Streamlined copper defenses make Bordetella pertussis reliant on custom-made operon

Abstract

Copper is both essential and toxic to living beings, which tightly controls its intracellular concentration. At the host-pathogen interface, copper is used by phagocytic cells to kill invading microorganisms. We investigated copper homeostasis in Bordetella pertussis, which lives in the human respiratory mucosa and has no environmental reservoir. B. pertussis has considerably streamlined copper homeostasis mechanisms relative to other Gram-negative bacteria. Its single remaining defense line consists of a metallochaperone diverted for copper passivation, CopZ, and two peroxide detoxification enzymes, PrxGrx and GorB, which together fight stresses encountered in phagocytic cells. Those proteins are encoded by an original, composite operon assembled in an environmental ancestor, which is under sensitive control by copper. This system appears to contribute to persistent infection in the nasal cavity of B. pertussis-infected mice. Combining responses to co-occurring stresses in a tailored operon reveals a strategy adopted by a host-restricted pathogen to optimize survival at minimal energy expenditure.

Fig. 1. Effect of copper on B. pertussis and B. bronchiseptica growth.

Growth curves of B. pertussis BPSM and B. bronchiseptica RB50 in SS medium supplemented (blue curves) or not (orange curves) with 5 mM CuSO₄ are shown. The turbidity of the cultures was measured using an Elocheck device. The values shown in ordinate (absorbency units) do not correspond to optical density measurements classically obtained with a spectrometer, since the Elocheck instrument uses distinct pathlength and wavelength.

Fig. 2. Copper regulation of homeostasis systems in B. pertussis and B. bronchiseptica.

(a, c, and d) RNAseq analyses of B. pertussis grown for 16 h in the presence of 2 mM CuSO₄ in SS medium (a), B. pertussis grown in SS medium and treated for 30 min with 2 mM of CuSO₄ (c), and B. bronchiseptica grown for 12 h in the presence of 2 mM CuSO₄ (d). Comparisons were made with bacteria grown in standard conditions. Each gene is represented by a dot. The x and y axes show absolute levels of gene expression in reads per kilobase per million base pairs (RPKM) in standard and copper conditions, respectively. The genes indicated in blue indicate genes of interest with the strongest regulation factors. The full sets of data are shown in Supplementary Tables S1, S2 and S3. (b) Summary of the transcriptomic and proteomic analyses performed after growing bacteria as in (a) and (c) in medium supplemented with 2 mM CuSO₄. Standard culture conditions were used for comparisons. * indicates small proteins difficult to detect by global proteomic approaches.

Fig. 3. Identification of CopZ in B. pertussis.

a Lysates of wild type B. pertussis (wt) or the deletion mutant (∆bp1727) grown in SS medium supplemented (+ Cu) or not with 2 mM CuSO₄ were subjected to SDS-PAGE electrophoresis in Tris-tricine gels for the detection of CopZ. The gel was stained with colloidal Coomassie blue dye. CopZ migrates below the 10 kDa band of the markers. b, c The protein band was cut from the gel, and CopZ was identified by mass fingerprinting analyses. The m/z ratios of the peptides identified by MALDI-TOF are shown in panel b, and the sequences of the peptides are in c. This experiment complements the proteomic analyses, because the small size of CopZ hampered its detection in global proteomic approaches (see Fig. 2b and Supplementary Data S4).

Fig. 4. Role of the operon in B. pertussis and in host–pathogen interactions.

a, b Growth yields of B. pertussis after 24 h in SS medium (a) or in SS medium devoid of glutathione (b), supplemented or not with 2 mM CuSO₄. The cultures were inoculated at initial OD₆₀₀ values of 0.1, and after 24 h the OD₆₀₀ was determined. Note the different y axis scales between (a) and (b). wt, wild type (parental) strain; ∆27, ∆28-29 and ∆27-29, KO mutants for bp1727 (copZ), bp1728-bp1729 (prxgrx-gorB) and the three genes, respectively. The various strains are represented using different symbols and shades of gray. ∆27-29/+27-29 represents the latter mutant complemented by expression of the operon at another chromosomal locus. The horizontal orange lines represent mean values. c Survival of the same strains to an oxidative shock of 30 min. d Intracellular survival of B. pertussis in THP1 macrophages. The horizontal orange lines represent mean values. See Supplementary Figure S6 for the data of the individual copZ and prxgrx-gorB KO mutants. e, f Colonization of mice lungs (e) and nasopharynxes (f) after nasal infections with the parental strain and the copZ-prxgrx-gorB KO mutant. The numbers of bacteria are indicated for each mouse and organ (black circles: parental bacteria, red squares: mutant). The lines connect the geometric means of the counts at each time point, and the dotted lines indicate the thresholds of detection. For all the assays, statistical analyses were performed using two-tailed Mann-Whitney tests (*, p < 0.05; ** p, <0.005). For panels a and b, 5 biologically independent samples were used. For panels c and d, 4 and 6 biological samples were used, respectively. For panels e and f, 5 and 4 animals per time point were used for the KO and wt strains, respectively.

Fig. 5. Activities of the proteins coded by the bp1727-1728-1729 (copZ-prxgrx-gorB) operon.

a The metal/protein ratios of recombinant CopZ were determined by ICP-AES. The apo form of CopZ was incubated or not with iron (gray) or copper (black), and both ions were measured. Three independent samples were processed. b Oxidation of NADPH by recombinant GorB over time. H₂O₂ was added at the indicated concentrations to generate the glutathione disulfide (GSSG) substrate of GorB, before adding the enzyme. The reaction was followed by the decrease of absorbency at 340 nm. c Plot of the reaction rate as a function of substrate concentration. A k_cat /Km value of 323,000 ± 46,000 M⁻¹s⁻¹ for GSSG was estimated based on Michaelis–Menten kinetics. One or two measurements were performed at each substrate concentration. d Effect of recombinant PrxGrx on the rate of oxidation of NADPH by GorB. The red and blue curves show reaction rates when GorB was present alone or when both enzymes were present, respectively. H₂O₂ was added last as a substrate of the first reaction. However, as H₂O₂ also generates GSSG, which initiates the second reaction, PrxGrx activity was detected by the increased rate of the reaction when both enzymes are present, and no enzymatic constants could be determined. (e) Schematic representation of the functions of the three proteins. By chelating Cu¹⁺, CopZ prevents the ion from generating hydroxyl radicals through the Fenton reaction. H₂O₂ is reduced to H₂O through the activity of PrxGrx, at the expense of reduced glutathione (GSH). The product of that reaction, glutathione disulfide, is reduced through the activity of GorB at the expense of NADPH.

Fig. 6. Regulation of the operon.

a, b qRT-PCR analyses of the parental and the cueR KO strains treated for 30 min with 2 mM CuSO₄ (a) or 10 mM H₂O₂ (b), showing the expression levels of prxgrx relative to untreated controls. Data were normalized with the housekeeping gene bp3416. Three biological replicates were performed. (c) EMSA with recombinant OxyR and DNA fragments of the cueR-copZ and copZ-prxgrx intergenic regions, IGR 1 and IGR 2, respectively. The uncropped gel is shown in Supplementary Figure S10. (d) Schematic representation of the locus, with sequences of the putative CueR and OxyR boxes. The site of transcription initiation was determined by 5’RACE (Supplementary Figure S5). The putative OxyR binding sites were identified by their similarity with the E. coli consensus sequences, and alignments of the Bordetella and Achromobacter sequences were used to build the consensus motif (Supplementary Figure S11).

Fig. 7. Upregulation of the operon in macrophages.

a Representative images of THP1 macrophages having engulfed B. pertussis harboring the mRPF1 gene under the control of the copZ-prxgrx-gorB operon promoter. Macrophages were either starved of copper using a chelator or treated with copper chloride prior to contact with bacteria. The bacteria are red, cell membranes are green, and nuclei are blue. In the two fields shown at the left, the bacteria were projected as objects, showing that similar numbers are present in both conditions. All image panels are shown at the same magnification. b Levels of fluorescence (arbitrary units) of intracellular bacteria in the two conditions. Statistical analyses were performed using a two-tailed Student T-test (****, p < 0.0001). The horizontal orange lines represent median values (n > 500 in both cases).