Photosensitizer-specific bacterial stress responses in Escherichia coli reveal distinct targets in photoinduced inactivation

Chittò, Marco; Tutschner, David; Dobrindt, Ulrich; Galstyan, Anzhela; Berger, Michael

doi:10.1038/s42003-025-08881-4

Download PDF

Article
Open access
Published: 01 October 2025

Photosensitizer-specific bacterial stress responses in Escherichia coli reveal distinct targets in photoinduced inactivation

Communications Biology volume 8, Article number: 1413 (2025) Cite this article

552 Accesses
1 Altmetric
Metrics details

Subjects

Abstract

The ongoing antibiotic crisis calls for alternative antimicrobial strategies. Antimicrobial photodynamic therapy (aPDT) offers a sustainable option, employing a light-activated photosensitizer (PS) to generate cytotoxic reactive oxygen species (ROS). The non-selective nature of these ROS helps minimize the risk of resistance development. Despite the development of numerous PSs, limited data on their exact mechanisms and bacterial targets still hinders broader clinical use. The focus of this study is to address this gap by capturing pathway-specific responses to sub-lethal photodynamic stress using a panel of transcriptional biosensors in E. coli K-12 strain MG1655. Our results indicate that methylene blue (MB) primarily causes oxidative stress in the cytosol while silicon phthalocyanine derivative (SiPc) induces envelope stress at physiological conditions. By monitoring well-characterized stress response pathways, our method offers a valuable tool for elucidating the physiological effects of aPDT and guiding more detailed mechanistic or transcriptomic studies.

Smart molecular designs and applications of activatable organic photosensitizers

Article 06 November 2024

The oxidative stress and metabolic response of Acinetobacter baumannii for aPDT multiple photosensitization

Article Open access 03 February 2022

Orthogonally conjugated phthalocyanine-porphyrin oligomer for NIR photothermal-photodynamic antibacterial treatment

Article Open access 14 March 2025

Introduction

Antimicrobial photodynamic therapy (aPDT) has emerged as a promising and rapidly expanding therapeutic modality for the treatment of various multidrug-resistant infections^1,2. In contrast to standard antibiotic treatment, there are several clear benefits of this approach. To the best of current knowledge, mutant bacteria resistant to aPDT have never been isolated or characterized³, the duration of treatment is generally quite short and aPDT targets the region of interest with little to no invasiveness^4,5.

The aPDT modality employs a light-activated compound called a photosensitizer (PS) that can either be introduced topically or systemically before tissue illumination with harmless visible light. Following absorption of a photon of light of the specific wavelength according to its´ set absorption spectrum, the PS is excited into a long-lived triplet state, which can react with nearby molecules by one of three proposed mechanisms. Type I mechanism results in an electron or hydrogen transfer, leading to the production of radicals. The Type II mechanism results in energy transfer directly to molecular oxygen yielding highly reactive singlet molecular oxygen, ¹O₂^6,7 with subsequent cytotoxic effects due to the potential of reactive oxygen species (ROS) reacting with nucleic acids, proteins, and/or cell membranes, thus destroying microorganisms in a very short time frame. A third, Type III, mechanism has recently been shown where aPDT can be potentiated by the simple addition of an inorganic salt such as potassium iodide⁶. This Type III mechanism involves photoinduced electron transfer that produces reactive inorganic radicals, enabling the inactivation of bacteria in the absence of oxygen⁷.

Recently, a proteomic approach has been used to characterize aPDT-induced damage to Staphylococcus aureus⁸. By analyzing the damaged proteins, it was found that the aPDT-induced damage to proteins is specific and likely dependent on the localization of the PS in the bacterial cell⁹. Using octacationic Zn(II)-phthalocyanine, it was demonstrated that the initial photodamage of E. coli cells occurs at the level of specific proteins in the outer membrane, thus promoting the penetration of the photosensitizer into the cytoplasmic membrane, where some enzymes critical for cell survival were inactivated¹⁰.

Since the ability of some PSs to bind to bacterial membranes shows a strong correlation with their antibacterial efficacy, it has been widely accepted—though not universally proven—that the bacterial membrane serves as a primary target in aPDT^11,12,13. This association has guided much of the research and development in aPDT, emphasizing the importance of membrane interactions in achieving effective bacterial inactivation^14,15. However, other studies have reported that ROS-induced damage to DNA, and the resulting disturbances in DNA replication and transcription processes, kill the cells¹⁶. Therefore, despite many efforts, it is still not clear which of these mechanisms and which targets are most important for microbial inactivation with an aPDT, and yet the elucidation of the mechanisms and targets of aPDT remains fundamentally important, as it is the key to further improve the antimicrobial activity and utilization of PSs¹⁷. Unfortunately, despite all these studies, there is still a lack of experimental data on the physiological response of bacteria to various PSs. Global approaches to address this, such as proteomic or transcriptomic analysis, can be time- and cost-prohibitive and, if pursued, may result in a very limited resolution due to the reduced number of samples that may be processed. Thus, making these experimental methods unsuitable for the screening of large libraries of PSs. To address this limitation, we therefore developed high-throughput biosensors for screening the impact of PSs on microbial physiological function. The assay is based on a set of chromosomally integrated promoter-yellow fluorescent protein (yfp) gene fusion modules in E. coli K-12 MG1655 that allow monitoring the expression of representative genes of major bacterial stress response pathways. In our study on the bacterial response to the photodynamic action, we chose two different classes of PSs: MB as one of the well-studied PSs that has been extensively used for the treatment of malignant diseases and infections¹⁸ and 2(3),9(10),16(17),23(24)-tetrakis-[3-(N-phenyl)pyridyloxy]-phthalocyaninato dihydroxy-silicon (IV) tetrabromide (SiPc)¹⁹ as a representative of positively charged, non-aggregated tetrapyrrole-based PS (Fig. 1). Both PSs absorb light in the visible electromagnetic spectrum with the absorption maxima at 664 nm (logε = 4.89) for MB and 678 nm (logε = 5.10) for SiPc and have comparable singlet oxygen quantum yields in organic solvents (Φ_Δ = 0.57 for MB and Φ_Δ = 0.59 for SiPc). Despite very similar photophysical properties, the photodynamic effect of these PSs was found to be very different²⁰. Whereas the photoactivity of MB against Gram-negative bacteria is very low, SiPc was found to decrease the bacterial viability at very low concentrations under the same irradiation conditions (Fig. S1, Supporting Information).

**Fig. 1: Photosensitizers and photochemical pathways for ROS generation.**

Our biosensors were also constructed to contain a constitutively expressed cyan fluorescent protein gene (Pfrr-cfp) for the normalization of the stress response, as the optical properties of the PSs prohibit the classical use of normalization of reporter gene expression to the optical density of the bacteria. Using these biosensors in combination with a set of regulator mutants, we have demonstrated in this study that MB activates the oxidative stress response via OxyR, whereas SiPc induces envelope stress in the bacteria via BaeR and CpxR in a concentration- and light-dependent manner.

Results

Construction and testing of the biosensors

The construction of our SOS biosensor is described elsewhere²¹. In order to measure a more comprehensive bacterial response to the PSs, we constructed a set of reporter strains in an identical manner that carry transcriptional fusions of yfp to the promoters of different stress response genes. We used yfp to replace the pH stress pathway gene gadA (coding for glutamate decarboxylase A)^22,23,24, tihe envelope stress pathway gene spy (coding for spheroplast protein Y)^25,26,27, the osmotic stress response pathway gene otsA (coding for trehalose-6-phosphate synthase, A subunit)²⁸ and the oxidative stress response pathway gene dps (coding for DNA-binding protein from starved cells)^29,30. In a second step, we PCR-amplified the stress gene promoter-yfp reporter construct from the chromosome and integrated it convergent to the constant part of the module that consisted of a transcriptional fusion of cfp to the promoter of the housekeeping gene frr (coding for ribosome-recycling factor)^31,32,33 in the λ attB site of the E. coli K-12 MG1655 chromosome (Fig. 2).

**Fig. 2: Schematic representation of an *E. coli* K-12 MG1655 *E. coli* cell with the relevant elements.**

The Pfrr-cfp fusion construct can serve as an alternative for the normalization of stress promoter response in cases where the optical properties of test substances do not allow for OD measurements^32,34. After Sanger sequencing of relevant module elements (chromosome-module and promoter-reporter gene junctions), we first tested the sensibility of the biosensor set containing the different modules to different concentrations of known inducers (Fig. 3).

**Fig. 3: Test of the reporter strains with known stressors.**

Briefly, we diluted the reporter strains 1:200 in 150 µl of M9 medium supplemented with casamino acids and 0.4% glucose in a 96-well plate, and grew the bacteria in the microplate reader to the logarithmic growth phase, recording optical density and fluorescence signals automatically. Once the bacteria reached logarithmic growth, we added 5 µl of either a dilution series of the stress-inducing chemical or the solvent of the stress-inducing chemical (control) and continued to record the response of the bacteria. This experimental setup allowed us to simultaneously determine (i) the minimal concentration of a test substance that affects the bacterial growth rate, (ii) the final optical density of the bacterial culture in the presence of a range of concentrations of a test substance (here defined as optical density at 20 h after the start of the experiment), (iii) the expression kinetics of stress response genes towards various concentrations of a test substance and, (iv) if a stress response towards a test substance is dependent on the concentration of the test substance.

pH stress biosensor (PgadA module)

We tested the pH stress biosensor (containing the gadA promoter module; PgadA module) by challenging the bacteria with acetic acid, as it was previously shown that gadA expression is strongly induced by acetate²¹. Adding acetic acid to 0.1% f.c. in the logarithmic growth phase (Fig. 3A, dashed line) had an impact on the growth rate of the bacteria. A mild reduction in the growth rate could still be observed at 0.05% acetic acid, but there was almost no difference at 0.025% acetic acid when compared to the control. The reduced growth rate at 0.1% acetic acid also correlated with the observed reduced final optical density at the end of the experiment (20 h), which was not the case at 0.05% acetic acid (Fig. 3B). In the absence of acetic acid stress, the YFP/CFP fluorescence signal ratio showed only a very mild increase over time from 170 to 250 min when the cells entered the stationary phase (Fig. 3C, control). This growth phase-dependent expression pattern is consistent with the previously described growth phase-and σ^S-dependent expression pattern of gadA²². In contrast to that, the YFP/CFP fluorescence signal ratio increased more strongly and in an acetic acid concentration-dependent manner after the addition of acetic acid (Fig. 3C, dashed line). This response lasted until the growth phase-dependent induction of the PgadA started (~250 min), and the overall YFP fluorescence signal was too strong to be recorded with the settings of the instrument around ~310 min. As the acetic acid stress response was immediately followed by the growth phase-dependent activation of PgadA, we calculated the total amount of YFP/CFP that was expressed for an hour following the addition of acetic acid (170 min). The total response of PgadA to the different concentrations of acetic acid was concentration-dependent (Fig. 3D) and significant enough to produce marked amounts of YFP/CFP at 0.025% acetic acid, which did not have a noticeable effect on growth rate or final optical density (Fig. 3A, B). We did not observe any concentration-dependent response of the other biosensors to the addition of acetic acid (Fig. S2, black arrows).

Envelope stress biosensor (Pspy module)

We tested the envelope stress biosensor (containing the spy promoter module; Pspy module) by challenging the bacteria with tannic acid, which was described as a strong inducer of spy expression³⁵. As shown in Fig. 3E, the addition of up to 67 µg/ml tannic acid (t = 170 min, dashed line) had only a mild impact on the growth rate and resulted in only a slight, concentration-dependent reduction of the final optical cell density at the end of the experiment (Fig. 3F). This finding was consistent with previous studies that showed growth inhibition at higher tannic acid concentrations³⁶. In our study, the YFP/CFP fluorescence signal ratio started to increase immediately after the addition of tannic acid, and the increase was concentration-dependent, lasting approximately to time point 270 min, after which no increase in the fluorescence signal was observed (Fig. 3G). In the absence of tannic acid, the Pspy module only expressed very small amounts of YFP/CFP over the whole course of the experiment, which was consistent with previous studies³⁵. As the response to tannic acid effectively ended at ~270 min, we used this endpoint for our calculations, calculating the total increase in YFP/CFP from the time point of induction until 270 min (Fig. 3H). We did not observe a concentration-dependent response of the other biosensors following the addition of tannic acid (Fig. S3, black arrows).

Osmotic stress biosensor (PotsA module)

We tested the osmotic stress biosensor (containing the otsA promoter module; PotsA module) by challenging the bacteria with different concentrations of NaCl (Fig. 3I–L). The challenge was relatively mild and affected the growth rate only at concentrations of 500 mM NaCl (Fig. 3I). However, the final OD at the end of the experiment was not affected (Fig. 3J). Similar to the PgadA module, the PotsA module started expressing YFP in the absence of an external stimulus at the transition to the stationary growth phase, as judged by the continuously increasing YFP/CFP ratio after ~230 min (compare control in Fig. 3K with control in Fig. 3C). The addition of NaCl to a final concentration of 500 mM, 250 mM, and 125 mM NaCl resulted in a concentration-dependent response of the PotsA module (Fig. 3K). As the osmotic stress response was followed by the growth phase-dependent activation of PotsA, we calculated the concentration-dependent response for the PotsA module for 1 h after the NaCl challenge (Fig. 3L). None of the other biosensors showed a concentration-dependent response to NaCl (Fig. S4).

SOS response biosensor (PrecA module)

We have previously tested the functionality of the SOS response biosensor module (containing the recA promoter module; PrecA module) by the assessment of an enzymatically inhibited SOS response²¹. Here, we tested the PrecA module by exposing the bacteria to different concentrations of the gyrase inhibitor sparfloxacin³⁷ (Fig. 3M–P, dashed lines). Sparfloxacin concentration-dependent effects on the growth rate were observable from approximately 200 min (Fig. 3M), and the final optical density was sparfloxacin concentration-dependent as well, as expected for an antibiotic (Fig. 3N). The response of the PrecA module to the sparfloxacin challenge at 170 min (dashed line) was rapid and lasted until ~300 min (Fig. 3O). In this period, the PrecA module of the control sample expressed only small amounts of YFP/CFP, whereas the total amount of YFP/CFP expressed in response to sparfloxacin was dependent on the concentration (Fig. 3P). None of the other biosensors was induced by the addition of sparfloxacin (Fig. S5). In contrast, the growth phase-dependent activation of the PgadA, PotsA, and Pdps modules showed a sparfloxacin-dependent reduction in YFP/CFP produced per period, as expected.

Oxidative stress biosensor (Pdps module)

We have previously tested the functionality of the oxidative stress biosensor (containing the dps promoter module; Pdps module) in E. coli K-12 extensively³⁰. The different concentrations of H₂O₂ tested here did not affect the growth rate (Fig. 3Q) and the final cell density at the end of the experiment (Fig. 3R). The Pdps module started to express YFP rapidly after the challenge with H₂O₂ as judged by the rapid increase in YFP/CFP (Fig. 3S) and the response lasted for ~30 min under these conditions, similar to what we have observed earlier³⁰. The growth phase-dependent induction of the Pdps module in the absence of H₂O₂ stress started around 220 min in this experiment, whereas the growth phase-dependent activation after the challenge with H₂O₂ was delayed (Fig. 3S). In order to separate the H₂O₂ response from the growth phase-dependent activation of the Pdps module, we calculated the total amount of YFP/CFP expressed for 30 min after induction. As for the other chemicals tested, the total response depended on the concentration of the test substance (Fig. 3T). The PgadA and PotsA modules were not induced by H₂O₂, but showed a delay in the growth phase-dependent activation, similar to the Pdps module (compare Fig. S6 to Fig. 3S). The H₂O₂ challenge did not induce the SOS biosensor as strongly as sparfloxacin (compare Fig. S6 to Fig. 3O). However, the YFP/CFP ratio was continuously increasing after the addition of H₂O₂ from 180 min to ~260 min, indicative of an induced bacterial subpopulation (see Fig. S11 of Supporting Information).

Photosensitizers induce different types of stress in E. coli

Oxidative stress induced by light-activated MB

Next, we tested different concentrations of MB for the activation of our biosensor set. In order to better separate a potential acid, osmotic, or oxidative stress response from the growth phase-dependent activation of the promoter, we added the test substance earlier (140 min; Fig. 4, dashed line). Without photoactivation, no effects of the tested concentrations of MB on the growth rate (Fig. 4A), final optical cell density (Fig. 4B), or induction of the oxidative stress biosensor were observed (Fig. 4 C). In addition, the growth phase-dependent activation of the module was not affected at all MB concentrations tested (Fig. 4C). MB also did not induce the rest of the biosensors without photoactivation at all concentrations tested (Fig. S7). In contrast, the Pdps module started to express YFP immediately after photoactivation of MB at all concentrations tested (note the YFP/CFP peak at 150 min in Fig. 4G). The final amount of YFP/CFP that was expressed in response to the photoactivation was dependent on the concentration of MB as well, however, higher at 1 µM MB than at 5 µM MB (Fig. 4H). There was only a mild effect on the growth rate of the bacteria observable after photoactivation of 5 µM MB (Fig. 4E), but no effect on final cell density (Fig. 4F) at all concentrations tested. We also tested the Pdps module on a plasmid³⁸ in an E. coli K-12 MG1655 rpoS mutant and observed induction after photoactivation of MB as well, which showed that MB is indeed activating the Pdps module via the σ⁷⁰ holoenzyme and not due to a general RpoS-mediated stress response (Fig. S8). None of the other biosensors showed induction after light activation of MB at all concentrations tested (Fig. S9). However, the PgadA and PotsA modules showed MB concentration-dependent time delay in growth phase-dependent activation, similar to what we have observed for H₂O₂ (compare PgadA and PotsA modules in Fig. S9 to Fig. S6).

**Fig. 4: MB is inducing oxidative stress in *E. coli* K-12 MG1655 after light activation.**

Envelope stress induced by light-activated SiPc

We next tested different concentrations of SiPc for the activation of our biosensor set. At all concentrations tested, SiPc did not activate the biosensors without photoactivation (Fig. S10), except for envelope stress. The Pspy module that showed a mild induction at 5 µM SiPc (Fig. 5C), resulted in slightly elevated total YFP/CFP expression as well (Fig. 5D). In addition, the growth rate was affected without photo activation at 5 µM SiPc (Fig. 5A), which also resulted in a reduced optical density at the end of the experiment (Fig. 5B). The other biosensors did not show any induction in the absence of photo activation, but the growth phase activation of the PgadA, PotsA and Pdps modules was either slightly delayed at 1 µM SiPc, or completely absent at 5 µM SiPc (Fig. S10). Photoactivation at a concentration of 5 µM SiPc caused an immediate stop of bacterial cell growth and/or cell division as judged by the stable optical density upon exposure to red light, whereas photoactivation at 1 µM SiPc had only a mild effect on the growth rate (Fig. 5E). However, in spite of the immediate and long-lasting effect on the growth of the bacteria (resuming growth was detectable at ~400 min), photoactivation of 5 µM SiPc reproducibly resulted in higher final optical densities at the end of the experiment than 5 µM SiPc without photo activation (compare Fig. 5B and 5F). Photoactivation resulted in an immediate induction of the Pspy module that lasted until ~200 min only at 1 µM SiPc, but not at the other concentrations tested (Fig. 5G). A significant total YFP/CFP expression in this period was also only observed for 1 µM SiPc (Fig. 5H). None of the other biosensors was induced after light activation of SiPc. In contrast, the growth phase-dependent activation of the PgadA, PotsA, and Pdps modules was either delayed or weaker at 1 µM SiPc, or completely absent at 5 µM SiPc in this period (Fig. S11).

**Fig. 5: SiPc is inducing envelope stress in *E. coli* K-12 MG1655.**

This is important, as higher concentrations of substances that kill bacteria are also very likely to interfere with the expression of reporter genes, as seen here at 5 µM MB and 5 µM SiPc (Figs. 4G and 5G).

Light-activated oxidative stress induced by MB is sensed by OxyR, envelope stress induced by SiPc is sensed by BaeR and CpxR

The fact that, except for the Pdps and Pspy modules, none of the other biosensors were induced following photoactivation of MB and SiPc indicated that the dps and spy promoters were not induced due to a general stress response. That the Pdps module was also induced by light activation of MB in an rpoS mutant background further supported the hypothesis that the Pdps module was specifically activated. However, if the activation of both modules by the stressors was specific, it should be absent in the absence of the relevant response regulators. In order to test this hypothesis, we decided to delete the regulators of Pdps and Pspy and in the oxidative stress and the envelope stress biosensor, respectively. Afterwards, we compared the response of the mutant biosensors to the PSs to the wild type.

Photoactivation of MB induces the Pdps module via OxyR

The dps promoter is regulated via OxyR and MntR^39,40. We deleted both genes in our oxidative stress biosensor strain by recombineering and repeated the experiment with photoactivation of 1 µM MB (Fig. 6). In the mntR-negative strain background, the response of the Pdps module to light activation of 1 µM MB was very similar to wild type and showed a prominent peak (Fig. 6A, 150 min) in Pdps module activity directly following the treatment (Fig. 6A, dashed line). This peak in promoter activity was completely absent in the control experiment (Fig. 6B). Notably, this peak in Pdps activity following the treatment was also completely absent in the oxyR mutant (Fig. 6C), and the module showed a very similar response in comparison to the water control experiment in the oxyR mutant (Fig. 6D). The total response of the modules to photoactivation of 1 µM MB was also similar in wild type and mntR mutant, but reduced to control levels in the oxyR mutant (compare Fig. 6E with Fig. 6F).

**Fig. 6: The MB-dependent induction of Pdps is not affected in mntR mutants, but absent in *oxyR* mutants.**

Photoactivation of SiPc induces the Pspy module via BaeR and CpxR

The spy promoter is regulated by the two-component system BaeSR and the regulator CpxR⁴¹. We deleted baeR and cpxR in our envelope stress biosensor strain by recombineering and repeated the experiment with photoactivation of 1 µM SiPc (Fig. 7). In the cpxR mutant (Fig. 7A), as well as in the baeR mutant (Fig. 7B) the Pspy promoter module response following photoactivation of SiPc was strongly reduced when compared to the wild-type. The total response of the Pspy module was reduced in the cpxR mutant relative to the wild-type (Fig. 7E) and not detectable in the baeR mutant, where the Pspy module response was indistinguishable from the water control (compare Fig. 7E, F).

**Fig. 7: The SiPc-dependent induction of Pspy is reduced in *cpxR* mutants and not detectable in *baeR* mutants.**

Discussion

Due to the current antibiotic crisis, aPDT has emerged as an alternative approach for antimicrobial therapy. However, the knowledge of the primary site of attack of photosensitizers is still limited⁴². Classical global approaches (e.g., proteomics or transcriptomics) used to identify sites of attack of different photosensitizers still have limitations with respect to the number of samples that can be processed. In addition, these approaches offer a very limited temporal resolution of the physiological response of the bacteria to the challenge. We therefore decided to analyze the response of a limited set of well-characterized stress gene promoter-reporter gene modules to a range of concentrations of PSs at a high temporal resolution. In order to avoid the limitations associated with the use of plasmids (such as the requirement for selection or stress-dependent copy number modifications)⁴³, we integrated our stress reporter modules in the λ attB site of the chromosome of the well-characterized E. coli K-12 strain MG1655 (Fig. 2). These modules contained a construct element that provides constitutive expression of cfp from the promoter of the housekeeping gene frr so that there is a normalization alternative available if the optical properties of a test substance do not permit optical density measurements^31,32,33. The second part of the modules was a construct that expresses YFP under control of a set of different stress-inducible promoters^{21,22,23,24,25,26,27}.

Our initial tests with substances that were described to activate the stress gene promoters showed that all biosensors used in this study exhibited a concentration-dependent effect to the challenges, albeit to varying degrees (Fig. 3). The absolute amount of YFP/CFP expressed in response to a challenge was dependent on the reporter module. For example, the maximal total amount of YFP/CFP expressed by the Pspy module was ~20 YFP/CFP at the highest tannic acid concentration (Fig. 3H), whereas the maximal total amount of YFP/CFP in response to NaCl was 1.8 YFP/CFP for the PotsA module (Fig. 3L). This result was expected, as the strength and extent of the response of the promoter is adjusted to the specific features of the corresponding stress response gene. These differences in fluorophore expression made it difficult to record the expression of the reporter gene for all stress reporter modules over the full time period of the experiment, as can be observed with the strong response of the PgadA module (Fig. 3C). Nevertheless, the detection of the stress response in logarithmic growth phase was readily possible for all biosensors used in this study with identical instrument settings (Fig. 3). In the future, N-terminal modifications of the coding sequence of the reporter gene could be used to further fine-tune the fluorophore expression of the modules to each other on the translational level⁴⁴. Another problem was the growth phase-dependent expression patterns of individual modules (see, for example, the activity of the control sample for PgadA and Pdps in Fig. 3C, S). One possibility to address this problem is to adjust the time point of the challenge in a way that allows for a better separation of the stress response from the growth phase-dependent activation of the module, which we have also done here (Figs. 3–6). Another possibility is the use of rpoS mutants, as long as the stress response of a module is not affected by the absence of RpoS. One such example is the Pdps module, which does not require RpoS for the OxyR-dependent activation in response to oxidative stress in the logarithmic growth phase (Fig. S7)³⁹. However, this will not be possible for the modules that are strictly dependent on RpoS, e.g., the PgadA and PotsA modules. Notably, these modules were reacting relatively slowly in response to known stressors (compare the rapid YFP/CFP increase of Pdps and Pspy module in Fig. 3S, G with the time-delayed increase in YFP/CFP for the PgadA and PotsA modules in Fig. 3C, K, respectively). Such a promoter kinetic would be consistent with an induction mechanism that first requires certain levels of expression and/or functional activation of an upstream regulator, here in particular RpoS that also results in a stress-independent, growth phase-dependent activation of the genes in the rpoS regulon⁴⁵. Another problem was the “cross-sensitivity” of E. coli stress gene promoters to a variety of stressors, especially if the promoter of a stress response gene is also recognized by RpoS. In such cases, a more comprehensive analysis of the response kinetics of the modules might help to dissect direct and indirect inducing effects of a stressor on a promoter, e.g., in the delayed responses of the PgadA and PotsA modules (see above).

Notably, the purpose of the functional tests here was merely to check, if we could measure a promoter response in high throughput in a 96-well plate format, e.g., an YFP/CFP signal significantly different from the control sample following the addition of a described test substance. Identically constructed reporter modules, e.g., for the stress genes relA, sulA, and dinB, did not show a measurable response to the described inducers. As these modules were therefore not suitable for us to detect the stressor within the framework of our experimental settings, we simply excluded them from our analysis in this instance. The use of 96-well plates for the assays usually allowed here for the rapid identification of a suitable concentration of a test substance for the read-out of a physiological response of the reporter modules without preliminary experiments. Testing a substance at a concentration that induces a physiological response of the bacteria, such as concentrations that do not cause a drop in OD, or an immediate and persistent growth arrest, was possible with just one test run. In order to confirm that a stress response following the addition of a test substance was indeed specific for the stress caused by the substance, it was necessary to measure the responses of several independent reporter modules and compare these to the responses of known stressors (compare e.g., Fig. 3 to Figs. S1–S5)^{22,23,24,25,26,27,28,29,30}. Whereas the specific activation of a module was already the first evidence that a certain substance was indeed inducing a specific type of stress in the bacteria, the absence of a promoter response in mutants of regulators could provide additional evidence that a promoter was indeed specifically activated/de-repressed in response to a stressor^{25,26,27,29,30}. Using a set of mutants of the regulators of Pspy⁴¹ and Pdps^39,40 modules, we showed here that SiPc activated Pspy via CpxR and BaeR, whereas MB induced Pdps via OxyR (Figs. 6 and 7). Future experiments will have to show whether other stress pathways that were not investigated here are also activated by SiPc or MB. Nevertheless, the parameters of the response of Pdps to MB and the response of Pspy to SiPc that were identified here are very likely to be helpful to effectively choose chemical concentrations and sampling time points for more comprehensive analyses of the global bacterial responses to MB and SiPc in aPDT by other methods, such as RNA-seq.

In aPDT, oxidative stress is believed to be a major factor that affects the viability of bacteria, and one of the protective cellular mechanisms is the DNA repair system. Remarkably different responses of the oxidative stress biosensor were detected for the two PSs used. A significant and dose-dependent increase in YFP production was observed with MB upon irradiation, whereas E. coli treated with SiPc did not activate the Pdps module. This could be related to the different mechanisms of action of the PSs. In diluted solutions, where MB is in a monomeric state, it acts as a type II photosensitizer generating a high amount of ¹O₂ similar to most of the tetrapyrrole-based PSs, including SiPc⁷_. However, pH⁴⁶ and different components of the biological media²⁰ can induce the formation of dimers and higher-order aggregates, where electron transfer reactions occur. This shifts the photosensitization mechanism to type I, promoting the generation of the superoxide anion radical O₂^∙-, which could enzymatically or through protonation be converted into other types of ROS, including highly cytotoxic hydrogen peroxide or hydroxyl radicals. Incubation of E. coli K-12 MG1655 with 5 µM of SiPc and irradiation resulted in an immediate stop of the increase of OD₅₉₅, whereas 1 µM SiPc had no marked effect on the growth rate (Fig. 5E). Nevertheless, the envelope stress biosensor was only induced at 1 µM SiPc, but not at higher, or lower concentrations of SiPc (Fig. 5G). This indicated that photoactivated SiPc, unlike MB, induced damage to the cell membrane, and that the concentration range in which a physiological envelope stress response of the bacteria to SiPc was both required and possible was relatively narrow. Interestingly, we observed a reduction of the growth rate at 5 µM SiPc in the absence of irradiation (Fig. 5A), which, in contrast to the irradiated samples, also resulted in a marked reduction of the final optical density at the end of the experiment (compare Fig. 5B to Fig. 5F). Under these conditions, the Pspy module showed activation after 5 µM SiPc treatment, indicating that SiPc was also, in the absence of light activation, to some extent affecting the normal membrane functions. This overall indicated that unlike MB, photoactivated SiPc mainly induced damage to the cell membrane. However, this was not surprising, because as a result of a very short lifetime of ¹O₂, aPDT-induced damage likely occurs only on the localization site of SiPc at concentrations that still allow for a physiological response of the cells⁴⁷. Indeed, the available literature strongly supports the hypothesis that hydrophobic tetrapyrrole-based PSs tend to accumulate in the cell membrane^48,49,50. Light activation of MB on the other hand is known to be mutagenic in E. coli since a long time and was shown to result in lesions of chromosomal DNA^51,52. The authors also reported no activation of the SOS system following aPDT with MB, which may be surprising, however, consistent with our data on the population level (Fig. S8). In order to damage DNA by ¹O₂ generation light activation of MB needs to occur in close spatial proximity. MB was shown to be a substrate for bacterial multi-drug efflux pumps, and the attachment of a multi-drug efflux pump inhibitor to MB was shown to enhance the efficacy of aPDT both in vitro and in vivo^53,54. Therefore, it is conceivable that due to its small size and hydrophilic characteristic, light activation of MB in contrast to SiPc indeed results in a substantial generation of ¹O₂ in the cytosol, which in turn results in an oxidative stress response and increase in the number of cells with activated SOS response within the bacterial population.

We have now identified both, the primary target compartment, as well as the reaction conditions at which PS-dependent modifications of components of the primary target compartment should be readily detectable. The latter is important, as reaction conditions that do not allow a physiological response of the bacteria are also likely to result in the disintegration of the cell, which could in turn permit the PS-dependent modification of cellular components that are not the primary target of the PSs. Therefore, an analysis of marker molecules of the primary target compartments should finally clarify the question, if SiPc-mediated damage is indeed mainly occurring at the membrane, whereas MB-mediated damage is rather found in the cytosol. In addition, our analysis revealed that SiPc was both inducing envelope stress in the absence of light activation as well as after light activation. This may be a basis for a potential resistance development to PSs and should eventually be taken into consideration in therapeutic regimes in the future as well.

Conclusion

This study demonstrates that chromosomally integrated stress reporter modules enable high-resolution, high-throughput assessment of bacterial stress responses to PSs used in aPDT. Using a panel of biosensors, we identified distinct stress signatures induced by MB and SiPc, suggesting different primary targets and mechanisms of action. MB primarily induced oxidative stress responses, consistent with cytosolic ¹O₂ generation, while SiPc mainly caused envelope stress, indicating membrane-associated damage. These insights provide a foundation for targeted investigations into PS mechanisms and support the development of optimized aPDT strategies with minimized resistance potential.

Methods

Bacterial strains used in this study

All bacterial strains used in this study are listed in Table S1.

Media and growth conditions

For the construction procedures, bacteria were grown in lysogeny broth (LB; containing 10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl) at 37 °C under aerobic shaking conditions. For solid medium, agar was added to a final concentration of 1.5% (w/v). If necessary, the medium was supplemented with kanamycin [30 µg/ml], ampicillin [100 μg/ml], and chloramphenicol [25 μg/ml]. For the stress response measurements, bacteria were grown in M9 minimal medium (M9 salts containing 33.9 g/L Na₂HPO₄, 15 g/L KH₂PO₄, 5 g/L NH₄Cl, 2.5 g/L NaCl) supplemented with 0.4% casamino acid and 0.4% glucose.

Oligonucleotides

All oligonucleotides used in this study are listed in Table S2.

Plasmids

All plasmids used in this study are listed in Table S3.

Construction of bacterial strains

Construction of E. coli K-12 MG1655 P_stress response gene-yfp-cat

For the construction of the stress gene promoter yfp fusions of gadA, spy and otsAB the pMB54³⁰ vector was used as template to generate PCR products in combination with the Q5® High-Fidelity DNA polymerase and primers, MC162/MC163, MC166/MC167 and MC178/MC179 and Q5® High-Fidelity DNA polymerase, respectively. The chromosomal fusions to the respective promoters were done in E. coli K-12 strain MG1655 by RedE/T recombineering⁵⁵. The verification of the correctness of the stress gene promoter-reporter gene fusion was afterwards done by amplifying the 5´ and 3´ junctions of the reporter and the chromosome by PCR. The 5´ junction PCR products (stress gene promoter – yfp) were additionally Sanger sequenced.

Construction of E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-P_stress response gene-yfp

For the construction of the final reporter modules, E. coli K-12 MG1655 Pstress response gene-yfp-cat were used as templates for PCR with primers MC164/MC206, MC168/MC207, MC180/MC209 and Q5® and High-Fidelity DNA polymerase. In addition, pMB54³⁰ was used as a template for PCR with the primers MBPD155/MC185 Q5® and High-Fidelity DNA polymerase. These intermediate templates were used in combination with the Q5® High-Fidelity DNA polymerase and the primers MC187/MC191, MC187/MC194, MC187/MC197, and MC187/MC183, respectively to generate DNA fragments for the final recombination step. The PCR products were then integrated convergent to the constant part of the module into the chromosome of E. coli K-12 MG1655 att::Pfrr-cfp-aph(3’)-Ia with RedE/T recombineering⁵⁵. The integration process resulted in the precise replacement of the aph(3’)-Ia gene with the Pstress response gene-yfp-cat fragments. The construction was completed by Sanger sequencing of the 5´and 3´ junctions amplified with the primers MBP206/MBP279.

Construction of E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pdps mntR::aph

E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pdps mntR::aph was constructed by transforming pKD46 into E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pdps. Afterward, we generated a PCR product using the primers MBP376/MBP377 and pKD4 as a template with the Q5® High-Fidelity DNA polymerase and replaced mntR with aph by recombineering⁵⁵. The 5´and 3´ junctions of chromosome and aph were afterwards verified by PCR with primers MBP378/UD2822 and MBP373/MBP379, respectively.

Construction of E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pdps oxyR::aph

E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pdps oxyR::aph was constructed by transforming pKD46 into E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pdps. Afterward, we generated a PCR product using primers MBP371 / MBP372 and pKD4 as template with Q5® High-Fidelity DNA polymerase and replaced oxyR by aph by recombineering⁵⁵. The 5´ and 3´ junctions of chromosome and aph were afterwards verified by PCR with primers MBP374/UD2822 and MBP373/MBP375, respectively.

Construction of E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pspy baeR::aph

E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pspy baeR::aph was constructed by transforming pKD46 into E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pspy. Afterward, we generated a PCR product using primers MBP384/MBP385 and pKD4 as template with Q5® High-Fidelity DNA polymerase and replaced baeR by aph by recombineering⁵⁵. The 5´ and 3´ junctions of chromosome and aph were afterwards verified by PCR with primers MBP386/UD2822 and MBP373/MBP387, respectively.

Construction of E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pspy cpxR::aph

E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pspy cpxR::aph was constructed by transforming pKD46 into E. coli K-12 MG1655 att::Pfrr-cfp-cat-yfp-Pspy. Afterward, we generated a PCR product using the primers MBP380/MBP381 and pKD4 as a template with the Q5® High-Fidelity DNA polymerase and replaced cpxR with aph by recombineering⁵⁵. The 5´and 3´ junctions of the chromosome and aph were afterwards verified by PCR with the primers MBP382/UD2822 and MBP373/MBP383, respectively.

Reporter module response measurements

The bacteria were grown in M9 medium supplemented with 0.4% casamino acids and 0.4% glucose in shaking conditions at 37 °C overnight in an Infors HT Multitron Standard incubator (Infors, Einsbach, Germany). The next day, the overnight cultures were inoculated 1:200 into 150 µl of freshly prepared M9 medium supplemented with 0.4% casamino acid and 0.4% glucose in black μ-clear plates with transparent bottom (Greiner Bio-One, Frickenhausen, Germany), covered with a transparent lid. Afterwards, the plate was incubated at 37 °C in a TECAN Infinite 200pro instrument (TECAN, Männedorf, Switzerland) for the indicated time with orbital shaking with an amplitude of 1.5 mm. Optical density was measured at 595 nm, and the fluorescence signal for CFP (excitation wavelength 442, emission wavelength 492 nm, detector gain 85) and for YFP (excitation wavelength 514 nm, emission wavelength 550 nm, detector gain 100) were recorded automatically every 10 min. At the indicated time point, 5 µl or the same volume of serial dilutions of the test substance, as well as a negative control (solvent) were added to the wells, and the measurement was continued until the time point of 20 h, which was the endpoint of each experiment. The optical density signal was corrected for the blank. The YFP signal was normalized to the CFP signal of each well for each time point. Promoter activity was calculated by subtracting the YFP/CFP value of the previous time point: promoter activity [d(YFP/CFP)]_n = (YFP/CFP)_n − (YFP/CFP)_n−1. For the photosensitizing agent, after adding the chemicals to the cultures, an extra 5 min shaking was required to properly mix the agent with the bacterial cells followed by 15 min of light exposure (660 nm ± 26 nm, 10 mW/cm², 9 J/cm²) to photo-activate the compounds.

Statistics and reproducibility

For the total response measurements, experiments were conducted in triplicate, and the results are expressed as mean ± standard deviation. Statistical data analysis was performed using GraphPad Prism 9 (GraphPad Software). Means were compared using a one-way ANOVA analysis across multiple groups. For all tests, a P value < 0.05 was considered statistically significant.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The source data for Figs. 3–7 are provided in Supplementary Data 1. The data that support other findings of this study are provided in the supporting information or available from the corresponding authors upon reasonable request.

References

Noimark, S., Dunnill, C. W. & Parkin, I. P. Shining light on materials—a self-sterilising revolution. Adv. Drug Deliv. Rev. 65, 570–580 (2013).
Article CAS PubMed Google Scholar
Wainwright, M. et al. Photoantimicrobials—are we afraid of the light?. Lancet Infect. Dis. 17, e49–e55 (2017).
Article PubMed Google Scholar
Youf, R. et al. Antimicrobial photodynamic therapy: latest developments with a focus on combinatory strategies. Pharmaceutics 2021, 13 (1995).
Leung, B. et al. Development of thermosensitive hydrogel containing methylene blue for topical antimicrobial photodynamic therapy. J. Photochem Photobio. B 203, 111776 (2020).
Article CAS Google Scholar
Kharkwal, G. B., Sharma, S. K., Huang, Y.-Y., Dai, T. & Hamblin, M. R. Photodynamic therapy for infections: clinical applications. Lasers Surg. Med. 43, 755–767 (2011).
Article PubMed PubMed Central Google Scholar
Wen, X. et al. Potassium iodide potentiates antimicrobial photodynamic inactivation mediated by rose bengal in in vitro and in vivo studies. Antimicrob. Agents Chemother. 61, e00467-17 (2017).
Article PubMed PubMed Central Google Scholar
Hamblin, M. R. & Abrahamse, H. Oxygen-independent antimicrobial photoinactivation: type III photochemical mechanism?. Antibiotics 9, 53 (2020).
Article CAS PubMed PubMed Central Google Scholar
Dosselli, R. et al. Molecular targets of antimicrobial photodynamic therapy identified by a proteomic approach. J. Proteom. 77, 329–343 (2012).
Article CAS Google Scholar
Alves, E. et al. Protein profiles of Escherichia coli and Staphylococcus warneri are altered by photosensitization with cationic porphyrins. Photochem. Photobio. Sci. 14, 1169–1178 (2015).
Article CAS Google Scholar
Segalla, A. et al. Photophysical, photochemical and antibacterial photosensitizing properties of a novel octacationic Zn(II)-phthalocyanine. Photochem. Photobiol. Sci. 1, 641–648 (2002).
Article CAS PubMed Google Scholar
Vieira, D. B. & Carmona-Ribeiro, A. M. Cationic lipids and surfactants as antifungal agents: mode of action. J. Antimicrob. Chemother. 58, 760–767 (2006).
Article CAS PubMed Google Scholar
Jia, H.-R., Zhu, Y.-X., Chen, Z. & Wu, F.-G. Cholesterol-assisted bacterial cell surface engineering for photodynamic inactivation of Gram-positive and Gram-negative bacteria. ACS Appl. Mater. Interfaces 9, 15943–15951 (2017).
Article CAS PubMed Google Scholar
Wang, B., Wang, M., Mikhailovsky, A., Wang, S. & Bazan, G. C. A membrane-intercalating conjugated oligoelectrolyte with high-efficiency photodynamic antimicrobial activity. Angew. Chem. Int Ed. Engl. 56, 5031–5034 (2017).
Article CAS PubMed Google Scholar
Galstyan, A. & Dobrindt, U. Breaching the wall: morphological control of efficacy of phthalocyanine-based photoantimicrobials. J. Mater. Chem. B 6, 4630–4637 (2018).
Article CAS PubMed Google Scholar
Galstyan, A., Ricker, A., Nüsse, H., Klingauf, J. & Dobrindt, U. Exploring the impact of coordination-driven self assembly on the antibacterial activity of low-symmetry phthalocyanines. ACS Appl. Bio Mater. 3, 400–411 (2020).
Article CAS PubMed Google Scholar
Alves, E. et al. An insight on bacterial cellular targets of photodynamic inactivation. Future Med Chem. 6, 141–164 (2014).
Article CAS PubMed Google Scholar
Galstyan, A., Schiller, R. & Dobrindt, U. Boronic acid functionalized photosensitizers: a strategy to target the surface of bacteria and implement active agents in polymer coatings. Angew. Chem. Int. Ed. 56, 10362–10366 (2017).
Article CAS Google Scholar
Tardivo, J. P. et al. Methylene blue in photodynamic therapy: from basic mechanisms to clinical applications. Photodiagnosis Photodyn. Ther. 2, 175–191 (2005).
Article CAS PubMed Google Scholar
Galstyan, A., Putze, J. & Dobrindt, U. Gaining access to bacteria through (reversible) control of lipophilicity. Chem. – A Eur. J. 24, 1178–1186 (2018).
Article CAS Google Scholar
Galstyan, A. & Dobrindt, U. Determining and unravelling origins of reduced photoinactivation efficacy of bacteria in milk. J. Photochem. Photobio. B 197, 111554 (2019).
Article CAS Google Scholar
Peng, L. et al. A robust one-step recombineering system for enterohemorrhagic Escherichia coli. Microorganisms 10, 1689 (2022).
Article CAS PubMed PubMed Central Google Scholar
Waterman, S. R. & Small, P. L. C. Transcriptional expression of Escherichia coli glutamate-dependent acid resistance genes gadA and gadBC in an hns rpoS mutant. J. Bacteriol. 185, 4644–4647 (2003).
Article CAS PubMed PubMed Central Google Scholar
Castanié-Cornet, M.-P. et al. Acid stress response in Escherichia coli: mechanism of regulation of gadA transcription by RcsB and GadE. Nucleic Acids Res. 38, 3546–3554 (2010).
Article PubMed PubMed Central Google Scholar
Arnold, C. N., McElhanon, J., Lee, A., Leonhart, R. & Siegele, D. A. Global analysis of Escherichia coli gene expression during the acetate-induced acid tolerance response. J. Bacteriol. 183, 2178–2186 (2001).
Article CAS PubMed PubMed Central Google Scholar
Srivastava, S. K., Lambadi, P. R., Ghosh, T., Pathania, R. & Navani, N. K. Genetic regulation of Spy gene expression in Escherichia coli in the presence of protein unfolding agent ethanol. Gene 548, 142–148 (2014).
Article CAS PubMed Google Scholar
Raffa, R. G. & Raivio, T. L. A third envelope stress signal transduction pathway in Escherichia coli. Mol. Microbiol 45, 1599–1611 (2002).
Article CAS PubMed Google Scholar
Raivio, T. L., Laird, M. W., Joly, J. C. & Silhavy, T. J. Tethering of CpxP to the inner membrane prevents spheroplast induction of the Cpx envelope stress response. Mol. Microbiol. 37, 1186–1197 (2000).
Article CAS PubMed Google Scholar
Gunasekera, T. S., Csonka, L. N. & Paliy, O. Genome-wide transcriptional responses of Escherichia coli K-12 to continuous osmotic and heat stresses. J. Bacteriol. 190, 3712–3720 (2008).
Article CAS PubMed PubMed Central Google Scholar
Calhoun, L. N. & Kwon, Y. M. Structure, function and regulation of the DNA-binding protein dps and its role in acid and oxidative stress resistance in Escherichia coli: a review. J. Appl Microbiol. 110, 375–386 (2011).
Article CAS PubMed Google Scholar
Berger, M. et al. Genes on a wire: the nucleoid-associated protein hu insulates transcription units in Escherichia coli. Sci. Rep. 6, 31512 (2016).
Article CAS PubMed PubMed Central Google Scholar
Liu, M. et al. Global transcriptional programs reveal a carbon source foraging strategy by Escherichia coli. J. Biol. Chem. 280, 15921–15927 (2005).
Article CAS PubMed Google Scholar
Chouikha, I., Bree, A., Moulin-Schouleur, M., Gilot, P. & Germon, P. Differential expression of iutA and ibeA in the early stages of infection by extra-intestinal pathogenic E. coli. Microbes Infect. 10, 432–438 (2008).
Article CAS PubMed Google Scholar
Janosi, L., Shimizu, I. & Kaji, A. Ribosome recycling factor (ribosome releasing factor) is essential for bacterial growth. Proc. Natl. Acad. Sci. USA 91, 4249–4253 (1994).
Article CAS PubMed PubMed Central Google Scholar
Heng, S. S. J., Chan, O. Y. W., Keng, B. M. H. & Ling, M. H. T. Glucan biosynthesis protein G is a suitable reference gene in Escherichia coli K-12. ISRN Microbiol. 2011, 469053 (2011).
Article PubMed PubMed Central Google Scholar
Zoetendal, E. G., Smith, A. H., Sundset, M. A. & Mackie, R. I. The BaeSR two-component regulatory system mediates resistance to condensed tannins in Escherichia coli. Appl. Environ. Microbiol. 74, 535–539 (2008).
Article CAS PubMed Google Scholar
Min, E. R., Pinchak, W. E., Anderson, R. C. & Callaway, T. R. Effect of tannins on the in vitro growth of Escherichia coli O157:H7 and in vivo growth of generic Escherichia coli excreted from steers. J. Food Prot. 70, 543–550 (2007).
Article CAS PubMed Google Scholar
Piddock, L. J. & Zhu, M. Mechanism of action of sparfloxacin against and mechanism of resistance in Gram-negative and Gram-positive bacteria. Antimicrob. Agents Chemother. 35, 2423–2427 (1991).
Article CAS PubMed PubMed Central Google Scholar
Schiller, P. et al. The superior adherence phenotype of E. coli O104:H4 is directly mediated by the aggregative adherence fimbriae type I. Virulence 12, 346–359 (2021).
Article CAS PubMed PubMed Central Google Scholar
Altuvia, S., Almirón, M., Huisman, G., Kolter, R. & Storz, G. The Dps promoter is activated by OxyR during growth and by IHF and sigma S in stationary phase. Mol. Microbiol. 13, 265–272 (1994).
Article CAS PubMed Google Scholar
Yamamoto, K., Ishihama, A., Busby, S. J. W. & Grainger, D. C. The Escherichia coli K-12 MntR miniregulon includes Dps, which encodes the major stationary-phase DNA-binding protein. J. Bacteriol. 193, 1477–1480 (2011).
Article CAS PubMed PubMed Central Google Scholar
Yamamoto, K., Ogasawara, H. & Ishihama, A. Involvement of multiple transcription factors for metal-induced spy gene expression in Escherichia coli. J. Biotechnol. 133, 196–200 (2008).
Article CAS PubMed Google Scholar
Sabino, C. P. et al. The biochemical mechanisms of antimicrobial photodynamic therapy. Photochem. Photobiol. 99, 742–750 (2023).
Article CAS PubMed Google Scholar
Berger, M., Aijaz, I., Berger, P., Dobrindt, U. & Koudelka, G. Transcriptional and translational inhibitors block SOS response and shiga toxin expression in enterohemorrhagic Escherichia coli. Sci. Rep. 9, 18777 (2019).
Article CAS PubMed PubMed Central Google Scholar
Kudla, G., Murray, A. W., Tollervey, D. & Plotkin, J. B. Coding-sequence determinants of gene expression in Escherichia coli. Science 324, 255–258 (2009).
Article CAS PubMed PubMed Central Google Scholar
Adams, J. et al. The timing of transcription of RpoS-dependent genes varies across multiple stresses in Escherichia coli K-12. mSystems 8, e0066323 (2023).
Article PubMed Google Scholar
Pottier, R., Bonneau, R. & Joussot-Dubien, J. pH dependence of singlet oxygen production in aqueous solutions using toluidine blue as a photosensitizer. Photochem. Photobio. 22, 59–61 (1975).
Article CAS Google Scholar
Dahl, T. A., Midden, W. R. & Hartman, P. E. Comparison of killing of Gram-negative and Gram-positive bacteria by pure singlet oxygen. J. Bacteriol. 171, 2188–2194 (1989).
Article CAS PubMed PubMed Central Google Scholar
Komagoe, K., Kato, H., Inoue, T. & Katsu, T. Continuous real-time monitoring of cationic porphyrin-induced photodynamic inactivation of bacterial membrane functions using electrochemical sensors. Photochem. Photobio. Sci. 10, 1181–1188 (2011).
Article CAS Google Scholar
Preuss, A. et al. Photoinactivation of Escherichia coli (SURE2) without intracellular uptake of the photosensitizer. J. Appl Microbiol. 114, 36–43 (2013).
Article CAS PubMed Google Scholar
Pudziuvyte, B. et al. Alterations of Escherichia coli envelope as a consequence of photosensitization with tetrakis(N-ethylpyridinium-4-yl)porphyrin tetratosylate. Photochem. Photobio. Sci. 10, 1046–1055 (2011).
Article CAS Google Scholar
Webb, R. B., Hass, B. S. & Kubitschek, H. E. Photodynamic effects of dyes on bacteria. II. genetic effects of broad-spectrum visible light in the presence of acridine dyes and methylene blue in chemostat cultures of Escherichia coli. Mutat. Res 59, 1–13 (1979).
Article CAS PubMed Google Scholar
Menezes, S., Capella, M. A. & Caldas, L. R. Photodynamic action of methylene blue: repair and mutation in Escherichia coli. J. Photochem Photobio. B 5, 505–517 (1990).
Article CAS Google Scholar
Tegos, G. P. & Hamblin, M. R. Phenothiazinium antimicrobial photosensitizers are substrates of bacterial multidrug resistance pumps. Antimicrob. Agents Chemother. 50, 196–203 (2006).
Article CAS PubMed PubMed Central Google Scholar
Rineh, A. et al. Attaching NorA efflux pump inhibitors to methylene blue enhances antimicrobial photodynamic inactivation of Escherichia coli and Acinetobacter baumannii in vitro and in vivo. Bioorg. Med Chem. Lett. 28, 2736–2740 (2018).
Article CAS PubMed PubMed Central Google Scholar
Datsenko, K. A. & Wanner, B. L. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640–6645 (2000).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

Our work was supported by the German Federal Ministry for Education and Research (grant no. 16GW0246) and by the Interdisciplinary Center for Clinical Research, Medical Faculty Münster (grant no. Dob2/022/16). D.T. was supported by the Medizinerkolleg (MedK) Münster program of the University of Münster. Support by the Münster Graduate School of Evolution (MGSE) to M.C. is gratefully acknowledged. A.G. thanks the German Research Foundation (GA 2362/6-1) and University of Duisburg Essen for the support. We thank Susanne Lindgren for the critical reading of the manuscript.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Author information

Authors and Affiliations

Institute of Hygiene, University of Münster, Münster, Germany
Marco Chittò, David Tutschner, Ulrich Dobrindt & Michael Berger
AO Research Institute Davos, Davos, Switzerland
Marco Chittò & Ulrich Dobrindt
University of Duisburg-Essen, Faculty of Chemistry, Center for Nanointegration Duisburg Essen (CENIDE), Center for Water and Environmental Research (ZWU) and Center for Molecular Biotechnology (ZMB), Universitätsstrasse 5, Essen, Germany
Anzhela Galstyan

Authors

Marco Chittò
View author publications
Search author on:PubMed Google Scholar
David Tutschner
View author publications
Search author on:PubMed Google Scholar
Ulrich Dobrindt
View author publications
Search author on:PubMed Google Scholar
Anzhela Galstyan
View author publications
Search author on:PubMed Google Scholar
Michael Berger
View author publications
Search author on:PubMed Google Scholar

Contributions

M.B., A.G., and U.D. conceived and designed the study. M.C., A.G., and D.T. performed experiments and analyzed the data. M.C., A.G., and M.B. wrote the manuscript. All authors approved the final version of the manuscript.

Corresponding authors

Correspondence to Anzhela Galstyan or Michael Berger.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Communications Biology thanks Hisato Kato and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Haichun Gao and Tobias Goris. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Materials

Supplementary Data 1

Reporting Summary

Transparent Peer Review file

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Chittò, M., Tutschner, D., Dobrindt, U. et al. Photosensitizer-specific bacterial stress responses in Escherichia coli reveal distinct targets in photoinduced inactivation. Commun Biol 8, 1413 (2025). https://doi.org/10.1038/s42003-025-08881-4

Download citation

Received: 17 February 2025
Accepted: 11 September 2025
Published: 01 October 2025
DOI: https://doi.org/10.1038/s42003-025-08881-4