Multifaceted Interplay between Hfq and the Small RNA GssA in Pseudomonas aeruginosa

Abstract

Behind the pathogenic lifestyle of Pseudomonas aeruginosa exists a complex regulatory network of intertwined switches at both the transcriptional and posttranscriptional levels. Major players that mediate translation regulation of several genes involved in host-P. aeruginosa interaction are small RNAs (sRNAs) and the Hfq protein. The canonical role of Hfq in sRNA-driven regulation is to act as a matchmaker between sRNAs and target mRNAs. Besides, the sRNA CrcZ is known to sequester Hfq and abrogate its function of translation repression of target mRNAs. In this study, we describe the novel sRNA GssA in the strain PA14 and its multifaceted interplay with Hfq. We show that GssA is multiresponsive to environmental and physiological signals and acts as an apical repressor of key bacterial functions in the human host such as the production of pyocyanin, utilization of glucose, and secretion of exotoxin A. We suggest that the main role of Hfq is not to directly assist GssA in its regulatory role but to repress GssA expression. In the case of pyocyanin production, we suggest that Hfq interplays with GssA also by converging a positive effect on this pathway. Furthermore, our results indicate that both Hfq and GssA play a positive role in anaerobic growth, possibly by regulating the respiratory chain. On the other hand, we show that GssA can modulate not only Hfq expression at both transcriptional and posttranscriptional levels but also that of CrcZ, thus potentially influencing the pleiotropic role of Hfq. IMPORTANCE The pathogenic lifestyle of the bacterium Pseudomonas aeruginosa, a leading cause of life-threatening infections in the airways of cystic fibrosis patients, is based on the fine regulation of virulence-associated factors. Regulatory small RNAs (sRNAs) and the RNA-binding protein Hfq are recognized key components within the P. aeruginosa regulatory networks involved in host-pathogen interaction. In this study, we characterized in the highly virulent P. aeruginosa strain PA14 the novel sRNA GssA. We found that it can establish a many-sided reciprocal interplay with Hfq which goes beyond the canonical mechanism of direct physical interaction that had previously been characterized for other sRNAs. Given that the Hfq-driven regulatory network of virulence factors is very broad and important for the progression of infection, we consider GssA as a new RNA target that can potentially be used to develop new antibacterial drugs.

Keywords: Hfq; Pseudomonas aeruginosa; exotoxin A; glucose utilization; posttranscriptional regulation; pyocyanin; sRNAs.

FIG 1

Genomic context of the gssA gene and analysis of the 5′ end of GssA RNA. (A) High-confidence prediction of the secondary structure of GssA by the RNAfold tool within the Vienna RNA Websuite (56). (B) Sequence of the PA14_39480-to-PA14_39500 intergenic region of PA14 where the gssA gene (indicated in boldface) is located. The mapped T1 5′ end of GssA is indicated, while the transcription terminator poly(T) tail is underlined. High-confidence −35 and −10 motifs upstream of T1 are indicated. These motifs were detected by SAPPHIRE (33), a web tool for σ⁷⁰ promoter prediction in Pseudomonas. (C) Northern blot and primer extension analyses of the 5′ end of GssA. Portions (10 μg) of total RNA from PA14 extracted at the end of the exponential growth were untreated (−) or treated (+) with terminator 5′-phosphate-dependent exonuclease (T-ex) and analyzed by Northern blotting. The arrowhead indicates the primary transcript of ~250 nt. The 5′ ends T1 and T2 were mapped by primer extension using 10 μg of total RNA extracted as described above, flanked by a TAGC sequencing ladder.

FIG 2

GssA expression is conditioned by temperature, availability of oxygen, planktonic versus aggregative forms of growth, lack of several σ factors, and inactivation of the hfq gene. (A) Northern blot analysis of the influence of temperature on GssA expression. PA14 cell samples were taken for total RNA extraction at mid (OD₆₀₀ of 0.8) and late (OD₆₀₀ of 1.8) exponential growth phase in BHI at 20°C, 37°C, or after 20 min of acclimation (AC) from 20 to 37°C. (B) Northern blot analysis of the influence of oxygen availability on GssA expression. PA14 cultures were grown in BHI at 37°C anaerobically (NO₃⁻), aerobically (O₂), and aerobically until reaching an OD₆₀₀ of 0.8 and then shifted to anaerobic conditions (O₂ → NO₃). Cell samples were taken for total RNA extraction at mid (OD₆₀₀ of 0.8) and late (OD₆₀₀ of 2.0) exponential growth phase and then 20 and 150 min after the shift from aerobic to anaerobic conditions (t₂₀ and t₁₅₀). (C) Effects of the type of growth, either planktonic or aggregative, on GssA abundance analyzed by Northern blotting and quantitative RT-PCR. Cell samples were taken for total RNA extraction from cultures grown at 37°C in liquid LB with shaking overnight (LSha), statically for 48 h (LSta), or on LB-agar in form of colony biofilm (CBio). The calculation by quantitative RT-PCR of the relative expression of GssA in LSha versus CBio and LSta was performed as described by the 2^−ΔΔCT method (57), first normalizing GssA amounts to 16S ribosome RNA (ΔC_T) and then relating the ΔC_T in CBio and LSta to that in LSha (ΔΔC_T). (D) Effects of inactivating several σ factors and hfq genes on GssA abundance. For total RNA extraction and quantitative RT-PCR, cell samples of a panel of previously published σ factor mutants of PA14 and PA14Δhfq were taken from CBio cultures at 37°C on LB-agar. The GssA levels in PA14 and the mutants are displayed graphically as 2^−ΔCT values. *, P < 0.05; ****, P < 0.0001 (calculated by one-way ANOVA with post hoc Tukey’s HSD test).

FIG 3

Deletion of GssA derepresses the production of pyocyanin and Pel. (A) Quantification of pyocyanin release in liquid LB by PA14, PA14ΔgssA, PA14Δhfq, and PA14ΔhfqΔgssA following growth with shaking overnight (LSha) or statically for 48 h (LSta). *, P < 0.05; **, P < 0.01; ***, P < 0.001 (calculated by one-way ANOVA with post hoc Tukey’s HSD test). (B) Visual comparison of pyocyanin release by PA14, PA14ΔgssA, PA14Δhfq, and PA14ΔhfqΔgssA following static growth in liquid LB for 48 h (LSta) or on LB-agar in form of colony biofilm (CBio). (C) Comparison of pigmentation intensity among spots of PA14, PA14ΔgssA, PA14Δhfq, and PA14ΔhfqΔgssA when grown on the surfaces of Congo red agar plates.

FIG 4

Deletion of hfq gene impairs glucose utilization and the simultaneous lack of GssA rescues this defect. (A) Portions (2 μL) of cultures of PA14, PA14ΔgssA, PA14Δhfq, and PA14ΔhfqΔgssA, serially diluted 10-fold, were spotted onto agar plates with LB and LB-glucose media and incubated at 37°C for 48 h. (B) Portions (2 μL) of cultures of PA14Δhfq and PA14ΔhfqΔgssA harboring the control vector pGM931, and PA14ΔhfqΔgssA harboring pGM-gssA, serially diluted 10-fold, were spotted onto agar plates with LB and LB-glucose media and incubated at 37°C for 72 h.

FIG 5

GssA expression is stimulated in glucose-containing media. PA14 cultures were grown overnight at 37°C in liquid with shaking in the indicated media, and then cell samples were taken for total RNA extraction. The expression of GssA and CrcZ in the different media relative to LB was calculated by quantitative RT-PCR using the 2^−ΔΔCT method. **, P < 0.01; ****, P < 0.0001 (calculated by one-way ANOVA with post hoc Tukey’s HSD test). Note that the GssA responsiveness to glucose in CBio and LSta is shown in Fig. S5.

FIG 6

Hfq and GssA have a convergent positive role in anaerobic growth. Portions (2 μL) of cultures of PA14, PA14ΔgssA, PA14Δhfq, and PA14ΔhfqΔgssA, serially diluted 10-fold, were spotted onto LB or LB-glucose agar plates supplemented with 100 mM KNO₃ to allow anaerobic respiration. Plates were incubated under aerobic (+O₂) or anaerobic (−O₂) conditions at 37°C for 48 to 72 h.

FIG 7

CrcZ levels are affected by the deletion of GssA in a growth medium-dependent manner. PA14 and PA14ΔgssA cultures were grown overnight at 37°C in liquid with shaking in the indicated media, and cell samples were taken for total RNA extraction. By quantitative RT-PCR, the expression of CrcZ in the different media relative to LB was calculated by using the 2^−ΔΔCT method. **, P < 0.01; ****, P < 0.0001 (calculated by one-way ANOVA with post hoc Tukey’s HSD test).

FIG 8

hfq mRNA levels are affected by the deletion of GssA in a growth condition-dependent manner. In LB or BHI at 37°C, PA14 and PA14ΔgssA cultures were grown overnight in liquid with shaking (LSha), overnight on the surface of medium-agar (CBio), or for 48 h in liquid statically (LSta), and then cell samples were taken for total RNA extraction. By quantitative RT-PCR, the levels of hfq mRNA in PA14 and PA14ΔgssA are displayed graphically as 2^−ΔCT values. *, P < 0.05; **, P < 0.01 (calculated by one-way ANOVA with post hoc Tukey’s HSD test).

FIG 9

GssA can enhance hfq mRNA translation in a growth-dependent manner. (A) Prediction by IntaRNA software (43) of the base-pairing interactions between GssA and hfq mRNA. Sequence coordinates are the same as in Fig. 1A for GssA and refer to the +1 translation start site for hfq mRNA. The base triplet change in GssA for generating the variant GssA_GUGmut is indicated, and the ribosome binding site (RBS_hfq) of hfq mRNA is highlighted. (B) Comparison of the fluorescence expressed by the translational fusion hfq::sfGFP in PA14 and PA14ΔgssA strains harboring pGM931 (–) and PA14ΔgssA harboring pGM-gssA (+). (C) Comparison of fluorescence resulting from the translational fusion hfq::sfGFP in PA14 combined with the control vector pGM931 and the plasmids pGM-gssA or pGM-gssA_GUGmut. The data are expressed in arbitrary units (AU) as the mean (n = 12) of the ratio FI₄₈₅/₅₃₅/Abs₅₉₅ ± the SD. ***, P < 0.001; ****, P < 0.0001 (calculated by one-way ANOVA with post hoc Tukey’s HSD test).

FIG 10

GssA can influence the levels of toxA mRNA and those of pvcB, ptxS, ptxR, and regA and toxR. In LB or BHI at 37°C, PA14 and PA14ΔgssA cultures were grown overnight in liquid with shaking (LSha), overnight on the surface of medium-agar (CBio), or for 48 h in liquid statically (LSta), and then cell samples were taken for total RNA extraction. The levels of mRNAs for the different genes in PA14 and PA14ΔgssA are displayed graphically as 2^−ΔCT values determined by quantitative RT-PCR. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 (calculated by one-way ANOVA with post hoc Tukey’s HSD test).

FIG 11

GssA and Hfq interplay in the regulation of secretion of exotoxin A. In LB at 37°C, PA14, PA14ΔgssA, PA14Δhfq, and PA14ΔhfqΔgssA cultures were grown overnight in liquid with shaking (LSha), overnight on the surface of medium-agar (CBio), or for 48 h in liquid statically (LSta). For LSha and LSta, cells were directly separated from the growth medium by centrifugation. For CBio, samples of cells were collected from the agar surface by inoculation loops, resuspended in PBS, and then pelleted by centrifugation. Cell lysates and supernatants (the growth media for LSha and LSta and PBS for CBio) were analyzed by Western blotting with anti-exotoxin A antibodies for detecting intracellular and extracellular levels of exotoxin A (ToxA), respectively.

FIG 12

Schematic representation of the interplay between GssA and Hfq. Hfq is seen in the middle as a central regulatory hub. One main effect of Hfq is to repress GssA and this can influence cellular functions related to the interaction with the host such as the secretion of exotoxin A and the utilization of glucose under the effect of stimuli that influence the regulatory activity of Hfq, including the decoy effect exerted by CrcZ. On the other hand, the expression of GssA is influenced by various environmental factors, including glucose, and by the physiological state dictated by planktonic or aggregative growth, and this is integrated with the regulation by Hfq. Another key function subject to an Hfq/GssA interplay is pyocyanin production. As for glucose utilization and exotoxin A secretion, GssA plays a repressive role in pyocyanin secretion that can be influenced by Hfq, as described above. However, Hfq seems to play a stimulatory GssA-independent role in pyocyanin secretion. Therefore, Hfq can make a positive double input converge on the pathway of pyocyanin production, one independent of and the other dependent on GssA. Reciprocally, GssA can influence Hfq expression at both transcription and posttranscription levels, also depending on the type of growth, planktonic or aggregative. The influence of GssA on Hfq would also pass through CrcZ whose expression is repressed by GssA under some conditions, such as in the presence of glucose and succinate. This could strengthen the CCR exerted by succinate on nonpreferred carbon (CS) sources, but also that of glucose on the same sources. Finally, since cellular functions such as anoxic biofilm formation and anaerobic respiratory chain are regulated by Hfq with CrcZ-mediated modulation (25), GssA may also affect anoxic biofilm and anaerobic respiration in response to host glucose concentration.