The effects of antidepressants fluoxetine, sertraline, and amitriptyline on the development of antibiotic resistance in Acinetobacter baumannii

This study investigates the effects of antidepressants fluoxetine, sertraline, and amitriptyline on the development of antibiotic resistance in clinical Acinetobacter baumannii isolates. The isolates were exposed to fluoxetine, sertraline, and amitriptyline for 30 days, respectively. The bacteria that developed resistance to gentamicin, imipenem, colistin, and ciprofloxacin were isolated and expression levels of some antibiotic-resistance genes were determined by quantitative reverse-transcriptase PCR. Before and after the exposure, minimum inhibitory concentration (MIC) values of the bacteria were determined by the microdilution method. The statistical analysis was performed using Student’s t test. A time-dependent increase was observed in the number of bacteria that developed resistance and increased the MIC value. After exposure to fluoxetine and sertraline, decreases were observed for efflux and outer membrane porin genes in isolates that developed colistin resistance, and increases were observed in isolates that developed ciprofloxacin resistance. These observations suggest that these antidepressants have similar effects on the development of resistance. While the exposure to fluoxetine did not result in the development of resistance to imipenem, it was observed after exposure to sertraline and amitriptyline, and a common decrease in ompA gene expression was determined in these isolates. To our knowledge, the comparative effects of selected antidepressants on the development of antibiotic resistance in A. baumannii are reported and presented in the literature here for the first time.

limited side effects (Do et al. 2017). A significant portion of the fluoxetine is excreted in the urine which leads to its accumulation in wastewater, rivers, and groundwater. Hence, attention has been focused on the toxicity and ecotoxicity of fluoxetine (Altamura et al. 1994;Brooks et al. 2003;Jin et al. 2018). The other commonly used SSRI in the treatment of depressive disorders is sertraline, of which less than 0.2% is excreted in urine (Wang et al. 2001). Although the negative effects of sertraline have been reported on algae, its effects on microbial communities have yet to be elucidated (Johnson et al. 2007;Yang et al. 2019). Amitriptyline classified in tricyclic antidepressants. Approximately, 1% of the drug is excreted unchanged in the urine (Şahin et al. 2008;Mastrianni et al. 2016).
Other than their essential therapeutic effects, the antimicrobial activities of certain antidepressants have also been reported (Munoz-Bellido et al. 2000;Ayaz et al. 2015;Laudy et al. 2017;de Sousa et al. 2018), and studies conducted in the past several years have shown that fluoxetine and sertraline have antibiotic-modulating activities as well as antimicrobial activities (Ayaz et al. 2015;de Sousa et al. 2018).
It should be noted that the long-term usage of non-antibiotic drugs may affect both the microflora members and environmental microorganisms due to their improper disposal and accumulation into wastewater, respectively. Antidepressants are among these drugs (Jin et al. 2018). The aim of this study is to investigate the effects of antidepressants fluoxetine, sertraline, and amitriptyline on the development of antibiotic resistance in A. baumannii. This study is one of the primary investigations that demonstrates the role of non-antibiotic drugs in antibiotic-resistance development.

Materials and methods
Bacterial strains and growth conditions Ten (gentamicin, imipenem, colistin, ciprofloxacin) susceptible and one resistant clinical A. baumannii isolates were used. A. baumannii ATCC 1709 (susceptible strain) and A. baumannii ATCC 1799 (resistant strain) were used as standard strains. Antibiotic susceptibilities of all test bacteria were confirmed using the disc diffusion and minimum inhibitory concentration (MIC) tests (European Committee on Antimicrobial Susceptibility Testing 2018). In susceptibility tests, Escherichia coli ATCC 25922 strain was used as the control.

Detection of ade efflux pump and outer membrane porin genes
The presence of ade efflux pump genes adeA, adeB, adeC, adeR, adeS, and outer membrane porin genes ompA, ompW in test bacteria were demonstrated by Polymerase Chain Reaction (PCR) method (Lopes 2012;McConnell et al. 2012;Jassim et al. 2016;Catel-Ferreira et al. 2016;Sepahvand et al. 2017). Gel electrophoresis was used for the detection of the amplification products visualized with SafeView Classic (ABM, Canada). The size of PCR products was compared with a 100 bp DNA Ladder (New England BioLabs, ABD).

Determination of minimum inhibitory concentration values
Before and after exposure to fluoxetine, sertraline, and amitriptyline, MIC values of test bacteria against gentamicin, imipenem, colistin, and ciprofloxacin were determined by the microdilution method (European Committee on Antimicrobial Susceptibility Testing 2018). Increase in the MIC values after exposure was detected. Each test bacterium was tested in triplicate.

Fluoxetine, sertraline, and amitriptyline exposure and isolation of resistant bacteria against test antibiotics
Initially, exposure concentrations of fluoxetine, sertraline, and amitriptyline were determined by MIC test (Clinical and Laboratory Standards Institute 2009). 40 μl suspension of test bacteria (adjusted to 10 8 -10 9 cfu/ml) was added into 3.96 ml LB containing the various concentrations of antidepressants, fluoxetine hydrochloride (Sigma) (2.5-5-10 μg ml −1 ), sertraline hydrochloride (Sigma) (1.25-2.5-5 μg ml −1 ) and amitriptyline hydrochloride (Sigma) (2.5-5-10 μg ml −1 ). After incubation at 37 °C for 18-24 h, 40 μl of the test bacteria cultures was transferred into 3.96 ml fresh LB containing the above-stated concentrations of fluoxetine, sertraline, and amitriptyline and incubated again. This subculturing procedure was repeated for 30 days incessantly. On the 5th, 10th, and 30th days of this procedure, 100 μl of exposed test bacteria samples was plated on the MHA plates with 4 μg ml −1 gentamicin, 8 μg ml −1 imipenem, 2 μg ml −1 colistin, and 1 μg ml −1 ciprofloxacin. After 24 h of incubation at 37 °C, the colonies formed on the MHA containing test antibiotics were considered as resistant. Test bacteria that were subcultured in LB broth without antidepressants were used as the control group (Jin et al. 2018).

Quantitative reverse-transcriptase PCR analysis of target genes
Expression levels of target genes of test bacteria were determined by quantitative reverse-transcriptase PCR (RT-qPCR). Pre-exposure and post-exposure expression levels of the ade efflux pump genes and outer membrane porin genes were compared (Eleaume and Jabbouri 2004;Kart et al. 2020). The total RNA was extracted using an RNA Isolation Kit (RNeasy Mini Kit, Hilden, Qiagen) according to the manufacturer's instructions. The total RNA was quantified in each sample using a NanoDrop spectrophotometer (Qiagen). cDNA was synthesized by NucleoGene RNA to cDNA Mix Kit (NucleoGene, Turkey) according to the manufacturer's instructions. A NucleoGene qPCRSYBR Green Master Mix Kit (2×) (NucleoGene, Turkey) was used for the determination of expression levels of the target genes. The RT-qPCR analysis was performed for each 20 µl PCR reaction mixture (10 µl NucleoGene qPCR Sybr Green Master Mix, 6.5 µl ddH 2 O, 2.5 µl cDNA template, 0.5 µl Forward Primer and 0.5 µl Reverse Primer). Relative gene expression levels were calculated with the 2 −ΔΔCT method using an RT2 Profiler PCR Array Data Analysis v3.5 (Qiagen, Hilden) analysis program. All data were normalized for 16S rRNA housekeeping gene levels.

Statistical analysis
Statistical data analysis was performed using the SPSS programme (Version 23, SPSS, Chicago, IL, USA). Student's t test was used for comparing the pre-and post-exposure expression levels of the target genes. P values < 0.05 were considered significant.

Detection of ade efflux pump and outer membrane porin genes
Agarose gel electrophoresis images of ade efflux pump and outer membrane porin genes are given in Fig. 1.

MIC values of fluoxetine, sertraline, and amitriptyline
MIC values of fluoxetine, sertraline, and amitriptyline are given in Supplementary Table. According to these values, exposure concentrations were determined as 2.5-5-10 μg ml −1 for fluoxetine hydrochloride and amitriptyline hydrochloride, 1.25-2.5-5 μg ml −1 for sertraline hydrochloride.

Antibiotic resistance development after exposure
The number of isolates that developed phenotypic resistance to gentamicin, imipenem, colistin, and ciprofloxacin on the 5th, 10th, and 30th days of the exposure are given in Fig. 2. Resistant isolates are indicated in Table 1.
Considering the results of phenotypic resistance development after exposure, it was observed that exposure to fluoxetine did not cause imipenem resistance. However, at the end of the fluoxetine exposure of the 30th day, an increase was observed in the number of isolates that developed resistance to ciprofloxacin and gentamicin. It was observed that exposure to the highest concentrations of sertraline and amitriptyline caused imipenem resistance in one isolate in each group at the end of the 5th day. In addition, at the end of the 30th day, imipenem resistance was observed in one of the isolates in the control group. Amitriptyline exposure caused the development of gentamicin resistance in all test bacteria, at the end of the 30th day. In addition, it was also determined that three isolates (7, 8, and 9) in the control group developed resistance to gentamicin. The resistance to ciprofloxacin and gentamicin largely developed after exposure to amitriptyline. Since colistin resistance was seen in all isolates in the control group, it could not be correlated with the resistance developed after antidepressant exposure. Jin et al. (2018) were the first to report that exposure to fluoxetine induces antibiotic resistance in E. coli. In this study, as a result of 10 h exposure of E. coli to fluoxetine, the expression levels of multidrug efflux pump genes acrB and acrD increased, while the expression levels of the outer membrane porin genes ompF and ompW decreased. The bacterium protected itself from exposure to fluoxetine by expelling the drug from the cell and preventing the entry of the drug into the cell. It was determined that exposure to fluoxetine for 30 days caused an increase in chloramphenicol, amoxicillin, and tetracycline resistance. In addition, these mutants exhibited multiple resistances against fluoroquinolones, aminoglycosides, and beta-lactams.
Our study is important in terms of investigating the effects of sertraline and fluoxetine, as both antidepressants have the same mechanism of action and amitriptyline which acts with a different mechanism, on the development of antibiotic resistance in A. baumannii for the first time in the literature.
To observe the effect of antidepressants on resistance genes, gene expression values of some test bacteria that developed resistance to test antibiotics after exposure to fluoxetine, sertraline, and amitriptyline were compared relative to controls that were not exposed to any antidepressant (Tables 2 and 4).
Decreases in the expression of all genes (except ompA and ompW for the number 12 strain) against ciprofloxacin were observed after exposure to amitriptyline, regardless of time. When all isolates with gentamicin resistance were evaluated together, the expression increase was determined only for the adeC and ompW genes. Much like the result obtained after exposure to sertraline, isolates that developed resistance to imipenem also found decreases in adeR and  ompA expressions. For colistin-resistant isolates, a consistent result could not be obtained between the resistance genes tested and resistance development (Table 2).
An increase in expression of adeR, adeS, and ompA, and ompW was observed against ciprofloxacin after 30 days of exposure compared to 5 days after sertraline exposure. Regardless of the time, reductions in all genes against colistin were observed, while reductions in adeR and ompW genes were observed against imipenem. For isolates that developed resistance to gentamicin, a consistent result could not be obtained between tested resistance genes and resistance development (Table 3).
After exposure to fluoxetine, increased expression was observed in adeR, adeS, ompA, and ompW against ciprofloxacin, and adeC, adeR, ompA, and ompW against gentamicin, after 30-day exposure compared to 5-day exposure. Decreases in the adeB, adeR, adeS, and ompW genes against colistin were detected regardless of time (Table 4).  In our study, expression values of A. baumannii ATCC 1709 and A. baumannii ATCC 1799 that did not develop resistance after exposure for the tested genes were compared (Fig. 3). In A. baumannii ATCC 1709, decreased expressions of adeR, adeS, and ompA for ciprofloxacin, adeC for gentamicin, and adeB, adeC, adeS, and adeR for imipenem were observed.
In this study, two isolates were selected for each antibiotic group from the isolates that developed resistance after exposure to different antidepressant concentrations for the 5th, 10th, and 30th days. In these isolates, mRNA expression levels of efflux pump genes adeA, adeB, adeC, adeR, adeS, and outer membrane porin protein genes ompA and ompW were measured using the RT-qPCR method. Expression data of isolates with resistance development were compared with control groups of the same isolate that were not exposed to antidepressants, thus the effects of resistance development on resistance genes were examined.
Although the mechanism of colistin resistance in A. baumannii is not known exactly, it is thought to occur as a result of modifications in the outer membrane lipopolysaccharides, resulting in decreased affinity of the antibiotic to Table 4 Relative fold change of resistance gene expression levels of some test bacteria that develop resistance after exposure to fluoxetine 4, 5, 6, 7, 12: susceptible test bacteria that develop phenotypic resistance. 11: resistant isolate. All isolates were compared with their controls not exposed to fluoxetine CIP Ciprofloxacin, GEN Gentamicin, IPM Imipenem, CST Colistin * Statistically insignificant (P ≥ 0.05)

GEN-S GEN-F GEN-A GEN-F CST-F CST-S IPM-A RelaƟve Fold Change (log2)
adeA adeB adeC adeR adeS ompA ompW Fig. 3 Comparison of gene expression values of A. baumannii ATCC 1709 and A. baumannii ATCC 1799 exposed to antidepressants its target site and decreased expression of ompW-like outer membrane porins (Dal et al. 2012). In our results, a decrease was detected in all genes of colistin-resistant isolates after sertraline exposure, and statistically significant decreases were observed in adeB, adeR, and adeS gene expressions, especially ompW against fluoxetine.
The AdeABC active drug pump system is responsible for the multidrug resistance of A. baumannii, and their overexpression causes the development of resistance against many antibiotics (Wieczorek et al. 2008). adeS, one of the regulatory genes, activates the transcription of adeABC pump proteins by causing phosphorylation of adeR (Dal et al. 2012). The study of Lari et al. (2018) found a significant correlation between ciprofloxacin resistance and upregulated adeB. In our results, in line with the literature, increases in adeR, adeS, ompA, and ompW expressions were observed in isolates that developed resistance to ciprofloxacin after exposure to both fluoxetine and sertraline. The opposite results were obtained after exposure to amitriptyline, and reductions were observed in all genes (except for the 12th strain). This may be because amitriptyline has a different mechanism of action compared to the other two antidepressants. Amitriptyline and fluoxetine played a similar role in the development of gentamicin resistance, causing an increase in adeC and ompW resistance genes. It is known that omp porin proteins found in the cell wall of A. baumannii prevent the entry of some molecules into the cell while preventing the exit of others from the cell, giving the cell wall selectivity. For this reason, the loss of these proteins in the outer membranes plays an important role in the emergence of antimicrobial resistance (Dal et al. 2012). While the exposure of our isolates to fluoxetine did not result in the development of resistance to imipenem, resistance to imipenem was observed after exposure to sertraline and amitriptyline, and a common decrease in the ompA gene expression was determined in these isolates.

Increase of the minimum inhibitory concentration values of test bacteria
Before and after exposure, MIC values of the bacteria against test antibiotics were given in Table 5. LB broth without any antidepressant was used as a control.
Considering the pre-and post-exposure MIC values of all test bacteria, it was observed that MIC values of isolates that developed imipenem resistance as a result of exposure to sertraline and amitriptyline increased > 128-fold compared to pre-exposure. The MIC value of the isolate in the control group, which was observed to develop phenotypical resistance to imipenem, did not increase compared to the preexposure. It was observed that MIC values of some isolates that developed resistance to gentamicin because of exposure to amitriptyline increased at rates ranging from 2-> 32-fold compared to pre-exposure. Besides MIC value of isolate number 10, which developed resistance to gentamicin as a result of exposure to fluoxetine, increased twofold compared to the pre-exposure. MIC value of isolate 7, which developed resistance to gentamicin after sertraline exposure, increased > 64-fold compared to pre-exposure. MIC values of some isolates that developed resistance to ciprofloxacin because of exposure to amitriptyline increased > 128-fold compared to pre-exposure. After exposure to fluoxetine, it was observed that the MIC value of isolate 12, which developed resistance to ciprofloxacin, increased > 4-fold compared to the pre-exposure. It was observed that the MIC values of isolates 6 and 7, which developed resistance to ciprofloxacin because of sertraline exposure, increased by 2-and > 128-fold, respectively, compared to pre-exposure. Since colistin resistance was seen in all isolates in the control group, it could not be correlated with the resistance developed after antidepressant exposure. However, an increase in the MIC values of some isolates was observed. When all MIC values of test bacteria are considered, we can say that exposure to amitriptyline causes the highest increase in MIC values, and exposure to fluoxetine causes the lowest increase.
With the extension of exposure time to fluoxetine, sertraline, and amitriptyline, the number of test bacteria that developed antibiotic resistance and the MIC values of some bacteria were increased. After exposure to fluoxetine and sertraline, decreases were observed for efflux and outer membrane porin proteins in isolates that developed colistin resistance, and increases were observed in isolates that developed ciprofloxacin resistance, suggesting that these antidepressants have similar effects in the development of antibiotic resistance. However, the study of two isolates for each antibiotic group causes limitations in the general interpretation of the results. In future studies, it will be possible to reach a clearer conclusion by examining these genes for more isolates that develop resistance.