Activity-Based Probe for Screening of High-Colibactin Producers from Clinical Samples

Article abstract of DOI:10.1021/acs.orglett.9b01345

While high-colibactin-producing Escherichia coli is thought to be associated with colorectal oncogenesis, this study is complicated part due to an inability to isolate colibactin adequately. Here, we created fluorescent probes activated by ClbP, the colibactin-maturing peptidase, to identify high-colibactin-producing strains. Our probe served as a valuable clinical diagnostic tool that allowed simple high-throughput diagnostic screening of clinical samples. Furthermore, the probe also allowed identification of high-colibactin producers that would help advance our understanding of colibactin biosynthesis.

Full text of DOI:10.1021/acs.orglett.9b01345

Letter
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Activity-Based Probe for Screening of High-Colibactin Producers
from Clinical Samples
†
†
‡
†
†
Yuichiro Hirayama, Yuta Tsunematsu, Yuko Yoshikawa, Ryota Tamafune, Nobuo Matsuzaki,
§
∥
∥
†
⊥
Yuji Iwashita, Ippei Ohnishi, Fumihiko Tanioka, Michio Sato, Noriyuki Miyoshi,
Michihiro Mutoh, Hideki Ishikawa, Haruhiko Sugimura, Keiji Wakabayashi,
and Kenji Watanabe*
#
@
§
∇
,
†
†
Department of Pharmaceutical Sciences, University of Shizuoka, Shizuoka 422-8526, Japan
‡
School of Veterinary Medicine, Faculty of Veterinary Science, Nippon Veterinary and Life Science University, Tokyo 180-8602,
Japan
§
Department of Tumor Pathology, Hamamatsu University School of Medicine, Shizuoka 431-3192, Japan
Division of Pathology, Iwata City Hospital, Iwata 438-8550, Japan
∥
⊥
Graduate School of Nutritional and Environmental Sciences, University of Shizuoka, Shizuoka 422-8526, Japan
#
Epidemiology and Prevention Division, Center for Public Health Sciences, National Cancer Center, Tokyo 104-0045, Japan
@
Department of Molecular-Targeting Cancer Prevention, Kyoto Prefectural University of Medicine, Kyoto 602-8566, Japan
Graduate Division of Nutritional and Environmental Sciences, University of Shizuoka, Shizuoka 422-8526, Japan
∇
*
S Supporting Information
ABSTRACT: While high-colibactin-producing Escherichia coli is thought to be
associated with colorectal oncogenesis, this study is complicated part due to an
inability to isolate colibactin adequately. Here, we created fluorescent probes activated
by ClbP, the colibactin-maturing peptidase, to identify high-colibactin-producing strains.
Our probe served as a valuable clinical diagnostic tool that allowed simple high-
throughput diagnostic screening of clinical samples. Furthermore, the probe also allowed
identification of high-colibactin producers that would help advance our understanding of
colibactin biosynthesis.
1
,7,9
n 2006, the gene cluster clb responsible for the biosynthesis
hybrid megasynthetases required for the production of Clb
I
of the genotoxin colibactin (Clb) was identified in the
(Figure 1A,B). The gene cluster also encodes many auxiliary
enzymes, including ClbP that is thought to activate a precursor
1
phylogenetic group B2 cell line of Escherichia coli. The gene
cluster was found not only in E. coli but also in other
10
into the genotoxic final product, based on the crystal
2
11
Enterobacteriaceae. The Clb-positive (clb+) E. coli was shown
structure of ClbP and peptidases ZmaM and XcnG with a
to induce DNA double-strand breaks in eukaryotic cells both in
vitro and in vivo. The affected cells can present signs of
low degree of sequence homology to ClbP (approximately 30%
identity) that are responsible for the formation of natural
3
12
13
incomplete DNA repair, which are marked by G2/M cell cycle
arrest and megalocytosis, chromosomal instability, the elevated
frequency of gene mutations, and anchorage-independent
colony formation, demonstrating the carcinogenic potential
products zwittermicin A and xenocoumacins, respectively,
from their respective N-acylated precursors. These peptidases
catalyze hydrolysis of precursor compounds to yield the final
bioactive molecules, furnishing the host microorganism a self-
resistance mechanism resembling a “prodrug strategy” in which
a bioactive compound is kept inactive by attachment of a
1
,3
of Clb. Nevertheless, Clb has never been purified despite
many attempts. Clb is thought to cause DNA interstrand cross-
14,15
4
protective group
(Figure 1B). Here, we present the
link, and structures of Clb derivatives that can alkylate DNA
5
−8
creation of fluorescent probes activated specifically by ClbP
for simple high-throughput screening (HTS) of E. coli isolates
have been described.
However, the precise structure of the
biologically active form of Clb remains elusive. Genetic and
functional analyses of the biosynthetic gene cluster revealed
that the cluster encodes polyketide synthases (PKSs), non-
ribosomal peptide synthetases (NRPSs), and PKS−NRPS
Received: April 17, 2019
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b01345
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
hydrolyzed by ClbP, releasing N-myristoyl-D-asparagine (N-
myr-Asn) and Clb. To detect the peptidase activity of ClbP
that is coupled with Clb biosynthesis, we chemically
synthesized Clb Probe 1 (1), 2 (2), and 3 (3), all containing
an N-myr-Asn moiety fused to either 7-amino-4-methylcou-
marin (1 and 2) or naphthalimide (3) as a terminal
fluorophore via a peptide bond (Figure 1C along with the
Supporting Information and Figures S16−S40). The fluo-
rescence signal intensifies upon release of the fluorophore from
the N-myr-Asn moiety through the cleavage of the peptide
bond by ClbP, allowing sensitive detection of Clb production
in E. coli. Probe 2 incorporated an L-alanine residue at the P1
position to mimic precolibactin more closely so that ClbP
might exhibit a greater affinity for the probe. After the release
of the N-myr-Asn moiety from 2 by ClbP, an alanine
16,17
aminopeptidase endogenous to E. coli
could remove the
L-alanine residue to activate the fluorophore. Probe 3 was
designed to examine the suitability of a naphthalimide
18
fluorophore previously developed elsewhere for our study.
Furthermore, to evaluate the specificity of ClbP at the β-
position of the substrate, we also prepared Clb CN probe 1 (4)
in which the D-asparagine residue of the N-myr-Asn moiety
was replaced with a nitrile-bearing unnatural D-amino acid
Figure 1. Proposed biosynthetic pathway of colibactin. (A) Gene
cluster of colibactin biosynthesis. Abbreviations: PKS, polyketide
synthase; NRPS, nonribosomal peptide synthetase. (B) Proposed
chemical structures of precolibactin and bioactive colibactin.
Precolibactin is a colibactin precursor containing an N-myristoyl-D-
asparagine (N-myr-Asn) residue (blue), predicted to be produced by
the combined actions of polyketide synthases (ClbC and -I),
nonribosomal peptide synthetases (ClbN, -H, and -J), PKS−NRPS
hybrid megasynthetases (ClbB and -K), and others (ClbD−G, -L, -O,
and -Q) that undergoes the removal of the N-myr-Asn residue via
hydrolysis by ClbP prior to maturation into the bioactive colibactin.
Abbreviation: PP, phosphopantetheinylation. (C) The 7-amino-4-
methylcoumarin conjugate of N-myr-Asn (Clb probe 1, 1) yields
blue-shifted fluorescence with an emission wavelength at 460 nm
upon cleavage of the fluorophore from the conjugate by ClbP. The
chemical structures of Clb probe 2 (2), probe 3 (3), and CN probe 1
(
Figure 1C and the Supporting Information and Figures S11−
S15).
To test whether the simplest probe 1 would allow
fluorescence-based detection of in vivo ClbP activity, we
prepared a BL21(DE3) E. coli strain heterologously producing
14,19−21
the recombinant ClbP
[BL21/ClbP (see Figures S2−
S6)] and a control BL21(DE3) strain without the clbP
expression plasmid (BL21/−) (see the Supporting Information
for details). We also isolated several E. coli strains from a
human colorectal cancer tissue. The presence of the clb gene
cluster in E. coli-4 (see the Supporting Information for details)
was confirmed by detection of the presence of the 16 clb genes
by polymerase chain reaction (PCR) (Figure S2A). Sub-
sequently, the presence of the clb gene cluster in other isolates
E. coli-38, -47, and -50 was confirmed by detection of the
presence of the clbB gene by PCR (Figure S2B). We created a
clbP-deficient mutant of E. coli-50 (E. coli-50/ΔclbP) as a
negative control of the Clb-producing strains (see the
Supporting Information for details). For the assay, E. coli
BL21(DE3) with or without a clbP expression plasmid, E. coli-
4, E. coli-50, E. coli-50/ΔclbP, and the known Clb producer
(4) are also shown.
from a clinical sample that allows fast determination of Clb
producers and identification of high-Clb producers.
At the initial stage of precolibactin (preClb) biosynthesis,
myristoyl-CoA and L-asparagine are shown to be condensed by
1
4
the amide bond forming activity of the NRPS, ClbN (Figure
B). In the subsequent step, preClb is thought to be
1
Figure 2. Screening of fluorescent probes 1−4 using ClbP-producing E. coli strains. The following strains were used: (i) Nissle1917 as a positive
control, (ii) E. coli-4, and (iii) E. coli-50. (iv) E. coli-50/ΔclbP and (v) growth medium only were used as negative controls. Probes 1, 2, and 4 were
incubated at 37 °C for 12 h (yellow), 24 h (blue), and 48 h (green), whereas 3 was incubated at 37 °C for 1, 4, and 7 days. Each data point is a
mean of triplicate measurements with an error bar indicating the standard error.
B
DOI: 10.1021/acs.orglett.9b01345
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 3. High-throughput screening of E. coli strains for differentiating clb+ and clb− isolates using fluorescent probe 1. Data points for isolates
identified by PCR (see the Supporting Information for details) as clb+ and clb− are colored blue and yellow, respectively. E. coli-50 (green) and E.
coli-50/ΔclbP (red arrow) were used as the positive and negative control of the assay, respectively. Each data point is a mean of triplicate
measurements with an error bar indicating the standard error.
9
Nissle 1917 were used. The fluorescence signal from 7-amino-
4
robust marker for detecting and preventing colorectal cancer.
However, a PCR-based labor-intensive, low-throughput
method is currently the sole approach to determine if the E.
coli in a clinical sample is clb+ or clb−. To demonstrate the
effectiveness of 1 in screening for Clb-producing E. coli isolates
from clinical samples simply and rapidly, more than 200
bacterial colonies recovered on MacConkey agar from a
homogenized surgical sample from a colorectal cancer patient
were subjected to the fluorescence assay (see the Supporting
Information for details). For the assay, the isolates were sorted
into clb+ and clb− groups based on the PCR-based test for the
presence of the Clb biosynthetic gene cluster on the genome.
The isolates were grown in LB liquid medium supplemented
with 1 in 96-well plates. HTS was carried out on the basis of
the fluorescence signal whose excitation and emission
wavelengths were 380 and 460 nm, respectively.
E. coli-50 and E. coli-50/ΔclbP were employed as the positive
and negative control, respectively (Figure S10). When the
results from the fluorescence assay and PCR typing were
combined, it became clear that the fluorescence assay with the
probe 1 could accurately identify clb+ isolates from the mixture
of isolates (Figure 3). The mean fluorescence intensities of the
clb+ and clb− isolates were 258.6 ± 15.79 for n = 61 and 21.85
± 4.007 for n = 42, respectively. An unpaired t test performed
on the averaged data yielded a two-tail P value of <0.0001,
indicating that the difference in the average fluorescence
intensities between the clb+ and clb− isolates is statistically
significant. Thus, the screening yielded a clearly distinguishable
difference in fluorescence intensity between the clb+ and clb−
isolates, which allowed quick identification of 61 isolates of the
103 representative isolates from the clinical sample analyzed in
this study.
-methylcoumarin was measured at excitation and emission
wavelengths of 380 and 460 nm, respectively. Significantly
enhanced fluorescence signals from the cultures of all clbP-
expressing strains in the presence of 1 indicated that the
fluorophore was successfully released from 1 by ClbP in vivo
(
Figure S7). Furthermore, extracts of the cultures of Clb
producers Nissle 1917, E. coli-4, and E. coli-50 were analyzed
by LC−HRMS for the formation of N-myr-Asn (Figure S8, i−
iii). The formation of N-myr-Asn was minimal with the E. coli-
5
0/ΔclbP negative control strain (Figure S8, iv). The results of
the LC−HRMS analysis were also exploited to quantitate the
amount of Clb produced by the E. coli strains. Known amounts
of the synthetic sample of N-myr-Asn were used to draw the
standard curve for calculating the amount of Clb present in the
sample analyzed (Figure S8). Using this method, we were able
to determine that Nissle 1917, E. coli-4, and E. coli-50
produced N-myr-Asn at 3.37, 15.90, and 96.45 μg/L of culture
under our experimental condition, respectively. The result
indicated that the newly isolated E. coli-50 could be producing
a substantially higher level of Clb than Nissle 1917.
Next, to examine the performance of the four probes, we
used the four probes, 1−4 against Nissle 1917, E. coli-4, E. coli-
0, and E. coli-50/ΔclbP, respectively, for the fluorescence
5
assay (see the Supporting Information for details). The results
indicated that 1 yielded the most sensitive response across all
strains tested, and the incubation period of 48 h was found to
produce superior intensity readouts in all cases (Figure 2). The
signal intensity obtained with the L-alanine-containing 2 was
less than half of that observed with 1, possibly because 2 is less
soluble than 1. Signals from 3 were too weak to be useful,
possibly because 3 was less soluble than 1. The total lack of a
signal from all strains with 4 strongly indicates that 4 is likely
not accepted as a substrate by ClbP. Among the strains
examined, Nissle 1917, E. coli-4, and E. coli-50 yielded the
high-intensity fluorescence signals with all functional probes
tested (Figure S7). Thus, 1 was selected as the probe of choice,
and E. coli-50 was chosen as the positive control of the assay
for the future studies.
Previous studies have indicated a strong link between Clb
and carcinogenesis and suggested that Clb-producing (clb+) E.
coli could be in part responsible for the colorectal onco-
1
,3,22
genesis.
However, it was also known that many of the E.
coli strains isolated from healthy individuals also carry the clb
1
biosynthetic gene cluster, and the Clb-producing E. coli strain
Nissle 1917 has been used widely as a probiotic for the
23
It has been reported that 66.7% of colorectal cancer patients
harbored clb+ E. coli, only 20.8% harbored non-inflammatory
bowel disease, and non-colorectal cancer control subjects
treatment of intestinal disorders. Therefore, we hypothesized
that only the high-Clb-producing E. coli strains might be
associated with colorectal oncogenesis. Thus, a fast and simple
method of screening for high-Clb-producing strains was
22
harbored clb+ E. coli, suggesting that clb+ E. coli could be a
C
DOI: 10.1021/acs.orglett.9b01345
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
thought to be valuable in detecting and possibly protecting
against E. coli-induced colorectal cancers.
Notes
The authors declare no competing financial interest.
The aim of this study is to design a fluorescence probe that
is activated specifically by an enzymatic activity associated with
Clb biosynthesis. Use of such a probe would allow character-
ization of the Clb biosynthetic activity in each of the large
number of E. coli isolates that can be collected from an
individual in a high-throughput fashion without requiring
highly specialized analytical instruments. We took advantage of
the prodrug strategy of the Clb biosynthesis, where the
peptidase ClbP is employed to remove the N-acyl protecting
group, N-myr-Asn, for maturation of preClb into the bioactive
final product. Condensation of a fluorescent group, such as
coumarin derivatives, with the N-myr-Asn moiety accom-
plished the synthesis of a probe that releases a fluorophore
specifically by cells expressing clbP, hence producing Clb. The
reliability of our probe was verified by both genotype- and
phenotype-based tests, namely PCR-based confirmation of the
presence of the clb gene cluster and LC−HRMS-based
confirmation of the release of the N-myr-Asn moiety. The
high fluorescence intensity in some of the clb− strains could be
due to the difference in the culture density or the pH of the
culture that may accelerate the background hydrolysis of the
probe. The probe can also be cleaved by nonspecific peptidases
present in E. coli. Thus, those strains that happen to produce
elevated levels of such peptidases could yield higher
fluorescence intensities. Using probe 1, we were able to isolate
E. coli-50, which was capable of producing 26-fold more Clb
than Nissle 1917. Identification of such high-Clb-producing
strains would greatly facilitate isolation and characterization of
Clb in detail, which remain an intractable task to date. The use
of our probe also made it possible to screen a large number of
samples within a short period of time using a simple
instrument. For example, we screened more than 200 E. coli
isolates in triplicate to determine whether they are clb+ or clb−
within a matter of hours using only a simple plate reader,
allowing such tests to be performed at clinics and hospitals
where PCR or LC−MS analyses are not possible. Taken
together, our probe has enabled swift identification of high-
Clb-producing E. coli strains that will not only contribute to
determining the exact structure of Clb and how this genotoxin
is biosynthesized but also lay the foundation for understanding
the involvement of clb+ E. coli in colorectal oncogenesis.
ACKNOWLEDGMENTS
■
The authors thank Dr. Shohachi Suzuki, Director of Iwata City
Hospital, for providing us with colorectal cancer tissue. The
authors also thank Dr. Keigo Matsumoto, Dr. Takako
Fukazawa, and Dr. Akihiro Uno at the Division of Digestive
Surgery in Iwata City Hospital and Dr. Shioto Suzuki, Mr.
Akira Kurita, Ms. Mari Muraki, and Mr. Kazuki Hirata at the
Division of Pathology in Iwata City Hospital. This work was
supported by the Development of Innovative Research on
Cancer Therapeutics from the Japan Agency for Medical
Research and Development (AMED) (K. Watanabe,
6ck0106243h0001), Innovative Areas from MEXT, Japan
K. Watanabe, 16H06449), the Takeda Science Foundation
K. Watanabe), the Institution of Fermentation at Osaka (K.
Watanabe), the Princess Takamatsu Cancer Research Fund (K.
Watanabe, 16-24825), and the Yakult Bio-Science Foundation
K. Watanabe).
1
(
(
(
REFERENCES
■
(
1) Nougayrede, J. P.; Homburg, S.; Taieb, F.; Boury, M.;
Brzuszkiewicz, E.; Gottschalk, G.; Buchrieser, C.; Hacker, J.;
Dobrindt, U.; Oswald, E. Science 2006, 313, 848−851.
(2) Putze, J.; Hennequin, C.; Nougayrede, J. P.; Zhang, W.;
Homburg, S.; Karch, H.; Bringer, M. A.; Fayolle, C.; Carniel, E.;
Rabsch, W.; Oelschlaeger, T. A.; Oswald, E.; Forestier, C.; Hacker, J.;
Dobrindt, U. Infect. Immun. 2009, 77, 4696−4703.
(
3) Cuevas-Ramos, G.; Petit, C. R.; Marcq, I.; Boury, M.; Oswald,
E.; Nougayrede, J. P. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 11537−
1542.
4) Bossuet-Greif, N.; Vignard, J.; Taieb, F.; Mirey, G.; Dubois, D.;
Petit, C.; Oswald, E.; Nougayrede, J. P. mBio 2018, 9, 02393-17.
1
(
̀
(5) Engel, P.; Vizcaino, M. I.; Crawford, J. M. Appl. Environ.
Microbiol. 2015, 81, 1502−1512.
(6) Healy, A. R.; Nikolayevskiy, H.; Patel, J. R.; Crawford, J. M.;
Herzon, S. B. J. Am. Chem. Soc. 2016, 138, 15563−15570.
(
7) Faïs, T.; Delmas, J.; Barnich, N.; Bonnet, R.; Dalmasso, G.
Toxins 2018, 10, No. E151.
8) Wilson, M. R.; Jiang, Y.; Villalta, P. W.; Stornetta, A.; Boudreau,
P. D.; Carra, A.; Brennan, C. A.; Chun, E.; Ngo, L.; Samson, L. D.;
Engelward, B. P.; Garrett, W. S.; Balbo, S.; Balskus, E. P. Science 2019,
63, No. eaar7785.
9) Homburg, S.; Oswald, E.; Hacker, J.; Dobrindt, U. FEMS
Microbiol. Lett. 2007, 275, 255−262.
10) Cougnoux, A.; Gibold, L.; Robin, F.; Dubois, D.; Pradel, N.;
(
́
3
(
ASSOCIATED CONTENT
■
(
*
S
Supporting Information
Darfeuille-Michaud, A.; Dalmasso, G.; Delmas, J.; Bonnet, R. J. Mol.
Biol. 2012, 424, 203−214.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.9b01345.
(
11) Dubois, D.; Baron, O.; Cougnoux, A.; Delmas, J.; Pradel, N.;
Boury, M.; Bouchon, B.; Bringer, M. A.; Nougayrede, J. P.; Oswald,
E.; Bonnet, R. J. Biol. Chem. 2011, 286, 35562−35570.
(12) Kevany, B. M.; Rasko, D. A.; Thomas, M. G. Appl. Environ.
Data from the NMR and MS determination of the
compounds and additional experimental information
Microbiol. 2009, 75, 1144−1155.
13) Reimer, D.; Pos, K. M.; Thines, M.; Grun, P.; Bode, H. B. Nat.
Chem. Biol. 2011, 7, 888−890.
14) Brotherton, C. A.; Balskus, E. P. J. Am. Chem. Soc. 2013, 135,
(
(PDF)
(
3
359−3362.
AUTHOR INFORMATION
■
*
(15) Bian, X.; Fu, J.; Plaza, A.; Herrmann, J.; Pistorius, D.; Stewart,
A. F.; Zhang, Y.; Muller, R. ChemBioChem 2013, 14, 1194−1197.
(
Corresponding Author
E-mail: kenji55@u-shizuoka-ken.ac.jp.
16) Lazdunski, C.; Busuttil, J.; Lazdunski, A. Eur. J. Biochem. 1975,
6
(
1
(
0, 363−369.
17) Gonzales, T.; Robert-Baudouy, J. FEMS Microbiol. Rev. 1996,
8, 319−344.
18) Zhang, L.; Duan, D.; Liu, Y.; Ge, C.; Cui, X.; Sun, J.; Fang, J. J.
Am. Chem. Soc. 2014, 136, 226−233.
ORCID
Noriyuki Miyoshi: 0000-0002-4233-4282
Kenji Watanabe: 0000-0002-0463-4831
D
DOI: 10.1021/acs.orglett.9b01345
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(
19) Vizcaino, M. I.; Engel, P.; Trautman, E.; Crawford, J. M. J. Am.
Chem. Soc. 2014, 136, 9244−9247.
20) Healy, A. R.; Vizcaino, M. I.; Crawford, J. M.; Herzon, S. B. J.
Am. Chem. Soc. 2016, 138, 5426−5432.
21) Trautman, E. P.; Healy, A. R.; Shine, E. E.; Herzon, S. B.;
Crawford, J. M. J. Am. Chem. Soc. 2017, 139, 4195−4201.
22) Arthur, J. C.; Perez-Chanona, E.; Muhlbauer, M.; Tomkovich,
(
(
(
S.; Uronis, J. M.; Fan, T. J.; Campbell, B. J.; Abujamel, T.; Dogan, B.;
Rogers, A. B.; Rhodes, J. M.; Stintzi, A.; Simpson, K. W.; Hansen, J. J.;
Keku, T. O.; Fodor, A. A.; Jobin, C. Science 2012, 338, 120−123.
(
23) Sonnenborn, U. FEMS Microbiol. Lett. 2016, 363, fnw212.
E
DOI: 10.1021/acs.orglett.9b01345
Org. Lett. XXXX, XXX, XXX−XXX