Thiazole orange-induced c-di-GMP quadruplex formation facilitates a simple fluorescent detection of this ubiquitous biofilm regulating molecule

Article abstract of DOI:10.1021/ja1091062

Recently, there has been an explosion of research activities in the cyclic dinucleotides field. Cyclic dinucleotides, such as c-di-GMP and c-di-AMP, have been shown to regulate bacterial virulence and biofilm formation. c-di-GMP can exist in different aggregate forms, and it has been demonstrated that the polymorphism of c-di-GMP is influenced by the nature of cation that is present in solution. In previous work, polymorphism of c-di-GMP could only be demonstrated at hundreds of micromolar concentrations of the dinucleotide, and it has been a matter of debate if polymorphism of c-di-GMP exists under in vivo conditions. In this Article, we demonstrate that c-di-GMP can form G-quadruplexes at low micromolar concentrations when aromatic molecules such as thiazole orange template the quadruplex formation. We then use this property of aromatic molecule-induced G-quadruplex formation of c-di-GMP to design a thiazole orange-based fluorescent detection of this important signaling molecule. We determine, using this thiazole orange assay on a crude bacterial cell lysate, that WspR D70E (a constitutively activated diguanylate cyclase) is functional in vivo when overexpressed in E. Coli. The intracellular concentration of c-di-GMP in an E. Coli cell that is overexpressed with WspR D70E is very high and can reach 2.92 mM.

Full text of DOI:10.1021/ja1091062

ARTICLE
pubs.acs.org/JACS
Thiazole Orange-Induced c-di-GMP Quadruplex Formation Facilitates
a Simple Fluorescent Detection of This Ubiquitous Biofilm
Regulating Molecule
†
†
†
‡,§
†
†
Shizuka Nakayama, Ilana Kelsey, Jingxin Wang, Kevin Roelofs, Bogdan Stefane, Yiling Luo,
‡
,§
,†,§
Vincent T. Lee, and Herman O. Sintim*
†
‡
§
Department of Chemistry and Biochemistry, Department of Molecular and Cell Biology, and Host-Pathogen Graduate Program,
University of Maryland, College Park, Maryland 20742, United States
S Supporting Information
b
ABSTRACT: Recently, there has been an explosion of research
activities in the cyclic dinucleotides field. Cyclic dinucleotides,
such as c-di-GMP and c-di-AMP, have been shown to regulate
bacterial virulence and biofilm formation. c-di-GMP can exist in
different aggregate forms, and it has been demonstrated that the
polymorphism of c-di-GMP is influenced by the nature of cation
that is present in solution. In previous work, polymorphism of
c-di-GMPcouldonlybe demonstratedathundredsofmicromolar
concentrations of the dinucleotide, and it has been a matter of
debate if polymorphism of c-di-GMP exists under in vivo condi-
tions. In this Article, we demonstrate that c-di-GMP can form
G-quadruplexes at low micromolarconcentrationswhen aromatic
molecules such as thiazole orange template the quadruplex
formation. We then use this property of aromatic molecule-induced G-quadruplex formation of c-di-GMP to design a thiazole
orange-based fluorescent detection of this important signaling molecule. We determine, using this thiazole orange assay on a crude
bacterial cell lysate, that WspR D70E (a constitutively activated diguanylate cyclase) is functional in vivo when overexpressed in E. Coli.
The intracellularconcentrationofc-di-GMP inanE. Coli cell that is overexpressed with WspR D70E is very highand can reach 2.92 mM.
’
INTRODUCTION
manifestation is bacterial resistance to conventional antibiotics.
Therefore, suppressing biofilms synthesis should make bacteria
more susceptible to current antibiotics. C-di-GMP signaling
initiates with synthesis by diguanylate cyclases (DGCs) and
terminates when degraded by phosphodiesterase (PDEs). In-
vestigations into all sequenced bacterial genomes reveal that all
genomes have at least one DGC or PDE, suggesting the evolu-
Over 70% of hospital infections are caused by biofilm forming
bacteria; yet there are no drugs currently in clinical use that
inhibit biofilm formation in bacteria. In the past decade, there has
been an explosion of research in quorum sensing (QS) and 3 ,5 -
cyclic diguanylic acid (c-di-GMP) signaling in bacteria. It is now
well appreciated that these two processes regulate both virulence
factors production and biofilm formation in a variety of bacteria,
including those of clinical relevance. Investigations that unravel
the molecular details of these two important transduction
processes in bacteria would lead to the identification of novel
drug targets for the development of next-generation anti-infec-
tives, which are not bacteriostatic but rather target bacterial
1
0
0
2
2
tionary importance of this signaling pathway in bacteria. Despite
the central role that c-di-GMP plays in bacteria, very little
information is known about the actual macromolecular targets
that c-di-GMP regulates. Only the metabolism proteins of c-di-
5
GMP (DGCs and PDEs) as well as a handful of adaptor proteins
6
and riboswitches have been characterized. Most of the adaptor
3
proteins that bind to c-di-GMP do not have enzymatic properties
on their own, but the effector proteins that these adaptor proteins
regulate are largely unknown. Additionally, the environmental
cues that regulate the metabolism of c-di-GMP are not well
characterized. The dearth of detailed information regarding c-di-
GMP signaling is probably due to a variety of factors, one
of which is the lack of specific chemical probes to readily detect
virulence and biofilm formation in bacteria. Molecules that
target bacterial pathogenesis rather than killing bacteria are
expected to provide less evolutionary pressure for bacteria to
develop resistance mechanisms.
C-di-GMP, a secondary messenger present uniquely and
ubiquitously in bacteria, plays a central role in bacterial biofilm
formation and regulation of virulence-related factors in diverse
4
bacteria. Biofilm formation encases the bacteria with a poly-
saccharide and proteinaceous matrix, resulting in enhanced
resistance to chemical and physical stress, and the clinical
Received: October 10, 2010
Published: March 08, 2011
r 2011 American Chemical Society
4856
dx.doi.org/10.1021/ja1091062
|J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Figure 1. TO is predicted to interact with c-di-GMP to give a 1:1, 1:4, or 1:8 complexes in the presence of monovalent cations. In scenarios whereby TO
is intercalated between the two planes of the G-tetrad (such as Q2), it is hypothesized that the fluorescence of TO will be enhanced.
c-di-GMP. Currently, c-di-GMP is detected in vitro and in vivo
via a tandem HPLC-MS protocol or the use of recombinant
fluorescent proteins, respectively. Bacterial cells or culture
such as duplex DNA, via association with the minor groove,
intercalation between base pairs, or association with the nega-
7
9a,11
tively charged phosphate backbone.
Out of these three
media may contain myriads of other nucleotides or other
modes of TO association with polynucleic acids, intercalation
and minor groove binding can lead to fluorescence enhancement
but not TO association with the phosphate backbone. Upon TO
confinement in nucleic acid cavities, the nonradiative decay
channel is blocked due to the molecular confinement. The
quantum yield of TO in confined cavities subsequently increases
8
compounds that might coelute with c-di-GMP. Therefore, good
resolving HPLC columns as well as sensitive MS instruments are
needed to detect c-di-GMP in crude cell lysates via the tandem
HPLC-MS protocol. The use of recombinant proteins to detect
c-di-GMP is more appropriate for in vivo applications. Because
this approach involves a transformation step, c-di-GMP cannot
be detected in bacteria that do not transform or do not express
the recombinant fluorescent protein very well. The aforemen-
tioned setbacks associated with current c-di-GMP detection
assay argue for the development of new detection platforms for
this signaling molecule. A detection method that uses simple
shelf-stable reagents without the need for sophisticated instru-
mentation or separation step will complement the current two
sate-of-the-art detection methods. Herein, we describe a simple
and specific fluorescent detection of c-di-GMP using thiazole
orange (TO, 1-methyl-4-[(3-methyl-2(3H)-benzothiazolylidene)-
methyl]quinolinium p-tosylate). TO is not generally fluorescent
in the presence of simple nucleotides such as GTP, GMP, or
cGMP. We demonstrate that c-di-GMP is an exception to this
rule and can be detected in vitro with this dye in a selective
fashion without the need for prior HPLC purification.
10
to up to 0.4. It therefore follows that nucleotides such as GTP
and cGMP, which do not have the ability to confine TO, cannot
be detected with TO. C-di-GMP, on the other hand, has been
shown to form aggregates such as tetramolecular quadruplexes
and octamolecular complexes in solutions containing mono-
þ 12
valent cations such as K . We rationalized that we could use
this unique property of c-di-GMP to form aggregates, for specific
detection in the presence of other guanine containing nucleotides
that do not generally form aggregates at micromolar concentra-
tions in water. Specifically, we hypothesized that in the presence
of potassium, sodium, or ammonium cations, c-di-GMP would
either form a 4:1 complex with TO (see Q1 or Q2 in Figure 1) or
an 8:1 complex with TO (see Q3 or Q4 in Figure 1), and as the
TO in these complexes would be restricted to freely rotate, the
nonradiative channel in excited TO would be closed and the
quantum yield of the confined TO would increase appreciably for
it to become fluorescent. Because the fluorescence enhancement
in TO is derived from restricted rotation in the molecule, it is
reasonable to assume that complexes Q2 and Q4 would be more
fluorescent than Q1 and Q3 because the TO in Q2 or Q4 is
embedded between two G-tetrad planes, whereas in Q1 or Q3
only one G-tetrad plane π-stacks with TO (Figure 1).
’
RESULTS AND DISCUSSION
Thiazole orange (TO) is a known fluorescent intercalator of
nucleic acids and can be used to detect different DNA and RNA
9
structures. TO has a low fluorescence quantum yield in aqueous
10
solution (Φf = 0.0002). TO can interact with polynucleotides,
4
857
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Figure 2. In the presence of monovalent cations, c-di-GMP interacts with TO and enhances the fluorescence of TO. The enhancement of TO
fluorescence was dependent on the nature and concentration of the cation. [c-di-GMP] = 20 μM, [thiazole orange] = 30 μM. Ex. 507 nm, em. 518-
7
4
00 nm. Buffer: 10 mM Tris-HCl (pH 7.5) containing (a) 1 M of NaCl, KCl, or NH OAc. (b) Fluorescence of TO-c-di-GMP complex in the presence
of different salt concentrations (0 mM, 50 mM, 150 mM, 250 mM, 500 mM, 1 M, or 2 M). (c) Q1 and Q2 are two possible modes of TO interaction with
c-di-GMP in the presence of a templating cation.
þ
Based on our proposal in Figure 1, among alkali metals, Li
that do not promote G-quadruplex formation
higher order aggregates of c-di-GMP or TO, which are less
fluorescent.
12,13
would not be
ideal for our detection scheme. In the c-di-GMP octamolecular
complex, the G-tetrads interlock each other. It can therefore be
argued that there is more space between the G-tetrad planes in
c-di-GMP tetramolecular complexes than in c-di-GMP octa-
molecular complexes. Consequently, intercalation into tetra-
molecular complexes might be easier than in the octamolecular
complex. Jones has shown that potassium promotes formation
of the c-di-GMP octamolecular complexes, whereas the propen-
sity for c-di-GMP to form octamolecular complexes is not high
In a 1:4 TO-c-di-GMP complex, it is reasonable to assume
that it is unlikely that both TO and the monovalent cations would
reside between the planes of the G-tetrad. Therefore, the likely
scenario is for the metal to reside between the G-tetrad planes,
whereas TO interacts with the G-tetrad plane via π-π interac-
15
tion (Q1, end-stacking mode, Figure 2) or whereby TO
intercalates between the two c-di-GMP G-tetrads and the
monovalent cation resides within the plane of the G-tetrad
16
(Q2, Figure 2). UV analysis of c-di-GMP interaction with
TO revealed that when c-di-GMP was added to TO, there was a
red shift in the TO absorption spectrum (Figure 3). This is an
indication of a π-π interaction between the guanine bases of
1
2
for sodium. Although it is known that potassium forms more
14
stable G-quadruplexes, we rationalized (on the basis of our
working model that TO would preferentially bind to c-di-GMP
tetramolecular complexes) that sodium might be the cation of
choice for c-di-GMP detection by TO. In agreement with this
working model, the detection of c-di-GMP with TO was most
17
c-di-GMP and TO.
The detection of c-di-GMP with TO is specific. In the
presence of increasing c-di-GMP concentration, the fluorescence
intensity of TO increased (see Figure 4a). The detection limit for
this assay is 5 μM. In line with expectation, a mixture containing
several nucleotides such as GTP, cGMP, ATP, etc. (see
Figure 4b) could not be detected with TO, whereas c-di-GMP
or a mixture of several nucleotides plus c-di-GMP could be
detected with TO. Importantly, the fluorescent signal of the
sample that contained only c-di-GMP and that which contained
several other nucleotides and c-di-GMP were similar (see
Figure 4b, compare lines a and c). At high concentrations
of both c-di-GMP and TO (300 μM), there was no need
for a spectrophotometer to observe complex formation (see
þ
sensitive in the presence of Na as compared to in the presence
þ
þ
of K or NH cation (see Figure 2).
4
The TO fluorescence in the TO-c-di-GMP complex was
dependent on the concentration of the monovalent cation. As the
concentration of the cation increased (up to 1 M), the fluores-
cence intensity also increased (see Figure 2b). This was expected
as higher cation concentration would promote more of the
higher aggregate of c-di-GMP, which is required for TO seques-
tration and fluorescence enhancement. Beyond 1 M cation
concentration, the fluorescence intensity of TO decreased. This
could be attributed to several reasons, such as the formation of
4
858
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Figure 3. UV spectra of TO þ c-di-GMP in buffer. Addition of monovalent cations to TO causes a red shift in the absorption spectrum. Buffer:
1
3
4
0 mM Tris-HCl (pH 7.5) containing 500 mM of one of these salts; NaCl, KCl, and NH OAc or no salt was added. [c-di-GMP] = 70 μM, [TO] =
0 μM.
Figure 4. (a) Dose-response for TO-c-di-GMP interaction. [TO] = 30 μM, buffer: 10 mM Tris-HCl (pH 7.5) containing 1 M NaCl. Ex. 507 nm, em.
5
1
18-700 nm. (b) Selectivity of this detection system. [c-di-GMP] = 20 μM, each [nucleotide] = 20 μM, buffer: 10 mM Tris-HCl (pH 7.5) containing
M NaCl, [TO] = 30 μM. (c) Visible difference between the presence and absence of c-di-GMP.
Figure 4c). c-di-AMP has also been shown to be a signaling
octamolecular complexes), it can be used to qualitatively deter-
mine if a G-quadruplex is present in solution. Jones has shown
that a positive CD peak at around 300 nm is indicative of
tetramolecular or octamolecular complex formation by c-di-
26
molecule in bacteria. Incubation of TO with c-di-AMP did not
lead to fluorescence enhancement of TO (see Supporting
Information, Figure S1). c-di-AMP does not form quadruplex,
and as we believe that the fluorescent enhancement of TO
by c-di-GMP is via intercalation between the G-tetrad plane,
the c-di-AMP result is not surprising and in fact augments our
hypothesis.
12
GMP. This CD signature, which indicates the presence of
c-di-GMP tetramolecular or octamolecular complexes, was cor-
roborated with NMR studies. Jones’ study used high concentra-
tions of c-di-GMP (hundreds of micromolar). The physiological
concentration of c-di-GMP ranges from high nanomolar to tens
of micromolar, so it is a matter of serious debate whether the
polymorphism of c-di-GMP is indeed relevant under in vivo
conditions. We therefore sought to establish if c-di-GMP can
form tetramolecular and octamolecular complexes at low micro-
molar ranges (0-100 μM). Between 0 and 100 μM of c-di-GMP,
Having established that TO can detect c-di-GMP in a con-
centration-dependent manner, we proceeded to gain some in-
sights into the aggregation state of the TO-c-di-GMP complex.
Circular dichroism (CD) is a powerful tool for identifying the
12,18
aggregation state of nucleic acids.
Although CD cannot give a
detailed molecular structure of G-quadruplex (tetramolecular or
4
859
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Figure 5. CD spectra of c-di-GMP alone and c-di-GMP-TO complexes. [c-di-GMP] = 70 μM, [TO] = 30 μM. Buffer: 10 mM Tris-HCl (pH 7.5)
containing (a) 1 M salt (NaCl, KCl, NH OAc, and LiCl) or no metal added, (b) KCl at various concentrations, 0, 50, 150, 250, 500 mM, 1 M, was added
4
to the buffer.
Figure 6. CD of TO-c-di-GMP complex in the presence of various monovalent cations at different temperatures. Buffer: 10 mM Tris-HCl (pH 7.5)
containing 1 M salt (NaCl, KCl, NH OAc, and LiCl). [c-di-GMP] = 70 μM, [TO] = 30 μM. The temperatures used were 10, 25, and 50 °C.
4
the amount of c-di-GMP tetraplexes or octaplexes formed was
minimal. In this concentration range, only potassium promoted
some G-quadruplex formation (see Figure 5a, light pink solid
line: 1 M KCl). However, upon the addition of TO to c-di-GMP
lines). It is noteworthy, that although G-quadruplexes formed by
c-di-GMP have been previously observed in solution, the con-
centrations of c-di-GMP that was required to form G-quadru-
plexes were several hundreds of micromolar and were not
physiologically relevant. Here, we show that in the presence of
aromatic molecules such as TO, c-di-GMP can form stable
G-quadruplexes at low micromolar concentrations.
þ
þ
þ
in buffers containing monovalent cations Na , K , NH , but
4
þ
not Li , a strong positive CD peak around 300 nm appeared in
the CD spectra (see Figure 5a, blue, red, purple, or green solid
4
860
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Figure 7. Job plot. Total [TO] þ [c-di-GMP] was fixed at 50 μM. The experiment was done in triplicate, and all of the triplicate data points are plotted
on the graphs. Buffer: 10 mM Tris-HCl (pH 7.5) containing 1 M salt (NaCl, KCl, and NH OAc). Ex. 507 nm, em. 533 nm.
4
Interestingly, although TO fluorescence was highest when
sodium was the templating cation, the CD intensity was highest
when potassium was the templating cation (compare Figures 2a
and 5a). Also, the maximum of the CD spectrum when sodium
was the templating cation is slightly red-shifted as compared to
those of potassium and ammonium (Figure 5a, compare the blue
dangerous to speculate the origin of this difference. We are
þ
however tempted to hypothesize that because Na is a smaller
þ
þ
þ
þ
cation, as compared to K or NH (size of Na = 0.95 Å, K =
4
þ
þ
1.33 Å, and NH4 = 1.48 Å), Na can nicely fit into the cavity in
the plane of the G-tetrad to form structures such as Q2 (see
þ
þ
Figures 1 and 2). On the other hand, K and NH , being bigger
4
þ
line (NaCl) with the red (KCl) or purple (NH OAc) lines).
than Na , prefer to sit between two G-tetrad planes to give other
4
Surprisingly, the TO-c-di-GMP G-quadruplexes in buffers
containing excess potassium or ammonium cations were stable
even at 50 °C, whereas in the presence of excess sodium, the c-di-
GMP G-quadruplex structure collapsed at 50 °C (see Figure 6).
complexes such as Q1, Q3, or Q4 (see Figure 1). As these
complexes are different, it is not surprising that the induced CD
in TO in these different complexes also differ.
To determine the stoichiometry between TO and c-di-GMP,
we performed a Job plot (Figure 7). The Job plot data indicate
that the stoichiometry of TO to c-di-GMP in the presence of
sodium or potassium or ammonium is 1:4.6, 1:4.9, and 1:6.1,
respectively. From these data, one can conclude that the major
TO-c-di-GMP aggregate species in buffers containing excess
sodium or potassium is likely to be a tetramolecular complex,
whereas more of higher order aggregates are present when
ammonium acts as the templating cation. As the fluorescent
intensity of TO decreases when the concentration of higher
aggregates of TO-c-di-GMP increases (compare Figures 2 and
7), we are tempted to speculate that the aggregate, which is
responsible for TO fluorescence enhancement, is the TO-c-di-
GMP G-quadruplex form (see Figure 1, Q1 and Q2). Further
work is however warranted to confirm this hypothesis.
þ
The lower stability of the c-di-GMP/TO/Na complex, as
þ
þ
compared to c-di-GMP/TO/K or c-di-GMP/TO/NH4 com-
plexes, seems to be at odds with our observation of fluorescence
enhancement, which we have attributed to the confinement of
TO within the G-quadruplex planes. If TO confinement as well
as the orientation of confinement both contribute to fluorescence
enhancement in TO, then the stability of the complex alone
cannot be used to predict which complex is going to be the most
fluorescent. Indeed, both CD and UV data (Figures 3, 5, and 6)
reveal that the orientation of TO in the c-di-GMP/TO complex,
in the presence of different metal salts, is different. TO is achiral
and therefore is not expected to have a CD spectrum. However,
upon binding to chiral molecules such as nucleic acids, TO can
exhibit induced CD. A CD band centered around 490-500 nm
has been suggested as providing circumstantial evidence of
As mentioned earlier in this Article, the main aim of this
project was to find a simple shelf-stable reagent for detecting c-di-
GMP, although the mechanistic insights that have been gained
from this study regarding the interactions of c-di-GMP and TO
are also interesting and important for the basic understanding of
how different nucleic acid structures interact with dye molecules.
10a,19
intercalation of cyanine dyes into polynucleotides.
We
observed a small induced CD between 490 and 540 nm (the
region where TO absorbs) only in the presence of sodium cation
but not in the presence of potassium or ammonium cations
(Figure 6). Without additional experimental evidence, it is
4
861
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Figure 8. (A) Enzymatic conversion of GTP into c-di-GMP by WspR D70E (DGC). Enzymatic reaction conditions: [WspR] = 1 μM, buffer, 10 mM
Tris-HCl (pH 8.0) containing 100 mM NaCl, 5 mM MgCl
2
, reaction temperature was 37 °C. Detection conditions: [TO] = 30 μM, buffer, 10 mM Tris-
HCl (pH 7.5) containing 1 M NaCl. (B) Fluorescence detection of c-di-GMP in E. Coli cell lysate (see Experimental Section for details). Detection
conditions: [TO] = 30 μM, buffer, 10 mM Tris-HCl (pH 7.5) containing 1 M NaCl. In line with expectation, bacteria that overexpressed
phosphordiesterase, RocR, or which contained an “empty” plasmid (pVL1321) did not produce detectable c-di-GMP.
Our group is interested in developing assays to quickly assay the
activity of c-di-GMP metabolism proteins. Previously, others
have done this particular assay using radiolabeled inorganic
polymorphism and metabolism as well as identify the key players
in this interesting signaling cascade. Fluorescent tools that detect
25
bioanalytes have played important roles in the advancement of
biology, and it is expected that specific tools that can be used to
characterize c-di-GMP will find broad utility in the c-di-GMP
field. In this Article, we show that TO can discriminate c-di-GMP
from other small nucleotides. This discovery could become
useful for in vitro assays whereby other macromolecules such
as DNA and RNA, which also bind to TO, can be separated from
the nucleotides via a simple filtration process. Our TO detection
assay for c-di-GMP is a nice complement of the tandem HPLC/
MS method as it does not require a separation/purification step.
We envisage potential future applications such as high-through-
put assays to investigate c-di-GMP binding molecules. Impor-
tantly, this work also reveals that the polymorphism of c-di-GMP
can be remarkably influenced by the presence of aromatic
molecules, which form higher aggregates with c-di-GMP. Our
fluorescence data reveal that even at low micromolar concentra-
tions (less than 10 μM), c-di-GMP can enhance the fluorescence
of TO. As it is likely that this enhancement of TO fluorescence by
c-di-GMP occurs via tetramolecular complex formation, this
work provides important evidence that the polymorphism of
c-di-GMP is relevant at physiological conditions. Aggregation of
c-di-GMP by aromatic molecules is bound to affect the intracel-
lular concentrations of monomeric and dimeric c-di-GMP con-
centrations, and one can speculate that aromatic molecules-
induced polymorphism of c-di-GMP could also be an important
regulatory mechanism in bacterial biofilm formation.
20
phosphate, which is not convenient to use due to safety reasons
21
or by HPLC detection of whole cell extracts. Toward devel-
oping a measurement system that does not rely on radioactivity
or separative methods, we have used our TO detection assay to
assay the conversion of GTP into c-di-GMP by the diguanylate
cyclase, WspR (see Figure 8A). To make meaningful conclusions
from such experiments, it is always necessary to determine if the
overexpressed protein is functional/active inside bacterial cells.
Using our simple TO detection assay, we have managed to
22
determine that WspR D70E (a constitutively activated DGC
that is overexpressed in E. Coli) is functional in vivo (see
Figure 8B). The overexpression of WspR D70E in E. Coli
5
resulted in the production of 8.79 ꢀ 10 c-di-GMP molecules
per cell. The typical E. Coli cell has the dimension of 2 μm ꢀ 0.5
-
15
μm ꢀ 0.5 μm and a corresponding volume of 0.5 ꢀ 10
L,
which implies that there is 2.92 mM c-di-GMP inside the E. Coli
cell (see the Supporting Information for the details). It is of note
that our TO detection of c-di-GMP was performed on crude cell
lysate, without any prior separation step. We also used HPLC
purification, followed by MS verification of the identity of c-di-
GMP, to quantify intracellular c-di-GMP concentrations of the
E. Coli (see the Supporting Information). The HPLC/MS
quantification method gave an estimated c-di-GMP concentra-
tion of 3.43 mM inside the E. Coli cell, remarkably similar to our
estimate of 2.92 mM using our TO assay. The estimate of
2
.92 mM c-di-GMP inside E. Coli is significantly higher
(
1000ꢀ) than the basal concentration of c-di-GMP in a variety
8,23
of bacteria (which is in the single digit micromolar range).
’
EXPERIMENTAL SECTION
This goes to show that DGCs, such as WspR D70E, can
synthesize c-di-GMP in bacteria in an uncontrolled fashion and
flush the cells with high concentrations of c-di-GMP.
General Methods for Optical Measurements. Absorbance
spectra were obtained on a JASCO V-630 spectrophotometer with 1 cm
path length cuvette. Fluorescence studies were performed on a Varian
Cary Eclipse fluorescence spectrophotometer with 1 cm path length
cuvette. CD experiments were performed by a JASCO J-81 spectro-
polarimeter with 1 cm path length cuvette. The concentration of a stock
solution of c-di-GMP and TO was determined by measuring the
absorbance at 260 nm for c-di-GMP and 501 nm for TO and using
21 600 and 63 000 M cm as molar extinction coefficients for c-di-
GMP and TO, respectively. TO was purchased from Sigma-Aldrich.
rGTP, rATP, rCTP, and rUTP were purchased from Promega. dATP,
dGTP, dCTP, and dTTP were purchased from New England Biolab.
’
CONCLUSION
It has now been over two decades since Benziman published
his seminal paper, which showed that c-di-GMP regulates
24
cellulose production in bacteria. Since then, it has been
subsequently shown that c-di-GMP is a master regulator of
several processes in bacteria and that unraveling the molecular
details of c-di-GMP signaling will no doubt reveal important
targets for antibiotic therapy. In pursuing this goal, it is im-
perative that we understand factors that affect c-di-GMP
-
1
-1
0
0
cGMP (guanosine 3 ,5 -cyclic mono phosphate) was purchased from
4
862
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
0
CALBIOCHEM. GMP (guanosine-5 -monophosphate) was purchased
from Amersco.
Schramm, V. L. Biochemistry 2006, 45, 12929–12941. (d) Ueda, A.;
Wood, T. K. PloS Pathogens 2009, 5, e1000483.
General Preparation of Sample before Measurements
Assay. c-di-GMP, water, buffer solution (pH 7.5), and salt solutions
were mixed, heated to 95 °C and kept at 95 °C for 5 min, and then cooled
back to room temperature and kept at room temperature for 15 min. TO
was then added to the mixture and incubated in the refrigerator at 4 °C
overnight (about 12 h).
(2) (a) Schirmer, T.; Jenal, U. Nat. Rev. Microbiol. 2009, 7, 724–735.
(
b) Galperin, M. Y. Curr. Opin. Microbiol. 2010, 13, 150–159. (c) Jonas,
€
K.; Melefors, O.; R €o mling, U. Future Microbiol. 2009, 4, 341–358. (d)
Pesavento, C.; Hengge, R. Curr. Opin. Microbiol. 2009, 12, 170–176. (e)
Krasteva, P. V.; Fong, J. C.; Shikuma, N. J.; Beyhan, S.; Navarro, M. V.;
Yildiz, F. H.; Sondermann, H. Science 2010, 327, 866–868. (f) Camilli,
A.; Bassler, B. L. Science 2006, 311, 1113–1116. (g) Tamayo, R.; Pratt,
J. T.; Camilli, A. Annu. Rev. Microbiol. 2007, 61, 131–148.
Measurement of Fluorescence. The instrument settings were
chosen as follows: λex = 508 nm (slit 5 nm), λem = 518-700 nm (slit
(3) (a) Sintim, H. O.; Smith, J. A.; Wang, J.; Nakayama, S.; Yan, L.
5
nm). The measurements were carried out at 10 °C.
Measurement of Thiazole Orange UV Spectra. The instru-
Future Med. Chem. 2010, 2, 1005–1035. (b) Beyhan, S.; Tischler, A. D.;
Camilli, A.; Yildiz, F. H. J. Bacteriol. 2006, 188, 3600–3613.
ment settings were chosen as follows: wavelength = 400-600 nm,
temperature was 10 °C. The concentration of c-di-GMP was 70 μM, TO
was 30 μM, and buffer was 10 mM Tris-HCl (pH 7.5) containing 0 mM
(4) (a) Hengge, R. Nat. Rev. Microbiol. 2009, 7, 263–273. (b)
R €o mling, U.; Gomelsky, M.; Galperin, M. Y. Mol. Microbiol. 2005,
57, 629–639.
or 1 M metal (NaCl, KCl, NH
Circular Dichroism Experiments (CD). The concentration of
c-di-GMP was 70 μM, TO was 30 μM, and buffer was 10 mM Tris-HCl
4
OAc, and LiCl).
(5) (a) Ramelot, T. A.; Yee, A.; Cort, J. R.; Semesi, A.; Arrowsmith,
C. H.; Kennedy, M. A. Proteins 2007, 66, 266–271. (b) Merighi, M.; Lee,
V. T.; Hyodo, M.; Hayakawa, Y.; Lory, S. Mol. Microbiol. 2007,
65, 876–895. (c) Freedman, J. C.; Rogers, E. A.; Kostick, J. L.; Zhang,
H.; Iyer, R.; Schwartz, I.; Marconi, R. T. FEMS Immunol. Med. Microbiol.
(pH 7.5) containing 0 mM, 50 mM, 150 mM, 250 mM, 500 mM, or 1 M
metal (NaCl, KCl, NH OAc, and LiCl). The measurement was per-
4
2
010, 58, 285–294.
formed at 10 °C. For stability study, 10, 25, and 50 °C were used as
analysis temperatures. Data pitch, 1 nm; scan speed, 50 nm/min;
response, 8 s; bandwidth, 1 nm.
Preparation of the Cell Lysate. E. Coli BL21(DE3) was used for
production of WspR(D70E), RocR from pVL1321 (a pET vector
derivative). Strains were grown up at 30 °C in Luria-Bertani (LB)
medium (3 L) with shaking, and IPTG induction was performed for 6 h.
Final OD (600 nm) was 2.8 for BL21(DE3) pVL1321-wspR, 6.8 for
BL21(DE3) pVL1321-rocR, and 6.0 BL21(DE3) pVL1321 vector con-
(6) (a) Lee, E. R.; Baker, J. L.; Weinberg, Z.; Sudarsan, N.; Breaker,
R. R. Science 2010, 329, 845–848. (b) Sudarsan, N.; Lee, E. R.; Weinberg,
Z.; Moy, R. H.; Kim, J. N.; Link, K. H.; Breaker, R. R. Science 2008,
321, 411–413. (c) Kulshina, N.; Baird, N. J.; Ferr ꢀe -D’Amar ꢀe , A. R. Nat.
Struct. Mol. Biol. 2009, 16, 1212–1217. (d) Smith, K. D.; Lipchock, S. V.;
Ames, T. D.; Wang, J.; Breaker, R. R.; Strobel, S. A. Nat. Struct. Mol. Biol.
2
009, 16, 1218–1223.
7) (a) Liu, X.; Beyhan, S.; Lim, B.; Linington, R. G.; Yildiz, F. H.
(
J. Bacteriol. 2010, 192, 4541–4552. (b) Simm, R.; Morr, M.; Remmin-
ghorst, U.; Andersson, M.; R €o mling, U. Anal. Biochem. 2009, 386, 53–58.
(c) Christen, M.; Kulasekara, H. D.; Christen, B.; Kulasekara, B. R.;
Hoffman, L. R.; Miller, S. I. Science 2010, 328, 1295–1297.
trol cells. From plating of serial dilution of the culture, each OD
9
(
600 nm) unit is equivalent to 1 ꢀ 10 colony forming units (CFU).
Cell were pelleted by centrifugation and resuspended to a final volume of
0 mL of 10 mM Tris-HCl (pH 8.0) containing 100 mM NaCl. 100 μL
4
(8) Hickman, J. W.; Harwood, C. S. Mol. Microbiol. 2008,
of 25 mg/mL Lysozyme was added to each resuspension. The resuspen-
sion was sonicated (20 s, three times) to disrupt membranes and allow
lysozyme to reach cell walls. 60% perchloric acid (final concentration
was 12%) was added to the resuspension to precipitate cellular macro-
molecules. The resuspension was incubated for 10 min on ice and
neutralized by 3 M KOH containing 0.4 M Tris and 2 M KCl. The
resuspension was centrifuged, and the collected supernatant was then
filtered by a 0.2 μm filter and 3 kD exclusion columns.
69, 376–389.
(9) (a) Biver, T.; García, B.; Leal, J. M.; Secco, F.; Turriani, E. Phys.
Chem. Chem. Phys. 2010, 12, 13309–13317. (b) Pu, F.; Huang, Z.; Ren,
J.; Qu, X. Anal. Chem. 2010, 82, 8211–8216. (c) Lubitz, I.; Zikich, D.;
Kotlyar, A. Biochemistry 2010, 49, 3567–3574. (d) Li, K.; Liu, B. Anal.
Chem. 2009, 81, 4099–4105. (e) Pei, R.; Rothman, J.; Xie, Y.; Stojanovic,
M. N. Nucleic Acids Res. 2009, 37, e59. (f) Yang, P.; De Cian, A.; Teulade-
Fichou, M. P.; Mergny, J. L.; Monchaud, D. Angew. Chem., Int. Ed. 2009,
48, 2188–2191.
(10) (a) Nygren, J.; Svanvik, N.; Kubista, M. Biopolymers 1998,
’
ASSOCIATED CONTENT
46, 39–51. (b) Karunakaran, V.; P ꢀe rez Lustres, J. L.; Zhao, L.; Ernsting,
N. P.; Seitz, O. J. Am. Chem. Soc. 2006, 128, 2954–2962.
(11) (a) Larsson, A.; Carlsson, C.; Jonsson, M.; Albinsson, B. J. Am.
Chem. Soc. 1994, 116, 8459–8465. (b) Netzel, T. L.; Nafisi, K.; Zhao, M.;
Lenhard, J. R.; Johnson, I. J. Phys. Chem. 1995, 99, 17936–17947.
S
Supporting Information. Additional figures and calcula-
b
tions. This material is available free of charge via the Internet at
http://pubs.acs.org.
(
12) Zhang, Z.; Kim, S.; Gaffney, B. L.; Jones, R. A. J. Am. Chem. Soc.
006, 128, 7015–7024.
13) Shim, J. W.; Tan, Q.; Gu, L. Q. Nucleic Acids Res. 2009,
7, 972–982.
14) (a) Sen, D.; Gilbert, W. Nature 1990, 344, 410–414. (b) Lee,
J. Y.; Yoon, J.; Kihm, H. W.; Kim, D. S. Biochemistry 2008,
7, 3389–3396. (c) Olsen, C. M.; Gmeiner, W. H.; Marky, L. A. J. Phys.
2
’
AUTHOR INFORMATION
(
3
Corresponding Author
hsintim@umd.edu
(
4
Chem. B 2006, 110, 6962–6969. (d) Hardin, C. C.; Perry, A. G.; White,
K. Biopolymers 2001, 56, 147–194. (e) Keniry, M. A. Biopolymers 2001,
56, 123–146. (f) Mariani, P.; Spinozzi, F.; Federiconi, F.; Ortore, M. G.;
Amenitsch, H.; Spindler, L.; Drevensek-Olenik, I. J. Nucleic Acids
’
ACKNOWLEDGMENT
The National Science Foundation Grant CHE 0746446
supported this work.
REFERENCES
2010, 472478. (g) Davis, J. T. Angew. Chem., Int. Ed. 2004,
43, 668–698. (h) Davis, J. T.; Spada, G. P. Chem. Soc. Rev. 2007,
36, 296–313.
’
(1) (a) Rasko, D. A.; Sperandio, V. Nat. Rev. Drug Discovery 2010,
(15) (a) Seo, Y. J.; Lee, I. J.; Kim, B. H. Bioorg. Med. Chem. Lett. 2008,
9
4
, 117–128. (b) Ng, W. L.; Bassler, B. L. Annu. Rev. Genet. 2009,
3, 197–222. (c) Singh, V.; Shi, W.; Almo, S. C.; Evans, G. B.; Furneaux,
18, 3910–3913. (b) Siddiqui-Jain, A.; Grand, C. L.; Bearss, D. J.; Hurley,
L. H. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 11593–11598. (c) Evans,
S. E.; Mendez, M. A.; Turner, K. B.; Keating, L. R.; Grimes, R. T.;
R. H.; Tyler, P. C.; Painter, G. F.; Lenz, D. H.; Mee, S.; Zheng, R.;
4
863
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864

Journal of the American Chemical Society
ARTICLE
Melchoir, S.; Szalai, V. A. J. Biol. Inorg. Chem. 2007, 12, 1235–1249.
(
d) Yang, Q.; Xiang, J.; Yang, S.; Li, Q.; Zhou, Q.; Guan, A.; Zhang, X.;
Zhang, H.; Tang, Y.; Xu, G. Nucleic Acids Res. 2010, 38, 1022–1033.
e) Gavathiotis, E.; Heald, R. A.; Stevens, M. F.; Searle, M. S. J. Mol. Biol.
(
2
003, 334, 25–36.
16) (a) Lubitz, I.; Borovok, N.; Kotlyar, A. Biochemistry 2007,
6, 12925–12929. (b) Nagesh, N.; Krishnaiah, A. J. Biochem. Biophys.
(
4
Methods 2003, 57, 65–74.
17) Ghasemi, J.; Ahmadi, Sh.; Ahmad, A. I.; Ghobadi, S. Appl.
Biochem. Biotechnol. 2008, 149, 9–22.
18) (a) Xu, Y.; Noguchi, Y.; Sugiyama, H. Bioorg. Med. Chem. 2006,
4, 5584–5591. (b) Sun, D.; Guo, K.; Rusche, J. J.; Hurley, L. H. Nucleic
(
(
1
Acids Res. 2005, 33, 6070–6080. (c) Oliviero, G.; Borbone, N.; Amato, J.;
D’Errico, S.; Galeone, A.; Piccialli, G.; Varra, M.; Mayol, L. Biopolymers
2
009, 91, 466–477. (d) Prislan, I.; Lah, J.; Vesnaver, G. J. Am. Chem. Soc.
2
008, 130, 14161–14169.
(
19) (a) Hannah, K. C.; Armitage, B. A. Acc. Chem. Res. 2004,
7, 845–853. (b) Thompson, M. Biomacromolecules 2007, 8, 3628–3633.
20) Hickman, J. W.; Tifrea, D. F.; Harwood, C. S. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 14422–14427.
21) Kulesekara, H.; Lee, V. T.; Brencic, A.; Liberati, N.; Urbach, J.;
3
(
(
Miyata, S.; Lee, D. G.; Neely, A. N.; Hyodo, M.; Hayakawa, Y.; Ausubel,
F. M.; Lory, S. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 2839–2844.
(
22) Merighi, M.; Lee, V. T.; Hyodo, M.; Hayakawa, Y.; Lory, S. Mol.
Microbiol. 2007, 65, 876–895.
23) (a) Weinhouse, H.; Sapir, S.; Amikam, D.; Shilo, Y.; Volman,
(
G.; Ohana, P.; Benziman, M. FEBS Lett. 1997, 416, 207–211. (b)
Spangler, C.; B €o hm, A.; Jenal, U.; Seifert, R.; Kaever, V. J. Microbiol.
Methods 2010, 81, 226–231.
(24) Ross, P.; Weinhouse, H.; Aloni, Y.; Michaeli, D.; Weinberger-
Ohana, P.; Mayer, R.; Braun, S.; De Vroom, E.; van der Marel, G. A.; van
Boom, J. H.; Benziman, M. Nature 1987, 325, 279–281.
(25) (a) Anslyn, E. V. J. Org. Chem. 2007, 72, 687–699. (b)
Kitamura, M.; Shabbir, S. H.; Anslyn, E. V. J. Org. Chem. 2009,
7
4, 4479–4489.
26) Witte, G.; Hartung, S.; B €u ttner, K.; Hopfner, K. P. Mol. Cell
008, 30, 167–178.
(
2
4
864
dx.doi.org/10.1021/ja1091062 |J. Am. Chem. Soc. 2011, 133, 4856–4864