A GGCT fluorogenic probe: Design, synthesis and application to cancer-related cells

Article abstract of DOI:10.1039/c5ob00086f

Cancer-related γ-glutamyl cyclotransferase (GGCT) specifically converts γ-glutamyl amino acids (γ-Glu-Xaa) into pyroglutamate and the corresponding amino acids (Xaa). Here we report a novel GGCT fluorogenic probe LISA-101 containing a masked O-acylated fluorophore resorufin on the side chain of the P′₁ amino acid (Xaa). Upon GGCT treatment, the P′₁ amino acid was liberated and spontaneously released the intact fluorophore. Thus, the fluorescence was regained. LISA-101 will expand the strategies for cancer studies.

Full text of DOI:10.1039/c5ob00086f

Organic &
Biomolecular Chemistry
View Article Online
View Journal
COMMUNICATION
A GGCT fluorogenic probe: design, synthesis and
application to cancer-related cells†
Cite this: DOI: 10.1039/c5ob00086f
Taku Yoshiya,*a Hiromi Ii,b Shugo Tsuda,a Susumu Kageyama,c Tatsuhiro Yoshikib
Received 15th January 2015,
Accepted 28th January 2015
and Yuji Nishiuchi*a,d
DOI: 10.1039/c5ob00086f
www.rsc.org/obc
Cancer-related γ-glutamyl cyclotransferase (GGCT) specifically
converts γ-glutamyl amino acids (γ-Glu-Xaa) into pyroglutamate
and the corresponding amino acids (Xaa). Here we report a novel
GGCT fluorogenic probe “LISA-101” containing a masked O-acy-
lated fluorophore “resorufin” on the side chain of the P’ amino
1
acid (Xaa). Upon GGCT treatment, the P’ amino acid was liberated
1
and spontaneously released the intact fluorophore. Thus, the
fluorescence was regained. LISA-101 will expand the strategies for
cancer studies.
Introduction
We have identified a human chromosome 7 ORF 24 (C7orf24)
which is a tumor-related protein.1 C7orf24 is involved in the
regulation of the glutathione homeostasis cycle as a γ-glutamyl
cyclotransferase (GGCT).2 GGCT specifically converts γ-gluta-
myl amino acids (γ-Glu-Xaa) into pyroglutamate (pyroGlu) and
the corresponding amino acids (Xaa) as shown in Scheme 1A.
In contrast, GGCT does not recognize γ-glutamyl peptides.
Although its definite role remains unclear, GGCT has been
found to be over-expressed in a range of cancers3 and silencing
of the gene by small interfering RNA (siRNA) showed an anti-
proliferative effect on cancer cell lines.1,4 Thus, it is believed
that GGCT accelerates cancer progression and its inhibitor
should function as an anti-cancer drug. Moreover, because the
glutathione cycle is a critical system for life, GGCT has recently
attracted the attention of researchers in areas of not only
cancer5 but also neurogenesis6 and botany.7 In particular, it is
interesting that systemic administration of siRNA of GGCT
could retard the tumor growth and induce necrosis of the
aPeptide Institute, Inc., Ibaraki, Osaka 567-0085, Japan.
E-mail: t.yoshiya@peptide.co.jp, yuji@peptide.co.jp
bDepartment of Clinical Oncology, Kyoto Pharmaceutical University, Kyoto,
Scheme 1 (A) Model scheme of the GGCT reaction. (B) Structures of
LISA compounds. (C) Mechanism of fluorogenic GGCT probes, illus-
trated using LISA-101.
Kyoto 607-8412, Japan
cDepartment of Urology, Shiga University of Medical Science, Otsu, Shiga 520-2192,
Japan
dGraduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5ob00086f
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
tumor tissue while showing no obvious toxicity to normal
tissues. Thus, GGCT is
treatments.
a promising target of cancer
To gain further understanding of GGCT, we designed the
first-in-class fluorogenic probe “LISA-4” to realize the concept
of the “CHELA (CHemical and Enzymatic Ligand Activation)
system” in 2013 (Scheme 1B).8 Prior to this, the singular sub-
strate preference of the enzyme had long hampered its chemi-
cal probe development. Namely, the presence of an α-carboxyl
group at the P′ site is essential for the enzyme–substrate inter-
1
Scheme 2 Synthesis of LISA-101. LISA-102 was synthesized in a similar
action. Thus, none of the conventional chromogenic/fluoro-
genic probes including γ-Glu-pNA could be processed by
GGCT. Our strategy to develop a fluorogenic probe is as
follows: (a) a canonical substrate-mimicking non-fluorescence
dipeptide “γ-Glu-Xaa”, where Xaa possesses a masked O-acy-
lated fluorophore on its side chain, is used; (b) GGCT-
mediated release of Xaa from the dipeptide probe causes its
intramolecular chemical cyclization reaction to liberate the
inherent fluorophore upon breaking the O-acyl structure
linking with the fluorophore; (c) consequently, the fluo-
rescence can be regained (Scheme 1C). For example, the
fluorogenic probe “LISA-4” [γ-Glu-Ser(CO-methylumbellifer-
one)] was effectively processed by GGCT, and the subsequent
chemical cyclization reaction immediately liberated the intact
manner.
fluorophore “methylumbelliferone”. As
a result, LISA-4
enabled visualization of GGCT activity for the first time. The
carbonate bond in this molecule, however, was not sufficiently
stable under neutral conditions, and this instability hampered
its application to cell-based assays. Here we report the design
and synthesis of a stable GGCT fluorogenic probe, LISA-101,
containing an N-ethyl urethane bond. This allows the probe to
be successfully applied to cell-based assays.
Fig. 1 Fluorescence spectra of resorufin, LISA-101 and LISA-102 (exci-
tation at 530 nm).
nant GGCT in the buffer (pH 8) was monitored by RP-HPLC
and fluorometry. While both probes were processed by GGCT
to liberate the fluorophore, LISA-101 was found to be a better
substrate for GGCT because the fluorophore could be released
faster from LISA-101 than from LISA-102 (Fig. 2). This differ-
ence might be attributed to the bulkiness of the N-alkyl struc-
ture. Additionally, non-enzymatic release of resorufin was not
observed from LISA-101 under this condition. In view of this,
we employed LISA-101 for additional examinations.
In the enzyme reaction with LISA-101, there was a time lag
between the decrease of the probe and the increase of resoru-
fin in the RP-HPLC detection. Thus, fluorescence intensity
measurement with a fluorometer could not be used to inter-
pret the enzyme reaction at that moment. This was due to the
slow conversion of the non-fluorescence intermediate 3 to the
intact fluorophore 5 in the assay medium. In fact, a significant
amount of intermediate 3 was detected in the course of
enzyme digestion by HPLC. This hampered direct monitoring
of the enzyme reaction with the guidance of fluorescence spec-
troscopy because the enzyme digest, i.e. intermediate 3, has no
fluorescence due to its O-acylated linkage. To quantify the
enzyme activity precisely by measuring fluorescence intensity,
intermediate 3 has to be completely converted to the cyclic
compound 4 and resorufin 5 via the cyclization reaction. In
general, cessation of the enzymatic reaction is achieved by
temperature elevation, pH change or denaturant addition.
Results and discussion
We designed a stable GGCT probe by modifying the structure
of LISA-4. The connection of the fluorophore with the side
chain of Xaa was replaced by an N-alkyl urethane structure,
which is known to be more stable than the carbonate bond
used in LISA-4. On the assumption of applying the probe to
cell-based imaging, the fluorophore was changed to resorufin
(emission λmax: 590 nm) which has a longer fluorescence wave-
length than the methylumbelliferone (emission λmax: 445 nm)
used in LISA-4. Consequently, we designed two probes,
LISA-101 [γ-Glu-A2bu(Nγ-Et-Nγ-CO-resorufin)] and LISA-102
[γ-Glu-A2bu(Nγ-iPr-Nγ-CO-resorufin)]. Both LISA-101 and
LISA-102 were readily synthesized (Scheme 2). As expected,
both compounds were sufficiently stable under neutral con-
ditions. Decomposition of LISA-101/102 in pH 8 buffers9 was
under the detection limit (<0.1 mol%) over 24 h at room tempe-
rature. We also confirmed that LISA-101/102 has virtually no
fluorescence compared with the strong fluorescence of resoru-
fin itself (Fig. 1). That is, the fluorescence intensity of resorufin
was more than hundredfold that of LISA-101/102 (excitation at
530 nm). The digestion progress of these probes with recombi-
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Communication
Fig. 2 LISA compounds were reacted with GGCT in pH 8 tris buffer at room temperature, and the reaction was monitored by HPLC [(A) LISA-101
and (B) LISA-102]. Resorufin was not released in the absence of GGCT (Fig. S1†).
temperature. The conversion rate was not affected even when
DMSO was added to the buffer. Raising the temperature of the
buffer containing 50% DMSO to 37 °C accelerated the conver-
sion rate, leading to completion of the conversion within 1 h.
On the other hand, we confirmed that GGCT was deactivated
and LISA-101 itself remained unchanged in this buffer con-
taining 50% DMSO (Fig. S2†). From these observations, we
concluded that the addition of DMSO to the GGCT reaction
mixture stops only the enzyme reaction, and after an
additional 1 h, GGCT activity can be quantified from the fluo-
rescence intensity.
Scheme 3 GGCT reaction of LISA-101 from the viewpoint of fluore-
scence. An intermediate amino acid is a product of GGCT treatment, but
does not possess fluorescence. Thus, the intermediate amino acid
should be converted to intact resorufin after the addition of DMSO,
According to the protocols thus established, we applied
LISA-101 for further GGCT assays (Fig. 3A). First, we validated
that GGCT activity can be quantitatively monitored using a
fluorometer. After LISA-101 had been treated with GGCT at
37 °C for 30 min, the addition of DMSO to stop the enzyme
activity uninterruptedly promoted the chemical cyclization
reaction to increase the fluorescence. An hour later at 37 °C,
the fluorescence intensity reached a plateau, suggesting that
the cyclization reaction was completed. By measuring the fluo-
rescence at this point, we could quantify GGCT activity. The
fluorescence intensities observed at these points were in pro-
under which condition the enzyme reaction has been stopped.
Taking into account the stability of the undigested probe 1
and the conversion propensity of intermediate 3 in the media
after terminating the enzyme activity, we tried adding DMSO
to the reaction mixture to stop the enzymatic reaction but not
prevent the subsequent chemical reaction (Scheme 3). We
examined the conversion efficiency of intermediate 3, which
had been synthesized separately, under neutral conditions
with or without an organic solvent. The results are summar-
ized in Table 1. In the buffer at pH 8, regardless of its salt,
intermediate 3 spontaneously cyclized and simultaneously
released resorufin with half-lives of around 30 min at room
Table 1 Half-lives of the cyclization reaction of the intermediate amino
acid 3 to afford intact fluorophore 5a
Buffer
Hepes
Tris
Phosphate
Tris
Trisb
Fig. 3 GGCT reaction in pH 8 tris buffer at 37 °C was monitored by
fluorometry with LISA-101. (A) The enzyme reaction was stopped at
30 min by the addition of DMSO, and the conversion of the intermediate
amino acid was completed by ca. 60 min (fluorescence intensity was
raised by the addition of DMSO). (B) Using the fluorescence intensity at
DMSO (%)
Half-life (min)
0
33
0
24
0
27
50
25
50
8.3
a The intermediate amino acid
synthesized and dissolved in pH 8.0 buffered solutions at room
temperature. Cyclization reaction (i.e. release of resorufin) was
monitored using HPLC. b Reaction at 37 °C.
3 derived from LISA-101 was
90 min, GGCT could be quantified. Dashed line is a regression line (R2
>
0.99).
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
cancer studies including those for GGCT inhibitor develop-
ment. Analyses of GGCT activity in plasma from patients with
several kinds of cancers are now in progress.
Acknowledgements
This work was supported in part by the Promotion Plan for the
Platform of Human Resource Development for Cancer, the
Japan Society for the Promotion of Science, and the Kyoto
Pharmaceutical University Fund for the Promotion of Scientific
Research.
Fig. 4 LISA-101 was applied to the cell-based assay in pH 8 tris buffer
at 37 °C with lysates of NHDF cells, which stably expresses GGCT or
EGFP. EGFP-expressing cell lysate was prepared as a negative control.
Notes and references
portion to the respective amounts of GGCT used (Fig. 3B).
Next, we confirmed the applicability of LISA-101 for cell-based
assays using the NHDF cell line, which is known to natively
express GGCT in a small quantity (Fig. 4). We compared two
transgenic cell lines: GGCT-expressing line and EGFP-expres-
sing line (negative control). The fluorescence increased when
the lysate of GGCT-expressing NHDF cells was applied, while a
limited rise was detected when the negative control cells were
applied. These results suggested that LISA-101 is a specific
probe for GGCT applicable to cell-based assays.
Finally, we compared the GGCT activities of MCF-7 and
NHDF cell lines. MCF-7 was adopted as a representation of
human cancer cell, and NHDF was adopted as a representation
of human normal cell. GGCT activities of cell lysates were com-
pared using LISA-101, and we confirmed that MCF-7 showed
100-times higher activity than NHDF. This result supports that
knockdown of GGCT has an efficacious effect against a drug-
resistant MCF-7 breast cancer cell without affecting normal
tissues.5b
1 S. Kageyama, H. Iwaki, H. Inoue, T. Isono, T. Yuasa,
M. Nogawa, T. Maekawa, M. Ueda, Y. Kajita, O. Ogawa,
J. Toguchida and T. Yoshiki, Proteomics: Clin. Appl., 2007, 1,
192–199.
2 (a) M. Orlowsli, P. G. Richman and A. Meister, Biochemistry,
1969, 8, 1048–1055; (b) P. G. Board, J. E. Smith and
K. Moore, J. Lab. Clin. Med., 1978, 91, 127–131;
(c) A. J. Oakley, T. Yamada, D. Liu, M. Coggan, A. G. Clark
and P. G. Board, J. Biol. Chem., 2008, 283, 22031–22042.
3 P. Gromov, I. Gromova, E. Friis, V. Timmermans-Wielenga,
F. Rank, R. Simon, G. Sauter and J. M. A. Moreira, J. Pro-
teome Res., 2010, 9, 3941–3953.
4 S. Hama, M. Arata, I. Nakamura, T. Kasetani, S. Itakura,
H. Tsuchiya, T. Yoshiki and K. Kogure, Cancer Gene Ther.,
2012, 19, 553–557.
5 (a) K. Takemura, H. Kawachi, Y. Eishi, K. Kitagaki, M. Negi,
M. Kobayashi, K. Uchida, J. Inoue, J. Inazawa, T. Kawano
and P. G. Board, Hum. Pathol., 2014, 45, 331–341; (b) R. Ran,
Y. Liu, H. Gao, Q. Kuang, Q. Zhang, J. Tang, H. Fu, Z. Zhang
and Q. He, J. Pharm. Sci., 2015, 104, 476–484.
6 Z. Chi, S. T. Byrne, A. Dolinko, M. M. Harraz, M. S. Kim,
G. Umanah, J. Zhong, R. Chen, J. Zhang, J. Xu, L. Chen,
A. Pandey, T. M. Dawson and V. L. Dawson, Cell Rep., 2014,
7, 681–688.
Conclusions
We designed and synthesized LISA-101 as a stable GGCT
fluorogenic probe. LISA-101 could be readily prepared and did
not possess fluorescence. When LISA-101 was treated with 7 B. Paulose, S. Chhikara, J. Coomey, H.-i. Jung,
GGCT, resorufin was released, and the fluorescence was
regained. Using the DMSO-stopped method, GGCT activity can
O. Vatamaniuk and O. P. Dhankher, Plant Cell, 2013, 25,
4580–4895.
be quantified by fluorescence intensity. Moreover, LISA-101 8 T. Yoshiya, S. Tsuda, M. Mochizuki, K. Hidaka, Y. Tsuda,
could be applied to the cell-based assay using a plate reader.
Much more GGCT activity could be detected in the cancer-
Y. Kiso, S. Kageyama, H. Ii, T. Yoshiki and Y. Nishiuchi,
ChemBioChem, 2013, 14, 2110–2113.
related cell compared with a normal cell line. These results 9 In this paper, all assays were conducted at pH 8, which is in
suggested that LISA-101 should be useful for GGCT-related
the optimum pH range of GGCT.2b
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015