Antimicrobial activity of prodigiosin is attributable to plasma-membrane damage

Abstract The bacterial pigment prodigiosin has various biological activities; it is, for instance, an effective antimicrobial. Here, we investigate the primary site targeted by prodigiosin, using the cells of microbial pathogens of humans as model systems: Candida albicans, Escherichia coli, Staphylococcus aureus. Inhibitory concentrations of prodigiosin; leakage of intracellular K+ ions, amino acids, proteins and sugars; impacts on activities of proteases, catalases and oxidases; and changes in surface appearance of pathogen cells were determined. Prodigiosin was highly inhibitory (30% growth rate reduction of C. albicans, E. coli, S. aureus at 0.3, 100 and 0.18 μg ml−1, respectively); caused leakage of intracellular substances (most severe in S. aureus); was highly inhibitory to each enzyme; and caused changes to S. aureus indicative of cell-surface damage. Collectively, these findings suggest that prodigiosin, log Poctanol–water 5.16, is not a toxin but is a hydrophobic stressor able to disrupt the plasma membrane via a chaotropicity-mediated mode-of-action.


Introduction
Micro-organisms synthesise structurally diverse pigments, including melanins, carotenoids, violacein, and prodigiosin; some of which exhibit a range of biological activities. Prodigiosin (Figure 1), synthesised by the Gram-negative bacterium Serratia marcescens, is an antimicrobial (Suryawanshi et al. 2014a; and has insecticidal (Suryawanshi et al. 2014b; and antiparasitic, anticancer, and immunomodulatory activities (Carlos et al. 2011;Suryawanshi 2015c). For some of these biological activities, mechanisms-of-action have been established. For instance, prodigiosin induces apoptosis in cancer cells (Dalili et al. 2012), can alter mitochondrial function and oxidative phosphorylation -giving rise to its antiparasitic activity (Carlos et al. 2011), alters production or function of key enzyme systems in insects (Suryawanshi et al. 2015b) and (in human cell lines) disrupts pH regulation by sequestering protons, disrupting electron transport chains (Roser et al. 2007). In contrast, there is a paucity of information on the mode-of-action of prodigiosin as an antimicrobial. The current study was carried out with the overall goal to identify the primary site targeted by prodigiosin in microbial cells, using the human pathogenic microbes Candida albicans (an ascomycetous yeast), Escherichia coli (a Gram-negative bacterium) and Staphylococcus aureus (a Gram-positive bacterium) as model systems. The specific aims were to: (i) determine inhibitory concentrations of prodigiosin (which cause a 30% growth rate inhibition of pathogens); (ii) quantify prodigiosin-induced leakage of intracellular substances (K + ions, sugars, amino acids and proteins); (iii) assess impacts on activities of microbial enzyme systems; and (iv) determine prodigiosin-induced changes in microbial cell surface.

Antimicrobial activity of prodigiosin
Prodigiosin was successfully isolated and purified from S. marcescens. In our previous study, prodigiosin produced and purified in this way was characterised by Fourier transform infrared (FTIR) spectroscopy and spectrophotometric analyses for UV absorbance, and the prodigiosin peak was visible at 535 nm (Suryawanshi, Patil, Borase, Narkhede, Stevenson, et al. 2015a). In addition, this prodigiosin exhibited considerable ferric-reducing activity, demonstrating its behaviour as an antioxidant. From the current study, the concentrations of this pigment corresponding to 30% growth rate reduction of C. albicans, E. coli and S. aureus were 0.3 ± 0.04, 100 ± 15 and 0.18 ± 0.02 μg ml −1 , respectively. Most microbial pigments that have been assayed for antibacterial activity have been found to inhibit Gram-positive bacteria, and relatively few inhibit Gram-negative species (Suryawanshi, Patil, et al. 2014). Bacteria such as Mycobacterium spp. are able to resist challenges from a variety of stressors and toxins/ toxicants (Santos et al. 2015), which may relate in part to the production of mycolic acids that are located in the cell membrane. For Gram-negative bacteria such as E. coli, it is likely that the lipopolysaccharide layer resists the penetration of inhibitory substances, including prodigiosin, inside the cell.

Prodigiosin-induced membrane leakage and enzyme activities
Atomic absorption spectroscopy was performed to detect the release of K + ions in control and prodigiosin-treated micro-organisms. There was some loss of K + ions from control cells, and ion release from prodigiosin-treated cells was expressed as a percentage of the former (Table 1). Generally, intracellular K + ions concentration is maintained by diverse types of proton pumps, so inactivation of these can lead to ionic imbalance (Clausen & Poulsen 2013). At concentrations of 1.5 to 3.0 μM, prodigiosin can inhibit K + -ATPase activity (Matsuya et al. 2000). In the current study, leakage of K + ions, sugars, amino acids and proteins was greatest for prodigiosin-treated cells of S. aureus (Table 1), and this is consistent with the low concentration of prodigiosin required to cause 30% growth rate inhibition, relative to those for the two other species assayed (see above). For prodigiosin-treated C. albicans cells, there was only a slight loss of sugars and no loss of amino acids or proteins was detected ( Table  1). The cell wall of C. albicans has a network of chitin attached to a matrix of (1 → 3)-β-dglucan, and the latter has a number of (1 → 6)-β-d-glucan branch points that play a role in linking with other (1 → 3)-β-d-glucan chains and mannoproteins. The protein components, both mannoprotein and non-mannoproteins, include ~40 moieties (Chauhan et al. 2002). Further research is needed to determine whether this complex, highly cross-linked structure has any role in preventing release of intracellular substances. There was no apparent inhibition of any of the enzyme activities assayed, regardless of microbial species (data not shown). Therefore, proteases, catalases and oxidases are not the apparent target for the antimicrobial action of prodigiosin.

Apparent changes to the cellular envelope
Control and prodigiosin-treated cells of S. aureus were examined using scanning electron microscopy ( Figure 2). The cells in the control are characterised by a homogeneous, smooth morphology whereas the surfaces of prodigiosin-treated cells are heterogeneous, highly textured with white dots which are approximately 1-2 nm in diameter. whereas it is unclear whether this appearance indicates prodigiosin-induced cell damage or a stress response to prodigiosin, these irregularities have been observed previously in cells of S. aureus exposed to taurine-5-bromosalicylaldehyde Schiff base or the antimicrobial peptide gramicidin (Mareike et al. 2010;yuan et al. 2014).
Collectively, chaotropic stressors are complex in as much as their biophysical modes-ofaction, the corresponding cellular stress responses and aspects of microbial ecology may differ (Cray, Bell, et al. 2013;Cray et al. 2015;. For instance, chaotropicity can determine the outcomes of competitive interactions (Cray et al. 2015;Cray et al. 2016), which microbes inhabit specific environments (Oren & Hallsworth 2014;lievens et al. 2015;yakimov et al. 2015;Paulussen et al. 2016;Fox-Powell et al. 2016) and the extent of biotic windows (Hallsworth et al. 2003a;williams & Hallsworth 2009) and can enable some microbes to dominate their respective habitats (Cray et al. 2013b;Cray et al. 2016). Furthermore, chaotropicity-mediated stresses cause changes in the cell at multiple levels and multiples sites (Hallsworth et al. 2003b;Cray et al. 2015). It may be, therefore, that this mode-of-action is the basis of prodigiosin-induced inhibition of proton pumps (Matsuya et al. 2000) and inability to maintain pH gradients (Roser et al. 2007). what is clear is that the primary target site of prodigiosin in microbial cells appears to be the plasma membrane rather than cytosolic enzyme systems. This finding also gives rise to further intriguing questions: how sensitive is the action of prodigiosin to extremes of -or minute changes in -temperature and water activity (Rummel et al. 2014;Stevenson et al. 2015a;Stevenson et al. 2015b); do chaotropic solutes and solvents (e.g. the biocide ethanol, Bell et al. 2013) cause structural changes in cells of S. aureus similar to those induced by prodigiosin (Figure 2(B)); and does prodigiosin induce a proteome response which is characteristic of chaotropicity-mediated stress (Hallsworth et al. 2003b;Bhaganna et al. 2016). Figure 2. cells of S. aureus without prodigiosin treatment (control; a) and with prodigiosin treatment (B). Prodigiosin was used at the concentration which caused a 30% inhibition of growth rate relative to the control (0.18 μg ml −1 ; see supplementary material). For selected cells, diameter is indicated (in nm).