of 12
A rhodium-cyanine fluorescent probe: detection and signaling of
mismatches in DNA
Adela Nano
,
Adam N. Boynton
, and
Jacqueline K. Barton
*
California Institute of Technology, Division of Chemistry and Chemical Engineering, Pasadena,
California 91125, United States
Abstract
We report a bifunctional fluorescent probe that combines a rhodium metalloinsertor with a cyanine
dye as the fluorescent reporter. The conjugate shows weak luminescence when free in solution or
with well matched DNA but exhibits a significant luminescence increase in the presence of a 27-
mer DNA duplex containing a central CC mismatch. DNA photocleavage experiments
demonstrate that, upon photoactivation, the conjugate cleaves the DNA backbone specifically near
the mismatch site on a 27-mer fragment, consistent with mismatch targeting. Fluorescence
titrations with the 27-mer duplex containing the CC mismatch reveal a DNA binding affinity of 3.1
× 10
6
M
−1
, similar to that of other rhodium metalloinsertors. Fluorescence titrations using
genomic DNA extracted from various cell lines demonstrate a clear discrimination in fluorescence
between those cell lines that are proficient or deficient in mismatch repair. This differential
luminescence reflects the sensitive detection of the mismatch repair-deficient phenotype.
Graphical abstract
Within the cell, the mismatch repair (MMR) machinery is critical for maintaining genomic
fidelity. Deficiencies in MMR result in the accumulation of base:base mismatches and
predispose the cell to cancerous transformation.
1
Various rhodium complexes have been
designed that target and bind single base pair mismatches with high specificity and
selectivity.
2
,
3
These rhodium compounds bear the sterically expansive 5,6-chrysene diimine
(
chrysi
) ligand, and preferentially target thermodynamically destabilized mismatches over
*
Corresponding Author: jkbarton@caltech.edu.
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website.
Experimental methods and supporting Figures (S1–S16) (PDF)
HHS Public Access
Author manuscript
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Published in final edited form as:
J Am Chem Soc
. 2017 December 06; 139(48): 17301–17304. doi:10.1021/jacs.7b10639.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
matched base pairs by a factor of >1000 through metalloinsertion.
4
,
5
In addition, these
complexes target and inhibit growth in MMR-deficient versus MMR-proficient cancer cells
with high selectivity.
2
,
6
Considering the strong association of MMR deficiency and cancer,
the development of early diagnostic tools for DNA mismatches and deficiencies in MMR
would be invaluable.
We have recently reported a class of ruthenium metalloinsertors that serve as luminescent
“light switches” for single base mismatches in dsDNA.
7
Luminescent Pt(II) complexes have
also been investigated as mismatch probes.
8
The use of small organic fluorophores for site-
specific targeting of DNA mismatches has been reported but still remains a challenging goal.
9
Since the rhodium metalloinsertors do not display luminescence but nonetheless show
remarkable mismatch selectivity, we have developed an alternative strategy for a
luminescent metalloinsertor, equipping the rhodium complex with an organic fluorophore to
serve as the optical reporter.
To achieve a fluorescent reporter of DNA mismatches, we have designed a bifunctional
conjugate,
RhCy3
, in which mismatch targeting is performed by the rhodium
metalloinsertor [Rh(phen)(chrysi)(dpa)]
3+
(
chrysi
= 5,6-chrysene diimine;
DPA
= 2,2
-
dipyridylamine) and the optical output by the cationic indole trimethine cyanine (Cy3)
fluorophore through a “light-up” effect (Figure 1). The two components are covalently
linked through a polyethylene glycol (PEG) linker, enhancing water solubility. Furthermore,
indocarbocyanines are known to have (i) high chemical- and photo-stability, (ii) low toxicity
and (iii) strong absorption in the visible range with a sharp increase in fluorescence when
interacting with DNA.
10
We rationalized that the conjugate would not bind tightly to well
matched DNA, resulting in weak luminescence. However, in the presence of a mismatched
DNA duplex, the conjugate would recognize the mismatched site and bind tightly to the
duplex; with Cy3 bound rigidly against the DNA groove, constraining rotation of Cy3
around the polymethine chain, an increased fluorescence from Cy3 would result. Minor
groove binding and DNA intercalation are binding modes that are both observed for
cyanines. Binding to DNA from the minor groove side is found both for the parent cyanine
10
and rhodium complex.
5
Thus, we synthesized and fully characterized the bifunctional fluorescent probe
RhCy3
containing a flexible PEG linker in a multistep procedure shown partially in Scheme 1. A
peptidic coupling performed between the primary amine
DPA-NH
2
and
Cy3-COOH
gave
DPA-Cy3
in 86% yield. The complexation of the latter with the rhodium precursor yielded
the final probe in good yield. Model compounds
[Rh(phen)(chrysi)(HDPA)]
3+
and the
Cy3-linker
(Figure 1), serving as spectroscopic references, were also prepared. The detailed
experimental protocols and full characterization of
RhCy3
and model compounds are
reported in the Supporting Information.
Electronic absorption profiles of the conjugate and spectroscopic references were obtained
in Tris buffer (5 mM Tris, 200 mM NaCl, pH = 7.4) at ambient temperature (see Table S1
and Figure S1). The UV-visible spectrum of
RhCy3
resembles the sum of the absorption
profiles of each component (Figure S1). The photoexcitation (
λ
Ex
= 520 nm) of
RhCy3
free
in Tris buffer yields a weak luminescence with a maximum emission centered at 570 nm
Nano et al.
Page 2
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
(Figure 2), slightly red-shifted compared to the spectroscopic reference
Cy3-linker
; RhCy3
is also weaker in intensity compared to the free dye as a result of static quenching by the Rh
complex. Significantly, the emission of the Rh probe (1 μM) is enhanced 9-fold in the
presence of a 27-mer dsDNA containing a central CC mismatch (MM DNA, 1 μM). For
these studies we utilized the highly destabilized CC mismatch; the parent Rh complex binds
80% of all mismatches depending upon their thermodynamic destabilization.
5
Little
fluorescence intensity increase is observed in the presence of fully well-matched (WM)
DNA (Figure 2). In addition, no mismatch-dependent luminescence is found for the
Cy3-
linker
lacking the metalloinsertor unit (Figure S2). We also examined the emission intensity
of
RhCy3
in the presence of bovine serum albumin (BSA) at different concentrations, and
no significant luminescence intensity increase was observed up to 50 equivalents of BSA
(data not shown).
Fluorescence titrations with increasing amounts of 27-mer dsDNA containing a CC
mismatch also show a strong increase in emission intensity (Figure 3), while with WM
duplex, a negligible fluorescence increase is seen. The data were fit to a one-site specific
binding equation to obtain the binding affinity of
RhCy3
,
K
B
(CC) = 3.1 × 10
6
M
−1
, to the
mismatched DNA sequence. This value is consistent with the binding affinities for other
metalloinsertors including
[Rh(phen)(chrysi)(HDPA)]Cl
3
,
11
indicating that the tethered
fluorophore does not significantly affect the specific DNA binding affinity of the final
conjugate.
Time-resolved fluorescence was also utilized to provide insight into the interaction of the
conjugate with DNA. Fluorescence decay measurements were performed as previously
described and data are gathered in Table 1.
12
The fluorescence decay profile of
RhCy3
when free in solution yields a monoexponential function with an excited state lifetime of 111
ps. The fluorescence lifetime of
RhCy3
in the presence of DNA (WM or MM) follows a
biexponential decay function. However, in the presence of WM DNA, there is only a 4%
extra population and clearly one major contribution from a population with a lifetime 116
ps. This value is in good agreement with the decay found for the free
RhCy3
indicating that
there is no significant interaction between the conjugate and WM DNA.
13
In the presence of MM DNA, the decay curve is biexponential with two excited state
lifetimes: a major contribution (85%) from a species with
τ
1
= 91 ps and a smaller
population (15%) with a significantly longer lifetime,
τ
2
= 308 ps, corresponding to the
DNA-bound
RhCy3
. Constraining the cyanine through DNA binding would be expected to
yield this increase in lifetime.
14
The specific mismatch targeting can also be evaluated through DNA photocleavage
experiments, since many of the Rh metalloinsertors promote site-specific DNA cleavage
with irradiation.
4
Following photolysis (320–440 nm), autoradiography, shown in Figure 4,
reveals specific photocleavage of the mismatched duplex at the mismatched site similarly to
[Rh(bpy)
2
(chrysi)]
3+
and no photocleavage in the presence of WM DNA. Additionally, the
Cy3-linker
lacking the metalloinsertor moiety shows no DNA damage.
Nano et al.
Page 3
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
The MMR machinery increases genome fidelity during the replication process, but if the
MMR machinery is defective, mismatches accumulate over time.
1
We therefore tested
whether our conjugate could distinguish any differences in mismatch frequency associated
with MMR-proficient versus MMR-deficient phenotypes using genomic DNA (gDNA)
extracted from different MMR-proficient and -deficient cell lines. In a prior investigation,
15
a rhodium(III) metalloinsertor bearing a benzo[
a
]phenazine-5,6-quinone diimine ligand was
incubated with gDNA extracts from various MMR-deficient and –proficient cell lines.
Indeed, following photo-activation, enhanced photocleavage was observed for those samples
containing DNA from the MMR-deficient cell lines. Based on microsatellite instability and
mutation frequency studies, there is an increase up to 1000-fold in mismatches in MMR-
deficient versus MMR-proficient cells.
16
,
1a
Remarkably,
RhCy3
fluorescence titrations
show a significant increase in fluorescence intensity in the presence of gDNA isolated from
MMR-deficient cells (HCT116O, DU145, and SKOV3) compared to gDNA isolated from
the MMR-proficient HCT116N cell line (Figure 5).
17
,
18
Thus, we find the conjugate to be
capable of differentiating the MMR-deficient versus MMR proficient phenotype, and the
results obtained for
RhCy3
correlate well with the reported mutation rates for these cell
lines.
15
,
17
,
19
This differential fluorescence enhancement reflects the remarkably high
mismatch specificity of rhodium metalloinsertors; in fact, luminescent Ru(II) light switch
complexes
7
do not exhibit comparable mismatch discrimination in genomic DNA samples.
Overall these results demonstrate the design and application of a Rh metalloinsertor-cyanine
conjugate in targeting mismatched DNA sites with a luminescent reporter of specific
binding. A remarkably high sensitivity of the metalloinsertor conjugate to the MMR-
deficient phenotype is observed which points to the application of the conjugate as a new
tool in the early diagnosis of mismatch-repair deficient cancers.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
Funding Sources
No competing financial interests have been declared.
We are grateful to the NIH for funding this work. We also thank the Beckman Institute Laser Resource Center
facilities and Dr. Jay R. Winkler for assistance.
References
1. (a) Loeb LA. Cancer Res. 2001; 61:3230–3239. [PubMed: 11309271] (b) Bhattacharya NP,
Skandalis A, Ganesh A, Groden J, Meuth M. Proc Natl Acad Sci USA. 1994; 91:6319–6323.
[PubMed: 8022779] (c) Strauss BS. Mutat Res. 1999; 437:195–203. [PubMed: 10592327]
2. Zeglis BM, Pierre VC, Barton JK. Chem Commun (Camb). 2007; 44:4565–4579.
3. Komor AC, Barton JK. J Am Chem Soc. 2014; 136:14160–14172. [PubMed: 25254630]
4. Jackson BA, Alekseyev VY, Barton JK. Biochemistry. 1999; 38:4655–4662. [PubMed: 10200152]
5. Zeglis BM, Pierre VC, Kaiser JT, Barton JK. Biochemistry. 2009; 48:4247–4253. [PubMed:
19374348]
Nano et al.
Page 4
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
6. Hart JR, Golebov O, Ernst RJ, Kirsch IR, Barton JK. Proc Natl Acad Sci U S A. 2006; 103:15359–
15363. [PubMed: 17030786]
7. (a) Boynton AN, Marcélis L, Barton JK. J Am Chem Soc. 2016; 138:5020–5023. [PubMed:
27068529] (b) Boynton AN, Marcélis L, McConnell AJ, Barton JK. Inorg Chem. 2017; 56:8381–
8389. [PubMed: 28657712]
8. Fung SK, Zou T, Cao B, Chen T, To WP, Yang C, Lok CN, Che CM. Nat Commun. 2016; 7:1–9.
9. (a) Sato Y, Honjo A, Ishikawa D, Nishizawa S, Teramae N. Chem Commun. 2011; 47:5885–5887.
(b) Arambula JF, Ramisetty SR, Baranger AM, Zimmerman SC. Proc Natl Acad Sci USA. 2009;
106:16068–16073. [PubMed: 19805260]
10. Cyanine’s emissive properties strongly depend on the photo-induced
cis
-
trans
isomerization of the
polymethine chain, which can be inhibited when cyanines are located in a rigid environment. See
ref.: Levitus M, Ranjit S. Q Rev of Biophys. 2011; 44:123–151. [PubMed: 21108866]
11. Komor AC, Schneider CJ, Weidman AG, Barton JK. J Am Chem Soc. 2012; 123:19223–19233.
12. Messina MS, Axtell JC, Wang Y, Chong P, Wixtrom AI, Kirlikovali KO, Upton BM, Hunter BM,
Shafaat OS, Khan SI, Winkler JR, Gray HB, Alexandrova AN, Maynard HD, Spokoyny AM. J Am
Chem Soc. 2016; 138:6952–6955. [PubMed: 27186856]
13. Static quenching of the cyanine by the Rh complex is evident for the conjugate free in solution.
14. Harvey BJ, Levitus M. J Fluoresc. 2009; 19:443–448. [PubMed: 18972191]
15. Junicke H, Hart JR, Kisko J, Glebov O, Kirsch IR, Barton JK. Proc Natl Acad Sci U S A. 2003;
100:3737–3742. [PubMed: 12610209]
16. (a) Iyer RR, Pluciennik A, Burdett V, Modrich PI. Chem Rev. 2006; 106:302–323. [PubMed:
16464007] (b) Tomlinson I, Sasieni P, Bodmer W. Am J Path. 2002; 160:755–758. [PubMed:
11891172] (c) Koi M, Umar A, Chauhan DP, CherianSPCarethers JM, Kunkel TA, Boland CR.
Cancer Res. 1994; 54:4308–4312. [PubMed: 8044777]
17. The fluorescence titrations with gDNA in conjunction with the data obtained from the fluorescence
titrations with 27-
mer
dsDNA provide a lower estimate of the number of mismatches for MMR-
deficient vs. MMR-proficient cells. Since the 27-
mer
DNA fluorescence titrations were performed
with dsDNA containing only CC mismatches, we assume comparable or lower binding affinities to
other (more thermodynamically stable) mismatches. Based on the fluorescence titrations with
genomic samples, we therefore estimate about 500-fold increase in mismatches in cells that are
MMR-deficient compared to the MMR-proficient analogue (SI, Section 8).
18. We also note that related metalloinsertors have been shown to bind abasic sites and single base
bulges, which may be relevant to insertions and deletions that are similarly repaired by MMR
machinery. See ref.: Zeglis BM, Boland JA, Barton JK. Biochemistry. 2009; 48:839–849.
[PubMed: 19146409]
19. Glaab WE, Tindall KR. Carcinogenesis. 1997; 18:1–8. [PubMed: 9054582]
Nano et al.
Page 5
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 1.
a) Chemical structure of
RhCy3
and its spectroscopic references. b) Illustration of the
interaction between the conjugate and DNA (well matched or mismatched).
Nano et al.
Page 6
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 2.
Steady-state emission spectra of
RhCy3
(1 μM) free in solution (black), or in the presence
of 1 μM DNA duplex oligomer: WM DNA (blue) or MM DNA (red). The DNA duplex used
was a 27-mer oligonucleotide with complement: 5
-GAC CAG CTT ATC A
C
C CCT AGA
TAA GCG-3
where the MM strand contains a (
C
) at the mismatched site versus (
G
). In
Tris buffer at 25 °C,
λ
Ex
= 520 nm.
Nano et al.
Page 7
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 3.
Fluorescence titrations with increasing amount of WM DNA (blue) and MM DNA (red).
[
RhCy3
] = 1 μM, in Tris at 25 °C.
λ
Ex
= 520 nm. The 27-mer DNA sequence: 5
-GAC
CAG CTT ATC A
C
C CCT AGA TAA GCG-3
where the MM strand comprises a (
C
) at the
mismatched site versus (
G)
for WM. Error bars calculated over three replicates. [DNA] is
per 27-mer sequence.
Nano et al.
Page 8
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 4.
Autoradiogram of a 20% polyacrylamide gel indicating photocleavage with MM DNA (lane
4). The dsDNA was a 27-mer 5
-
32
P-GAC CAG CTT ATC A
C
C CCT AGA TAA GCG-3
where the MM strand comprises a (
C
) at the mismatched site versus (
G)
for the WM. The
sample concentrations were 10 μM in 20 mM NaCl, 10 mM NaPi, pH 7.1. Irradiations were
carried out with a solar simulator (see SI). The arrow indicates the photocleaved DNA
fragment nearby the mismatched site.
Nano et al.
Page 9
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 5.
RhCy3
fluorescence titrations with increasing amounts of gDNA extracted from HCT116N,
HCT116O, SKOV3 and DU145 cell lines. The fluorescence was measured in Tris buffer at
25 °C.
λ
Ex
= 520 nm. [
RhCy3
] = 1μM. Error bars were calculated over two replicates.
[gDNA] is per base pairs.
Nano et al.
Page 10
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Scheme 1.
Synthetic strategy for
RhCy3
. i) HBTU, DIPEA, DMF 0°C, 1.5–2 hrs., 86% yield. ii)
CH
3
CN/H
2
O/EtOH, 95 °C, overnight, then anion exchange with MgCl
2
, 40% yield.
Nano et al.
Page 11
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Nano et al.
Page 12
Table 1
Excited state lifetimes (
τ
) of RhCy3 with WM and MM DNA (1:1 ratio) or without DNA.
Compound
τ
1,
ps
a
τ
2
, ps
RhCy3
No DNA
111
-
WM DNA
116
1,340
(96%)
b
(4%)
MM DNA
91
308
(85%)
b
(15%)
a
Errors ± 5%. Measurements were carried out in Tris buffer (aerated) using 1 μM Rh and oligomers.
b
In the presence of WM or MM DNA, two lifetimes were obtained resulting from a biexponential decay function. Percentages reflect the relative
contributions of each lifetime component to the overall decay.
J Am Chem Soc
. Author manuscript; available in PMC 2018 April 10.