A One-Pot Synthesis of Highly Functionalized Purines

Article abstract of DOI:10.1021/acs.orglett.7b03209

Highly substituted purines were synthesized in good to high yields through a one-pot straightforward metal-free scalable method, using the Traube synthesis adapted to Vilsmeier-type reagents. From 5-amino-4-chloropyrimidines, new 9-aryl-substituted chloropurines and intermediates for peptide nucleic acid synthesis were prepared. Variant procedures allowing a rapid synthesis of ribonucleosides and 7-benzylpurine from 5-amidino-6-aminopyrimidines are also reported to illustrate the high potential of this versatile toolbox. This route appears to be particularly interesting in the field of nucleic acids for a direct and rapid access to various new 8-alkylpurine nucleosides.

Full text of DOI:10.1021/acs.orglett.7b03209

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX-XXX
A One-Pot Synthesis of Highly Functionalized Purines
Renaud Zelli,† Wael Zeinyeh,†,‡ Romain Haudecoeur,† Julien Alliot,† Benjamin Boucherle,†
̈
Isabelle Callebaut,§ and Jean-Luc Decout*,†
́
†Univ. Grenoble Alpes, CNRS, DPM, 38000 Grenoble, France
‡Univ. Grenoble Alpes, CEA, CNRS, BIG-LCBM, 38000 Grenoble, France
§IMPMC, Sorbonne Universites
́ ́
, UPMC Univ. Paris 06, UMR CNRS 7590, Museum National d’Histoire Naturelle, IRD UMR 206,
IUC, Case 115, 4 Place Jussieu, 75005 Paris Cedex 05, France
S
* Supporting Information
ABSTRACT: Highly substituted purines were synthesized in good
to high yields through a one-pot straightforward metal-free scalable
method, using the Traube synthesis adapted to Vilsmeier-type
reagents. From 5-amino-4-chloropyrimidines, new 9-aryl-substituted
chloropurines and intermediates for peptide nucleic acid synthesis
were prepared. Variant procedures allowing a rapid synthesis of
ribonucleosides and 7-benzylpurine from 5-amidino-6-aminopyrimi-
dines are also reported to illustrate the high potential of this versatile
toolbox. This route appears to be particularly interesting in the field
of nucleic acids for a direct and rapid access to various new 8-
alkylpurine nucleosides.
n recent decades, the purine scaffold has contributed
substituted purines using various electrophiles (see Supporting
Information, SI).4 9H-Purines were also extensively modified,
especially chloropurines, for further functionalization. However,
access to some chemical spaces remains challenging, jeopard-
izing the development of pertinent analogues in structure−
activity relationship processes. 9-Arylation of purines is an
emblematic example of such difficulties, as chemists still
struggle with inefficient reactions and narrow substrate scopes.
The 9-arylation is indeed usually performed through either (1)
copper-catalyzed C−N cross-coupling reaction of purines5 or
(2) reaction of an arylamine with a 5-amino-4-chloropyrimi-
dines before ring closure by an orthoester derivative,6 an acyl
chloride,7 or more recently via an additional prior formylation
step.8 These pathways present drawbacks, such as restricted
substrate scope, low yields, and multiple steps, especially for
Chan−Lam reactions. Beyond the case of aryl groups, the
synthesis of 9-alkylpurines or purine-based nucleosides via
direct substitution also causes difficulty, as the reaction leads to
mixtures of N3-, N7-, and N9-substitutions, with a loss of yield
and purification issues. In addition, all these methods do not
provide a concomitant easy access to diversity at position 8, and
a bromination/substitution sequence is generally needed for
reaching 8-substituted purines of interest.
I
intensively to the development of both biotechnological
tools and bioactive compounds.1 Purine-based analogues have
also been extensively used in clinics since the approval of the
antiviral drug, Vidarabine, in 1976. Many drugs have been
designed from analogy with natural purine-based nucleosides
and nucleobases.2 They traditionally come with structural
diversity at position 9 that is often linked to various
carbohydrates, carbocycles, or alkyl groups (Figure 1).
Since the preparation of purine itself by Fisher in 1898 and
the first general synthesis of purines from 4,5-diaminopyr-
imidines and formic acid by Traube, many variants of the
original synthesis3 were developed that led to diversely
The Vilsmeier’s reagent, i.e. (chloromethylene)-
dimethyliminium chloride, has been used for generating
formamidine groups, often as an amino protecting group,9 or
to generate formamido groups10 and also sporadically as
intermediates for the synthesis of various 1,3-diazole hetero-
Figure 1. Examples of bioactive purine-based compounds with various
structures and activities.
Received: October 14, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b03209
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
aromatic rings upon cyclization (e.g., imidazoles, benzimida-
zoles, 9H-purines).11 However, the few reported reactions were
associated either with the formation of naked rings starting
from o-diamino or amino-o-nitro derivatives, without simulta-
neous N-functionalization, or with extreme pressure, temper-
ature, and reaction time conditions, incompatible with most of
the classical functional groups. Using 5-amino-4-chloropyrimi-
dines, we considered that an amidine group embedded in the
amino substituent could easily undergo cyclization right after
substitution of the chlorine atom by an amine, resulting in the
formation of 9-substituted 6-chloropurines preferably under
mild acidic conditions (Scheme 1). Indeed, under basic
conditions, dimethylamine released in the final cyclization/
aromatization step or formed under drastic conditions from
DMF, if chosen as a solvent,8,12 could react in SNAr reactions
from a chlorine atom carried by the starting or intermediate
pyrimidines and/or the formed purines. The conditions of this
reaction sequence thus appeared as fully compatible for
building a one-pot strategy, directly from the starting
pyrimidine to the final substituted purine. Adjustments were
envisaged for reacting starting pyrimidines with electrophiles at
positions 4 (Variant A) or 5 (Variant B), leading to a
straightforward access to nucleosides and 7-substituted purines.
In all cases, the 8-substitution of the purines could be managed
by using modified Vilsmeier-type reagents generated in parallel.
Thus, we report a new one-pot, metal-free reaction toolbox
based on Vilsmeier’s reagent reactivity, dedicated to the
synthesis of extensively substituted purines with high yields,
an associated extremely wide directly reachable diversity at
positions 2, 6, 7, 8 and/or 9, and a privileged access to 2- and/
or 6-halogenated purines that allow further fine-tuned structural
modifications.
Scheme 1. A New Strategy for Purine Synthesis
A typical procedure involves 5-amino-4,6-dichloropyrimi-
dines 1−3 and an amine refluxed in dioxane for 16 h. The
reaction was monitored by TLC to follow the disappearance of
the starting material and to witness aromatic nucleophilic
substitution of the chloro group by the amine. The Vilsmeier’s
Scheme 2. Substrate Scope of the One-Pot Substitution/Cyclization Sequence
a
b
Modified procedure via formation of 5-amidino-2,4,6-trichloropyrimidine; see ref 14 and Supporting Information (SI). DIPEA (2.0 equiv) was
c
d
used instead of p-toluenesulfonic acid. Concomitant protection of the 2-amino group form occurred under the reported conditions. Modified
procedure via prior formation of a Vilsmeier-type reagent; see ref 16 and SI.
B
DOI: 10.1021/acs.orglett.7b03209
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
reagent (2 equiv) was then added, immediately inducing the
formation of the amidine group and the subsequent cyclization
in less than 30 min at 25 °C. After rapid workup and
purification, the reaction sequence afforded the desired purines
in good to very good yields. Overall, although the three steps
described above, i.e. nucleophilic substitution, formation of the
amidine group, and cyclization, were conducted in one pot
without any intermediate purification, very few byproduct
formation was observed, leading to easy isolation of the purines
through a final chromatography. Several experimental con-
ditions were modulated in order to assess the robustness of the
reaction. The influences of water traces and oxygen were
evaluated using nonanhydrous solvents and an ambient
atmosphere, resulting in no yield loss (2 equiv of Vilsmeier’s
reagent). Comparative assays performed from 50 mg and 1 g of
starting pyrimidine demonstrated the efficacy of the reaction at
different operating scales. Overall, these results emphasized a
low sensitivity to moisture, oxygen, and scale.
byproducts) while POCl3, SOCl2, and N,N-dimethylacetamide
dimethyl acetal led mainly to lower yields with byproduct
formation.
The synthesis of modified nucleosides, a topic of high
importance, can be accessible from aminosugars according to
the developed Vilsmeier-based methodology. However, the
preparation of aminosugars can be hard and lengthy. We thus
addressed this hurdle by developing a first Variant “A” of the
method (Scheme 3), mainly for rapid access to 8-substituted
Scheme 3. Substrate Scope of “Variant A” Sequence
The scope of the reaction was investigated using compound
1 and several aromatic amines carrying various substituents for
synthesizing purines 4−17 (Scheme 2). 3- and 4-Chloro, 3,5-
dichloro-, and 4-bromo-anilines and naked aniline were all
proven effective substrates, yielding corresponding purines
(respectively 4, 5, 7, 8, and 9) through the one-pot process.
Unsurprisingly, 2-chloroaniline as the starting amine showed far
less reactivity; however, the corresponding purine 6 was
isolated with an acceptable yield in a development context.
The reaction was then performed with a variety of anilines for
testing the preparation of purines substituted by electron-
donating (10−12) or electron-withdrawing groups (13−16) or
heterocyclic moieties (17) at position 9. High yields were
obtained, except for 6 for which it was limited by steric effect in
the first step, indicating low influence of the nucleophilicity of
the amine, even when considering Hammett constant extrema
(e.g., p-NO2Ph and p-OMePh, 15 and 12, respectively).13
Diversely 2-substituted pyrimidines, with R2 = Me, NH2, Cl,14
were also reacted with p-substituted anilines and gave yields
that remain good (18−23). Aliphatic amines reacted effectively
under alkaline conditions during the first step to give purines
24−27. The use of glycine tert-butyl esters readily resulted in
the production of building blocks (26 and 27) for peptide
nucleic acid (PNA) synthesis. This shows the expected
compatibility of the method with the presence of acid-labile
groups since the release of dimethylamine during the
cyclization/aromatization process limits the acidification of
the medium.
The preparation of a key intermediate 23 in the synthesis of
SR 3029 (Figure 1), a commercially available kinase inhibitor
with potent antiproliferative properties, aptly illustrates the
potential of the present method. Compound 23 was obtained
with 95% yield14 as compared to the 18% yield described for its
synthesis from 2,6-dichloropurine by the Chan−Lam coupling
reaction.15
For C8 functionalization, Vilsmeier-type reagents were
prepared separately using oxalyl chloride and a suitable amide
and were added at 35 °C to the reaction mixture right after the
first step of the sequence.16 The cyclization step was then
performed at 80 °C, as it appeared to be less reactive than for
classical Vilsmeier’s reagent. Thus, the method was found
tolerant to sterical hindrance, as methyl, dimethylamino,
phenyl, and 4-chlorophenyl groups were all introduced with
good yields (28−32). Oxalyl chloride presented the best
characteristics in a one-pot context (i.e., formation of gaseous
ribonucleosides. The reaction of 6-amino-5-amidinopyrimidines
33−35 with electrophiles provided the purine N9-sugar, thereby
opening a route to nucleosides. Precursors of adenosine,
ascamycin, and guanosine derivatives 36−38 were synthesized
in only one step from amidines 33−35, 1-O-acetyl-2,3,5-tri-O-
benzoyl-β-D-ribose, and TMSOTf, with moderate yields
through nonoptimized silyl-Hilbert−Johnson17 procedures
(replacement of TMSOTf by SnCl4 led to similar yields).
Interestingly, the method allowed the straightforward intro-
duction of a substituent at position 8, as demonstrated with the
synthesis of the 8-methyl nucleoside 37. This procedure of
glycosylation was highly selective since only one ribonucleoside
1
was detected by H NMR.
In the presence of TMSOTf, pyrimidine 33 led exclusively to
6-chloropurine that appeared in additional experiments to be
unreactive in the glycosylation reaction without silylation (SI).
Thus, the formation of 9H-6-chloropurines can compete with
the oxocarbenium formation necessary for pyrimidine glyco-
sylation, thereby limiting the yields. Under the same conditions,
pyrimidine 34 was converted to nucleoside 37, isolated in 42%
yield as the sole detected nucleoside (NMR), and unreacted 34
was recovered (55%), showing in this case a weak yield
limitation via 9H-purine formation. These results demonstrate
that 5-amidino-4-aminopyrimidines can be used for preparing
rapidly purine nucleosides via a highly selective one-pot
glycosylation−cyclization process involving primarily the
reaction of the silylated pyrimidine amino group with the
activated sugar.18
Finally, 7-benzylpurine was prepared through a second
Variant “B” involving (1) selective N5-alkylation of 5-amino-
4,6-dichloropyrimidines and amination, according to Liu et
al.,19 and (2) addition of a Vilsmeier-type reagent resulting in a
quantitative purine scaffold formation (SI).
In summary, the versatile one-pot synthesis of highly
substituted purines and its variant routes A and B reported
herein highlight the large chemical potential of the method
made of Vilsmeier-type reagents and pyrimidines. Such a
toolbox should receive many applications in the search for new
biological tools and drugs, strongly broadening the accessible
C
DOI: 10.1021/acs.orglett.7b03209
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Andreasen, J.; Shimpukade, B.; Ulven, T. Green Chem. 2013, 15, 336−
chemical diversity. The one-pot synthesis is compatible with
the presence of acid-labile groups and avoids the release of
nucleophilic species able to substitute the 6-chlorine atom of
the prepared purines in contrast to other methods of purine
synthesis from pyrimidines that release water or alcohols. In the
pyrimidine intermediates, the high reactivity of the primary 5-
or 6-amino group in comparison to that of the secondary 6- or
5-amino group introduced through SNAr reaction allows a
highly selective Vilsmeier reaction.
340.
(6) (a) Aguado, L.; Camarasa, M.-J.; Per
Chem. 2009, 11, 210−212. (b) Aguado, L.; Canela, M.-D.; Thibaut, H.
J.; Priego, E.-M.; Camarasa, M.-J.; Leyssen, P.; Neyts, J.; Perez-Perez,
́ ́
ez-Perez, M.-J. J. Comb.
́
́
M.-J. Eur. J. Med. Chem. 2012, 49, 279−288.
(7) Yang, J.; Dang, Q.; Liu, J.; Wei, Z.; Wu, J.; Bai, X. J. Comb. Chem.
2005, 7, 474−482.
(8) Dejmek, M.; Kovackova, S.; Zbornikova, E.; Hrebabecky, H.;
Sala, M.; Dracinsky, M.; Nencka, R. RSC Adv. 2012, 2, 6970−6980.
(9) Toste, D.; McNulty, J.; Still, I. W. J. Synth. Commun. 1994, 24,
1617−1624.
(10) Daluge, S. M.; Martin, M. T.; Sickles, B. R.; Livingston, D. A.
Nucleosides, Nucleotides Nucleic Acids 2000, 19, 297−327.
(11) (a) Pawar, V. G.; De Borggraeve, W. M.; Robeyns, K.; Van
Meervelt, L.; Compernolle, F.; Hoornaert, G. Tetrahedron Lett. 2006,
47, 5451−5453. (b) Dohle, W.; Staubitz, A.; Knochel, P. Chem. - Eur. J.
2003, 9, 5323−5331. (c) Stucky, G.; Griffiths, G. PCT WO96/21664,
1996.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.7b03209.
Experimental procedures and spectral data (PDF)
(12) Cechova, L.; Jansa, P.; Sala, M.; Dracinsky, M.; Holy, A.; Janeba,
Z. Tetrahedron 2011, 67, 866−871.
AUTHOR INFORMATION
Corresponding Author
■
(13) The reaction of formamidine of 1 with 4-bromoaniline in a 1:1
ratio in the presence of benzenesulfonic acid led to the expected
purine 8 and its 6-(bromophenyl)amino derivative in a 10:90 ratio.
When using formamidine of 1 in excess (5 equiv), complete reaction
*E-mail: Jean-Luc.Decout@univ-grenoble-alpes.fr.
ORCID
1
was observed by H NMR to lead to 8 and 6-substituted 9-arylpurine
Romain Haudecoeur: 0000-0002-6271-4717
Benjamin Boucherle: 0000-0002-1174-4449
in a 70:30 ratio.
(14) This compound was prepared from N-(2-amino-4,6-dichlor-
opyrimin-5-yl)-N′,N′-dimethylformamidine which was converted to
the corresponding 2,4,6-trichloropyrimidine through diazotization
(44% yield); procedure adapted from: (a) Krchnak, V.; Arnold, Z.
Collect. Czech. Chem. Commun. 1975, 40, 1390−1395. (b) Nara, S. J.;
Jha, M.; Brinkhorst, J.; Zemanek, T. J.; Pratt, D. A. J. Org. Chem. 2008,
73, 9326−9333. This compound showed high reactivity and selectivity
in nucleophilic substitution with arylamines and allowed more efficient
reactions with o-substituted anilines (21 vs 6).
Jean-Luc Decout: 0000-0001-7058-1304
́
Notes
The authors declare the following competing financial
interest(s): R.Z., W.Z., R.H., B.B., and J.L.D. are co-inventors
of the related patent application PCT/IB2017/000688
(mentioned in the manuscript).
(15) (a) Bibian, M.; Rahaim, R. J.; Choi, J. Y.; Noguchi, Y.; Schurer,
̈
ACKNOWLEDGMENTS
■
S.; Chen, W.; Nakanishi, S.; Licht, K.; Rosenberg, L. H.; Li, L.; Feng,
Y.; Cameron, M. D.; Duckett, D. R.; Cleveland, J. L.; Roush, W. R.
Bioorg. Med. Chem. Lett. 2013, 23, 4374−4380. (b) Roush, W. R.;
Duckett, D. R.; Cleveland, J. L.; Rosenberg, L. H. PCT WO 2017/
066055, 2017.
(16) For 8-substituted purines, a rapid workup was performed after
the alkylation step, and the crude product in glyme was reacted with
the Vilsmeier-type reagent, prepared right before by adding oxalyl
chloride in a solution of amide in 1,2-dichloroethane.
(17) Niedballa, U.; Vorbrueggen, H. J. Org. Chem. 1976, 41, 2084−
2086.
(18) The reaction of the anomeric isomer mixture of 2,3,5-tri-O-
benzoyl-1-chlororibose and freshly silylated pyrimidine 33 led to the β-
ribonucleoside 36 in 34% yield without formation of 9H-purine
demonstrating purine formation from the silylated pyrimidine.
(19) Liu, J.; Dang, Q.; Wei, Z.; Shi, F.; Bai, X. J. Comb. Chem. 2006, 8,
410−416.
The authors are grateful to association “Vaincre La
Mucoviscidose” (B.B., J.A., and R.Z. Grants RF20140501061,
RF20150501421, RF20160501676) and ANR (Labex Arcane
ANR-11-LABX-0003-01, J.A. and W.Z. grants) for financial
support.
REFERENCES
(1) Rosemeyer, H. Chem. Biodiversity 2004, 1, 361−401.
■
(2) (a) De Clercq, E.; Li, G. Clin. Microbiol. Rev. 2016, 29, 695−747.
(b) Jordheim, L. P.; Durantel, D.; Zoulim, F.; Dumontet, C. Nat. Rev.
Drug Discovery 2013, 12, 447−464. (c) Pautus, S.; Sehr, P.; Lewis, J.;
Fortune,
́
A.; Wolkerstorfer, A.; Szolar, O.; Guilligay, D.; Lunardi, T.;
out, J.-L.; Cusack, S. J. Med. Chem. 2013, 56, 8915−8930.
Dec
́
(d) Serpi, M.; Ferrari, V.; Pertusati, F. J. Med. Chem. 2016, 59, 10343−
10382. (e) Robak, P.; Robak, T. Cancer Treat. Rev. 2013, 39, 851−861.
(f) Sharma, S.; Singh, J.; Ojha, R.; Singh, H.; Kaur, M.; Bedi, P. M. S.;
Nepali, K. Eur. J. Med. Chem. 2016, 112, 298−346.
(3) Traube, W. Ber. Dtsch. Chem. Ges. 1900, 33, 1371.
(4) (a) Shaw, E. J. Biol. Chem. 1950, 185, 439−447. (b) Lister, J. H.
The Chemistry of Heterocyclic Compounds: The Purines, Supplement 1;
Taylor, E. C., Ed.; Wiley: New York, 1996. (c) Seela, F.; Ramzaeva, N.;
Rosemeyer, H. Product class 17: Purines. In Science of Synthesis, Vol.
16; Yamamoto, Y., Ed.; Thieme: Stuttgart, 2003; pp 945−1108.
(5) (a) Chan, D. M.; Monaco, K. L.; Wang, R.-P.; Winters, M. P.
Tetrahedron Lett. 1998, 39, 2933−2936. (b) Evans, D. A.; Katz, J. L.;
West, T. R. Tetrahedron Lett. 1998, 39, 2937−2940. (c) Lam, P. Y.;
Clark, C. G.; Saubern, S.; Adams, J.; Winters, M. P.; Chan, D. M.;
Combs, A. Tetrahedron Lett. 1998, 39, 2941−2944. (d) Bakkestuen, A.
K.; Gundersen, L.-L. Tetrahedron Lett. 2003, 44, 3359−3362. (e) Niu,
H.-Y.; Xia, C.; Qu, G.-R.; Zhang, Q.; Jiang, Y.; Mao, R.-Z.; Li, D.-Y.;
Guo, H.-M. Org. Biomol. Chem. 2011, 9, 5039−5042. (f) Engel-
D
DOI: 10.1021/acs.orglett.7b03209
Org. Lett. XXXX, XXX, XXX−XXX