16S rRNA methyltransferase KsgA contributes to oxidative stress and antibiotic resistance in Pseudomonas aeruginosa Authors: Kamonwan Phatinuwat, Sopapan Atichartpongkul, Watthanachai Jumpathong, Skorn Mongkolsuk, Mayuree Fuangthong Published: 2024 ABSTRACT AbstractRibosomal RNA (rRNA) modifications are involved in multiple biological processes. KsgA is a 16S rRNA adenine dimethyltransferase that methylates at the adenines 1518 and 1519 (A1518/1519) positions, which are located near the ribosome decoding center. These methylations are conserved and important for ribosome biogenesis and protein translation. In this study, we demonstrated the absence of A1518/1519 methylation in the 16S rRNA of a Pseudomonas aeruginosa ksgA mutant. Biolog phenotypic microarrays were used to screen the phenotypes of the ksgA mutant against various antimicrobial agents. The loss of ksgA led to increased sensitivity to menadione, a superoxide generator, which was, at least in part, attributed to decreased in a superoxide dismutase (SOD) activity. Interestingly, the decrease in SOD activity in the ksgA mutant was linked to a decrease in the SodM protein levels, but not the sodM mRNA levels. Furthermore, the ksgA mutant strain exhibited sensitivity to hygromycin B and tylosin antibiotics. The tylosin-sensitive phenotype was correlated with decreased transcriptional levels of tufA, tufB, and tsf, which encode elongation factors. Additionally, the ksgA mutant showed resistance to kasugamycin. Collectively, these findings highlight the role of KsgA in oxidative stress responses and antibiotic sensitivity in P. aeruginosa.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-024-78296-4.Keywords: Pseudomonas aeruginosa, rRNA methylation, ksgA, Oxidative stress, AntibioticsSubject terms: RNA, Bacterial genes, Gene expression, Gene regulation Introduction 16S ribosomal RNA (rRNA) modifications play a crucial role in the biogenesis of the bacterial 30S ribosomal subunit1. Three major types of bacterial rRNA posttranscriptional modifications are known: pseudouridine (Ψ), nucleotide base methylation, and 2′-O-ribose methylation2,3. These modified nucleotides are primarily clustered around the decoding center, the peptidyl transferase center, and the nascent peptide exit site of the ribosome, suggesting their importance in protein translation3,4. Many studies have demonstrated that the loss of rRNA modifications leads to changes in the structure of active site5,6, which subsequently slows translation rates and reduces translation fidelity3,7–9. Among the enzymes responsible for these modifications, KsgA has been reported to participate in ribosome biosynthesis and maintain ribosome functionality in Escherichia coli and Thermus thermophilus4,5,10. KsgA transfers two methyl groups from the methyl donor S-adenosyl-L-methionine to 16 S rRNA at A1518 and A1519 (A1518/1519), which are located in helix 45 of the 16S rRNA5,11. The crystal structure of the 30S subunit of ksgA mutant showed a loss of hydrogen-bonding interactions between helices 44 and 45 of 16S rRNA, and U793 (located in helix 24a) was not occupied by the methylated adenosines5. These findings indicate that methylated adenosines are important for the proper packaging of helices 24a, 44, and 455. The ksgA mutant reduced translational fidelity in cases of stop-codon read-through and frameshifting, suggesting that methylated A1518/1519 are crucial for protein translation12. KsgA orthologs are conserved among bacteria, archaea, and eukaryotes13. Previous research on ksgA in certain bacteria has shown a pleiotropic phenotype associated with ksgA deletion. Notably, a ksgA mutation has been found to confer resistance to the aminoglycoside antibiotic kasugamycin in E. coli, Neisseria gonorrhoeae, Salmonella enterica, and Chlamydia trachomatis14–17. The lack of ksgA also reduced the virulence of Staphylococcus aureus in silkworms12. The S. aureus ksgA mutant showed increased sensitivity to oxidative stress (menadione and hydrogen peroxide), decreased survival rate in murine macrophages, and reduced translational fidelity12. Moreover, the lack of ksgA in S. enteritidis results in reduced virulence in chickens17. The impact of ksgA in Pseudomonas aeruginosa, which is ubiquitous in diverse environments including the human body18, in response to antimicrobial agents remains unknown. We hypothesized that ribosomes containing unmethylated adenines in 16S rRNA may broadly affect protein synthesis, particularly during exposure to antimicrobial agents in P. aeruginosa. We utilized Biolog phenotypic microarrays (PMs) to assess the phenotypic response of the P. aeruginosa ksgA mutant against a variety of antimicrobial agents. The results indicated that the P. aeruginosa ksgA mutant exhibited sensitivity to menadione, hygromycin B, and tylosin. Materials and methods Bacterial strains and plasmids Pseudomonas aeruginosa strain UCBPP-PA14 (PA14) was used as the wild-type strain. All Escherichia coli and Pseudomonas strains were grown at 37 °C in lysogeny medium (Lennox, USA). The following concentrations of antibiotics were used: 200 µg/mL carbenicillin (Cb) and 75 µg/mL gentamicin (Gm) for P. aeruginosa; 15 µg/mL Gm, 100 µg/mL ampicillin (Ap), and 15 µg/mL kanamycin (Km) for E. coli. All bacterial strains, plasmids, and primers used in this study are listed in Table S1 and Table S2, respectively. Construction of the ksgA mutant and ksgA complemented strains The upstream and downstream regions of ksgA were amplified using PCR from gDNA of wild type with the primer pairs BT7760/BT7761 and BT7762/BT7763, respectively. The loxP with the Gm resistance cassette (loxGm) region was amplified using PCR from vector pCM35119 with primer pair LoxP R/LoxPL-0. These fragments were cloned into pKNOCK-Ap20 to yield pKNOCK-ksgA-loxGm. This construct was transferred by transformation into E. coli BW20767 and then by conjugation into PA14. The vector pCM15719, expressing Cre-recombinase, was transferred into the recombinant PA14 to remove the loxGm region. The ksgA mutant strain was verified using PCR and Southern blot analysis. The full-length (FL) ksgA gene was amplified via PCR from gDNA of wild type with primer pair BT7808/BT6206 and was then cloned into the pUC18 mini-Tn7T-Gm vector21 to produce pUC18-ksgA. Co-transformation of pUC18-ksgA and pTNS2 into the ksgA mutant strain resulted in the insertion of the DNA region between Tn7L and Tn7R into a specific site after glmS in the chromosome. pTNS2 is a helper plasmid that encodes only the specific TnsABC + D transposition pathway. The complemented strain containing a single copy of ksgA was verified using PCR. Primer extension analysis Primer extension analysis following a previously established protocol22 involved reverse transcription of P. aeruginosa total RNA using a radioactively the labeled primer BT8042 complementary to 16S rRNA. A sequencing ladder was prepared using a Thermo Sequenase Cycle Sequencing Kit (Applied biosystems) with the primer BT8042. The reverse transcription reactions were performed using SuperScript III reverse transcriptase (Invitrogen). Extended products were electrophoresed in 10% polyacrylamide gel electrophoresis with 6 M urea and visualized using a Typhoon phosphor imager (Cytiva). Analysis of the modified rRNA fragment using an orbitrap mass spectrometer The 16S rRNA was isolated by electrophoresis in a 4% polyacrylamide gel containing 7 M urea. Approximately 100 fmol of 16S rRNAs were then digested with RNase T1 (Thermo Scientific) at 37ºC for 1 h23. The digested rRNA was purified using Strata™ solid-phase extraction columns (Phenomenex) and directly infused for ionization mass analysis using an orbitrap mass spectrometer (Q Exactive Focus, Thermo Scientific). The mass spectrometer was operated under the negative ion mode with the scan m/z range of 600–1800, and scan parameters including AGC target, 1 × 106; maximum injection time, 55 ms; and resolution, 35,000. The hot electrospray (HESI) was set up as follows: sheath gas flow rate, 7 (arbitrary unit); auxiliary gas flow rate, 3 (arbitrary unit); spray voltage, 3.00 kV; capillary temperature, 275 °C; auxiliary gas heater temperature, 100 °C; and S-lens radio frequency level, 70.0 (arbitrary unit). Prior to direct infusion (flow rate of 0.100 mL/min), the purified rRNAs were prepared as previously published method24 with slight modifications, in which both purified rRNAs were dissolved in a solution of 1:1 acetonitrile/water in the presence of 1%v/v 1,1,1,3,3,3-Hexafluoro-2-propanol, 10 mM imidazole, and 23 µM formic acid. Phenotypic microarrays (Biolog) Biolog phenotypic microarrays (PMs) assays were performed following the manufacturer’s protocol (Biolog). Colonies of wild-type and ksgA mutant strains were resuspended in IF-10 solution to 85% transmittance. The suspension was 200-fold diluted in IF-10 solution with dye A. Then, 100 µL of the culture was inoculated into each well of Biolog PM11-20 plates and incubated in a Biolog Omnilog reader at 37 °C for 24 h. The Biolog OmniLog PM software was used to analyze cell respiration kinetics and calculated the ΔAUC value for each well, defined as ΔAUC = AUCMT-AUCWT, where AUCMT and AUCWT are the areas of the kinetic curves for the ksgA mutant and wild type, respectively. Minimum inhibitory concentration (MIC) assay The MICs of all chemicals were determined using the broth microdilution method25,26. Cell cultures were adjusted to the MacFarland 0.5 standard, and diluted 20-fold in normal saline. Then, 10 µL of the adjusted cultures were added to a 96-wells plate containing serial dilutions of the test chemicals (kasugamycin, hygromycin B, cefsulodin, tylosin, and sodium bromate). Results were detected using a microplate spectrophotometer at OD600nm after incubation at 37 °C overnight. MIC was defined as the lowest concentration that completely inhibited the bacterial growth. Oxidative stress sensitivity test using a killing assay The cell cultures in the exponential growth phase were diluted to OD600nm of 0.1 and treated with 400 mM menadione for 1 h. The cell cultures were serially diluted 10-fold in normal saline; and 10 µL of each dilution was dropped on LB agar plates. The cultures were incubated at 37 °C overnight. The colonies were counted; and the percent survival of treated cultures was determined relative to the percent survival of untreated cultures. Superoxide dismutase (SOD) enzyme activity assay Total SOD activity was monitored using xanthine–xanthine oxidase-coupled reduction of cytochrome c27. Cell cultures in the stationary phase were incubated with 100 mM menadione for 30 min at 37 ºC. The cell pellets were lysed by sonication. SOD activity was measured in a solution containing 50 mM potassium phosphate buffer (pH 7.5), 30 µM ferricytochrome c, and 100 µM xanthine. Ferricytochrome c reduction was measured at an absorbance of 550 nm. One unit of SOD activity per mg of protein (U/mg) is defined as the amount required to inhibit the reduction rate of ferricytochrome c by 50%. Western blot analysis Streptavidin (strep)-tagged full-length gene fragments were amplified using PCR with specific primers: BT8123/BT8671 for sodM and BT8672/BT8671 for sodB. The fragments were cloned into the pBBR1-MCS4 vector and transformed into wild-type, ksgA mutant, and complemented strains. Exponential phase cultures (OD600 of about 0.4) were incubated with 100 mM menadione for 30 min. Total protein was extracted and separated using 15% SDS-PAGE for SodM and Native-PAGE for SodB. Proteins were transferred to a polyvinylidene difluoride membrane, blocked with TBST buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.05% Tween 20) containing 5% skim milk at 4 ºC overnight, and then incubated with anti-strep antibody (Abcam) for 1.5 h. After washing with TBST, the membrane was incubated with goat anti-rabbit horseradish peroxidase antibody (Enzo Life Sciences). The target protein was detected using a chemiluminescent ECL kit (Cytiva). Semiquantitative real-time PCR Exponential phase cultures (OD600 of about 0.4) were used for total RNA extraction. For sod genes expression, the exponential phase cells were treated with or without 100 mM menadione for 10 min at 37 °C. For elongation factor genes, untreated exponential phase cells were used. Total RNA was treated with DNase (Thermo Fisher Scientific) and used to synthesize cDNA using RevertAid Reverse Transcriptase (Thermo Fisher Scientific). Semiquantitative real-time PCR was performed according to the manufacturer’s instructions using KAPA SYBR FAST reagent (KAPA Biosystems). The specific primer pairs for the genes of interest are presented in Table S2. The 16S rRNA gene was used as an internal control to normalize mRNA expression. Results P. aeruginosa KsgA encodes a 16S rRNA dimethyltransferase P. aeruginosa PA14 KsgA exhibited 50% identity and 79% similarity with E. coli K-12 KsgA (Supplementary Fig. S1). The E. coli KsgA has been experimentally identified as a 16S rRNA dimethyltransferase11,28 that specifically modifies nucleotides A1518 and A1519 at position N6 (Fig. 1a)29. The P. aeruginosa KsgA presents the hallmark sequences for binding to the methyl group donor S-adenosyl-L-methionine (SAM), shown as motifs I to VIII (Supplementary Fig. S1)16. Therefore, the function of the P. aeruginosa KsgA was predicted as a methyl transferase that methylates adenine at positions 1518/1519 in 16S rRNA. In the present study, a primer extension analysis was performed to probe the role of P. aeruginosa ksgA in this methylation. The m26A product of the KsgA reaction inhibits strand elongation by polymerase30; therefore, a short DNA product is expected when A1518/1519 methylation is present in the 16S rRNA. As expected, the reverse transcriptase was primarily stopped at guanosine 1520 when using RNA from wild-type and complemented P. aeruginosa strains (Fig. 1b). A band corresponding to the stop at adenine 1519 was also observed. However, reverse transcriptase activity was not stopped at either guanosine 1520 or 1519 when using RNA from the ksgA mutant strain, suggesting that 16S rRNA isolated from the ksgA mutant lacks methylation at A1518/1519 (Fig. 1b). The DNA product observed in the 16S rRNA isolated from the ksgA mutant arose due to the presence of m3U at position 1498. ksgA is responsible for dimethyladenosine formation in 16S rRNA. The N6, N6 dimethyladenosine biosynthesis in E. coli (a). Modifications at A1518/1519 of 16S rRNA were examined using primer extension analysis (b). The sequencing products were electrophoresed in G, A, T, and C wells. The nucleotides were numbered according to the Escherichia coli rRNA sequence numbering system. For a full primer extension image, see Supplementary Figure S3. The mass spectrum analysis of 16S rRNA fragments from the ksgA mutant (c) and wild-type strains (d). The AACCUGp oligonucleotides from the ksgA mutant undergo the loss of phosphate (-H2PO3) and a non-methylated adenosine nucleotide (-Ap). The m26Am26ACCUGp oligonucleotides from the wild type undergo the loss of a dimethylated adenosine nucleotide (-m26Ap). We also confirmed that ksgA is responsible for the methylation of A1518/1519 on 16S rRNA by analyzing the digested 16S rRNA fragments from the wild-type and ksgA mutant strains using an orbitrap mass spectrometer. The mixture of digested 16S rRNA fragments from the ksgA mutant contained a doubly charged negative ion ([M – 2H]2–obs = 967.6275, [M – 2H]2-calc for C57H71O42N23P6 = 967.6271, mass error = 0.4 ppm) in a hexamer (AACCUGp, MW 1936.260), with no detection of the wild-type 16S rRNA fragment (Fig. 1c). Upon ionization at the ESI probe, each monomer of the AACCUGp oligonucleotide was fragmented from 5’ to 3’ without induction by the higher collision-induced dissociation (HCD) cell, which enabled determination of the sequence of the ksgA mutant 16S rRNA without conducting MS/MS fragmentation using higher-energy collisional dissociation (HCD). As shown in Fig. 1c, the loss of phosphate from the parent ksgA oligonucleotide yielded the observed m/z value of 927.6050. The m/z transition from 927.6050 to 803.1037 (w5 ion) clearly indicates the loss of a non-methylated adenosine nucleotide. Another loss of a non-methylated adenosine nucleotide was again observed from the m/z transition of 803.1037 (w5 ion) to 638.5762 (w4 ion). The mass errors of all the daughter ions were calculated, as shown in Supplementary Table S3. The MS analysis of the 16S rRNA of the wild type clearly revealed the doubly charged negative ion ([M – 2 H]2–obs = 995.6587, [M – 2 H]2-calc for C61H79O42N23P6 = 995.6584, mass error = 0.3 ppm) in the hexamer containing m26A (m26Am26ACCUGp, MW 1992.321) (Fig. 1d). The w4 and w5 ions of the wild-type 16S rRNA had molecular masses of 638.5752obs and 815.1086obs, respectively, suggesting the loss of a dimethylated adenosine nucleotide (m26A). Ideally, the expected m/z value of w5 is 817.1170; however, the detected m/z w5 of 815.1086 corresponds to the species that loses four additional hydrogen atoms (Supplementary Fig. S2). Another possibility might be the abstraction of four additional protons from the phosphate groups simultaneously with the generation of cation radicals on nitrogen atoms (Supplementary Fig. S2). At this point, the MS results strongly indicate that ksgA encodes a methyltransferase responsible for the dimethylation of adenosine of 16S rRNA at positions A1518/1519. Phenotypic screening of the ksgA mutant for antimicrobial sensitivity using Biolog phenotypic microarrays Biolog phenotype microarrays facilitate the study of the potential function of an unknown gene. The Biolog PMs 11–20 were used to screen phenotypes of the ksgA mutant in P. aeruginosa PA14, while the P. aeruginosa PA14 wild type was used as a control strain. The PM 11–20 tests 240 antimicrobial agents at 4 concentrations for each antimicrobial agent (Supplementary Fig. S4 and Table S4). Cell respiration (NADH production) was used as the endpoint measurement for cell survival. We observed that the ksgA mutant in P. aeruginosa was susceptible to multiple antibiotics (Fig. 2). The areas under the cell respiratory kinetic curve (AUCs) were compared between the ksgA mutant and the wild type (∆AUC) (Fig. 2). The top 5 antimicrobial agents that showed high |∆AUC| values were hygromycin B (-13,892.70), cefsulodin (-11,926.67), tylosin (-10,166.70), sodium bromate (-7,178.67), and menadione (4,675.00) (Table 1). All 5 antimicrobial agents showed negative ∆AUC values, suggesting a sensitive phenotype of the ksgA mutant. The growth kinetic response curves of wild-type and ksgA mutant strains to these antimicrobial agents are shown in Fig. 2b-f. Confirmation experiments were performed for each antimicrobial agent using a broth microdilution minimum inhibitory assay (MIC) or a killing assay (Table 2and Fig. 3a). However, the ksgA mutant showed a confirmed sensitivity toward only three antimicrobial agents, namely menadione, tylosin, and hygromycin B. The phenotype of thePseudomonas aeruginosa ksgA mutant against antimicrobial agents. Heatmaps of phenotypic array profiles performed using Biolog phenotype microarrays (PMs) (a). The column represents the difference of the area under the curve of the ksgA mutant compared to the wild type (∆AUC). Blue represents high relative growth (ksgA mutant gains the phenotype), while yellow represents low relative growth (ksgA mutant lost the phenotype) under antimicrobial reagent conditions. Four concentrations of each antimicrobial agent from lowest to highest concentrations: 1, 2, 3, and 4, were tested. Growth kinetic response curves of wild-type and ksgA mutant strains in response to hygromycin B (b), cefsulodin (c), tylosin (d), sodium bromate (e), and menadione (f) exposures. The OmniLog unit is defined as 500 times the OD, as described by Biolog Inc. Table 1The antimicrobial agents and different AUCs of the top 5 antimicrobial agents that showed high |∆AUC| values. Asterisks denote a significant difference by t-test (*p ≤ 0.05, ***p ≤ 0.001, and ns indicates not significant).Antimicrobial agentsSubtractive difference(AUC of ksgA mutant-WT)Hygromycin B− 13,892.70***Cefsulodin− 11,926.67*Tylosin− 10,166.70*Sodium bromate− 7,178.67nsMenadione− 4,675.00* The antimicrobial agents and different AUCs of the top 5 antimicrobial agents that showed high |∆AUC| values. Asterisks denote a significant difference by t-test (*p ≤ 0.05, ***p ≤ 0.001, and ns indicates not significant). Table 2The median MIC for hygromycin B and tylosin in the wild-type, ksgA mutant, and complemented strains.MIC (µg/mL)Wild typeksgA mutantksgA complementedHygromycin B32832Cefsulodin1.51.51.5Tylosin1,0245121,024Sodium bromate16,00016,00016,000Kasugamycin2502,000250 The median MIC for hygromycin B and tylosin in the wild-type, ksgA mutant, and complemented strains. ksgA affects superoxide dismutase (SOD) activity. The menadione-sensitive phenotype in the ksgA mutant strain (a). Graph shows survival of the wild-type, ksgA mutant, and complemented strains after exposure to 400 mM menadione. Culture without menadione exposure was used as a control. Total SOD activity of the wild-type, ksgA mutant, and complemented strains after menadione exposure (b). Relative mRNA expression levels of sodM (c) and sodB (d) in the wild-type, ksgA mutant, and complemented strains. Relative SodM (e) and SodB (f) protein expression levels in wild-type, ksgA mutant, and ksgA complemented strains. The band intensities were measured using ImageQuant TL 7.0. Western blot analysis of SodM (g) and SodB (h) proteins. The amount of total protein loaded in each gel is shown as a loading control. All graphs and statistical analysis were performed using GraphPad Prism 9.0.0. Data are expressed as means ± standard deviations of at least three biological replicates. Asterisks denote a significant difference by one-way ANOVA (*p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001). For full western blot image, see Supplementary Figure S6. SOD enzymatic activity and SodM protein levels are reduced in the ksgA mutant strain The sensitivity of the ksgA mutant strain to menadione was confirmed using an assay of oxidative stress sensitivity by killing. The percent relative survival after menadione exposure was lower for the ksgA mutant strain (7.1%) than for the wild-type (21.7%) and complemented (25.1%) strains (Fig. 3a). Menadione is a well-known superoxide anion generator through redox cycling reactions31. Superoxide dismutase (SOD) is the major enzyme that rapidly catalyzes the conversion of superoxide radicals to molecular oxygen and hydrogen peroxide32. Loss of KsgA may therefore affect SOD activity, thereby reducing the capacity of P. aeruginosa to defend against menadione. We tested this possibility by comparing the SOD enzymatic activity of the wild-type, ksgA mutant, and complemented strains. In the unexposed condition, the SOD enzymatic activities of the wild-type, ksgA mutant, and complemented strains showed no statistically significant differences (27, 24, and 27 units/mg protein, respectively; Fig. 3b). Exposure to 100 mM menadione, a concentration that shows no difference of growth for the ksgA mutant in comparison to the wild type (Supplementary Fig. S5a), resulted in significantly lower SOD enzymatic activity in the ksgA mutant (28 units/mg protein) than in the wild-type (39 units/mg protein) and complemented (37 units/mg protein) strains (Fig. 3b). This observation suggests that the lack of ksgA may contribute to a decrease in SOD activity, thereby a decrease in the menadione detoxification capability in P. aeruginosa. The SOD levels in the ksgA mutant may not be sufficient to detoxify menadione, resulting in the sensitive phenotype of the ksgA mutant. P. aeruginosa possesses two superoxide dismutases, sodB and sodM, which encode the iron- and manganese-cofactor SODs, respectively33–35. We used qRT-PCR to test whether the menadione sensitivity of the ksgA mutant was due to the sodM and/or sodB mRNA levels. During the menadione treatment, the transcriptional levels of sodM and sodB in the ksgA mutant did not differ significantly from those of the wild type, indicating that the sodM and sodB transcriptional levels were ksgA independent and did not contribute to the low SOD activity (Fig. 3c and d). Subsequent western blot analysis of the levels of SodM and SodB proteins revealed a significantly lower level of SodM protein in the ksgA mutant than in the wild-type strain after 100 mM menadione exposure (Fig. 3e and g). By contrast, the SodB protein levels were similar in both the ksgA mutant strain and the wild-type strain, either with or without menadione exposure (Fig. 3f and h). These findings suggest that KsgA might mediate the menadione stress response in P. aeruginosa via SodM at the translational level. The absence of ksgA contributes to an altered antibiotic response Among four antimicrobial agents that showed high |∆AUC| values in the Biolog phenotype microarrays, only the sensitive phenotypes to hygromycin B and tylosin antibiotics were confirmed by MIC assays (Table 2). The MIC for hygromycin B was 4-fold lower for the ksgA mutant strain (8 µg/mL) than for the wild-type (32 µg/mL) or complemented strains (32 µg/mL); and the MIC for tylosin was 2-fold lower in the ksgA mutant (512 µg/mL) than in the wild-type (1,024 µg/mL) and complemented (1,024 µg/mL) strains. The growth curves of the wild-type, ksgA mutant, and complemented strains showed similar growth characteristics supporting that the sensitive phenotypes toward hygromycin B and tylosin are not from growth defect (Supplementary Fig. S5b). Moreover, the ksgA mutation in other organisms (E. coli, Neisseria gonorrhoeae, Salmonella enterica, and Chlamydia trachomatis) has shown conserved kasugamycin-resistance phenotypes14–17. In the present study, the P. aeruginosa ksgA mutant strain also showed an 8-fold increase in the MIC of kasugamycin (Table 2). Tylosin, a bactericidal macrolide used in veterinary medicine to treat bacterial infections, is effective against both Gram-positive and Gram-negative bacteria36. Tylosin inhibits protein synthesis by binding to the 50 S ribosomal subunit37. We hypothesized that the tylosin sensitivity of the ksgA mutant strain was linked to reduced expression of tylosin resistance-related genes, such as genes encoding elongation factor Tu and elongation factor Ts. The expression of these genes at mRNA and protein levels were highly up-regulated in tylosin-resistant Streptococcus suis38. The overexpression of elongation factors was suggested to facilitate tRNA binding to ribosomes, catalyze ribosomal translocation, and coordinate tRNA movement during the development of tylosin resistance38,39. In P. aeruginosa, the elongation factor Tu is encoded by paralogous genes tufA and tufB, and elongation factor Ts is encoded by tsf. Analysis of the mRNA levels of tufA, tufB, and tsf using quantitative RT-PCR revealed 0.71, 0.76, and 0.58-fold lower mRNA levels for tufA, tufB, and tsf in the ksgA mutant than in the wild-type and complemented strains (Fig. 4). The lack of ksgA therefore contributed to decreased tufA, tufB, and tsf expression at the transcription level and altered the tylosin-sensitive phenotype in the ksgA mutant. Relative mRNA expression levels oftsf, tufA, and tufB in wild-type, ksgA mutant, and complemented strains. Data are expressed as means ± standard deviations of five biological replicates. The statistical analysis was performed using GraphPad Prism 9.0.0. Asterisks denote a significant difference by one-way ANOVA (*p ≤ 0.05). Discussion In this study, we revealed that P. aeruginosa KsgA is involved in A1518 and A1519 methylation on 16S rRNA, as supported by primer extension and MS analysis. The methylation at A1518/1519 was experimentally confirmed in E. coli40 and S. aureus12, and these methylations are considered important for 30S subunit biosynthesis5. KsgA first binds to 17S rRNA and methylates helix 45 at A1518/151941. RNase then cleaves 17S rRNA to produce 16S rRNA, and the precursor rRNAs and proteins are assembled to generate the ribosome41,42. The crystal structure of the 16S rRNA from wild-type T. thermophilus HB8 showed hydrogen bonding interactions between the parts of helix 44 and helix 45 in the decoding center. In the absence of methylation at adenines A1518/1519, as in the ksgA mutant, the hydrogen-bonding interactions between helix 44 and helix 45 did not occur, resulting in a shift in the N6 atoms of A1518 and A1519 of approximately 5Å and 4Å, respectively5. The methylations of A1518/1519 also led to unbonding between the A1516 and A1519 loop. This caused the loop to open, thereby facilitating hydrogen-bonding interactions between helix 44 and helix 45, which stabilized the 30S ribosomal subunit5. Our results indicate that ksgA mutant is sensitive to menadione in P. aeruginosa, which is consistent with previous reports of the menadione-sensitive phenotype in S. aureus12. The observations in the present study suggest that the KsgA might mediate the menadione stress response in P. aeruginosa through SodM at the translational level. Interestingly, we also observed reduced SOD activity in the ksgA mutant toward paraquat, which is another superoxide anion producer (Supplementary Fig. S9). Superoxide dismutase (SOD) serves as the first line of defense against superoxide stress by converting superoxide into hydrogen peroxide and oxygen, thus protecting cells from the toxic effects of superoxide33. Pseudomonas aeruginosa possesses two types of superoxide dismutases: manganese-cofactored SOD (Mn-SOD, encoded by sodM) and iron-cofactored SOD (Fe-SOD, encoded by sodB). The expression of these SODs is regulated by environmental factors. Fe-SOD is predominantly expressed under high extracellular iron concentration. In contrast, when iron availability is limited, Fe-SOD expression decreases and Mn-SOD production increases33. Moreover, Mn-SOD is expressed during the stationary phase and can be induced by menadione (a superoxide generator)43. Methylation of 16S rRNA at the ribosome at A1518 and A1519 near the decoding center results in a conformational change in that region5 (Supplementary Fig. S8a) and likely interferes with the binding of antibiotics, such as kasugamycin and hygromycin B. The aminoglycoside kasugamycin inhibits the initiation of translation by blocking formylmethionyl-tRNA binding to the 30S subunit11. The crystallization of the 30S ribosomal subunit of T. thermophilus HB8 ksgA mutant indicates the absence of hydrogen bonding between helix 44 and 45 of 16S rRNA, which confers a slight conformational change in the kasugamycin binding pocket5. Therefore, the lack of A1518/1519 methylation that leads to kasugamycin resistance most likely has indirect effects on the bases involved in kasugamycin binding44,45. The resistance to the aminoglycoside antibiotic kasugamycin is conserved in diverse organisms, including E. coli, N. gonorrhoeae, S. enterica, and C. trachomatis14–17. In this study, the P. aeruginosa ksgA mutant strain also showed resistance to kasugamycin (Table 2). Hygromycin B binds to the bacterial ribosome 30S ribosomal subunit at 16S rRNA residues and interferes with the translocation of ribosomes, ultimately leading to the inhibition of protein synthesis in bacterial cells46. The A1518/1519 residues of 16S rRNA are adjacent to the hygromycin B binding pocket47. We speculated that the absence of methylation at A1518/1519 in 16S rRNA may cause a slight conformational change in the hygromycin B binding site that increases the binding affinity, resulting in increased susceptibility to hygromycin B. Structures of the 30S ribosome containing 16S rRNA from T. thermophilus HB8 were available for the wild-type (PDB ID: 1J5E48) and the ksgA mutant strains (PDB ID: 3OTO5). The residues within the hygromycin B binding pocket of both T. thermophilus HB8 and P. aeruginosa were identical (Supplementary Fig. S7). Molecular docking revealed free energies of binding of -9.05 and − 10.18 kcal/mol for hygromycin B to 16S rRNA of the T. thermophilus wild type and ksgA mutant, respectively (Supplementary Fig. S8). The estimated inhibition constants (Ki) for the ksgA mutant was approximately 6.7-fold lower than that for the wild type. These results may indicate that hygromycin B binds more tightly to the 16S rRNA ksgA mutant than to wild-type 16S rRNA. However, further experimental validation is required to confirm these findings. Tylosin is a macrolide antibiotic approved as a veterinary medicine to control infectious diseases, including mastitis, liver abscesses, and dysentery, in cattle, poultry, pigs, and other farm animals37. It has a broad spectrum of activity against Gram-positive bacteria and a limited range of Gram-negative bacteria37,49. Tylosin is commonly used to reduce the incidence of liver abscesses (containing Pseudomonas spp.)50. Tylosin inhibits protein synthesis by binding to the 50 S ribosome at the exit tunnel of the ribosome, thereby inhibiting bacterial survival37. During protein synthesis, elongation factor Tu (EF-Tu) promotes the binding of aminoacyl-tRNA to the ribosome, while elongation factor Ts (EF-Ts) restores EF-Tu to its active form51,52. Elongation factor G catalyzes the movement of tRNAs and mRNA during the translocation step in the ribosome39. FolD is involved in the synthesis of a formyl donor for the formylation of initiator tRNA39. The overexpression of elongation factors could therefore facilitate tRNA binding to ribosomes, catalyze ribosomal translocation, and coordinate tRNA movement39. The protein and mRNA levels of elongation factors Ts (tsf), Tu (tuf), and G (fusA) were upregulated in tylosin-resistant Streptococcus suis38. Similar results were obtained in tylosin-resistant Mycoplasma gallisepticum, which also showed upregulation of the protein levels of elongation factor (Tu and G) and FolD39. Here, we demonstrated a decrease in the mRNA levels for elongation factors (tuf and tsf) (Fig. 4), and these decreases may contribute to tylosin sensitivity in the ksgA mutant of P. aeruginosa. In conclusion, this study demonstrated that P. aeruginosa KsgA is involved in the A1518 and A1519 methylations of 16S rRNA, which is a component of the 30S ribosome. The sensitive phenotype of the ksgA mutant toward three antimicrobial agents-menadione, tylosin, and hygromycin B-was observed. The menadione-sensitive phenotype is likely a result of decreased superoxide dismutase (SOD) activity and reduced SodM at the translational level. The ksgA mutant also demonstrates a reduction in the transcriptional levels of tylosin-resistance related genes that may contribute to the observed sensitivity to tylosin in the ksgA mutant P. aeruginosa strain. In addition, the P. aeruginosa ksgA mutant also exhibited resistance to kasugamycin. Taken together, the findings support an involvement of ksgA in oxidative stress and antibiotics resistance. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements The authors are grateful to Weerachai Tanboon (Laboratory of Biotechnology, Chulabhorn Research Institute) and Sakuna Wan-erlor (Laboratory of Chemical Sciences, Chulabhorn Graduate Institute) for their technical assistance. The authors also thank Dr. Rapeepat Sangsuwan (Laboratory of Natural Products, Chulabhorn Research Institute) for her valuable comments. Author contributions KP: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. SA: Conceptualization, Formal analysis, Investigation, Methodology, Writing – review & editing. WJ: Investigation, Formal analysis, Methodology, Visualization, Writing – review & editing. SM: Funding acquisition, Writing – review & editing. MF: Conceptualization, Funding acquisition, Supervision, Project administration, Writing – review & editing. Funding This work was supported by Chulabhorn Graduate Institute Scholarship (CGS): CGS(2020)/13, Thailand Science Research and Innovation (TSRI), Chulabhorn Research Institute (Grant no. 48293/4691984), and in part by the grant from Center of Excellence on Environmental Health and Toxicology (EHT), OPS, Ministry of Higher Education, Science, Research and Innovation. Data availability All data generated or analyzed during this study are included in this published article and its supplementary information files. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data generated or analyzed during this study are included in this published article and its supplementary information files.