Quorum-sensing effector pyocyanin but not farnesol & acyl homoserine lactone exhibit antibacterial activity.

Sir,

Quorum sensing (QS) is a mechanism of cell-to-cell communication in microbes, involving cell-density-dependent production of small signalling molecules that regulate a wide array of traits including cell morphology, motility, expression of virulence factors and drug resistance1234. Amongst bacteria, the major QS signals are linear/cyclic oligopeptides (in Gram-positive bacteria), acyl homoserine lactones (in Gram-negative bacteria), furanone derivatives (in both) and others such as quinolone, phenazines, indole and cis-2-decenoic acid in selected species12. In fungi, QS signals have been well characterized for Candida spp., and primarily include farnesol and tyrosol that reciprocally regulate the yeast-to-hyphal transition3. A lipoxygenase-derived mediator and pantothenic acid derivative have also been described in Aspergillus flavus and Cryptococcus neoformans, respectively34.

Over the past several years, QS-based interference has emerged as an important strategy against microbial infections123. This therapeutic approach involves the use of certain natural or synthetic compounds1235, and even repurposed drugs6, to inhibit the QS and thereby the associated virulence mechanisms. Alternatively, many of the QS effector molecules are important mediators of both intra- and interspecies microbial communication123, and themselves display antimicrobial or antibiofilm activity against unrelated species78910111213141516. These include farnesol, which is a sesquiterpene alcohol produced by Candida spp. and is also a component of many essential oils3; long-chain (C10 to C14) acyl homoserine lactones are synthesized by Gram-negative bacteria12 and pyocyanin, a blue-green N-methyl-1-hydroxyphenazine pigment that is a terminal signalling factor in the quorum-sensing network of Pseudomonas aeruginosa7. While farnesol reportedly exhibits antimicrobial function by compromising the cell membrane integrity, biofilm formation or virulence-factor production9101112, long-chain acyl homoserine lactones may disrupt the bacterial cell wall and virulence gene expression131415, and pyocyanin interferes in the electron transport chain during respiration, leading to the production of toxic oxygen radicals7816. Although the inhibitory activity of these agents has been explored in the literature against many Gram-positive and Gram-negative bacterial pathogens78910111213141516, majority of the studies have tested one or few strains only, especially the standard strains. Considering the emergent drug resistance crisis, the present work aimed to evaluate the antimicrobial potential of these three compounds against culture-type strains and multiple clinical isolates of four common nosocomial pathogens that are frequently antibiotic resistant: Staphylococcus epidermidis, Staphylococcus aureus, Escherichia coli and Klebsiella pneumoniae.

The study was undertaken in the department of Microbial Biotechnology, Panjab University, Chandigarh, India, from January 2016 to May 2017. S. epidermidis (ATCC 35984), S. aureus (ATCC 29213), E. coli (ATCC 25922) and K. pneumoniae (ATCC 700603) were used as the culture-type strains. Clinical isolates of S. epidermidis (n=14), S. aureus (n=10), E. coli (n=3) and K. pneumoniae (n=11) (Table I) were from our laboratory collection and had been obtained from catheter-related bloodstream infections in a previous study17. The bacterial identification was performed by biochemical testing and subsequently confirmed by mass spectrometry (MALDI Biotyper, Bruker Daltonics, Bremen, Germany). The strains were preserved in the brain–heart infusion broth (Hi-Media, Mumbai) with 15 per cent glycerol (SRL Pvt, Ltd., Mumbai) at −20°C. Antibiotic susceptibility testing was performed as per the guidelines of the Clinical and Laboratory Standards Institute (CLSI)1819 and the zones of growth inhibition (ZOI) were interpreted accordingly.

Table I

Antibiotic susceptibility of Staphylococcus epidermidis, Staphylococcus aureus, Escherichia coli and Klebsiella pneumoniae culture-type strains and clinical isolates used

Bacterial species	Antibiotic susceptibility
	Cefotaxime (30 µg)	Imipenem (10 µg)	Oxacillin#	Amikacin (30 µg)	Ciprofloxacin (5 µg)	Chloramphenicol (30 µg)	Vancomycin#
S. epidermidis
CI-6*	S	ND	S	S	S	S	S
ATCC 12228	S	ND	S	R	R	S	S
ATCC 35984, CI-1, CI-2, CI-3, CI-5, CI-7, CI-10, CI-12	R	ND	R	S	S	S	S
CI-4, CI-8	R	ND	R	S	I	S	S
CI-9, CI-13, CI-14	R	ND	R	S	R	S	S
CI-11	R	ND	R	S	R	R	S
S. aureus
ATCC 29213, ATCC 25923, CI-1, CI-5	S	ND	S	S	S	S	S
CI-9*	S	ND	S	S	I	S	S
CI-10*	S	ND	S	S	R	S	S
CI-8	S	ND	S	S	I	R	S
CI-2, CI-3	R	ND	R	S	I	S	S
CI-4, CI-7	R	ND	R	S	R	S	S
CI-6	R	ND	R	S	R	R	S
E. coli
ATCC 25922	S	S	ND	S	S	S	ND
CI-1* , CI-2, CI-3*	R	S	ND	S	R	S	ND
K. pneumoniae
CI-2, CI-4	S	S	ND	S	S	S	ND
ATCC 700603	I	S	ND	S	S	I	ND
CI-9*	R	S	ND	S	I	S	ND
CI-10	R	S	ND	S	R	I	ND
CI-6, CI-11	R	S	ND	R	R	S	ND
CI-3	R	S	ND	R	R	I	ND
CI-7, CI-8	R	R	ND	R	S	S	ND
CI-1, CI-5	R	R	ND	R	R	R	ND

The antibiotic susceptibility of these isolates has been reported previously in Singh et al17, 2016. Cefotaxime, imipenem, amikacin, ciprofloxacin and chloramphenicol were tested by disk diffusion. Vancomycin and oxacillin in staphylococci were tested by vancomycin agar screen test (brain heart infusion agar with 6 µg/ml vancomycin) and oxacillin agar screen test (Mueller Hinton agar with 4% NaCl and 6 µg/ml oxacillin), respectively. S, susceptible; I, intermediate; R, resistant; ND, not done (imipenem for S. epidermidis and S. aureus; vancomycin and oxacillin for E. coli and K. pneumoniae)

The stock solutions of pyocyanin and N-(3-oxododecanoyl)-L-homoserine lactone (3-Oxo-C12-HSL; Sigma-Aldrich, Bengaluru) were prepared in dimethyl sulfoxide (SRL Pvt., Ltd., Mumbai), and that of trans, trans-farnesol (Sigma-Aldrich, Bengaluru) was prepared in methanol (Merck Millipore, Bengaluru) as per the manufacturer’s instructions. The antimicrobial activity of these agents was determined by disk-diffusion assay, according to the CLSI guidelines given for antibiotics1819. Disk contents ranging from 5 to 30 µg (pyocyanin) or up to 100 µg (farnesol and 3-Oxo-C12-HSL) were used. The diameters of ZOI were measured up to the nearest whole millimetres. Control set-ups comprising dimethyl sulphoxide and methanol were set up in parallel to confirm that the observed antibacterial activity was indeed resulting from the test agent. The ZOIs of test compounds were interpreted for activity according to the method of Quinto and Santos20, wherein the zone sizes <10, 10-13, 14-19 and >19 were considered as inactive, partially active, active and very active, respectively. In addition, the minimum inhibitory concentrations (MICs) of all the three compounds against the culture-type strains were determined by broth microdilution method1819. The experiments were performed in triplicate and repeated thrice to confirm the observations. Statistical significance was determined using analysis of variance test for normal distribution. The data were represented as mean±standard deviation (SD).

Pyocyanin exhibited a significant, dose-dependent antibacterial activity against culture-type strains of S. epidermidis, S. aureus and E. coli by disk-diffusion assay (Table II). The MICs of pyocyanin were found to be four, eight and 32 μg/ml for S. epidermidis, S. aureus and E. coli, respectively, by broth microdilution method. Notably, pyocyanin was also observed to be highly active against all the tested clinical isolates of these three species (Table III), including the isolates resistant to antibiotic classes such as β-lactams, fluoroquinolone (ciprofloxacin) and phenicol (chloramphenicol), as well as the multidrug-resistant isolates such as methicillin-resistant staphylococci (Tables I and III). The average ZOI diameters in clinical isolates of S. epidermidis, S. aureus and E. coli were 29±3, 24±1 and 18±1 mm (mean±SD), respectively, after exposure to 30 μg pyocyanin by disk diffusion (Table III). A statistical comparison revealed that this inhibitory effect was highest against S. epidermidis, followed by S. aureus and then E. coli (P<0.001). It should be noted, however, that only a limited number of E. coli cultures were tested in the present study.

Table II

Antibacterial activity of pyocyanin against culture-type strains of S. epidermidis, S. aureus, E. coli and K. pneumoniae, tested using disk-diffusion assay with pyocyanin content ranging from 5 to 30 µg

Pyocyanin disk content (µg)	ZOI (mm; mean±SD)
	S. epidermidis (ATCC 35984)	S. aureus (ATCC 29213)	E. coli (ATCC 25922)	K. pneumoniae (ATCC 700603)
5	12±1	14±2	11±3	0
10	18±1	18±1	17±2	0
20	24±1	21±1	19±4	9±2
30	29±2	22±1	24±2	9±1

The results are represented as the diameters of the ZOI, rounded up to the nearest whole millimetres. ZOI, zones of growth inhibition; SD, standard deviation

Table III

Antibacterial activity of pyocyanin against clinical isolates of S. epidermidis, S. aureus, E. coli and K. pneumoniae, tested using disk-diffusion assay with pyocyanin content of 30 µg

Bacterial strain/isolate	ZOI with 30-µg pyocyanin (mm; mean±SD)
S. epidermidis
CI-1	31±1
CI-2	31±2
CI-3	28±1
CI-4	31±6
CI-5	31±1
CI-6	31±1
CI-7	29±1
CI-8	33±4
CI-9	25±5
CI-10	26±1
CI-11	27±1
CI-12	20±4
CI-13	28±2
CI-14	28±2
S. aureus
CI-1	29±2
CI-2	22±1
CI-3	24±1
CI-4	25±2
CI-5	29±2
CI-6	26±1
CI-7	24±1
CI-8	23±1
CI-9	19±1
CI-10	24±4
E. coli
CI-1	18±4
CI-2	19±1
CI-3	18±3
K. pneumoniae
CI-1	10±0
CI-2	10±0
CI-3	10±0
CI-4	10±0
CI-5	18±1
CI-6	9±1
CI-7	10±1
CI-8	16±1
CI-9	10±1
CI-10	8±1
CI-11	9±1

The results are represented as the diameters of the ZOI, rounded up to the nearest whole millimetres

To assess if pyocyanin was auto-inhibitory for the producing species, its activity was also tested against a P. aeruginosa strain (ATCC 27583), which was found to be resistant to its action. These results thus affirm the role of pyocyanin in conferring a competitive advantage to P. aeruginosa during interspecies interaction in human infections and corroborate with some of the previous publications wherein P. aeruginosa cells or purified pyocyanin were reported to exhibit antagonistic behaviour against staphylococci and E. coli7816. Other studies, however, indicate that the competitive advantage of P. aeruginosa over staphylococci is independent of pyocyanin2122 and rather mediated by certain polysaccharides23.

In comparison to the other species, K. pneumoniae was inhibited by pyocyanin to a lesser extent (Tables II and III). The average ZOI diameter in culture-type strain and clinical isolates was 11±3 mm after exposure to 30 μg pyocyanin. Increasing the dose of pyocyanin up to 50 or 100 μg did not improve the inhibition zone size. Furthermore, the different isolates of K. pneumoniae tested varied in their susceptibility to pyocyanin. The compound was inactive against four, partially active against six and active against two of the K. pneumoniae tested, as observed by disk diffusion. The MIC of pyocyanin against K. pneumoniae ATCC 700603 was found to be >64 μg/ml. Contrary to the generally accepted notion, this decreased sensitivity of K. pneumoniae was not solely the result of its Gram-negative nature, as all the E. coli isolates tested were observed to be sensitive to pyocyanin. Instead, it could have resulted from the thick capsule that K. pneumoniae harbours, which limits the cell permeability. Alternatively, an increased expression of enzymes such as catalase and superoxide dismutase, which inactivate the reactive oxygen intermediates, may contribute to the decreased pyocyanin efficacy7. This reduced susceptibility to pyocyanin may contribute to the ability of K. pneumoniae to co-exist with P. aeruginosa during mixed-species infections24. In contrast, S. aureus, being sensitive to pyocyanin and other respiratory inhibitors produced by P. aeruginosa, often exists in temporal or spatial segregation, or may form electron-deficient small-colony variants to counter these respiratory toxins during co-existance2526.

Neither 3-Oxo-C12-HSL nor farnesol demonstrated any antibacterial activity against S. epidermidis, S. aureus, E. coli and K. pneumoniae even when tested up to a dose of 100 μg by disk diffusion (ZOI, 0 mm) or up to 64 μg/ml by broth microdilution in the present study. These results differ from many of the previous reports, which describe the antibacterial properties of farnesol and long-chain acyl homoserine lactones11131415. This is likely due to differences in the experimental protocol employed. In the present work, antimicrobial susceptibility testing was performed using the assay parameters recommended by CLSI for clinical testing so as to obtain results that are physiologically relevant. The earlier studies have followed variable methodologies, including the culture media, inoculum density and the method of antibacterial activity assessment11131415; these parameters are known to influence the results of susceptibility testing1819. To validate our findings, we tested two additional culture-type strains (S. epidermidis ATCC 12228 and S. aureus ATCC 25923) used in earlier studies1314. While pyocyanin (30 μg) exhibited antibacterial activity against these strains (S. epidermidis ATCC 12228, 15±2 mm; S. aureus ATCC 25923, 24±2 mm), 3-Oxo-C12-HSL and farnesol were found to be ineffective against both these strains as well (ZOI, 0 mm up to 100 μg disk content), under our tested conditions.

In conclusion, our results indicated that of the three quorum-sensing effectors tested, only pyocyanin exhibited a potent antibacterial activity. The inhibitory effect was the maximum against S. epidermidis, followed by S. aureus and E. coli, including their drug-resistant clinical isolates. K. pneumoniae was inhibited to a lesser extent, and the various isolates tested differed in their susceptibility to pyocyanin. In contrast, farnesol and 3-Oxo-C12-HSL did not show any growth inhibitory activity against the four bacterial species under the tested conditions, suggesting that these quorum-sensing effectors do not have an antagonistic action, and would not be potentially antimicrobial.