Optimization of a solid-phase synthesis of a PNA monomer.

Article abstract of DOI:10.1021/ol016822y

A new scheme for the synthesis of peptide nucleic acid (PNA) is described. First, a resin-bound amino acid is alkylated under Fukuyama-Mitsunobu conditions. A nucleobase is then incorporated via an acid fluoride. Subsequently, oligomerization may be achie

Full text of DOI:10.1021/ol016822y

ORGANIC
LETTERS
2001
Vol. 3, No. 24
3931-3934
Optimization of a Solid-Phase Synthesis
of a PNA Monomer
Russell D. Viirre and Robert H. E. Hudson*
Department of Chemistry, The UniVersity of Western Ontario, London,
Ontario N6A 5B7, Canada
rhhudson@uwo.ca
Received September 26, 2001
ABSTRACT
A new scheme for the synthesis of peptide nucleic acid (PNA) is described. First, a resin-bound amino acid is alkylated under Fukuyama−
Mitsunobu conditions. A nucleobase is then incorporated via an acid fluoride. Subsequently, oligomerization may be achieved by deprotection,
coupling of another amino acid, and repetition of the cycle. Each step of the submonomer synthesis has been optimized to provide essentially
quantitative yield.
Peptide nucleic acid or PNA is perhaps one of the most
successful synthetic analogues of DNA or RNA. Since its
introduction in 1991,1 there have been over 600 publications
about PNA, from such diverse fields as chemistry, biochem-
istry, biotechnology, and medicine.2
There are however a few drawbacks which complicate the
direct use of PNA in many applications, the most severe
being low water solubility and poor cellular uptake. These
problems are most frequently addressed by the conjugation
of various moieties to the end of a PNA strand. An alternative
to this would be to construct a PNA with an N-(2-
aminoethyl)X backbone, where X, rather than glycine, could
be any readily available R-amino acid. Such modifications
can impart chirality, charge, or functionality, all without
disturbing the bond configuration required for the desired
binding properties.
While some PNA oligomers of this sort have been studied,3
we feel there is room for further investigation in this area.
The construction of such oligomers first requires the synthesis
of the appropriate PNA monomers. This can either be done
by reductive alkylation4 or Fukuyama-Mitsunobu alkylation5
of the desired R-amino acid, followed by acylation with the
desired nucleobase-acetic acid derivative. These processes
can be tedious and time-consuming. In addition, when one
The success of PNA as a nucleic acid mimic lies in its
structure. The sugar-phosphate backbone of natural nucleic
acids is replaced by a polyamide backbone consisting of
N-(2-aminoethyl)glycine units. A nucleobase is connected
to the internal nitrogen via a methylene-carbonyl linker. This
structure maintains the bond spacing of DNA/RNA (six
covalent bonds along the backbone, with the nucleobase three
bonds away from the backbone). As well, the amide bonds
confer some degree of conformational rigidity to a PNA
strand, aiding in complex formation. As a result of this
design, PNA forms exceptionally stable and sequence-
specific duplexes and triplexes with complementary DNA,
RNA, or PNA. Also, PNA is stable to degradation by
nucleases and proteases.
(3) Sforza, S.; Haaima, G.; Marchelli, R.; Nielsen, P. E. Eur. J. Org.
Chem. 1999, 194.
(4) Dueholm, K. L.; Egholm, M.; Buchardt, O. Org. Prep. Proc. Int.
(1) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science 1991,
254, 1497.
(2) For more about the applications, properties, and standard syntheses
of PNA, see the following review and references therein: Nielsen, P. E.
Acc. Chem. Res. 1999, 32, 624.
1993, 30, 1927.
(5) Fukuyama, T.; Jow, C.-K.; Cheung, M. Tetrahedron Lett. 1995, 36,
6373.
10.1021/ol016822y CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/08/2001

considers the multitude of readily available R-amino acids,
natural nucleobases, and nucleobase analogues, the number
of possible monomers to synthesize becomes unmanageable.
As a solution to this problem, we envision a submonomer
synthesis of PNA oligomers (Scheme 1). On the basis of
be removable under orthogonal conditions, all of which must
not affect the resin linkage itself. Moreover, even the
nucleobase derivatives must be protected from Mitsunobu
alkylation. In addition, since the cycle is to be repeated
several times, each step must proceed in very good yield
with no resin-bound byproducts and, preferably, as quickly
as possible. The purpose of this Letter is to communicate
the results of our efforts to optimize each step in the
submonomer synthesis scheme.
Scheme 1. Submonomer Synthesis Cycle Based on
Fukuyama-Mitsunobu Amine Synthesis6 a
Our resin of choice contains the p-benzyloxybenzyl alcohol
(Wang) linker, which offers enhanced stability toward
nucleophiles (such as thiophenolate ion, required in the
desulfonylation step) over the standard Merrifield resin, while
also remaining stable to the dilute acid required for the
detritylation step in our synthetic scheme. In addition, the
Wang resin is extremely well-suited for solid-phase optimi-
zation studies. Perhaps one of the more challenging aspects
to solid-phase chemistry is the monitoring of reactions.
Reactions on the Wang resin were conveniently and quan-
titatively monitored by periodically removing a small amount
(5-10 mg) of resin from the reaction, thoroughly rinsing
and drying, and then cleaving directly into a deuterated
1
solvent, collecting the H NMR spectrum and comparing
integrations of the compounds of interest.
The first reaction to be studied was sulfonylation. While
many solvents, bases, and concentrations were tried, it was
found that 2 equiv each of 2-nitrobenzenesulfonyl chloride
(nosyl chloride) and N,N-diisopropylethylamine (DIEA) at
a concentration of 0.2 M in dichloromethane gave the best
results. Beginning with the commercially available Fmoc-
glycine derivatized Wang resin, the Fmoc group was
removed under standard conditions, and the nosyl chloride
and DIEA solution was added. Every 2 min for 10 min, a
small amount of resin was removed from the reaction and
thoroughly rinsed. After 10 min, the resin was rinsed and
fresh reagents were added for a further 10 min. The 10
samples were dried and cleaved using CF3COOD/CDCl3, and
a (a) o-NsCl, DIEA, CH2Cl2; (b) DMTNHCH2CH2OH, TMAD,
PBu3, DIEA, THF; (c) KSPh, NMP, H2O; (d) BaseCH2COF, DIEA,
DMAP, NMP; (e) TFA, MeOH, CH2Cl2; (f) Fmoc-amino acid,
peptide coupling; (g) piperidine, DMF. X ) O or NH.
the Fukuyama-Mitsunobu alkylation, a resin-bound amino
acid is first sulfonylated and then alkylated under Mitsunobu
conditions with an N-protected 2-aminoethanol and then
desulfonylated. The secondary amino group is then acylated
with the desired protected nucleobase-acetic acid derivative,
yielding the resin-bound PNA monomer 1.
1
the H NMR spectra were recorded. The results are shown
in Figure 1a. Interestingly, the reaction was very rapid, rising
to nearly 90% completion within the first 2 min. However,
the starting material, glycine, was not completely consumed
until 4 min after the second addition of reagents. Perhaps
even more interesting though was that the complete disap-
pearance of glycine did not correspond to complete reaction,
as dinosylated glycine was formed, amounting to nearly 20%
of the material by the end of 20 min. To our knowledge,
there has been no previous report of a side reaction of this
kind. This could either be due to the use of milder
sulfonylation conditions or the fact that the dinosylated
product could be hydrolyzed to the desired mononosyl
compound on extractive workup or HPLC analysis or
purification. Nevertheless, under these conditions dinosyla-
tion does occur, and if left untreated, the disulfonylated
material would not be alkylated in the subsequent Mitsunobu
reaction, significantly decreasing the yield of the cycle.
Fortunately, a very brief treatment with piperidine solution
completely converts all dinosylated product to the desired
mononosylated product (see Figure 1b). Considering all of
At this point, rather than cleaving and purifying the
monomer for use in an oligomer synthesis, the next monomer
is constructed on the first one by deprotection of the amino
group, coupling of another amino acid, which is deprotected,
and repetition of the cycle. A PNA oligomer synthesis is
thus carried out, with each step incorporating a moiety
smaller than a single monomer or a submonomer.7 While
PNA monomers have been constructed via similar scheme,8
certain challenges are associated with the synthesis of
oligomers by this method. First, all protecting groups must
(6) The following abbreviations are used: o-Ns or nosyl (2-nitrobenzene-
sulfonyl); DEAD (diethyl azodicarboxylate); DIAD diisopropyl azodicar-
boxylate), TMAD [N,N,N′,N′-tetramethyl azodicarboxamide or 1,1′-azobis-
(N,N-dimethylformamide)]; PyBroP (bromo-tris-pyrrolidino-phosphonium
hefaxfluorophosphate); PyBOP (benzotriazole-1-yl-oxy-tris-pyrrolidino-
phosphonium hexafluorophosphate); HBTU [2-(1H-benzotriazole-1-yl)-
1,1,3,3-tetramethyluronium hexafluorophosphate]; HATU [2-(7-aza-1H-
benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate].
(7) For a different submonomer synthesis of PNA, see: Richter, L. S.;
Zuckermann, R. N. Bioorg. Med. Chem. Lett. 1995, 5, 1159.
(8) Davis, P. W.; Swayze, E. E. Biotechnol. Bioeng. 2000, 71, 19.
3932
Org. Lett., Vol. 3, No. 24, 2001

each of DMT-ethanolamine, TMAD, and PBu3 and 10 equiv
of DIEA for an hour, removing a small amount of resin every
10 min, followed by a second treatment with fresh reagents
for a second hour. Again, the extent of reaction was
determined by cleaving from the resin directly into deuterated
solvent and comparing 1H NMR signal integrations of
alkylated and nonalkylated products. For comparison, the
experiment was repeated with DEAD and with DIAD. The
results are shown in Figure 2. As can be seen, the TMAD
Figure 1. (a) Proportions of glycine derivatives at 2 min intervals
in the nosylation of glycine-derivatized Wang resin, as determined
by integration of the methylene peaks in the 1H NMR spectrum of
the crude cleavage material: 9, unreacted glycine, δ 4.10 ppm; 2,
mononosyl glycine, δ 4.16 ppm; ×, dinosyl glycine, δ 4.94 ppm.
1
(b) i. H NMR spectrum of the product after 2 × 10 min under
nosylation conditions (recorded in 20% CF3COOD/CDCl3). The
arrows indicate the signals due to the undesired dinosylated product.
ii. 1H NMR spectrum of material cleaved from the same resin after
a 1 min treatment with 30% piperidine in DMF (recorded in 7.5%
CF3COOH/CDCl3). Unlabeled peaks are due to the desired mono-
nosylated glycine.
Figure 2. Conditions for optimization of the Mitsunobu reaction,
1
and results, as determined by integration of the H NMR signals
of the indicated protons.
the above, nosylations are now routinely carried out with
two 3 min treatments of nosyl chloride/DIEA, followed by
a 1 min treatment with 30% piperidine in DMF, affording
95-100% conversion.
alkylation proceeds nearly linearly for the first 40 min, after
which the effectiveness of the reagent tapers off. The
azodiester reagents, on the other hand, seem to lose their
effectiveness after only 10 min under these conditions. Given
the above results, our protocol is to use two 40 min
treatments with TMAD, which routinely affords essentially
quantitative conversion.
The last stage of the Fukuyama-Mitsunobu synthesis to
be optimized was the desulfonylation reaction. To examine
this reaction by H NMR, the DMT protecting group first
had to be removed, to clarify the aromatic region of the NMR
spectrum. This was accomplished by successive 30 s
treatments with TFA/MeOH/CH2Cl2 (1:2:97).11 By the third
treatment, no orange color, characteristic of the DMT cation,
was generated, which was taken as an indication of complete
detritylation. The primary amino group was then acetylated.
While many reagents were tested for the denosylation
reaction, we found that 0.5 M KSPh in NMP/H2O (24:1)
The next reaction studied was the Mitsunobu alkylation.9
The protecting group of choice for the nitrogen of the
aminoethyl group is the dimethoxytrityl or DMT group. This
was chosen for its stability under basic/nucleophilic condi-
tions, its ease of removal under brief acid treatment (through
which the resin linkage is stable), and for the ease of
synthesis of DMT-ethanolamine. Mitsunobu alkylation of
nosyl glycine by DMT-ethanolamine was attempted under
a number of conditions. The standard conditions (DEAD/
PPh3) were found to be completely ineffective. The results
were improved with the use of TMAD/PBu3,10 but even this
reagent system did not afford complete alkylation within a
reasonable time period. Only upon addition of excess base
(DIEA) was the reaction seen to proceed to completion
within a few hours. To determine how quickly the reaction
proceeds, resin-bound nosyl glycine was treated with 3 equiv
1
(9) Mitsunobu, O. Synthesis 1981, 1.
(10) Tsunoda, T.; Otsuka, J.; Yamamiya, Y,; Ito, S. Chem. Lett. 1994,
539.
(11) Alsina, J.; Giralt, E.; Albericio, F. Tetrahedron Lett. 1996, 37, 4195.
Org. Lett., Vol. 3, No. 24, 2001
3933

gave the best results. The N-(acetamidoethyl)-N-nosylglycine
resin was treated with 7 equiv of the KSPh solution for 2 ×
Scheme 2. Nucleobase Submonomer Incorporation
1
Experimenta
30 min, sampling every 10 min, for analysis by H NMR.
This time, the very first sample (t ) 10 min) showed no
sign of the DMT or nosyl groups, confirming both that the
visual test for detritylation was accurate and that denosylation
occurs rapidly and is complete within 10 min. Subsequent
experiments have shown that denosylation is also complete
within 10 min when the DMT protecting group is in place.
The last step of the submonomer synthesis scheme to
require attention was the incorporation of the nucleobase
submonomer. It should be noted that even thymine which
remains unprotected for standard PNA syntheses requires a
protecting group for our synthesis scheme, as the N3 position
is efficiently alkylated under our Mitsunobu conditions (see
Supporting Information).
a 2, DIEA, catalytic DMAP, NMP, various time intervals; (b)
TFA, MeOH, CH2Cl2, 3 × 30 s; (c) Ac2O, DIEA, CH2Cl2, 1 hl (d)
95% TFA/H2O, 16 h, evaporate, then 5% CF3COOD/DMSO-d6.
Investigation of a few different protecting groups led us
to the p-methoxybenzyl (PMB) group. This can be easily
prepared and can be quantitatively deprotected by treatment
with 0.6 M AlCl3 in anisole followed by alkaline aqueous
workup.12
any R-amino acid and any nucleobase at any position in the
oligomer. This scheme includes a protection strategy in which
all deprotection conditions are completely compatible with
other protecting groups which must remain in place, includ-
ing the nucleobase protecting group and the resin linkage.
We have optimized each step in the scheme so that it
proceeds as quickly and cleanly as possible. In particular,
with respect to the Fukuyama-Mitsunobu process, we have
discovered the following: (1) Nosylation can be made to
occur extremely rapidly, but undesired dinosylated product
is formed, which can be reversed by a brief treatment with
piperidine; (2) sluggish Mitsunobu reactions can be acceler-
ated by the addition of excess base; (3) desulfonylation can
also be made to occur quite rapidly. Work is currently
underway to demonstrate the cyclic yield and overall
efficiency of the scheme by the preparation and quantitation
of a deprotected PNA oligomer. Also, our results toward the
incorporation of other nucleobases and R-amino acids will
be reported in due course.
Incorporation of the PMB-protected thymine submonomer
proved to be a difficult step to optimize. Although it appears
to be a straightforward acylation, the common acylating
agents PyBroP, PyBOP, HBTU, and HATU were all inef-
fective, perhaps due to the steric effect of the DMT protecting
group. The acid fluoride13 was prepared, isolated, and used
in combination with DIEA and catalytic DMAP to afford
the desired product. This step was also optimized according
to Scheme 2. Resin-bound N-(DMT-aminoethyl)glycine was
treated with a 0.2 M solution of the acid fluoride and DIEA
with a catalytic amount of DMAP for 1 h. Every 15 min, a
small amount of resin was removed from the reaction,
detritylated under the conditions described above, and then
acetylated under standard conditions, cleaved, and examined
1
by H NMR. A sample of the expected byproduct, the
peracetylated PNA backbone, was prepared by standard
solution-phase methods in order to determine its precise
1
chemical shifts. The H NMR spectrum showed no sign of
this compound, consisting mainly of the desired PMB-
protected, acetylated thymine PNA monomer. Acylation by
the acid fluoride then is concluded to be complete within 15
min.
Acknowledgment. This work was supported by the
National Sciences and Engineering Research Council (NSERC)
and UWO. R.D.V. is thankful for scholarship support from
the Ontario Government and NSERC.
In summary, we have demonstrated that solution-phase
1H NMR can be conveniently used to monitor solid-phase
reactions on Wang resin. We have devised a submonomer
synthesis strategy for PNA which can in principle incorporate
Supporting Information Available: Experimental pro-
cedures and spectral data for key compounds, as well as
experimental procedures for optimization experiments. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(12) Mellor, B. J.; Thomas, E. J. J. Chem. Soc., Perkin Trans. 1 1998,
747.
(13) Carpino, L. A.; Sadat-Aalaee, D.; Chao, H. G.; DeSelms, R. H. J.
Am. Chem. Soc. 1990, 112, 9651.
OL016822Y
3934
Org. Lett., Vol. 3, No. 24, 2001