RNA-templated molecule release induced protein expression in bacterial cells

Article abstract of DOI:10.1039/c2cc37826d

We have developed a system to release a biologically active molecule in response to the sequence of a target gene. The releasing system, which was triggered by the reduction of an azidomethyl group, was successfully applied to protein expression induced by the release of IPTG triggered by endogenous RNA in bacterial cells. This journal is

Full text of DOI:10.1039/c2cc37826d

View Article Online
View Journal | View Issue
This article is part of the
Nucleic acids: new life, new materials
web‐themed issue
Guest edited by:
Mike Gait
Medical
Research
Council,
Ned Seeman David Liu
Oliver Seitz
Humboldt‐
Universität zu University of
Berlin,
Makoto
Komiyama
Jason
New York
University,
USA
Harvard
University,
USA
Micklefield
University of
Manchester,
UK
Tsukuba,
Japan
Cambridge, UK
Germany
All articles in this issue will be gathered online at
www.rsc.org/nucleic_acids

ChemComm
COMMUNICATION
RNA-templated molecule release induced protein
expression in bacterial cells†‡
Aya Shibata,a Yoshihiro Ito*a and Hiroshi Abe*ab
Cite this: Chem. Commun.,
2013, 49, 270
Received 29th October 2012,
Accepted 13th November 2012
DOI: 10.1039/c2cc37826d
www.rsc.org/chemcomm
We have developed a system to release a biologically active molecule of the lac operator.12 The lac repressor usually binds to the lac
in response to the sequence of a target gene. The releasing system, operator in the plasmid vector and target protein overexpression is
which was triggered by the reduction of an azidomethyl group, was repressed. When the IPTG modified probe was introduced into
successfully applied to protein expression induced by the release of Escherichia coli (E. coli), the templated Staudinger reaction occurs on
IPTG triggered by endogenous RNA in bacterial cells.
the target RNA, followed by the elimination reaction. Next the
elimination reaction proceeds to release IPTG inside the cells. Upon
The ideal drug should function by selectively targeting cells that are IPTG binding with a lac repressor, it activates the promoter and
disease-causing organisms, or cure diseased cells or cancerous induces the expression of GFP. Therefore, we can evaluate the
cells.1 A smart concept has been proposed by Taylor and Ma in efficiency of the molecular release based on genetic information
2000,2 which involves drug release based on genetic information, in by monitoring the fluorescence signal in living cells.
other words, nucleic acid-templated drug release. The concept offers
We have designed and synthesized two different DNA probes that
a number of approaches including increasing drug selectivity for recognize different sites of the target gene (Fig. 1 and 2); one probe
chemotherapy,3 bacterial-directed activation4 and nucleic acid sen- has an azidomethyl-caged immolative linker containing a model
sing that involves the detection of RNA in cells or genetic diagnosis. molecule at the 30 terminal end, whereas the other probe has
Various chemistries have been reported. A major focus of DNA/RNA- triphenylphosphine (TPP) as the reducing reagent at the 50 terminal
templated reactions has been on nucleic acid sensing in vitro and end. The functional molecules containing hydroxy or amino groups
in vivo. These reactions are nucleophilic ligation reactions,5 template can be conjugated to the a-hydroxy group of the mandelate core of
native chemical ligations6 and template Staudinger reactions.7 Some the immolative group by a carbonate ester or carbamate linkages,
of methods have been applied to molecular releasing in vitro. Those respectively, and are released from the probes triggered by the
are photochemical reactions,8 hydrolysis reactions2,9 and the Staudinger reaction.10
template Staudinger reactions.10 The nucleic acid dependent
Initially, we tested whether the probe can release the molecule by
chemical photocatalyst successfully generated toxic singlet oxygen reduction of the azidomethyl group in vitro. In this design the probe
by continuous photo-irradiation in living cells.11
emits a fluorescence signal upon releasing the quencher molecule
Here, we present a RNA-templated molecular releasing system
that causes the expression of a protein in living cells. In order to
develop a system to release a molecule in response to a target RNA
sequence, we selected isopropyl-b-D-thiogalactoside (IPTG) as the
release molecule (Fig. 1). The addition of IPTG to bacteria is a
long-standing approach to induce expression of plasmid based
genes for the production of recombinant proteins under the control
a Nano Medical Engineering Laboratory, Advance Science Institute, RIKEN, 2-1,
Hirosawa, Wako, Saitama 351-0198, Japan. E-mail: h-abe@riken.jp,
y-ito@riken.jp
b PRESTO, Japan Science and Technology Agency, 4-1-8, Honcho, Kawaguchi,
Saitama 332-0012, Japan
† This article is part of the ‘Nucleic acids: new life, new materials’ web themed
issue.
‡ Electronic supplementary information (ESI) available: Experimental methods,
Schemes S1 and S2, Fig. S1–S3. See DOI: 10.1039/c2cc37826d
Fig. 1 IPTG release by the templated reaction induced expression of GFP.
c
270 Chem. Commun., 2013, 49, 270--272
This journal is The Royal Society of Chemistry 2013

Communication
ChemComm
Fig. 2 Time course of the fluorescence intensity in the reaction between 50 nM ON 1 and 250 nM ON 4 with a 50 nM matched (ON 6) or mismatched template (ON 10)
in 20 mM Tris-HCl buffer (pH 7.2) containing 100 mM MgCl2 and 1 mg mlꢀ1 BSA. The reaction was monitored by excitation at 490 nm and emission at 522 nm.
(Fig. 2). p-Methyl red was used as the quencher. The quencher and (n = 4) in 60 min (‘‘n’’ is the number of nucleotide gaps). The
fluorophore molecules are positioned close to each other through increasing numbers of nucleotide gaps reduced the efficiency of
the immolative linker on the DNA probe. The molecular releasing molecular release. In addition, the no gapped template (n = 0) gave a
probe (ON 1) has p-methyl red, which was connected to a caged yield of 23% and the releasing rate was slower than the one or two
linker and fluorescein. When p-methyl red was released by reduction gapped template used. The immolative group was conjugated to the
of the azidomethyl group, efficiency of the fluorescence resonance DNA probe by a rigid acetylene linker. So the steric hindrance
energy transfer (FRET) decreased and the fluorescence signal was between the molecular releasing and reducing probes contributed
observed to increase. The synthesis of the azidomethyl-caged immo- to the low releasing rate on the no gapped template. When the one
lative linker 12 containing p-methyl red is shown in Scheme S1 nucleotide gapped template was used, the releasing rate was the
(ESI‡), and the probes and target sequences are shown in Fig. 2 and largest when compared with other conditions. Thus, subsequent
Fig. S1 (ESI‡). When a high concentration of dithiothreitol was experiments were carried out using the one nucleotide gapped
added to a solution of ON 1, a strong fluorescence signal appeared template.
as a result of releasing p-methyl red on ON 1 (Fig. S2, ESI‡). Thus, we
Next, we tested the fluorescence detection of 23S rRNA in the
have confirmed that the reduction of the azidomethyl group became E. coli JM109 (DE3) strain using a molecular releasing probe.
a trigger of molecular release. We then tested whether the molecular Probes were designed for targeting the sequence 968–998 on
releasing probe can release p-methyl red in response to the target the 23S rRNA, which is known to be accessible by the standard
sequence. The molecular releasing probe (ON 1) and the TPP probe FISH probe.14 The molecular releasing probe and the TPP probe
(ON 4) incubated with or without template DNA, and the fluore- with a matched sequence are both 15 bases long (Fig. S1, ESI‡).
scence signal at the emission wavelength of 522 nm was monitored The TPP probe with a scrambled sequence was used as a
over 60 min (Fig. 2). In the presence of the match sequence template negative control. The targeting sequence had one nucleotide
DNA (ON 6), the fluorescence signal increased in a time dependent gap between the molecular releasing and TPP probes. The cells
manner and reached saturation at 60 min. The releasing yield at were incubated in the buffer containing low concentration of
60 min was 62%. The rate is relatively slow compared with previously SDS in the presence of the probes for 60 min at 37 1C and
reported reactions.2,5–7,9–11 In a previous study, we reported that the analyzed by flow cytometry (Fig. S4, ESI‡). The set of matched
reduction of an azidomethyl group on the template was completed sequence probes offered significantly higher fluorescence signals
in 3 min.7c Therefore, the rate-limiting step of molecular release than the controls. This result shows that molecule release was
appears to be the elimination step that follows the Staudinger performed in response to the target RNA in bacterial cell.
reaction. Conversely, in the absence of the template, no significant
Finally, we tried the expression of the protein based on genetic
increase in fluorescence was observed and the yield was 3%. In the information in living cells. The synthesis of IPTG based caged linker
presence of a single-base mismatched template (ON 10), the reaction 14 is shown in Scheme S2 (ESI‡). The 6-hydroxymethyl group of the
yield was only 5%. These results indicated that the probes were galactosyl moiety of b-galactosides appears to be of considerable
selective.
importance.12 Substitution of the 6-hydroxy group of galactose with,
The rates of the DNA-templated chemical reactions were depen- for example, a methyl group destroys its induction capabilities.
dent on the distance between the reactive groups.9a,13 We then Therefore, caged linker 14 cannot function as an inducer. The
examined the influence of the distance between the caged linker release of IPTG from the molecular releasing probe (ON 13,
and reducing reagent on the efficiency of molecular release (Fig. S3, Fig. S1, ESI‡) was found by expression of GFP in E. coli JM109
ESI‡). As expected, changing the distance between the caged linker (DE3). The cells were incubated with the buffer in the presence of
and reducing reagent influenced the rate of molecular release. The the molecular releasing probes for 30 min, suspended in an LB/Amp
releasing yield was 62% (n = 1), 51% (n = 2), 27% (n = 3) or 28% medium and incubated for 24 h at 37 1C. After 24 h, these cells are
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 270--272 271

ChemComm
Communication
Fig. 3 Detection of the expression of GFP under the control of the lac operator in E. coli JM109 (DE3) cells. (A) Flow cytometry histogram showing cell-count frequency
vs. fluorescence intensity for each probe. These histograms correspond to cells that were treated by no probe (control); by IPTG (positive control); by matched
probes; and by scramble probes. (B) Medians of fluorescence intensity and signal-to-background ratios calculated from the flow cytometry histogram. (C) Imaging
of GFP in E. coli JM109 (DE3) cells. Microscope settings were as follows: excitation: 488 bandpass filter; emission: 500/10 bandpass filter.
analyzed by flow cytometry and a fluorescence microscope. The flow
H.A. was financially supported by MEXT, NEDO and PRE-
cytometry histogram is presented in Fig. 3A and the median STO. A.S. was financially supported by a Grant-in-Aid for Young
values of fluorescence intensity are plotted in Fig. 3B. The values Scientists (B) and the Special Postdoctoral Researcher Program
on the bars indicate the signal-to-background (S/B) ratio, which of RIKEN. We are grateful to the support of BSI’s Research
is defined as the relative ratios of fluorescence intensity to Resources Center for mass spectrum analysis and BSI-Olympus
the background signal from the cell (no IPTG or probes). The S/B Collaboration Center for imaging equipment.
ratios were 1.79 ꢁ 0.35 and 0.86 ꢁ 0.45 for matched and scrambled
Notes and references
sequences, respectively. Fig. 3C shows the fluorescence images and
bright field images of the cells. A strong fluorescent signal was
observed in the case of the matched probe pair. Conversely,
a negligible signal was observed from the scrambled probe
pair. These results clearly showed that the expression of GFP was
induced by the release of IPTG based on gene information in
the cells.
In this study, we chose the 23S rRNA as a target of the templated
reaction. The growing bacterial cells contain about 70 000 ribo-
somes15 and the cell volumes in E. coli were 0.6–0.7 mm.3,16 From
these values, the concentration of 23S rRNA in the E. coli is
calculated to be B180 mM. The concentration of IPTG required for
50% of the operator DNA to be released from the lac repressor–
operator complex is 2.2 mM.17 Thus, in this model experiment, the
templated reaction could produce the required amount of IPTG to
induce protein expression. While rRNA represents 80–85% of intra-
cellular RNA, mRNA represents only 1–4% of the total RNA content.
Therefore, in the case of using this releasing system as target mRNA,
it is necessary to select a molecule that is effective at lower
concentration (order of BnM; for example dexamethasone18 or
doxorubicin3).
1 F. Kratz, I. A. Muller, C. Ryppa and A. Warnecke, ChemMedChem,
2008, 3, 20–53.
2 Z. C. Ma and J. S. Taylor, Proc. Natl. Acad. Sci. U. S. A., 2000, 97,
11159–11163.
3 M. Shamis, H. N. Lode and D. Shabat, J. Am. Chem. Soc., 2004, 126,
1726–1731.
4 R. C. Kelly, M. E. Bolitho, D. A. Higgins, W. Y. Lu, W. L. Ng,
P. D. Jeffrey, J. D. Rabinowitz, M. F. Semmelhack, F. M. Hughson
and B. L. Bassler, Nat. Chem. Biol., 2009, 5, 891–895.
5 For examples of nucleophilic ligation reactions, see: (a) S. Sando
and E. T. Kool, J. Am. Chem. Soc., 2002, 124, 2096–2097; (b) H. Abe
and E. T. Kool, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 263–268.
6 T. N. Grossmann and O. Seitz, J. Am. Chem. Soc., 2006, 128,
15596–15597.
7 For examples of Staudinger reactions, see: (a) J. F. Cai, X. X. Li, X. Yue
and J. S. Taylor, J. Am. Chem. Soc., 2004, 126, 16324–16325;
(b) Z. L. Pianowski and N. Winssinger, Chem. Commun., 2007,
3820–3822; (c) K. Furukawa, H. Abe, K. Hibino, Y. Sako, S. Tsuneda
and Y. Ito, Bioconjugate Chem., 2009, 20, 1026–1036; (d) R. M. Franzini
and E. T. Kool, J. Am. Chem. Soc., 2009, 131, 16021–16023.
8 For examples of photochemical reactions, see: A. Okamoto,
K. Tanabe, T. Inasaki and I. Saito, Angew. Chem., Int. Ed., 2003, 42,
2502–2504.
9 For examples of hydrolysis reactions, see: (a) Z. C. Ma and J. S.
Taylor, Bioconjugate Chem., 2003, 14, 679–683; (b) I. Boll, R. Kramer,
J. Brunner and A. Mokhir, J. Am. Chem. Soc., 2005, 127, 7849–7856.
10 K. Gorska, A. Manicardi, S. Barluenga and N. Winssinger, Chem.
Commun., 2011, 47, 4364–4366.
In conclusion, we have developed a novel molecular release
´
system based on genetic information. The system was based on 11 D. Arian, E. Clo, K. V. Gothelf and A. Mokhir, Chem.–Eur. J., 2010, 16,
288–295.
the Staudinger reduction and 1,6-elimination. The molecular
releasing probes were designed and synthesized for this chemistry.
12 W. Boos, P. Schaedel and K. Wallenfe, Eur. J. Biochem., 1967, 1,
382–394.
After reduction of the azide group to the amino group, the molecular 13 Z. J. Gartner and D. R. Liu, J. Am. Chem. Soc., 2001, 123, 6961–6963.
14 K. S. Ko, J. Kruse and N. L. Pohl, Org. Lett., 2003, 5, 1781–1783.
15 R. L. Gourse, T. Gaal, M. S. Bartlett, J. A. Appleman and W. Ross,
releasing probe released the quencher or IPTG. The system was
shown to detect both DNA and RNA sequences and be sensitive to a
Annu. Rev. Microbiol., 1996, 50, 645–677.
single base mismatch. We successfully applied the probe to the 16 H. E. Kubitschek, J. Bacteriol., 1990, 172, 94–101.
17 J. Xu, S. Liu, M. Z. Chen, J. P. Ma and K. S. Matthews, Biochemistry,
2011, 50, 9002–9013.
18 L. E. Holz, K. P. Jakobsen, J. Van Snick, F. Cormont and W. A. Sewell,
release of a functional molecule in response to endogenous RNA,
and the released molecule induced protein expression in
bacterial cells.
J. Inflamm., 2005, 2, 3.
c
This journal is The Royal Society of Chemistry 2013
272 Chem. Commun., 2013, 49, 270--272