Synthesis of a Fluorescent Analogue of Paclitaxel That Selectively Binds Microtubules and Sensitively Detects Efflux by P-Glycoprotein

Article abstract of DOI:10.1002/anie.201703298

The anticancer drug paclitaxel (Taxol) exhibits paradoxical and poorly understood effects against slow-growing tumors. To investigate its biological activity, fluorophores such as Oregon Green have been linked to this drug. However, this modification incr

Full text of DOI:10.1002/anie.201703298

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201703298
German Edition: DOI: 10.1002/ange.201703298
Fluorescent Probes
Synthesis of a Fluorescent Analogue of Paclitaxel that Selectively Binds
Microtubules and Sensitively Detects Efflux by P-Glycoprotein
+
+
Molly M. Lee , Zhe Gao , and Blake R. Peterson*
th
Dedicated to Professor FranÅois Diederich on the occasion of his 65 birthday
Abstract: The anticancer drug paclitaxel (Taxol) exhibits
paradoxical and poorly understood effects against slow-grow-
ing tumors. To investigate its biological activity, fluorophores
such as Oregon Green have been linked to this drug. However,
this modification increases its polarity by approximately 1000-
fold and reduces the toxicity of Taxol towards cancer cell lines
by over 200-fold. To construct more drug-like fluorescent
probes suitable for imaging by confocal microscopy and
analysis by flow cytometry, we synthesized derivatives of Taxol
linked to the drug-like fluorophore Pacific Blue (PB). We
found that PB-Gly-Taxol bound the target protein b-tubulin
with both high affinity in vitro and high specificity in living
cells, exhibited substantial cytotoxicity towards HeLa cells, and
was a highly sensitive substrate of the multidrug resistance
transporter P-glycoprotein (P-gp).
P
aclitaxel (Taxol, Figure 1), a natural product first isolated
from the bark of the yew tree Taxus brevifolia, is one of the
Figure 1. Structures of paclitaxel (Taxol), the commercially available
green fluorescent taxoid Flutax-2, and Pacific Blue derivatives (1–3).
most effective treatments for ovarian, breast, and lung
[
1]
cancers. By binding the protein b-tubulin, this drug stabil-
izes microtubules in the cytosol, inhibits microtubule poly-
merization dynamics, and triggers cellular death through
position through a b-Ala ester to the fluorophore Oregon
Green (OG). This b-Ala linker was first used by Nicolaou and
[1]
[8]
apoptosis. Although Taxol is widely used as a therapeutic,
the mechanisms underlying its selective cytotoxicity towards
cancer cells remain incompletely understood. Taxol and
related compounds exhibit antimitotic activity against rapidly
dividing cancer cell lines and xenografts that double every 1–
co-workers to prepare some of the earliest fluorescent
[
14]
taxoids. Esters derived from l-Ala
ents
and other substitu-
[
5,10]
have also been used to link fluorophores to Taxol at
the 7-position. The term Flutax-2 has also been used to
describe a related probe where Taxol is linked to OG via an l-
[
2]
[7]
1
2 days, thus suggesting that effects against rapidly prolif-
Ala ester [here termed Flutax-2 (l-Ala)]. The side chain of
[
16,17]
erating cells might provide selectivity for tumors in human
patients. However, these drugs also show major effects against
slow-growing tumors (median doubling time of ca. 147 days)
Taxol has also been linked to the BODIPY fluorophore,
but these probes are generally not considered suitable for
imaging in living cells.
[
2]
in patients, while sparing rapidly proliferating normal cells
in the bone marrow, gut, and other tissues. This surprising
selectivity has been termed the “proliferation-rate para-
At physiological pH, OG exists as a highly polar dianion
[
18]
(Figure 1). In Flutax-2, this increases the polarity of Taxol
by almost 1000-fold, altering its calculated octanol–water
distribution/partition coefficient from cLogDpH7.4 = 2.9
(Taxol) to cLogDpH7.4 = 0.0 (Flutax-2). Given that most
small-molecule drugs are moderately hydrophobic
(cLogDpH7.4 ꢀ 2), which facilitates passive diffusion across
membranes, Flutax-2 is not very drug-like. To provide more
drug-like probes, we report here studies of Taxol linked to the
[
3]
dox”.
To probe the anticancer effects of Taxol, fluorescent
[4–15]
analogues
have been reported. The most extensively
studied derivative, a commercially available compound
termed Flutax-2 (Figure 1), comprises Taxol linked at the 7-
[
19]
fluorophore Pacific Blue (PB, Figure 1), which is substan-
tially smaller and less polar than OG (cLogDpH7.4 = 2.0 for
PB-Gly-Taxol (1); Figure 1).
[
+]
[+]
[
*] Dr. M. M. Lee, Z. Gao, Prof. Dr. B. R. Peterson
Department of Medicinal Chemistry, The University of Kansas
Lawrence, KS 66045 (USA)
To examine whether modification of Taxol with PB might
more effectively recapitulate its biological activities, we
E-mail: brpeters@ku.edu
+
[
] These authors contributed equally to this work.
[
20]
synthesized probes 1–3 (Figure 1) via the known
TBS-
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
https://doi.org/10.1002/anie.201703298.
protected derivative 4 (Scheme 1). PB was chosen because,
unlike most other coumarins, its photophysical and chemical
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
[7]
and 265 Æ 54 nm for PB-GABA-Taxol (3) in GAB buffer.
Addition of excess paclitaxel (10 mm) blocked this fluores-
cence enhancement, thus demonstrating that 1–3 bind b-
tubulin at the same site as paclitaxel (data not shown).
To attempt to rationalize differences in the affinities of 1–
[
25]
3
, these compounds were docked, using Autodock vina, to
[
26]
the Taxol-binding site from a recent cryo-EM structure of
b-tubulin bound to Taxol. These modeling studies suggested
that linked fluorophores may favorably insert into a pocket
near the Taxol-binding site (Figure 3). Although docking did
Scheme 1. Synthesis of PB-Taxols 1–3.
À1
À1
properties (lEx = 400 nm, lEm = 447 nm, e = 29500m cm ,
Figure 3. Comparison of a published structure (PDB ID: 3J6G) of Taxol
bound to b-tubulin (A) with a model (B) of b-tubulin docked to 1. In
(B), a docking pose of 1 is shown where the side-chain orientations
are similar to those of Taxol in panel (A), and the PB moiety of
1 engages a neighboring pocket on the protein surface.
[
19]
QY= 0.75, phenol pK = 3.7)
allow it to be efficiently
excited at 405 nm with violet lasers found on many confocal
microscopes and flow cytometers. Additionally, PB can be
readily synthesized in gram quantities. Because taxoids
modified at the 7-position can retain high affinity for micro-
a
[21]
[
22]
tubules, PB was linked at this position via amino acids that
differ subtly in the number of methylenes between the amine
and the carbonyl. Optical spectroscopy confirmed that the
fluorescence properties of 1–3 are similar to PB, whereas
Flutax-2 is similar to OG (Figure S1 in the Supporting
Information).
not readily correlate the higher affinity of 1 compared to 2
and 3 with greater or lesser engagement of this pocket by the
fluorophore, the affinities determined experimentally suggest
that conformational restriction of the longer and more
flexible linkers of 2 and 3 might reduce the affinity upon
binding of PB to the same pocket.
Flutax-2 (l-Ala) has been reported to bind human b-
tubulin with an apparent K = 14 nm, as quantified by changes
The cytotoxicity of fluorescent taxoids towards HeLa
cervical carcinoma cells (Figure 4) and multidrug-resistant
HCT-15 colorectal adenocarcinoma cells (Figure S3) was
evaluated. These cell lines were treated for 48 h, cellular
d
[
23]
in fluorescence anisotropy. BODIPY-modified paclitaxel is
known to exhibit enhanced fluorescence upon binding to this
[
24]
protein, and we measured the affinities of 1–3 for cross-
[
7]
linked microtubules from bovine brain using a fluorescence-
enhancement method (Figure 2). Curve fitting of the equi-
librium binding curves yielded apparent K values of 34 Æ
d
6
nm for PB-Gly-Taxol (1), 63 Æ 8 nm for PB-b-Ala-Taxol (2),
Figure 2. Quantification of the affinities (K ) of 1–3 (25 nm) for cross-
d
linked microtubules by enhancement of fluorescence. Binding studies
were conducted in aqueous GAB buffer (pH 6.5). PB was excited at
Figure 4. Analysis of cytotoxicity. HeLa cells were treated with com-
pounds in the absence (A) or presence (B) of verapamil (25 mm) for
48 h and cellular viability measured by flow cytometry.
4
05 nm and emission was measured at and above 450 nm.
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!

Angewandte
Communications
Chemie
viability was analyzed by flow cytometry, and Taxol was used
as a positive control. HCT-15 cells overexpress P-glycoprotein
P-gp, MDR1, ABCB1), and because Taxol is a substrate of
(
[
27]
this efflux transporter, we further measured cytotoxicity in
the presence of the P-gp inhibitor verapamil
[
28]
(Figure 4,
panel B, and Figure S3). Among the fluorescent probes, 1 was
uniquely toxic, with IC = 120 nm in the absence of and IC =
5
0
50
6
0 nm in the presence of verapamil (25 mm) in HeLa cells. In
HCT-15 cells, verapamil (25 mm) enhanced the toxicity of 1 by
1-fold (from IC = 3.7 mm to IC = 90 nm). Control experi-
4
5
0
50
ments confirmed that verapamil itself (IC > 75 mm at 48 h,
5
0
Figure S3) did not contribute to these cytotoxic effects. HeLa
[29]
cells express low levels of P-gp compared to HCT-15 cells,
and this enhanced cytotoxicity mediated by verapamil
suggested that 1–3 might be highly efficient substrates of
this drug efflux transporter. Flutax-2 is known to be a sub-
[
30]
strate of P-gp, but in HeLa cells treated with verapamil,
Flutax-2 was the least cytotoxic (IC = 1310 nm), likely due to
5
0
its higher polarity and associated lower cellular permeability.
Flutax-2 (l-Ala) has been previously reported to exhibit
cytotoxic IC50 values of 800 nm against drug-sensitive A2780
cells and more than 20 mm against drug-resistant A2780AD
[
9]
cells. Because 1 appeared to most closely mimic Taxol, it
became the focus of further studies.
To examine the subcellular distribution of 1, we imaged
HeLa cells by confocal laser scanning microscopy (Figure 5).
After treating cells with 1 (1 mm, 1 h), co-treatment with
verapamil dramatically enhanced cellular fluorescence in
a dose-dependent manner (Figures 5 and S2), revealing
intricate networks of microtubules in living cells. The distinct
spectral profiles of PB and OG were used to further examine
the colocalization of 1 and Flutax-2 (Figures 5C). Whereas
1
bound microtubules with very high specificity, Flutax-2
additionally conferred punctate fluorescence that did not
colocalize with 1. This lower specificity of Flutax-2 is
consistent with a prior report of its off-target accumulation
in the Golgi apparatus.
Figure 5. A–C) Confocal laser scanning and differential interference
[10]
contrast (DIC) microscopy of HeLa cells treated with 1 (1 mm, 1 h).
A,B) Images of cells in the absence (A) and in the presence (B) of
verapamil (100 mm). In the inset of (A), the brightness was enhanced
to reveal the weak fluorescence of 1 resulting from efflux. C) Cells were
further treated with verapamil (25 mm) and Flutax-2 (5 mm) to examine
specificity. Arrows indicate punctate fluorescence of Flutax-2. Scale
bar: 25 mm.
Overexpression of P-gp frequently confers resistance to
[
31]
the antiproliferative effects of Taxol. To further investigate
whether 1 is a substrate of P-gp, we transiently transfected
[
32]
PC-3 cells, which lack this transporter,
with a plasmid
(
pHaMDR-EGFP) encoding P-gp fused to enhanced green
fluorescent protein (EGFP). Unlike HeLa cells, imaging of
PC-3 cells treated with 1 (1 mm) revealed strong blue
fluorescence in the absence of verapamil (Figure S4). How-
ever, in cells expressing green fluorescent P-gp-EGFP,
a decrease in blue fluorescence was observed that was
dependent on the level of P-gp-EGFP expression (Figure S4),
thus indicating that 1 is a potent substrate of this transporter.
Treatment with verapamil reversed this effect by blocking P-
gp to prevent the efflux of 1.
presence of verapamil (at 25 or 100 mm), the fluorescence of
1 increased by 7-fold at 25 mm and 15-fold at 100 mm
verapamil. Moreover, in HeLa cells, 1 could detect low
levels of P-gp activity that were undetectable by Rho123
(Figures 6D and Figure S5), enhancing the fluorescence by
10-fold at 25 mm verapamil and 23-fold at 100 mm verapamil,
thus indicating that 1 is a uniquely sensitive sensor of this
efflux transporter.
In contrast to Flutax-2, which is only weakly cytotoxic and
exhibits low cellular permeability due to the high polarity of
the appended OG fluorophore, the more drug-like PB-Gly-
Taxol (1) substantially recapitulates the cytotoxic, tubulin-
binding and P-gp-mediated efflux activity of the parent
anticancer drug. Because PB can be efficiently detected and
analyzed by confocal microscopy and flow cytometry, this
We additionally investigated the efflux of 1 in HCT-15
[
29]
cells that express high levels of P-gp. Compound 1 was
compared with rhodamine 123 (Rho123), a common P-gp
[
33]
substrate that accumulates in mitochondria. As shown in
Figure 6, the cells were analyzed after treatment with
1
(1 mm), Rho123 (1 mm), and verapamil (0, 25, 100 mm) by
confocal microscopy and flow cytometry. Whereas the
fluorescence of Rho123 increased by only 3-fold in the
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
Keywords: antitumor agents · confocal microscopy ·
drug design · fluorescent probes · paclitaxel
[
[
1] C. C. Rohena, S. L. Mooberry, Nat. Prod. Rep. 2014, 31, 335 –
55.
2] E. Komlodi-Pasztor, D. Sackett, J. Wilkerson, T. Fojo, Nat. Rev.
Clin. Oncol. 2011, 8, 244 – 250.
3
[
[
3] T. J. Mitchison, Mol. Biol. Cell 2012, 23, 1 – 6.
4] I. Barasoain, J. F. Diaz, J. M. Andreu, Methods Cell Biol. 2010,
95, 353 – 372.
[
5] J. F. Dꢀaz, I. Barasoain, A. A. Souto, F. Amat-Guerri, J. M.
Andreu, J. Biol. Chem. 2005, 280, 3928 – 3937.
[
6] I. Barasoain, A. M. Garcꢀa-Carril, R. Matesanz, G. Maccari, C.
Trigili, M. Mori, J.-Z. Shi, W.-S. Fang, J. M. Andreu, M. Botta,
J. F. Dꢀaz, Chem. Biol. 2010, 17, 243 – 253.
[
[
[
7] J. F. Diaz, R. Strobe, Y. Engelborghs, A. A. Souto, J. M. Andreu,
J. Biol. Chem. 2000, 275, 26265 – 26276.
8] R. K. Guy, Z. A. Scott, R. D. Sloboda, K. C. Nicolaou, Chem.
Biol. 1996, 3, 1021 – 1031.
9] X. Li, I. Barasoain, R. Matesanz, J. F. Diaz, W. S. Fang, Bioorg.
Med. Chem. Lett. 2009, 19, 751 – 754.
[
[
10] S. Duchi, P. Dambruoso, E. Martella, G. Sotgiu, A. Guerrini, E.
Lucarelli, A. Pessina, V. Cocce, A. Bonomi, G. Varchi, Biocon-
jugate Chem. 2014, 25, 649 – 655.
11] G. Lukinavi cˇ ius, L. Reymond, E. DꢁEste, A. Masharina, F.
Gottfert, H. Ta, A. Gꢂther, M. Fournier, S. Rizzo, H. Waldmann,
C. Blaukopf, C. Sommer, D. W. Gerlich, H.-D. Arndt, S. W. Hell,
K. Johnsson, Nat. Methods 2014, 11, 731 – 733.
[
[
12] S. Sengupta, T. C. Boge, G. I. Georg, R. H. Himes, Biochemistry
1995, 34, 11889 – 11894.
13] Y. Han, A. G. Chaudhary, M. D. Chordia, D. L. Sackett, B.
Perez-Ramirez, D. G. Kingston, S. Bane, Biochemistry 1996, 35,
14173 – 14183.
[
14] A. A. Souto, A. U. Acuna, J. M. Andreu, I. Barasoain, M. Abal,
F. AmatGuerri, Angew. Chem. Int. Ed. Engl. 1995, 34, 2710 –
2712; Angew. Chem. 1995, 107, 2910 – 2912.
Figure 6. A,B) Confocal laser scanning and DIC microscopy of HCT-15
cells treated with the P-gp substrates rhodamine 123 (A) and 1 (B).
C,D) Analysis of HCT-15 (C) and HeLa (D) cells treated with rhod-
amine 123 (left panels) or 1 (right panels) by flow cytometry. Concen-
trations of added verapamil are shown. Scale bar: 25 mm.
[15] E. Baloglu, D. G. Kingston, P. Patel, S. K. Chatterjee, S. L. Bane,
Bioorg. Med. Chem. Lett. 2001, 11, 2249 – 2252.
[16] C. Martin, J. Walker, A. Rothnie, R. Callaghan, Br. J. Cancer
2003, 89, 1581 – 1589.
[
17] B. Gao, A. Russell, J. Beesley, X. Q. Chen, S. Healey, M.
Henderson, M. Wong, C. Emmanuel, L. Galletta, S. E. Johnatty,
D. Bowtell, M. Haber, M. Norris, P. Harnett, G. Chenevix-
Trench, R. L. Balleine, A. deFazio, Sci. Rep. 2014, 4, 4669.
18] W. C. Sun, K. R. Gee, D. H. Klaubert, R. P. Haugland, J. Org.
Chem. 1997, 62, 6469 – 6475.
probe offers a new tool for studies of the paradoxical
anticancer effects of Taxol.
[
[
Acknowledgements
19] W.-C. Sun, K. R. Gee, R. P. Haugland, Bioorg. Med. Chem. Lett.
1998, 8, 3107 – 3110.
We thank Prof. Dr. Liang Xu for the HCT-15 cells, Prof. Dr.
Matthew Levy for the PC-3 cells, and Dr. Michael Gottesman
for the pHaMDR-EGFP plasmid. B. Peterson acknowledges
support from the NIH (P20-GM103638, P20-GM103418, and
R01-CA211720), and the G. Harold and Leila Y. Mathers
Charitable Foundation. MML acknowledges support from the
NIH Dynamic Aspects of Chemical Biology Training Grant at
the University of Kansas (T32-GM08545).
[20] N. F. Magri, D. G. I. Kingston, J. Nat. Prod. 1988, 51, 298 – 306.
[
[
21] M. M. Lee, B. R. Peterson, ACS Omega 2016, 1, 1266 – 1276.
22] X. Xiao, J. Wu, C. Trigili, H. Chen, J. W. K. Chu, Y. Zhao, P. Lu,
L. Sheng, Y. Li, F. J. Sharom, I. Barasoain, J. F. Diaz, W.-s. Fang,
Bioorg. Med. Chem. Lett. 2011, 21, 4852 – 4856.
[
[
[
23] R. M. Buey, E. Calvo, I. Barasoain, O. Pineda, M. C. Edler, R.
Matesanz, G. Cerezo, C. D. Vanderwal, B. W. Day, E. J. Sor-
ensen, J. A. Lopez, J. M. Andreu, E. Hamel, J. F. Diaz, Nat.
Chem. Biol. 2007, 3, 117 – 125.
24] R. Rouzier, R. Rajan, P. Wagner, K. R. Hess, D. L. Gold, J. Stec,
M. Ayers, J. S. Ross, P. Zhang, T. A. Buchholz, H. Kuerer, M.
Green, B. Arun, G. N. Hortobagyi, W. F. Symmans, L. Pusztai,
Proc. Natl. Acad. Sci. USA 2005, 102, 8315 – 8320.
Conflict of interest
25] O. Trott, A. J. Olson, J. Comput. Chem. 2010, 31, 455 – 461.
The authors declare no conflict of interest.
[26] G. M. Alushin, G. C. Lander, E. H. Kellogg, R. Zhang, D. Baker,
E. Nogales, Cell 2014, 157, 1117 – 1129.
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!

Angewandte
Communications
Chemie
[
[
[
[
[
27] S. H. Jang, M. G. Wientjes, J. L.-S. Au, J. Pharmacol. Exp. Ther.
001, 298, 1236 – 1242.
28] T. Tsuruo, H. Iida, S. Tsukagoshi, Y. Sakurai, Cancer Res. 1981,
1, 1967 – 1972.
29] I. V. Lebedeva, P. Pande, W. F. Patton, PLoS ONE 2011, 6,
e22429.
30] G. Patwardhan, V. Gupta, J. Huang, X. Gu, Y. Y. Liu, Biochem.
Pharmacol. 2010, 80, 72 – 79.
31] P. D. W. Eckford, F. J. Sharom, Chem. Rev. 2009, 109, 2989 –
[32] J. P. van Brussel, G. J. van Steenbrugge, J. C. Romijn, F. H.
2
Schrçder, G. H. J. Mickisch, Eur. J. Cancer 1999, 35, 664 – 671.
[33] S. Forster, A. E. Thumser, S. R. Hood, N. Plant, PLoS ONE
2012, 7, e33253.
4
Manuscript received: March 29, 2017
3011.
Final Article published: && &&, &&&&
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Fluorescent Probes
The fluorophore normally linked to the
anticancer drug paclitaxel (Taxol) to
investigate its biological activity signifi-
cantly alters its properties. To construct
more drug-like fluorescent probes suita-
ble for confocal microscopy and flow
cytometry, Taxol was linked to the drug-
like fluorophore Pacific Blue (PB). PB-Gly-
Taxol bound the target protein b-tubulin
with high affinity and high specificity, and
exhibited substantial cytotoxicity towards
HeLa cells.
M. M. Lee, Z. Gao,
B. R. Peterson*
&&&& — &&&&
Synthesis of a Fluorescent Analogue of
Paclitaxel that Selectively Binds
Microtubules and Sensitively Detects
Efflux by P-Glycoprotein
6
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!