Hydrophilic modifications in peptide nucleic acid - Synthesis and properties of PNA possessing 5-hydroxymethyluracil and 5-hydroxymethylcytosine

Article abstract of DOI:10.1139/V07-030

We have investigated the chemistry for the incorporation of C5-hydroxymethyl-uracil and -cytosine in peptide nucleic acid (PNA) and the subsequent effect of this modification on PNA hybridization behavior. Largely based on literature precedent, we prepared a peptide nucleic acid monomer, possessing 5-hydroxymethyuracil, which was compatible with Fmoc-based oligopeptide synthesis. An improved, large-scale synthesis of 5-hydroxymethylcytosine was developed, as a starting point for the synthesis of a monomer containing this nucleobase. In each case, the hydroxyl group was blocked as a t-butyldiphenylsilyl ether, and the exocyclic amino group of cytosine was additionally blocked with the benzoyl-group. The modified monomers were incorporated into isolated positions in the oligomer sequence using standard protocols. The modified oligomers showed that the 5-hydroxymethyl group is compatible with triplex and duplex formation.

Full text of DOI:10.1139/V07-030

302
Hydrophilic modifications in peptide nucleic acid —
Synthesis and properties of PNA possessing 5-
hydroxymethyluracil and 5-hydroxymethylcytosine1
Robert H.E. Hudson, Yuhong Liu, and Filip Wojciechowski
Abstract: We have investigated the chemistry for the incorporation of C5-hydroxymethyl-uracil and -cytosine in pep-
tide nucleic acid (PNA) and the subsequent effect of this modification on PNA hybridization behavior. Largely based
on literature precedent, we prepared a peptide nucleic acid monomer, possessing 5-hydroxymethyuracil, which was
compatible with Fmoc-based oligopeptide synthesis. An improved, large-scale synthesis of 5-hydroxymethylcytosine
was developed, as a starting point for the synthesis of a monomer containing this nucleobase. In each case, the
hydroxyl group was blocked as a t-butyldiphenylsilyl ether, and the exocyclic amino group of cytosine was additionally
blocked with the benzoyl-group. The modified monomers were incorporated into isolated positions in the oligomer se-
quence using standard protocols. The modified oligomers showed that the 5-hydroxymethyl group is compatible with
triplex and duplex formation.
Key words: peptide nucleic acid, hydroxymethyluracil, hydroxymethylcytosine, modified nucleobase, hybridization.
Résumé : On a étudié la chimie de l’incorporation d’un C5-hydroxyméthyl-uracile et d’une -cytosine dans un acide
nucléique peptidique (ANP) ainsi que l’effet de cette modification sur le comportement d’hybridisation de l’ANP. En se
basant principalement sur les précédents trouvés dans la littérature, on a préparé un acide nucléique peptidique mono-
mère possédant un groupe 5-hydroxyméthyluracile qui était compatible avec une synthèse d’oligopeptide basée sur
l’utilisation d’un groupe Fmoc. On a développé une synthèse à grande échelle améliorée de la 5-hydroxyméthylcytosine
qui a été utilisée comme point de départ pour la synthèse d’un monomère contenant cette nucléobase. Dans chaque cas,
le groupe hydroxyle a été bloqué par un éther t-butyldiphénylsilyle et on a de plus bloqué le groupe amino exocyclique
de la cytosine à l’aide d’un groupe benzoyle. On a incorporé les monomères modifiés dans des positions isolées de la
séquence de l’oligomère en faisant appel aux protocoles habituels. Les oligomères modifiés ont permis de montrer que
le groupe 5-hydroxyméthyle est compatible avec la formation d’un triplex et d’un duplex.
Mots-clés : acide nucléique peptidique, hydroxyméthyluracile, hydroxyméthylcytosine, nucléobase modifiée, hybridisation.
[Traduit par la Rédaction] Hudson et al. 312
Introduction
both in vitro diagnostic applications and in vivo therapeutic
use (2).
Ready access to synthetic oligonucleotides has enabled
much of the tremendous progress witnessed in molecular bi-
ology, forensics, diagnostics, and related areas over the past
couple of decades. Some applications of oligonucleotides,
especially in hostile environments containing DNA or RNA
degrading enzymes, require structural modifications to resist
degradation. Out of endeavors to make robust oligo-
nucleotide analogs, peptide nucleic acid (PNA) was born (1).
PNA is an oligonucleotide mimic based on a polyamide re-
peat unit. It exhibits very tight and specific binding to com-
plementary nucleic acids and continues to be of interest for
Although the original design of PNA “works” quite well
(see Fig. 1), possessing excellent hybridization affinity and
selectivity, often-cited limitations are its poor water solubil-
ity, poor ability to remain in solution and a tendency to be
adsorbed onto surfaces, sequence-dependent aggregation,
and a general difficulty in penetrating cellular membranes.
These biophysical properties are related in various degrees
to the non-ionic structure of PNA. Judicious chemical modi-
fications can improve these biophysical properties while
maintaining the highly desirable hybridization properties.
For example, the aqueous solubility of PNA is improved by
the attachment of a cationic amino acid, such as lysine, to
one or both ends of the oligomer (3). Alternatively, introduc-
tion of charged (guanidinium) groups along the backbone
structure improves the aqueous solubility and also facilitates
cellular-membrane penetration (4). Another strategy to im-
prove the solubility of PNA is to introduce hydrophilic
nucleobase replacements (5).
Received 22 December 2006. Accepted 5 February 2007.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 5 May 2007.
R.H.E. Hudson,2 Y. Liu, and F. Wojciechowski.
Department of Chemistry, The University of Western Ontario,
London, ON N6A 5B7, Canada.
1This article is part of a Special Issue dedicated to Professor
Kelvin K. Ogilvie.
The modular nature of PNA is particularly amenable to
structural variation, and this has led to the synthesis of a
wide variety of modified PNAs. We have chosen to investi-
2Corresponding author (e-mail: rhhudson@uwo.ca).
Can. J. Chem. 85: 302–312 (2007)
doi:10.1139/V07-030
© 2007 NRC Canada

Hudson et al.
303
Fig. 1. Structures of DNA and PNA.
of signals are indicated where appropriate (s: single, d: dou-
blet, t: triplet, q: quartet, m: mulitplet, br: broad, v: very).
Melting points were determined on a Gallenkamp melting-
point apparatus and are reported uncorrected.
5-Hydroxymethyluracil (1)
The synthesis of 5-hmU followed our previously reported
method (7). Specifically, triethylamine (42 mL, 300 mmol)
was added to a suspension of uracil (22.4 g, 200 mmol) and
paraformaldehyde (18.0 g, 600 mmol) in water (600 mL).
The mixture turned into a clear solution after heating to
60 °C. The solution was stirred at 60 °C overnight after
which time the water was removed under vacuum until the
volume reached 60 mL. To this residue, 60 mL of ethanol
(95%) was added, and the resultant solution was left to stand
at room temperature for 30 min. Once a precipitate formed,
the mixture was placed in the refrigerator at 4 °C for several
hours, a white solid was collected by filtration, and was
washed with cold ethanol. The filtrate was cooled to –20 °C,
and a second crop of 1 was collected. Further crops were
yielded by reducing the volume of the filtrate by one half
and diluting with an equal volume of 95% ethanol, followed
by cooling to –20 °C. White solid; 26.1 g (92%); mp (from
ethanol) 360 °C dec (lit. 300 °C dec.) (13). TLC (CHCl3–
gate nucleobase modifications as a way to introduce addi-
tional functionality, which may be distributed throughout the
PNA oligomer without changing the N-2-aminoethylglycine
backbone structure. Our approach, as reported herein, was to
make a conservative structural modification that added hy-
drophilic character, which was synthetically accessible and
would be sterically tolerated in a duplex or triplex structure.
The 5-hydroxymethyl modification of pyrimidines (Fig. 2)
fulfilled these requirements.
The 5-hydroxymethyl modification constitutes a relatively
minor structural change to the nucleobases, representing the
replacement of a hydrogen on 5-methyluracil (thymine) or 5-
methylcytosine by a hydroxyl. The resulting 5-substituent is
placed in the major groove and is expected to be sterically
well tolerated (6). Other advantages are that the chemistry is
accessible and starts with inexpensive reagents, and substitu-
tion does not produce a new stereocenter and thereby avoids
problems of enantiomeric purity in the product.
1
CH3OH, 4:1): Rf 0.16. H NMR (400.1 MHz, DMSO-d6,
ppm) δ: 11.04 (s, 1H, NH), 10.69 (d, J = 5.20 Hz, 1H, NH),
7.22 (d, J = 5.20 Hz, 1H, H6), 4.85 (t, J = 5.60 Hz, 1H,
OH), 4.09 (d, J = 4.80 Hz, 2H, –CH2–). 13C NMR
(100.6 MHz, DMSO-d6, ppm) δ: 163.6, 151.2, 138.0, 112.6,
55.8. HR-MS (EI) calcd. for C5H6O3N2: 142.0378; found:
142.0376.
Ethyl(5-hydroxymethyluracil-1-yl)acetate (2)
To a suspension of 5-hydroxymethyluracil (22.4 g,
158 mmol) and K2CO3 (21.8 g, 158 mmol) in DMF
(475 mL) was added ethyl bromoacetate (98%, 28.3 g,
166 mmol). The reaction mixture was stirred overnight after
which time the K2CO3 was removed by filtration, and the
DMF was removed under vacuum, thus yielding a white
solid. The solid was washed with H2O (180 mL) by stirring
the slurry for approximately 2 h. The first crop of 2 (21.4 g)
was obtained by filtration and drying. Further crops were re-
covered from the filtrate by reducing its volume. White
solid; 82% (29.5 g); mp 178–180 °C. TLC (CHCl3 – CH3OH,
Experimental
General
All reagents and solvents were obtained from common
commercial sources and used without further purification,
unless indicated. Anhydrous solvents were prepared by pass-
ing the reagent-grade solvent through columns of activated
alumina (Innovative Technologies). Chromatography was
carried out using silica gel 60 (particle size 0.063–
0.200 mm; Merck). TLC was carried out on plastic sheets,
silica gel 60 F254, layer thickness of 0.2 mm (Merck). NMR
1
4:1): Rf 0.46. H NMR (400.1 MHz, DMSO-d6, ppm) δ:
11.42 (s, NH), 7.53 (s, 1H, H6), 5.07 (t, J = 5.60 Hz, 1H,
OH), 4.54 (s, 2H, –CH2–), 4.13 (m, 4H, overlapping –OCH2–),
1.21 (t, J = 7.20 Hz, 3H, –CH3). 13C NMR (100.6 MHz,
DMSO-d6, ppm) δ: 168.1, 163.0, 150.7, 141.7, 113.4, 61.2,
55.9, 48.8, 14.2. HR-MS (EI) calcd. for C9H12O5N2:
228.0746; found: 228.0748.
1
spectra were recorded on Mercury 400 (400.1 MHz for H;
100.6 MHz for 13C) and Inova 400 spectrometers. Chemical
shift values, reported in ppm, were referenced by the spec-
trometer to the residual proton in the deuterated solvents
used to record the spectra. Resonances, attributed to con-
formers due to restricted rotation around amide bonds (i.e.,
rotamers), are reported as major (ma) and minor (mi) in the
1H and 13C NMR spectra. Assignments are shown for se-
lected resonances that have been unambiguously identified
by the use of 2D NMR correlation experiments. Multiplicity
Ethyl[5-(t-butyldiphenylsilanyloxymethyl)uracil-1-
yl]acetate (3)
To a solution of ethyl(5-hydroxymethyluracil-1-yl)acetate
(5.01 g, 22 mmol) and imidazole (3.29 g, 48 mmol) in DMF
(110 mL) was added tert-butyldiphenylchlorosilane
(TBDPS-Cl, 98%, 6.64 g, 23.7 mmol) with stirring. The so-
lution was stirred overnight after which time the reaction
was stopped by adding 5% NaHCO3 in H2O (11 mL). The
solvents were removed under vacuum to give a white solid.
© 2007 NRC Canada

304
Can. J. Chem. Vol. 85, 2007
Fig. 2. (a) Expected base-pairing of adenine (A, DNA strand) with 5-hydroxymethyluracil-containing PNA in duplex formation (black)
and triplex formation (black and grey). (b) Expected base-pairing between guanine (G, DNA strand) with 5-hydroxymethylcytosine-
containing PNA in duplex formation (black) and triplex formation (black and grey).
This material was dissolved in dichloromethane (250 mL),
washed with H2O (100 mL × 3), and the organic phase was
separated and dried over sodium sulfate. The crude product
was obtained after removing the dicholormethane by vac-
uum and was subsequently purified by trituration with hex-
anes to remove the silanol byproduct (TBDPS-OH; Rf 0.96;
ethyl acetate–DCM, 1:2). The yield was 99% (10.2 g); mp
added aq. NaHCO3 (3 equiv., 7.4 g/120 mL) with rapid stir-
ring for 50 min. The reaction mixture was diluted with
400 mL H2O, which produced an instant white precipitate
that redissolved on the addition of DCM (350 mL). This
aqueous phase was extracted with DCM (2 × 300 mL), and
then the DCM phase was dried over sodium sulfate, acidi-
fied with Et2O – HCl (5 mol/L), and cooled in the freezer
overnight. The desired product was collected by filtration:
white solid; 7.0 g, 61%; mp 110–113 °C. Further product
can be obtained by reducing the volume of the filtrate. TLC
(CH3COOEt – CH3OH, 95:5; visualized with UV and 0.3%
1
168–170 °C. TLC (ethyl acetate – DCM, 1:2): Rf 0.71. H
NMR (400.1 MHz, DMSO-d6, ppm) δ: 11.52 (s, NH), 7.65–
7.67 (m, 4H, Ar–H), 7.62 (s, 1H, C6–H), 7.43–7.48 (m, 6H,
Ar–H), 4.59 (s, 2H, N–CH2–), 4.36 (s, 2H, O–CH2Ar), 4.16
(q, J = 7.20 Hz, 2H, O–CH2–), 1.21 (t, J = 7.20 Hz, 3H, C–
CH3), 1.00 (s, 9H, SiC–CH3). 13C NMR (100.6 MHz,
DMSO-d6, ppm) δ: 168.3 (C7), 163.0 (C2), 150.8 (C1),
142.6 (C4), 135.2 (C13), 132.7 (C12), 130.0 (C15), 128.0
(C14), 112.0 (C3), 61.3 (C8) 58.9 (C5), 48.9 (C6), 26.7
(C11), 18.9 (C10), 14.1 (C9). HR-MS (EI) calcd. for
C25H30O5N2Si: [M + H]+, 467.2002; found: 467.1994.
1
ninhydrin (purple)): Rf 0.24. H NMR (400.1 MHz, DMSO-
d6, ppm) δ: 9.44 (br s, 2H, H2N+), 7.90 (d, J = 7.20 Hz, 2H,
Ar), 7.70 (d, J = 7.20 Hz, 2H, Ar), 7.58 (t, J = 5.60 Hz,
CONH), 7.42 (t, J = 7.20 Hz, 2H, Ar), 7.33 (t, J = 7.60 Hz,
2H, Ar), 4.33 (d, J = 6.80 Hz, 2H, –CH2–O), 4.23 (t, J =
6.80 Hz, 1H, Ar2–CH–), 4.00 (s, 2H, N–CH2–), 3.74 (s, 3H,
O–CH3), 3.33 (overlapping with water, –CH2–CH2–), 3.02
(t, J = 6.00 Hz, 1.8H, ma, –CH2–CH2–), 2.90 (mi, 0.2H). 13
C
NMR (100.6 MHz, DMSO-d6, ppm) δ: 167.1, 156.3, 143.9,
140.8, 127.7, 127.2, 125.3, 120.2, 65.7, 52.7, 46.7, 46.6,
46.5, 36.6. MS (EI) calcd. for [C20H22O4N2HCl – HCl]+:
354.16; found: 354.1.
Byproducts identified as dibenzofulvene: Rf 0.96 (ethyl
acetate – methanol, 95:5), and 2-oxo-3-N-(Fmoc)piperazine:
Rf 0.42 (CHCl3 – CH3OH – CH3COOH, 90:8:2). Mp 172–
174 °C.
[5-(t-butyldiphenylsilanyloxymethyl)uracil-1-yl]acetic
acid (4)
To
a
solution of ethyl[5-(t-butyldiphenylsilanyloxy-
methyl)uracil-1-yl]acetate (10.6 g, 22.7 mmol) in THF
(204 mL) and H2O (22 mL) at 0 °C was added 2.5 mol/L
NaOH (47 mL, 6 equiv.). The mixture was stirred at 0 °C for
20 min, then 1 mol/L HCl (170 mL) was added to stop the
reaction. The aqueous phase was extracted with ethyl acetate
four times (400 mL × 1, 100 mL × 3). The ethyl acetate
phase was washed with H2O (100 mL × 2) and then dried
over sodium sulfate. The solvent was removed under vacuum
to yield the title compound: white solid; 8.2 g, 96%; mp
Methyl(2-(((9H-fluoren-9-yl)methoxy)car-
bonyl)aminoethyl)glycinate (6)
The hydrochloride salt of the title compound (3.0 g,
7.8 mmol) was suspended in dichloromethane (350 mL) and
shaken against a saturated aq. NaHCO3 solution (150 mL ×
3) to liberate the free base. The phases were separated, and
the combined aqueous phases were back-extracted with di-
chloromethane (100 mL × 3). The organic phase was dried
over sodium sulfate, and the solvent was removed under vac-
uum to yield the free base as a white wax-like product:
2.70 g (99%); mp 76–78 °C. TLC (ethyl acetate – methanol,
95:5; visualized by UV and 0.3% ninhydrin (purple)): Rf
1
169–170 °C. TLC (ethyl acetate – DCM, 1:2): Rf 0.71. H
NMR (400.1 MHz, DMSO-d6, ppm) δ: 13.14 (br s, 1H,
COOH), 11.47 (s, 1H, NH), 7.65–7.67 (m, 4H, Ar), 7.62 (s,
1H, C6–H), 7.41–7.50 (m, 6H, Ar), 4.49 (s, 2H, N–CH2–),
4.35 (s, 2H, O–CH2–), 1.00 (s, 9H, –CH3). 13C NMR
(100.6 MHz, DMSO-d6, ppm) δ: 169.8, 163.0, 150.9, 142.8,
135.2, 132.8, 130.0, 128.0, 111.8, 58.9, 48.8, 26.7, 18.9. MS
(ESI) calcd. for [C23H26O5N2Si + Na]+: 461.15; found: 461.1.
1
Methyl(N-(2-((9H-fluoren-9-yl)methoxy)car-
bonyl)aminoethylglycinate hydrochloride (5)
To a suspension of methyl 2-aminoethylglycinate dihydro-
chloride (6.0 g, 20.8 mmol) and Fmoc-N-hydroxy-
succinimide (10.8 g, 1.1 equiv.) in dioxane (100 mL) was
0.30. H NMR (400.1 MHz, CDCl3, ppm) δ: 7.76 (d, J =
7.60 Hz, 2H, Ar), 7.61 (d, J = 7.60 Hz, 2H, Ar), 7.40 (t, J =
7.60 Hz, 2H, Ar), 7.32 (t, J = 7.20 Hz, 2H, Ar), 5.36 (br, ma,
CONH), 5.12 (br, mi, CONH), 4.40 (d, J = 7.20 Hz, 2H, –
CH2–O), 4.22 (t, J = 6.80 Hz, 1H, Ar2–CH–), 3.73 (s, 3H,
© 2007 NRC Canada

Hudson et al.
305
Fig. 3. Structure of the 5-hydroxymethyluracil (hmU) PNA
monomer ester (R = –CH3) and acid (R = –H), defining the arbi-
trary numbering system used in NMR signal assignments.
O–CH3), 3.42 (s, 2H, N–CH2–C(O)), 3.27 (q, J = 5.60 Hz,
ma, CONH–CH2–), 3.17 (mi, CONH–CH2–), 2.75 (t, J =
5.60 Hz, ma, –CH2–N), 2.62 (mi, –CH2–N). 13C NMR
(100.6 MHz, CDCl3, ppm) δ: 173.1, 156.7, 144.2, 141.5,
127.8, 127.2, 125.2, 120.1, 40.8, 66.7, 52.0, 50.4, 48.8, 47.4.
HR-MS (EI) calcd. for C20H22O4N2: 354.1580; found:
354.1581.
Fmoc hmU-OMe (7)
To a suspension of methyl(2-(((9H-fluoren-9-yl)meth-
oxy)carbonyl)aminoethyl)glycinate (6) (1.61 g, 4.5 mmol)
and the nucleobase acetic acid derivative 4 (3.18 g,
7.3 mmol) in DMF (45 mL) was added 1-ethyl-3-(3′-
dimethylaminopropyl)carbodiimide hydrochloride (EDC,
1.39 g, 7.3 mmol) in two portions within 0.5 h. The mixture
was stirred for 16 h at room temperature, and then the reac-
tion mixture was diluted with dichloromethane (1000 mL).
This solution was washed with 0.05 mol/L HCl (200 mL ×
4), saturated NaHCO3 (200 mL × 3), and H2O (450 mL × 7),
respectively. The organic layer was separated and dried over
Na2SO4, and the solvent was removed under vacuum at
room temperature. The crude product (white foam) was puri-
fied by gradient column chromatography (DCM, ethyl ace-
tate, ethyl acetate – methanol (95:5), respectively) to yield
the title compound as a white foam (Fig. 3): 3.10 g, 88%
yield. TLC (ethyl acetate – methanol, 95:5; visualized by
UV and 0.3% ninhydrin (purple)): Rf 0.63. 1H NMR
(400.1 MHz, CDCl3, ppm) δ: 7.19–7.77 (m, overapping 19H,
Ar), 6.06 (br t, ma, backbone NH) and 5.45 (br t, mi, back-
bone NH), 4.51–4.54 (br s, overlap H5 and its rotamer),
4.51–4.54 (br s, ma, H6) and 4.37–4.44 (overlap, mi, H6),
4.37–4.44 (overlap, H23 and its rotamer), 4.17–4.20 (m,
H22), 4.17–4.20 (mi, H27) and 4.07 (s, ma, H27),3 3.74 (s,
ma, H29) and 3.78 (s, mi, H29), 3.57 (br s, ma, 1.6 H, H26)
and 2.88 (br s, mi, 0.4 H, H26), 3.45 (br m, ma, H25) and
3.38 (br m, mi, H25), 1.08 (s, ma, H11) and 1.09 (s, mi,
H11). 13C NMR (100.6 MHz, CDCl3, ppm) δ: 170.1 (C28),
167.3 (C7), 162.7 (C2), 157.0 (C24), 151.2 (C1), 144.0
(C16), 141.4 (C17), 141.2 (C4), 135.7 (C13), 133.0 (C12),
130.0 (C15), 128.0 (C14), 127.8 (C19), 127.2 (C20), 125.2
(C21), 120.1 (C18), 114.2 (C3), 66.9 (C23), 59.1 (C5), 53.0
(r), 52.6 (C29), 50.5 (r), 49.0 (C27), 48.9 (C26), 48.6 (r),
48.5 (C6), 47.3 (C22), 39.2 (r), 39.6 (C25), 27.1 (C11), 19.4
(C10). HR-MS (ESI-TOF) calcd. for [C43H46O8N4Si + Na]+:
797.2983; found: 797.3022.
removed under vacuum at room temperature to yield an off-
white (yellowish) foam (Fig. 3). The product was purified
by column chromatography using a gradient of solvents
(DCM; DCM–methanol (90:10), (85:15), and (78:22); re-
spectively) to give the title compound as an off-white solid:
2.60 g, 90%; mp 216–21s7 °C. TLC (CHCl3 – CH3OH –
CH3COOH, 90:8:2; visualized with UV and 0.3% ninhydrin
1
(grey)): Rf 0.08. H NMR (400.1 MHz, DMSO-d6, ppm) δ:
11.39 (br s, uracil NH) and 11.36 (br s, uracil NH), 7.29–
7.88 (overlapping signals, 18H Ar, backbone NH), 4.75 (s,
mi, 0.2H, H6) and 4.56 (s, ma, 1.8H, H6), 4.23–4.35 (m,
overlap, 2H, H5) and (r, H23), 4.17–4.20 (m, overlap, H23,
H22), 3.95 (s, 2H, H27),3 3.44 (br m, mi, H26) and 3.38
(br m, ma, H26), 3.36 (br m, mi, H25) and 3.16 (br m, ma,
H25), 0.98 (s, 9H, H11). 13C NMR (100.6 MHz, DMSO-d6,
ppm) δ: 171.5 (C28), 166.9 (C7), 163.0 (C2), 156.2 (C24),
150.9 (C1), 143.9 (C16), 143.3 (C4), 140.7 (C17), 135.2
(C13), 132.8 (C12), 129.9 (C15), 128.0 (C14), 127.6 (C19),
127.1 (C20), 125.3 (C21), 120.1 (C18), 111.4 (C3), 65.6
(C23), 58.9 (C5), 51.4 (r), 48.5 (C27), 48.1 (C6), 47.6 (r),
46.9 (C26), 46.7 (C22), 38.7 (r), 38.0 (C25), 26.7 (C11),
18.8 (C10). HR-MS (ESI-TOF) calcd. for [C42H44O8N4Si +
Na]+: 783.28261; found: 783.2841.
5-Hydroxymethylcytosine (9)
Fmoc-hmU-OH (8)
Triethylamine (16.5 mL, 118 mmol) was added to a sus-
pension of cytosine (5.27 g, 47.4 mmol) and paraform-
aldehyde (3.28 g, 109 mmol) in water (140 mL). The
mixture was stirred and heated under reflux for 72 h. After
24 h and 32 h of reaction, a second and third portion of
paraformaldehyde (1.14 g, 38.0 mmol) and (0.57 g,
19.0 mmol) was added, respectively. After completion of the
reaction, the volatile components (water and triethylamine)
were removed under vacuum to give a dark-coloured residue
(12.8 g), which represented greater than 100% theoretical
yield. This residue was dissolved to produce a clear solution
by heating with 35 mL of 95% ethanol and 18–24 mL H2O,
depending on solubility. The solution was left at room tem-
The ester precursor to the PNA monomer acid (2.93 g,
3.78 mmol) was dissolved in THF (32 mL), and then H2O
(21 mL) was added to produce a turbid mixture. The mixture
was cooled to 0 °C, and 2.5 mol/L NaOH (12 mL, 8 equiv.)
was added. The mixture was stirred for 5 min at 0 °C during
which time the mixture cleared to produce a homogeneous
solution. The reaction was stopped by addition to a separa-
tory funnel containing a mixture of ethyl acetate (400 mL)
and 0.3 mol/L HCl (200 mL). The acidic aqueous phase was
extracted by ethyl acetate (150 mL × 3). The ethyl acetate
washes were combined (~850 mL), washed with H2O
(100 mL × 3), and dried over Na2SO4. The ethyl acetate was
3 Variable position (3.85 – 4.20 ppm), depending on the protonation state of the carboxylic acid.
© 2007 NRC Canada

306
Can. J. Chem. Vol. 85, 2007
perature for a short time during which a precipitate formed.
Further precipitation was induced by cooling to 4 °C. The
first crop of 5-hydroxymethylcytosine, 4.2 g (5-hmC, yellow
solid, purity ≈ 90% (wt%), 7%–8% (wt%) cytosine, as deter-
O–CH2–Ar), 4.29 (s, 2H, N4–CH2–), 4.16 (q, J = 7.1, 4H,
2 × –CH2–), 1.22 (t, J = 7.2, 6H, 2 × –CH3,). 13C NMR
(100.6 MHz, DMSO-d6, ppm) δ: 185.2, 170.9, 167.9, 145.9,
144.6, 108.8, 61.6, 61.5, 56.1, 49.7, 49.0, 14.1.
1
mined by H NMR spectroscopy) was collected by filtration
and washed three times with 20 mL of 95% ethanol. This
first crop represents 57% chemical yield of 9 from cytosine.
Reduction in the volume of the filtrate gave subsequent
crops of 5-hmC. The later crops contained proportionally
more cytosine and other unidentified compounds. Nonethe-
less, these impure fractions of 5-hmC were suitable for fur-
ther use. The total yield of 5-hmC calculated from all
Ethyl[5-(t-butyldiphenylsilanyloxymethyl)cytosin-1-
yl]acetate (11)
Crude ethyl(5-hydroxymethylcytosin-1-yl)acetate (10),
which contained 5-hmC, was used as the starting material
for this transformation. As a byproduct, 5-(t-butyldiphenyl-
silanyloxymethyl)cytosine was obtained, which could be
alkylated by ethyl bromoacetate to give the title compound,
as a supplementary route to the following one.
1
recovered fractions, based on determining the purity by H
NMR spectroscopy, was 85%. This experiment was repeated
many times and proved to be reproducible. Mp 300 °C dec.
TLC (t-butyl alcohol – methyl ethyl ketone – H2O –
NH4OH, 4:3:2:1): Rf 0.32, compared with Rf 0.46 for cyto-
As an example, a particularly impure sample was utilized.
To a solution of crude ethyl(5-hydroxymethylcytosin-1-
yl)acetate (10) (5.08 g, 48% (wt%), ~11 mmol) and
imidazole (3.66 g, 54 mmol) in DMF (115 mL) was added
TBDPS-Cl (8 mL, 27.0 mmol) with stirring. The solution
was stirred overnight, and then a saturated solution of
NaHCO3 (10 mL) was added to stop the reaction. The DMF
and H2O were removed under vacuum to yield a dark-brown
oil, which was dissolved in ethyl acetate (300 mL) and
washed with H2O (150 mL × 4). Unreacted starting materi-
als (9 and 10), as well as dialkylated hmC, were removed by
the aqueous washes. Compound 9 was recovered by back-
extraction with ethyl acetate, the efficacy of which was mon-
itored by TLC. The ethyl acetate phase was dried over
Na2SO4, and the solvent was removed under vacuum to yield
compound 9 along with silylated 5-hmC. The title com-
pound 11 was isolated as a slightly orange solid, 40% yield,
by flash column chromatography, using a gradient of sol-
vents (hexane, ethyl acetate, and ethyl acetate – methanol
(96:4) for 9; ethyl acetate – methanol (75:25) for 5-(t-butyl-
diphenylsilanyloxymethyl)cytosine). Mp 183–185 °C, as iso-
lated from chromatography. TLC (ethyl acetate – methanol,
1
sine. H NMR (400.1 MHz, DMSO-d6, ppm) δ: 10.60 (v br,
ring NH), 7.28 (s, 1H, C6–H), 7.10 (br, NH2), 6.54 (br,
NH2), 4.99 (v br, OH), 4.16 (s, 2H, O–CH2–Ar). 13C NMR
(100.6 MHz, DMSO-d6, ppm) δ: 165.6, 157.0, 140.8, 104.9,
57.1. HR-MS (EI) calcd. for [C5H7O2N3 + H]+: 142.0616;
found; 142.0620.
Ethyl(5-hydroxymethylcytosin-1-yl)acetate (10)
To a suspension of crude 5-hydroxymethylcytosine (9)
(8.97 g, 68% (wt%), 43 mmol) and K2CO3 (5.97 g,
43.2 mmol) in dry DMF (140 mL), was added ethyl
bromoacetate (5.6 mL, 98% pure, 47.5 mmol). The reaction
mixture was stirred overnight, and a dark-coloured oil was
obtained after removal of the K2CO3 by filtration through a
pad of Celite and evaporation of DMF under vacuum. Ethyl
acetate (120 mL) was added to triturate the residue, which
was sonicated (often 1–2 h, sometimes up to 5 h), and then a
brown solid was isolated by filtration. The solid crude prod-
uct was washed with a small amount of ethyl acetate (50 mL ×
3). All of the mass was accounted for in the crude material,
which was found to contain the desired product (10, 60%
(wt%)) along with the dialkylated byproduct (19% (wt%)),
some unreacted starting material 9 (~20% (wt%); Rf 0.06;
DCM – CH3OH, 4:1), and small amounts of unidentified
1
95:5) Rf 0.18. H NMR (400.1 MHz, CDCl3, ppm) δ: 7.64
(d, J = 8.00 Hz, 4H, Ar), 7.38–7.48 (m, 6H, Ar), 6.48 (s, 1H,
C6–H), 4.42 (s, 2H, SiO–CH2–Ar), 4.31 (s, 2H, –CH2–N),
4.19 (q, J = 6.40 Hz, 2H, –O–CH2–), 1.26 (t, J = 6.40 Hz,
3H, –CH3), 1.04 (s, 9H, SiC–CH3). 13C NMR (100.6 MHz,
CDCl3, ppm) δ: 143.6, 168.2, 166.0, 156.2, 135.8, 132.7,
130.3, 128.1, 105.0, 62.0, 61.4, 50.2, 27.0, 19.3, 14.3. HR-
MS (EI) calcd. for [C25H31O4N3Si + H]+: 466.2162; found:
466.2155.
1
compounds, as estimated by examination of H NMR spec-
tra. The crude material was suitable for the next reaction.
The title compound 10 could also be purified by solvent gra-
dient silica-gel column chromatography (DCM; DCM –
CH3OH (90:10), (85:15), and (80:20); respectively). Mp
175–177 °C, as isolated from chromatography. TLC (DCM –
Identified
byproduct:
5-(t-butyldiphenylsilanyloxy-
methyl)cytosine; white solid. TLC (ethyl acetate – methanol,
1
95:5): Rf 0. H NMR (400.1 MHz, DMSO-d6, ppm) δ: 7.63
1
(d, J = 6.00 Hz, 4H, Ar), 7.42–7.50 (m, 6H, Ar), 7.10 (s, 1H,
C6–H), 4.46 (s, 2H, SiO–CH2–Ar), 1.00 (s, 9H, SiC–CH3).
MS (TOF-MS, ES+) calcd. for [C21H25O2N3Si + H]+:
380.1794; found: 380.1794.
CH3OH, 4:1): Rf 0.25. H NMR (399.8 MHz, DMSO-d6,
ppm) δ: 7.54 (s, 1H, C6–H), 7.33 (br s, NH2), 6.62 (br s,
NH2), 5.11 (t, J = 5.20 Hz, OH), 4.45 (s, 2H, –CH2–N), 4.17
(d, J = 5.20 Hz, 2H, O–CH2–Ar), 4.11 (q, J = 7.20 Hz, 2H,
O–CH2–), 1.19 (t, J = 7.20 Hz, 3H, –CH3). 13C NMR
(100.6 MHz, DMSO-d6, ppm) δ: 168.8, 165.0, 155.7, 144.2,
105.7, 60.9, 57.1, 50.0, 14.1. HR-MS (EI) calcd. for
[C9H13O4N3]+: 227.0906; found: 227.0905.
Ethyl[4-benzoyl-5-(t-
butyldiphenylsilanyloxymethyl)cytosin-1-yl]acetate (12)
To
a
solution of ethyl[5-(t-butyldiphenylsilanyloxy-
Identified byproduct: N1,N4-bisalkylated nucleobase:
ethyl(N4-(2-ethoxy-2-oxoethyl)-5-(hydroxymethylcytosin-
1yl)acetate. TLC (DCM – CH3OH, 4:1): Rf 0.52. H NMR
(400.1 MHz, DMSO-d6, ppm) δ: 7.88 (s, 1H, C6–H), 5.44 (t,
J = 5.2, –OH), 4.72 (s, 2H, N1–CH2–), 4.32 (d, J = 5.3, 2H,
methyl)cytosin-1-yl]acetate (3.42 g, 7.3 mmol) in anhyd.
pyridine (35 mL) was added benzoyl chloride (1 mL,
8.0 mmol) with stirring. The solution was stirred overnight,
ethyl acetate (300 mL) was added, and the solution was
washed with 1 mol/L HCl (200 mL × 2) to remove most of
1
© 2007 NRC Canada

Hudson et al.
307
the pyridine.4 The ethyl acetate phase was subsequently
washed with 0.02 mol/L HCl (200 mL × 3) and then water
(200 mL × 3) to remove traces pyridine. The acidic aqueous
phase was back-extracted by ethyl acetate (200 mL). The
ethyl acetate phases were combined and dried over Na2SO4.
The crude product had the appearance of an orange oil after
removal of the solvent. Column chromatography (hexane,
hexane–EtOAc (95:5) and (92:8) for 12; EtOAc – CH3OH
(95:5) to recover 11; Rf 0.04 (EtOAc – DCM, 1:6) was used
to purify the crude material. White solid; 92% (4.16 g, after
column-chromatography purification); mp 130–132 °C. TLC
(EtOAc – DCM, 1:6): Rf 0.76. 1H NMR (400.1 MHz,
CDCl3, ppm) δ: 13.19 (s, 1H, NH), 8.03 (d, J = 7.20 Hz, 2H,
Ar), 7.63 (d, J = 6.40 Hz, 4H, Ar), 7.28–7.40 (m, 10H, Ar
and C6–H), 4.70 (s, 2H, O–CH2–Ar), 4.43 (s, 2H, N–CH2–),
4.20 (q, J = 7.20 Hz, 2H, O–CH2–), 1.24 (t, J = 7.20 Hz, 3H,
–CH3), 1.04 (s, 9H, –Si–CCH3). 13C NMR (100.6 MHz,
CDCl3, ppm) δ: 179.6, 167.2, 158.2, 148.5, 141.3, 136.8,
135.6, 133.0, 132.6, 130.2, 130.0, 128.2, 128.1, 115.2, 62.4,
59.4, 49.9, 27.0, 19.5, 14.3. HR-MS (EI) calcd. for
[C32H35O5N3Si]+: 569.2346; found: 569.2352.
stirred overnight. The precipitated DCU was filtered out by
passing the solution through a pad of Celite. The filtrate was
diluted with DCM (450 mL) and washed with acidic water
twice (H2O, 250 mL; 1 mol/L HCl, 6 mL) to remove the
unreacted backbone submonomer, then washed with
NaHCO3 (satd. 250 mL × 2) and H2O (300 mL × 3), respec-
tively. All aqueous phases were back-extracted with DCM.
The combined organic phases were dried over Na2SO4, and
the solvent was removed under vacuum (RT). The crude
product (yellow oil) was purified by column chromatogra-
phy (hexane, hexane – ethyl acetate (80:20), (60:40), and
(10:90) for the title compound, whereas ethyl acetate –
methanol (98:2) and (95:5) was used to recover the back-
bone). White solid; 3.61 g, (58%); mp 125–126.5 °C (soft-
ens at 92 °C). TLC (ethyl acetate – methanol, 95:5;
visualized with UV and 0.3% ninhydrin (green)): Rf 0.62. 1H
NMR (400.1 MHz, CDCl3, ppm) δ: 13.26 (v br, NH), 8.13
(d, J = 7.20 Hz, 2H, H32), 7.29–7.74 (m, 22H, aromatic pro-
tons), 6.06 (br t, ma, backbone NH) and 5.48 (br t, J =
4.80 Hz, mi, backbone NH), 4.77 (s, 2H, H5), 4.64 (s, ma,
H6) and 4.52 (s, mi, H6), 4.47 (d, J = 6.80 Hz, ma, H23)
and 4.40 (d, J = 7.20 Hz, mi, H23), 4.22–4.25 (m, 1H, H22),
4.09 (s, ma, H27) and 4.22–4.25 (m, mi, H27), 3.73 (s, ma,
H29) and 3.79 (mi H29), 3.73 (br, ma, H26) and 2.95 (br, mi
H26), 3.46 (br, ma, H25) and 3.41 (br, mi H25), 1.15 (s, mi,
H11) and 1.13 (s, ma, H11). 13C NMR (100.6 MHz, CDCl3,
ppm) δ: 179.3 (C30), 169.9 (C28), 166.9 (C7), 158.5 (C2),
156.9 (C24), 148.5 (C1), 143.8 (C16), 142.8 (C4), 141.2
(C17), 136.7 (C31), 135.5 (C13), 133.0 (C12), 132.4 (C34),
130.0 (C15), 129.8 (C32), 128.1 (C33), 127.9 (C14), 127.7
(C19), 127.1 (C20), 125.1 (C21), 119.9 (C18), 114.2 (C3),
66.7 (C23), 59.3 (C5), 52.4 (ma, C29) and 52.9 (mi, C29),
48.9 (ma, C6) and 49.1 (mi, C6), 48.8 (ma, C27) and 50.3
(mi, C27), 48.7 (C26), 47.2 (C22), 39.0 (mi, C25) and 39.4
(ma, C25), 26.9 (C11), 19.3 (C10). HR-MS (ESI-TOF)
[4-N-Benzoyl-5-(t-butyldiphenylsilanyloxymethyl)cytosin-
1-yl]acetic acid (13)
To a solution of 12 (2.78 g, 4.88 mmol) in THF (55 mL)
and H2O (14 mL) at 4 °C was added 2.5 mol/L NaOH
(16 mL, 8 equiv.). The mixture was stirred at 0 °C for
15 min, and the reaction was stopped by decanting into a
separation funnel containing ethyl acetate (280 mL) and
0.25 mol/L HCl (190 mL) in which the reaction mixture was
acidified, and 13 was obtained. The ethyl acetate phase was
washed with H2O (140 mL × 2). The aqueous phase was ex-
tracted with ethyl acetate (280 mL × 2). The ethyl acetate
phases were combined and washed with H2O (400 mL),
dried over Na2SO4, and the solvent was removed under vac-
uum to yield the title compound as a white foam: 2.74 g
(99% yield). TLC (ethyl acetate – DCM, 1:6; same eluent
calcd. for [C50H51O8N5Si
900.3430.
+
Na]+: 900.3405; found:
1
used for 12): Rf 0. H NMR (400.1 MHz, CDCl3, ppm) δ:
8.06 (d, J = 7.20 Hz, 2H, Ar), 7.68 (m, 4H, Ar), 7.34–7.49
(m, 10H, Ar and C6–H), 4.73 (s, 2H, O–CH2–Ar), 4.54 (s,
2H, –CH2–N), 1.09 (s, 9H, SiC–CH3). 13C NMR
(100.6 MHz, CDCl3, ppm) δ: 177.7, 170.8, 158.7, 149.6,
142.8, 136.2, 135.7, 132.9, 132.8, 130.2, 129.8, 128.4,
128.1, 114.5, 59.6, 50.5, 27.1, 19.5. HR-MS (ESI-TOF, posi-
tive ion mode) calcd. for [C30H31O5N3Si + H]+: 542.2111;
found: 542.2091.
Fmoc-hmC-OH (15)
Compound 14 (3.40 g, 3.9 mmol) (Fig. 4) was dissolved
in THF (42 mL) and H2O (11 mL). The solution was cooled
to 4 °C in an ice-bath, and then 2.5 mol/L NaOH (12.1 mL,
8 equiv.) was added. The solution was stirred for 5 min at
4 °C, and then the reaction was stopped by decanting it into
a separation funnel containing ethyl acetate (500 mL) and
acidified water (36 mL of 1 mol/L HCl and 200 mL of
H2O). The ethyl acetate phase was washed with H2O
(200 mL × 2). The aqueous phase was extracted with ethyl
acetate (150 mL × 2). The organic phase was separated and
dried over Na2SO4. The solvent was removed under vacuum
(RT). The crude product (off-white solid, 3.7 g) was purified
by flash column chromatography (hexane, hexane – ethyl ac-
etate (50:50), ethyl acetate, and ethyl acetate – methanol
(95:5), (80:20), and (75:25)). White solid; 2.95 g (90%); mp
186–187 °C, as isolated from chromatography (softens at
130 °C). TLC (EtOAc – CH3OH, 95:5; visualized with UV
Fmoc hmC-OMe (14)
Compound 13 (3.83 g, 7.1 mmol) and hydroxybenzotri-
azole (HOBt, 1.03 g, 7.6 mmol) were dissolved in anhyd.
DMF (39 mL). The solution was cooled to 0 °C. After add-
ing N,N′-dicyclohexylcarbodiiimide (DCC, 1.58 g, 7.6 mmol)
in one portion, the mixture was taken off the ice and stirred
at room temperature for 1 h during which time a thick white
precipitate of dicyclohexylurea (DCU) was formed.
Methyl(2-(((9H-fluoren-9-yl)methoxy)carbonyl)aminoethyl)gly-
cinate (2.52 g, 7.1 mmol) was added, and the mixture was
1
and 0.3% ninhydrin (green)): Rf 0.04. H NMR (400.1 MHz,
4 Caution: care must be taken in the amount of HCl used, which should be calculated based on complete neutralization of pyridine; excess
acid removed the N4-Bz group.
© 2007 NRC Canada

308
Can. J. Chem. Vol. 85, 2007
Fig. 4. Structure of the 5-hydroxymethylcytosine (hmC) PNA
monomer ester (R = –CH3) and acid (R = –H), defining the arbi-
trary numbering system used in NMR signal assignments.
Oligomer P4, which contains an hmC residue, was addi-
tionally treated with concentrated aq. ammonia (1 mL, 55 °C
for 12 h). The solution was dried under reduced pressure to
give an off-white solid.
Oligomer purification
The crude, solid oligomers were dissolved in aq. TFA
(0.1%), filtered, and purified by RP-HPLC, using a 250 mm ×
4.6 mm, C18-bonded phase, 300 Å pore, and 5 µm particle-
size column.
Characterization of PNAs
Mass spectral analysis of oligomers was performed at the
UWO Biological Mass Spectrometry Laboratory. Electrospray
ionization time-of-flight mass spectrometry (ESI-TOF MS)
was performed with mass calibration provided by myoglobin
(M+ = 16951.49 Da, average mass) or histatin 5 (M+
3036.33 Da, monoisotopic mass) (see Table 1).
=
Binding analysis
Poly(riboadenylic acid) (poly(rA)) was purchased from
Sigma (Milwaukee, WI). Mixed sequence oligonucleotides
were synthesized and purified by standard methods or pur-
chased from the University Core DNA Services (Calgary,
AB). Thermal denaturation of PNA:NA complexes
(~2 µmol/L in nucleobases) was measured at 260 nm with a
temperature ramp of 0.5 °C/min under ionic conditions
(150 mmol/L NaCl, 10 mmol/L Na2HPO4, and 0.1 mmol/L
EDTA at pH 7.0). A minimum of two independent samples
were run many times, and the Tm was estimated by deter-
mining the maximum in the first derivative of the denatur-
ation curve. The error associated with the Tm values is
0.5 °C.
DMSO-d6, ppm) δ: 7.96 (d, J = 7.60 Hz, 2H, H32), 7.31–
7.89 (22H, aromatic protons), 7.28–7.55 (overlap, 1H, back-
bone NH), 4.92 (s, mi, H6) and 4.72 (s, ma, H6), 4.60 (s, mi,
H5) and 4.58 (s, ma, H5), 4.34 (d, J = 6.80 Hz, mi, H23)
and 4.24 (d, J = 6.40 Hz, ma, H23), 4.19–4.20 (m, 1H,
H22), 3.99–4.01 (br, depending on the state of protonation,
2H, H27), 3.47 (b, ma, H26) and 3.30 (br m, mi, H26), 3.39
(br m, ma, H25) and 3.17 (br m, mi, H25), 0.97 (s, 9H,
H11). 13C NMR (100.6 MHz, DMSO-d6, ppm) δ: 172.1
(C30), 171.5 (C28), 167.3 (C7), 158.7 (C2), 156.6 (C24),
148.7 (C1), 146.3 (C4), 144.0 (C16), 140.8 (C17), 136.3
(C31), 135.3 (C13), 132.8 (C12), 132.5 (C34), 130.0 (C32),
129.3 (C15), 128.2 (C33), 128.0 (C14), 127.7 (C19), 127.2
(C20), 125.3 (C21), 120.1 (C18), 111.9 (C3), 65.7(C23),
59.2(C5), 51.1 (mi, C27) and 49.4 (ma, C27), 48.6 (C6),
47.7 (mi, C26) and 47.1 (ma, C26), 46.9 (C22), 38.2 (C25),
26.7 (C11), 18.9 (C10). HR-MS (ESI-TOF) calcd. for
[C49H49O8N5Si – H]+: 862.32722; found: 862.3262.
Results and discussion
The goal of this work was to achieve the synthesis of 5-
hydroxymethyl-derivatized uracil and cytosine PNA mono-
mers, incorporate these modified bases into PNA, and study
their impact on hybridization. As a starting point, we were
interested in a scalable and straightforward approach to the
synthesis of the monomers. Key considerations were on the
preparation of the 5-hydromethyluracil and -cytosine and
subsequently the use of an appropriate protecting-group
strategy for the hydroxyl functional group.
Synthesis of PNA oligomers
PNA oligomers were synthesized on a 5 µmol scale using
the standard FastMoc module on an Applied Biosystems
433A synthesizer employing a NovaSyn TGR (Rink amide)
resin. The hexameric PNA (P1) was synthesized using the
Rink amide resin with a loading of 0.23 mmol/g. Oligomers
(P2–P5) were prepared on a Rink amide resin preloaded
with lysine (approximate loading: 0.065 mmol/g).
5-Hydroxymethyluracil-containing PNA
The synthesis of 5-hydroxymethyluracil monomer (hmU),
suitable for Fmoc-based peptide chemistry, is outlined in
Scheme 1. We took the approach of preparing the appropri-
ate (pyrimidin-1-yl)acetic acid ester derivative, which would
be condensed with the methyl[N-2-(Fmoc-amino)ethyl]glycinate
“backbone” submonomer in a convergent fashion, which we
have previously reported (8). The synthesis starts with the
inexpensive uracil and a one-step hydroxymethylation. De-
spite the potential biological importance of 5-hydroxy-
methyluracil and nucleoside derivatives thereof (9–12),5
Resin cleavage and deprotection
To the resin bound oligomer (P1–P5) was added an ice-
cold solution of TFA (950 µL) and triethylsilane (25 µL).
The resin was agitated for 2 h. After filtration, TFA was re-
moved with a stream of nitrogen, and the oligomer was pre-
cipitated, as a white solid, with cold diethyl ether (15 mL).
Under these conditions, the TBDPS protecting group is also
removed.
5 5-hmU is a component of bacteriophage DNA, as well, there exists hmU specific binding proteins. The DNA of the dinoflagellate
Gyrodinium cohnii contains 5-hmU. Glycosylated hmU is found in Typanosoma Brucei, the causative organism of African sleeping sick-
ness. 5hmU is caused by oxidative damage to DNA
© 2007 NRC Canada

Hudson et al.
309
Table 1. Mass spectral characterization of PNA oligomers.
Oligomer
Sequence
Found
Calculated
P1
P2
P3
P4
P5
Ac-(hmU)T(hmU)T(hmU)T-Lys-NH2
H-Gly-GTA GAT CCC T-Lys-NH2
H-Gly-GTA GAT CTC T-Lys-NH2
H-Gly-GTA GAT C(hmC)C T-Lys- NH2
H-Gly-GTA GAT C(hmU)C T-Lys- NH2
1832.50
2887.86
2902.78
2917.68
2918.93
1832.76
2887.80
2902.82
2917.83
2918.82
Note: Abbreviations in use: 5-hydroxymethyluracil (hmU), 5-hydroxymethylcytosine (hmC).
Scheme 1. Reagents and conditions: (a) (HCHO)n, 0.5 mol/L Et3N, 60 °C, 16 h; (b) BrCH2CO2Et, K2CO3, DMF, RT, 16 h;
(c) TBDPS-Cl, imidazole, DMF, RT, 16 h; (d) 0.43 mol/L NaOH, 0 °C, 20 min; (e) methyl[N-2-(Fmoc-amino)ethyl]glycinate (6), EDC,
DMF, RT, 16 h; (f) (i) 0.46 mol/L NaOH, 0 °C, 5 min; (ii) H+.
there exist but a few synthetic routes to the modified
nucleobase. Hydroxymethyluracil has been prepared by the
reaction of uracil with paraformaldehyde in water, using an
inorganic base (13, 14). However, we wished to avoid the
potential problem of separating the nucleobase from inor-
ganic salts and were inspired by the report of triethylamine-
catalyzed hydroxymethylation of the deoxynucleoside (15).
The reaction of paraformaldehyde with uracil in aq. NEt3
(0.5 mol/L) proceeds smoothly overnight, and the product
conveniently crystallizes from the reaction solvent.
As with uracil, thymine, and related derivatives, we next
proceeded to the alkylation of the nucleobase with ethyl
bromoacetate in anhyd. DMF with K2CO3 (16), suspecting
that we would achieve the desired chemoselectivity, based
on the pKa differences between the 5-hydroxyl group and
the lactam/imide groups of the heterocycle. We observed
complete selectivity for N-alkylation and recovered the de-
sired N1-derivative in good yield. The only byproduct that
we were able to identify from this reaction was the N1,N3-
bisalkylated nucleobase.
as well as orthogonality to the Fmoc- group and the methyl
ester of the backbone. Initially, we prepared the t-
butyldimethylsilyl ether, but this proved not robust enough
in model studies; thus, we turned to the t-butyldiphenylsilyl
ether. This group was stable to alkyl ester hydrolysis
(0.43 mol/L NaOH, THF – H2O, 0 °C, >40 min), Fmoc-
removal (30% piperidine in DMF, 40 min), and normal
workup conditions (pH 1–12), yet was completely and rap-
idly removed by treatment with TFA. After settling on a
hydroxyl-protecting group, standard transformations yielded
the hmU monomer 8 with some noteworthy points. The con-
densation of the backbone submonomer 6 with the modified
nucleobases 4 was mediated by the use of the carbodiimide
reagent EDC without the use of additional base or additives.
As well, the best yield and least elimination of the Fmoc-
group were obtained when the acid component was used in
excess (1.6 equiv.) relative to the backbone submonomer
free base. In this instance, the isolated yields of 7 were reli-
ably high (88%) with a minimal loss of the base-labile
Fmoc-group (<4%).
Next, we chose a silyl ether as the masking group for the
5-hydroxy functionality for ease of installation and removal
The penultimate step, the saponification of the monomer
ester (7), was also carefully done to avoid loss of the Fmoc-
© 2007 NRC Canada

310
Can. J. Chem. Vol. 85, 2007
Scheme 2. Reagents and conditions: (a) (HCHO)n, 0.75 mol/L Et3N, H2O, 3 d, 95 °C; (b) K2CO3, DMF, BrCH2CO2Et, 24 h;
(c) TBDPS-Cl, imidazole, DMF, 16 h; (d) BzCl, pyridine; (e) 0.46 mol/L NaOH, 15 min, 4 °C; ( f ) 6, DCC, HOBT, DMF; (g)
(i) 0.46 mol/L NaOH, 0 °C, 5 min; (ii) H+.
group. The reaction conditions (8 equiv. of NaOH,
0.46 mol/L, 5 min) were optimized to give complete hydro-
lysis, while minimizing concomitant removal of the Fmoc-
group. Nonetheless, some Fmoc- groups were lost (4%–8%),
yet the final monomer was recovered in 90% yield from col-
umn chromatography. The optimized conditions for the con-
densation and hydrolysis reactions were then used for the
synthesis of the hmC monomer 15.
action solution by crystallization. Although later crops are
significantly contaminated with unreacted cytosine (up to
approximately 20% (wt%)), they are still suitable for use in
the next step.
The next necessary transformations are installation of
nucleobase-protecting groups and alkylation. For cytosine,
either of two routes may be used: N4-protection followed by
N1-alkylation in the presence of either K2CO3 or NaH (22,
23); or the reverse, N1-alkylation in the presence of NaH
and then protection of the exocyclic amino group (24). Both
routes were attempted and gave identical yields; thus, the
alkylation was simply performed in DMF with K2CO3, as
was done with 5-hydroxymethyluracil.
Following alkylation, the nuclebase was protected, first by
silylation of the hydroxyl group and then benzoylation of the
exocyclic amino group. The silylation step gave a poor yield,
likely by interference by the adjacent amino group, which
could not be improved even by use of an excess TBDPS-Cl
(2 equiv.). The remaining transformation to yield the hmC
PNA monomer 15 proceeded without incident.
5-Hydroxymethylcytosine-containing PNA
Although 5-hydroxymethylcytosine has been known for
some time in the context of a modified base found in the
DNA of the Escherichia coli T-even bacteriophages (17) and
in the nucleoside antibiotic mildiomycin (18), there has been
no report of its synthesis from cytosine, via hydroxy-
methylation, until quite recently. Aside from isolation from
these natural sources, early syntheses are based on multistep
de novo heterocycle formation (19).
Our initial attempts to produce hmC under similar condi-
tions of reaction to those used for the synthesis of hmU were
unproductive. During this period of time, Abdel-Rahman
and El Ashry reported a small-scale (1 mmol) hydroxy-
methylation of cytosine with paraformaldehyde in 0.5 N
KOH solution under microwave irradiation (20). Our at-
tempts to repeat this synthesis failed. However the report of
the reaction of 2-deoxycytosine monophosphate with
paraformaldehyde in the presence of 5 mol/L KOH (75 °C,
4 d) (12% yield) (21) motivated our further efforts on the di-
rect hydroxymethylation of cytosine. Although the reaction
of cytosine in aq. KOH with paraformaldehyde failed miser-
ably, hmC was formed in good chemical yield (~85%) when
triethylamine was employed along with higher reaction tem-
perature and longer reaction time. We found that a molar ex-
cess of paraformaldehyde was necessary for the reaction to
proceed in good chemical yields, but the addition must be
made portion-wise during the course of the reaction. For in-
stance, addition of greater than 3 equiv. of paraformaldehyde
at the beginning of the reaction causes the formation of un-
desired and unidentified byproducts and made the recovery
of hmC by recrystallization unfeasible. The reaction com-
pletely fails to give the desired product at 10 equiv. of
paraformaldehyde. To overcome these problems, the amount
of paraformaldehyde was carefully controlled and added in
several portions during the reaction ((HCHO)n – cytosine
(mol) 3.5:1), Scheme 2. The hmC is recovered from the re-
Synthesis and hybridization properties of PNA
containing C5-hydroxymethylpyrimidines
The hmU and hmC PNA monomers were used for both
automated and manual PNA-oligomer synthesis. Initially,
standard conditions were employed (resin loading
0.23 mmol/g), and this was suitable for the incorporation of
hmU in isolated positions (P1) by automated methods.
However, upon attempts to prepare contiguously modified
oligomers by manual methods, the synthesis failed after 5 or
6 residues. Although these oligomers were identified by
mass spectrometry, they were inseparable, and further se-
quence elongation was not possible. Therefore, we turned to
lower-loading resin (0.06 mmol/g), which did not resolve
this problem. However, the lower-loading resin was subse-
quently used to prepare singly modified oligomers P2–P4,
which allowed us to evaluate the effect of the hmU or hmC
substitution for T or C, respectively.
The oligomers were examined for their ability to bind
complementary nucleic acids (Table 2). The poly(hmU/T)
hexamer was used for triplex formation with poly(rA),
wherein two strands of PNA bind to the RNA strand, thus
giving each complex a total of six modifications. Compared
with the unmodified oligomer, there was a slight drop in Tm
(–1.5 °C) per insert. Interestingly, the same modest Tm drop
© 2007 NRC Canada

Hudson et al.
Table 2. Summary of Tm values for complexes formed between PNA and complementary DNA or RNA.
311
Oligomer
Sequence
Complement
Tm (°C)
∆Tm/modd (°C)
a
P0
P1
P2
Ac-TTT TTT-K-NH2
poly(rA)
poly(rA)
DNAb
62.0
53.0
50.7
—
Ac- (hmU)T(hmU)T(hmU)T-Lys-NH2
H-Gly-GTA GAT CCC T-Lys-NH2
–1.5
—
P3
P4
H-Gly-GTA GAT CTC T-Lys-NH
H-Gly-GTA GAT C(hmC)C T-Lys- NH2
DNAc
47.3
49.3
—
–1.4
DNAb
P5
H-Gly-GTA GAT C(hmU)C T-Lys- NH2
DNAc
46.0
–1.3
aFrom ref. 7.
bComplementary deoxyoligonucleotide sequence: 5′-AGA GAT CTA C-3′.
cComplementary deoxyoligonucleotide sequence: 5′-AGG GAT CTA C-3′.
d∆Tm/mod: the difference in the melting temperature of modified oligomer compared with the unmodified oligomer on a per-substitution
basis.
was observed in the singly modified duplexes (P4 and P5),
regardless of the identity of the nucleobase. These investiga-
tions represent the first study of the effect of 5-hydroxy-
methylpyrimidines on complex stability in either DNA or
analogs. Previous studies have focused on the glycosylated
derivatives (25), wherein only the diaminoglycoside conju-
gate was stabilizing towards duplex formation (26, 27). Rel-
evant studies to 5-hydroxymethylcytosine have been focused
on the interaction between oligonucleotides bearing this
modification and enzymes and have not included Tm mea-
surements (28, 29).
In summary, we have described a convenient, large-scale
synthesis of 5-hydroxymethylcytosine and 5-hydroxymethyl-
uracil and the transformation of these modified nucleobases
in monomers compatible with PNA synthesis. The modifica-
tions were placed in isolated positions in sequences capable
of either triplex formation or duplex formation. Dense
functionalization of PNA oligomers was not achieved, likely
because of the hydrophobic/steric nature of the protected
hmU and hmC monomers. A single modification had a small
negative impact on the stability of the complex, independent
of the particular nucleobase, and multiple substitutions had
an additive effect. Overall, the substitution of a hmU or hmC
for T or C, respectively, is tolerated in both the context of a
duplex or a triplex, within the sequences examined. Further
investigation on the optimization of the protecting-group
strategy and oligomerization chemistry is needed to fully
evaluate this modification and its impact on the biophysical
properties of PNA.
References
1. P.E. Nielsen, M. Egholm, R.H. Berg, and O. Buchardt. Science
(Washington, D.C.), 254, 1497 (1991).
2. P.E. Nielsen, U. Koppelhus, and F. Beck. In Pseudo-peptides
in drug discovery. Edited by P.E. Nielsen. Wiley-VCH,
Weinheim, Germany. 2004. Chap. 4.
3. F. Beck. In Methods in molecular biology: Peptide nucleic ac-
ids, methods and protocols. Vol. 208. Edited by P.E. Nielsen.
Humana Press Inc., Totowa, NJ, USA. 2002. Chap. 2.
4. P. Zhou, A. Dragulescu-Andrasi, B. Bhattacharya, H. O’Keefe,
P. Vatta, J.J. Hyldig-Nielsen, and D.H. Ly. Bioorg. Med.
Chem. Lett. 16, 4931 (2006).
5. B.D. Gildea, S. Casey, J. MacNeil, H. Perry-O’Keefe, D.
Sørensen, and J.M Coull. Tetrahedron Lett. 39, 7255 (1998).
6. G.R. Revandkar and T.S. Rao. In Comprehensive natural prod-
ucts: DNA and aspects of molecular biology. Vol. 7. Edited by
E.T. Kool. Elsevier Sciences Ltd., Oxford, UK. 2002. Chap. 2.
7. R.H.E. Hudson and F. Wojciechowski. Can. J. Chem. 83, 1731
(2005).
8. R.H.E. Hudson, M. Goncharenko, A. Wallman, and F.
Wojciechowski. Syn. Lett. 9, 1442, (2005).
9. (a) G.G. Johnson and E.P. Geiduschek. Biochemistry, 16, 1473
(1977); (b) J.R. Greene and E.P. Geiduschek. EMBO J. 4,
1345 (1985).
10. P.M.M. Rae. Proc. Natl. Acad. Sci. U.S.A. 70, 1141 (1973).
11. F. VanLeeuwen, M.C. Taylor, A. Mondragon, H. Moreau, W.
Gibson, R. Kieft, and P. Borst. Proc. Natl. Acad. Sci. U.S.A.
95, 2366, (1998).
12. M. Berger, J. Cadet, and J. Ulrich. Can. J. Chem. 63, 6, (1985).
13. R.E. Cline, R.M. Fink, and K. Fink. J. Am. Chem. Soc. 81,
2521 (1959).
14. U. Chiacchio, A. Corsaro, J. Mates, P. Merino, A. Piperno, A.
Rescifina, G. Romeo, R. Giovanni, and T.T. Roberto. Tetrahe-
dron, 59, 4733, (2003).
Acknowledgements
The authors thank the Natural Sciences and Engineering
Research Council of Canada (NSERC) and the Academic
Development Fund at The University of Western Ontario for
support of this work. Prof. G. Lajoie and Ms. Paula Pittock
of the Biological Mass Spectrometry Laboratories (UWO)
are thanked for their expert assistance with mass spectral
characterization of the polyamides. Mr. Doug Hairsine, De-
partment of Chemistry, is thanked for performing HR-MS of
small molecules. Mr. David Dodd is acknowledged for per-
forming polyacrylamide gel electrophoresis of PNA (data
not shown).
15. C.J. LaFrancois, J. Fujimoto, and L.C. Sowers. Chem. Res.
Toxicol. 11, 75, (1998).
16. R.H.E. Hudson and J.M. Moszynski. Synlett, 18, 2997, (2006).
17. G.R. Wyatt and S.S. Cohen. Nature (London), 170, 1072,
(1952).
18. S. Harada and T. Kishi. J. Antibiot. (Tokyo), 31, 519, (1978).
19. T.L.V Ulbricht and C.C. Price. J. Am. Chem. Soc. 21, 567,
(1956), and refs. cited therein.
20. A.A.-H. Abdel-Rahman and E.-S.H. El Ashry. Synlett, 12,
2043, (2002).
21. A.H. Alegria. Biochim. Biophys. Acta, 149, 317, (1967).
© 2007 NRC Canada

312
Can. J. Chem. Vol. 85, 2007
22. K.L. Dueholm, M. Egholm, C. Behrens, L. Christensen, H.F.
Hansen, T. Vulpius, K.H. Petersen, R.H. Berg, P.E. Nielsen,
and O. Buchardt. J. Org. Chem. 59, 5767, (1994).
23. Z. Timar, L. Kovacs, G. Kovacs, and Z. Schmel. J. Chem.
Soc., Perkin Trans. 1, 19, (2000).
26. J. Hunziker. Bioorg. Med. Chem. Lett. 9, 201, (1999).
27. R. Tona, R. Bertolini, and J. Hunziker. Org. Lett. 2, 1693,
(2000).
28. S. Tardy-Planechaud, J. Fujimoto, S.S. Lin, and L.C. Sowers.
Nucleic Acids Res. 25, 553, (1997).
24. L. Christensen, H.F. Hansen, T. Koch, and P.E. Nielsen. Nu-
cleic Acids Res. 26, 2735, (1998).
25. No Tm data reported: J.J. Turner, N.J. Meeuwenoord, A. Rood,
P. Borst, G.A. van der Marcel, and J.H. van Boom. Eur. J. Org.
Chem. 3832, (2003).
29. M. de Kort, P.C. de Visser, J. Kurzeck, N.J. Meeuwenoord,
G.A. van der Marcel, W. Rüger, and J.H. van Boom. Eur. J.
Org. Chem. 2075, (2001).
© 2007 NRC Canada