SEARCH WITHIN CONTENT
ERIC G. ROMANOWSKI / KARA M. LEHNER / NATALIE C. MARTIN / KRIYA R. PATEL / JAKE D. CALLAGHAN / NICHOLAS A. STELLA / ROBERT M.Q. SHANKS *
Citation Information : Polish Journal of Microbiology. Volume 68, Issue 1, Pages 43-50, DOI: https://doi.org/10.21307/pjm-2019-005
License : (CC-BY-NC-ND 4.0)
Published Online: 27-March-2019
Several biotypes of the Gram-negative bacterium
Microbial secondary metabolites include crucial medicines such as antibiotics and anti-inflammatory compounds (Vining 1990). Secondary metabolites influence microbial physiology and behaviors, for example, through regulation of outer membrane vesicle formation (Hoefler et al. 2017), quorum sensing signaling (Waters and Bassler 2005; Barriuso et al. 2018), cellular motility (Daniels et al. 2004), metal acquisition (Demain and Fang 2000; Khan et al. 2018), and spore formation (Demain and Fang 2000).
Some of the most studied secondary metabolites are pigments such as violacein and prodigiosin (Williamson et al. 2006; Choi et al. 2017). Both of these pigments are antimicrobial and therefore likely contri bute to microbe-microbe interactions (Danevcic et al. 2016a; Danevcic et al. 2016b; Im et al. 2017; Hage-Hulsmann et al. 2018). This study focuses on the regulation of prodigioisin, a red tri-pyrole pigment made by Serratia species and a handful of other organisms. While the biological role for prodigiosin is not clearly elucidated, the Haddix group has provided evidence suggesting that prodigiosin is used to finely tune internal energy levels within the bacterium (Haddix et al. 2008; Haddix and Shanks 2018). With respect to mammalian cells, prodigiosin has multiple biological activities including anticancer and autophagy modulating properties (Cheng et al. 2017; Klein et al. 2017; Lin et al. 2017; Klein et al. 2018). Therefore, a thorough understanding of how this secondary metabolite is regulated can be useful for industrial prodigiosin production as well as giving insight into how the organism interacts with the environment.
A variety of transcription factors and environmental stimuli have been shown to influence prodigiosin biosynthesis (Williamson et al. 2006). Among the earliest described is thermoregulation of prodigiosin production by S. marcescens (Blizzard and Peterson 1963; Williams et al. 1965; Tanaka et al. 2004). At 37°C pigment biosynthesis is strongly inhibited, whereas it is synthesized at temperatures of ~ 32°C and below. In this study we used a molecular genetic approach to investigate fundamental questions of whether pro digiosin is thermoregulated at the transcriptional level or at the post-transcriptional level and what transcription factors are involved in this process. We conclude that for S. marcescens strain PIC3611, thermoregulation is controlled primarily at the transcriptional level and this requires the HexS transcription factor.
Bacterial strains and growth. Bacterial strains are listed in Table I. Bacteria were grown in LB broth (Bertani 1951), supplemented with antibiotics when appropriate: tetracycline (10 µg/ml), kanamycin (100 µg/ml), carbenicillin (50 µg/ml), and gentamicin (10 µg/ml) for S. marcescens and ampicillin (100 µg/ml) and kanamycin (50 µg/ml) for Escherichia coli. For liquid broth, bacteria were aerated using a TC-7 tissue culture roller (New Brunswick Inc). Conjugation was used to transform bacteria as follows. Aliquots (500 µl) of donor and acceptor overnight cultures were pelleted by centrifugation, combined, spotted on an LB agar plate, incubated for ~ 18 hours, and streaked for single colonies on selective medium. Macrocolonies were generated by spotting 10 µl of culture, which was grown overnight in LB broth, on LB agar plates and incubating at 30 and 37°C for 24 hours.
Molecular biology. Plasmids were made using yeast homologous recombination as previously described (Shanks et al. 2006; Shanks et al. 2009). Primer sequences are available upon request. A marnier transposon delivery plasmid with a promoterless luxCDABE (lux) reporter was derived from pSC189 (Chiang and Rubin 2002) with yeast replication machinery from pMQ132 (Shanks et al. 2009) and the lux operon from pGRL1 (Benedetti et al. 2012). The resulting plasmid, pMQ690, was moved into S. marcescens by conjugation. Six pigmentless colonies were chosen from approximately 5000 transposon mutants and these were analyzed for lux operon expression using a luminometer (Perkin Elmer Wallac 1450-021). Two of the isolates generated light due to the orientation of the transposon in the genome and the mutations were mapped and the insertion junction was sequenced as previously described (Chiang and Rubin 2002). The two insertions mapped within the pig operon which rendered the isolates unable to make pigment, and oriented downstream of the pig promoter so that light can be measured as a reporter of transcription.
To make pig promoter reporter plasmid, pMQ713, the pig promoter from strain PIC3611 was cloned upstream of the lux operon on an oripBBR1-based low copy plasmid, pMQ670 (R. Shanks, to be described elsewhere) as noted above. This episomal plasmid was used to transform a variety of strains described herein by conjugation.
Reporter analysis and prodigiosin assay. Bacteria were grown for 18–20 hours in LB medium with aeration and the culture density was read at 600 nm with a SpectraMax M3 spectrophotometer using a 1 cm cuvette. Luminescence was measured from 150 µl aliquot of culture in a black, opaque, 96-well plate with a luminometer as noted above. The ratio of luminescence arbitrary units to OD600 was calculated as relative luminescence units (RLU). For HcRed fluorescence, a BioTek Synergy 2 plate reader was used to read fluorescence and optical density (OD600) of 150 µl samples of overnight cultures. The ratio of fluorescence to optical density was reported as relative fluorescence units (RFU).
To measure prodigiosin from macrocolonies, colonies were scraped from agar plates using a toothpick and suspended in 1.2 ml of saline. A 0.2 ml sample was used to measure OD600 in a 1 cm cuvette, and the remaining 1 ml was pelleted by centrifugation, prodigiosin was extracted using acidified ethanol as previously described (Slater et al. 2003), and prodigiosin was measured using a spectrophotometer at 534 nm.
Transcriptional control of pigment biogenesis. Typical of pigmented S. marcescens strains, strain PIC3611 does not generate prodigiosin when grown at 37°C but does so at lower temperatures, such as 30°C and below (Fig. 1A). To analyze transcription from the pigment biosynthetic operon (pig), we decided to make a luxCDABE (lux) reporter strain. To achieve this, we generated a promoterless lux mariner transposon delivery plasmid and introduced it into the chromosome of S. marcescens strain PIC3611. Following transposon mutagenesis, pigmentless colonies were screened with a luminometer for light production after growth at 30°C. Two isolates were found, one mapped to pigB (bp 1730) and another to pigF (bp 1074) in the pig operon. Cultures of these two strains were grown with aeration at 30°C and 37°C, and light production was measured (Fig. 1B). After 18 hours of growth in LB medium, there was a significantly higher level of the signal from both cultures incubated at 30°C, compared to the same strains grown at 37°C. A similar pattern was observed for both pigB and pigF-lux isolates, although there was a higher level of expression from the pigB insertion. This result suggests that temperature regulation of pigment production is, at least in part, controlled at the transcriptional level.
To test whether induced pig operon transcription at 37°C could restore prodigiosin biosynthesis, the chromosomal pigA-N operon was placed under control of the inducible PBAD operon of wild-type strain PIC3611 by integration of plasmid pMQ262. Prodigiosin biosynthesis of the resulting strain was tightly regulated by addition of L-arabinose to the growth medium. In the absence of L-arabinose, no visible pigmentation was observed, even at 30°C (Fig. 2A). When L-arabinose was added, pigmentation increased in a dose dependent manner at 30°C (Fig. 2A). When tested at 37°C, L-arabinose was able to induce pigmentation in the WT with PBAD promoting pig expression (Fig. 2B). Arabinose (2 mM) added to the wild-type strain without pMQ262 did not alter pigmentation (data not shown). This result suggests that transcription of the pigA-N operon is sufficient to produce pigmentation even at non-permissive temperatures. This suggests that lack of pig operon transcription at 37°C rather than unknown mechanism(s) of post-transcriptional control are responsible for the pigmentation defect at 37°C.
HexS is necessary to inhibit pigmentation at 37°C. The CRP and HexS transcription factors are potent inhibitors of secondary metabolite biosynthesis in strain PIC3611 and other strains, such that crp and hexS mutants are more highly pigmented than the isogenic wild-type strains (Kalivoda et al. 2010; Stella et al. 2012; Stella and Shanks 2014; Shanks et al. 2017b). It was hypothesized that HexS and CRP were responsible for inhibiting the prodigiosin production at 37°C. To test a role for these transcription factors in thermoregulation, we used well-defined strains with crp and hexS mutations in the PIC3611 strain background. The prodigiosin phenotypes of these specific strains were previously complemented in trans with the respective genes on plasmids (Kalivoda et al. 2010; Stella and Shanks 2014). The crp mutant was only slightly pigmented at 37°C (Fig. 3A), whereas the hexS mutant was pigmented at both 30°C and 37°C (Fig. 3B). Pigment from the bacteria on agar plates was measured (Fig. 3C). The hexS mutant produced significantly more prodigiosin at 37°C than either wild type or the crp mutant. The crp mutant did have a ~7.5-fold increase over wild-type prodigiosin levels, but the difference was not significant. Additionally, in liquid cultures grown at 37°C, the hexS but not the crp mutant was pigmented. In this case, the wild 30°C type bacteria grown in liquid culture had an A534/OD600 value of 0.003 ± 0.001, the crp mutant generated 0.004 ± 0.002, and the hexS produced significantly more 0.012 ± 0.004 (n = 6, p < 0.001, ANOVA with Tukey’s post-test). These results indicate that HexS has a prominent role in temperature-dependent pigment inhibition. Consistently, the hexS gene was found in a genetic screen in which it produced slightly pigmented colonies at 37°C in S. marcescens strain 274 (Tanikawa et al. 2006).
To further analyze the impact of HexS on pig transcription, a plasmid-borne pig-lux reporter was generated and used to test the prediction that pig transcription is increased at 37°C in the hexS mutant relative to the wild type. While the plasmid-borne reporter has a much higher background than the chromosomal reporters (Fig. 3D versus Fig. 1B) it was much easier to move the reporter plasmid to different strain backgrounds. A greater than10-fold reduction in RLU measured from the pig-lux reporter was measured from the wild type grown at 37°C compared to 30°C. Furthermore, expression from the wild type at 37°C was indistinguishable from pig-lux measured from an eepR mutant grown at 30°C (Fig. 3D). This mutant is unable to make pigment because it lacks, EepR, a positive regulator of pig transcription. These data validated the reporter construct, so it was moved into the hexS mutant. There was no significant difference in pig-lux measured from the hexS mutant at 30°C versus 37°C (p > 0.05 ANOVA with Tukey’s post-test), although there was a slight downward trend.
Pigment thermoregulation is EepR and PigP independent. In previous studies, it was determined that the transcription factor genes eepR and pigP were more highly expressed in crp and hexS mutants and that EepR and PigP were positive regulators of pig transcription in S. marcescens and another Serratia species (Fineran et al. 2005; Shanks et al. 2013; Stella et al. 2015; Shanks et al. 2017a). A model depicting the interaction between these transcription factors and the pig operon is shown in Fig. 4. Based on these data, we predicted that the increased expression of eepR and pigP in the crp and hexS mutants could be responsible for the increased pigmentation at 37°C. The wild-type strain with either a vector or wild-type eepR and pigP on a plasmid was grown at 30°C and 37°C, and while more pigmentation was observed at 30°C with the eepR and pigP plasmids, no pigmentation was observed at 37°C (Fig. 5). These plasmids were previously shown to complement eepR and pigP mutant phenotypes and are under control of the E. coli lac promoter (Shanks et al. 2013). We confirmed that the lac promoter is highly expressed in S. marcescens at both 30°C and 37°C using a transcriptional hcred fluorescent reporter-fusion. Although fluorescence was 2.7-fold higher at 30°C than 37°C (10745 ± 1019 RFU at 30°C versus 3997 ± 947 RFU at 37°C), at both temperatures fluorescence was much higher than background levels of ~ 10 RFU at both temperatures. Together these control experiments suggest that the plasmids are functional and expressing the genes at both temperatures. Moreover, data presented here suggest that thermoregulation of pigmentation is largely independent of EepR and PigP and is partially due to direct repression of pig by HexS at 37°C.
The results of our study indicate that thermoregulation predominantly occurs at the transcriptional level. Our study: 1) correlated the nonpermissive temperature with a lack of pig transcription, and 2) demonstrated that artificial induction of pig operon transcription was sufficient for pigmentation at 37°C. This latter result indicated that the prodigiosin biosynthetic genes were translated and function at the nonpermissive temperature and that the chemical substrates needed to generate prodigiosin were present in the cell at the non-permissive temperature. Whereas post-transcriptional thermoregulation may be occurring, it was below the level of detection of the assays used in this study. Therefore, the findings in this study support the conclusion that thermoregulation of prodigiosin biosynthesis by S. marcescens occurs predominantly at the transcriptional level.
There are a large number of other stimuli that influence prodigiosin production at the transcriptional level, such as sugars, phosphate, ATP, cyclic nucleotides, indole, and quorum sensing molecules in S. marcescens and other Serratia species, presumably through the action of transcription factors that respond to the stimuli (Williamson et al. 2006; Haddix et al. 2008; Fender et al. 2012; Hidalgo-Romano et al. 2014; Haddix and Shanks 2018). It is possible that altered concentrations of one or more of these factors at 37°C is responsible for the pigment inhibition effect. Negative regulators of S. marcescens prodigiosin production include CRP (Kalivoda et al. 2010; Stella and Shanks 2014), HexS (Tanikawa et al. 2006; Stella et al. 2012), and RssAB (Horng et al. 2010). Previously, Horng and colleagues (2010) tested whether mutation of the rssBA genes would restore pigmentation at 37°C, and found that the CH-1 strain of S. marcescens with mutation of rssA or both rssA and rssB did not produce prodigiosin at 37°C. The authors concluded that prodigiosin production is inhibited by an unknown mechanism independent of the RssB/A two-component system (Horng et al. 2010). Here, we found that mutation of crp allowed production of faintly red colonies at 37°C, but liquid cultures were not red at 37°C. This may be due to a number of different variables between growth in liquid and on plates such as oxygen levels, as oxygen is necessary for pigment production (Heinemann et al. 1970). A similar pattern was observed at 30°C with a pigP mutant of strain PIC3611 and gumB and eepR mutants of the strain K904 background, all of which produced pink colonies on plates, but no detectable pigmentation in aerated liquid cultures (Shanks et al. 2013; Stella et al. 2015; Stella et al. 2018). Another example is evident with the wild-type strain grown at 30°C on LB agar plates becomes bright red (Fig. 1A) whereas in LB broth, the bacteria are clearly less pigmented (Fig. 2A). Evidence suggests that CRP does not directly bind to the pigment biosynthetic operon promoter (Kalivoda et al. 2010), implying that one or more of the many genes regulated by CRP, in turn, alters pig operon expression at 37°C. The CRP protein is implicated in heat shock responses in E. coli and other bacteria, indicating that a crp mutation can impact bacterial temperature responses (Nagai et al. 1990; Choudhary et al. 2014). Here, we found a relatively minor role for CRP, with some pigment being observed on agar plates in the crp mutant at 37°C.
The hexS gene of S. marcescens was originally isolated by Tanikawa and colleagues (2006) in a screen for mutations, in strain 274, that allowed prodigiosin and serrawettin W1 (also known as serratamolide) production at 37°C. The hexS mutant produced a faintly red colony at 37°C and overproduced progidiosin at 30°C. This result was different from our study where the hexS mutant produced a robustly red colony at 37°C on the same medium over a similar time frame suggesting that there may be some differences between pigment regulation between strains, as has been shown for cyclic-AMP regulation of prodigiosion (Stella and Shanks 2014). Nevertheless, these studies confirm that HexS is a key transcription factor in the inhibition of pigment biosynthesis at 37°C.
Given the potential of prodigiosin and similar compounds as antimicrobial, antitumor, and anti-inflammatory therapeutics (Perez-Tomas et al. 2003; Williamson et al. 2006; Perez-Tomas and Vinas 2010), understanding regulation of the natural biosynthesis of these compounds can inform their industrial production. This study provides evidence that barriers to prodigiosin production at 37°C can be overcome at the transcriptional level, and that mutations such as in hexS can be used to increase prodigiosin production, and provides insight into the basic biological question of secondary metabolite thermoregulation by S. marcescens.
The authors would like to thank Mohammed AlHigaylan and Kimberly Brothers for critical reading of the manuscript. This work was supported by NIH Core Vision Research grant EY08098, NIH grant AI085570, unrestricted funds from the Research to Prevent Blindness, and the Eye and Ear Foundation of Pittsburgh.
Thermoregulation of pigmentation by A. Pigment production of the wild-type strain PIC3611 grown for 18 h at 30°C and 37°C streaked for single colonies on an LB agar plate. B. Thermoregulation of
Thermoregulation of pigmentation by
A. Pigment production of the wild-type strain PIC3611 grown for 18 h at 30°C and 37°C streaked for single colonies on an LB agar plate. B. Thermoregulation of
Induction of A. Pigment production at 30°C in liquid culture after 18 h of growth at 30°C. The wild-type strain was modified by recombination of pMQ262 through which an L-arabinose-inducible promoter replaces the
A. Pigment production at 30°C in liquid culture after 18 h of growth at 30°C. The wild-type strain was modified by recombination of pMQ262 through which an L-arabinose-inducible promoter replaces the
The A and B. Pigment production at 30°C and 37°C after 18 h of growth at 30°C. The
A and B. Pigment production at 30°C and 37°C after 18 h of growth at 30°C. The
Model genetic circuit used in this study. In this regulatory circuit, the role of several transcriptional regulators in control of the
Model genetic circuit used in this study. In this regulatory circuit, the role of several transcriptional regulators in control of the
Multicopy expression of
Multicopy expression of