Burkholderia cenocepacia H111 was obtained from the Belgian co-ordinated collection of microorganisms. Cultures were grown in either LB-broth (Millers), or a modified Glycerol-glycerophosphate medium (GMM) with or without the addition of zinc [25]. Cultures were routinely grown aerobically at 37°C within an orbital shaking incubator.

The composition of GGM medium was as follows: 40 mM MES, 18.7 mM NH₄Cl, 13.4 mM KCl, 7.64 mM β-glycerophosphate, 5.0 mM glycerol, 4.99 mM K₂SO₄, 1.0 mM MgCl₂, 0.134 mM EDTA, 68.0 µM CaCl₂·2H₂0, 18.5 µM FeCl₃·6H₂O, 12.28 µM ZnSO₄·7H₂O, 1.62 µM H₃BO₃, 0.5 µM MnCl₂·4H₂O, 587 nM CuCl₂·2H₂O, 344 nM Co(NO₃)₂·6H₂O, and 80.9 nM (NH₄)6Mo₇O₂₄·4H₂O, in MilliQ (18 MΩ cm⁻¹) water. To prepare the medium, bulk components (MES, NH₄Cl, KCl, K₂SO₄, glycerol, H₃BO₃), dissolved in MilliQ water were passed through a Chelex 100 column (Bio Rad, United Kingdom) to remove contaminating metal cations, following the manufacturer’s instructions. CaCl₂ was subsequently added, and the solution sterilised in polycarbonate bottles using a microwave, to avoid introduction of contaminating zinc [26]. After sterilisation, trace metals dissolved in EDTA, MgCl₂, and β-glycerophosphate were added by filter sterilisation using 0.2 µm PTFE syringe filters. Syringes without rubber seals were used to minimise zinc contamination. Chemicals were purchased from either Merck or Fisher Scientific and were trace metal grade where available. To further minimise the potential for zinc contamination, culture flasks, measuring cylinders, and other plasticware was washed in 3% trace metal grade nitric acid overnight, and extensively rinsed using MilliQ water (18 MΩ cm⁻¹) before use.

Ten mL of overnight cultures of B. cenocepacia H111 in LB, were harvested by centrifugation at 3,000 g for 15 min, and the cell pellets washed by gently resuspending in GGM without trace metal additions but supplemented with 1 mM EDTA to remove surface adsorbed metals. Cells were pelleted again and further washed in trace metal free GGM to remove EDTA. After a final centrifugation step, cells were resuspended in 10 ml of complete GGM with or without the addition of zinc. Twenty µL of the resuspended culture was added to 180 µl of fresh GGM in 96-well plates. Plates were sealed using Breathe-Easy^® plate seals and incubated at 37°C with orbital shaking at 200 rpm in a CLARIOstar plate reader (BMG LABTECH). Optical density (600 nm) measurements were taken every 10 min for 20 h.

B. cenocepacia H111 was cultured in zinc depleted and zinc replete GGM and harvested at late exponential growth phase, by centrifugation at 3,000 g. Cell pellets were washed with EDTA as described above to remove any surface adsorbed metals. After washing, the cells were collected by centrifugation and dried to constant weight before being digested in 70% trace metal grade nitric acid for 7 days at room temperature. Following complete digestion, samples were diluted to 2% nitric acid in MilliQ water and passed through 0.45 µm PTFE filters. ICP-MS analysis was performed using a NexION 300X instrument (PerkinElmer). Calibration was performed with an external ⁶⁶Zn standard across the 0–1,000 ppb range. ⁴⁵Sc was used as the internal standard. The instrument was operated in Helium KED mode to remove any polyatomic interferences. Five technical replicates were taken for each sample analysed, and four biological replicates were performed.

RNA-sequencing and differential gene expression analysis was performed by GENEWIZ™. B. cenocepacia H111 was grown in zinc depleted and zinc replete GGM to mid-exponential growth phase, as described above. Cells were harvested by centrifugation at 3,000 g, and cell pellets stored at −80°C for 1 week before shipping on dry-ice to GENEWIZ™ for RNA extraction, RNA library preparation and RNA-sequencing. Total RNA was extracted using a Qiagen RNeasy Plus Mini Kit. rRNA depletion was performed using Qiagen FastSelect rRNA bacteria kit. RNA quantity and integrity were determined using a Qubit 2.0 Fluorometer, and Agilent 4200 Tapestation. RNA sequencing libraries of 150 bp reads, were prepared using a NEBNext Ultra II RNA library kit. Samples were paired end sequenced using an Illumina Hiseq instrument.

Data processing was performed by GENEWIZ™. RNA-sequence reads were trimmed using Trimmomatic v. 0.36, removing adapter sequences. The trimmed reads were mapped to the B. cenocepacia H111 reference genome [27], using STAR aligner v. 2.5.2b. Unique gene hit counts were calculated by using featureCounts v.1.5.2. Only unique reads within exon regions were counted. Gene hit counts were used for downstream differential expression analysis. Using DESeq2, a comparison of gene expression between cells grown in zinc replete and zinc depleted GGM was performed. p-values and Log2 fold changes were calculated using the Wald test. Genes with adjusted p-values < 0.05 and absolute log2 fold changes > 1 were called as being differentially expressed.

RNA was extracted from cell pellets using a GeneJET RNA Purification kit (Thermo Scientific™) and used a template for RT-qPCR. Gene sequences were extracted from the Burkholderia genome database [27], and primers designed using Primer3 v. 4.1.0 [28]. A full list of primers used is given in Supplementary Table S1. RT-qPCR was performed using Power SYBR™ green RNA-to-C_T 1-step kit (Applied Biosytems™), following the manufacturer’s instructions. PCR was performed using a Quantstudio 3 real time PCR system (Applied Biosystems™). Cycling was as follows: 1 cycle of 48°C for 30 min, 1 cycle of 95°C for 10 min, 40 cycles of 95°C for 15 s and 60°C for 60 s. Relative levels of gene expression were calculated using the 2^−ΔΔCt method using the gyrB gene as an internal control.

B. cenocepacia H111 was cultured in zinc replete and depleted GGM and growth monitored using OD₆₀₀ measurements as a proxy for cell numbers. The data (Figure 1A) reveals no significant difference in growth between the two conditions evaluated. The duration of the lag phase is comparable, and cells grown in zinc depleted and replete media reached similar final cell numbers with an average OD₆₀₀ of 0.54 and 0.59, respectively at 20 h. Average doubling times during exponential growth across both conditions was 226 min. That compares to a doubling time in LB-broth of 108 min (data not shown). Growth rates in LB in our lab are comparable to previously reported doubling times for Bcc species of between 70–186 min [29]. The result of this experiment is significant for two reasons. Firstly, the data shows that B. cenocepacia H111 is well adapted for growth in zinc-depleted environments such as those expected to be encountered in the host, and secondly because it demonstrates that GGM-medium is suitable for trace metal studies in Burkholderia species more generally.

The effect of zinc availability on total cellular zinc quotas was determined using Inductively coupled plasma mass spectrometry (ICP-MS). The data shows that that cellular zinc quotas are significantly reduced when B. cenocepacia H111 cells are grown in zinc depleted GGM compared to when they are grown in the replete medium (Figure 1B). Cells grown in the zinc replete GGM contained an average of 38.10 µg of zinc per g of dry cell weight, compared to just 11.35 µg of zinc per g of dry cell weight in cells grown in the zinc depleted medium. This result is surprising when taken in the context of the previous growth experiment that showed zinc depletion had no measurable effect on growth rates (Figure 1A), and together, these data suggest that B. cenocepacia H111 may either maintain significant zinc stores within the cell during replete conditions, or use “zinc sparing” as a mechanism to reduce overall cellular demand for zinc, when bioavailability is low. This phenomenon has been observed in other bacterial species where the expression of genes encoding non-essential zinc dependent proteins is repressed in response to zinc depletion [30].

RNA-sequencing was used to measure any changes in gene expression that occur in B. cenocepacia H111 in response to zinc depletion. A total of 201 genes showed significant differential expression, with 85 upregulated and 116 downregulated genes (Figure 2), representing 2.95% of predicted genes. The full list of significantly differentially expressed genes is given in Supplementary Table S2, and the 33 genes that had more than a Log2 fold increase in expression, are also listed in Table 1. The gene with the greatest fold increase in expression (locus tag I35_RS28540) encodes a hypothetical cytoplasmic protein [27]. A protein BLAST of the translated gene sequence reveals that the gene encodes a protein with significant sequence similarity to members of the COG0523 subfamily of P-loop GTPases that are predicted to play a role in metal homeostasis [31]. In the opportunistic pathogen Acinetobacter baumannii, the orthologous gene zigA is zinc binding and required for growth under zinc limited conditions [32]. ZigA from A. baumannii and the putative COG0523 protein from B. cenocepacia H111 share 75% sequence identity, and a conserved metal binding motif (CICC). Closer examination of the genomic neighbourhood of the COG0523 gene in B. cenocepacia, reveals that it sits in a cluster of several genes that are upregulated in response to zinc depletion (Figure 3; Table 1; Supplementary Table S2), and includes genes that encode the usually zinc dependent proteins carbonic anhydrase, dihydrooratase, and 6-carboxy-5,6,7,8-tetrahydropterin synthase (QueD), as well as FolE2, and Acyl-coA dehydrogenase.This data points to a key role for COG0523 in the response of B. cenocepacia to zinc limitation.

The RNA-seq data also revealed a set of upregulated genes encoding a “cation ABC transporter,” and a “Fur-family transcriptional regulator”, which based on protein BLAST searches are likely to encode the high affinity zinc uptake transporter ZnuABC and the zinc transcriptional repressor protein Zur. Figure 4A shows the genomic arrangement of these genes, and a putative Zur binding site that was identified upstream of zur that matches exactly to a consensus Zur-box motif for Burkholderia taken from the RegPrecise database (Figure 4B) [33]. It should be noted that it can be difficult to accurately determine the metal specificity of proteins using bioinformatics, and this is the first experimental evidence that this is the ZnuABC system in B. cenocepacia.

Other upregulated genes identified that may play a role in zinc uptake from depleted environments include two TonB-dependent outer membrane receptors and a Major Facilitator Superfamily (MFS) transporter (Table 1), although a specific role for these transporters in zinc uptake remains to be experimentally confirmed.

Intriguingly, the expression of genes related to phenylacetic acid degradation were significantly upregulated too (Table 1), with this catabolic pathway required for the full pathogenicity of B. cenocepacia in a Caenorhabditis elegans infection model [34].

Amongst the genes that are downregulated in response to zinc depletion are some genes encoding usually zinc dependent proteins including aldolase, peptidase M23 and a dihydroorotase (Table 2; Supplementary Table S2), providing further evidence that B. cenocepacia H111 uses a mechanism of zinc sparing to reduce overall cellular demand for zinc during conditions of zinc deprivation. Also noteworthy was the downregulation of a membrane ion permease and a predicted metal transporting ATPase, possibly to reduce zinc efflux from the cell (Table 2; Supplementary Table S2).

The RNA-seq data were validated using RT-qPCR to independently measure changes in the expression of six genes in response to zinc depletion. Several of the genes selected (tonB-dependent receptor, COG0523, zur, znuA and znuC), are putative components of the zinc homeostasis system in B. cenocepacia H111 identified in this study. Good agreement was observed between the two data sets (Figure 5).

• B. cenocepacia is an opportunistic pathogen, responsible for severe respiratory infections in people with cystic fibrosis.

• B. cenocepacia is remarkably well adapted for growth in zinc poor environments.