Mode of action of the 2-phenylquinoline efflux inhibitor PQQ4R against Escherichia coli

Bacteria and growth conditions

The strains included in the study were the wild-type E. coli K-12 AG100 (argE3 thi-1 rpsL xyl mtlΔ (gal-uvrB)supE44); the AcrAB pump-deficient E. coli AG100A (ΔacrAB), and the AcrAB overexpressing E. coli AG100_tet. The strain AG100A is a derivative of AG100 and has the AcrAB system inactivated due to an insertion of the transposon Tn903 (ΔacrAB::Tn903 Kan^r) (George & Levy, 1983; Okusu, Ma & Nikaido, 1996). AG100_tet is a derivative of AG100 obtained by continuous exposure to increasing concentrations of tetracycline (TET) (Viveiros et al., 2005). The strains were grown in Luria-Bertani (LB) broth at 37 °C with shaking. AG100A was grown in presence kanamycin (KAN) at 100 µg/ml to maintain the transposon and AG100_tet was grown in media supplemented with TET at 8 µg/ml to maintain the overexpression of efflux pumps.

Chemicals

Ofloxacin (OFX), oxacillin (OXA), KAN, TET, ethidium bromide (EtBr), CPZ, PAβN, carbonyl cyanide-m-chlorophenylhydrazone (CCCP), glucose, and phosphate-buffered saline (PBS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). The 2-phenylquinoline PQQ4R was synthesized as previously described (Sabatini et al., 2013).

Antibacterial activity evaluation

The minimum inhibitory concentrations (MICs) were determined using the broth microdilution method according to the CLSI guidelines (Clinical and Laboratory Standards Institute, 2014). The synergistic activity between the efflux inhibitors, antibiotics and EtBr was evaluated by checkerboard assays as previously described (Pillai, Moellering & Eliopoulos, 2005; Coelho et al., 2015). Two-fold serial dilutions of the compounds were made to achieve the following concentrations: 320 µM–40 µM. The MICs were determined as the lowest concentration at which no visible growth was observed after 18 h of incubation at 37 °C. The assays were performed in triplicate and the MIC value was given as result of at least two concordant values.

Evaluation of efflux activity by real-time fluorometry

The effect of the inhibitors on EtBr accumulation and efflux was assessed by fluorometry as previously described (Viveiros et al., 2008; Viveiros et al., 2010). The strains were grown until an OD_600nm of 0.6 at 37 °C with shaking. After, the cells were collected by centrifugation at 16,060× g for 3 min, washed in PBS and centrifuged again.

For the accumulation assays, the OD_600nm of the cell suspension was adjusted to 0.6 by adding PBS, allowing the assay to run with a final OD of 0.3. To determine the EtBr concentration at which influx and efflux are in equilibrium, the accumulation assays were performed in presence of increasing concentrations of the dye. The assays were prepared to a final volume of 100 µl containing 50 µl of the cellular suspension (final OD_600nm of 0.3) plus 50 µl of EtBr solutions to final concentrations ranging from 0.0625 to 5 µg/ml. The effect of the inhibitors on the accumulation of EtBr was evaluated in a final volume of 100 µl containing 50 µl of the cellular suspension (final OD_600nm of 0.3) and 50 µl of a solution containing EtBr at the equilibrium concentration and the compounds to a final concentration of 80 µM for AG100 and AG100_tet, and 20 µM for AG100A. The molar concentrations used represent $\frac{1}{4}$ or less of the MIC determined for each compound against each strain (see Table 1) in order to guarantee that the real-time efflux inhibitory effects measured were not due to any antimicrobial effect of the compound. The assays were conducted in a Rotor-Gene 3000 (Corbett Research, Sydney, Australia) at 37 °C, and the fluorescence acquired at 530/585 nm at the end of every 60 s, for 30 min. The activity of the compounds on the accumulation of EtBr was evaluated by the relative final fluorescence (RFF) index according to the formula: RFF = (RF_treated– RF_untreated)/RF_untreated. In this formulae the RF_treated corresponds to the fluorescence at the last time point of the EtBr accumulation curve (minute 30) in the presence of an inhibitor and the RF_untreated corresponds to the fluorescence at the last time point of the EtBr accumulation curve of the control tube (Machado et al., 2011). The experiments were done in triplicate and the RFF values are presented as the average of three independent assays (±SD).

For the efflux assays, the strains were exposed to conditions that promote maximum accumulation of EtBr, i.e., EtBr at the equilibrium concentration for each strain, no glucose, presence of the efflux inhibitors, and incubation at room temperature for 1 h (Viveiros et al., 2010). The OD_600nm of the cell suspension was adjusted to 0.3 and incubated with EtBr under the conditions described above. Aliquots of 50 µl of the cells were transferred to tubes containing 50 µl of each efflux inhibitor at 80 µM (AG100 and AG100_tet) or 20 µM (AG100A) without EtBr. Control tubes with only cells and cells with and without 0.4% glucose were included. The fluorescence was measured in a Rotor-Gene 3000 and the data was acquired every 30 s for 30 min at 37 °C. The efflux activity was quantified by comparing the fluorescence data obtained under conditions that promote efflux (presence of glucose and absence of efflux inhibitor) with the data from the control in which the bacteria are under conditions of no efflux (presence of an inhibitor and no energy source). The relative fluorescence corresponds to the ratio of the fluorescence that remains per unit of time, relatively to the EtBr-loaded cells (Viveiros et al., 2008; Viveiros et al., 2010).

Time-kill kinetics

The determination of the killing activity of PQQ4R was analysed by time-kill assays as previously described (Pillai, Moellering & Eliopoulos, 2005) with slight modifications. Briefly, AG100 was grown until OD_600nm of 0.6 at 37 °C with shaking. Exponential bacterial cultures were diluted in Mueller-Hinton Broth (MHB) to a cell density of 1 × 10⁵ cells/ml and 500 µl added to test tubes containing MHB and PQQ4R at the desired concentration. The compound was added to each tube to achieve the final concentrations of 8× to 0.5× the MIC. A drug-free control was included in the assay to monitor the normal growth of the strain. Cultures were sampled for CFU determination after 0, 1, 2, 3, 4, 5, 6, and 24 h of incubation at 37 °C with shaking. For CFU determination, 10-fold serial dilutions were made in a saline solution and 20 µl of each solution was placed onto the surface of a Mueller–Hinton agar plate. The colonies were counted after the incubation of the plates at 37 °C for 24 h. The limit of detection of the assay was 17 CFU/ml. Each assay was repeated at least twice.

Membrane potential assay

The effect of PQQ4R on the membrane potential was measured using the BacLight Bacterial Membrane Potential Kit (Molecular Probes, Life Technologies) according to the manufacturer’s instructions. CPZ and PAβN were tested for comparison. AG100 was grown at 37 °C with shaking until reach OD_600nm 0.6. After this, the cells were washed in PBS and diluted to 1 × 10⁷ CFU/ml with PBS. PQQ4R was added to the cell suspension at concentrations from 640 to 80 µM. The samples were transferred into black flat bottom 96-well plates, 30 µM of DiOC₂ (3) was added to the mixture, and the plates were incubated during 30 min in the dark. The fluorescence was measured using a Synergy HT multi-mode microplate reader (BioTek Instruments Inc, Vermont, USA) with the filters 485/20 (excitation) and 528/20 (emission) for green and 590/35 (emission) for red. CCCP was used as positive control at 156 µM (half MIC), since it eradicates the proton gradient, eliminating the membrane potential. The red to green ratio was determined and normalized against the emission from the DiOC₂ (3) blank well and the results are presented as the percentage of depolarized membranes (±SD) compared with the drug-free control.

Membrane permeability assay

The evaluation of the membrane integrity was done using the Live/Dead BacLight Bacterial Viability Kit (Molecular Probes, Life Technologies) according to the manufacturer’s instructions. AG100 was grown at 37 °C with shaking until OD_600nm of 0.6. Samples of 500 µl were incubated with PQQ4R at concentrations from 640 to 80 µM during 1 h at room temperature. The cells were collected by centrifugation at 16,060× g for 10 min, the supernatant was discarded and the pellet resuspended in the same volume of a saline solution. Then, 100 µl of the cell suspension were transferred into black flat bottom 96-well plates and the dyes propidium iodide and SYTO 9 (ratio 1:1) were added to each well to stain the cells. The plate was incubated during 15 min at room temperature in the dark. The fluorescence was measured using a Synergy HT multi-mode microplate reader with the filters 485/20 (excitation) and 528/20 (emission) for green and 590/35 (emission) for red. The green to red ratio was determined and the results are presented as the percentage of intact membranes (±SD) compared with the control (no treatment). CPZ and PAβN were included for comparison.

Intracellular ATP levels determination

The ATP levels were measured using the ATP Determination Kit (Invitrogen, Life Technologies, Paisley, UK) according to the manufacturer’s instructions. AG100 was grown until OD_600nm of 0.6 at 37 °C with shaking. Exponential bacterial cultures were diluted in MHB to a cell density of 1 × 10⁵ cells/ml and 500 µl added to test tubes containing MHB and PQQ4R at half MIC and at the MIC. A drug-free control was included in the assay to monitor the normal growth of the strain. Aliquots of bacteria were collected at 0, 1, 2, 3, 4, 5, 6, and 24 h of incubation at 37 °C, with shaking, inactivated by heating and immediately deep frozen. The cell lysates were transferred into white flat bottom 96-well plates and the ATP content measured using a Synergy HT multi-mode microplate reader and expressed as relative luminescence units. CPZ and PAβN were included for comparison.

Cytotoxicity against human monocyte-derived macrophages

The blood was collected from healthy donors and the peripheral blood mononuclear cells were isolated by Ficoll-Paque Plus (GE Healthcare, Freiburg, Germany) density gradient centrifugation as previously described (Machado et al., 2016). Briefly, monocytes were differentiated into macrophages during 7 days in RPMI-1640 medium with 10% fetal calf serum (FCS), 1% GlutaMAX™, 1 mM sodium pyruvate, 10 mM HEPES at pH 7.4, 100 IU/ml penicillin and 100 µg/ml streptomycin (Gibco, Life Technologies), and 20 ng/ml M-CSF (Immunotools, Friesoythe, Germany) and incubated at 37 °C in a 5% CO₂ atmosphere. Fresh medium was added at day 4 post isolation. The effect of the compounds on human monocyte-derived macrophages was evaluated using the AlamarBlue method (Molecular Probes, Life Technologies) (O’Brien et al., 2000), and by measuring the release of the lactate dehydrogenase (LDH) into the culture supernatants using the Pierce LDH Cytotoxicity Assay (ThermoFisher Scientific, Waltham, MA, USA) (Chan, Moriwaki & Rosa, 2013) according to the manufacturer’s instructions. Briefly, 5 × 10⁴ cells were seeded in 96-well microplates treated with the compounds, and incubated at 37 °C with 5% CO₂. After 72 h of treatment, the cell viability was assessed. For the AlamarBlue method, the dye was added to each well to a final concentration of 10% and incubated overnight at 37 °C and 5% CO₂. The fluorescence was measured with a 540/35 excitation filter and a 590/20 emission filter in a Synergy HT multi-mode microplate reader. The release of LDH from damaged cells into culture supernatants, as an indicator of cytotoxicity, was measured in serum-free medium. Here the amount of enzyme activity correlates to the number of damaged cells. The specific lysis was calculated as follows: (treated cells—spontaneous LDH release)/ (maximum LDH release—spontaneous LDH release) ×100. The inhibitory concentration IC₅₀ (50%) value was calculated using GraphPad Prism V5.01 software (La Jolla, USA). The IC₁₀ (10%) and IC₉₀ (90%) were calculated using the GraphPad program “ECanything” available online (Graph Pad Software, 2016).

Statistical analysis

Statistical analysis was carried out using the Student’s t-test. A P value <0.05 was considered statistically significant and highly significant when ^∗∗P < 0.01 and ^∗∗∗P < 0.001 (two-tailed tested).

Synergistic activity of PQQ4R in combination with antimicrobials

The MICs of the antibiotics, which are substrates of the AcrAB efflux pump, the efflux substrate EtBr, PQQ4R, and for comparison, PAβN and CPZ, known E. coli efflux inhibitors, were determined for the wild-type AG100, the pump deficient AG100A, and the pump-overexpressing AG100_tet (Table 1). AG100A is more susceptible to the selected antibiotics and EtBr than the wild-type, due to the absence of a functional AcrAB, reconfirming that these antibiotics are substrates of this pump (Piddock, 2006; Viveiros et al., 2005). The susceptibility data also showed that neither PQQ4R nor PAβN and CPZ had, by themselves, antimicrobial activity against the E. coli strains. Moreover, PQQ4R MIC decreased 8-fold against AG100A once compared with the wild-type parental strain, indicating that, PQQ4R might be a substrate of the AcrAB system. Similar results were observed for PAβN.

To assess whether PQQ4R potentiated the activity of the tested antibiotics against E. coli, the MICs of OFX, OXA, and TET were determined in the presence of PQQ4R. PAβN and CPZ were tested at the same concentrations for a direct comparison of their activities (Table 2). Against the wild-type strain, PQQ4R, at concentrations ranging from 40 to 160 µM, produced a 2-fold reduction in OFX and TET MICs, while a 2- or 4-fold reduction on the MICs of EtBr at 80 and 160 µM, respectively, were observed. On the contrary, no effect was observed for OXA. Interestingly, PQQ4R at concentrations of 80 and 160 µM, significantly increased the antibacterial activity of OFX and TET in the AcrAB-overexpressing strain AG100_tet, (AcrAB overexpressed 6–10 times compared with the isogenic wild-type strain—Viveiros et al., 2005; Viveiros et al., 2007) causing a 4-fold decrease in their MICs, with a marginal effect on the MIC of EtBr. In accordance with the increased susceptibility detected above, the pump deficient AG100A did not grow in presence of PQQ4R at concentrations above 40 µM and this concentration had no effect on the MIC of the antibiotics. However, this concentration caused a ≈8-fold reduction in the EtBr MIC for AG100A, confirming that other efflux pumps, that extrude EtBr (AcrEF, among others), are active in this strain (Viveiros et al., 2005; Viveiros et al., 2007).

The presence of PAβN significantly reduced the MICs of OXA and TET on the three strains. PAβN reduced the MIC of TET and OXA on the pump-deficient strain by 16-fold and 4-fold, respectively; however, it showed no effect on OFX and EtBr MICs. Conversely, the exposure to PAβN produced a 128-fold decreased in the MIC of EtBr against the AcrAB-overexpressing strain, while no change was observed in the MIC of the wild-type strain. The inability of PAβN to reduce the MIC of EtBr in wild-type strains has already been described for E. coli and P. aeruginosa (Lomovskaya et al., 2001; Kern et al., 2006). This result indicates that PAβN is acting as a pump competitor and not as an efflux inhibitor like PQQ4R and CPZ. The activity of CPZ against the pump-deficient strain was similar to that of PQQ4R, with the exception that CPZ reduced the MIC of TET by 4-fold. CPZ significantly reduced the MICs of OFX, OXA, TET and EtBr on the wild-type and the AcrAB-overexpressing strain.

Efflux inhibitory activity of PQQ4R

The MIC results give an indirect measure of the potential efflux activity of the strains and the effect of the inhibitors on this activity. The ability of PQQ4R to inhibit E. coli efflux systems was confirmed by real-time fluorometry using EtBr, a broad efflux pump substrate (Viveiros et al., 2008; Viveiros et al., 2010). Initially, it was determined the equilibrium concentration at which the influx of EtBr equals its efflux. The accumulation of EtBr started at concentrations above 1 µg/ml for AG100, 0.25 µg/ml for AG100A, and 2 µg/ml for AG100_tet (Fig. S1). Using these concentrations, we evaluated the ability of PQQ4R, at sub-inhibitory concentrations (<1/8 MIC), to promote the intracellular accumulation of EtBr (Fig. 1) and calculated the RFFs (Table 3). The RFF value is a measure of how effective the compound is on the inhibition of the EtBr efflux (at a given concentration) by comparison of the final fluorescence at the last time point (60 min) of the treated cells with the cells treated only with EtBr. An index of activity above zero indicated that the cells accumulate more EtBr under the condition used than those of the control (non-treated cells taken as 0). In case of negative RFF values, these indicated that the treated cells accumulated less EtBr than those of the control condition. Values above 1 in the presence of the efflux inhibitors were considered enhanced accumulation of EtBr inside the cells. CPZ and PAβN were included at the same concentrations for a direct comparison of their inhibitory potencies.

At 80 µM, PQQ4R showed an RFF of 13.1 against the wild-type strain, indicating a strong ability to interfere with EtBr efflux compared to CPZ and PAβN, that showed RFF values of 8.1 and 0.7, respectively (Fig. 1A; Table 3). This means that PQQ4R promotes the higher accumulation levels of EtBr compared to CPZ followed by PAβN. In the AcrAB-deficient strain there was a small increase in the accumulation of EtBr in the presence of PQQ4R (RFF = 1.5) and CPZ (RFF = 2), whereas almost no effect was observed with PAβN (RFF = 0.1) (Fig. 1B; Table 3). These results showed once again that AcrAB is the main efflux system that pumps out EtBr in E. coli, but that other efflux pumps susceptible to PQQ4R and CPZ are pumping out EtBr as described before (Viveiros et al., 2005; Viveiros et al., 2007). For the AcrAB-overexpressing strain AG100_tet (AcrAB is 6–10 times overexpressed compared to the wild-type strain), the accumulation of EtBr in the presence of PQQ4R corresponded to a RFF of 5.2 (Fig. 1C; Table 3). The EtBr accumulation in this strain increased significantly when compared to AG100A, but, for the same molar concentration of inhibitor tested (80 µM), did not reach the same levels of inhibition observed for AG100, since the AcrAB system is 6–10 times more expressed in the AG100_tet strain compared with the AG100 (Viveiros et al., 2008; Viveiros et al., 2010). These results, obtained in a set of three isogenic E. coli that differ only by the absence/presence or expression level of the AcrAB system, indicated that PQQ4R promotes accumulation (i.e., interferes with the efflux) of EtBr in E. coli mainly thought the inhibition of the AcrAB efflux system. In the presence of CPZ and PAβN, the EtBr accumulation in the AG100_tet reaches similar levels (RFF 7.9 and 1.3, respectively) to those obtained for the wild-type strain (RFF 8.1 and 0.7, respectively), being less effective in the inhibition of efflux activity than PQQ4R.

To confirm the results obtained above on the accumulation of EtBr promoted by the inhibitors, we performed efflux assays for each strain. Each strain was subject to conditions that promote significant EtBr accumulation over a period of 60 min at room temperature: in the absence of glucose and presence of PQQ4R, CPZ or PAβN. After the maximum accumulation has been reached, EtBr and the efflux inhibitors were washed-out and the cells were subsequently re-suspended in new buffer with and without glucose and the inhibitor. As showed in Fig. 2, the efflux take place readily in presence of glucose at 37 °C, an activity that is inhibited in the presence of PQQ4R and CPZ and to a lesser extent with PAβN. These results confirmed that PQQ4R, CPZ, and PAβN inhibit the efflux of EtBr and this inhibitory effect is transient when the cells are washed-out of the inhibitor and an energy source is given to promote active efflux reenergizing the cells (Fig. 2).

Membrane depolarization and cell viability

To test the hypothesis that the interference with the bacterial energy metabolism is the cause of the efflux inhibition by PQQ4R, the effect of this compound on the bacterial membrane potential was evaluated using the BacLight Bacterial Membrane Potential Kit. Thirteen minutes after adding PQQ4R at any of the concentrations tested, the percentage of depolarized cells was over 85% (Fig. 3A). The protonophore CCCP was used as depolarization control and, as expected, the membrane potential collapsed when the cells were incubated for the same time with CCCP (over 95%). These results showed that PQQ4R depolarizes E. coli membranes by interfering with the gradient of protons through the cell membrane. Then, to evaluate if the depolarization of the cells, promoted by PQQ4R, results on a transient effect on cells with an intact membrane or alters the membrane integrity, the Live/Dead BacLight Bacterial Viability Kit was used. This assay uses the fluorescent stain SYTO 9 that penetrates in all bacterial membranes and stains the cells in green, and propidium iodide that penetrates only in the cells with permeabilized membranes. The combination of the two stains produces red fluorescent cells and these latter ones are considered dead by cell damage. The integrity of E. coli membranes was evaluated, by fluorometry, after exposure to PQQ4R at sub-inhibitory (1/4 and 1/2 MIC) and bactericidal concentrations (MIC), during 1 h. The results showed that at half MIC, PQQ4R caused cell damage, reducing the membrane integrity by 28% (Fig. 3B). This increased membrane permeability was accompanied by a reduction in the bacterial viability of approx. 1 log (Fig. 4—see also kinetics of bacterial killing). Membrane integrity was then evaluated on the cells exposed to PQQ4R at 80 µM (≈1/8 MIC), the same concentration used in the real-time fluorometry accumulation and efflux assays, for 1 h and the results were assessed as described before. At 80 µM, PQQ4R did not alter the cell membrane integrity (Fig. 3B). To evaluate if the EtBr accumulation observed in presence of CPZ and PAβN was due to membrane permeability we tested both compounds also at 80 µM. The results showed that the membrane permeability increased by 18% and 14% with CPZ and PAβN, respectively. Membrane depolarization occurs in presence of CPZ at 80 µM similarly to that caused by PQQ4R, and PAβN induced 14% membrane depolarization at 80 µM. Overall, these results showed that PQQ4R causes transient membrane depolarization without affecting membrane integrity or cell viability at low concentrations (80 µM - 1/8 MIC- cell viability at this concentration 14.34 log₁₀ CFU/ml—Fig. 4). In contrast, and as expected by the previous results, at bactericidal concentrations, PQQ4R acts as a membrane-destabilizing agent, which is also consistent with the cell death at this concentration (Fig. 4). The gathering of these results points towards the depolarization of the cell membrane as the main mechanism for PQQ4R’s transient efflux-inhibition activity.

PQQ4R effect on intracellular ATP levels

Destabilization of the membrane functions can impair the respiratory chain functions and consequently reduce the ATP levels. To evaluate whether the exposure to PQQ4R could have an effect on the intracellular ATP levels, E. coli AG100 was exposed to PQQ4R and the intracellular ATP levels measured during 24 h at either sub-inhibitory (320 µM –1/2 MIC) or bactericidal concentrations (640 µM—MIC) (Fig. 5). The ATP levels remained constant during the first 3 h of exposure at both concentrations and at the same level of the drug-free control excluding the abrupt ATP depletion from being the direct cause of the efflux-inhibition previously seen with the EtBr accumulation and efflux assays, and supports a slow but steady ATP production impairment. After 4 h of exposure, the ATP levels of the drug-free control increased but the ATP levels of the drug-containing tubes remained practically constant after the addition of PQQ4R and until the end of the assay, with the exception of the cells exposed to half MIC. In this case, the ATP production increased after 24 h of exposure but was 80% less than that of the drug-free control.

Kinetics of PQQ4R bacterial killing

To characterize the killing effect of PQQ4R, we measured its bacterial killing activity against AG100 through time-kill studies (Fig. 4). PQQ4R at half MIC had no effect on the viability of E. coli after 24 h of exposure. Exposure to PQQ4R reduced the viability of E. coli to zero after 3 h at the MIC (256 µg/ml–640 µM) while at 2× MIC PQQ4R reduced the viability to zero after 1 h of exposure. After 24 h of exposure, all cultures remained negative at these concentrations. These results showed that PQQ4R killing activity is rapid reaching 100% lethality at the MIC, a concentration that was also found to be bactericidal (minimal bactericidal concentration—MBC).

Spectrum of activity of PQQ4R

The antimicrobial activity of PQQ4R was evaluated against other bacterial species to assess its antibacterial specificity: the Gram-negative bacteria Acinetobacter baumannii, Salmonella enterica serovar Enteritidis, Klebsiella pneumoniae and Enterobacter aerogenes, the Gram-positive S. aureus, and the acid-fast bacteria Mycobacterium smegmatis, M. avium, and M. tuberculosis. The results are depicted in Table 4 and showed that PQQ4R possessed moderate antibacterial activity towards Gram-positive (average MIC 50 µg/ml) and acid-fast bacteria (average MIC 32 µg/ml) and reduced antibacterial activity against Gram-negative bacteria (average MIC ≥128 µg/ml).

Cytotoxicity

Finally, PQQ4R was evaluated for its toxicity against human monocyte-derived macrophages using CPZ and PAβN for comparison. After 72 h of exposure, the viability of the macrophages was assessed using the AlamarBlue method. The comparative activities, CC₉₀, CC₅₀ and CC₁₀ values, are presented in Table 5. The dose–response curves are shown in Fig. 6. PQQ4R showed to be more toxic (CC₅₀ 10.83 µM) when compared to CPZ (CC₅₀ 22.24 µM) and PAβN (CC₅₀1269 µM). Afterwards, the amount of LDH release was measured in the cultures supernatants (Fig. 7). The LDH is an intracellular enzyme that is released in the culture media as a consequence of damaged cell membranes (Chan, Moriwaki & Rosa, 2013). The treatment with PQQ4R at 0.625 µM–2.5 µM caused 1%–4.6% increase in LDH release compared with the non-treated cells; at 5 µM LDH release increases by 52.45% and to 72%–78% at higher concentrations. Comparatively, CPZ increases LDH release from 0.95% at 1.25 µM to 24% at 20 µM. At concentrations above, the amount of LDH detected in the culture supernatant was above 95%. Concerning PAβN, the amount of LDH detected varied from 1.08% to 8% at concentrations from 10 µM to 640 µM. At concentrations above, PAβN caused 66%–68% LDH release. The results obtained with the LDH method corroborated the results obtained with the AlamarBlue method.