mechanistic studies

Having developed a generic set of conditions for the direct arylation of adenosine and 2′-deoxyadenosine we felt that some of the observations made in the study warranted further investigation. Several parameters have been probed to assess their impact on the benchmark reaction (Scheme 2) with a view to gaining additional insight into the reaction mechanism.

Mechanistic discussion

Adenosine, as a substrate, is a complicated ligand with several possible coordination modes to palladium.35, 35(a), 35(b), 35(c) Monomers such as I are known (Fig. 6), as are dimeric and higher order complexes, where N7 is the dominant ligating atom. Similar copper complexes can also be formed (e.g., II).

In terms of the reaction mechanism an electrophilic substitution, via intermediate III, can be proposed, in accord with the proposals made by Miura (in the arylation of thiazoles and

Conclusion

In summary, valuable reaction conditions for the direct aryl functionalisation of unprotected adenosines have been developed. Deglycosylation has been identified as a major issue using similar conditions to those described previously for protected adenines and other direct arylation processes. The above synthetic protocol allows C8-arylated adenosines to be obtained in one step. In Suzuki–Miyaura cross-couplings there is an absolute requirement for the prefunctionalisation of adenosine

General details

Proton (1H, 400 MHz) NMR spectra were recorded on an Oxford AS400 spectrometer. Samples were prepared using approximately 10 mg of compound dissolved in 0.7 ml of DMSO-d6. Chemical shifts were referenced to residual undeuterated DMSO in DMSO-d6 at δ=2.5 ppm. All spectra were reprocessed using MestRec version 4.9.9.6 on a PC (SineBell apodization was used to obtain detailed proton spin–spin coupling patterns and constants). Carbon (13C, 100.6 MHz) NMR spectra were recorded on the same NMR

Acknowledgements

The authors wish to thank Dr. D.A. Ashford for a preliminary reaction analysis by reverse-phase liquid chromatography and mass spectrometry, Ms. M. Stark for technical assistance with the transmission electron microscopy, and Dr. R.G. Sturmey and Professor H.J. Leese for the use of an Agilent 1100 high performance liquid chromatography system. Mr. P. Ellis, Dr. P. Sehnal and Dr. A.F. Lee (York) are thanked for providing characterised Pd–PVP colloids and for valuable discussions. We are grateful

References and notes (46)

  • E. De Clercq

    Biochem. Pharmacol.

    (1984)
    N. Minakawa et al.

    Nucleosides Nucleotides

    (1996)
    E. De Clercq et al.

    Proc. Natl. Acad. Sci. U.S.A.

    (1979)
    S. Manfredini et al.

    J. Med. Chem.

    (1995)
    M. Hocek et al.

    J. Med. Chem.

    (2000)
  • A. Collier et al.

    Org. Biomol. Chem.

    (2006)
    A. Collier et al.

    Chem. Commun.

    (2008)
  • For general reviews,...K. Godula et al.

    Science

    (2006)
    L.-C. Campeau et al.

    Chem. Commun.

    (2006)
    A.R. Dick et al.

    Tetrahedron

    (2006)
    J.-Q. Yu et al.

    Org. Biomol. Chem.

    (2006)
  • J.A. Widegren et al.

    J. Mol. Catal. A: Chem.

    (2003)
  • B.S. Lane et al.

    Org. Lett.

    (2004)
  • M.T. Reetz et al.

    Chem. Commun.

    (2004)
    I.J.S. Fairlamb et al.

    Dalton Trans.

    (2004)
    J.G. de Vries

    Dalton Trans.

    (2006)
    I.J.S. Fairlamb et al.

    New J. Chem.

    (2006)
  • D. Camboli et al.

    Inorg. Chim. Acta

    (1983)
  • K. Burgess et al.

    Chem. Rev.

    (2000)
  • I.J.S. Fairlamb

    Tetrahedron

    (2005)
  • L.A. Agrofoglio et al.

    Chem. Rev.

    (2003)
    M. Hocek

    Eur. J. Org. Chem.

    (2003)
    M.K. Lakshman

    Curr. Org. Synth.

    (2005)
  • A.G. Firth et al.

    Tetrahedron Lett.

    (2006)
    A. Collier et al.

    Synth. Commun.

    (2006)
    K.H. Shaughnessy

    Eur. J. Org. Chem.

    (2006)
  • P. Capek et al.

    Org. Biomol. Chem.

    (2006)
  • M. Catellani et al.

    Eur. J. Org. Chem.

    (2007)
  • I.J.S. Fairlamb

    Chem. Soc. Rev.

    (2007)
  • H.A. Chiong et al.

    Org. Lett.

    (2007)
    I.V. Seregin et al.

    Chem. Soc. Rev.

    (2007)
  • I. Čerňa et al.

    Org. Lett.

    (2006)
  • I. Čerňa et al.

    Chem. Commun.

    (2007)
  • S. Schröter et al.

    Tetrahedron

    (2005)
  • K. Shen et al.

    Tetrahedron

    (2007)
  • F.N. Ngassa et al.

    Org. Lett.

    (2006)
  • F. Bellina et al.

    Eur. J. Org. Chem.

    (2006)
    F. Bellina et al.

    Eur. J. Org. Chem.

    (2006)
    F. Bellina et al.

    Tetrahedron

    (2007)
  • Cited by (83)

    • Application of water-soluble palladium-catalyst systems for introduction of C-C bonds in nucleosides

      2018, Palladium-Catalyzed Modification of Nucleosides, Nucleotides and Oligonucleotides
    • C-H bond functionalization strategies for modification of nucleosides

      2018, Palladium-Catalyzed Modification of Nucleosides, Nucleotides and Oligonucleotides
    • Nondirected C-H Bond Functionalizations of (Hetero)arenes

      2017, Strategies for Palladium-Catalyzed Non-directed and Directed C bond H Bond Functionalization
    View all citing articles on Scopus
    View full text