1
Suppo
rting
Infor
mation
Inorgan
ic
-
Organ
ic Hyb
rid Fluorescent Binary Probe for DNA Detection
Based on
Spin
-
Forbidden Reson
ance Energy
Tran
sfer
An
gel A. Martí,
†
Cindy A. Pu
ckett,
‡
Joanne Dyer,
†,§
Nathan Stevens,
†
Steffen
Jockusch,
†
Jingyue Ju,
||,
⊥
Jacqueli
ne K. Ba
rton,
‡
Nicholas J. Turro
†,
⊥
,
*
†
Department of Chemistry and
⊥
Department of Chemical Engineering Columbia University, New
York, NY 10027.
‡
Division of Chemistry and Chemical Engineering, California Institute of
Technology, Pasadena, California 91125
.
||
Columbia Genome Center, Columbia University College
of Physicians and Surgeons, New York, NY 10032.
Exp
erimental procedure
Prob
e sequence
The probe
seque
nces
are com
plementary
to
a region
of
Aplysia
californica
sensorin mRNA
.
A region
low
in seco
nda
ry structure was selected as the target for the
binary probe
based on
the mode
led seconda
ry structure. The mode
ling
details have been
repor
ted elsewhere
.
1
Ru
-
probe
:
5’
-
AAG TTG ATC AAG TTG GT
-
(Ru(bpy
’)(DIP)
2
2+
)
-
3'
Alexa
-
Probe
:
5’
-
AAG
TTG ATC AAG TTG GT
-
(Alexa488)
-
3'
Cy5
-
Probe
-
1:
5’
-
Cy5
-
TAT GTT TCA CTG GAT GA
-
3’
Cy5
-
Probe
-
2:
5’
-
Cy5
-
ATG
TT
T
CA
C
TG
G
AT
G
A
-
3’
Cy5
-
Probe
-
3:
5’
-
Cy5
-
TTC ACT GGA TGA
-
3’
Target: 5’
-
TCA TCC AGT GAA ACA TAC AGC ACC AAC TTG ATC AAC TT
-
3’
Probe
Systems
PS1: Ru
-
P
robe
+ Cy5
-
Probe
-
1
PS2: Ru
-
Probe
+ Cy5
-
Probe
-
2
PS3: Ru
-
Probe
+ Cy5
-
Probe
-
3
PS4: Alexa
-
Probe
+ Cy5
-
Probe
-
3
Prob
e syn
thesis
[Ru(DIP)
2
bpy’]Cl
2
.
Ru(DIP)
2
Cl
2
was synt
hesized in analogous
fashion
to the publ
ished
synt
hesis of Ru(bpy)
2
Cl
2
.
2
Ru(DIP)
2
bpy’
2+
was prepared
by
refluxi
ng
41
mg
of
Ru(DIP)
2
Cl
2
(49
μmol) and
16.
4 mg (64
μmol) of 4
-
(3
-
carboxyp
ropyl
)
-
4’
-
methyl
-
2,2’
-
bipyr
idine (prepared
acc
ording
to
the publ
ished
procedur
e
3
) in
10
mL of 1
-
1
ethanol
/water for 3 h. The mixture was cool
ed to room
temperature and
the ethanol
remove
d in vacuo.
The solution
was diluted with water (20 mL) and
filtered. The
com
plex was precipitated as the PF
6
-
salt by
addi
tion
of NH
4
PF
6
, then returned to the Cl
-
salt using
a Sepha
dex DEAE anion
exchange
column. ESI
-
MS: m/z = 511
.
Ru
-
prob
e
.
Ru(DIP)
2
bpy’
2+
was tethered to the 3’
-
end
of DNA by
first coupl
ing the
com
plex to amine
-
modi
fied beads, follow
ed by
DNA synt
hesis and
cleavage of the Ru
-
2
DNA conj
uga
te from
the beads.
4
The FMOC group was remove
d from 3’
-
amino
-
modi
fier
C7 CPG 500
beads (Glen Research) by
incuba
tion with 20%
piperidine in DMF for 15
min. The beads were rinsed with DMF and
CH
3
CN, dri
ed in vacuo,
then placed unde
r
Ar. To the beads (2 μmol), [Ru(DIP)
2
bpy’
]Cl
2
(4.5 mg, 4 μmol), HBTU (1.5 mg, 4
μmol), HOBT (0.6 mg, 4 μmol), and
DIEA (2 μL, 12
μmol) in anhydr
ous
DMF (1.5 mL)
were adde
d. The reaction
mixture was shaken for 30
min. at room
temperature. The
beads were rinsed with DMF, CH
3
CN, and
CH
2
Cl
2
, then divided into two aliquot
s and
transferred into two DNA synt
hesis colum
ns. DNA was synt
hesized using
an ABI 3400
DNA synt
hesizer. The DNA was cleaved from
the beads and
deprotected wit
h conc
.
NH
4
OH (2 h at RT, 6 h at 60
°C
).
The Ru
-
DNA conj
uga
te was pur
ified by
HPLC using
a
gradient of 5:95
to 65:
35
(acetoni
trile/50
mM ammoni
um
acetate) ove
r 30
min. The
DMT was remove
d with 80%
acetic acid for 15
min. at room
temperature, follow
ed by
addi
tion
of ethanol
, and
remova
l of solvent in vacuo.
The Ru
-
DNA conj
uga
te was
pur
ified onc
e more by HPLC. MALDI
-
TOF: 647
3 (obs
’d), 6477
(calc’d).
Cy5
-
prob
es
.
DNA was synt
hesized using
‘ultramild’ reagents with Cy5 was adde
d at the
5’
-
end,
using
a Cy5
phos
phor
amidite (Glen Research). The MMT group
was remove
d by
the DNA synt
hesizer. The DNA was cleaved and deprotected with 0.05
M pot
assium
carbona
te in methanol
for 4 h at room
temperature. To the sup
ernatant, 1.5 equi
valents
by
vol
um
e 2
M TEAA was
adde
d. The solution
was conc
entrated in vacuo
and
desalted
using
a Nap10
colum
n (GE Healthcare), eluting
with water. The Cy5
-
DNA conj
uga
te
was pur
ified by
HPLC
using
a gradient of 5:95 to 65:
35
(acetoni
trile/50
mM ammoni
um
acetate) ove
r 30
min. MALDI
-
TO
F: Cy5
-
Probe
-
1, 5748
(obs
’d), 5749
(calc’d); Cy5
-
Probe
-
2, 5442
(obs
’d), 5445
(calc’d); Cy5
-
Probe
-
3, 4193
(obs
’d), 4194
(calc’d).
Stead
y
-
state and time
-
resolved luminescence exp
eriments
Steady
-
state lum
inescence spectra were performed
on
a FL3
-
22
Fluo
rolog
-
3
spectrom
eter (J. Y. Horiba, Edison,
NJ, USA) in qua
rtz cuve
ttes of 0.4 mm path lengt
h.
In a typi
cal expe
riment, the spectra were obt
ained in solutions
of 0.5
μ
M Ru
-
probe
and
0.5
μ
M Cy5
-
probe
in 250
μ
L of 10
mM Tris
-
HCl, 400
mM NaCl, 5m
M MgC
l
2
pH
7.
5.
Target was adde
d in 1:1 propor
tion
(probe
s:target) to a final target conc
entration
of 0.5
μ
M. The expe
riments using
Alexa
-
Cy5
system (PS4) were performed similarly and
the
details are repor
ted
elsewhere
.
5
The steady
-
state spectra were corrected
for spectral
efficiencies of the mo
noc
hrom
ator and
PMT.
Time
-
resolved expe
riments were performed on an OB920
singl
e
-
phot
on
count
ing
spectrom
eter (Edinbu
rgh
Analytical Instrum
ents) with a Picoqua
nt 460
nm
pulsed LED
or 659
nm
diode
laser as excitation
sour
ce. Expone
ntial fits were obt
ained
with a program
include
d with the instrum
ent (F900,
v6.
42)
. Iterative reconvol
ution
with the instrum
ent
respons
e func
tion
(IRF) was employe
d for Cy5
decays, using
a time window
of 20
ns.
Lifetimes invol
ving
long
-
lived
Ru(bpy’
)(DIP)
2
2+
were obt
ained
by
tail
fit of the
expone
ntial decay after the IRF.
3
Stead
y
-
state luminescence spectra of PS1, PS2 and PS3
1.0
0.8
0.6
0.4
0.2
0.0
Normalized Intensity
800
750
700
650
600
550
Wavelength (nm)
1.0
0.8
0.6
0.4
0.2
0.0
Normalized Intensity
800
750
700
650
600
550
Wavelength (nm)
1.0
0.8
0.6
0.4
0.2
0.0
Normalized Intensity
800
750
700
650
600
550
Wavelength (nm)
(
a
)
P
S
1
(
b
)
P
S
2
(
c
)
P
S
3
S
/
B
=
1
0
S
/
B
=
1
2
S
/
B
=
1
4
Figu
re S1.
(a) PS1, (b) PS2, and
(c) PS3 with (
⎯
) and
without
target (
⎯
). The S/B ratio
seems to increase as the num
ber of bases in the Cy5
-
probe
decr
ease. The S/B ratio
depends
on
the extent of the increase in the signa
l of Cy5
and
the decrease in the signa
l of
Ru(bpy’
)(DIP)
2
. In buf
fer solution,
most of the backgr
ound
fluor
escence is due
to non
-
specific Cy5
fluor
escence. This fluor
escence may be due
t
o direct excitation
of the Cy5
-
probe
at the excitation
wavelengt
h of the Ru donor
or to RET due
to non
-
specific bindi
ng
of Ru(bpy’
)(DIP)
2
2+
to DNA. Ru(bpy’
)(DIP)
2
2+
is know
n to have a high
affinity for
DNA, which can cause intermolecular interaction with a
vicinal Cy5
-
probe
. Figur
e S1
show
s that this interaction
might
be less stable as the DNA seque
nce in the probe
becom
e
shor
ter. This woul
d decrease the fluor
escence intensity of the Cy5
-
probe
in the absence
of target increasing
the S/B ratio, cons
istently
with the spectra in Figur
e S1.
Time decay
an
alysis for the different PS
Legend
R
= Ru(bpy’
)(DIP)
2
2+
R
= Ru
-
probe
n
R
= Ru
-
probe
seque
nce without
Ru(bpy’
)(DIP)
2
2+
C
= Cy5
C
1
= Cy5
-
probe
1
C
2
= Cy5
-
probe
2
C
3
= Cy5
-
probe
3
n
C
1
= Cy5
-
probe
1 seque
nce without
Cy5
n
C
2
= Cy5
-
probe
2 seque
nce without
Cy5
n
C
3
= Cy5
-
probe
3 seque
nce without
Cy5
4
Tab
le S1.
Lifetime data for different PS
Entry
τ
exc.
460/em. 615
,
a
ns
τ
exc.
460/em. 667
,
a,b
ns
τ
exc.
659/em. 680
,
a
ns
(abundance, %)
1
R
696
-
-
2
R
1780
-
-
3
R
1780
-
-
4
C
1
R
1790
-
1.1 (37)
2.0 (63)
5
C
2
R
1730
-
1.1 (35)
2.1 (65)
6
C
3
R
1770
-
0.9 (43)
1.8 (57)
7
C
1
R
1620
1620
44
1.1 (34)
2.1 (66)
8
C
2
R
1600
1600
45
1.1 (28)
2.1 (72)
9
C
3
R
1680
1680
69
1.3 (33)
2.2 (67)
10
C
-
-
0.4 (21)
1.0 (79)
11
C
1
-
-
1.0
(35)
2.0 (64)
12
C
2
-
-
1.1 (40)
2.2 (60)
13
C
3
-
-
0.9 (48)
1.8 (52)
14
C
1
n
R
-
-
1.1 (41)
2.1 (59)
15
C
2
n
R
-
-
1.0 (31)
2.1 (69)
16
C
3
n
R
-
-
0.9 (44)
1.8 (56)
17
C
1
n
R
-
-
1.2 (42)
2.1 (58)
18
C
2
n
R
-
-
1.0 (36)
2.0 (64)
5
19
C
3
n
R
-
-
1.2 (37)
2.0 (63)
20
n
C
1
R
1780
-
-
21
n
C
2
R
1760
-
-
22
n
C
3
R
1750
-
-
23
n
C
1
R
1930
-
-
24
n
C
2
R
1850
-
-
25
n
C
3
R
1790
-
-
a
Uncertainty of ca.
±
10%
b
Tail exponential and biexponential fit starting after the decay of the IRF were used to extract lifetime
parameters. Since the first moments of the transient can not be use
d in the fitting routine due to overlap
with the IRF, a good portion of the contribution of the short
-
lived component is not taken into account,
affecting the determination of accurate preexponential factors. For this reason, preexponential factors are
not
reported for the lifetimes determined for the RET process (Exc 460 nm, Em. 667 nm).
The lum
inescence lifetime of Ru(bpy’
)(DIP)
2
2+
in solution
is 696
ns (Table S1,
Entry 1). The linka
ge of the com
plex to the DNA produc
e an increase in the lifetime of
the
Ru(bpy’
)(DIP)
2
2+
to 1780
ns (Entry 2), probably by
partially protecting
it from
que
nching
by
molecular oxyge
n. Addi
tion
of target DNA to the Ru
-
probe
doe
s not
change
signi
ficantly the lifetime of the probe
(Entry 3). How
ever, the addi
tion of the nC
1
-
probe
, nC
2
-
probe
, and
nC
3
-
probe
(the Cy5
-
probe seque
nces without
the Cy5)
in the
presence of target increase the lifetimes to 1930
ns, 1850
ns, and 1790
ns, respectively
(Entry 23
-
25)
. The decreasing
lifetimes are probably caused by
the decreasing
chain
lengt
h
of the Cy5
-
probe
s, which decreases the distance from
the Ru
-
probe
. This sugge
sts
that the ruthenium
com
plex doe
s not
interact strongl
y with the doubl
e strande
d (dd)
DNA
formed when Ru
-
probe
binds
to target but
it doe
s interact with the dd
DNA formed by
Cy5
-
probe
seque
nce and
the target.
In cont
rast, the Cy5
fluor
ophor
e fits to a biexpon
ential func
tion
with lifetime of
1.0 ns and
0.4 ns when free in solution
and
when excited at 659
nm
(Entry 10)
that
virtually doubl
es when it is bound
to the probe
seque
nce
(Entry 11
-
13)
. Addi
tion
of the
Ru
-
probe
, nR
-
probe
(the Ru
-
probe
without
the Ru(bpy’
)(DIP)
2
2+
) and
the target seque
nce
doe
s not
change
signi
ficantly the lifetime of Cy5
(Entry 14
-
19)
.
The lum
inescence lifetimes of the Ru
-
probe
with Cy5
-
probe
s (PS1, PS2, an
d PS3)
free in solution (Entry 4
-
6), show
values similar to that of just the Ru
-
probe
in solution
(Entry 2). Addi
tion
of target to PS1, PS2, and
PS3 produc
es an increase in the obs
erved
fluor
escence decay lifetime of Cy5
(667
nm
) in the three PS (Entry 7
-
9
), when excited at
460
nm
. This decay lifetime is more than 20
times longe
r than the fluor
escence of the
probe
excited directly (Entry 17
-
19) and
it is due to the SF
-
RET from the ruthenium
6
com
plex to Cy5.
Since the SF
-
RET is a slow
process, the energy
is t
ransferred gradua
lly
causing
a delay in the fluor
escence emission.
This fluor
escence delay increases as the SF
-
RET rate cons
tant decrease. Based on
Förster’s RET theory, the RET rate cons
tant
decreases as the distance between the RET pair increase.
6
It can be also not
ed from
Table
S1 that the delayed Cy5
lifetime (Exc. 460
nm
, Em. 667
nm
) increases as the Cy5
-
probe
s
get separated from
the Ru
-
probe
from
44
ns to 45
ns to 69
ns when the num
ber of
nuc
leotides in between the probe
s increases
from
5 to 6 to 10.
A com
put
er simulation
using
the progr
am FRETview
7
(Figur
e S2) shows delayed fluor
es
cence Cy5
lifetimes
from
62
to 76
ns when the distance between the probe
s is between 28
to 29
Å, which is a
reasona
ble estimation
for the distance of the probe
s when PS3 is hybr
idized to target. The
longe
r com
pone
nt (1.6
-
1.7
μ
s) obs
erved exciting at 460
nm
and
moni
toring
at 667
nm
(Entry 7
-
9), decays slight
ly faster than Ru
-
probe
free in solution,
and
is obs
erved also at
Ru(bpy’
)(DIP)
2
2+
emission
wavelengt
h (615
nm
) indi
cating
that it corresponds
to Ru
-
probe
not
bound
to target.
(
a
)
(
b
)
Figu
re S2.
Screen shoot
from
the FRETView simulation
of the RET com
pone
nts for PS3
(a) at a distance of 28
Å and
(b) at a distance of 29
Å:
⎯
donor
fluor
escence in the
absence of target,
⎯
donor
fluor
escence in the presence of target,
⎯
acceptor
-
delayed
fluor
escence in the prese
nce of target. The donor
decay lifetime in the presence of target
(
⎯
)
is not
detected in our
expe
rimental transients since the RET process is ove
r 95%
efficient.
Determination
of optimum time window
for TRES exp
eriments
O
pt
i
m
i
z
a
t
i
on
Ra
t
i
o=
A
c
c
e
pt
or
de
c
a
y
w
i
t
h
T
a
r
ge
t
A
c
c
e
pt
or
de
c
a
y w
i
t
hout
T
a
r
ge
t
D
onor
de
c
a
y w
i
t
h T
a
r
ge
t
D
onor
de
c
a
y w
i
t
hout
T
a
r
ge
t
The
opt
imization
ratio has the pur
pos
e of helping to determine the time window
for opt
imum
S/B ratio. The num
erator of the equation
takes into account
the increase in