U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Probe Reports from the NIH Molecular Libraries Program [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2010-.

Cover of Probe Reports from the NIH Molecular Libraries Program

Probe Reports from the NIH Molecular Libraries Program [Internet].

Show details

Discovery of Two, Structurally Distinct Agonists of Vibrio cholerae Quorum Sensing Acting via the CqsS Membrane Receptor

, , , , , , , , , , , , , , , , , , , and .

Author Information and Affiliations

Received: ; Last Update: September 3, 2013.

Quorum sensing (QS) is a process of bacterial cell-to-cell communication that relies upon recognition of extracellular signaling molecules called autoinducers. QS allows bacteria to synchronize their behavior in response to changes in the population density and species composition of the proximal bacterial community. Known behaviors regulated by QS include bioluminescence, sporulation, virulence factor production, and biofilm formation. We carried out a high throughput screen (HTS) to identify small molecules that modulate QS in a modified V. cholerae strain carrying a luciferase operon; activation of the quorum pathway is accompanied by light production. 352,083 compounds from the NIH-MLPCN compound library were evaluated. Potential QS modulators were characterized via additional bacterium-based epistatic assays to elucidate their mode of action. We report the discovery and development of two, structurally distinct, small-molecule probes (ML343 and ML344) shown to be agonists of Vibrio cholerae CqsS, a transmembrane receptor. ML343 and ML344 should greatly expand the general understanding of how QS membrane receptors productively interact with ligands and how they relay information from the external environment. In addition these compounds could lead to the development of antibacterial drugs designed to interfere with QS which could have enormous ramifications for improving human health.

Assigned Assay Grant No: R03 MH094166-01

Screening Center Name & PI: Broad institute Probe Development Center, Stuart L. Schreiber, PhD

Chemistry Center Name & PI: Broad institute Probe Development Center, Stuart L. Schreiber, PhD

Assay Submitter & Institution: Bonnie L. Bassler, PhD., Princeton University, Princeton, NJ

PubChem Summary Bioassay Identifier (AID): 588521

Probe Structure & Characteristics

ML344.

ML344

PubChem
CID		TargetsEC50 (μM)
[SID, AID]		Anti-TargetIC50 (μM)
[SID, AID]		Fold Selective*
CID 4727493V. cholerae BH15780.2
[SID 152199102, AID 651816]		HeLa
cytotoxicity		>30
[SID 152199102, AID 651864]		150x
V. cholerae BH1651>30
[SID 152199102, AID 651847]		N/A
V. cholerae DH231>30
[SID 152199102, AID 651808]		N/A
V. cholerae WN11030.06
[SID 152199102, AID 651841]		500x
*

Selectivity = Anti-target IC50/Target EC50

ML343.

ML343

PubChem
CID		TargetsEC50 (μM)
[SID, AID]		Anti-TargetIC50 (μM)
[SID, AID]		Fold Selective*
CID 57525036V. cholerae BH15780.5
[SID 137282727,AID 651816]		HeLa
cytotoxicity		>30
[SID 137282727, AID 651864]		60x
V. cholerae BH1651>30
[SID 137282727,AID 651847		N/A
V. cholerae DH231>30
[SID 137282727, AID 651808]		N/A
V. cholerae WN11030.3
[SID 137282727, AID 651841]		100x
*

Selectivity = Anti-target IC50/Target EC50

1. Recommendations for the Scientific Use of these Probes

Discovery of quorum sensing (QS) receptor agonists will permit more precise control of Vibrio cholerae QS through the perturbation of distinct steps within the signaling cascade. This, in turn, would enable elucidation of the underlying principles governing ligand-receptor interactions utilized by transmembrane histidine kinases, such as the V. cholerae QS membrane receptor CqsS. Upon ligand binding, CqsS undergoes a number of conformational changes that are critical to its numerous functions. The newly identified probes may shed light on the species-specific nature of QS ligands and possibly provide a clearer understanding of the inherent kinase and phosphatase activity of CqsS [1].

Crystallography efforts to visualize the CqsS transmembrane domain are underway, and these high-quality probes will be evaluated for possible co-crystallization. Concurrently, directed mutagenesis will be applied to understand how these probes interact with CqsS. Since there is some capacity for interspecies communication amongst the different Vibrio species [2], ML343 and ML344 can be tested for QS agonism with other Vibrio species. Identification of CqsS-agonist binding motifs will establish a framework for the development of antibacterial drugs designed to interfere with quorum sensing. These probes can also test the utility of quorum pathway activation as a means of treating cholera via reduction of pathogen’s virulence in vivo. There are several animal models available with which such studies may be conducted [3–5]. If such an approach is efficacious, these potential therapeutics, operating in a completely unexploited target space, would have enormous ramifications upon improving human health.

2. Materials and Methods

See subsections for a description of the materials and methods used for each assay.

Materials and Reagents

  • CellTiter-Glo® Luminescent Cell Viability Assay was purchased from Promega (Catalog No. G7573; Madison, WI)

Bacterial strains & Cell Lines

The following cell lines were used in this study:

  • Vibrio cholerae BH1578; a genetically modified strain lacking LuxS and CqsA autoinducer synthases that was provided by the Bassler lab. This strain was used in the primary assay.
  • Vibrio cholerae BH1651; a genetically modified strain with constitutively active LuxO that was provided by the Bassler lab.
  • Vibrio cholerae DH231; a genetically modified strain lacking the CqsS receptor that was provided by the Bassler lab.
  • Vibrio cholerae WN1103; a genetically modified strain lacking the LuxQ receptor that was provided by the Bassler lab.
  • HeLa obtained from ATCC (Catalog Number CCL-2; Manassas, VA) is a human epithelial adenocarcinoma cell line used for mammalian cytotoxicity profiling

Note: All of the V. cholerae strains also carry the heterologous V. harveyi luxCDABE luciferase operon

Table 1Strains of V. cholera used in assays

StrainGenotypeReporterSuggested Mechanism for Inducers
BH1578ΔluxS ΔcqsAV. harveyi luciferase operonMultiple Mechanisms possible
BH1651luxOD47EV. harveyi luciferase operonLuxO inhibitor or downstream
DH231ΔluxS ΔcqsSV. harveyi luciferase operonLuxPQ agonist or downstream
WN1103ΔluxQ ΔcqsAV. harveyi luciferase operonCqsS agonist or LuxO inhibition

2.1. Assays

2.1.1. Primary assay – V. cholerae BH1578 bioluminescence inducer assay (AID 588346, AID 602243, AID 624270, AID 651809, AID 651816)

A modified strain of Vibrio cholerae used in this assay uses light production to indicate quorum sensing induction. Vibrio cholerae is not naturally bioluminescent but the closely related species Vibrio harveyi produces light when the population is at a high density (i.e. a quorum is sensed). The heterologous Vibrio harveyi luciferase operon (luxCDABE) was cloned into the Vibrio cholerae C6706 O1 El Tor bacterial strain on the pBB1 cosmid to create a bioluminescence assay strain. This operon is activated by the endogenous V. cholerae quorum sensing pathway [2]. The BH1578 strain is a cqsA, luxS double mutant that lacks both autoinducer synthases (CAI-1 and AI-2). BH1578 does not generate light in the absence of exogenous autoinducers but bioluminescence can be stimulated up to 10,000-fold by adding 1 μM (saturating) CAI-1. On day 0, a colony of bacteria was picked into 50 mL Luria Broth with 10 μg/mL tetracycline and cultured overnight at 30 °C. On day 1, bacterial density was determined by spectrometry, and the OD600 was adjusted to 0.3. 20 μL of Luria Broth with 10 μg/mL tetracycline was added per well into white, opaque 384 well plates. Compounds and controls were added by pin transfer method. The HTS used 150 nL of compound for screening at 20 μM, and the retest assays used 100 nL of compound. After 6 hours, the luminescence signal was determined with the Perkin-Elmer EnVision plate reader. In addition to luminescence, the confluency of each well was measured at an absorbance of 600 nM. Primary HTS data were analyzed in Genedata Screener Assay Analyzer. All values were normalized against DMSO treated samples and the positive control (1 μM CAI-1, CID 24892809). For the HTS, the average of two replicates was used to rank order activity and to choose compounds for retests. For dose studies, percent (%) activity was determined for each concentration, and the concentration response curves (CRCs) were generated with Genedata Screener’s Condoseo.

2.1.2. Secondary assay – HeLa cytotoxicity assay (AID 624140, AID 651774, AID 651864)

HeLa cells were treated with compounds for 24 hours, and then cell viability was measured using the CellTiter-Glo Assay (Promega), a luciferase-based reagent that measures cellular ATP levels. The compounds were tested at different concentrations to determine IC50 values. Compounds that were inactive (IC50 ≥ 30 μM) in this assay were considered for probe development. Data were normalized against DMSO in Genedata Screener’s Assay Analyzer. Curves were generated with Genedata Screener’s Condoseo and showed percent (%) activity for the individual doses.

2.1.3. Secondary assay – V. cholerae BH1651 LuxO inhibitor assay (AID 624254, AID 624269, AID 651847)

A modified strain of Vibrio cholerae used in this assay uses light production to indicate quorum sensing induction. Vibrio cholerae is not naturally bioluminescent but the closely related species Vibrio harveyi produces light when the population is at a high density (i.e. a quorum is sensed). The heterologous Vibrio harveyi luciferase operon (luxCDABE) was cloned into the Vibrio cholerae C6706 O1 El Tor bacterial strain on the pBB1 cosmid to create a bioluminescence assay strain. This operon is activated by the endogenous V. cholerae quorum sensing pathway [2]. The BH1651 strain is a luxOD47E mutant where LuxOD47E mimics the behavior of phosphorylated LuxO, rendering LuxOD47E constitutively active within the QS pathway. BH1651 does not generate light but any compound that inhibits LuxO or works downstream of LuxO will induce light production. On day 0, a colony of bacteria was picked into 50 mL Luria Broth with 10 μg/mL tetracycline and cultured overnight at 30 °C. On day 1, bacterial density was determined by spectrometry, and the OD600 was adjusted to 0.3. 20 μL of Luria Broth with 10 μg/mL tetracycline was added per well into white, opaque 384 well plates. Compounds and controls were added by pin transfer method. After 6 hours, the luminescence signal was determined with the Perkin-Elmer EnVision plate reader. In addition to luminescence, the confluency of each well was measured at an absorbance of 600 nM. Data were analyzed in Genedata Screener Assay Analyzer. All values were normalized against DMSO treated samples. Percent (%) activity was determined for each concentration, and the concentration response curves (CRCs) were generated with Genedata Screener’s Condoseo.

2.1.4. Secondary assay – V. cholerae DH231 sensor mechanism assay (AID 624281, AID 651808)

A modified strain of Vibrio cholerae used in this assay uses light production to indicate quorum sensing induction. Vibrio cholerae is not naturally bioluminescent but the closely related species Vibrio harveyi produces light when the population is at a high density (i.e. a quorum is sensed). The heterologous Vibrio harveyi luciferase operon (luxCDABE) was cloned into the Vibrio cholerae C6706 O1 El Tor bacterial strain on the pBB1 cosmid to create a bioluminescence assay strain. This operon is activated by the endogenous V. cholerae quorum sensing pathway [2]. The DH231 strain is a luxS and cqsS double deletion mutant. DH231 does not generate light but any compound that agonizes the receptor LuxQ will induce light production, and CqsS agonists will have no activity in this assay. On day 0, a colony of bacteria was picked into 50 mL Luria Broth with 10 μg/mL tetracycline and cultured overnight at 30 °C. On day 1, bacterial density was determined by spectrometry, and the OD600 was adjusted to 0.3. 20 μL of Luria Broth with 10 μg/mL tetracycline was added per well into white, opaque 384 well plates. Compounds and controls were added by pin transfer method. After 6 hours, the luminescence signal was determined with the Perkin-Elmer EnVision plate reader. In addition to luminescence, the confluency of each well was measured at an absorbance of 600 nM. Data were analyzed in Genedata Screener Assay Analyzer. All values were normalized against DMSO treated samples. Percent (%) activity was determined for each concentration, and the concentration response curves (CRCs) were generated with Genedata Screener’s Condoseo.

2.1.5. Secondary assay – V. cholerae WN1103 sensor mechanism assay (AID 624275, AID 651841)

A modified strain of Vibrio cholerae used in this assay uses light production to indicate quorum sensing induction. Vibrio cholerae is not naturally bioluminescent but the closely related species Vibrio harveyi produces light when the population is at a high density (i.e. a quorum is sensed). The heterologous Vibrio harveyi luciferase operon (luxCDABE) was cloned into the Vibrio cholerae C6706 O1 El Tor bacterial strain on the pBB1 cosmid to create a bioluminescence assay strain. This operon is activated by the endogenous V. cholerae quorum sensing pathway [2]. The WN1103 strain is a luxQ and cqsA double deletion mutant. DH231 does not generate light but any compound that agonizes the CqsS receptor will induce light production and LuxQ agonists will have no activity in this assay. On day 0, a colony of bacteria was picked into 50 mL Luria Broth with 10 μg/mL tetracycline and cultured overnight at 30 °C. On day 1, bacterial density was determined by spectrometry, and the OD600 was adjusted to 0.3. 20 μL of Luria Broth with 10 μg/mL tetracycline was added per well into white, opaque 384 well plates. Compounds and controls were added by pin transfer method. After 6 hours, the luminescence signal was determined with the Perkin-Elmer EnVision plate reader. In addition to luminescence, the confluency of each well was measured at an absorbance of 600 nM. Data were analyzed in Genedata Screener Assay Analyzer. All values were normalized against DMSO and 1 μM CAI-1 treated samples. Percent (%) activity was determined for each concentration, and the concentration response curves (CRCs) were generated with Genedata Screener’s Condoseo.

2.1.6. Secondary assay – CqsS in vitro auto-phosphorylation assay (AID 651905)

Kinase assays were performed with inverted E. coli membranes expressing recombinant Vibrio cholerae wildtype CqsS. Reactions were carried out in phosphorylation buffer (50 mM Tris pH 8.0, 100 mM KCl, 5 mM MgCl2, and 10% (v/v) glycerol) and were initiated with the addition of 100 μM ATP and 2 uCi [γ-32ATP] (stock 3000 Ci mmol−1, Perkin Elmer). DMSO, CAI-1 or CqS agonists were added for a final concentration of 100 μM ten minutes prior to ATP addition. Reactions were incubated at room temperature and terminated after two minutes with SDS-PAGE loading buffer. Reaction products were separated using SDS-PAGE and the gels were dried at 80 °C on filter paper under a vacuum. The gels were exposed to a phosphoscreen for two hours and subsequently analyzed using Typhoon 9400 scanner and ImageQuant software [10].

2.2. Probe Chemical Characterization

After preparation as described in Section 2.3, the probes ML344 and ML343 were analyzed by UPLC, 1H and 13C NMR spectroscopy, and high-resolution mass spectrometry. The data obtained from NMR and mass spectroscopy are consistent with the structure of the probe, and UPLC indicates an isolated purity of greater than 99% for both probes. The 1,5-substitution pattern of ML344 was unambiguously established by 2D-NMR spectroscopy. The solubility of probe ML344 was experimentally determined to be >100 μM in phosphate buffered saline (PBS) with 1% (v/v) DMSO. Under identical conditions, the solubility of probe ML343 was measured to be 1.4 μM. The solubility of ML343 was also measured using the Luria Broth (LB) media utilized for the primary assay. Table 2 summarizes the solubility of ML343 and several analogs in both PBS and LB media. In general, LB media is a better solvent than PBS for the biphenyl compounds.

Table 2. Solubility of ML343 and Analogs in PBS and LB Assay Media.

Table 2

Solubility of ML343 and Analogs in PBS and LB Assay Media.

Both probes are stable in PBS solution (>99% remaining after a 48-hour incubation at 23 °C). The data from the PBS stability assay is provided in Figure 1. The probes are also stable to human and murine plasma exposure, with greater than 84% remaining after incubation at 37 °C for 5 hours. No significant reaction with glutathione (GSH) was observed for either compound. Table 3 summarizes the various stability assays performed. The physical properties of probes ML344 and ML343 are summarized in Table 4.

Figure 1. Stability of Probes ML344 and ML343 in PBS Buffer (pH 7.4, 23°C).

Figure 1

Stability of Probes ML344 and ML343 in PBS Buffer (pH 7.4, 23°C).

Table 3. Plasma Stability and Plasma Binding of Probes ML344 and ML343.

Table 3

Plasma Stability and Plasma Binding of Probes ML344 and ML343.

Table 4. Summary of Probe Properties Computed for Probes ML344 and ML343.

Table 4

Summary of Probe Properties Computed for Probes ML344 and ML343.

2.3. Probe Preparation

Probe 1 (ML344) and most associated analogs were prepared according to the route outlined below in Scheme 1. 4-Isopropylbenzaldehyde (2) was stirred with 5-aminotetrazole (3) in the presence of sodium cyanoborohydride and acetic acid to give the coupled amine 4. The tetrazole adduct was then reacted with ethyl iodide and potassium carbonate to give a mixture of regioisomers. Column chromatography removed the undesired 2,5-regioisomer to afford the probe ML344.

Scheme 1. Synthesis of Probe ML344.

Scheme 1

Synthesis of Probe ML344.

Probe 2 (ML343) and its analogs were prepared by amide coupling of the appropriate phenyl acetic acids with the corresponding amine (Scheme 2). 4-Biphenyl acetic acid 6 was coupled to cyclobutylamine 7 with the assistance of 4-(N,N-dimethylamino)pyridine and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide. ML343 was isolated by selective precipitation from the reaction mixture.

Scheme 2. Synthesis of Probe ML343.

Scheme 2

Synthesis of Probe ML343.

3. Results

3.1. Dose Response Curves for Probe

Figure 2. ML344 and ML343 were tested across a range of concentrations up to 35 μM in the primary assay and a HeLa cytotoxicity assay.

Figure 2ML344 and ML343 were tested across a range of concentrations up to 35 μM in the primary assay and a HeLa cytotoxicity assay

Concentration response curves were generated with Genedata Screener Condeseo and show normalized percent activity for the individual doses. ML344 BH1578 assay (AID 651816), IC50 = 0.17 μM (A); ML343 BH1578 assay (AID 651816), IC50 = 0.56 μM (B); ML344 HeLa CellTiter-Glo (AID 651864), IC50 ≥ 35 μM (C); ML343 HeLa CellTiter-Glo (AID 651864), IC50 ≥ 35 μM (D). O = replicate 1, Δ = replicate 2.

3.2. Cellular Activity

The primary assay was performed with whole bacterial cells. A secondary screen evaluating toxicity utilized mammalian HeLa cells, and additional phenotypic assays evaluated the activation of QS pathways in live Vibrio cholerae mutants. An overview of these assays is provided above in Section 2.1. The probes ML344 and ML343 satisfy the cellular activity criteria specified for this project.

3.3. Profiling Assays

Both probes were screened for binding to a broad panel of receptors and ion channels using the Eurofins Panlabs’ LeadProfilingScreen (Catalog No. 68). The results of this screen are summarized below in Figures 3 and 4. ML344 shows less than 50% inhibition of all 67 targets evaluated (Figure 3). Its strongest activity is against the melatonin MT1 receptor, which exhibits 44% inhibition when treated with 10 μM ML344. Similarly, ML343 does not show significant activity in this screening panel (Figure 4). The serotonin 5-HT2B receptor is inhibited by 45% using 10 μM ML343, and this represents the most potent interaction between ML343 and the 67 targets screened.

Figure 3. Probe ML344 was evaluated for inhibition of 67 mammalian membrane receptors, ion channels, and membrane transporters.

Figure 3

Probe ML344 was evaluated for inhibition of 67 mammalian membrane receptors, ion channels, and membrane transporters. A test concentration of 10 μM was used, and each experiment was performed in duplicate (n=1).

Figure 4. Probe ML343 was evaluated for inhibition of 67 mammalian membrane receptors, ion channels, and membrane transporters.

Figure 4

Probe ML343 was evaluated for inhibition of 67 mammalian membrane receptors, ion channels, and membrane transporters. A test concentration of 10 μM was used, and each experiment was performed in duplicate (n=1).

4. Discussion

4.1. Comparison to Existing Art and How the New Probe is an Improvement

Currently, the only known agonists of CqsS are the natural ligand CAI-1 and structurally related analogs [2,13]. While ML344 and ML343 are less potent than CAI-1 in the primary assay (relative to EC50 values), they elicit a stronger response from BH1578 cells as indicated by the stronger observed luminescence (Table 5). The SAR associated with CAI-1 is steep [2,13], and not much structural information is known regarding its CqsS binding pocket. Additionally, the underlying structure of CAI-1 promotes micelle formation that complicates its usage in vitro and leads to increased variability in the cell-based assays. In contrast, the distinct architectures of ML344 and ML343 do not suggest that either compound will form micelles in vitro. It is anticipated that these probes will provide valuable information regarding available binding pockets of CqsS and guide the development of the next generation of quorum sensing focused therapeutics.

Table 5. Comparison of Probes ML344 and ML343 to Natural CqsS Ligand CAI-1.

Table 5

Comparison of Probes ML344 and ML343 to Natural CqsS Ligand CAI-1.

5. References

1.
Ng WL, Wei Y, Perez LJ, Cong J, Long T, Koch M, Semmelhack MF, Wingreen NS, Bassler BL. Probing bacterial transmembrane histidine kinase receptor-ligand interactions with natural and synthetic molecules. Proc Natl Acad Sci U S A. 2010 Mar 23;107(12):5575–80. Epub 2010 Mar 8. [PMC free article: PMC2851778] [PubMed: 20212168]
2.
Ng WL, Perez LJ, Wei Y, Kraml C, Semmelhack MF, Bassler BL. Signal production and detection specificity in Vibrio CqsA/CqsS quorum sensing systems. Mol Microbiol. 2011 Mar;79(6):1407–17. Epub 2011 Jan 26. [PMC free article: PMC3285556] [PubMed: 21219472] [CrossRef]
3.
Dutta NK, Habbu MK. Experimental cholera in infant rabbits: a method for chemotherapeutic investigation. Br J Pharmacol Chemother. 1955 Jun;10(2):153–9. [PMC free article: PMC1509487] [PubMed: 14389652]
4.
Aziz KM, Mohsin AK, Hare WK, Phillips RA. Using the rat as a cholera “model” Nature. 1968 Nov 23;220(5169):814–5. [PubMed: 5725345]
5.
Klose KE. The suckling mouse model of cholera. Trends Microbiol. 2000 Apr;8(4):189–91. Review. [PubMed: 10754579]
6.
Casino P, Rubio V, Marina A. The mechanism of signal transduction by two-component systems. Curr Opin Struct Biol. 2010 Dec;20(6):763–71. Epub 2010 Oct 13. Review. [PubMed: 20951027]
7.
Rutherford ST, Bassler BL. Bacterial quorum sensing: Its role in virulence and possibilities for its control. Cold Spring Harb. Perspect. Med. 2012;2:a012427. [PMC free article: PMC3543102] [PubMed: 23125205]
8.
Miller MB, Skorupski K, Lenz DH, Taylor RK, Bassler BL. Parallel quorum sensing systems converge to regulate virulence in Vibrio cholerae. Cell. 2002 Aug 9;110(3):303–14. [PubMed: 12176318]
9.
Higgins DA, Pomianek ME, Kraml CM, Taylor RK, Semmelhack MF, Bassler BL. The major Vibrio cholerae autoinducer and its role in virulence factor production. Nature. 2007 Dec 6;450(7171):883–6. Epub 2007 Nov 14. [PubMed: 18004304]
10.
Wei Y, Ng WL, Cong J, Bassler BL. Ligand and antagonist driven regulation of the Vibrio cholerae quorum sensing receptor CqsS. Mol Microbiol. 2012 Mar;83(6):1095–108. Epub 2012 Feb 14. [PMC free article: PMC3310172] [PubMed: 22295878] [CrossRef]
11.
Lenz DH, Mok KC, Linney BN, Kulkarni RV, Wingreen NS, Bassler BL. The small chaperone Hfq and multiple small RNAs control quorum sensing in Vibrio harveyi and Vibrio cholerae. Cell. 2004;118:69–82. [PubMed: 15242645]
12.
Zhu J, Mekalanos JJ. Quorum sensing-dependent biofilms enhance colonization in Vibrio cholerae. Dev Cell. 2003 Oct;5(4):647–56. [PubMed: 14536065]
13.
Perez LJ, Ng WL, Marano P, Brook K, Bassler BL, Semmelhack MF. Role of the CAI-1 Fatty Acid tail in the Vibrio cholerae quorum sensing response. J Med Chem. 2012 Nov 8; [Epub ahead of print] [PMC free article: PMC3798069] [PubMed: 23092313]

Views

  • PubReader
  • Print View
  • Cite this Page

In this Page

Related information

Similar articles in PubMed

See reviews...See all...

Recent Activity

ClearTurn OffTurn On

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...