Photoelectrocyclization as an activation mechanism for organelle-specific live-cell imaging probes

Article abstract of DOI:10.1002/anie.201502403

Photoactivatable fluorophores are useful tools in live-cell imaging owing to their potential for precise spatial and temporal control. In this report, a new photoactivatable organelle-specific live-cell imaging probe based on a 6π electrocyclization/oxida

Full text of DOI:10.1002/anie.201502403

Angewandte
Chemie
International Edition: DOI: 10.1002/anie.201502403
German Edition: DOI: 10.1002/ange.201502403
Fluorescent Probes
Hot Paper
Photoelectrocyclization as an Activation Mechanism for Organelle-
Specific Live-Cell Imaging Probes**
Mai N. Tran and David M. Chenoweth*
Abstract: Photoactivatable fluorophores are useful tools in
live-cell imaging owing to their potential for precise spatial and
temporal control. In this report, a new photoactivatable
organelle-specific live-cell imaging probe based on a 6p elec-
trocyclization/oxidation mechanism is described. It is shown
that this new probe is water-soluble, non-cytotoxic, cell-
permeable, and useful for mitochondrial imaging. The probe
displays large Stokes shifts in both pre-activated and activated
forms, allowing simultaneous use with common dyes and
fluorescent proteins. Sequential single-cell activation experi-
ments in dense cellular environments demonstrate high spatial
precision and utility in single- or multi-cell labeling experi-
ments.
Herein, we present a new intracellular imaging probe that
utilizes a photoactivation mechanism based on a photochemi-
cally allowed 6p electrocyclization/oxidation reaction. This
mechanism expands the toolbox for intracellular photoacti-
vatable probes, providing an emissive pre- and post-activated
form that can be tracked by fluorescence microscopy. To
demonstrate this strategy, we introduce a water-soluble, non-
cytotoxic, and readily cell-permeable photoactivatable fluo-
rescent probe for mitochondrial-specific live-cell imaging. We
compare the cytotoxicity of this probe to a commonly used
and commercially available mitochondrial dye.[8] We also
demonstrate its utility in spatiotemporally defined live-cell
imaging of mitochondria in dense cellular populations
through sequential single-cell activation experiments.
P
hotoactivatable fluorescent probes are powerful tools for
First, we developed a rapid and efficient synthesis of the
desired photoactivatable probe starting with commercially
available 2-cyanopyridine (Supporting Information, Scheme
S1). Following a similar route to that reported in our previous
studies,[9] we elaborated 2-cyanopyridine to thioketone 1.
Homocoupling of 1 using copper powder in toluene at reflux
yielded 1,1’-diazaxanthilidene 2 as an interconverting mixture
of the E and Z isomers in 84% yield (Scheme 1). Methylation
of 2 with dimethyl sulfate afforded the desired monomethyl-
ated product (E)-3/(Z)-3 in 96% yield. The overall yield of
the probe from the commercially available starting materials
studying biological systems owing to the high spatial and
temporal control afforded by light. Photoactivatable probes
can be genetically encoded proteins[1] or small molecules.[2]
Each of these technologies comes with its own benefits and
limitations and may be thought of as complimentary to one
another depending on the application. The protein “Kaede”
represents the first discovery of a photoactivatable fluores-
cent protein, and much progress has been made in recent
years with similar genetically encoded approaches.[3] Small-
molecule photoactivatable probes have also found wide-
spread use, and they generally rely on a small number of
photoactivation mechanisms, such as photoisomerization,[4]
photouncaging,[5] photodecomposition of azides,[6] or photo-
click reactions.[7] These strategies primarily depend on the
conversion from an initial non-fluorescent state into a fluo-
rescent state.
[*] M. N. Tran, Prof. D. M. Chenoweth
Department of Chemistry, University of Pennsylvania
231 South 34th Street, Philadelphia, PA 19104 (USA)
E-mail: dcheno@sas.upenn.edu
[**] This work was supported by funding from the University of
Pennsylvania. We thank the Vietnam Education Foundation for
funding (VEF Fellowship to M.N.T.) and the NSF for support of the
X-ray diffractometer (Grant CHE-0840438). We thank Dr. George
Furst and Dr. Rakesh Kohli for assistance in obtaining the high-
resolution NMR and mass spectral data, respectively. We are
grateful to Dr. Patrick Carroll for X-ray crystallographic assistance.
We thank Tom Troxler for assistance with the TCSPC measurements
collected at the Ultrafast Optical Processes Laboratory at the
University of Pennsylvania (supported by the NIH, Grant
P41GM104605). We thank Mike Lampson and The Penn CDB
Microscopy Core facility for microscope use.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201502403.
Scheme 1. Synthesis of (E)-3/(Z)-3 and crystal structures of (E)-2 and
(Z)-3.
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
after six steps was 24% (Scheme S1, Scheme 1). Single
crystals of (E)-2 and (Z)-3 were obtained by slow evaporation
from chloroform/methanol (1:1) and water solutions, respec-
tively. Both structures adopted an anti-folded conformation,
despite the increased steric hindrance from methylation in
1
(Z)-3 (Scheme 1). Variable-temperature H NMR spectros-
copy[10] of (E)-2/(Z)-2 in deuterated N,N-dimethylformamide
confirmed the dynamic interconversion of the E and Z
isomers at room temperature, which is consistent with our
previous studies (Figure S2).[9] At low temperature, separate
sets of sharp and well-resolved signals were observed for the
E and Z isomers of 3. Incremental heating resulted in
significant signal broadening followed by sharpening and
resolved coupling at high temperature. This result indicates
rapid exchange on the NMR timescale. Coalescence was
observed at 458C, and the activation energy was determined
to be DG° = 15.7 kcalmolÀ1 (Figure S1).
Upon irradiation at 365 nm using a Rayonet photoreactor,
the (E)-3/(Z)-3 mixture was found to undergo photocycliza-
tion followed by oxidation to yield photoproduct 4 (Fig-
ure 1A). To promote the photocyclization step of (E)-3/(Z)-3,
both the anti-folded conformation and dynamic interconver-
sion are crucial. According to the Woodward–Hoffmann
rules, electrocyclization of the 1,3,5-hexatriene moiety of
(Z)-3 (4n + 2 electron system) is thermally allowed in
a disrotatory manner and photochemically allowed in a con-
rotatory manner.[11] The anti-folded conformation of (Z)-3 is
preorganized for conrotatory cyclization, satisfying the
requirement for the photochemical reaction. This conclusion
was further confirmed by TD-DFT calculations[12] on (Z)-3.
As is evident from the calculated molecular orbitals, a con-
rotatory cyclization leads to constructive interactions and is
photochemically favored (Figure 1B). The structure of (E)-3
is disfavored for electrocyclization but E–Z isomerization
leads to the formation of (Z)-3, facilitating the photoreaction.
As a result, the dihydrophenanthrene intermediate was
attained, which was followed by oxidation to give photo-
product 4. The dihydrophenanthrene intermediate was not
isolable from the reaction mixture.
Compound (E)-3/(Z)-3 is fluorescent with a large Stokes
shift (135 nm). Photoproduct 4 is red-shifted in both absorb-
ance and emission by approximately 100 nm while a large
Stokes shift of 108 nm is retained (Figure 1C). Live-cell
imaging studies of (E)-3/(Z)-3 were performed with HeLa
cells, and specific sub-cellular localization was observed,
consistent with mitochondrial uptake. An engineered HeLa
cell line that expresses GFP-labeled proteins specifically
localized to the outer mitochondrial membrane (mito-GFP
cell line, Figure S3), as introduced in one of our previous
studies,[5b] was used as a control to confirm the mitochondrial
localization of (E)-3/(Z)-3 in this report. The localization
statistics were also compared to those of a commonly used
and commercially available MitoTracker dye (Figure S4).[8]
Our control cell line consisted of a population of HeLa cells
expressing a GFP fusion protein localized specifically to the
outer mitochondrial membrane (mito-GFP cell line) in
addition to a population of non-GFP-expressing HeLa cells,
which served as an internal control for compound localization
in the absence of the GFP signal. The two channels 405/635
Figure 1. A) E–Z interconversion and photocyclization/oxidation reac-
tion of (E)-3/(Z)-3. B) Idealized and calculated LUMO of (Z)-3 (TD-
DFT B3LYP, 6-311+G(2d,p) basis set). Note the orbital preorganiza-
tion for photochemically favored conrotatory cyclization. C) Absorb-
ance and emission spectra of (E)-3/(Z)-3 [lmax =408 nm (abs.),
543 nm (em.), e=11040mÀ1 cmÀ1, FF =0.021, t=0.97 ns] and 4
[lmax =504 nm (abs.), 612 nm (em.), e=11496mÀ1 cmÀ1, FF =0.100,
t=4.77 ns] in water.
and 488/525 were used to detect (E)-3/(Z)-3 and GFP,
respectively, with no bleed-through observed. All images
were recorded at the same brightness and contrast settings.
Incubating mito-GFP cells with (E)-3/(Z)-3 allowed for
colocalization to be assessed (Figure 2A and Figure S3).
Colocalization statistics calculated over multiple frames for
a total of 80 cells showed significant overlap (Pearsonꢀs
coefficient: 0.81 Æ 0.02; Mandersꢀ coefficients: 0.98 Æ 0.01 and
1.00 Æ 0.00; Spearman correlation: 0.88 Æ 0.02). High Man-
dersꢀ coefficients point to near exclusive mitochondrial
localization of (E)-3/(Z)-3 in GFP-positive cells. Variations
in intensity between localized (E)-3/(Z)-3 and GFP resulted
in a slightly lower Pearsonꢀs coefficient. This small deviation
may be attributed to the difference between internal mito-
chondrial localization of (E)-3/(Z)-3 versus external mito-
chondrial membrane localization of GFP. Intensity profiles
across multiple cells are shown in Figure 2B. Additional
colocalization studies were carried out using commercially
available Mitotracker Red,[8] and similar results were
observed when comparing the colocalization with GFP-
labeled mitochondria (Figure S4). High signal-to-noise
2
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
photoconversion compared to a standard sample of
photoproduct 4 (Figure 2C). Cytotoxicity studies
on HeLa cells over 24, 48, and 72 hours confirmed
that (E)-3/(Z)-3 had very low cytotoxicity. For
comparison, MitoTracker Red FM, one of the
most commonly used mitochondrial imaging
probes, showed significant cytotoxicity over the
same time course at a five-fold lower concentration
(Figure 2D).
By combining the observed photoreaction (Fig-
ure 1A) with the ability to selectively label mito-
chondria for live-cell imaging (Figure 2A), intra-
cellular photoactivation of (E)-3/(Z)-3 was
attempted. As the absorbance and emission spectra
of 4 are red-shifted by approximately 100 nm
compared to those of (E)-3/(Z)-3, the individual
species can be selectively excited. Alternating use
of the two channels (i.e., 444/525 for (E)-3/(Z)-3
and 488/632 for 4) resulted in the photoconversion
of (E)-3/(Z)-3 into 4 in live cells. A significant
increase in the fluorescence intensity of 4 was
achieved after 30 seconds, using 200 total scans
(Figure 3A,B). To further investigate the spatial
selectivity, a targeted 405 nm laser (CrystaLaser,
DL405-050-O) was used to target smaller areas of
the cells. Confocal images were captured using
a 488 nm excitation laser and a 632 nm emission
filter. Immediately after the first activation, a bright
signal was detected at the targeted region. The
tubular morphology of the mitochondria was
observed, and the remaining portions of the mito-
chondria remained dim for an additional ten
seconds using the 488 nm excitation laser (Fig-
ure 3C).
Finally, we wanted to show that individual cells
can be photoactivated in extremely crowded con-
fluent cellular environments. A normalized inten-
sity plot for the sequential photoactivation of two
adjacent cells is shown in Figure 3C. A standard
405 nm laser commonly used for photobleaching
experiments was used for sequential photoactiva-
tion with excellent spatial control over individual
cells in crowded environments (Figure 3D). Using
50 pulses with a length of 20 s, the mitochondrial
probe 3 was photoconverted into 4. Two individual
cells out of 20 cells were sequentially photoacti-
vated (cell 7 and cell 19, Figure 3D).
In conclusion, we have developed a photoacti-
vatable probe that is based on a 6p electrocycliza-
tion/oxidation mechanism. The probe is water-
soluble, cell-permeable, non-cytotoxic, and can
selectively stain mitochondria. This photoactivation
strategy should be broadly applicable to many dye
classes, and further studies are underway to develop
Figure 2. A) Confocal images of mito-GFP cells (right), HeLa cells stained with
(E)-3/(Z)-3 (middle), and mito-GFP cells stained with (E)-3/(Z)-3 (left). Compound
(E)-3/(Z)-3 was observed at 405/635 nm while GFP was observed at 488/525 nm.
All images were kept at the same contrast/brightness, and bleed-through between
channels was not detected. Scale bar: 10 mm. B) Line plot of the fluorescence
intensity across multiple cells showing high colocalization and signal-to-noise
ratios for both GFP and (E)-3/(Z)-3. In the colocalization plot in the bottom left of
panel A, the nuclear regions in adjacent cells are designated with “n”. Pearson’s
colocalization coefficient: 0.87Æ0.01; Manders’ overlap coefficients: 0.98Æ0.01
and 1.00Æ0.00. C) HPLC chromatogram of (E)-3/(Z)-3 (1 mm solution in water)
after ten days under ambient light (top chromatogram) and HPLC chromatogram
of 4 (bottom chromatogram). No decomposition products or the photoproduct of
(E)-3/(Z)-3 were detected. D) The cell viability was determined using HeLa cells in
the presence of (E)-3/(Z)-3 (5 mm) or MitoTracker Red FM (1 mm).
ratios were observed in the intensity line plots for both GFP
and (E)-3/(Z)-3 (Figure 2B). The stability of a 1 mm aqueous
solution of (E)-3/(Z)-3 was assessed under ambient light
conditions at 258C for ten days. HPLC analysis of the
resulting solution revealed no evidence for decomposition or
new cellular imaging tools.
Keywords: electrocyclization · fluorescent probes ·
imaging agents · photochemistry
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
2010, 285, 11607 – 11616; c) V. V. Verkhusha, A. Sorkin, Chem.
Biol. 2005, 12, 279 – 285; d) G. H. Patterson, J. Lippincott-
Schwartz, Methods 2004, 32, 445 – 450; e) D. M. Shcherbakova,
V. V. Verkhusha, Curr. Opin. Chem. Biol. 2014, 20, 60 – 68; f) V.
Adam, R. Berardozzi, M. Byrdin, D. Bourgeois, Curr. Opin.
Chem. Biol. 2014, 20, 92 – 102.
[2] a) E. Lacivita, M. Leopoldo, F. Berardi, N. A. Colabufo, R.
Perrone, Curr. Med. Chem. 2012, 19, 4731 – 4741; b) M. Fernꢁn-
dez-Suꢁrez, A. Y. Ting, Nat. Rev. Mol. Cell Biol. 2008, 9, 929 –
943.
[3] a) R. Ando, H. Hama, M. Yamamoto-Hino, H. Mizuno, A.
Miyawaki, Proc. Natl. Acad. Sci. USA 2002, 99, 12651 – 12656;
b) M. Tomura, N. Yoshida, J. Tanaka, S. Karasawa, Y. Miwa, A.
Miyawaki, O. Kanagawa, Proc. Natl. Acad. Sci. USA 2008, 105,
10871 – 10876; c) I. Hayashi, H. Mizuno, K. I. Tong, T. Furuta, F.
Tanaka, M. Yoshimura, A. Miyawaki, M. Ikura, J. Mol. Biol.
2007, 372, 918 – 926; d) P. S. Dittrich, S. P. Schꢂfer, P. Schwille,
Biophys. J. 2005, 89, 3446 – 3455.
[4] a) S. Y. Cho, Y. K. Song, J. G. Kim, S. Y. Oh, C. M. Chung,
Tetrahedron Lett. 2009, 50, 4769 – 4772; b) N. Gagey, P. Neveu, C.
Benbrahim, B. Goetz, I. Aujard, J. B. Baudin, L. Jullien, J. Am.
Chem. Soc. 2007, 129, 9986 – 9998.
[5] a) T. Faal, P. T. Wong, S. Tang, A. Coulter, Y. Chen, C. H. Tu,
J. R. Baker, S. K. Choi, M. A. Inlay, Mol. BioSyst. 2015, 11, 783 –
790; b) E. R. Ballister, C. Aonbangkhen, A. M. Mayo, M. A.
Lampson, D. M. Chenoweth, Nat. Commun. 2014, 5, 5475; c) T.
Kobayashi, T. Komatsu, M. Kamiya, C. Campos, M. Gonzalez-
Gaitan, T. Terai, K. Hanaoka, T. Nagano, Y. Urano, J. Am.
Chem. Soc. 2012, 134, 11153 – 11160.
[6] S. J. Lord, N. R. Conley, H. L. D. Lee, R. Samuel, N. Liu, R. J.
Twieg, W. E. Moerner, J. Am. Chem. Soc. 2008, 130, 9204 – 9205.
[7] a) R. K. V. Lim, Q. Lin, Acc. Chem. Res. 2011, 44, 828 – 839;
b) Z. Yu, L. Y. Ho, Q. Lin, J. Am. Chem. Soc. 2011, 133, 11912 –
11915; c) P. An, Z. Yu, Q. Lin, Org. Lett. 2013, 15, 5496 – 5499.
[8] I. Johnson, M. T. Z. Spence, Molecular probes handbook: a guide
to fluorescent probes and labeling technologies, 11th ed., Life
Technologies Corp., 2010.
[9] R. A. Rarig, M. N. Tran, D. M. Chenoweth, J. Am. Chem. Soc.
2013, 135, 9213 – 9219.
[10] J. Sandstrçm, Dynamic NMR spectroscopy, Academic Press,
London, New York, 1982.
Figure 3. A) Confocal microscopy images of HeLa cells stained with
(E)-3/(Z)-3, observed at 444/525 nm and 488/632 nm over 30 s with
200 alternating 150 ms pulses. B) Average intensity of the 444/525 and
488/632 channels after 200 alternating pulses. The data from each of
the two channels were fitted to a one-phase exponential curve, and the
obtained rate constants and half-lives were 0.060 sÀ1 and 11.61 s for
the 444/525 channel and 0.030 sÀ1 and 23.05 s for the 488/632
channel. C) Normalized intensity plot for the sequential photoactiva-
tion of two cells. D) Selective sequential photoactivation of two cells
(7 and 19) among a field of view containing 20 confluent cells. Each
cell was activated by fifty 20 s pulses of a normal 405 nm FRAP laser.
Dashed lines indicate cell boundaries. Scale bar: 10 mm.
[11] a) E. V. Anslyn, D. A. Dougherty, Modern physical organic
chemistry, University Science Books, Sausalito, 2006; b) I.
Fleming, Molecular Orbitals and Organic Chemical Reactions,
Wiley, Chichester, 2009.
[12] Gaussian 09 (Revision B.01 ed.), M. J. T. Frisch et al., Wall-
ingford, CT, 2010.
[1] a) K. A. Lukyanov, D. M. Chudakov, S. Lukyanov, V. V. Verkhu-
sha, Nat. Rev. Mol. Cell Biol. 2005, 6, 885 – 891; b) A. Welman,
A. Serrels, V. G. Brunton, M. Ditzel, M. C. Frame, J. Biol. Chem.
Received: March 14, 2015
Published online: && &&, &&&&
4
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Fluorescent Probes
M. N. Tran,
D. M. Chenoweth*
&&&& — &&&&
Photoelectrocyclization as an Activation
Mechanism for Organelle-Specific Live-
Cell Imaging Probes
A photoactivatable organelle-specific live-
cell imaging probe based on a 6p electro-
cyclization/oxidation mechanism is de-
scribed. This probe is water-soluble, non-
cytotoxic, cell-permeable, and can be
used for mitochondrial imaging.
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!