Synthesis by conjugate radical addition of new heterocyclic amino acids with nucleic acid bases in their side chains

Article abstract of DOI:10.1039/b006843h

N-(2-Iodoethyl) and N-(3-iodopropyl)pyrimidines and purines undergo stereoselective conjugate radical addition with an optically active oxazolidinone acceptor to give syn-adducts that can be converted into pyrimidine and purine amino acids.

Full text of DOI:10.1039/b006843h

Synthesis by conjugate radical addition of new heterocyclic amino acids with
nucleic acid bases in their side chains
Raymond C. F. Jones,*† Didier J. C. Berthelot and James N. Iley*
Department of Chemistry, The Open University, Walton Hall, Milton Keynes, UK MK7 6AA
Received (in Cambridge, UK) 21st August 2000, Accepted 20th September 2000
First published as an Advance Article on the web 13th October 2000
N-(2-Iodoethyl) and N-(3-iodopropyl)pyrimidines and pur-
ines undergo stereoselective conjugate radical addition with
an optically active oxazolidinone acceptor to give syn-
adducts that can be converted into pyrimidine and purine
amino acids.
The radical precursors were haloalkyl pyrimidines and
purines, prepared from the appropriately protected heterocyclic
base and an w-haloalcohol.11 Thus 3-benzoylthymine 912 was
coupled with 2-bromoethanol or 3-bromopropanol (DIAD,
Ph3P) to afford the 1-(w-bromoalkyl) derivatives 10a,b,
respectively (Scheme 3). Attempts to generate radicals from
these bromides proved fruitless, so they were converted directly
to the iodoalkyl compounds 10c,d, respectively (NaI, propa-
none reflux; 85, 87% from 9).
Peptide-based nucleic acid analogues (PNAs) have attracted
much attention as molecules with the potential to interact with
nucleic acid chains.1 Suggested applications include antisense
properties.2 Nielsen’s PNA has been shown to form duplexes
with the complementary DNAs.1 DNA recognition using
analogues with a ‘real’ peptide backbone has, however, proved
more elusive. Substituted alanine oligomers 1 (B = pyrimidine
or purine base) and homologues 2 (termed a-PNA3) do not
demonstrate hybridisation with DNA and insufficient flexibility
of the polypeptide chain has been suggested as the cause,4
whereas triplex formation between tetrapeptides of type 2 and
poly(dT) or poly(dU) has been reported.5 Our interest in
unusual amino acids led us to propose the homologous amino
acids 4 carrying the nucleic acid bases with a 3- or 4-methylene
tether to the peptide backbone, as components for PNA variant
3. Residues 4 are also analogues of natural pyrimidine and
purine amino acids.6 We report here our flexible methodology
based on stereospecific radical chemistry.7
Scheme 3 Reagents: i, HO(CH2)n+1Br, PriO2CNNNCO2Pri, Ph3P; ii, Nal,
Me2CO reflux; iii, Method A: 10 (1 mol equiv.), Bu3SnH (1 mol equiv.),
AIBN (0.1 mol equiv.), toluene reflux; Method b: 8 (2 mol equiv.), Bu3
SnCl (0.3 mol equiv.), NaBH3CN (2 mol equiv.), AIBN (0.1 mol equiv.).
In contrast to published routes to residues with C2 tethers,3,5,8
we determined to link preformed heterocycles with the peptide
backbone by forming a carbon–carbon bond in the tether, and
proposed to generate the C(b)–C(g) bond by conjugate radical
addition to chiral acceptor 7 (Scheme 1).9 The (S)-acceptor 8
was prepared from S-methyl-(R)-cysteine (Scheme 2) by
adaptation of a published sequence to the N-benzoyl analogue.10
syn-Sulfone 7 was formed as a 10+1 mixture with its anti-
diastereoisomer 6 [57% overall from S-methyl-(R)-cysteine]
and easily separated by column chromatography.10 The syn-
configuration was supported inter alia by mutual NOE
enhancements between C-2(H) and C-4(H). Base treatment
afforded (S)-oxazolidinone 8 as a crystalline solid.
Iodide 10c was treated under two protocols differing in the
method for radical generation;9 method A: with oxazolidinone
8 (1 mol equiv.) in toluene at reflux containing AIBN (0.1 mol
equiv.) and dropwise addition of Bu3SnH (1 mol equiv.); or
method B: with oxazolidinone 8 (2 mol equiv.), Bu3SnCl (0.3
mol equiv.), NaBH3CN (2 mol equiv.) and AIBN (0.1 mol
equiv.) in tert-BuOH at reflux. Method A afforded the
conjugate addition product 11a (26%) and reduction product
12a (24%), whereas method B after 16 h afforded 54% of
conjugate addition products consisting of the adduct 11a (24%)
and the 3-debenzoylated derivative 11b (30%), with no reduced
material. The extent of debenzoylation was time dependent; a
reaction time of 40 h led to 11b as the sole addition product
(47%). This suggests the deacylation may be via hydride-
mediated reduction of the out-of-plane benzoyl carbonyl group,
a possibility supported by an observed decrease in debenzoyla-
tion when less NaBH3CN is used in method B, and that
debenzoylation of 10 occurs in the presence of NaBH3CN
alone.13 When 1-iodopropylthymine derivative 10d was treated
under method B, adduct 11c was not found and deacylated
adduct 11d was isolated (25%) along with reduction product
12b (75%).
Scheme 1
We elected to extend these standard protocols (method B
preferred) to other pyrimidines and purines rather than optimise
each conjugate addition. Thus 3-benzoyluracil 1312 was
Scheme 2 Reagents: i, NaOH aq.; ButCHO, Dean-Stark; iii, PhCH2OCOCl
(ZCl); iv, oxone®, MeCN–H2O; v, DBU.
† Current address: Department of Chemistry, Loughborough University,
Leicestershire, UK LE11 3TU; e-mail r.c.f.jones@lboro.ac.uk
DOI: 10.1039/b006843h
Chem. Commun., 2000, 2131–2132
This journal is © The Royal Society of Chemistry 2000
2131

View Article Online
Scheme 4 Reagents: i, ii as Scheme 3; iii, Method B.
converted into the 1-(iodoalkyl) derivatives 14a,b (Scheme 4).
Method B applied to 14a afforded addition product 15a (41%)
and reduction product 12c (46%); when the reaction was left for
2 days, deacylated addition product 15b (51%) was isolated.
Homologue 14b gave debenzoylated adduct 15d (44%) with
reduced material 12d (51%).‡ In the purine series, the
9-(iodoalkyl)adenines 17a,b were prepared from (2-methylpro-
pionyl)adenine 1614 (Scheme 5). Using method B, iodoethyl
compound 17a led to the expected mixture of conjugate
addition [40%; acylated 18a (26%) and deacylated 18b (14%)]
and reduction [36%; acylated 19a (17%) and deacylated 19b
(19%)]. Iodopropyl derivative 17b likewise gave adducts [22%;
acylated 18c (12%) and deacylated 18d (10%)] and reduced
compounds [34%; acylated 19c (11%) and deacylated 19d
(23%)]. Finally, a protected guanine 20a15 was converted into
the 9-iodoethyl derivative 20b (Scheme 6) and method A led to
adduct 21 (21%) and reduction to 20c (20%).
acids 22h–n, 23e–g and 24b, analysed by esterification (AcCl,
EtOH, reflux) and subsequent conversion to the Mosher amides
(R-3,3,3-trifluoro-2-methoxy-2-phenylpropanoyl chloride, pyr-
idine);16 19F NMR spectroscopy revealed, e.g. 86–91% e.e. for
the amino acids 22i,j,l,m, and 23f.
We have thus made available a range of novel pyrimidinyl
and purinyl amino acids for application, for example, in PNA
variants.
We thank the Open University for financial support (com-
petitive studentship to D. J. C. B.) and the EPSRC National
Mass Spectrometry Service Centre (Swansea) for some MS
data.
Notes and references
‡ The yield of 15d could be increased to 62% by using 5 mol equiv. of
acceptor 8 in method B, but we more usually used 2 mol equiv. of this
valuable optically active intermediate. When less than 2 mol equiv.
NaBH3CN was used, some of the benzoylated adduct 15c was isolated.
1 For recent reviews, see: B. Hyrup and P. E. Nielsen, Bioorg. Med.
Chem., 1996, 4, 5; P. E. Nielsen and G. Haaima, Chem. Soc. Rev., 1997,
26, 73.
2 H. K. Larsen, T. Bentin and P. E. Nielsen, Biochim. Biophys. Acta, 1999,
1489, 159.
Scheme 5 Reagents: i, ii as Scheme 3; iii, Method B.
3 N. M. Howarth and L. P. G. Wakelin, J. Org. Chem., 1997, 62, 5441.
4 See: M. Kuwahara, M. Arimitsu and M. Sisido, Tetrahedron, 1999, 55,
10067, and refs. therein.
5 T. Yamakazi, K. Komatsu, H. Umemiya, Y. Hashimoto, K. Shudo and
H. Kagechika, Tetrahedron Lett., 1997, 38, 8363.
6 See, for example: R. M. Adlington, J. E. Baldwin, D. Catterick and G. J.
Pritchard, J. Chem. Soc., Perkin Trans. 1, 1999, 855.
7 See: C. J. Easton, Chem. Rev., 1997, 97, 53, for a review of radical
reactions in amino acid synthesis.
8 A. Lenzi, G. Reginato and M. Taddei, Tetrahedron Lett., 1995, 36,
1713.
Scheme 6 Reagents: i, ii as Scheme 3; iii, Method A.
9 J. R. Axon and A. L. J. Beckwith, J. Chem. Soc., Chem. Commun., 1995,
549, and refs. therein.
10 S. G. Pyne, B. Dikic, P. A. Gordon, B. W. Skelton and A. H. White,
Aust. J. Chem., 1993, 46, 73.
11 Cf. G. Cadet, C.-S. Chan, R. Y. Daniel, C. P. Davis, D. Guiadeen, G.
Rodriguez, T. Thomas, S. Walcott and P. Scheiner, J. Org. Chem., 1998,
63, 4574.
12 K. A. Cruickshank, J. Jiricny and C. B. Reese, Tetrahedron Lett., 1984,
25, 681.
13 Cf. S. Kim, T. A. Lee and Y. Song, Synlett, 1998, 471, for radical
addition to imides.
14 J. Zhou, K. Bouhadir, T. R. Webb and P. B. Shevlin, Tetrahedron Lett.,
1997, 38, 4037.
15 J. Zhou, J.-Y. Tsai, K. Bouhadir and P. B. Shevlin, Synth. Commun.,
1999, 29, 3003.
16 J. A. Dale, D. L. Dull and H. S. Mosher, J. Org. Chem., 1969, 34,
2543.
The illustrated conjugate radical addition products were all
syn-adducts, as determined by NOE studies [enhancements
between C-2(H) and C-4(H)]. Only one diastereoisomer was
1
visible in the H NMR spectra at 300 MHz. All of these syn-
oxazolidinones could be easily and efficiently converted into N-
benzyloxycarbonyl-(S)-amino acids (suitable for peptide cou-
pling) by base hydrolysis (LiOH, aq. THF, 0 °C, 30–60 min;
70–98%). Thus the three thymine-substituted Z-amino acids
22a–c (having 3- or 4-carbon tethers for the pyrimidine) were
prepared from the adducts 11a,b,d, respectively. The uracil Z-
amino acids 22d–g were likewise prepared from adducts 15a–d,
respectively, as were adenine derivatives 23a–d (from 18a–d,
respectively) and guanine Z-amino acid 24a (from 21). To
monitor optical purity, the Z group was removed by hydro-
genolysis (Pd–C, EtOH–H2O; 60–80%) to afford the amino
2132
Chem. Commun., 2000, 2131–2132