Communications
The interaction between Pna1 and Pna2–4 was next
investigated by fluorescence spectroscopy in the presence and
in the absence of various types of DNA. Three different DNA
sequences (Quad1–3; Table 1) were tested for their capacity
to template the fluorogenic reaction of cyanine dye forma-
tion. Quad1 corresponds to the previously reported
ckit21T[7b–c] quadruplex-forming sequence with two addi-
tional single-stranded flanking arms, located downstream and
upstream of the quadruplex, and complementary to Pna1 and
Pna2–4, respectively. Quad2 differs from Quad1 by only four
G!A mutations to prevent G-quadruplex formation. Quad3
contains the same ckit21T sequence as Quad1 but the
quadruplex-forming motif is now flanked with randomized
single-stranded arms that are not complementary to the
fluorogenic PNAs Pna1–4. Each DNA (200 mm) was folded in
a potassium phosphate buffer (10 mm, pH 7.4) that also
contained 100 mm KCl.[22] Under such conditions, Quad1 and
Quad3 formed a parallel-stranded quadruplex (see the
Supporting Information, Figure S1) whereas Quad2 remained
single-stranded. Briefly, a stoichiometric mixture of aldehyde
(Pna1) and indoline (Pna2, Pna3, or Pna4) in potassium-
containing buffer was incubated at room temperature in the
presence or in the absence of an equimolar amount of folded
DNA. The reaction of cyanine dye formation was then
monitored by fluorescence spectroscopy at different time
points. First, the reaction between the PNA aldehyde (Pna1)
and the PNA indoline (Pna2) was investigated. Interestingly,
only very moderate fluorescence was detectable when work-
ing at a PNA strand concentration up to 500 nm. However,
when adding a stoichiometric amount (500 nm) of folded
quadruplex Quad1 to the previous mixture, a strong fluores-
cence signal instantaneously appeared which increased up to
after 2 h (Figure 2), at which time equilibrium was finally
reached (Supporting Information, Figure S2). At equilibrium,
a 45-fold increase in fluorescence intensity was observed
compared to the quadruplex-free experiment.
Figure 2. a) Fluorescence emission spectra (lexc =540 nm) of a mixture
of Pna1, Pna2, and Quad1 (500 nm each) in potassium phosphate
buffer (10 mm, pH 7.4) and 100 mm KCl after 10 min, 1 h, and 2 h
(bottom to top) at RT. b) Fluorescence emission spectra
(lexc =540 nm) of a mixture of Pna1 and Pna2 (500 nm each) in
*
potassium phosphate buffer and in the absence ( ) or in the presence
!
&
^
of 500 nm of Quad1 ( ), Quad2 ( ), or Quad3 ( ). Fluorescence
A similar trend, although of weaker intensity, was also
observed when decreasing the PNA and DNA concentrations
down to 200 nm (Supporting Information, Figure S3). To
demonstrate that the efficiency of the fluorogenic reaction
was indeed linked to quadruplex formation, the same
stoichiometric mixture of Pna1 and Pna2 (500 nm each) was
reacted in potassium phosphate buffer and in the presence of
either Quad2 or Quad3. Key mutations of the ckit21T
sequence to prevent quadruplex formation resulted in a
complete inhibition of the fluorogenic reaction. Random-
ization of the quadruplex flanking sequences to prevent
PNA:DNA hybridization also led to a significant inhibition
compared to the reaction templated by Quad1. These results
are consistent with the proposed model suggesting that
hybridization of the both aldehyde and indole PNAs to the
quadruplex flanking regions associated with folding of the
central DNA sequence into a quadruplex conformation are
the only conditions that bring both reactive groups in close
enough proximity to form the fluorescent cyanine dye. If only
one of those requirements is satisfied, no or little reaction will
take place.
spectra were recorded after 2 h.
investigated. Pna3 differs from Pna2 by two extra methylene
groups between the heterocycle and the PNA scaffold.
Although
a specific quadruplex-templating effect was
observed when mixing Pna1 and Pna3 in the presence of
Quad1 which was similar to that obtained with Pna2, it was
significantly weaker, thus suggesting the influence of the
linker (for example flexibility) on the reaction efficiency
(Supporting Information, Figure S4).
An interesting intrinsic property of cyanine dyes is the
possibility to tune their spectroscopic properties by varying
either the nature of the nitrogen-containing heterocycles or
the length of the polymethine chain between them. To shift
our quadruplex-specific fluorescent biosensor toward longer
wavelengths, Pna4 was synthesized, which differs from Pna2
by the substitution of the indoline moiety by a benz[e]indoline
(Table 1). Reaction of Pna4 with Pna1 was then expected to
generate an unsymmetrical cyanine dye absorbing and emit-
ting at significantly longer wavelengths than the symmetrical
dye formed upon reaction between Pna1 and Pna2
(Scheme 1).[23] Although no reaction was observed when
The influence of the linker between the PNA and the
indoline on the efficiency of the fluorogenic reaction was then
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 2738 –2742