Synthesis of a ferrocenyl uracil PNA monomer for insertion into PNA sequences

Article abstract of DOI:10.1016/j.jorganchem.2008.05.009

The deprotection of the tert-butyl group of a ferrocenyl uracil Peptide Nucleic Acid (PNA) monomer, Fmoc-aeg(R)-O^tBu (1) was achieved using a two step synthesis involving hydrolysis in basic conditions to give first the zwitterion of ⁺NH₃-aeg(R)-O^- (7). Compound 7 was reacted in situ with N-(9-fluorenylmethoxycarbonyloxy)succinimide to obtain the expected compound Fmoc-aeg(R)-OH (2) (Abbreviations: Aeg = (2-aminoethyl)-glycine; Fmoc = 9-fluorenylmethoxycarbonyl; O^tBu = tert-butyl; R = 5-(N-ferroce-nylmethylbenzamido)uracyl). Crown Copyright

Full text of DOI:10.1016/j.jorganchem.2008.05.009

Journal of Organometallic Chemistry 693 (2008) 2478–2482
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis of a ferrocenyl uracil PNA monomer for insertion into PNA sequences
*
Gilles Gasser, Leone Spiccia
School of Chemistry, Monash University, Clayton, Vic. 3800, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
The deprotection of the tert-butyl group of a ferrocenyl uracil Peptide Nucleic Acid (PNA) monomer,
Fmoc-aeg(R)-OtBu (1) was achieved using a two step synthesis involving hydrolysis in basic conditions
to give first the zwitterion of +NH3-aeg(R)-Oꢀ (7). Compound 7 was reacted in situ with N-(9-fluorenyl-
methoxycarbonyloxy)succinimide to obtain the expected compound Fmoc-aeg(R)-OH (2) (Abbreviations:
Aeg = (2-aminoethyl)-glycine; Fmoc = 9-fluorenylmethoxycarbonyl; OtBu = tert-butyl; R = 5-(N-ferroce-
nylmethylbenzamido)uracyl).
Received 7 March 2008
Received in revised form 28 April 2008
Accepted 6 May 2008
Available online 13 May 2008
Keywords:
Peptide nucleic acids (PNA)
Synthesis
Crown Copyright Ó 2008 Published by Elsevier B.V. All rights reserved.
Ferrocene
Bioorganometallic chemistry
Metallocenes
Acidic medium hydrolysis
1. Introduction
have been successfully inserted within a PNA oligomer sequence
(instead of amino terminus, as reported by Metzler-Nolte et al.
Insertion of organometallics into peptide nucleic acids (PNA) [1]
is currently a field of active research [2–14]. The idea of coupling
the remarkable properties of PNA, such as high binding affinity
for DNA/RNA strands, high chemical stability and resistance to
nucleases [15], with those of organometallics is attractive from
the point of view of the development of DNA biosensors [16].
Due to its well established chemistry, availability, reversible elec-
trochemistry and chemical stability [17], ferrocene (Fc) was the
first organometallic moiety to be attached to a PNA monomer
[10], and then to a PNA sequence [9]. The method used by Met-
zler-Nolte and co-workers involving the coupling of ferrocenecar-
boxylic acid to the amino end of a PNA sequence allows insertion
of the organometallic moiety at the amino terminus only. Signifi-
cant flexibility in bioconjugate design could be achieved if the
organometallic moiety could be inserted at any selected position
within the PNA sequence. From the perspective of electrochemical
DNA/RNA biosensor applications, this would allow, for example,
the incorporation of a redox-active unit in closer proximity to a
base mismatch. In an effort to achieve this insertion, Hudson
et al. [8], Maiorana et al. [3–6,13] and our group [18] have reported
the synthesis of PNA monomers containing an organometallic moi-
ety that could be potentially inserted anywhere within the PNA se-
quence. When Fc was used, it was attached either directly to the
base uracil [8,18] or to the PNA backbone [4,6,13]. To the best of
our knowledge, however, none of these ferrocenyl PNA monomers
[9]). During our attempts to achieve this, we encountered synthetic
problems derived from the instability of ferrocene bearing PNA
monomers in acidic media. We report the successful synthesis of
a ferrocenyl PNA monomer (2 in Fig. 1) that could be inserted into
PNA sequences using standard solid phase protocols and the char-
acterisation of the side-products (3–5 in Fig. 1) isolated during our
unsuccessful endeavours to obtain 2. The choice of an appropriate
solid support for ferrocene to be inserted into a PNA sequence is
also discussed.
2. Experimental
2.1. Materials
All chemicals were of reagent grade quality or better, obtained
from commercial suppliers and used without further purification.
Solvents were used as received or dried over 4 Å molecular sieves
or dried following literature procedures [19]. High purity nitrogen
gas was used directly from the reticulated system.
2.2. Instrumentation
1H and 13C NMR spectra were recorded in deuterated solvents
on either
a Bruker AC200, a Bruker DPX300 or an Avance
DRX400 Bruker spectrometer at 30 °C. The chemical shifts, d, are
reported in ppm (parts per million). Tetramethylsilane (TMS) or
the residual solvent peaks have been used as an internal reference.
* Corresponding author. Fax: +61 3 9905 4597.
E-mail address: leone.spiccia@sci.monash.edu.au (L. Spiccia).
0022-328X/$ - see front matter Crown Copyright Ó 2008 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.05.009

G. Gasser, L. Spiccia / Journal of Organometallic Chemistry 693 (2008) 2478–2482
2479
O
O
O
N
O
N
HN
N
NH
NH
Fe
OH
O
Fe
O
O
O
O
O
O
O
N
N
O
N
OR
O
N
H
OR
H
1: R = C(CH3)3
2: R = H
3:
5
R = C(CH3)3
4: R = H
Fig. 1. Structures of 1–5.
The abbreviations for the peak multiplicities are as follows: s (sin-
glet), d (doublet), dd (doublet of doublets), t (triplet), q (quartet), m
(multiplet) and br (broad). 2D experiments were conducted and
analysed on XWinNMR. The 1H and 13C NMR spectra of compounds
2–4, in particular, were difficult to assign because of the presence
of rotamers. This was attempted with the aid of the 2D plots (COSY,
HMQC and HMBC), but the assignments should not be treated as
definitive. Infrared spectra were recorded using KBr pellets using
a Shimadzu FTIR-8400S spectrophotometer at 1 cmꢀ1 resolution.
Electrospray mass spectra were recorded on Micromass Platform
II Quadrupole Mass Spectrometer fitted with an electrospray
source. Accurate mass spectra were recorded on a Bruker BioApex
II 47e FT-ICR MS fitted with an Analytica Electrospray Source. Sam-
1142 w, 1104 w, 1024 w, 960 w, 793 w, 740 w, 492 w. 1H
NMR Spectrum (acetone-d6): 3.10–3.45 (m, 4H, NH–CH2–CH2
and CH2–CH2–N), 3.95–4.00 (m, 4H, 2 ꢁ CH Cp ring and N–CH2–
CON), 4.05 (s, 5H, 5 ꢁ CH Cp), 4.10–4.30 (m, 6H, 2 ꢁ CH Cp,
Fmoc–CH–CH2O and 2 ꢁ Fmoc–CH–CH2O and 1 ꢁ Cp–CH2–N),
4.48 (d, 2J(H–C–H) = 16.3 Hz, 1H, 1 ꢁ N–CH2–COOH), 4.65 (d,
2J(H–C–H) = 16.3 Hz, 1H, 1 ꢁ N–CH2–COOH), 4.94 (d, 2J(H–C–
H) = 13.5 Hz, 1H, 1 ꢁ Cp–CH2–N), 6.46 (min) and 6.60 (maj) (rota-
mers, br s, 1H, CH2–NH–COO), 7.08-7.35 (rotamers, m, 10H,
4 ꢁ CHFmoc arom and 5 ꢁ CH arom and C@CHAN), 7.57 (m, 2H,
CHFmoc arom), 7.75 (m, 2H, CHFmoc arom), 10.14 (min) and
10.17 (maj) (rotamers, s, 1H, CO–NH–CO). 13C NMR Spectrum
(acetone-d6): d 39.96 (NH–CH2–CH2), 48.20 (Fmoc–CH–CH2O),
48.43, 48.53 and 49.04 (br, corresponds to Cp–CH2–N, CH2–CH2–
N, N–CH2–CON and N-CH2–COOH), 67.21 (min) and 67.30 (maj)
(rotamers, Fmoc–CH–CH2O), 69.14 (CH Cp), 69.33 (broad, CH
Cp), 70.87 (CH Cp), 83.76 (C Cp), 120.90 (CH Fmoc), 126.24 (CH
Fmoc), 128.05 (CH arom and CH Fmoc), 128.62 (b, CH arom and
CH Fmoc), 130.32 (CH arom), 137.73 (rotamers, CO–C benzyl),
142.22 (CFmoc), 145.23 (maj) and 145.29 (min) (rotamers,
CFmoc), 146.92 (C@CHAN), 151.14 (NH–CO–N), 157.61 (NH–
COO–CH2), 161.76 (C–CO–NH), 167.70 (min) and 167.79 (maj)
(N–CH2–CON), 168.37 (CH2–COOH), 170.58 (min) and 171.42
(maj) (N–CO–benzyl). We were unable to locate the signal for
(NAC@CH), probably due to a long relaxation time of this carbon.
Electrospray mass spectrum (m/z): Negative mode: 808 [MꢀH]ꢀ
(100%). High Resolution ESI Mass Determination: Found:
832.2046; calcd for C43H39FeN5NaO8, 832.2038.
ples were introduced by syringe pump at 1
voltage of 200 V was applied.
lL/min. A capillary
2.3. Synthesis
2.3.1. tert-Butyl-2-(N-(2-(((9H-fluoren-9-yl)methoxy)carbonylamino)
ethyl)-2-(5-(N-ferrocenylmethylbenzamido)-2,4-dioxo-3,4-
dihydropyrimidin-1(2H)-yl)acetamido)acetate (1)
Compound 1 was prepared following the procedure published
by Spiccia et al. [18]. The spectroscopic data of the product
matched that reported previously [18].
2.3.2. 2-(N-(2-(((9H-fluoren-9-yl)methoxy)carbonylamino)ethyl)-2-
(5-(N-ferrocenylmethylbenzamido)-2,4-dioxo-3,4-dihydropyrimidin-
1(2H)-yl)acetamido)acetic acid (2)
Compound 1 (50.0 mg, 0.06 mmol) was stirred in a mixture of
aqueous 2 M NaOH:acetone 1:1 (5 mL of each) for 5 h at r.t. The
solution was neutralised with 20% aqueous solution of citric acid
and the pH was raised to 8 with an aqueous solution of sat. NaH-
CO3. Water (5 mL) and acetone (5 mL) were added to the mixture
to solubilise the precipitate formed. N-(9-fluorenylmethoxycar-
bonyloxy)-succinimide (19.2 mg, 0.06 mmol) was then added
and the solution was stirred for 24 h at r.t. The pH was then ad-
justed to 4 with a 20% aqueous solution of citric acid. Following
evaporation of acetone, the resulting aqueous phase was ex-
tracted with dichloromethane (150 mL). The organic phase was
dried with Na2SO4, filtered and evaporated to dryness to give an
orange oil. Chromatography on a silica column using a gradient
ranging from hexane:ethyl acetate 1:1 to acetone:water 5:1 as
the eluent was performed. An orange fraction was collected and
evaporated to dryness. The residual solid was dissolved in ethyl
acetate (25 mL) and the organic phase washed with water
(10 mL), dried with Na2SO4, filtered and evaporated to dryness
to give 2 as an orange solid. Yield: 18 mg (38%). Selected IR bands
2.3.3. 2-(N-(2-(((9H-fluoren-9-yl)methoxy)carbonylamino)ethyl)-2-
(5-benzamido-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-
yl)acetamido)acetic acid (4)
Compound 1 (50.0 mg, 0.06 mmol) was dissolved in CH2Cl2
(2.5 mL) and a solution of 4 M HCl in dioxane (0.35 mL) was added
to the orange solution. The solution was stirred for 48h at r.t. A green
solution and a white precipitate were obtained. Cold diethyl ether
(5 mL) was added to the mixture to complete the precipitation of
product. The white solid, 2-(N-(2-(((9H-fluoren-9-yl)methoxy)car-
bonylamino)ethyl)-2-(5-benzamido-2,4-dioxo-3,4-dihydropyrimi-
din-1(2H)-yl)acetamido)acetic acid (4), was then filtered, washed
with ether (3 ꢁ 5 mL) and dried. Yield: (24 mg, 67%). Selected IR
bands (KBr; m
, cmꢀ1): 3391 w br, 3181 w br, 3072 w br, 2922 w br,
1625–1750 s br, 1542 m, 1462 m, 1416 w, 1384 m, 1244 m, 1230
m, 1196 w, 1153 w, 1104 w, 1001 w, 888 w, 759 w, 742 w, 707 w,
622 w. 1H NMR Spectrum (acetone-d6): 3.15–3.55 (m, 4H, NH–
CH2–CH2 and CH2–CH2–N), 4.07 (maj) (rotamers, s, N–CH2–CON),
4.15–4.23 (m, Fmoc–CH–CH2O and rotamers (min) N–CH2–CON),
4.30 (min) and 4.32 (maj) (rotamers, m, 2H, Fmoc–CH–CH2O),
4.65 (min) and 4.80 (max) (s, 2H, N–CH2–COOH), 6.42 (min) and
(KBr;
1625–1750 s br, 1541 w, 1461 m, 1407 m, 1246 m, 1230 m,
m
, cmꢀ1): 3396 w br, 3177 w br, 3072 w br, 2924 w br,

2480
G. Gasser, L. Spiccia / Journal of Organometallic Chemistry 693 (2008) 2478–2482
6.66 (maj) (rotamers, br s, 1H, CH2-NH–COO), 7.18–7.31 (m, 4H, CH
Fmoc arom), 7.38–7-47 (m, 3H, CH arom), 7.56–7.61 (m, 2H, CH
Fmoc arom), 7.70-7.83 (m, 4H, 2 ꢁ CH arom and 2 ꢁ CH Fmoc arom),
8.39 (min) and 8.41 (maj) (rotamers, s, 1H, C@CHAN), 8.46 (min)
and 8.48 (maj) (rotamers, s, 1H, Ar–CO–NH), 10.54 (s, 1H, CO–NH–
CO). 13C NMR Spectrum (acetone-d6): 39.53 (min) and 39.90 (maj)
(rotamers, NH–CH2–CH2), 48.13 (Fmoc-CH–CH2O), 48.50 (maj)
and 48.61 (min) (rotamers, CH2–CH2–N), 49.68 (br, corresponded
to two different carbons, N–CH2–CON and N–CH2–COOH), 67.08
(maj) and 67.27 (min) (rotamers, Fmoc–CH–CH2O), 114.94 (maj)
and 115.02 (min) (rotamers, NAC@CH), 120.82 (CH Fmoc arom),
126.20 (CH Fmoc arom), 127.99 (corresponded to two different car-
bons, CH Fmoc arom and CH arom), 128.52 (CH Fmoc arom), 129.62
(CH arom), 132.80 (CH arom), 133.40 (min) and 133.68 (maj) (rota-
mers, C@CHAN), 134.19 (min) and 135.20 (maj) (rotamers, CO–C
arom), 142.14 (CFmoc), 145.23 (C Fmoc), 150.13 (NH–CO–N),
161.22 (NH–COO–CH2), 165.98 (HC@CACOANH), 168.18 (N–CH2–
CON), 168.69 (NH–CO–benzyl), 171.09 (CH2–COOH). Electrospray
mass spectrum (m/z): 610 [MꢀH]ꢀ (100%). High Resolution ESI Mass
Determination: Found: 610.1961; calcd for C32H28N5O8, 610.1938.
3. Results and discussion
When using the Fmoc (9-fluorenylmethoxycarbonyl)/Bhoc
(benzhydryloxycarbonyl) strategy to prepare peptide nucleic acids
(PNAs), the primary amino group of the PNA monomers is pro-
tected with a Fmoc group and the exocyclic amino group of the
adenine (A), guanine (G) and cytosine (C) monomers with the Bhoc
group [21]. The carboxylic acid group is unprotected, ready to be
activated by an activator agent in the presence of a base (generally
a mixture of diisopropylamine and lutidine). Optimisation of the
synthesis protocols found that O-(7-azabenzotriazol-1-yl)-N,N,N0,
N0-tetramethyluronium hexafluorophosphate (HATU) gave the
highest average coupling yields and therefore HATU is now
generally preferred over others activator agents, such as O-(ben-
zotriazol-1-yl)-N,N,N0,N0-tetramethyluronium
hexafluorophos-
phate (HBTU) for the synthesis of PNAs [22]. Comprehensive
reports on the different PNA synthesis methods, as well as apprais-
als of their advantages and disadvantages can be found in the liter-
ature [21–24]. We recently reported a new ferrocenyl uracil PNA
monomer, tert-butyl-2-(N-(2-(((9H-fluoren-9-yl)methoxy)carbon-
ylamino)ethyl)-2-(5-(N-ferrocenylmethylbenzamido)-2,4-dioxo-
3,4-dihydropyrimidin-1(2H)-yl)acetamido)acetate (1 in Fig. 1),
which was developed to incorporate ferrocenyl units into PNA se-
quences. Cleavage of the tert-butyl group of 1 to give 2-(N-(2-
(((9H-fluoren-9-yl)methoxy)carbonylamino)ethyl)-2-(5-(N-ferr-
ocenylmethylbenzamido)-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-
yl)acetamido)acetic acid (2) (Fig. 1) is necessary before it can be
used in either automated PNA synthesis [21] or manual solid phase
synthesis (for a recent example of manual synthesis of metal-la-
belled PNAs, see [25]).
The first synthetic method used to cleave the tert-butyl group of
1 was a mixture of trifluoroacetic acid (TFA):m-cresol 4:1 (v/v),
which is typically employed during standard PNA synthesis proce-
dures that cleave both the PNA from the resin and the Bhoc groups
from the exocyclic amine of the base [21]. However, when applied
to 1, our mixture turned from orange to green, indicative of oxida-
tion to the ferricenium ion. Milder acidic conditions, such as dilut-
ing the TFA in dichloromethane or the use of another scavenger
(triethylsilane) did not yield the desired product; either the initial
orange mixture turned green or only the starting material was
recovered. Oxidation of ferrocenyl compounds in biomolecules
has previously been reported by Tartar et al., when they synthes-
ised a ferrocene containing pentapeptide on an automated peptide
synthesiser [26]. These workers showed that the addition of ascor-
bic acid facilitated the formation of the desired ferrocenyl peptide
[26]. In our case, the addition of even a large excess of antioxidant
agents, such as ascorbic acid and Na2SO3, prevented the mixture
from turning green but in all cases the expected product 2 was
not obtained.
Use of 4 M HCl in dioxane further diluted in CH2Cl2 led to the
isolation of tert-butyl 2-(N-(2-(((9H-fluoren-9-yl)methoxy)carbon-
ylamino)ethyl)-2-(5-benzamido-2,4-dioxo-3,4-dihydropyrimidin-
1(2H)-yl)acetamido)acetate (3) and 2-(N-(2-(((9H-fluoren-9-
yl)methoxy)carbonylamino)ethyl)-2-(5-benzamido-2,4-dioxo-3,4-
dihydropyrimidin-1(2H)-yl)acetamido)acetic acid (4) (Fig. 1),
indicative of the cleavage of the ferrocenyl moiety from the mole-
cule. Evidence of the presence of 3 and 4 were given by ESI–MS
spectrometry with a peak at m/z = 690 and 610, corresponding to
[M+Na]+ and [MꢀH]ꢀ, respectively, and the concordance of the
High Resolution ESI-Mass spectra. The cleavage of the ferrocenyl
moiety was further confirmed by the disappearance of characteris-
tics 1H and 13C NMR signals (see SI for 1H NMR spectra of 3 and 4).
Furthermore, the two signals in the 1H spectrum corresponding to
the magnetically inequivalent protons of the CH2 linking the ferro-
cene to the amino uracil derivative [18] (in particular, a signal at
2.3.4. tert-Butyl-2-(N-(2-(((9H-fluoren-9-yl)methoxy)carbonylamino)
ethyl)-2-(5-benzamido-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-
yl)acetamido)acetate (3) and hydroxymethylferrocene (5)
The filtrate obtained in the synthesis of 4 was evaporated to
dryness and the orange residual solid purified by column chro-
matography on silica with a gradient from hexane:ethyl acetate
2:1 to hexane:ethyl acetate 1:3 as the eluent. An orange solid,
hydroxymethylferrocene (5) (Rf(5) = 0.76 in hexane:ethyl acetate
1:1), and a white solid, (3) (Rf(3) = 0.72 in hexane:ethyl acetate
1:3) were isolated and identified. Estimated yields for 5 (3 mg,
24%) and for 3 (4 mg, 10%). Characterisation of 3. Selected IR
bands (KBr;
m
, cmꢀ1): 3200 w br, 3049 w, 2924 m, 2860 w,
1625–1750 s br, 1544 m, 1456 s, 1411 m, 1383 m, 1246 m,
1154 m, 1104 w, 1027 w, 800 w, 760 w, 742 w, 705 w. 1H
NMR Spectrum (acetone-d6): 1.33 (maj) and 1.42 (min) (s, 9H,
C(CH3)3), 3.10–3.55 (m, 4H, NH–CH2–CH2 and CH2–CH2–N),
3.95 (maj) (rotamers, s, N–CH2–CON), 4.10-4.20 (m, Fmoc–CH–
CH2O and rotamers (min) N–CH2–CON), 4.21 (min) and 4.31
(maj) (rotamers, m, 2H, Fmoc–CH–CH2O), 4.63 (maj) and 4.79
(min) (rotamers, s, 2H, N–CH2–COOC(CH3)3), 6.41 (min) and
6.64 (maj) (rotamers, br s, 1H, CH2–NH–COO), 7.18–7.32 (m,
4H, CH Fmoc arom), 7.39–7.49 (m, 3H, CH arom), 7.57–7.62
(m, 2H, CH Fmoc arom), 7.71–7.83 (m, 4H, 2 ꢁ CH arom and
2 ꢁ CH Fmoc arom), 8.39 (maj) and 8.41 (min) (rotamers, 1H,
C@CHAN), 8.46 (min) and 8.48 (maj) (rotamers, s, 1H, Ar–CO–
NH), 10.53 (s, 1H, CO–NH–CO). 13C NMR Spectrum (acetone-
d6): d 28.57 (COO–C(CH3)3, 39.91 (min) and 40.31 (maj) (rota-
mers, NH–CH2–CH2)), 48.48 (Fmoc-CH–CH2O), 48.88 (maj) and
48.92 (min) (rotamers, CH2–CH2–N, 49.04 (maj) (rotamers, N–
CH2–CON), 50.05 (maj) and 50.15 (min) (N–CH2–COOC(CH3)3),
51.58 (min) (rotamers, N–CH2–CON), 67.38 (maj) and 67.54
(min) (rotamers, Fmoc–CH–CH2O), 82.30 (COO–C(CH3)3), 115.25
(maj) and 115.33 (min) (NAC@CH), 121.10 (min) and 121.13
(maj) (CH Fmoc), 126.50 (CH Fmoc), 128.33 (CH Fmoc), 128.84
(CH Fmoc), 129.94 (CH arom, corresponds to two different car-
bons), 133.10 (CH arom), 133.78 (min) and 134.10 (maj) (rota-
mers, C@CHAN), 135.27 (CO–C arom), 142.49 (CFmoc), 145.55
CFmoc), 150.44 (NH–CO–N), 161.58 (NH–COO–CH2), 165.88 (C–
CO–NH), 168.38 (N–CH2–CON), 169.00 (NH–CO–benzyl), 169.62
(CH2–COO–C(CH3)3). Electrospray mass spectrum (m/z): 690
[M+Na]ꢀ (100%). High Resolution ESI Mass Determination:
Found: 690.2535; calcd for C36H37N5O8Na, 690.2540. Character-
isation data of 5. The analytical data of the products matched
that reported previously [20].

G. Gasser, L. Spiccia / Journal of Organometallic Chemistry 693 (2008) 2478–2482
2481
Fe
O
O
O
N
O
N
HN
N
NH
NH
Fe
+
O
O
O
O
O
O
Fe
H
Fmoc
N
Fmoc
N
N
H
OC(CH3)3
N
H
OC(CH3)3
1
6
3
H
O
- H+
O
+ H2O
H
H
Fe
Fe
Fe
6
5
Scheme 1. Proposed mechanism of the formation of 3 and 5 from 1.
ca. 5 ppm with a geminal coupling constant of 13.7 Hz) were no
longer present in the spectra of both 3 and 4. The main difference
between the 1H spectrum of 3 and 4 are the two singlets at 1.33
(maj. rotamer) and 1.42 ppm (min. rotamer) corresponding to the
protons of the tert-butyl group in 3 that are absent in 4. An inter-
esting feature in the 1H NMR spectra of 3 and 4 is the presence of
two singlets at 8.39 and 8.41 ppm (rotamers) corresponding to the
CH proton on the uracil base. These resonances are shifted signifi-
cantly downfield upon the removal of the ferrocenyl moiety com-
pared to the 1H NMR spectrum of 1, where they appear at 7.11 and
7.19 ppm [18].
The isolation of 3 and 4 implies the formation of other ferroce-
nyl derivatives. In fact, the known ferrocenyl compound, hydrox-
ymethylferrocene (5 in Fig. 1) was isolated from the reaction
mixture. Another less polar ferrocenyl compound was also present
but could not be identified. Interestingly, we recently found that 5
was a side-product of the coupling of (ferrocenylmethyl)trimethy-
lammonium iodide and 1,4,7-(triformyl)-1,4,7,10-tetraazacyclod-
odecane to give 1-(ferrocenemethyl)-4,7,10-(triformyl)-1,4,7,
10-tetraazacyclododecane in aqueous solution [27]. A possible
pathway for the formation of 3–5 is presented in Scheme 1 and
involves the simultaneous formation of 3 and the carbocation 6
which has been previously reported to be stabilised in acidic media
[28–31]. Interestingly, the fact that both 3 and 4 were formed sug-
gests that the cleavage of ferrocenyl group is faster than the cleav-
age of the tert-butyl group as 2 would have been isolated if the
working under argon in degassed solvent, was previously reported
by Metzler-Nolte et al., but the decomposition compounds were
not reported [32]. Maiorana et al. recently reported similar difficul-
ties to ours when synthesising a triferrocenyl thymine PNA
monomer. The instability of ferrocene in acidic media led this
group to use basic conditions [13]. In light of these findings, we
decided to change our synthetic strategy. The new approach was
similar to that reported by Metzler-Nolte et al. during their first
description of the automated synthesis of a ferrocenyl PNA se-
quence [9], in which acidic conditions were not used because they
employed highly cross-linked polystyrene beads as the solid sup-
port that was functionalised with a glycine moiety via a p-hydrox-
ymethylbenzoic acid linker. Importantly, the removal of the PNA
sequence from this solid support does not require TFA, but metha-
nolic ammonia which was tolerated by their ferrocenyl PNA se-
quence. This same strategy was used to incorporate an amino
acid bearing ferrocenyl unit into oligopeptides [33]. With this in
mind, aqueous 2 M NaOH diluted in acetone was employed to
cleave the tert-butyl group of 1. This also cleaved the base sensitive
Fmoc protecting group, forming the zwitterion of 2-(N-(2-amino-
ethyl)-2-(5-(N-ferrocenylmethylbenzamido)-2,4-dioxo-3,4-dihy-
dropyrimidin-1(2H)-yl)acetamido)acetic acid (7) when neutralised
to pH 7 with aqueous 20% citric acid solution (Scheme 2).
Compound 7 was not isolated but was directly reacted with one
equivalent of N-(9-fluorenylmethoxycarbonyloxy)succinimide
(Fmoc-OSu) at pH 8 (adjusted with saturated NaHCO3 (aq) as
reported by Coull et al. for an adenine PNA monomer [34]). The
expected compound 2 was obtained after purification by column
chromatography on silica (see SI). Evidence of the successful
synthesis of 2 is given by the ESI–MS with a peak at m/z 808 cor-
responding to the [MꢀH]ꢀ ion, the concordance of the High Reso-
lution ESI–MS data (found: 832.2046; calcd for C43H39FeN5NaO8,
832.2038) and the disappearance in the 1H and 13C NMR spectra
of the proton and carbon signals, respectively, corresponding to
cleavage of the ester was faster. The p-system of the ferrocene in
1 is probably responsible for the (hyper-) sensitivity to hydrolysis.
The carbocation 6 is then highly susceptible to reaction with nucle-
ophiles and will react readily with a water molecule to give 5. The
hygroscopic character of the HCl in dioxane solution would provide
sufficient water for this reaction to take place.
Decomposition of ferrocenyl amino acid derivatives in a mix-
ture of 50% TFA in CH2Cl2, despite the use of ascorbic acid and
O
O
O
O
O
N
O
N
N
N
N
(a)
NH
(b)
NH
NH
Fe
Fe
Fe
N
O
O
O
O
O
O
O
O
O
Fmoc
N
Fmoc
N
N
N
OC(CH3)3
N
H
OH
O
H3N
H
1
2
7
Scheme 2. (a) i. 2M NaOH(aq): acetone 1:1 (v/v), 5 h, r.t.; ii. 20% citric acid aqueous solution until pH 7; (b) sat. NaHCO3(aq), Fmoc-OSu, 24 h, r.t., Total yield. 38% over 2 steps.

2482
G. Gasser, L. Spiccia / Journal of Organometallic Chemistry 693 (2008) 2478–2482
the tert-butyl group. As for 3 and 4, a very broad band between
1625 and 1750 cmꢀ1 is observed in the IR spectrum of 2, which
is attributed to the mCO stretch of several different types of carbonyl
References
[1] P.E. Nielsen, M. Egholm, R.H. Berg, O. Buchardt, Science 254 (1991) 1497–1500.
[2] N. Metzler-Nolte, Conjugates of peptides and PNA with organometallic
functional groups present in the compound. The presence of rota-
mers, which were detected by NMR spectroscopy, gives rise to
additional carbonyl vibrations and adds complexity to this region
of the spectrum. The 1H NMR spectrum exhibits some other inter-
esting features. Contrary to that observed for 3 and 4 (see above),
no singlets at 8.39 and 8.41 ppm (rotamers) corresponding to the
alkene CH proton on the uracil base are observed for 2. They appear
in the multiplet at 7.05–7.35 ppm. This is consistent with that
found for 1 for which the peaks were observed at 7.11 and
7.19 ppm [18]. As for 1, the two protons of the CH2 linking the fer-
rocene to the uracil base are not magnetically equivalent and
therefore appeared as two doublets with a high value geminal con-
stant (ꢂ14 Hz). Contrary to what was found for 1, the N–CH2–
COOH protons in 2 appeared as two doublets (at 4.48 and
4.65 ppm with a geminal coupling constant of about 16 Hz), and
not as a single singlet [18]. The assignment of these two protons
was aided by 2D NMR spectroscopy which indicated that these
two protons correlated to each other (COSY) and to the same car-
bon (HMQC).
complexes:
Bioorganometallics:
Weinheim, 2006.
syntheses
and
applications,
Labelling,
in:
Medicine,
G.
Jaouen
(Ed.),
Biomolecules,
Wiley-VCH,
[3] C. Baldoli, S. Maiorana, E. Licandro, G. Zinzalla, D. Perdicchia, Org. Lett. 4 (24)
(2002) 4341–4344.
[4] C. Baldoli, P. Cerea, C. Giannini, E. Licandro, C. Rigamonti, S. Maiorana, Synlett
13 (2005) 1984–1994.
[5] C. Baldoli, C. Giannini, E. Licandro, S. Maiorana, G. Zinzalla, Synlett 6 (2004)
1044–1048.
[6] C. Baldoli, L. Falciola, E. Licandro, S. Maiorana, P. Mussini, P. Ramani, C.
Rigamonti, G. Zinzalla, J. Organomet. Chem. 689 (2004) 4791–4802.
[7] A. Maurer, H.-B. Kraatz, N. Metzler-Nolte, Eur. J. Inorg. Chem. (2005) 3207–
3210.
[8] R.H.E. Hudson, G. Li, J. Tse, Tetrahedron Lett. 43 (2002) 1381–1386.
[9] J.C. Verheijen, G.A. van der Marel, J.H. van Boom, N. Metzler-Nolte,
Bioconjuguate Chem. 11 (6) (2000) 741–743.
[10] A. Hess, N. Metzler-Nolte, Chem. Commun. (1999) 885–886.
[11] C. Baldoli, E. Licandro, S. Maiorana, D. Resemini, C. Rigamonti, L. Falciola, M.
Longhi, P.R. Mussini, J. Electroanal. Chem. 585 (2005) 197–205.
[12] R. Hamzavi, T. Happ, K. Weitershaus, N. Metzler-Nolte, J. Organomet. Chem.
689 (25) (2004) 4745–4750.
[13] C. Baldoli, C. Rigamonti, S. Maiorana, E. Licandro, L. Falciola, P.R. Mussini,
Chem. Eur. J. 12 (15) (2006) 4091–4100.
[14] D.R. van Staveren, N. Metzler-Nolte, Chem. Rev. 104 (2004) 5931–5985. and
references therein.
[15] P.E. Nielsen, M. Egholm, An introduction to PNA, in: P.E. Nielsen, M. Egholm
(Eds.), Peptide Nucleic Acids, Protocols and Applications, Horizon Scientific
Press, Wymondham, UK, 1999, pp. 1–19.
4. Conclusion
[16] J. Wang, PNA biosensor for nucleic acid detection, in: P.E. Nielsen, M. Egholm
(Eds.), Peptide Nucleic Acids, Protocols and Applications, Horizon Scientific
Press, Wymondham, Uk, 1999, pp. 155–161.
[17] J.H.R. Tucker, S.R. Collinson, Chem. Soc. Rev. 31 (2002) 147–156. and
references therein.
A new ferrocenyl PNA monomer (2) that could potentially be in-
serted at any chosen position within a PNA sequence has been pre-
pared via a one-pot two step synthesis involving the simultaneous
saponification of the tert-butyl group and the cleavage of the Fmoc
[18] G. Gasser, M.J. Belousoff, A.M. Bond, L. Spiccia, J. Org. Chem 71 (2006) 7565–
7573.
protecting
group
of
tert-butyl-2-(N-(2-(((9H-fluoren-9-
[19] W.L.F. Armarego, D.D. Perrin, Purification of Laboratory Chemicals, 4th ed.,
Butterworth-Heinemann, Oxford, UK, 1996.
yl)methoxy)carbonylamino)ethyl)-2-(5-(N-ferrocenylmethylben-
zamido)-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)ace-
tate (1), followed by the reprotection of the amino end of the PNA
backbone. Initial attempts using acidic conditions were found to be
unsuccessful, leading to either no reaction or to ferrocenylmethyl
decomposition and indicated that the introduction of a ferrocenyl
PNA monomer into a PNA sequence by automation and the subse-
quent cleavage from the resin is a real chemical challenge. While
the ferrocenyl PNA monomer 2 is unsuitable for use in PNA synthe-
sis that applies the XAL/PAL resin [21], these limitations may be
overcome by using Fmoc/N-acyl protected PNA monomers
[35,36] and highly cross-linked polystyrene beads functionalised
with a glycine moiety as the solid support, as reported by Met-
zler-Nolte et al. [9] This approach allows basic conditions (etha-
nolic ammonia) to be used to cleave the PNA from the resin.
[20] The 1H NMR spectrum obtained in our laboratories is in accordance with this
one furnished by Aldrich, Inc.
[21] R. Casale, I.S. Jensen, M. Egholm, Synthesis of PNA oligomers by Fmoc
chemistry, in: P.E. Nielsen, M. Egholm (Eds.), Peptide Nucleic Acids, Protocol
and Applications, Horizon Scientific Press, Wymondham, UK, 1999, pp. 39–50.
[22] T. Koch, PNA oligomers synthesis by Boc chemistry, in: P.E. Nielsen, M. Egholm
(Eds.), Peptide Nucleic Acids, Protocols and Applications, Horizon Scientific
Press, Wymondham, UK, 1999, pp. 21–37.
[23] E. Uhlmann, B. Greiner, G. Breiphol, PNA/DNA chimeras, in: P.E. Nielsen, M.
Egholm (Eds.), Peptide Nucleic Acids, Protocols and Applications, Horizon
Scientific Press, Wymondham, UK, 1999, pp. 51–70.
[24] R. Vinayak, Synthesis of PNA–DNA chimera by MMT chemistry, in: P.E.
Nielsen, M. Egholm (Eds.), Peptide Nucleic Acids, Protocols and Applications,
Horizon Scientific Press, Wymondham, UK, 1999, pp. 71–80.
[25] S.I. Kirin, I. Ott, R. Gust, W. Mier, T. Weyhermueller, N. Metzler-Nolte, Angew.
Chem., Int. Ed. 47 (2008) 955–959.
[26] E. Cuingnet, C. Sergheraert, A. Tartar, M. Dautrevaux, J. Organomet. Chem. 195
(1980) 325.
[27] G. Gasser, A.J. Fischmann, C.M. Forsyth, L. Spiccia, J. Organomet. Chem. 692
(2007) 3835–3840.
ˇ
´
[28] V. Kovac, V. Rapic, I. Sušnik, M. Šuprina, J. Organomet. Chem 530 (1997) 149–
158.
Acknowledgements
[29] R. Herrmann, I. Ugi, Tetrahedron 37 (1981) 1001–1009.
[30] E.A. Hill, R. Wiesner, J. Am. Chem. Soc. 91 (1969) 509–510.
[31] C.A. Bunton, N. Carrasco, W.E. Watts, J. Chem. Soc., Perkin Trans. II (1979)
1267–1273.
This work was supported by the Swiss National Science Founda-
tion (SNSF), through the award of a Swiss Fellowship for Prospec-
tive Researchers Grant to G.G. (PBNE2-106771) and the
Australian Research Council (ARC) through the Australian Centre
for Electromaterials Science.
[32] J. Sehnert, A. Hess, N. Metzler-Nolte, J. of Organomet. Chem (2001) 637–639.
349–355 and references therein.
´
´
[33] L. Barišic, V. Rapic, N. Metzler-Nolte, Eur. J. Inorg. Chem (2006) 4019–4021.
[34] J.M. Coull, M. Egholm, R.P. Hodge, M. Ismail, S.B. Rajur, Synthons for the
synthesis and deprotection of peptide nucleic acids under mild conditions.
United States Patent, 6,133,444, 2000.
Appendix A. Supplementary material
[35] F. Bergmann, W. Bannwarth, S. Tam, Tetrahedron Lett. 36 (1995) 6823–6826.
[36] Z. Timar, L. Kovacs, G. Kovacs, Z. Schmel, J. Chem. Soc., Perkin Trans. 1 1 (2000)
19–26.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jorganchem.2008.05.009.