To generate the PA14 Δbc₁ mutant, a lambda-red recombinase method was used (Bru et al., 2019; Datsenko and Wanner, 2000). P. aeruginosa PA14 cells were transformed with pUCP18-RedS as previously described (Lesic and Rahme, 2008) and selected in LB medium + carbenicillin 300 µg/mL. The lambda-red system in cells harboring pUCP18-RedS was induced with 10 mM arabinose. Cells were pelleted at 16,000 × g for 10 minutes at room temperature and were washed with 300 mM sucrose. Electrocompetent cells were transformed with the mutagenic amplicon into pUCP18-RedS-positive cells in a 0.4 cm gap cuvette at 2500 V using a MicroPulser Electroporator (Bio-Rad, Hercules, CA) as previously reported (Bru et al., 2019). For the mutagenic amplicon, regions of 500–650 bp upstream and downstream of the operon encoding cytochrome bc₁ (petA-cytB-cytC₁) were amplified using the primer pairs listed in Table 1. The gentamicin cassette of pAS03 was amplified using the primers Gm_pAS03_F and Gm_pAS03_R (Table 1). The three amplicons were combined and amplified via fusion PCR using the Bc₁ Upstream Bc1_Mut_F and Bc₁ Downstream Bc1_Mut_R primers. Mutants were selected on gentamicin plates (30 µg/mL) and verified by PCR using the primers Bc1_F and Bc1_R and by nanopore bacterial genomic sequencing (Eurofins Genomics, Louisville, KY). Our sequencing data show a single elimination of the bc₁ operon with no insertions or deletions in any other site. The genome sequence of the mutant is available upon request.

P. aeruginosa strains PA14 (wild-type) and PA14 Δbc₁ were grown in LB Broth (Miller), modified artificial urinary medium (mAUM), and lung medium (LM) at 37°C with shaking at 250 RPM as reported previously (Hu et al., 2024; Liang et al., 2020; Ruhluel et al., 2022). mAUM and LM were prepared as described previously (Hu et al., 2024; Liang et al., 2020; Ruhluel et al., 2022). Growth was assessed by following the absorbance of cell cultures at 620 nm using an 800 TS microplate reader (BioTek, Winooski, VT). Maximum growth, duplication time, and lag phase duration were calculated by fitting the growth data to logistic functions, as previously detailed in our work (Liang et al., 2020; Kahm et al., 2010). Experiments were conducted in triplicate, with results reported as mean ± SD, n = 3. Statistical significance between PA14 and PA14 Δbc₁ was assessed using t-test analysis.

5637 human urinary bladder carcinoma cells (ATCC HTB-9) were cultured in RPMI 1640 medium supplemented with 10% Fetal Bovine Serum (FBS). Following treatment with 0.25% trypsin-EDTA, 2.0 x 10⁵ cells were seeded in 24 well plates and incubated for 48 hours at 37 °C with 5% CO₂ in a humidified atmosphere prior to infection. P. aeruginosa strains PA14 and PA14 Δbc₁ were grown for 24 hours in LB at 37 °C under continuous agitation (250 RPM). Bacterial cells were centrifuged at 16,000 × g for 10 minutes at room temperature and washed with PBS. Monolayers of 5637 cells were washed with PBS, and fresh RPMI 1640 supplemented with 10% FBS was added. Bladder cells were then infected with PA14 or PA14 Δbc₁ at a MOI of 10 for 2, 4, and 6 hours. At each time point, cells were washed with PBS and stained with a 1:1 mixture of PBS:trypan blue (0.4%). Bright-field images were collected using a Keyence BZ-X710 Fluorescence Microscope (Itasca, IL) with a cooled monochrome CCD camera and Nikon Plan Fluor ELWD 20x/0.45 Ph1 ADM, ∞/0–2 WD 8.2–6.9 lens. Quantification of cell viability post-infection was performed using the BZ-X analyzer (version 01.03.00.05) and ImageJ v1.54p (Abramoff et al., 2004), with assessors blinded to outcomes. Viability was quantified by calculating the percentage of live cells (unstained) from the total cells [unstained plus stained (dead)]. Results are reported as mean ± SD, n = 3, with statistical significance between PA14 and PA14 Δbc₁ assessed via t-test.

Following 5637 cell infection with P. aeruginosa strains PA14 or PA14 Δbc₁, expression of the type III secretion system (T3SS) proteins exoT, exoU, and exoY was assessed in the media of the infected cell culture at 4 hours post-infection (hpi). Supernatant total RNA was extracted using TRIzol LS (Thermo Fisher Scientific, Waltham, MA) and quantified on a Qubit 3.0 fluorometer using the AccuBlue Broad Range RNA Quantitation Kit (Biotium, Fremont, CA), followed by cDNA synthesis with the ProtoScript II First Strand cDNA Synthesis Kit using the Random Primer Mix (New England Biolabs, Ipswich, MA). qPCR was subsequently performed using the primer pairs listed in Table 2, designed using Primer 3 software (Koressaar and Remm, 2007; Untergasser et al., 2012; Kõressaar et al., 2018). P. aeruginosa rpoD was utilized as a housekeeping gene due to its vital role in promoter identification to facilitate constitutive gene expression (Potvin et al., 2008; Savli et al., 2003). cDNA (100 ng) amplification and quantification were performed on a Bio-Rad CFX Connect Real-Time PCR Detection System using the iTaq Universal SYBR Green Supermix kit (Bio-Rad, Hercules, CA). mRNA expression was determined via a 2-ΔΔCT method comparing rpoD, exoT, exoU, and exoY expression in the presence of P. aeruginosa PA14 or PA14 Δbc₁ infection via cycle threshold analysis (Livak and Schmittgen, 2001). Data are presented as mean ± SE, n = 9, with statistical significance determined via t-test.

P. aeruginosa infection in mice was performed using a sepsis model, as described previously (Six et al., 2021). PA14 and PA14 Δbc₁ were grown in LB medium for 24 hours, followed by centrifugation at 16,000 × g for 10 minutes at room temperature. Cells were washed with PBS and eight-week-old female BALB/cJ mice (The Jackson Laboratory, Bar Harbor, ME) were inoculated with 50 µl of PBS (n = 7), 1.25-1.64 x 10⁷ CFUs of PA14 (n = 17), or 1.50-1.60 x 10⁷ CFUs of PA14 Δbc₁ (n = 14) intravenously through lateral tail vein using a 29G syringe, according to Illinois Institute of Technology IACUC-approved protocols. Animals were provided with unrestricted access to food and water and kept on a standard 12 h light/dark cycle with lights off at 7:00 p.m. Animals’ weights and survival were monitored twice daily for one week post-infection. Survival outcomes were visualized with Kaplan-Meier curves and statistically evaluated using the log-rank test implemented in OriginLab (version Origin 2025b (10.25), OriginLab Corporation, Northampton, MA).

Biofilm formation was assessed for P. aeruginosa strains PA14 and PA14 Δbc₁, as described previously (O’Toole, 2011). Bacterial overnight cultures were grown in LB at 37 °C with shaking (250 RPM). One milliliter of bacteria culture was centrifuged at 16,000 × g for 10 minutes at room temperature, and cells were washed with PBS twice. Bacterial pellets were resuspended in 1 mL of PBS and adjusted to OD₆₀₀ of 1.0. This suspension was further diluted to an OD₆₀₀ of 0.01 in LB, mAUM, or LM. 100 uL of each suspension was transferred to individual wells of 96-well plates. Experiments were performed three times using 10 technical replicates per condition and strain. Plates were incubated at 37 °C for 48 hours without shaking. Following incubation, supernatant was removed, and wells were washed three times with PBS and allowed to air dry. Biofilm was stained with 1% filtered crystal violet for 30 minutes, dye was removed, and wells were washed with ultrapure water. Retained crystal violet was solubilized with 33% acetic acid and absorbance was read at 595 nm using an 800 TS microplate reader (BioTek, Winooski, VT). Statistical significance between PA14 and PA14 Δbc₁ was determined via t-test.

ATP detection was performed using the BacTiter-Glo™ Microbial Cell Viability Assay (Promega, Madison, WI) according to manufacturer’s instructions. Cultures of P. aeruginosa strains PA14 and PA14 Δbc₁ harvested during both logarithmic and stationary phases following growth in LB, mAUM, and LM were vortexed and combined with BacTiter-Glo™ reagent in a 1:1 ratio in 96-well white opaque-walled plates. For cultures grown in LB, mAUM, and LM, logarithmic growth was defined at 7, 7, and 16 hours post-inoculation, respectively, whereas stationary phase was designated at 24, 12, and 24 hours for LB, mAUM, and LM, respectively. Culture reagent mixtures were incubated for five minutes and luminescence was recorded. ATP standard curves were prepared in LB, mAUM, and LM and obtained values were interpolated based on the appropriate media to obtain moles of ATP in each sample. Protein was quantified via BCA assay (Thermo Fisher Scientific, cat. no. 23225, Waltham, MA) and ATP quantity was subsequently normalized to protein concentration. Experiment was performed in triplicate, with results reported as mean ± SD, n = 3. Statistical significance between PA14 and PA14 Δbc₁ was determined using a t-test.

ROS detection was carried out according to Fraser-Pitt et al. (Fraser-Pitt et al., 2018). P. aeruginosa strains PA14 and PA14 Δbc₁ were grown in LB, mAUM, and LM at 37 °C with shaking at 250 RPM and harvested in the logarithmic and stationary phases of growth. Samples were subsequently centrifuged at 16,000 × g for 10 minutes and washed with PBS at room temperature. H₂DCFDA (2’,7’-dichlorodihydrofluorescein diacetate) (Invitrogen, cat. no. D399, Waltham, MA) was added to 100 µL cell samples at a concentration of 1 µM in a 96-well black clear bottom plate and fluorescence was assessed at 480/520 nm excitation/emission using a SpectraMax iD5 Multi-Mode Microplate Reader (Molecular Devices, San Jose, CA). Experiments were performed in triplicate. Data were normalized to sample protein concentration, quantified via BCA assay, and reported as relative light units (RLU)/µg protein. Statistical significance was determined via t-test analysis.

Membrane potential was assessed with JC-1 dye (Invitrogen, cat. no. T3168, Waltham, MA) in P. aeruginosa strains PA14 and PA14 Δbc₁ harvested during both logarithmic and stationary phases of growth in LB, mAUM, and LM. JC-1 (10 µM) was added to harvested bacteria (2.5 µg of protein) in a 96-well black clear bottom plate and incubated for 10 minutes at room temperature in the dark, with or without oligomycin (1 µM) and carbonyl cyanide m-chlorophenyl hydrazone (CCCP) (5 µM). The conditions in which membrane potential formation were analyzed include basal metabolic state (control), presence of oligomycin (1 µM) (Symersky et al., 2012) to induce hyperpolarization by inhibiting ATP synthesis, and the presence of the protonophore CCCP (5 µM) (Daniels et al., 1981) to depolarize the membrane. Fluorescence was measured at 480/520 and 535/596 nm excitation/emission for the green and red fluorescent states of the dye. Results were presented as the ratio of red to green fluorescence. Cell autofluorescence (red and green) was subtracted for all samples. After conducting the experiment in triplicate, the data were analyzed using t-test analysis.

Membranes were obtained from P. aeruginosa PA14 Δbc₁ grown in mAUM during the logarithmic stage of growth, as previously described (Hu et al., 2024; Liang et al., 2020). Cell pellets were washed with KHE buffer (150 mM KCl, 20 mM HEPES, 1 mM EDTA, pH 7.5) and stored at -80 °C until further use. Bacterial cells were thawed and resuspended in KHE buffer with 5 mM MgCl₂, 10 µg/mL DNase, and 1 mM PMSF. Cells were ruptured at 16,000 PSI using an Avestin Emulsiflex-C5 high pressure homogenizer (Ottawa, ON, Canada). Debris was separated from membranes via centrifugation at 10,000 × g (4 °C) and cell membranes were subsequently collected from supernatant via ultracentrifugation at 100,000 × g (4 °C). Membrane fractions were washed with KHE buffer and stored at -80 °C.

The respiratory activity of P. aeruginosa PA14 Δbc₁ membranes harvested in the logarithmic phase following growth in mAUM was measured in a customized 1.6 mL glass oximetric chamber using a Clark-type oxygen electrode (YSI 5300) at 37 °C using the methods previously detailed (Hu et al., 2024; Liang et al., 2020) in triplicate. Assays were performed in KHE buffer with and without 1 µM rotenone to inhibit complex I, with 200 µM NADH or deamino-NADH to test for NADH oxidase activity, 10 mM lactate to test for lactate dehydrogenase activity, and 1 mM succinate to test for succinate dehydrogenase activity. Ubiquinol oxidase activity was determined using 50 µM ubiquinone-1, 500 µM 1,4-Dithiothreitol (DTT), and various concentrations of KCN to inhibit terminal oxidases.

As reported in our previous work, cytochrome bc₁ is critical for aerobic metabolism, ATP synthesis, and ion pumping in P. aeruginosa (Hu et al., 2024; Liang et al., 2020). To assess the role of this enzyme in supporting proliferation under physiologically relevant conditions, growth of wild-type PA14 and the deletion mutant was assessed in media resembling the environment of common sites of infection, such as the urinary tract (mAUM) and the respiratory epithelium (LM), and compared to growth in nutrient-rich laboratory LB medium (Figure 1). Previous work was carried out in a PA14 Δbc₁ transposon mutant (Hu et al., 2024), which was not completely characterized at the genomic level, and may carry additional genetic modifications. In this work, we characterized the PA14 strain in which the bc₁ gene was specifically eliminated using the lamba-red recombination system. Our group has previously validated mAUM for the analysis of adaptations in urine-like conditions (Liang et al., 2020), while LM formulation has been used to study the growth of other pathogens in synthetic lung media (Ashworth et al., 2024; Ruhluel et al., 2022). Determining P. aeruginosa growth in urine-like medium and lung medium is critical for understanding how this pathogen adapts to the unique nutrient conditions of these environments that mimic infection sites, which can inform treatment strategies for chronic and acute urinary and respiratory infections such as (CA)UTIs and cystic fibrosis infections.

The growth of the wild-type and Δbc₁ mutant was followed in LB (Figure 1A), mAUM (Figure 1B), and LM (Figure 1C). In LB, the mutant strain exhibited significantly lower growth and reached a maximum OD₆₂₀ 32% lower than the wild-type (Figure 1D). Moreover, duplication time was reduced by 33% in the mutant (Figure 1E) and the lag phase duration was increased 1.5-fold (Figure 1F). These results confirm the important role of the bc₁ complex in this growth condition, as shown from our previous work (Hu et al., 2024). Similarly, in urine-like conditions (mAUM), maximal bacterial growth was decreased in the mutant, approximately 16% (Figures 1B, D). Moreover, the duplication rate was nearly half compared to wild-type strain (Figure 1E), while its lag phase was almost doubled (Figure 1F). The most pronounced difference in growth between the wild-type and the mutant is observed in LM (Figure 1C). The mutant had a 50% reduction in total growth (Figure 1D), more than doubling of the lag phase (Figure 1F), and a reduction in the growth rate >50% (Figure 1E). The behavior exhibited by the mutant in these conditions suggests a hindered capacity to produce energy under physiologically relevant environments. This impairment may be due to the disruption of the electron transport chain in the mutant, highlighting the essential role of cytochrome bc₁ in maintaining energy metabolism across diverse conditions. Collectively, these findings emphasize the importance of cytochrome bc₁ as a key respiratory component that enables P. aeruginosa to adapt and proliferate in both urinary environments and nutrient-limited lung conditions.

To characterize the role of cytochrome bc₁ in P. aeruginosa virulence and pathogenicity, 5637 human epithelial bladder cells were infected with either wild-type or Δbc₁ mutant strains, and bladder cell viability was assessed (Figure 2B). At 2 hpi, neither strain impacted bladder cell survival, indicating that this timeframe is necessary for the bacterial cells to express virulence factors. At 4 hpi, the viability of cells infected with wild-type P. aeruginosa drops sharply to 23% (Figures 2A, B). On the other hand, 62% of cells infected with the mutant strain are still viable (Figures 2A, B). By the 6-hour time point, the wild-type strain eliminated all viable human cells, while in the mutant 12% of the infected cells are still viable. Even though cellular viability decreases over time in both conditions, virulence is clearly diminished by the lack of cytochrome bc₁, which may be due to alterations in bacterial virulence factor production. Thus, we sought to investigate how the mutation attenuates pathogenicity during host cell infection. Following 6-hour bladder cell infection with PA14 or PA14 Δbc₁ at MOI 10, culture media was collected, filtered (0.2 µm), and applied to uninfected bladder epithelial cells for 6 hours. Subsequent evaluation of cellular viability shows that neither wild-type nor Δbc₁ mutant (Figures 2C, D) produce soluble factors that induce bladder cell death in the conditions tested, as the bladder cells exposed to the filtered media retained 100% cell viability (Figure 2D). This suggests that the virulence factor(s) downregulated in the mutant requires cell-to-cell contact-mediated mechanisms.

P. aeruginosa possesses a type III secretion system (T3SS) that is critical for contact-based cytotoxicity, enabling the injection of proteins into the host cell cytoplasm (Hauser, 2009; Stover et al., 2000). PA14 secretes exoT, exoU, and exoY effector proteins, which function distinctly yet complementarily throughout the infection process (Filloux, 2011). ExoU is the major virulence factor in vivo and in vitro due to its phospholipase activity, which compromises the host membrane, while exoT and exoY have been shown to play a minimal role during infection (Hauser, 2009; Kazmierczak and Engel, 2002; Krall et al., 2000; Lee et al., 2005; Phillips et al., 2003; Sato et al., 2003). Relative fold-change analysis of T3SS effector gene expression (Figure 3A) showed a 53% decrease in exoT and a 60% decrease in exoU expression in the deletion mutant strain. However, exoY expression remained unchanged between the strains. Interestingly, it was previously reported that P. aeruginosa strains lacking the gene encoding for exoY or failing to express the protein were more toxic to lung epithelial cells than strains expressing the toxin (Silistre et al., 2021), indicating that exoY presence is not detrimental to host cells. These results demonstrate that deletion of cytochrome bc₁ severely hinders P. aeruginosa virulence (Figures 2B, 3A) by downregulating expression of the most cytotoxic effectors exoT and exoU, plausibly due to its compromised respiratory chain function and thus bioenergetic processes. The P. aeruginosa T3SS needle complex relies on PscN, an ATPase that drives the translocation of effector proteins into host cells (Hauser, 2009). Deletion of cytochrome bc₁ in PA14 disrupts oxidative phosphorylation and diminishes cellular ATP generation (see below), thereby limiting the energy supply required for T3SS function. As a result, the molecular syringe may operate less efficiently, leading to reduced effector protein secretion and attenuated host cell damage.

To further evaluate the impact of cytochrome bc₁ deletion on P. aeruginosa pathogenicity, the effects of the mutant were assessed on a systemic infection model in mice, evaluating survival following infection with wild-type or Δbc₁ cells. As shown in Figure 3B, <20% of mice infected with wild-type PA14 survived beyond seven days post-infection. Remarkably, deletion of cytochrome bc₁ increased survival to 93%, indicating that this enzyme is essential for virulence in vivo. The deletion mutant is completely avirulent, as the survivability did not differ statistically from that of the PBS control group (Figure 3B). This effect may be attributed to the bacterium’s energy deficit paired with lack of virulence factor production, resulting in an inability to resist host immune response. Indeed, exoU has been associated with phagocyte killing (Diaz et al., 2008), thereby compromising host defense mechanisms. Consequently, diminished exoU expression in the Δbc₁ mutant strain may mitigate virulence and would promote bacterial clearance in mice.

Biofilm is a critical virulence factor of P. aeruginosa, contributing to both infection establishment and multidrug resistance in different tissues, including the lung and urinary tract (Maurice et al., 2018). In this work, we further characterized the role of the bc₁ complex in pathogenicity by evaluating biofilm formation in physiologically relevant media. As shown in Figures 3C, in LB medium, both wild-type and mutant strains produce comparable amounts of biofilm. In contrast, a significant difference between the wild-type and mutant is observed in LM, where the mutant had a seven-fold decrease in biofilm formation. Interestingly, in mAUM, the cells could not produce significant amounts of biofilm. These findings underscore the critical role of the bc₁ complex in facilitating biofilm formation in LM, suggesting that the enzyme may serve as a promising therapeutic target for disrupting P. aeruginosa in biofilm-associated infections, which may be particularly relevant in lung infections, in conditions such as cystic fibrosis.

As shown in our previous studies, the electron transport chain is essential for ATP generation in P. aeruginosa, contributing to the formation of the proton gradient, which drives energy production (Mitchell, 2011). ATP levels of PA14 and PA14 Δbc₁ were measured in cells cultured in LB, mAUM, and LM to determine the relevance of cytochrome bc₁ in ATP generation under physiologically relevant conditions (Figure 4A). In laboratory LB media, no difference was found in ATP content between the wild-type strain and mutant in stationary or logarithmic phases, which is consistent with the minor changes in growth found in this condition. However, in mAUM, the ATP content of the Δbc₁ mutant was 44% less than the wild-type strain during stationary growth. This effect was more pronounced during the logarithmic phase of growth in mAUM, as the mutant produced 92% less ATP than the wild-type strain. In lung media, ATP content was severely decreased in the Δbc₁ mutant in both stationary and logarithmic phases, by 86% and 79% compared to the wild-type strain, respectively. These results clearly demonstrate that the bc₁ complex, while not absolutely essential for growth, is critical in P. aeruginosa bioenergetics and physiology. Remarkably, the effects of the mutant on ATP content closely correlate with the effects seen in the growth parameters under physiologic culture conditions as shown in Figure 1. Altogether, the data indicate that cytochrome bc₁ is highly relevant to energy generation in physiologic media, which explains the almost complete lack of virulence of the mutant in animal models.

As reported in the literature, cytochrome bc₁ is one of the main contributors to reactive oxygen species production (Brand, 2016; Seaver and Imlay, 2004). To fully characterize the metabolic role of the bc₁ complex in P. aeruginosa, ROS production was evaluated in the mutant and wild-type strain. Our results show that ROS production was significantly modified in the mutant strain in all three media tested (Figure 4B). In LB medium during stationary and logarithmic stages of growth, ROS production by PA14 Δbc₁ significantly decreased compared to the wild-type strain, by 78% and 60%, respectively. In LM, there was no difference between the two strains during logarithmic growth, while a nearly 10-fold decrease in ROS production was observed in the mutant strain during the stationary phase. The decrease in ROS production by the mutant in these conditions is expected, as the bc₁ complex is a main contributing factor to oxygen radicals (Pagacz et al., 2021). Surprisingly, in mAUM, ROS production by the mutant strain increased significantly compared with PA14, showing a 3.8- and 2.2-fold increase in stationary and logarithmic growth phases, respectively. This increase likely reflects the pathogen’s reliance on cytochrome bc₁ in mAUM, where it drives ion pumping and respiration (Hu et al., 2024). The lack of the Pseudomonas quinolone signal (PQS) in the mutant (Soh et al., 2021), which is linked to the generation of anti-oxidative responses (Häussler and Becker, 2008), may also contribute to this surge in ROS in mAUM.

As discussed above, our previous work shows that cytochrome bc₁ is a critical contributor to ion pumping in the P. aeruginosa respiratory chain (Hu et al., 2024), which links the operation of the respiratory chain to ATP production. To evaluate the role of cytochrome bc₁ on membrane potential, the fluorescent signal of the voltage-dependent JC-1 dye was studied in the wild-type and Δbc₁ mutant in LB, mAUM, and LM (Figures 4C-E). The addition of oligomycin produced relatively small changes in membrane potential and thus the data were omitted from this work for clarity. In LB, wild-type cells produced a significant membrane potential in the stationary phase, which decreased in the logarithmic phase. In this condition, no significant differences were found in the Δbc₁ mutant (Figure 4C), which correlates with the small differences observed in growth and ATP content. This effect could be due to the significant role that other respiratory enzymes, such as NQR, bd oxidase, and bo₃ oxidase, play in membrane potential generation, which could compensate for lack of the bc₁ complex, as discussed in a previous work (Hu et al., 2024). In mAUM (Figure 4D), the membrane potential produced by wild-type cells during stationary phase nearly doubled compared to that of LB, and did not decrease in the logarithmic state, which could be due to the activation of the respiratory chain observed in these conditions (Hu et al., 2024). Remarkably, in mAUM, the Δbc₁ mutant did not exhibit a difference in membrane potential compared to the wild-type strain. Finally, membrane potential measurements for cells cultured in LM were carried out in fresh LM instead of PBS, which was used for other conditions, as LM-grown cells displayed an undetectable membrane potential when measured in PBS. The data shown in Figure 4E indicate that in this condition, the cells produce a small but detectable membrane potential, which was significantly increased in the mutant during logarithmic growth. Experiments were also carried out with DiOC₂(3), another widely used membrane potential indicator (Novo et al., 1999; Pan et al., 2017), obtaining similar results (not shown). It is likely that the nutrient composition of LM does not allow for a very active metabolism, reflected in the lowest growth rate of all conditions (approximately 20% compared to LB medium). In spite of the significant changes in the metabolic parameters that we have measured, including ATP content, in the case of membrane potential, the mutant only showed a significant increase compared to the wild-type strain during logarithmic growth in LM, which indicates that the respiratory chain in the mutant is adapted to the lack of the bc₁ complex in most conditions, likely by decreasing the expression of pathways that require active H⁺- transport, which would result in diminished growth and an avirulent phenotype.

As shown in this manuscript and our previous work, the bc₁ complex has a critical role in cell physiology. Nonetheless, this enzyme is not essential for P. aeruginosa, as the cells are able to adapt to its elimination and can grow in all physiologic media tested, although at lower rates. However, the significant metabolic challenge imposed by this deletion has a deleterious effect on virulence. To understand the adaptability of P. aeruginosa, the respiratory chain of the mutant was fully characterized as previously described (Hu et al., 2024; Liang et al., 2020). P. aeruginosa has three NADH:ubiquinone oxidoreductases: Complex I, NQR, and NDH-2, which are not homologous and can be distinguished by their substrate and inhibitor sensitivity (Degli Esposti, 1998; Matsushita et al., 1987; Yagi, 1991; Zambrano and Kolter, 1993; Zhou et al., 1999; Zickermann et al., 2000). NQR and complex I, which are proton pumps (Raba et al., 2018), are capable of utilizing both NADH and deamino-NADH (Matsushita et al., 1987; Zhou et al., 1999). Complex I is able to do so without showing substrate preference, and is inhibited by rotenone (Degli Esposti, 1998; Zickermann et al., 2000), while NQR is rotenone insensitive (Zhou et al., 1999). In contrast, NDH-2, which does not have proton pumping activity, specifically uses NADH as a substrate and is resistant to rotenone (Matsushita et al., 1987; Yagi, 1991; Zambrano and Kolter, 1993). Thus, the activities of these enzymes can be differentiated by measuring the respiratory activity with NADH, deamino-NADH, and rotenone, as showed previously by our group (Liang et al., 2020). Figure 5A depicts the oxygen consumption rate (OCR) of P. aeruginosa Δbc₁ membranes harvested in mAUM during the logarithmic growth stage using NADH, deamino-NADH, succinate, and lactate as substrates. These substrates are employed by the main dehydrogenases involved in the pathogen’s respiratory chain in mAUM, contributing to the ubiquinone/ubiquinol pool that can ultimately be utilized by the terminal oxidases (Hu et al., 2024; Liang et al., 2020). In comparison to the succinate and lactate dehydrogenase activities, the NADH oxidase activity in the mutant exhibits the highest OCR, indicating its critical role in respiration. NADH activity was mostly insensitive to rotenone, indicating that complex I is minimally involved, facilitating approximately 20% of electron transfer. Moreover, deamino-NADH oxidation was relatively lower compared to the wild-type cells (Hu et al., 2024; Liang et al., 2020), revealing that NQR and complex I contribute <35% of the NADH dehydrogenase activity in the cytochrome bc₁ mutant, and that the mutant mainly relies on NDH-2, supporting around 65% of electron transfer from NADH. This result contrasts sharply with wild-type P. aeruginosa, in which 75% of the NADH dehydrogenase activity is provided by NQR (Hu et al., 2024; Liang et al., 2020).

To assess the respiratory contribution of terminal oxidases in the cytochrome bc₁ mutant, respiratory ubiquinol oxidase relative activities were analyzed through KCN titration, as previously described (Figure 5B) (González-Montalvo et al., 2024; Hu et al., 2024; Liang et al., 2020). In addition to the bc₁ complex, which feeds cytochrome c oxidases, the respiratory chain of P. aeruginosa contains two ubiquinol oxidases: cytochrome bo₃ oxidase (Arai, 2011), which is sensitive to cyanide in the 10-30 μM range (Liang et al., 2020), and the cyanide-insensitive bd oxidase, which can be inhibited in the mM range (Arai et al., 2014; Liang et al., 2020). In logarithmic phase cells cultured in mAUM, the primary oxidase employed by the Δbc₁ mutant is the cyanide-insensitive bd oxidase, carrying out nearly 70% of the activity, while the remaining 30% can be attributed to bo₃ oxidase. According to our data, the respiratory chain in the wild-type is composed of NQR and bc₁ in addition to cytochrome oxidases, and in the mutant it switches to NDH-2 and bd oxidase, which significantly reduces the ability of the respiratory chain to pump protons and produce ATP (see below).

In our previous work, we reported that the growth of a PA14 cytochrome bc₁ transposon mutant (Liberati et al., 2006) showed significant defects in maximum growth and growth rate, and an increase in the lag phase (Hu et al., 2024). In this study, we obtained a deletion mutant of cytochrome bc₁ in PA14 using the lambda-red recombinase method (Datsenko and Wanner, 2000), which specifically eliminates the gene of interest and minimizes unintended effects of transposon mutagenesis, which can inactivate neighboring genes or regulatory sequences (Datsenko and Wanner, 2000). Moreover, we expanded our studies to LM (Ruhluel et al., 2022) to understand the effects of the mutant in the highly relevant lung environment, particularly for opportunistic infections in cystic fibrosis. Our results show significant growth defects in the mutant, including an increase in the lag phase (Figure 1F) and decreases in the growth rate (Figure 1E) and yield (Figure 1D) in rich laboratory LB media (Figure 1A), which were considerably more pronounced in physiologic media. The differences observed among culture media may be due to their specific composition. For instance, mAUM contains lactate and citrate as main carbon sources (González-Montalvo et al., 2024; Liang et al., 2020), while LM is composed mostly of amino acids (Ruhluel et al., 2022). In both mAUM and LM, an active Krebs cycle and respiratory chain are required for growth, underscoring the relevance of the bc₁ complex in these conditions due to its involvement in electron transport. In mAUM, P. aeruginosa relies on a tightly coupled, oxidative TCA cycle with strong gluconeogenic flux and high energetic efficiency, which could be related to faster growth. In contrast, in LM the metabolism becomes flexible and nitrogen-driven, with amino acid catabolism feeding multiple TCA entry points. In our previous work, the relative contribution of respiratory enzymes to ATP production was calculated, demonstrating that the bc₁ complex is the main respiratory enzyme in mAUM, contributing more than 50% of the respiratory chain H⁺-pumping activity (Hu et al., 2024). The relevance of the bc₁ complex in P. aeruginosa is confirmed in this work by the slow growth of the mutant in mAUM (Figures 1B, E). Remarkably, our data show that the bc₁ complex is also critical for proliferation in the lung-mimicking environment, where a sharp reduction in growth was observed in the mutant in comparison to the wild-type strain (Figures 1C, E). Relative to mAUM, the wild-type strain grew to higher OD₆₂₀ in LM but exhibited a lower growth rate. In the mutant, we observed an extended lag phase, indicating that cytochrome bc₁ plays important roles in adaptability to nutritional sources, particularly to the amino acid-rich lung media, which mirrors the conditions typically observed in cystic fibrosis lungs (Barth and Pitt, 1996). Amino acid catabolism feeds directly into the Krebs cycle, generating reducing equivalents that drive the electron transport chain. Thus, in an amino acid-rich medium such as LM, the cytochrome bc₁ mutation is expected to produce a significant growth defect.

Biofilms are hallmark contributors to the persistence of chronic urinary and pulmonary infections and are an important virulence factor of P. aeruginosa (Folkesson et al., 2012; Mancuso et al., 2023). Our results show that in LB, the wild-type strain was able to produce a significant amount of biofilm, which was not affected by the lack of the bc₁ complex. On the other hand, in mAUM, PA14 produced moderate amounts of biofilm, approximately 28% compared to the LB condition. Interestingly, in mAUM, the Δbc₁ mutant showed a decrease of 50% in biofilm production compared to the wild-type (Figure 3C). Indeed, a previous report indicates that in urea-rich environments, such as those in the urinary tract and mAUM, quorum sensing is repressed in P. aeruginosa (Cole et al., 2018), which plays an essential role in biofilm development (Miranda et al., 2022). In LM, the wild-type strain showed nearly a 3-fold increase in biofilm production compared to LB that could be due to high mucin content in LM, which may aid in bacterial attachment and aggregation, mimicking cystic fibrosis environments (Landry et al., 2006). Remarkably, biofilm production was severely diminished in the mutant strain (seven-fold reduction). The attenuation in biofilm production by the Δbc₁ mutant in physiologic media can be attributed to a decrease in electron flow through cytochrome bc₁, which diminishes both aerobic respiration via terminal cytochrome c oxidases and anaerobic respiration via denitrification. Several reports indicate that in the lung environment, especially during cystic fibrosis and in biofilms, P. aeruginosa switches its metabolism and is able to generate energy anaerobically via denitrification (Hassett et al., 2002; Ojoo et al., 2005; Williams et al., 2007), which employs nitrate and nitrite reductases dependent on cytochrome bc₁ activity (Arai, 2011; Williams et al., 2007). A previous report also demonstrates that inactivation of the cytochrome c₁ subunit of cytochrome bc₁ halted P. aeruginosa growth in anaerobic conditions using nitrite as an electron acceptor (Hasegawa et al., 2003). Alternatively, although the arginine deiminase (ADI) pathway is also utilized by P. aeruginosa under conditions when oxygen and nitrate are not present (Williams et al., 2007), increased expression of key ADI enzymes has been observed in CF lungs, indicating that this pathway is also employed during such infections (Hogardt and Heesemann, 2013). However, our data demonstrate that denitrification via the cytochrome bc₁-dependent respiratory chain is the predominant pathway supporting energy generation in P. aeruginosa biofilms under the tested conditions, as exemplified by the decreased biofilm produced by the mutant. As biofilm formation is a key virulence factor in P. aeruginosa in both pulmonary and urinary tract infections, reducing biofilm is critical for decreasing persistence, reducing colonization, and enhancing antibiotic efficacy (Narten et al., 2012; Ciofu et al., 2017; Soares et al., 2020). By identifying the metabolic relevance of cytochrome bc₁ for P. aeruginosa biofilm, we have demonstrated that therapeutic targeting of cytochrome bc₁ holds promise, particularly in the pulmonary environment of cystic fibrosis patients, which may also enhance current antibiotic efficacy via synergistic interactions. Our data and previous findings highlight the potential of cytochrome bc₁ as an antibiotic target that can be exploited in both aerobic and anaerobic conditions.

Our data show that deletion of cytochrome bc₁ reduces P. aeruginosa growth in physiologic conditions and decreases the virulence of cells in an in vitro infection model of 5637 human urinary bladder carcinoma cells, a system that recapitulates part of the urinary tract infection. In these conditions, the mutant has a significant attenuation of virulence, as there is a 39% difference between the death of Δbc₁-infected and PA14-infected cells (Figures 2A, B), which is not mediated by the secretion of virulence factors, but is contact-based (Figures 2C, D). qPCR experiments show that the mutant decreases expression of the essential T3SS components exoT and exoU, while exoY remains unchanged (Figure 3A). ExoU is encoded on the PAPI-2 pathogenicity island in PA14, which contains other non-exoU virulence genes (Harrison et al., 2010). The decrease of exoU expression in the mutant strain suggests that other virulence genes encoded within PAPI-2 may also have downregulated expression. Moreover, previous findings revealed that the respiratory chain is critical for P. aeruginosa pathogenicity in a Galleria mellonella model (Torres et al., 2019). Prior reports also show that the elimination of the bc₁ complex decreases P. aeruginosa virulence factor production and hinders motility (Shen et al., 2021). While these reports show a similar trend, the mechanism of bc₁ virulence regulation has remained unclear until now. In this work, we have shown that elimination of the bc₁ complex reduces ATP content by 44-92%, which can explain the reduced virulence, as the T3SS is a molecular motor that requires ATP for its function (Hauser, 2009). Thus, factors that affect energy production should have a direct effect on contact-based virulence. In this report, we extended the studies carried out on invertebrate models by investigating the role of cytochrome bc₁ in the virulence of P. aeruginosa in a clinically-relevant in vivo systemic murine infection model. Our results show that while the wild-type strain is highly virulent, producing >80% mortality in less than a week, the PA14 Δbc₁ strain is completely avirulent, demonstrating the critical role of the enzyme’s contribution to P. aeruginosa pathogenicity (Figure 3B).

It has previously been suggested that the highly branched respiratory chain of P. aeruginosa, in particular the diversity of terminal oxidases, is advantageous for pathogenicity (Hirai et al., 2016; Jo et al., 2014) and survival in different environments, such as biofilms, in which oxygen gradients are established (Jo et al., 2022). However, our data challenges this hypothesis and shows that cytochrome bc₁ is the most important respiratory enzyme in most conditions, including planktonic growth in physiologic media, and in biofilms. The critical role of the bc₁ complex is likely due to its involvement in supporting the activity of the terminal cytochrome c oxidases: caa₃, cbb₃-1, and cbb₃-2, which contribute the greatest amount of usable energy for the pathogen under aerobic conditions (Williams et al., 2007; Liang et al., 2020; Srivastav et al., 2025). Indeed, cytochrome c oxidases have been reported as essential for P. aeruginosa PA14 biofilm and pathogenicity in Caenorhabditis elegans (Jo et al., 2017), as deletion of subunit CcoN4, which makes complexes with cbb₃–1 and cbb₃-2 (Hirai et al., 2016), impairs biofilm architecture and improves nematode survival (Jo et al., 2017). Furthermore, elimination of bc₁’s Rieske subunit gene has been reported to interfere with the pathogen’s PQS, hindering biofilm formation (Soh et al., 2021). In addition to its critical role in aerobic conditions, the bc₁ complex also has a fundamental role under anaerobic environments, as it supports cytochrome c reduction, which is required for nitrate and nitrite-based anaerobic respiration (Williams et al., 2007). Taken together, the data strongly support a critical role of the bc₁ complex in virulence, supporting the bioenergetic pathways of the cell necessary for the survival of the pathogen or the expression or secretion of virulence factors.

Our data show that in both physiologic conditions, mAUM and LM, the ATP content of wild-type PA14 cells increased 2.4-4.3 times compared to LB, which suggests that the respiratory metabolism is activated. Indeed, we reported that cells grown in mAUM show an increase of about 3 times in the respiratory activity (Hu et al., 2024), consistent with the observed increase in ATP content in this condition (Figure 4A). Based on this result, we propose that the respiratory activity of cells grown in LM could also be activated and that the bc₁ complex would be even more important in this condition. Our results show that the elimination of cytochrome bc₁ significantly decreased ATP in the physiologically relevant mAUM and LM media in both growth phases. These results independently corroborate the hypothesis proposed by our group that electron transfer through cytochrome bc₁ (and the associated terminal oxidases) is the most important energy-generating pathway in P. aeruginosa (Hu et al., 2024; Liang et al., 2020). This is also consistent with the enzyme’s pumping stoichiometry, as bc₁ and cytochrome c oxidases pump 6 protons through the membrane per electron pair (Mitchell, 2011), compared to 2 and 4 protons per electron pair for bd (D’mello et al., 1996) and bo₃ (Graf et al., 2019) oxidases, respectively (Hu et al., 2024). Interestingly, we found a significant increase in the membrane potential of the mutant compared to the wild-type during logarithmic growth in LM. Membrane potential is a complex metabolic parameter that depends on many factors, including respiratory activity, membrane permeability (regulated by membrane lipid composition), and numerous gradient-consuming processes (ATP synthesis, active transport, motility, etc.). It is likely that due to the limited ability of the respiratory chain to produce a membrane potential, the gradient-consuming steps are significantly downregulated, which would allow the cell to maintain a stable membrane potential but would affect downstream homeostatic processes. However, further experiments are required to clarify this issue.

In addition to the respiratory metabolism, the bc₁ complex would also impact ROS production by the cell (Borisov et al., 2021), as this complex is known to be the main contributor within the electron transport chain (Chance and Hollunger, 1960). Quantification of ROS production demonstrates that the wild-type strain produces significantly more ROS in LB media during both growth phases and LM during stationary phase (Figure 4B), consistent with this complex as a major source of ROS. However, in mAUM the mutant produced significantly more ROS compared to the wild-type due to accumulation of electrons in the quinol pool, leading to redirection via the bd oxidase. Thus, the excess electrons can react with oxygen, generating superoxide, linking the decreased ATP observed by the mutant in mAUM to the observed increase in ROS. Moreover, other reports have shown that mutants lacking bc₁ do not produce PQS signaling (Soh et al., 2021), which induces an antioxidant response (Häussler and Becker, 2008). This deficiency could account for the increase in ROS production in the mutant strain cultured in mAUM in comparison to the wild-type in both growth phases. During the logarithmic phase in LM, there is no significant difference in ROS production between the wild-type and mutant strains, which can be attributed to both being highly metabolically active during these conditions, prioritizing energy and resources for growth. Furthermore, the extended lag phase and reduced growth rate observed in both strains suggests that ROS mechanisms may be more apparent under non-proliferative or stress-inducing conditions.

The data presented here, as well as previous reports in the literature, indicate that the bc₁ complex is critical but not essential for growth, which suggests that the mutant is able to use other metabolic pathways to compensate for its absence. In this report, we show that the mutant compensates for lack of the bc₁ complex by using bd oxidase, which intriguingly only pumps 2 H⁺ per electron pair (D’mello et al., 1996), instead of bo₃ oxidase, which pumps twice as many protons (Graf et al., 2019). In addition, the cells shift from NQR, which pumps two cations (Na⁺ or protons) per electron pair (Raba et al., 2018), to NDH-2, which is not a proton pump (Heikal et al., 2014), as the primary NADH dehydrogenase. Thus, the respiratory chain of the mutant (NDH-2 plus bd oxidase) pumps a total of 2 protons per electron pair, while the wild-type chain (NQR, bc₁, and cytochrome c oxidases) pumps 8 protons per electron pair. Assuming that the metabolic rates through the Krebs cycle are similar in the mutant, this would represent a predicted decrease of 4 times in the ATP production in the bc₁ mutant, which is close to the observed experimental values. Furthermore, our observations regarding growth behavior may also be attributed to lack of ATP production by the mutant, causing the bacteria to compensate by utilizing fermentation or the ADI pathway (Williams et al., 2007). This process does not contribute to formation of the proton gradient, and thus may not be enough to sustain virulence. However, in LB media, we did not see a significant difference in ATP production between the wild-type and mutant strains, which may be due to reliance on fermentation or substrate-level phosphorylation. This drastic decrease in ATP production explains a severe decrease in virulence by the mutant cells in vitro and an almost complete lack of virulence in the animal model.