Introduction

According to the European Union Commission and World Health Organisation, environmental pollution by pharmaceutical residues is now one of the increasingly significant problems. This fact is forcing scientists to conduct more extensive ecotoxicological studies on the toxicity, genotoxicity and effects of drugs and their metabolites on living organisms. Microbial whole-cell biosensors have emerged as a promising tool for pre-screening the effects of such pollutants on living organisms.

Different antibiotics (ampicillin, kanamycin, nalidixic acid) and anti-inflammatory and analgesic drugs are commonly used to treat human and animal bacterial infections and inflammation1. Nalidixic acid, ampicillin and kanamycin are antibiotics detected in sewage and water worldwide. Nalidixic acid comes from the group of quinolones, which play an important part in bacterial infection treatment. The antibacterial mechanism of action of nalidixic acid is by interfering with bacterial DNA replication through inhibition of DNA synthesis. Ampicillin belongs to the β-lactam antibiotics of the aminopenicillin group, a broad-spectrum antibiotic that is sensitive to bacterial β-lactamases. The mechanism of action of ampicillin is by interfering with the bacterial cell wall biosynthesis. Kanamycin belongs to the group of aminoglycoside antibiotics and it exerts bactericidal activity by binding directly to the 16 S rRNA in the 30 S subunit of the ribosome and interfering with the translocation step of the translation process2. Due to the frequent use of analysed drugs in treatment, their metabolites end up in municipal sewage and become a water contamination problem. The presence of antibiotic residues in water brings the risk of resistance of pathogenic bacteria, as well as harmful effects on beneficial bacteria that inhabit the digestive system of mammals3. The solution to the problem of detecting harmful substances and their metabolites in water is bacterial biosensors4,5. Biosensors that use the entire microorganism (e.g. bacterial cells) as a biological component are called whole-cell biosensors6. Their action is based on physiological7, and biochemical8,9 changes occurring within the microorganism affected by the environmental change10. These are specific reactions between available molecules and receptors, as well the reactions involving nucleic acids and complementary sequences11. Bacterial biosensors are often based on the respiratory and metabolic processes of the bacterial cell, and the substance detected by the biosensor is usually a substrate or inhibitor of these processes12. As a result of the reactions taking place within the bacterial cell, a signal is created in the form of emission or absorption of light, uptake of gases, or their release13,14. At this time, the transducer transforms the biological signal into a measurable signal, which is amplified and processed and can be used for analyses15.

The advantages of whole-cell biosensors are low price16, versatility17, and high sensitivity18, as well as the ability to adapt to various environmental conditions19, and the ability to modify the body depending on the detected substance20. The bacterial biosensors have the possibility of genetic modification of bacteria using mutations and DNA recombination21. Bacterial cells are also an exceptionally good source of obtaining intracellular enzymes12. Bacteria used as biosensors have plasmids with a regulatory promoter and a reporter protein that may occur naturally in the cell or be the result of genetic modification. As a result of contact of the biosensor with a given substance, the regulatory promoter is activated and then the regulatory protein produces a detectable, measurable signal22. This operating scheme is used by bioluminescent biosensors, due to the possibility of monitoring changes in real time23. They are a real solution for warning about the inflow of compounds that are highly toxic to microorganisms used in wastewater treatment24,25. Luminescent biosensors are characterized by fast response and high sensitivity12. In biosensors which use the luciferase gene, lux genes that encode bacterial luciferase, genes are located in a plasmid construct with a specifically induced promoter. Expression of the lux operon produces all the necessary components needed for the emission of light energy by bacteria26. These genes require active cellular metabolism. There are two patterns of analyte detection using luminescence-based biosensors: analyte-induced, which results in increased lux gene expression, or when the analyte suppresses cell life processes, resulting in decreased lux gene expression27. The detected substance can affect receptors on the cell wall surface or penetrate it and then bind to a specific receptor located inside the bacteria cell. In both cases, the signal is transmitted to the promoter, which is a consequence of the transcription of reporter genes (luxCDABE). With increasing luciferase synthesis, luminescence increases, which is a measurable signal. The signal coming from biosensors with the lux gene depends on the concentration of the tested substance (analyte). Low concentrations do not affect bacterial cells, and the luminescence is constant. At higher analyte concentrations, the promoter is activated, increasing luminescence. When the concentration of the analyte is very high and it is also toxic to the cell, vital functions are interrupted and the luminescence decreases28.

Reporter genes that are used in luminescent biosensors are designed to transform the cell’s response into a measurable signal. This is a process that is major for the operation of the biosensor. The most important genes used in bacterial biosensors are the lacZ gene encoding the enzyme β-galactosidase28, the crt operon responsible for the synthesis of carotenoid29, gfp (green fluorescent protein)30,31, lux genes (encoding bacterial luciferase)32, and luc genes (encoding eukaryotic luciferases)28. In the presented work we used E. coli luminescent bioreporter in the stationary phase.

Bacterial luciferase genes (luxCDABE) are obtained from strains of luminescent bacteria, including species such as Vibrio fischeri, Vibrio harveyi, Photobacterium phosphoreum, and Photobacterium leiognathi. In all species mentioned, the luxA and lux B genes encode luciferase subunits, while luxCDE encodes a fatty acid reductase complex33. Bacterial luciferases are heterodimers that oxidize long-chain aldehydes with atmospheric oxygen using reduced flavin mononucleotide32. The reaction results in the emission of blue-green light.

Results and discussion

This work studied the expression of four different gene constructs in E. coli SM lux biosensor strains—E. coli SM342 recA:luxCDABE, E. coli SM335 soxS:luxCDABE, E. coli SM343 micF:luxCDABE and E. coli SM346 rpoB:luxCDABE after exposure to three different antibiotics (nalidixic acid, ampicillin, kanamycin) respectively, and the anti-inflammatory drug—diclofenac. Together with a zero trial (without drugs)—five different series of results were obtained, depending on the gene construct used in the E. coli SM lux biosensor strains and the substance tested.

As a result of contact of the E. coli SM lux biosensor strains with a studied substance, the regulatory promoter is activated and then stimulation of the regulatory protein, which produces a detectable, measurable luminescence signal22. Signal induction was analysed for the concentration of the substance (analyte) incubated with every gene construct and the incubation time. It is a real solution for warning about the inflow of compounds that are highly toxic to microorganisms used in wastewater treatment.

The expression of recA:luxCDABE in E. coli SM342 lux biosensor after antibiotics—nalidixic acid

Incubation of the E. coli SM342 recA:luxCDABE strain in various concentrations of nalidixic acid results in the expression of the used gene construct. Gene expression occurs under the influence of each of the tested concentrations, increasing with the duration of incubation (Fig. 1). The highest increase in recA promoter induction in E. coli SM342 recA:luxCDABE culture was recorded at the concentrations of 10 µg/mL of nalidixic acid, after 30 min (3.5%) and 1 h (5.4%). The reaction was stronger with increasing concentration, where 10 µg/mL was the threshold of the positive reaction. The strongest inhibition of recA promoter after in E. coli SM342 recA:luxCDABE culture was recorded after 1 h at the concentrations from 0,625 to 5 µg/mL (respectively 67.5% and 56.03%) of nalidixic acid (Fig. 1). These are statistically significant differences (Supplementary Fig. S1a, Supplementary Fig. S2a). With bioreporters, a positive reaction shows that there is stress in the cell, whereas a negative reaction can be caused by, for example, the cell dying or inhibition of other cell processes. Nalidixic acid, or 1-ethyl-7-methyl-4-oxo-(1,8)naphthyridine-3-carboxylic acid, comes from the group of quinolones, which play an important part in bacterial infection treatment. This is confirmed by the research results where luminescent biosensor is characterized by fast response “time 0” and high sensitivity12. Nalidixic acid affects DNA gyrase in bacterial cells. Inhibitors bind to and stabilize a transient covalent DNA protein intermediate in the gyrase reaction cycle, termed the cleavage complex. Genotoxicity results from the conversion of cleavage complexes into DNA breaks by an as-yet-unknown mechanism. The use of quinolone antibiotics induces the SOS response in bacterial cells. Moreover, for a response to occur during the action of nalidixic acid, RecA protein and RecBC proteins must be present in the cell34. The RecA protein is a not genetically engineered bacterial protein that enhances the interaction and exchange of homologous DNA strands. It is sensitive to genotoxic stress35. It regulates repair mechanisms in the cell, catalyzes the formation of base pairs between a single-stranded DNA molecule and its complementary double-stranded DNA sequence, and induces the SOS response36. The RecA protein is an activator of SOS operons, which is an example of an adaptive response of bacteria to stress. SOS is activated when single-stranded DNA (ssDNA), oligonucleotides or free nucleotide triphosphates appear in the cell, which indicates inhibition of replication. When there is a significant damage in the cell, the signal sent by ssDNA and ATP causes a change in the conformation of the RecA protein, which stimulates the autoproteolysis of the LexA protein and the induction of the SOS system15. When the DNA damage signal disappears, the RecA protein returns to its inactive conformation, which no longer causes autoproteolysis of LexA. It has been shown that the concentration of RecA protein in cells reaches a maximum after induction of the SOS response, which occurs much later than the derepression of the recA promoter. The recA gene expression increase counteracts the RecA protein concentration weakening in cells, which is caused by cell fibrosis37.

Fig. 1
figure 1

Signal induction [%] pattern (standardised for control—100%) of the recA promoter after E. coli SM342 recA:luxCDABE incubation with nalidixic acid (NA). Standard error bars indicate the standard error from the mean of four independent replicates.

Full size image

Descriptive statistics are presented in Supplementary Table S1. Variation in the induction value of the recA promoter in E. coli SM342 recA:luxCDABE after incubation with nalidixic acid depended on the concentration of the analyte and the incubation time of the strain (Fig. 1). Statistically significant differences occur between the incubation time of the E. coli strain versus to the control (“time 0”), respectively, to time after 30 min. (p = 0.000207) and after 1 h (p = 0.000059), and after 30 min. and after 1 h (p = 0.000013) (Supplementary Table S2). Taking into account the concentration of the analyte on the variability of promoter induction versus the control (0 µg/ml), statistically significant differences occur in concentrations 10 µg/mL (p = 0.015825) and 0.625 µg/mL (p = 0.020013) and between concentrations 2.5 µg/mL and 10 µg/mL (p = 0.037487) (Supplementary Table S2). The strongest gene expression over time was observed during incubation of the E. coli strain with the selected gene construct in the concentration of 10 µg/mL of the analyte (Supplementary Fig. S1b, Supplementary Fig S2b). It was found that after 30 min, and after 1 h of incubation at a concentration of 10 µg/ml, the stimulation of the promoter was the highest (Fig. 1). The RecA protein is the main regulator of SOS in the bacterial cell and works through three different mechanisms. These proteins catalyze the autoproteolysis of the transcription repressor LexA, inducing the expression of over 40 SOS genes. The inductive signal is DNA damage encountered at the replication fork. RecA also mediates recombinant DNA repair at double- and single-strand break sites.

The lowest gene expression and a negative reaction compared to the control (0 µg/mL) was obtained for nalidixic acid at a concentration of 0.625 µg/mL (67.5%) and after 1-h incubation of E. coli SM342 recA:luxCDABE biosensor strain (Fig. 1). The inhibition of recA promoter and luminescence signal in the analyte’s lowest concentration (0.625 ug/mL) showed that it affects bacterial cycles and processes. Nalidixic acid is an antibiotic used in medicine, mainly in urinary tract bacterial infections38, and has high antimicrobial activity. It is detected in hospital sewage in concentrations of up to several μg/l, also in industrial sewage, and even in treated sewage. The consequence is that the antibiotic and its residues enter the aquatic environment. As a result of the common use of this antibiotic, bacteria have developed drug resistance39,40. To improve the effect of nalidixic acid, changes were introduced in the basic quinolone nucleus to synthesize new derivatives of nalidixic acid, 1,2,4-triazole, diacyl, and sulfonyl acyl hydrazine. The molecules produced by combining different pharmacophores show a better antimicrobial profile than the parent compound41.

The expression soxS:luxCDABE in E. coli SM335 lux biosensor after ampicillin

Incubation of the E. coli SM335 strain with the soxS:luxCDABE gene construct in various concentrations of ampicillin causes gene expression at each of the tested concentrations (Supplementary Fig. S3). The contact of the biosensor with the ampicillin (after 30 min of incubation) produces strong soxS promoter inhibition in E. coli SM335 soxS:luxCDABE culture for all concentrations except 10 µg/mL (Fig. 2, Supplementary Fig. S4). At a concentration of ampicillin of 10 µg/mL, the stimulation of the soxS promoter in E. coli SM335 soxS:luxCDABE culture was recorded, at each time tested versus to control (“time 0”). Descriptive statistics are presented in Supplementary Table S1. Statistically significant differences occur between the incubation time of the E. coli strain versus to control occur, after 1 h (p = 0.000318), and time after 30 min. and after 1 h (p = 0.003874) (Supplementary Table S3). Taking into account the concentration of the analyte on the variability of promoter induction, statistically significant differences occur between concentration 0.625 µg/mL to concentrations 2,50 µg/mL (p = 0.033544) and 5 µg/mL (p = 0.0195518) and between concentration 5 µg/mL and 10 µg/mL (p = 0.041470) (Supplementary Table S3).

Fig. 2
figure 2

Signal induction [%] pattern (standardised for control—100%) of the soxS promoter after E. coli SM335 soxS:luxCDABE incubation with ampicillin (AM). Standard error bars indicate the standard error from the mean of four independent replicates.

Full size image

The highest increase in soxS promoter induction in E. coli SM335 soxS:luxCDABE culture values were achieved at ampicillin concentrations lowest of 0.625 µg/mL and 1.25 µg/mL after 1 h versus to control (Fig. 2). This indicates the stimulation reaction of the soxS promoter in E. coli SM335 soxS:luxCDABE culture and increase luminescence under the influence of this analyte (Supplementary Fig. S3). The antibacterial effect of ampicillin is based on blocking one of the last stages of bacterial cell wall biosynthesis, the so-called transpeptidation, i.e. the process of cross-linking the original linear structure of the bacterial cell wall. Ampicillin binds to penicillin-binding proteins (PBPs) and inactivates them on the inner bacterial cell membrane. Inactivation of PBP disrupts the cross-linking of peptidoglycan chains necessary for the strength and stiffness of the bacterial cell wall. The ampicillin inhibits bacterial transpeptidase, which hinders the synthesis of the bacterial cell wall. This action activates autolytic enzymes in the bacterial cell wall, causing cell wall lysis and destruction of bacterial cells42. Ampicillin is a semi-synthetic penicillin antibiotic a β-lactam from the aminopenicillin group used to treat bacterial infections such as pneumonia and bronchitis, urinary tract inflammation, salmonellosis, and inflammation of the inner layer of the heart wall (endocardium). Moreover, ampicillin can bind tissue proteins in the body to form a complete antigen, which causes allergic reactions in people43. The popularity of this antibiotic has contributed to the emergence of bacterial strains that are fully resistant to its effects. The analysis of municipal wastewater showed that the number of drug-resistance genes reaches 9.50 × 105 gene copies/mL of municipal wastewater, which exceeds the number of gene copies in pharmaceutical wastewater44,45. Antibiotic degradation products are regularly detected in sewage, surface water, groundwater, fishponds, hospital effluent, and even seawater. Antibiotic residues are formed by biodegradation and chemical degradation, as well as during wastewater treatment processes46.

The lowest increase in soxS promoter induction in E. coli SM335 soxS:luxCDABE culture values were achieved at ampicillin concentrations of 2.5 µg/mL and 5 µg/mL, at each time tested versus to control (Fig. 2, Supplementary Fig. S4b). This inhibition of the soxS promoter leads to inhibition of lux gene expression and a decrease in luminescence signal. The soxS regulator is common in bacteria and it is an activator of the transcription of over 100 diverse genes involved in resistance to antibiotics. SoxS activates transcription through a “pre-recruitment” mechanism, where newly synthesized soxS molecules form binary complexes with RNA polymerase in solution, then scan soxS-dependent promoters and further activate transcription47. The soxS regulatory protein belongs to the group of AraC/XylS transcriptional regulators found in Enterobacteriaceae. Bacteria with the above-mentioned transcription regulators have three common regulatory functions: carbon metabolism, pathogenesis, and stress response48. These regulators play a particular role in antibiotic resistance, the pathogenicity of bacterial species, biofilm formation, and motility49,50.

The response of the biosensor to incubation in various concentrations of ampicillin is characterized by an increased reaction in promoter soxS during the last measurement (after 1 h), relative to the control, except for the concentrations of 2.5 µg/mL and of 5 µg/mL (Fig. 2). It was found that ampicillin concentrations of 10 µg/ml, 1.25 µg/mL and 0.625 µg/mL caused the highest increase in induction soxS promoter after 1 h, by more than 9.9%, 61.5% and 67.5%, respectively, compared to the control (0 µg/ml). The highest increase in incubation soxS promoter after 1 h, may indicate that ampicillin shows the highest activity after a longer incubation time (Fig. 2). It has been shown that during the 3-month freezing of meat from pigs treated with ampicillin at – 20 °C and − 75 °C, the drug concentration in the meat does not decrease. Only after 8 months of storing meat at − 20 °C and − 75 °C did the concentration of ampicillin decrease51. Ampicillin in a changed form can remain in the animal’s body52, enter the muscles, or be partially excreted with milk by the mammary glands53. Because human and animal bodies are unable to completely metabolize ampicillin, using it results in residues of the substance in food of animal origin or the aquatic environment52.

The expression of micF:luxCDABE in E. coli SM343 lux biosensor after antibiotics—kanamycin

Incubation of the E. coli SM343 micF:luxCDABE strain in various concentrations of kanamycin causes the biosensor to respond (Fig. 3). The micF gene from E. coli bacteria can control the expression of other genes and specifically inhibits the expression of the target ompF gene54. MicF RNA, a small regulatory RNA found in bacteria, posttranscriptionally regulates the expression of outer membrane protein F (OmpF) by interacting with the ompF mRNA 5’UTR55. Studies show that the micF gene is present in Gram-negative bacterial species such as Salmonella typhimurium, Klebsiella pneumoniae, and Pseudomonas aeruginosa. Salmonella typhi, Salmonella typhimurium, and Klebsiella pneumoniae DNA sequences show over 96% homology in the micF gene compared to the E. coli micF sequence56.

Fig. 3
figure 3

Signal induction [%] pattern (standardised for control—100%) of the micF promoter after E. coli SM343 micF:luxCDABE incubation with kanamycin (KA). Standard error bars indicate the standard error from the mean of four independent replicates.

Full size image

The expression of micF:luxCDABE in E. coli SM343 lux biosensor after kanamycin was during each of three measurements time "0", after 30 min. and 1 h. Statistically significant differences occur between the incubation time of the E. coli strain versus the control (“time 0”), respectively, to time after 30 min. (p = 0.000423) and after 1 h (0.005685), and time after 30 min. and after 1 h (p = 0.000060) (Supplementary Table S4). Kanamycin is an aminoglycoside antibiotic, whose main mechanism of action is related to the disruption of bacterial protein synthesis during the transfer of amino acids from aminoacyl-tRNA to ribosomes. Kanamycin promotes the retention of amino acids on the ribosome of aminoacyl-tRNAs that do not correspond to the codons formed in the A site of the ribosome. This coding synthesizes incorrect polypeptides with many errors, making kanamycin cytotoxic to bacterial cells.

In the first contact of the biosensor with the analyte in time "0", for each concentration, the induction of promoter micF is higher than the other two measurements (after 30 min and after 1 h) (Supplementary Fig. S5a, Supplementary Fig. S6a). After 30 min for each analyte concentration, an inhibition reaction of promoter micF and a decrease in luminescence was detected (Fig. 3). The strongest luminescence signal inhibition was demonstrated for kanamycin concentrations of 0.625–5 µg/mL after 30 min. 1-h E. coli SM343 micF:luxCDABE culture incubation with kanamycin at a concentration of 0.625 µg/mL and 1.25 µg/mL, triggered the micF promoter stimulation and lux gene expression growth and luminescence emission up to 73.9% and 68.4%, respectively.

Considering the concentration of kanamycin, differences are not statistically significant (Supplementary Fig. S5b, Supplementary Fig. S6b). Kanamycin is resistant to alkaline and acidic environments and is also thermally stable. To hydrolyze kanamycin needed a temperature not less than 100 °C, but according to studies on kanamycin removal, a removal efficiency of 96.4% was achieved at pH = 9 and a temperature of 160 °C. Therefore, it is very important to test industrial and drinking water for the presence of antibiotics and to monitor water sources in real time for kanamycin content57,58.

In the analysis of the micF promoter induction in comparison to the control (0 µg/ml), the strongest luminescence signal inhibition was demonstrated for kanamycin concentrations of 0.625 µg/mL to 5 µg/mL after 30 min. For micF:luxCDABE in E. coli SM343 lux biosensor, the highest increase in gene expression and luminescence emission up to 73.9% was found after 1 h at a concentration of 0,625 µg/mL. Kanamycin is produced by microbial fermentation59 and is commonly used in veterinary medicine to treat cattle. Kanamycin residues are detected in food of animal origin, honey, milk and meat. In addition to food, kanamycin residues also end up in the aquatic environment. Consuming food containing kanamycin residues leads to antibiotic resistance, kidney disease and hearing damage60,61.

The expression rpoB:luxCDABE in E. coli SM346 lux biosensor after diclofenac

Incubation of the E. coli SM346 rpoB:luxCDABE strain in various concentrations of diclofenac results in gene expression at each of the tested concentrations (Fig. 4). These differences were found to be statistically significant (Supplementary Table S5). The rpoB gene encodes the β subunit of DNA-dependent RNA polymerase62, and mutations within its sequence have been found to cause resistance to rifampicin, which belongs to the group of ansamycin antibiotics. Rifampicin is an antibiotic that binds to bacterial RNA polymerase, interfering with DNA transcription. Rifampicin resistance mutations in rpoB have been shown to be responsible for severe effects on transcription, cell fitness, bacterial stress response, and virulence63. Rifampicin resistance is common in several microorganisms and is usually caused by point mutations in the gene encoding the β subunit of RNA polymerase. Furthermore, rpoB DNA contains a conserved region in the bacterial cell that can be used for species classification. Using it, it is possible to identify intestinal bacteria, Mycobacterium, various species of spirochetes and Legionella species64.

Fig. 4
figure 4

Signal induction [%] pattern (standardised for control—100%) of the rpoB promoter after E. coli SM346 rpoB:luxCDABE incubation with diclofenac (DI). Standard error bars indicate the standard error from the mean of four independent replicates.

Full size image

The response of the biosensor on diclofenac is characterized by the positive induction reaction of promoter rpoB after 30 min. for each concentration, except 10 µg/mL, compared to the other two measurements (in “time 0” and after 1 h) (Supplementary Fig. S7a, Supplementary Fig. S8a). The induction of the rpoB promoter over time is statistically significant (Supplementary Table S5). The highest values of rpoB promoter induction were recorded for diclofenac concentration of 1.25 µg/mL after 30 min. and after 1 h, up to 27.4% and 68.6%, respectively (Fig. 4, Supplementary Fig. S7b, Supplementary Fig. S8b). After 30 min. high values of rpoB promoter induction were in the range from 19.1% at the concentration of 0.625 µg/mL to 27.4% at the concentration of 1.25 µg/mL (Fig. 4). After 1 h highest values of rpoB promoter induction were in the range from 60.8% at the concentration of 0.625 µg/mL to 68.6% at the concentration of 1.25 µg/mL (Fig. 4). Diclofenac, or 2-(2,6-dichlorineanilin)phenylacetic acid, is a non-steroidal anti-inflammatory and analgesic drug. The mechanism of its action is based on the inhibition of the enzyme cyclooxygenase or prostaglandin synthase, which catalyzes the conversion of arachidonic acid to prostaglandin by reducing the synthesis of prostanoids. Both enzymes (cyclooxygenases) have different activities in the inflammatory process and regulation of local homeostasis, depending on the effect of prostanoids on the kidneys, platelets and the vascular system65.

The lowest values induction promoter rpoB were recorded for the concentration of 10 µg/mL for measurement after 30 min. and after 1 h (Fig. 4). It was found that the strongest gene expression occurs during incubation of the bacterial strain at a concentration of 1.25 µg/mL diclofenac. Diclofenac is excreted from the body with urine, in its original form or as metabolites, and then gets into sewage, where it is partially removed. Another reason for the presence of diclofenac in sewage is its improper disposal as solid waste. Unfortunately, the removal of diclofenac residues from wastewater is still not effective enough, so it is often carried out in water resources66,67. The maximum detection limit for diclofenac in municipal sewage is approximately 7 μg/l. In a long-term study of effluent from three sewage treatment plants lasting 10 months, diclofenac was detected at average concentrations of 1–10 μg/l, while in an EU-wide study monitoring effluent from sewage treatment plants, a maximum concentration of 174 ng/l was found, with mean values of 50 ng/l23,68.

Analysis of the biosensor response relative to the control (0 µg/mL) shows that the highest values of rpoB promoter induction (up to 60–69%) occurs when the bacterial strain is incubated at a concentration of 1.25 µg/mL diclofenac after 30 min and after 1 h (Fig. 4). The strongest luminescence signal inhibition was demonstrated for diclofenac concentrations of 2.5–5 µg/mL after 1 h, up to 34.7% and 34.1%, respectively. These were the lowest induction values promoter rpoB and the decrease in luminescence compared to the control (Fig. 4).

Comparing the results obtained in our work with the results of antibiotic studies using microbial E. coli biosensors with plasmid gene constructs containing recA, micF, soxS and rpoB promoters in their structure4,8, we found pronounced differences in promoter induction values and intensity of the luminescent cellular response (this is also explained in the Materials and Methods section of the paper). These discrepancies are due to the different growth phase of the E. coli cultures used in the study. In the experiments described in the present study, the E. coli cultures were in the stationary growth phase and in the work of Urban et al.4 and Melamed et al.8 in a logarithmic growth phase.

It is known that antibiotics target cellular biosynthetic processes. In the stationary growth phase, bacterial cells are no longer growing and dividing due to the lack of oxygen and/or nutrients, so those processes that are inhibited by antibiotics in the exponential growth phase no longer occur, and therefore, the stress level is low. Furthermore, luminescence requires energy, and it is doubtful that there is still glucose in the medium as an energy source.

The mechanism of antibiotics and diclofenac action in E. coli

Various species of bacteria can be used to construct a bacterial biosensor, but the most common are biosensors based on E. coli cells because they are the best-known gram-negative bacteria. E. coli bacteria with lux genes are used as semi-specific or induced sensors in response to damage to cell components. Toxic substances damage three types of biomolecules: DNA, proteins, and lipids. In E. coli cells, there are several regulons in which genes are simultaneously transcribed after the appearance of toxic factors. The SOS regulon (reacting to DNA damage), the “heat shock” regulon (reacting to protein damage), the regulon whose genes are expressed in the event of cell membrane damage, and two oxidative stress regulons: soxRS and oxyRS28.

The studies showed that the appearance of an analyte (stressor) in various concentrations triggers a cascade of reactions leading to the expression of a reporter gene and the emission of a cellular signal. The possible mechanism of antibiotics and diclofenac action in E. coli was presented in Fig. 5. The organised heat shock proteins network in oxidative stress response and recA, SoxS, micF and rpoB genes activity (Drugs-Reactive Oxygene Species (ROS)- Heat Shock Proteins (HSPs), notation in brief Drugs-ROS-HSPs) was described in five important stages (Fig. 5).

Fig. 5
figure 5

The possible mechanism of antibiotics and diclofenac action in E. coli. The role of organised heat shock proteins network in oxidative stress response and recA, SoxS, micF and rpoB genes activity (Drugs-ROS-HSPs). (1) Drugs stress signal in E. coli; (2) Increase in ROS synthesis and oxidative stress activation, (3) Activation of oxidative stress-related cellular signal transduction pathways, including activation of the heat shock factor-1 (HSF-1), (4) Stimulation of expression of heat shock proteins genes and HSPs synthesis. HSPs have a cytoprotective function by reducing ROS levels and protecting nucleic acids and cellular proteins; HSPs may have a cytoprotective and supportive role in the activity of gene expression products recA, SoxS, micF and rpoB and the repair of damage after E. coli cell exposure to test drugs.; (5) Oxidative stress affects miRNA activity, which, can regulate the expression of many genes, including recA, SoxS, micF and rpoB and other genes whose expression products are involved in the antioxidant defence of the cell69,70,71. Abbreviations: CAT-catalase, GSH—Glutathione, GSHT—Glutathione transferase, miRNA—microRNA.

Full size image

Other genes with different activity and protein expression are also used in bacterial cells. The lacZ gene is also used in bacterial biosensors, but it is not popular due to low sensitivity, low efficiency, and length of the enzymatic reaction28. The lacZ gene encoding the enzyme β-galactosidase was isolated from the genome of E. coli bacteria. This enzyme hydrolyzes lactose into glucose and galactose. Enzyme activity is measured using photocolorimetry. When β-galactosidase is produced, X-gal is hydrolyzed to form 5-bromine-4-chlorineindoxyl, which spontaneously dimerizes to form an insoluble blue pigment called 5,5′-dibromine-4,4′-dichloroine-indigo.

The ctr genes synthesise carotenoids in bacterial cells29 and their activity is also measured using photocolorimetry. Bacteria that produce a visible red pigment (deinoxanthin) are used as biosensors, where the ctr gene has been removed from the bacterial chromosome and then inserted into a plasmid under the control of an inducible promoter. The disadvantage of biosensors with ctr genes is the reaction time and the complexity of the process because the dye is produced as a result of the reaction involving many proteins.

The next gene used in bacterial biosensors is a green fluorescent protein (gfp) that comes from the jellyfish species Aequorea victoria. GFP proteins are an invaluable tool for visualizing cellular structures and explaining the biochemical transport of single cells into entire organisms. The main gfp motif is the p-hydroxybenzylidene-2,3-dimethylimidazolinone chromophore30. For the reason, no substrate is needed to activate fluorescence, gfp may have an advantage over other genes. However, for fluorescence to occur, the protein must develop sufficiently. Expression of fluorescent proteins involves transcription and translation. After translation, a fluorescent protein chromophore is formed. The time required to form the chromophore of a fluorescent protein is referred to as the maturation time. Genetic engineering allowed the optimization of gfp features, such as higher quantum yield and increased photostability31.

Bacterial whole-cell biosensors are a valuable tool for detecting and assessing the toxicity of pollutants in the natural environment, and for monitoring the level of pharmaceuticals in surface waters, soil, air, sewage, and food14,67. They are an irreplaceable tool in research on gene expression, promoter activity, and regulation of bacterial cell response under stress conditions. Microbial biosensors with gfp and lux reporter genes are used in monitoring metals in contaminated soils, polychlorinated biphenyls (PCBs) in sewage, organic pollutants in treated sewage, chromium, arsenic, lead, pesticides and petroleum hydrocarbons in water7,11. E. coli strains with the recA:luxCDABE plasmid gene construct are used to determine the genotoxicity of many chemical compounds, including antibiotics8, phenol, chromium (Cr6+), lead (Pb2+) and the potential of UV radiation to affect bacterial cells. In environmental monitoring, E. coli recA:luxCDABE strains are used to determine the genotoxicity of crude oil in surface water samples and the genotoxicity of municipal sewage11,13. In the work of Zappi et al.19 two microbial E. coli biosensors containing the gfp and lux genes were used to determine the toxicity of nitrification inhibitors (allylthiourea, phenol and copper) in industrial and municipal sewage. Microbial biosensors with lux and gfp reporter genes are used to determine the cytotoxicity of drugs and other chemical compounds7. According to the International Agency for Research on Cancer72 it is recommended that carcinogenic, mutagenic or teratogenic substances should be eliminated from treated wastewater. The organic compounds with toxic effects and endocrinal active compounds are important for human health73, and some drugs detected in wastewater, such as anticancer drugs74 and hormones (17β-estradiol)75 have these properties. The risks associated with the presence of pharmaceuticals in the environment cause their transport into the environment should be limited and constant monitoring of environmental matrices should be carried out76. Currently, bio-membrane-based methods and AOPs (Advanced Oxidation Processes) processes are becoming increasingly important in this area.

Conclusion

In the present study, the effect of selected antibiotics nalidixic acid, ampicillin, kanamycin, and antiinflammatory drug—diclofenac on E. coli SM biosensor strains, which contain pBRlux-trp:recA::luxCDABE plasmid constructs with recA, SoxS, micF and rpoB promoters and luxCDABE reporter gene was examined. The genotoxicity of the test nalidixic acid was measured based on the induction of genotoxin and DNA damage-inducible recA promoter in E. coli SM342 recA:luxCDABE biosensor strain. Using E. coli SM335 soxS:luxCDABE and E. coli SM343 micF:luxCDABE, the effects of ampicillin and kanamycin on the promoters of soxS and micF genes involved in the regulation of expression of many genes, including genes under SoxS regulation and genes involved in the removal of antibiotics, were investigated. Using the E. coli strain SM346 rpoB:luxCDABE the effect of the diclofenac tested on the RNA polymerase gene promoter (beta subunit) was determined. In these experiments, E. coli luminescent bioreporter in the stationary phase was used and the luminescence response was very low compared to the data in the cited scientific publications.

The results obtained in the present study are the basis for the following conclusions:

  1. 1.

    All of the drugs tested interacted with different intensities in E. coli SM strains on cellular pathways involving recA, SoxS, micF and rpoB gene promoters, indicating different sensitivity of the gene constructs tested to the drugs. Variation in the induction value of test promoters recA, SoxS, micF and rpoB in E. coli SM strains after incubation with drugs depended on the concentration of the analyte and the incubation time of the strain.

  2. 2.

    The highest increase in recA promoter induction in E. coli SM342 recA:luxCDABE culture was recorded at the concentrations of 10 µg/mL of nalidixic acid, after 30 min (3.5%) and 1 h (5.4%). The reaction was stronger with increasing concentration, where 10 µg/mL was the threshold of the positive reaction.

  3. 3.

    Among the drugs tested, one-hour incubation of E. coli soxS:luxCDABE with ampicillin at concentrations of 10 µg/ml, 1.25 µg/ml and 0.625 µg/ml induced the soxS promoter most intensively, by more than 9.9%, 61.5% and 67.5%, respectively, compared to the control (0 µg/ml).

  4. 4.

    In the analysis of the micF promoter induction in E. coli micF:luxCDABE in comparison with control (0 µg/ml), the highest biosensor response up to more than 68.4% and 73.9% was shown in the case of kanamycin concentrations of 1.25 and 0.625 µg/ml after 1 h.

  5. 5.

    Diclofenac tested at a concentration of 1.25 µg/ml after 1 h induced the rpoB promoter in E. coli rpoB:luxCDABE more intensively by more than 69% compared to the control (0 µg/ml).

  6. 6.

    The results obtained indicate that the E. coli biosensor strains with plasmid vectors containing recA, SoxS, micF and rpoB promoters and luxCDABE reporter gene used in this study can be used in gene expression analysis and as biosensors in the detection, genotoxicity testing and investigation of the toxicity effects of pharmacological contaminants present in the environment.

Materials and methods

Material

Research about the expression level of plasmid constructs with recA, soxS, micF and rpoB promoters in transcriptional fusion with luxCDABE reporter gene in E. coli SM strains was conducted using luminometric indication. E. coli SM strains used in this study are presented in Table 1. A genetic map of the pBRlux-trp plasmid with the luxCDABE reporter gene used in this study is presented in Fig. 6. E. coli SM strains containing the luxCDABE plasmids with recA, soxS, micF and rpoB promoters are from the collection of Prof. Shimshon Belkin’s laboratory, Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel.

Table 1 Characterization of the E. coli strains used in the experiments described in this work8.
Full size table
Fig. 6
figure 6

Genetic map of the pBRlux-trp plasmid with the luxCDABE reporter gene and with the relevant restriction sites8.

Full size image

In selecting the promoters for the study, we were guided by their sensitivity to antibiotic stress-induced changes in the cell, as documented earlier in scientific studies8,77,78. These include DNA damage (recA promoter), the participation of dual transcriptional activator in the removal of antibiotics (soxS promoter), a small regulatory RNA that regulates of the translation of the OmpF porin (micF promoter), the regulation of RNA polymerase, beta subunit (rpoB promoter)8,77. It is known that some of the promoters e.g. recA are active at a certain level in the cell all the time. Furthermore, in E. coli cells in stationary growth, promoters can be induced by metabolic products of the bacteria that accumulate in the culture medium. Another factor capable of inducing the promoters under study and inducing significant luminescence in bacterial cells is reactive oxygen species (ROS), whose presence is always found at some level in the cells of aerobic organisms79,80.

The analyses were performed using the GloMax*-Multi Detection System plate reader from Promega. The experiments were carried out in the culture medium and a specific dose of active substances (analytes). Among the antibiotics used were nalidixic acid, ampicillin, kanamycin, and the inflammatory drug diclofenac (Table 2). NA—a known genotoxin in this paper was used as a standard. The mechanism of action of NA is through inhibition of DNA replication in bacterial cells. Studies using E. coli strains or other bacteria often use nalidixic acid as a standard, especially in experiments with E. coli strains containing plasmids with a recA promoter and a specific reporter gene, e.g. lux or gfp16. β-lactam antibiotics (and ampicillin is one of these) are also among the most frequently detected contaminants of pharmaceutical origin in wastewater, water and dairy products78.

Table 2 Chemical structure of the drugs used in the study.
Full size table

All the tested drugs were dissolved in dimethyl sulfoxide (DMSO, Chempur). Each substance was prepared with a series of dilutions in DMSO with a dilution factor of q = 2. In all experiments, the drugs were tested at final concentrations of 10 µg/mL, 5 µg/mL, 2.5 µg/mL, 1.25 µg/mL, 0.625 µg/mL and 0—only DMSO control sample. Four separate series of studies were performed using four different gene constructs in E. coli SM cells. For each series of studies, the gene construct contained a different promoter (recA, soxS, micF and rpoB), corresponding to the tested substance.

In the earlier works76,77,78 the E. coli strains used in that work were already evaluated. However, the bacteria were in a different growth phase, a logarithmic one, whereas in our work we investigated how the sensitivity of luminescent E. coli biosensor strains changes in the stationary phase. This is because it is known that E. coli sensitivity to antibiotics is variable depending on the growth phase. The cultures tested in our work containing plasmids with the luxCDABE reporter gene were in the stationary growth phase. Furthermore, in our experiments, we used different concentrations of the test drugs and different incubation times to determine the intensity of the luminescence response. These methodological differences affected the results obtained4,8,77.

Effects of test drugs on induction of recA, soxS, micF and rpoB promoters in the pBRlux-trp-luxCDABE plasmid in E. coli SM strains

The effect of test drugs was determined based on the induction parameters of the recA, soxS, micF and rpoB promoters in E. coli SM strains with the luxCDABE reporter gene in the plasmid vector pBRlux-trp. The determination of the effect is based on the quantitative determination of the intensity of luminescence of E. coli SM recA:luxCDABE, micF:luxCDABE, soxS:luxCDABE and rpoB:luxCDABE cultures incubated for a specified time with the tested active substances of the drugs compared to the control culture. The luminescent signal is proportional to the level of induction of a given promoter and the response potency of the active drug substances tested on bacterial cells. E.coli SM/pBRlux-trp strains with recA, soxS, micF and rpoB promoters were used for 24 h in LB broth at 37 °C. Luminescence measurement and data analysis were performed according to the method of Melamed et al.8 and Kessler et al.76. Amount of 100 µl overnight broth culture of E. coli SM/pBRlux-trp strains with recA, soxS, micF and rpoB promoters was placed on a 96-well white plate (Grainer Bio-one, Germany) with a transparent bottom for simultaneous measurement of absorbance and luminescence (in wells A to F). The culture was repeated four times in horizontal orientation.

According to the scheme (Table 3), an equal volume of DMSO was added to 50 µl of culture of biosensor E. coli strains, so that the test volume was 100 µl. This volume of added DMSO had no toxic effect on the bacterial cells.

Table 3 Scheme for adding broth culture and diluting the compound on the plate.
Full size table

The prepared plate was placed in the measuring device. The measurement of luminescence and absorbance was performed at a wavelength of λ = 600 nm. The first measurement was performed after preparing the plate and was marked as “time 0”, the next measurement was performed 30 min. after preparing the plate, and the next one after 1 h after preparing the plate. After obtaining the luminescence and absorbance values, the luminescence (L) was calculated using the formula:

$$L=\frac{l}{Abs}$$

L—luminescence, l—luminescence value from the luminometer, Abs—absorbance value for λ = 600 nm.

The effects of tested drugs on E. coli SM cells with plasmid gene constructs containing recA, soxS, micF and rpoB promoters with luxCDABE reporter gene are presented as luminescence values (L) of test and control samples at time 0 and after 30 min and 1 h and as a percentage (%) increase/decrease in the luminescence of treated samples compared to the control (0 µg/mL).

Statistical analysis

Statistica ver. 13.3 (TIBCO Software Inc., 1984–2017) was used for the statistical evaluation of the expression of different gene constructs in Escherichia coli SM lux biosensor after drugs exposure. Statistically significant differences in the induction values ​​of different promoters under the influence of active substances of tested drugs were assessed, taking into account different concentrations of analytes and time of promoter induction. A parametric T-test for dependent samples was performed for repeated measurements, with the condition of normal distribution using the Shapiro–Wilk W test and the lack of significant differences between variances. The T-test was used to assess differences between means in individual repetitions, as dependent samples “before” and “after” drug administration. Analysis of variance (ANOVA) compared descriptive statistics and correlations for dependent variables in each of the groups defined by one grouping variable, i.e. time or analyte concentration. Descriptive statistics, tests and statistical analyses are presented in the supplementary material. Values ​​of p < 0.05 were regarded as significant and p < 0.01 as highly significant.