The Small RNA NcS25 Regulates Biological Amine-Transporting Outer Membrane Porin BCAL3473 in Burkholderia cenocepacia

ABSTRACT Regulation of porin expression in bacteria is complex and often involves small-RNA regulators. Several small-RNA regulators have been described for Burkholderia cenocepacia, and this study aimed to characterize the biological role of the conserved small RNA NcS25 and its cognate target, outer membrane protein BCAL3473. The B. cenocepacia genome carries a large number of genes encoding porins with yet-uncharacterized functions. Expression of the porin BCAL3473 is strongly repressed by NcS25 and activated by other factors, such as a LysR-type regulator and nitrogen-depleted growth conditions. The porin is involved in transport of arginine, tyrosine, tyramine, and putrescine across the outer membrane. Porin BCAL3473, with NcS25 as a major regulator, plays an important role in the nitrogen metabolism of B. cenocepacia. IMPORTANCE Burkholderia cenocepacia is a Gram-negative bacterium which causes infections in immunocompromised individuals and in people with cystic fibrosis. A low outer membrane permeability is one of the factors giving it a high level of innate resistance to antibiotics. Porins provide selective permeability for nutrients, and antibiotics can also traverse the outer membrane by this means. Knowing the properties and specificities of porin channels is therefore important for understanding resistance mechanisms and for developing new antibiotics and could help in overcoming permeability issues in antibiotic treatment.


RESULTS
Outer membrane porins are conserved predicted targets of NcS25. NcS25 is an 84-nucleotide (nt) noncoding RNA, as determined by differential transcriptome sequencing (dRNA-Seq) (22), and ncS25 is located in the intergenic region between a gene encoding a conserved hypothetical protein (BCAL3007) and a gene encoding a porin protein (BCAL3008) (Fig. 1A) (22). ncS25 and its upstream genomic context are conserved within the genera Burkholderia, Paraburkholderia, and Pandoraea. Located downstream of the ncS25 sequence is a gene annotated as hypothetical protein with which it shares a bidirectional rho-independent terminator, followed by a gene annotated as RNA-binding chaperone protein, or in Pandoraea spp. a gene annotated as a sigma 54 (s 54 )-dependent Fis family transcriptional regulator (Fig. 1A).
Ten genomes each of the genera Burkholderia, Paraburkholderia, and Pandoraea were computationally screened for conserved putative targets of NcS25 using CopraRNA (25). This software tool takes the energy required to unfold the secondary structure of sRNA and target as well as conservation of targets into account and combines the results with phylogenetic information (25). The screening was limited to the region 200 nt upstream to 100 nt downstream of gene translation start sites, and the output comprised hits conserved among at least 50% of input strains. In the genus Burkholderia, the 100 hits with the highest probability included 14 genes encoding porin proteins, 13 of which belonged to the large family of homologous porin proteins found in Burkholderia spp. The gene with the most extensive interaction region was porin BCAL3473, a member of this porin protein family.
The combined CopraRNA P value for the predicted interaction between NcS25 and the BCAL3473 homolog in Burkholderia spp. is 4 orders of magnitude lower than for the next-likeliest interaction predicted by this tool (see Table S1 in the supplemental material), setting this particular predicted target apart from the rest. This is similar for the respective BCAL3473 homologs in Paraburkholderia spp. and Pandoraea spp. (Table S1).
The putative SD sequence of BCAL3473 is conserved among all genera which harbor ncS25, and in all cases, a 7-nt stretch within the SD sequence has a perfect complement within its cognate NcS25 ( Fig. 1B; Fig. S1). Another interaction region is predicted for the start of the BCAL3473 coding sequence; here, the complementarity is not always perfect. The region between the SD sequence and the translation start site of BCAL3473 is not conserved, nor is the rest of the 59 untranslated region (UTR).
BCAL3473 is 383 amino acids long with a size of 39.91 kDa, and it forms a 16-stranded beta-barrel to span the OM. The 59 UTR of BCAL3473 is 87 bases long (26). The predicted interaction region in B. cenocepacia spans from nucleotide position 214 in the 59 UTR of BCAL3473 to 16 nt into the beginning of the gene ( Fig. 2A).
sRNA NcS25 represses expression of porin BCAL3473. To confirm the regulation of targets by NcS25, an ncS25 deletion mutant of B. cenocepacia J2315 (DncS25) was constructed. The DncS25 strain had no growth defect compared to the wild type in rich mineral medium (Fig. S2).
The expression of six putative targets was analyzed by qPCR, comparing the wild type and deletion mutant in planktonic culture and in biofilms. Besides porin BCAL3473, three other porins with a similar predicted sRNA-mRNA interaction that included the region around the gene start site were tested (BCAL2615, BCAM1787, and BCAL0594), as well as the putative target porin encoded adjacent to ncS25 (BCAL3008) and a phosphotransferase system II (PTSII) transport protein (BCAL0781) (Fig. S3).
Of these genes, only BCAL3473 was significantly upregulated in the DncS25 mutant, and quite considerably, 83-fold in planktonic cultures and 27-fold in biofilms (Table 1). This could be complemented by expressing NcS25 in trans from its native promoter (Fig. 2C). When a plasmid with a mutated derivative of ncS25 in which several bases in the putative interaction region were exchanged was expressed in trans ( Fig. 2A and B), complementation did not occur (Fig. 2C). Expression of NcS25 in trans in the DncS25 mutant was confirmed by qPCR and was comparable to wild-type expression, for plasmid pncS25 as well as for pncS25mut (quantification cycle [C q ] values: 20.1 for wild type, 19.1 for the DncS25 mutant carrying pncS25, and 18.5 for the DncS25 mutant carrying pncS25mut).
Experiments with translational reporter gene fusions to confirm regulation of gene expression with sRNA are usually performed with a plasmid system in which both sRNA and target are overexpressed from separate promoters. In B. cenocepacia J2315, the high innate antibiotic resistance renders most resistance markers unsuitable; moreover, only one useable promoter system and compatible plasmid origin of replication are available. For that reason, only the putative target region was overexpressed, fused to mCherry, and under the control of a rhamnose-inducible promoter. This construct was then transformed into wild-type J2315 and J2315 DncS25.
Fluorescence production was slightly accelerated in the DncS25 strain compared to the wild type (Fig. 2D), confirming the regulation of BCAL3473 by NcS25. In positive controls with the intact mCherry gene, fluorescence production was equal in both the wild-type and DncS25 strains (Fig. S4).
Nitrogen depletion induces expression of porin BCAL3473. Screening the literature for expression data for porin BCAL3473 revealed that the porin was downregulated when s 54 factor enhancer binding protein NtrC was deleted in B. cenocepacia H111 (27) and upregulated in Burkholderia glumae when the regulator GvmR (homolog to BCAL0461) was deleted (28). To explore whether these factors also play a role in BCAL3473 regulation in B. cenocepacia J2315, we investigated expression of NcS25, BCAL3473, and other putative targets under conditions of nitrogen depletion and in a BCAL0461 knockdown mutant.
Under nitrogen depletion, of the tested genes, only those encoding porins BCAL3473 and BCAL2615 were upregulated (Table 1), in the wild-type and DncS25 strains. As a positive control for response to nitrogen depletion, expression of two genes known to be induced by nitrogen depletion and exhibiting a s 54 binding site (29), a urea ABC transporter (BCAL3098), and nitrogen regulatory protein P-II (BCAL0729), was also assessed. As expected, these genes were strongly upregulated under nitrogen depletion (Table 1), whereas NcS25 did not change expression under this condition.
The promoter region of porin BCAL3473 contains sequences identical to the 212 and 224 elements of the wider consensus sequence for the s 54 binding sites of B. cenocepacia (29), but with an extended spacer sequence (Fig. 3). The promoter region of BCAL2615 contains only the tetranucleotide of the 212 element. No sequences similar to these motifs were detected upstream of ncS25 or the other quantitative PCR (qPCR)-tested targets.
Knocking down expression of LysR-type regulator BCAL0461 affected the expression of all tested porins, with porin BCAL3473 downregulated to a large extent (Table 1). NcS25 was not differentially expressed in the knockdown mutant (Table 1). The predicted secondary structure is identical for both sequences, whereas the predicted interaction with the BCAL3473 mRNA (indicated in orange on the secondary structure) is much shorter for NcS25mut than for native NcS25. The part of NcS25 interacting with the SD sequence is situated in a single-stranded section and can serve as a seed region for interaction initiation. The translation start site is depicted in red. Nucleotides exchanged in ncS25 to create ncS25mut are indicated in blue. Mutations were chosen to have no impact on the computationally predicted secondary structure of the sRNA. Numbers indicate the position relative to the translation start site for mRNA and sRNA nucleotide position. The predicted free energy of the interaction is shown in parentheses. (C) Expression of BCAL3473 in the DncS25 mutant, compared to wild-type J2315, is strongly increased in planktonic cultures grown in rich mineral medium. This was complemented by expression of the native NcS25 in trans, whereas in trans expression of NcS25mut did not have this effect. Error bars depict standard deviations, and asterisks denote statistically significant changes (P # 0.05) compared to wild-type B. cenocepacia J2315 vector control (VC). The data are derived from three biological replicates with two technical replicates each. (D) Fluorescence produced by translational reporter gene fusion confirms regulation of BCAL3473 by NcS25. The 59 UTR and first 84 nt of the coding sequence of BCAL3473 were fused to mCherry and the construct transformed into wild-type B. cenocepacia and the DncS25 mutant. Incubation was performed with and without 0.1% rhamnose (rha). The error bars represent the standard deviations for three biological replicates.
Role and Regulation of B. cenocepacia Porin BCAL3473 mSphere Biological amines are among the putative substrates of porin BCAL3473. Porin BCAL3473 was deleted from the genomes of B. cenocepacia strains J2315 and K56-2, and complementation mutants were constructed for expressing BCAL3473 in trans under the control of a rhamnose-inducible promoter ( Table 2). In rich mineral medium, mutants had no growth defect compared to the respective wild-type strain (Fig. S2).
Initial screening for possible substrates of porin BCAL3473 was performed with Biolog carbon source plates PM1 and PM2A and with B. cenocepacia strain K56-2, which is from the same clonal complex as strain J2315 (30). Strain J2315 produced a uniformly low signal in all test wells of the Biolog plates and in the negative-control well and was therefore not used for this assay.
Tyramine (the product of decarboxylation of tyrosine), arginine, and putrescine were the only substrates on these plates for which deletion mutant K-DBCAL3473 showed strongly reduced respiration, complemented by expressing BCAL3473 in trans (Table S2). To test further compounds and to confirm the Biolog results, we determined growth curves in an ammonium-free mineral medium with glucose as the carbon source and a variety of nitrogen compounds as the sole source of nitrogen, or with compounds other than glucose as the sole source of carbon. Growth of K-DBCAL3473 was attenuated on tyrosine, which is not included in the Biolog plates that were used, and on arginine, tyramine, and putrescine. All these phenotypes could be complemented (Fig. 4).
Strain J2315 did not grow on tyrosine and tyramine as sole sources of nitrogen, possibly because of a point mutation in hmgA (31). Instead, it produced a brown pigment when either of these compounds was present in the medium (32). Growth with putres-  cine or arginine as the sole source of nitrogen was attenuated in J-DBCAL3473 (Fig. S2) and slightly increased with putrescine in DncS25. Growth on all other major amino acids, including L-ornithine and L-homoserine, and on purine and pyrimidine bases was not affected by deletion of porin BCAL3473. Utilization of the aromatic compounds 4-hydroxybenzoate, phenylacetic acid, and 4-hydroxyphenylacetic acid, as well as phenylpropionic acid and 4-hydroxyphenylpropionic acid, was unaffected by porin deletion. Utilization of the nitrogen compounds urea, ammonium, and diethanolamine was also unaffected (data not shown). The MICs of the antibiotics amikacin, tobramycin, azithromycin ciprofloxacin, chloramphenicol, tetracycline, imipenem, meropenem, and ceftazidime for J-DBCAL3473 did not change compared to those for the wild-type strain (Table S3). Porin protein BCAL3473 is therefore likely not involved in transporting these compounds across the OM.

DISCUSSION
In the present study, we aimed to elucidate the biological role of NcS25, a conserved small RNA highly expressed in B. cenocepacia biofilms (22). Computational analysis of the genome context and of putative targets with complementarity to the NcS25 sequence pointed to a role in regulating porin expression. NcS25 is encoded adjacent to a porin throughout the genera in which NcS25 occurs. A conserved syntenic context points to Plasmids pSCrhaB2 Expression vector, ori pBBR1 , rhaR, rhaS, P rhaB , dhfr 56 pSCrhaM2 Derivative of pSCrhaB2, SD sequence and start codons removed 38 pncS25 pSCrhaM2 with ncS25 including 167 nt upstream sequence This study pncS25-mut Derivative of pncS25 with point mutations in the predicted interaction region This study pL3473 pSCrhaB2 with porin BCAL3473 under a rhamnose-inducible promoter This study pGPI-SceI-XCm Suicide vector for allelic replacement, I-SceI restriction site, ori R6K , dhfr, Cm r 53 pDAI-SceI-SacB I-SceI nuclease, Tet r , counterselectable marker SacB, ori pBBR1 53 pSC200 Suicide vector for promoter exchange, ori R6K , rhaR, rhaS, P rhaB , dhfr 57 pSC200-L0461 pSC200 with fragment of regulator BCAL0461 This study pRK2013 Helper plasmid for triparental mating, ori colE1, RK2 derivative, Km r , mob 1 , tra 1 55 pSCmCh-pos pSCrhaB2 with fluorescent protein mCherry This study pSCmCh-neg pSCrhaM2 with mCherry This study pSCmChL3473 pSCrhaM2 with mCherry fused to the 59 UTR and partial coding sequence of BCAL3473 This study affiliation to a certain regulon or biological response, and sRNAs and their flanking genes are often involved in the same metabolic process or stress response (14). Even if NcS25 does not regulate the porin adjacent to which it is encoded, both might still be part of the same regulatory circuit. The relatively high number of porin proteins among the targets predicted with high probability also pointed to a role of NcS25 in OM porin protein regulation, as many well-investigated sRNAs regulate multiple genes with similar function (25). While several predicted target porins were tested for differential regulation in DncS25, only one of these changed expression. This is not surprising, as computational target prediction based on sequence complementarity often results in many false-positive hits (25). On the other hand, since trans-encoded sRNAs often have multiple targets, it is likely that other genes besides BCAL3473 are targeted by NcS25, either by base-pairing or other mechanisms, and possibly under different physiological conditions. Levels of BCAL3473 mRNA were drastically increased when ncS25 was deleted from the genome; the large fold change increase shows that NcS25 is effective at repressing the expression of porin BCAL3473. The predicted interaction region is relatively large, longer than for other typical sRNAs which are involved in posttranscriptional regulation of porin proteins. It is also larger than that for the other predicted targets, which is the underlying reason for the large jump in CopraRNA P values from the first on the list to the next. RhyB, an sRNA regulating OM proteins in E. coli, needs only a 7-bp pairing region to effectively repress its targets (33). A 7-nt region complementary to the SD sequence in the target mRNA could be the seed of the interaction, as this region is located in a singlestranded part of the folded NcS25. This seed could then initiate a larger interaction, even though this would involve opening part of the secondary structure of NcS25. The free energy left after unfolding and hybridization shows that this is possible.
The conservation of the putative interaction points to a conserved mechanism of action that includes inhibiting translation initiation by occluding the SD. It is possible that Role and Regulation of B. cenocepacia Porin BCAL3473 mSphere the relatively long predicted interaction of NcS25 with the BCAL3473 mRNA in B. cenocepacia also induces mRNA degradation by forming an RNA-RNA duplex and thereby recruiting RNase E to the target mRNA (34), which would help to explain the large effect NcS25 has on BCAL3473 mRNA expression. In comparison to the large effect that ncS25 deletion has on the mRNA level, the effect seen in translational reporter gene fusions appears to be relatively small. It is likely that the abundance of NcS25 is not large enough compared to that of the target supplied in trans from the plasmid to allow for a bigger effect in this assay. Ideally, the sRNA is also oversupplied in assays like this, which was not possible in this case for lack of suitable vectors. Nonetheless, despite the small difference, the results from this assay confirm that BCAL3473 is regulated by NcS25.
Besides NcS25, other factors also regulate BCAL3473 expression. BCAL3473 is activated by LysR-type regulator BCAL0461, which makes it inversely regulated to the homologous porin in B. glumae, where the regulator GvmR acts as a repressor (28).
Porin BCAL3473 is upregulated under nitrogen depletion, independently of the presence of NcS25. This observation might be linked to the similarity of the BCAL3473 promoter region with the conserved s 54 binding site. Sigma subunits of RNA polymerase are required for selective promoter recognition and transcription initiation. The s 54 subunit (RpoN) forms the s 54 RNA polymerase holoenzyme, which binds to conserved 212 and 224 promoter elements (35). The s 54 holoenzyme represses expression from promoters until ATP hydrolysis by an enhancer binding protein (such as NtrC) activates transcription (36).
In B. cenocepacia H111, sigma binding sites were enriched in the promoter regions of genes induced by nitrogen limitation (29); the s 54 factor therefore has a role in B. cenocepacia in the response to nitrogen limitation. BCAL3473 was one of the genes upregulated by nitrogen limitation in strain H111; moreover, the porin was upregulated in a NtrC deletion mutant (27).
The BCAL3473 promoter region contains elements with the same conserved 212 and 224 sequence elements as in the consensus sequence for s 54 -dependent binding of B. cenocepacia, but with an extended spacer sequence. Spacer length between the two conserved elements is normally conserved, a longer spacer could disrupt binding and impair repression by s 54 holoenzyme (37). Our results, together with the observations in B. cenocepacia strain H111, suggest that s 54 holoenzyme could bind, though possibly inefficiently, to the promoter region of BCAL3473.
However, expression of porin BCAL3473 does not respond specifically to nitrogen limitation. Porin BCAL3473 was constitutively expressed in LB broth and other rich media (38,39), in line with an incomplete s 54 -dependent repression. Likewise, NcS25 was constitutively expressed, and no evidence for differential expression was found under the conditions tested. sRNAs are often expressed only under certain stress conditions (14,15), but this does not seem to be the case for NcS25. NcS25 could constitutively downregulate porin BCAL3473 and thus contribute to an equilibrium of porins present in the OM and to homeostasis of cell metabolism, important for fitness of the bacterium.
The outer membrane permeability of Burkholderia spp. for antibiotics is relatively low; in Burkholderia cepacia, it was found to be approximately 10 times lower than that of E. coli and similar to that of P. aeruginosa (40). The properties of Burkholderia porins should therefore promote membrane impermeability. The major porin isolated from B. cepacia, OpcP (BCAM1931 in B. cenocepacia J2315 and Opm38 in Burkholderia thailandensis and Burkholderia pseudomallei [41]), is a member of the large family of homologous porins and is relatively small (40,42). BCAM1931 is the porin with the highest expression in B. cenocepacia J2315, as shown by RNA sequencing (38), and could therefore also be the major porin in this species. Moreover, BCAM1931 is essential in several Burkholderia spp. (43), which makes it the only bacterial porin protein shown to be essential for growth in several independent studies using transposon mutagenesis combined with high-throughput sequencing (43)(44)(45)(46). The high expression and the essentiality of BCAM1931 suggest that it has an additional biological role besides its function as an OM channel, possibly in maintaining OM integrity (43).
Porin BCAL3473 mRNA is more than 250-fold less abundant than BCAM1931 (38), in part due to the repression by NcS25. It is involved in arginine and tyrosine utilization and can also transport biological amines tyramine and putrescine. Aromatic compounds similar to tyrosine and tyramine, but without amine groups, were not affected. Thus, porin BCAL3473 seems to be specific for amine-containing compounds with a variety of structures. Porins are passive diffusion channels and often not specific for a single compound (4). On the other hand, amino acids such as arginine can be transported by more than one porin, as observed in P. aeruginosa (4,47). This could also be the case in B. cenocepacia, since the porin deletion mutants were still growing on arginine and tyrosine, albeit more slowly. In B. cenocepacia K-56-2 porin BCAL3474 appears to be essential for growth on tyramine, and in B. cenocepacia J2315 for growth on putrescine. The accelerated growth of the DncS25 mutant on putrescine shows that increased expression of BCAL3473 can accelerate transport of this compound. These two compounds could therefore be specifically transported by porin BCAL3473.
Conclusion. The present study represents, to the best of our knowledge, the first characterization of the substrate specificity of an OM porin protein in B. cenocepacia. Porin BCAL3473 is involved in transport of arginine, tyrosine, tyramine, and putrescine across the OM, which makes it an important part of nitrogen metabolism in B. cenocepacia. Porin BCAL3473 is constitutively expressed and simultaneously strongly repressed by NcS25, possibly to dampen unwanted translation that would otherwise lead to imbalances in porins present in the OM. Based on the conservation of the interaction regions between sRNA and porin mRNA, this regulation appears to be conserved throughout several genera of Burkholderiaceae, suggesting that the downregulation of BCAL3473 by NcS25 is important for fitness in these bacteria. The regulation of porins is probably as complex in B. cenocepacia as in other bacteria, and this study is just the first step toward elucidating the specificity and regulation of the many porins present in B. cenocepacia.

MATERIALS AND METHODS
Strains and media. Strains and plasmids used in this study are listed in Table 2. Strains were routinely cultured in LB broth or agar supplemented with antibiotics when necessary (50 mg/mL trimethoprim, 50 kanamycin, or 20 mg/mL tetracycline for Escherichia coli strains; 800 mg/mL trimethoprim or 250 mg/mL tetracycline for B. cenocepacia strains). For experiments, a phosphate-buffered mineral medium was used, either with or without ammonium (2 g/L NH 4  . Organic components were 5 g/L glycerol (Scharlab), 5 g/L yeast extract (Lab M), and 2 g/L Bacto peptone (BD Biosciences) when a medium with high concentrations of carbohydrates and amino acids was required (referred to as rich mineral medium). To grow strains under nitrogen depletion, mineral medium without ammonium was supplemented with 25 mM glucose as the carbon source. Media were supplemented with 600 mg/mL (strain J2315) or 200 mg/mL (strain K56-2) trimethoprim as the selective antibiotic when appropriate. Gene expression from plasmids was induced by adding rhamnose (Sigma) to a final concentration of 0.2% (wt/vol). All incubations were performed at 37°C.
Computational methods. Homologous sRNA and 59 UTR sequences were searched for using BLASTn (48), with the following adapted parameters: word size, 7; match/mismatch score, 3/22; gap existence cost, 2; gap extension cost, 2. Homologous porin genes were screened for using BLASTp and the Burkholderia genome database (12). Genes were defined as homologous to BCAL3473 at .95% query coverage and .60% amino acid identity and defined as belonging to the same porin family at .80% query coverage and .30% amino acid identity.
Computational target prediction was performed with CopraRNA (25), using default parameters. This algorithm takes accessibility of interaction sites and conservation of putative targets into account. Sequences from 200 nt upstream to 100 nt downstream of the first nucleotide of annotated genes were considered for target prediction. Only genes which were conserved in at least 50% of the input strains were reported as output. The results from LocARNA were then used as input for the RILogo software (51) to create sequence logos depicting conservation. Interactions between sRNAs and mRNAs were predicted with the IntaRNA tool (52). Construction of mutants. ncS25 was deleted from the B. cenocepacia J2315 genome by a method based on allelic recombination (53). Inserts flanking the sequence region targeted for deletion were amplified using PrimeSTAR GXL high-fidelity DNA polymerase (TaKaRa Bio) and cloned into the suicide vector pGPI-SceI-XCm. After successful recombination, pGPI-SceI-XCm was removed from the genome by introducing a double-strand break using the endonuclease Sce-I from the vector pDAI-SceI-SacB, followed by homologous recombination. Deletion mutants were screened for by PCR using primers annealing adjacent to the deleted sequence. Correct insert sequences and DNA junctions were confirmed by Sanger sequencing. All transformations of B. cenocepacia strains were performed by triparental mating (54,55).
Porin BCAL3473 was deleted from B. cenocepacia J2315 and K56-2 using the same method, resulting in mutants J-DBCAL3473 and K-DBCAL3473.
J-DBCAL3473 and K-DBCAL3473 were complemented by cloning the entire gene into vector pSCrhaB2 (56), where its expression is under the control of a rhamnose-inducible promoter, resulting in vector pL3473.
For complementing DncS25, ncS25, including 167 nt directly upstream and presumably containing the native promoter, was cloned into a modified version of pSCrhaB2 without the start codon and SD sequence, pSCrhaM2 (38), resulting in plasmid pncS25. This construct was then used as the template for further modifications by reverse PCR of the entire plasmid, resulting in pncS25mut, with point mutations in the predicted interaction region.
A knockdown mutant of regulator BCAL0461 was constructed by cloning a 376-nt fragment of the 59 end of BCAL0461 into the suicide vector pSC200 (57). The resulting plasmid, pSC200-L0461, was transformed into strain J2315, and ex-transformants were screened for correct insertion by PCR using primers annealing upstream of the BCAL0461 fragment in the vector and downstream of the fragment in the genome.
To construct a translational reporter gene fusion, the 59 UTR and the start of the coding sequence of BCAL3473 were fused to the fluorescent protein mCherry and cloned into the one available plasmid usable for cloning and in trans expression in B. cenocepacia. The fused protein would probably be transported into the periplasm, because the target sequence contains a transport tag. mCherry, unlike green fluorescent protein (GFP), is stable in the periplasm.
The entire gene for fluorescent protein mCherry was amplified from pmCherry (Clontech) and cloned into pSCrhaB2, resulting in plasmid pSCmCh-pos, which was used as a positive control for fluorescence protein production. mCherry without a start codon was amplified and cloned into pSCrhaM2, resulting in plasmid pSCmCh-neg, which was used for translational fusions and for negative controls. The entire 59 UTR and the first 84 nt of the BCAL3473 gene were cloned into plasmid pSCmCh-neg, resulting in plasmid pSCmChL3473. All constructs are under the control of a rhamnose-inducible promoter. The three plasmids were transformed into wild-type J2315 and its DncS25 mutant. Plasmid pSCmCh-neg was used as a negative control for fluorescence production without fused gene fragments.
Amplification of all plasmid inserts was performed with PrimeSTAR GXL high-fidelity DNA polymerase (TaKaRa Bio). The sequences of all plasmid inserts were verified by Sanger sequencing. All primer sequences are listed in Table S4.
Biolog screen and growth curves in microtiter plates. Biolog PM1 and PM2A plates were used as carbon sources according to the manufacturer's instruction with modifications. Cells were grown over night in LB broth in the presence of trimethoprim, washed, and resuspended in physiological saline (0.9% [wt/vol] NaCl). Cells were resuspended in IF-0 inoculating fluid to a final optical density (OD) of 0.1, corresponding to 10 8 CFU/mL. Biolog redox dye mix A (1Â) was added as a redox indicator and rhamnose (0.2% [wt/vol]) was added as an inducer to the IF-0 inoculating fluid in all experiments, and 100 mL per well was added to the Biolog plates. Absorption was read at 580 nm in a microplate reader after 24 h of incubation at 37°C (Victor Nivo, Perkin Elmer).
Growth on compounds as sole sources of nitrogen was tested in mineral medium lacking ammonium and containing 25 mM glucose as the carbon source. Nitrogen-containing compounds were added at a 2 mM final concentration. Growth on compounds as sole sources of carbon was tested in mineral medium containing ammonium. The test compound was added at a 5, 10, or 20 mM final concentration depending on the molecular weight and solubility of the compound, aiming at approximately 60 mM carbon equivalents.
Growth curves were determined in round-bottom 96-well microtiter plates, filled with 200 mL medium per well. Bacterial strains were grown overnight to an OD of 1.0 in the mineral medium without ammonium on 25 mM glucose and 0.5% Bacto peptone, to minimize nitrogen compound carryover, and then diluted to an OD of 0.001 (10 6 CFU/mL).
Absorption was read at 600 nm every 30 min in a microplate reader set to 37°C (Victor Nivo, Perkin Elmer). The MICs of antibiotics were determined in Mueller-Hinton broth in 96-well plates in duplicate, according to the EUCAST broth microdilution protocol. Wells with visible turbidity were scored as positive.
Translational reporter gene fusion assay. Mutant strains were grown overnight in mineral medium containing 2 g/L of each glucose and glycerol, supplemented with trimethoprim. The cultures were harvested by centrifugation and washed with the same mineral medium without trimethoprim. Cultures were normalized to an OD of 1.0, and 100 mL was added to the wells of a microtiter plate. One hundred microliters of LB broth with rhamnose to a final concentration of 0.1% was added to each well. The plate was incubated at 37°C, and fluorescence (excitation wavelength, 555 nm, and emission wavelength, 635 nm) was determined every 20 min using a microtiter plate reader (Envision multilabel reader; PerkinElmer).
qPCR. For evaluating gene expression in the DncS25 strain and the BCAL0461 knockdown mutant, cells were grown in rich mineral medium. Glass flasks (250 mL) with 25 mL medium were inoculated with a liquid overnight culture to approximately 4 Â 10 7 CFU/mL and incubated in a shaking incubator at 100 rpm until an OD of 0.5 was reached.
For evaluating the response to nitrogen starvation, strains were grown overnight in LB and then centrifuged, washed, and resuspended in mineral medium without ammonium. The washed cells were set to an OD of 1.0 and supplemented with either 25 mM glucose (nitrogen-depleted condition) or 25 mM glucose, 35 mM ammonium chloride, and 0.5% Bacto peptone (nitrogen-replete condition). Twenty-five milliliters of cells was further incubated in flasks on a shaker and harvested after 3 h.
Cells were snap cooled, pelleted by centrifugation at 4°C and stored at 280°C for a maximum of 1 week. RNA extraction, cDNA generation, and qPCR were performed as described previously (23). C q values were normalized to that of the control gene rpoD (BCAM0918), which had shown stable expression in previous studies (22)(23)(24).
One-way analysis of variance (ANOVA) with a Tukey post hoc test using SPSS v. 25 was performed to determine statistical significance. If variances were not equal, a nonparametric Kruskal-Wallis test was applied.
Primer sequences are listed in Table S4.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.