XU Xiaoxiao, LIN Hong, GUO Jiamin, LIU Pei, and SUN Haixin, *

Subinhibitory Levels of Fluoroquinolones Result in Enrichment of the Membrane Proteome of

XU Xiaoxiao1), 3), LIN Hong2), GUO Jiamin1), 3), LIU Pei1), 3), and SUN Haixin1), 3), *

1),,266071,2),,,266003,3),,266071,

is a common marine foodborne pathogen. In this study, antibiotics ciprofloxacin and enrofloxacin were used to induce drug-resistance inThe differentially expressed proteins (DEPs) were analyzed and compared with those in the bacteria cultured without antibiotics. The primary proteomic alterations were in the levels of cell membrane components and pro- teins related to lysine and folic acid biosynthesis, which were all significantly up-regulated. The minimal inhibitory concentrations (MIC) for both test drugs were elevated to 10 μg mL?1following serial passaging. These results indicated that, for both ciprofloxacin and enrofloxacin, drug-resistance were developed even in the subinhibitory levels and the primary response was a major alteration in the cell membrane proteome. These changes were similar to those observed incultured with super-MIC levels of these anti- biotics. The current study provides a theoretical basis for in-depth study of the related changes of marine foodborne pathogens in sub- inhibitory concentrations that are commonly found.

subinhibitory environment; marine foodborne pathogen; proteomics; fluoroquinolone

1 Introduction

The problem of drug contamination of marine food is becoming increasingly serious. Maintaining marine food sa- fety from foodborne pathogens is an increasing global chal- lenge. Enterotoxigenic food poisoning due to the ubiquitous Gram-positive bacteriumis of great concern since this organism can cause purulent infections, pneumonia, myocarditis, pericarditis and other diseases in humans and animals (Liu., 2022). The current preven- tion and treatment methods ofinfections rely pri- marily on-lactams, macrolides, glycopeptides and other antibiotics (Mlynarczyk-Bonikowska., 2022). Impor- tantly, drug resistance inis currently on the rise (Guo., 2020) and the environments where drugs are widely used can also provide antibiotic cross-resistance to non-target microorganisms (Dawan., 2020). Fluoroqui- nolone (FQ) drugs such as ciprofloxacin (CIP) and enroflo- xacin (ENR) are widely used and even abused, causing se- rious marine food contamination. However, whether FQ- resistance has been induced inexposed to subin- hibitory levels of CIP and ENR is currently unknown. Such information can be helpful for risk assessment of bacterial contaminants in marine foods and in other areas such as market and farm environments (Qiao, 2021).

Current research of drug resistance in microorganisms is primarily based on the use of DNA or RNA amplification methods using the polymerase chain reaction (PCR) and whole genome sequencing methods (Wang., 2021). Proteomic analyses have not been utilized to investigate the drug subinhibitory environments of non-targeted microor- ganisms. Proteomics can identify specific expressed pro- teins that are directly related to specific antibiotic-resistance phenotype. In this study, we applied quantitative proteomic analysis to examine the alterations inexposed to subinhibitory levels of ciprofloxacin and enrofloxacin.

2 Materials and Methods

2.1 Materials

The FQ sensitivestrain C2CC21600 was do- nated by Shandong Seatone Detection Evaluation Co., Ltd. CIP and ENR powders were obtained from commercial sup- pliers. Mueller-Hinton (MH) medium and Luria Bertani (LB) medium were purchased from Qingdao Hopebio Biotech- nology (Qingdao, China).

2.2 Methods

2.2.1 MIC determinations and resistant strain cultures

Minimum inhibitory concentration (MIC) experiments were carried out by NCCLS microbroth dilution method as previously described (Shen., 2019). In brief, CIP and ENR were serially diluted in MH liquid medium to final concentrations of 40, 20, 10, 5, 2.5, 1.25, 0.625, 0.313, 0.156, 0.078 and 0.039 μg mL?1. MH without antibiotics was used as a positive control. The cultures were incubated at 37℃ for 18 h and the MIC values were recorded.were cultivated in LB containing sub-minimal inhibitory concen- trations (1/2 MIC) of CIP and ENR with shaking at 150 r min?1at 37℃ overnight, and serially passaged, while main- taining the same culture conditions for each generation, and the MIC of each generation was measured until the strain developed CIP and ENR resistance (refer to NCCLS drug susceptibility testing standards).

2.2.2 Liquid Chromatography-Mass Spectrometer analysis

Whole protein from antibiotic-resistant strains were iso- lated as described previously (Xie., 2012). In brief, the test strains were inoculated in LB medium containing 5μg mL?1ciprofloxacin and enrofloxacin and cultured overnight as described above. Then they were subcultured and grown to logarithmic growth period (OD600was about 1.0). The bacterial cells were collected with centrifugation at 4℃.

Samples for Liquid Chromatography-Mass Spectrome- ter (LC-MS) analysis utilized 300 μL of SDT lysate that was sonicated in a boiling water bath for 3 min and centrifuged at 16000 ×for 20 min. Supernatants were collected and to- tal protein content was quantified using BCA method while 10 μg protein was used for SDS-PAGE analysis. Totally 300 μg protein samples were used for filtration-aided sample preparation (FASP) enzymatic hydrolysis overnight. The re- sulting peptide mixture was desalted by passing through a C18 cartridge and was lyophilized under vacuum. The pow- ders were reconstituted with 0.1% trifluoroacetic acid and used for chromatographic separation with an Easy nLC 1200 system (Thermo Scientific, Pittsburg, PA, USA), DDA (Da- ta Dependent Acquisition) mass spectrometry analysis was performed on a Q-Exactive HF-X mass spectrometer (Ther- mo Scientific, Pittsburg, PA, USA) after peptide separation (Cox., 2014).

2.2.3 Bioinformatics

Protein identification from MS data was carried out us- ing the Proteome Discoverer software supplied with the ins- trument (Thermo Scientific). Significantly differentially ex- pressed proteins (DEP) were classified as those with ex- pression differences > 1.5-fold at a statistical significance le- vel of< 0.05 compared with proteins isolated from con- trol group. GO enrichment analysis and KEGG pathway en- richments were performed on DEP.

3 Results

3.1 MIC and Total Protein Analyses

The strain ofwe used for our experiments pos- sessed MIC values of 1.25 μg mL?1for both ciprofloxacin and enrofloxacin. The MIC values for both these drugs in- creased to 10 μg mL?1following 20 serial passages in the presence of 1/2 MIC concentrations (Table 1). This indicated that drug resistance was developed in the presence of sub- MIC levels of these drugs. We then analyzed the total pro- tein content for the pre- and post-induced strains and ob- tained typical protein profiles using SDS-PAGE. The ma- jority of the proteins migrated with molecular weights in the range of 15 – 100 kDa. Interestingly, the profiles from the drug-resistant isolates were increased significantly in the 25 and 40 kDa regions and decreased significantly at 70 kDa (Fig.1). Someefflux pump proteins, such as NorA MFS, are a class of fluoroquinolone-resistant efflux pump proteins with a molecular weight of 42.32 kDa as described previously by Zárate. (2019). In this study, protein en- richment in the 40 kDa regions has been linked to the in- creases of efflux pump proteins inunder subinhi- bitory environments. Our results were consistent with this previous study.

Table 1 MIC changes of bacteria after drug resistance induction in vitro

Fig.1 Total protein extracts taken from S. aureus cultured with or without 1/2 MIC levels of CIP and ENR. M, Molecular weight protein standards; 0, control (no antibiotic) group; 1, CIP group; 2, ENR group. Each set is represented in triplicate samples on the gel, respectively. Areas of interest indicated in the text are denoted by arrowheads.

3.2 Proteomic Analysis

3.2.1 Differential protein expression analysis

We used our protein samples to identify differentially ex- pressed proteins (DEP) and identified a total of 3925 DEPs (1934 up-regulated and 1991 down-regulated) in the CIP group. Similarly, the ENR group contained a total of 3869 DEPs that included 1912 up-regulated and 1957 down-re- gulated proteins. The similarity of the numbers for the up- and down-regulated proteins in the CIP and ENR groups indicated thathad similar drug-resistance mecha- nisms to CIP and ENR.

3.2.2 DEPs subcellular locations analysis

Subcellular localization analysis of DEPs revealed simi- lar numbers and proportions of DEPs induced by CIP and ENR, with most DEPs significantly elevated in the cell mem- brane and cytoplasm as well as in translating ribosomes (Fig. 2). These results were consistent with a previous study that indicated elevated efflux pump (membrane) expression was related to the formation of bacterial biofilms (Tretter., 2012). Biofilm formation and the direct excretion of drugs by efflux pumps can greatly improve bacterial resistance to antibiotics. Our results indicated that the majority of DEPs were present in the cell membrane, indicating it is a primary response of the bacteria to exposure to sub-inhibitory levels of CIP and ENR (Cox., 2014). These results were con- sistent with the SDS-PAGE analysis results.

Fig.2 Subcellular localization of S. aureus DEPs after cultured with 1/2 MIC levels of (A) CIP and (B) ENR respectively.

3.2.3 DEPs GO enrichment analysis

We used GO enrichment analysis to analyze DEPs frac- tions from three aspects: biological process (BP), molecu- lar function (MF) and cellular component (CC). The classifications of theDEPs we identified in CIP cul- tures were primarily involved in amino acid and carboxy- lic acid metabolism of BP, catalytic activity and transferase activity (MF), and cell membrane and cytoplasm compo- nents (CC) (Fig.3A). DEPs from the ENR cultures were si- milar, including small molecule and carbohydrate metabo- lism (BP), catalytic and transferase activity (MF), and cell membrane components (CC) (Fig.3B). These results were consistent with subcellular localization analysis, indicating that the generation of drug resistance was directly related to the formation of efflux pump proteins.

FQ resistance in bacteria under non-subinhibitory environ- ment is often the result of key mutations in the amino acid sequences encoded by GyrA and ParC genes (Hooper, 2015).The mutation at the acidic residues can reduce the overall catalytic activity by 5 – 10 times. As the structure of the en- coded DNA gyrase changes, the combination of drugs and target enzymes will be affected, resulting in drug resistance (Aldred., 2014; Yang., 2020). We found that DEPs were enriched in pathways related to amino acid metabo- lism and synthesis while the affected MF category included catalytic activity, indicating that the drug-resistance deve- loped in subinhibitory environment is consistent with that in non-subinhibitory environment.

3.2.4 DEPs KEGG pathway analysis

Analysis of DEPs in KEGG pathway from the CIP group indicated the up-regulation of lysine biosynthesis and other metabolic pathways, as well as the down-regulation of car- bon metabolism (Fig.4A). Similarly, DEPs for the ENR group were primarily enriched in folic acid biosynthesis and other metabolic pathways, while many proteins in carbon metabolic pathways were down-regulated (Fig.4B). Lysine is involved in energy metabolism and protein and central carbon metabolism, which usually include glycolysis, the pentose phosphate pathway and tricarboxylic acid cycle. This amino acid is a primary energy source required by or- ganisms and provides anabolic precursors (Zhao., 2010). We found that following CIP induction, lysine synthesis- related proteins were significantly up-regulated and carbon metabolism-related proteins were significantly down-regu- lated. These alterations suggested that energy metabolism was also associated with drug resistance under these subin- hibitory conditions. These findings differed with the increase of enzymes related to folic acid metabolism observed in the ENR experimental group. Folic acid is a necessary precur- sor of purine and pyrimidine synthesis and one-carbon me- tabolism, and plays an important role in nucleic acid meta- bolism (Czeczot., 2008). These alterations suggested that nucleic acid metabolism was also involved in drug re- sistance under these subinhibitory conditions.

Fig.3 GO enrichment of S. aureus DEPs after cultured with 1/2 MIC levels of (A) CIP and (B) ENR respectively.

3.2.5 DEPs functional protein interaction networks

Proteins do not exist independently in organisms but in- teract in complex networks to carry out biological functions. In the current study, we examined protein-protein interac- tions using the STRING database. The majority of the DEPs we identified participated in amino acid, nucleic acid meta- bolic pathways and other pathways related to energy meta- bolism (Figs.5A and B). The results of functional protein interaction networks were consistent with the results of GO and KEEG analyses.

4 Discussion

is a common marine foodborne pathogenic mi- croorganism, while the enterotoxin produced in its propa- gation is a pathogenic factor causing food poisoning. In ad- dition to marine food safety problems, marine foodborne pa- thogenic bacteria may produce cross-contamination through the food chain. Presently, with the heavy use of different drugs, untargeted drugs may cause cross-resistance of mi- croorganisms. We examined the changes of DEPs incells serially passaged in the presence of sub-MIC le- vels of CIP and ENR, respectively. The primary antibacte- rial mechanisms used byagainst long term expo- sure to FQs fall into 3 categories: 1) altering cell membrane permeability (Craft., 2019), 2) enhancement of active efflux pump activity that reduces net drug uptake (Hashi- zume., 2017; Lekshmi., 2018; Costa., 2019; Dashtbani-Roozbehani., 2021), (3) mutations in bac- terial target enzymes such as GyrA and ParC (Lee., 2016; Foster., 2017; Phillips-Jones., 2018). In our study, we examined the actual expression of proteins involved in these processes but used sub-inhibitory levels of CIP and ENR. We found thatcould develop drug resistance even in this subinhibitory environment and the mechanisms were similar to those found in studies us- ing high FQ levels. The quantitative proteomic analysis ap- proach indicated that the majority of induced proteins were the components of the membrane and cell wall, which was also reflected in the expressions of amino acids, nucleic acids and energy metabolic pathways. These were directly comparable to the results from previous studies with a high concentration of drug (Thai., 2017). Our study provides a theoretical basis for further elucidating related alterations in the proteomic profiles of foodborne pathogenic bacteria, providing a theoretical support for risk assessments. Cur- rently, there is a lack of studies on the effect of low drug le- vels that are commonly found in marine food and within marine environments. Our future studies will examine DEPs inexposed to trace drug levels.

Fig.4 KEGG enrichment of S. aureus DEPs after cultured with 1/2 MIC levels of (A) CIP and (B) ENR respectively.

Fig.5 Functional interaction networks of S. aureus DEPs after cultured with 1/2 MIC levels of (A) CIP and (B) ENR respectively.

Acknowledgement

This research was funded by the Professional Innovation and Integration Project of Qingdao University (2020).

Aldred, K. J., Kerns, R. J., and Osheroff, N., 2014. Mechanism of quinolone action and resistance., 53 (10): 1565- 1574, https://doi.org/10.1021/bi5000564.

Costa, S. S., Sobkowiak, B., Parreira, R., Edgeworth, J. D., Vivei- ros, M., Clark, T. G.,., 2019. Genetic diversity of norA, coding for a main efflux pump of., 9: 710, https://doi.org/10.3389/fgene.2018. 00710.

Cox, J., Hein, M. Y., Luber, C. A., Paron, I., Nagaraj, N., and Mann, M., 2014. Accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction, termed MaxLFQ., 13 (9): 2513-2526, https://doi.org/10.1074/mcp.M113.031591.

Craft, K. M., Nguyen, J. M., Berg, L. J., and Townsend, S. D., 2019. Methicillin-resistant(MRSA): Antibio- tic-resistance and the biofilm phenotype., 10 (8): 1231-1241, https://doi.org/10.1039/c9md00044e.

Czeczot, H., 2008. Kwas foliowy w fizjologii i patologii (Folic acid in physiology and pathology)., 62: 405-419.

Dashtbani-Roozbehani, A., and Brown, M. H., 2021. Efflux pump mediated antimicrobial resistance byin health- related environments: Challenges and the quest for inhibition., 10 (12): 1502, https://doi.org/10.3390/antibiotics 10121502.

Dawan, J., and Ahn, J., 2020. Assessment of cross-resistance po- tential to serial antibiotic treatments in antibiotic-resistant., 148: 104478, https://doi.org/10.1016/j.micpath.2020.104478.

Foster, T. J., 2017. Antibiotic resistance in. Current status and future prospects., 41 (3): 430-449, https://doi.org/10.1093/femsre/fux007.

Guo, Y., Song, G., Sun, M., Wang, J., and Wang, Y., 2020. Preva- lence and therapies of antibiotic-resistance in., 10: 107, https://doi.org/10.3389/fcimb.2020.00107.

Hashizume, H., Takahashi, Y., Masuda, T., Ohba, S. I., Ohishi, T., Kawada, M.,., 2017.efficacy of β-lactam/tripro- peptin C in a mouse septicemia model and the mechanism of reverse β-lactam resistance in methicillin-resistantmediated by tripropeptin C., 71: 79-85, https://doi.org/10.1038/ja.2017.88.

Hooper, D. C., and Jacoby, G. A., 2015. Mechanisms of drug re- sistance: Quinolone resistance., 1354 (1): 12-31, https://doi.org/10.1111/nyas. 12830.

Lee, Y. D., and Park, J. H., 2016. Phage conversion for β-lactam antibiotic resistance offrom foods., 26 (2): 263-269, https:// doi.org/10.4014/jmb.1508.08042.

Lekshmi, M., Ammini, P., Adjei, J., Sanford, L. M., Shrestha, U., Kumar, S.,., 2018. Modulation of antimicrobial efflux pumps of the major facilitator superfamily in., 4 (1): 1-18, https://doi.org/10.3934/ microbiol.2018.1.1.

Liu, C., Shen, Y., Yang, M., Chi, K., and Guo, N., 2022. Hazard of staphylococcal enterotoxins in food and promising strategies for natural products against virulence., 70 (8): 2450-2465, https://doi.org/10. 1021/acs.jafc.1c06773.

Mlynarczyk-Bonikowska, B., Kowalewski, C., Krolak-Ulinska, A., and Marusza, W., 2022. Molecular mechanisms of drug re- sistance in., 23 (15): 8088, https://doi.org/10.3390/ijms 23158088.

Phillips-Jones, M. K., and Harding, S. E., 2018. Antimicrobial resistance (AMR) nanomachines-mechanisms for fluoroquino- lone and glycopeptide recognition, efflux and/or deactivation., 10 (2): 347-362, https://doi.org/10.1007/ s12551-018-0404-9.

Qiao, Z., 2021. Epidemiological study of intestinalfrom different sources. Master thesis. Yangzhou University (in Chi- nese).

Shen, Y. B., Shi, X. M., Shen, J. Z., Wang, Y., and Wang, S. L., 2019. Application of whole genome sequencing and bioinfor- matics analysis in the study of bacterial drug resistance., 35 (4): 541-557 (in Chinese with English absract).

Thai, V. C., Lim, T. K., Le, K., Lin, Q., and Nguyen, T., 2017. iTRAQ-based proteome analysis of fluoroquinolone-resistant., 8: 82-89, https://doi.org/10.1016/j.jgar.2016.11.003.

Tretter, E. M., and Berger, J. M., 2012. Mechanisms for defining supercoiling set point of DNA gyrase orthologs: II. The shape of the GyrA subunit C-terminal domain (CTD) is not a sole de- terminant for controlling supercoiling efficiency., 287 (22): 18645-18654, https://doi.org/ 10.1074/jbc.M112.345736.

Wang, G., Song, G., and Xu, Y., 2021. A rapid antimicrobial sus- ceptibility test forusing a broth micro- dilution combined with MALDI TOF MS., 14: 1823-1831, https://doi.org/10.2147/IDR.S305 280.

Xie, Y. Z., Mi, K., Li, J., Chen, Z. X., Peng, H., Ma, C. X.,., 2012. Establishment of two-dimensional electrophoresis assay for proteome analysis ofO157:H7.

, 48 (10): 31-33 (in Chinese with English absract).

Yang, J. J., Cheng, A., Tai, H. M., Chang, L. W., Hsu, M. C., and Sheng, W. H., 2020. Selected mutations by nemonoxacin and fluoroquinolone exposure among relevant gram-positive bac- terial strains in Taiwan., 26 (2): 110- 117, https://doi.org/10.1089/mdr.2019.0048.

Zárate, S. G., Morales, P., ?widerek, K., Bolanos-Garcia, V. M., and Bastida, A., 2019. A molecular modeling approach to iden- tify novel inhibitors of the major facilitator superfamily of ef- flux pump transporters., 8 (1): 25, https://doi.org/10. 3390/antibiotics8010025.

Zhao, S., Xu, W., Jiang, W., Yu, W., Lin, Y., Zhang, T.,., 2010. Regulation of cellular metabolism by protein lysine acetylation., 327 (5968): 1000-1004, https://doi.org/10.1126/science. 1179689.

(July 6, 2022;

August 19, 2022;

January 9, 2023)

? Ocean University of China, Science Press and Springer-Verlag GmbH Germany 2023

. E-mail: hisense_sun@sina.com

(Edited by Qiu Yantao)