Metabolomic profiling of Campylobacter jejuni with resistance gene ermB by ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry and tandem quadrupole mass spectrometry

Accepted Manuscript Metabolomic profiling of Campylobacter jejuni with resistance gene ermB by ultra-high performance liquid chromatographyquadrupole time-of-flight mass spectrometry and tandem quadrupole mass spectrometry

Qin Fu, Dejun Liu, Yingyu Wang, Xiaowei Li, Lina Wang, Fugen Yu, Jianzhong Shen, Xi Xia PII: DOI: Reference:

S1570-0232(17)31822-6 https://doi.org/10.1016/j.jchromb.2018.02.009 CHROMB 21030

To appear in: Received date: Revised date: Accepted date:

24 October 2017 7 February 2018 9 February 2018

Please cite this article as: Qin Fu, Dejun Liu, Yingyu Wang, Xiaowei Li, Lina Wang, Fugen Yu, Jianzhong Shen, Xi Xia , Metabolomic profiling of Campylobacter jejuni with resistance gene ermB by ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry and tandem quadrupole mass spectrometry. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Chromb(2017), https://doi.org/10.1016/j.jchromb.2018.02.009

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Metabolomic profiling of Campylobacter jejuni with resistance gene ermB by ultra-high performance liquid chromatography-quadrupole time-of-flight mass

PT

spectrometry and tandem quadrupole mass spectrometry

Beijing Advanced Innovation Center for Food Nutrition and Human Health, College of Veterinary

NU

a

SC

Jianzhong Shena,b*, Xi Xiaa,b*

RI

Qin Fua, Dejun Liua, Yingyu Wanga, Xiaowei Lia,b, Lina Wanga, Fugen Yua,

Medicine, China Agricultural University, Beijing 100193, China Beijing Key Laboratory of Detection Technology for Animal-Derived Food Safety and Beijing

MA

b

AC

CE

PT E

D

Laboratory for Food Quality and Safety, China Agricultural University, Beijing 100193, China

* Corresponding author. Tel: +86-10-62732802. Fax: +86-10-62731201. E-mail: [email protected] (JZ. Shen), [email protected] (X. Xia).

1

ACCEPTED MANUSCRIPT

Abstract The metabolome changes of Campylobacter jejuni with resistant gene ermB remain unclear. Here, we described an untargeted

metabolomic

workflow based on ultra-high performance liquid

chromatography-quadrupole time-of-flight mass spectrometry to investigate the metabolites

PT

perturbations mediated by ermB in C. jejuni. After optimization of extractants and chromatographic

RI

conditions, the combination of 100% methanol extraction with a 12 min gradient by C18 column was

SC

adopted for untargeted metabolomic profiling in reversed phase separation. Meanwhile, 60% methanol extraction followed by a 14 min separation using hydrophilic interaction chromatography column was

NU

suitable to complementally expand the metabolite coverage of C. jejuni. Multivariate statistical analysis

MA

was performed by means of orthogonal projection to latent structures−discriminant analysis to select metabolic features. The selected features were further confirmed by ultra-high performance liquid

D

chromatography-tandem quadrupole mass spectrometry. A total of thirty-six differential metabolites

PT E

between the susceptible strain (C. jejuni NCTC 11168) and resistant stain (C. jejuni NCTC 11168 with ermB) were identified. These pivotal metabolites were primarily participated in biological processes as

CE

cell signaling, membrane integrity/stability, fuel and energy source/storage and nutrient. The biofilm

AC

formation capability of resistant strain was inferior to that of susceptible strain, confirming the influence of ermB on membrane integrity/stability of C. jejuni. Our findings revealed important metabolic regulatory pathways associated with resistant C. jejuni with ermB. Keywords: metabolomics; Campylobacter jejuni; ermB; drug resistance; ultra-high performance liquid chromatography; mass spectrometry

2

ACCEPTED MANUSCRIPT

1. Introduction Campylobacter jejuni, as an epidemic zoonotic pathogenic microorganism, is the most common cause of gastroenteritis and enterocolitis in humans, especially in industrialized countries [1,2]. Besides, it can trigger long-term sequelae such as reactive arthritis and Guillain–Barré syndrom (GBS) through

PT

molecular simulation mechanism, and GBS is related to serious post-infection neurological

RI

complications [3]. An estimated 2.5 million people are infected with this bacteria every year in the

SC

USA and 1% population is infected in Western Europe [4]. Due to the short duration, clinically mild and self-limiting, there is rarely need for clinical treatment of infections. Nevertheless, for

NU

immunocompromised and serious bloody diarrhea patients, treatment by antibacterials is essential [5].

MA

However, antimicrobial resistance in C. jejuni has increased significantly, presenting a major public health threat [6,7].

D

Macrolides are recognized as the effective antibiotics to confront the infection of C. jejuni. The

PT E

resistance to macrolides in C. jejuni was occurred due to the extensive and long-term use of these drugs both in human and veterinary medicine, as well as feed additives for growth promoting purposes [8,9].

CE

As one of the main resistance mechanisms, erm genes encode methyltransferase, which catalyzed

AC

dimethylation of an adenine residue A2058 in 23S ribosomal RNA [10]. This methylation modification was located in the macrolides-binding site, which overlapped with that of two other classes of antibiotics, lincosamides and streptogramins B, conferring resistance to these antibiotics in many bacterial species [11-14]. Although more than 40 types of erm genes have been identified, little is known about the metabolome changes associated with the resistant erm gene expression in the bacteria [15,16]. Characterizing the global metabolite alterations in resistant strain with erm may lead to further understanding of the antibiotic resistance mechanisms.

3

ACCEPTED MANUSCRIPT

In this study, using the resistance gene ermB as a model, we developed an untargeted metabolomic strategy based on ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UHPLC-QTOF/MS) to investigate the metabolites expressed at different levels and associated biofunctions as well as pathways mediated by ermB gene in C. jejuni. The sampling and

PT

extraction methods were optimized for metabolomic profiling. The potential metabolic features

RI

screened out by UHPLC-QTOF/MS were further validated using UHPLC coupled to tandem

SC

quadrupole mass spectrometry (MS/MS). Our findings revealed multiple metabolic pathway changes in the C. jejuni with ermB.

NU

2. Experimental

MA

2.1 Chemicals and reagents

Mueller-Hinton Agar (MHA), ammonium acetate and leucine enkephalin were obtained from

D

Sigma-Aldrich (St. Louis, MO, USA). HPLC grade acetonitrile, methanol, acetone and formic acid

PT E

were purchased from Fisher Scientific Inc. (Pittsburgh, PA, USA). Ultra-pure water was produced by Milli-Q Plus water purification system (Millipore, Bedford, MA, USA). Sodium chloride was obtained

AC

China).

CE

from Beijing Chemical Co. (Beijing, China). Crystal violet stain solution was from Solarbio (Beijing,

2.2 Cell culture and metabolite extraction The parent strain C. jejuni NCTC 11168 (CJ-11168) and resistant strain C. jejuni NCTC 11168 with ermB (CJ-ermB) were both cultivated in MHA plate under the microaerobic environment (85% N 2, 10% CO2, 5% O2) at 42°C. The minimum inhibition concentration of CJ-ermB was 128 µg/mL for erythromycin. Cells at the bacterial exponential growth phase were collected, transferred into 1 mL 0.85% NaCl, and followed by centrifugation at 10000 g, -10°C for 10 min. The supernatants were

4

ACCEPTED MANUSCRIPT

discarded, and the cell pellets were washed two more times with 0.85% NaCl. Twenty mg of cell pellets (wet weight) were prepared as a single sample. Six replicates were prepared for both CJ-11168 and CJ-ermB. Flash freezing in liquid nitrogen (-196°C) was adopted to quench the metabolism and decrease the possibility of cell lysis. The metabolites were extracted with methanol for reversed phase

PT

separation or 60% methanol for hydrophilic interaction chromatography (HILIC) separation by

RI

freezing-thawing cycles. The cell pellet was vortex-mixed with 1 mL of extractant, put in the liquid

SC

nitrogen for 2 min for freezing, and then thawed on the ice. The mixture was centrifuged at 10000 g for 10 min at -10°C, and the supernatant was transferred to another clean centrifuge tube. The extraction

NU

was repeated for three times. Mixed standard solution containing sulfadiazine, difloxacin, monensin,

MA

tylosin, and chloramphenicol was spiked into the combined supernatant for quality control (QC) at a final concentration of 500 ng/mL. The extracts were evaporated to dryness by a vacuum rotation

D

evaporator and dissolved in 500 µL of final solution by 1 min vortexing. The final solution was

PT E

acetonitrile:water (50:50, v/v, containing 0.1% formic acid) for reversed phase separation, or acetonitrile:0.5 mM ammonium acetate (95:5, v/v) for HILIC separation. The reconstituted solution

CE

was centrifuged at 20000 g for 10 min at -10°C before injection into the UHPLC-MS system.

AC

Mixed standard solution containing sulfadiazine, difloxacin, monensin, tylosin, and chloramphenicol was spiked into the individual study samples for quality control (QC) at a final concentration of 500 ng/mL. QC sample was extracted and processed as the above extracting procedures for metabolites. 2.3 Untargeted metabolomic analysis Samples were analyzed using a UHPLC-QTOF/MS (Acquity-Synapt HDMS, Waters, Milford, MA, USA) and separated by reversed phase column and HILIC column, respectively. For the reversed phase separation, a BEH Shield RP18 column (2.1×50 mm i.d., 1.7 µm) was used at 30°C with mobile phase

5

ACCEPTED MANUSCRIPT

A (0.1% formic acid in water) and phase B (0.1% formic acid in acetonitrile). The gradient conditions were optimized as follows: 0−1 min, 2−5% B; 1−8 min, 5−90% B; 8−10 min, 90−100% B; 10−10.1 min, 100−2% B; 10.1-12 min, 2% B. For the HILIC separation, a BEH Amide column (50×2.1 mm i.d., 1.7 µm) was adopted with mobile phase A (95% acetonitrile + 5% 0.5 mM ammonium acetate) and

PT

phase B (50% acetonitrile + 50% 0.5 mM ammonium acetate). The gradient conditions were optimized

RI

as follows: 0−0.5 min, 1% B; 0.5−10 min, 1−100% B; 10−11 min, 100% B; 11−11.1 min, 100-1% B;

SC

11.1−14 min, 1% B. The flow rate was 0.3 mL/min. The injection volume was 10 μL. The MS system was operated separately in electrospray ionization positive (ESI+) and negative (ESI−)

NU

mode with MS resolution of 10000 FWHM for each sample in reversed phase and HILIC separation.

MA

The mass range was set at m/z 100−1000 Da in the full-scan mode. The optimized ESI parameters were as follows: capillary voltage, 3 kV; cone voltage, 35 V; source temperature, 100°C; desolvation

D

temperature, 350°C; desolvation gas flow, 600 L/h. For accurate mass measurement, leucine

PT E

enkephalin was used as the lock spray standard ([M+H] +=556.2771; [M−H]−=554.2615) at a concentration of 100 ng/mL under a flow rate of 50 μL/min.

CE

2.4 Metabolomic data processing

AC

UHPLC-QTOF/MS raw data files were processed with MarkerLynx XS software (Waters). Peak picking parameters were as follows: intensity threshold, 100; mass window, 0.02; retention time (RT) window, 0.1; noise elimination level, 6.0; peak widths automatically detection. Multivariate statistical analysis was performed by EZinfo software (Waters) for orthogonal projection to latent structures−discriminant analysis (OPLS−DA) to maximize the differences of metabolic profiles between parent and resistant strains. The combined S plots and variable importance in the projection (VIP) plots from the OPLS−DA analysis were used to select distinct variables as potential biomarkers

6

ACCEPTED MANUSCRIPT (VIP scores≧2.0, SIMICA-P≦-0.9 or ≧0.9). 2.5 Quantification of metabolite features To verify the reliability of UHPLC-QTOF/MS analysis, relative quantification of metabolite features was performed by UHPLC-triple quadrupole mass spectrometry (MS/MS) (Acquity-Quattro LC,

PT

Waters) with ESI multiple reaction monitoring (MRM) detection. The LC separation conditions were

RI

identical to untargeted profiling. Typical MS source parameters were as follows: capillary voltage, 3.5

SC

kV; source temperature, 100°C; desolvation temperature, 350°C; desolvation gas flow, 600 L/h. The transitions as well as corresponding collision energy of each feature were summarized in

NU

supplementary material Table S1. The fold change (FC) was calculated by comparing the mean

MA

response of each features in CJ-ermB versus that of CJ-11168. The features with FC≧1.5 were selected for further identification.

D

2.6 Metabolites identification

PT E

Identification of selected features was firstly performed by elemental composition analysis. The mass error and isotopic pattern of the selected ion (i-fit) were used to assess the accuracy of potential

CE

formulas. After that, Mass Fragment (Waters) was applied to facilitate the MS/MS fragment ion

AC

analysis. Later, databases query containing Lipdmaps, Markers, Kyoto Encyclopedia of Genes and Genomes (KEGG), Metlin, Chemspider, Human Metabolome Database (HMDB), Massbank was performed with a mass tolerant of 0.01 Da. Finally, the identities of specific metabolites were confirmed by comparing their MS/MS spectra and chromatographic retention behavior with that of commercial standards. 2.7 Biofilm formation assay The biofilm formation estimation was as the modified protocols based on descriptions by Reeser [17],

7

ACCEPTED MANUSCRIPT specifically, 96-well plates were inoculated with 200 μL CJ-11168 and CJ-ermB to an OD600 of 0.025 and incubated at microaerophilic conditions at 37 °C for 18 h, 24 h, 36 h, 48 h and 72 h. The culture medium was removed and the 96-well plates were dried at 55°C for 30 min. After that, 200 μL 0.1% crystal violet was added to the wells and dyed for 5 min at room temperature. The unbound crystal

PT

violet was removed and the wells were washed twice with H2O. The wells were dried at 55°C for 15

RI

min, and bound crystal violet was decolorized with 200 μL 80% ethanol-20% acetone. The absorbance

SC

of solution at 570 nm was determined by a microplate reader (SpectraMax M5) to determine biofilm formation.

NU

3 Results and discussion

MA

3.1 Optimization of metabolites extraction and chromatographic separation Metabolites extraction is a key step in sample preparation of metabolomic profiling, impacting

D

analytical quality and depth of metabolite coverage [18]. Pure methanol was recommended as the

PT E

preferable extractant for metabolites in the previous report, whereas aqueous extracts also showed a significant superiority [19,20]. In our preliminary experiment, the effects of 100% methanol and 60%

CE

methanol as extractant were evaluated by UHPLC-QTOF/MS with reversed phase column. Comparing

AC

the peak picking results, we found that 100% methanol tended to detect more potential features (Fig. 1a). Meanwhile, chromatographic separation also has significant influence on metabolites detection [21]. After assessing the extracted ion chromatograms of most of features, we identified that the peak width was between 0.17 min and 0.25 min and the peak shapes were satisfactory with few tailing, when the gradient conditions ranged from 10 min to 16 min. Then we compared four gradient separation programs (10 min, 12 min, 14 min, and 16 min) and the specific conditions were described in supplementary material Table S2. As shown in Fig. 1b, the 12 min gradient was the optimal program

8

ACCEPTED MANUSCRIPT

for detecting the maximum features. C18 column is commonly used in the separation of medium and weak polar organic compounds. However, strong polar metabolites are not retained well on reversed phase column [22]. In this study, HILIC analysis was performed using a BEH Amide column in parallel with reversed phase separation.

PT

Accordingly, the extractant for HILIC analysis should be optimized for extracting more polar

RI

metabolites. In the literature, 50% methanol was recommended for the metabolites extraction of

SC

leishmania parasites by HILIC analysis [23]. Therefore, we investigated different extractants (20%, 40%, 50%, 60%, 80%, 100% methanol) and gradient separation programs (10 min, 12 min, 14 min, 16

NU

min) for HILIC analysis (supplementary material Table S3). 60% methanol and 14 min gradient

MA

yielded the best result (Fig. 1c and d). According to the final identification results (Table 1), 50% of differential metabolites were respectively detected by C18 column and Amide column, demonstrating

D

the importance of simultaneous application of reversed phase and HILIC separation in metabolomic

PT E

profiling.

3.2 Metabolomic analysis and confirmation

CE

We used the optimized workflow to perform the metabolomic profiling of CJ-11168 and CJ-ermB.

AC

Total ion chromatograms of reversed phase and HILIC separation with positive and negative mode were showed in supplementary material Fig. S1 and Fig. S2. QC results showed that the retention times of all the fortified standards were almost identical and their average mass error was below 10 ppm, with the relative standard deviation (RSD) of the six replicates lower than 6.64%, which proved satisfactory accuracy and repeatability of the untargeted metabolomic profiling (supplementary material Table S4). After raw data processing, the detected ions were transferred to multivariate statistical analysis. Principal Component Analysis was first employed to obtain an overview of the data

9

ACCEPTED MANUSCRIPT

structure, and the results indicated that the susceptible strain and resistant strain could not be separated using such unsupervised analysis strategy (Fig. S3, S4). In order to investigate deeper potential differences between the susceptible and resistant strains, a supervised OPLS-DA was then performed on the basis of all detected ions. The resistant strain was completely separated from susceptible strain

PT

in the scores plots (Fig. 2a, b). From the S plots of OPLS-DA analysis, the ions far away from the

RI

origin were regarded as the differential metabolites (Fig. 2c, d). Sixty-six potential features were

SC

selected from OPLS-DA analysis, according to the criteria described in section 2.4. The selected features were further validated by UHPLC-MS/MS through MRM acquiring. The LC

NU

conditions were completely same as UHPLC-QTOF/MS analysis, so we could easily determine the

MA

retention time of each feature. As for the transition parameters, one precursor ion and two product ions for each feature were also determined on the basis of QTOF/MS spectra. Corresponding cone voltage

D

and collision energy were optimized for individual targeted compound to obtain the best sensitivity. As

PT E

showed in Table 1 and Table S5, forty-eight features, approximately 72% of the selected, were validated regarding variation trend and fold change, indicating the necessity of UHPLC-MS/MS

CE

validation. The coefficient of variance (CV) not exceeding 23% for targeted analysis of compounds is

AC

recommended by European Commission Decision A higher CV value is acceptable for untargeted analysis, but the result with CV>50% is unreliable [24,25]. In this study, the CV of forty-eight validated features was no more than 37%, and 66% of the selected features had a CV≦23%, demonstrating reliability of the quantification results. 3.3 Metabolites identification and pathways analysis As summarized in Table 1, thirty-six metabolites, 75% of the validated features, were identified by combining elemental composition analysis, mass fragment analysis, and databases query results. The

10

ACCEPTED MANUSCRIPT

authentic commercial standards were used for the confirmation of the putative metabolites. Among the identified compounds, five standards were commercial available, and the identities of all the five corresponding putative metabolites were confirmed according to the results of UHPLC-QTOF/MS analysis. The result of myricanone was exhibited as an example shown in Fig. 3. The retention time of

PT

putative metabolite in sample was highly coincident with that of myricanone, and their MS/MS spectra

RI

seemed to parallel very closely.

SC

According to the investigation of the biofunctions and pathways of the differentially expressed metabolites in HMDB and KEGG, these metabolites mainly participated in cell signaling, membrane

NU

integrity/stability, fuel and energy source/storage, and nutrient (Fig. 4a and Table S6). They were

MA

primarily located on the membrane, extracellular and cytoplasm (Fig. 4b and Table S6). It has been speculated that unspecified fitness cost should be paid for the expression of the erm genes. 27 Therefore,

D

we compared the capability of the CJ-11168 and CJ-ermB stains to form biofilm by crystal violet

PT E

staining. As shown in Fig. 5, it is obvious that the biofilm-forming capability of CJ-ermB was inferior to that of CJ-11168, which demonstrated the influence of ermB on membrane integrity/stability. The

CE

association of up-regulated and down-regulated metabolites with ermB resistant mechanism was not

AC

further investigated in this work and should be verified in future studies. 4. Conclusions

In summary, we developed a metabolomics method based on UHPLC-QTOF/MS and UHPLC-MS/MS to characterize C. jejuni wirh ermB. The extractants and chromatographic separation conditions were optimized. After multivariate statistical analysis of raw data and validation by relative quantification, forty-eight differential metabolites between CJ-11168 and CJ-ermB were selected, and thirty-six of them were identified successfully, which were primarily associated with cell signaling, membrane

11

ACCEPTED MANUSCRIPT

integrity/stability, fuel and energy source/storage, and nutrient. The biofilm-formating capabity of CJ-ermB was relatively inferior than CJ-11168. These results gave us new insights into ermB mediated resistance mechanism of C. jejuni. Acknowledgments

PT

This work was supported by the National Natural Science Foundation of China [grant number

AC

CE

PT E

D

MA

NU

SC

RI

31772794, 31530076] and Chinese Universities Scientific Fund [grant number 2017QC188].

12

ACCEPTED MANUSCRIPT References: [1] A.A. Saleha, G.C. Mead, A.L. Ibrahim, Campylobacter jejuni in poultry production and processing in relation to public health, World Poultry Sci. J. 54 (1998) 49-58. [2] J.I. Dasti, A.M. Tareen, R. Lugert, A.E. Zautner, U. Gross, Campylobacter jejuni: A brief overview

PT

on pathogenicity-associated factors and disease-mediating mechanisms, Int. J. Med. Microbiol 300 (2010) 205-211.

RI

[3] I. Nachamkin, B.M. Allos, T. Ho, Campylobacter species and Guillain-Barre syndrome, Clin.

SC

Microbiol. Rev. 11 (1998) 555.

NU

[4] P.S. Mead, L. Slutsker, P.M. Griffin, R.V. Tauxe, Food-related illness and death in the United States - Reply, Emerg. Infect. Dis. 5 (1999) 841-842.

MA

[5] B.M. Allos, Campylobacter jejuni infections: Update on emerging issues and trends, Clin. Infect. Dis. 32 (2001) 1201-1206.

PT E

D

[6] E. Burow, C. Simoneit, B.A. Tenhagen, A. Kaesbohrer, Oral antimicrobials increase antimicrobial resistance in porcine E. Coli - a systematic review, Prev. Ver. Med. 113 (2014) 364-375.

CE

[7] T.H. Akwar, C. Poppe, J. Wilson, R.J. Reid-Smith, M. Dyck, J. Waddington, D. Shang, N. Dassie, S.A. McEwen, Risk factors for antimicrobial resistance among fecal Escherichia coli from residents

AC

on forty-three swine farms, Microb. Drug Resist. 13 (2007) 69-76. [8] V. Bianchini, M. Luini, L. Borella, A. Parisi, R. Jonas, S. Kittl, P. Kuhnert, Genotypes and Antibiotic Resistances of Campylobacter jejuni Isolates from Cattle and Pigeons in Dairy Farms, Int. J. Env. Res. Pun. He. 11 (2014) 7154-7162. [9] S. Pyorala, K.E. Baptiste, B. Cary, E. van Duijkeren, C. Greko, M.A. Moreno, M. Constanca Matias Ferreira Pomba, M. Rantala, M. Ruzauskas, P. Sanders, E.J. Threlfall, J. Torren-Edo, K. Torneke, Macrolides and lincosamides in cattle and pigs: Use and development of antimicrobial resistance, 13

ACCEPTED MANUSCRIPT

Vet. J. 200 (2014) 230-239. [10] S. Qin, Y. Wang, Q. Zhang, M. Zhang, F. Deng, Z. Shen, C. Wu, S. Wang, J. Zhang, J. Shen, Report of ribosomal RNA methylase gene erm(B) in multidrug-resistant Campylobacter coli, J. Antimicrob. Chemoth. 69 (2014) 964-968. Saribas,

F.

Tunckanat,

A.

Pinar,

Prevalence

of

erm

genes

encoding

PT

[11] Z.

RI

macrolide-lincosamide-streptogramin (MLS) resistance among clinical isolates of Staphylococcus

SC

aureus in a Turkish university hospital, Clin. Microbiol. Infec. 12 (2006) 797-799. [12] Y. Min, E. Yoon, A. Kwon, M. Shim, E. Choi, Alterations in regulatory regions of erm(B) genes

NU

from clinical isolates of enterococci resistant to telithromycin, Arch. Pharm. Res. 34 (2011)

MA

2149-2154.

[13] S. Hung, S. Wang, C. Tu, Y. Tsai, S. Chuang, M. Shieh, P. Liu, W. Wang, Antibiotic susceptibility

PT E

176 (2008) 197-204.

D

and prevalence of erythromycin ribosomal methylase gene, erm(B) in Streptococcus spp., Vet. J.

[14] F. Deng, J. Shen, M. Zhang, C. Wu, Q. Zhang, Y. Wang, Constitutive and inducible expression of

CE

the rRNA methylase gene erm(B) in campylobacter, Antimicrob. Agents Ch. 59 (2015) 6661-6664.

AC

[15] S. Wendlandt, S. HeSS, J. Li, A.T. FeSSler, Y. Wang, K. Kadlec, C. Gallert, S. Schwarz, Detection of the macrolide-lincosamide-streptogramin B resistance gene erm(44) and a novel erm(44) variant in staphylococci from aquatic environments, FEMS Microbiol. Ecol. 91 (2015) v90. [16] L.B. Jensen, N. Frimodt-Moller, F.M. Aarestrup, Presence of erm gene classes in Gram-positive bacteria of animal and human origin in Denmark, FEMS Microbiol. Lett. 170 (1999) 151-158. [17] R.J. Reeser, R.T. Medler, S.J. Billington, B.H. Jost, L.A. Joens, Characterization of Campylobacter jejuni biofilms under defined growth conditions, Appl. Environ. Microb. 73 (2007) 1908-1913.

14

ACCEPTED MANUSCRIPT

[18] D. Vuckovic, Current trends and challenges in sample preparation for global metabolomics using liquid chromatography-mass spectrometry, Anal. Bioanal. Chem. 403 (2012) 1523-1548. [19] C.L. Winder, W.B. Dunn, S. Schuler, D. Broadhurst, R. Jarvis, G.M. Stephens, R. Goodacre, Global metabolic profiling of Escherichia coli cultures: An evaluation of methods for quenching and

PT

extraction of intracellular metabolites, Anal. Chem. 80 (2008) 2939-2948.

RI

[20] A.O. Nwadinigwe, Antimicrobial activities of methanol and aqueous extracts of the stem of

SC

Bryophyllum pinnatum Kurz (Crassulaceae), Afr. J. Biotechnol. 10 (2011) 16342-16346. [21] G. Paglia, M. Magnusdottir, S. Thorlacius, O.E. Sigurjonsson, S. Gudmundsson, B.O. Palsson, I.

NU

Thiele, Intracellular metabolite profiling of platelets: Evaluation of extraction processes and

MA

chromatographic strategies, J. Chromatogr. B 898 (2012) 111-120. [22] P. Chocholous, J. Vackova, I. Sramkova, D. Satinsky, P. Solich, Advantages of core-shell particle

D

columns in Sequential Injection Chromatography for determination of phenolic acids, Talanta 103

PT E

(2013) 221-227.

[23] R. Zhang, D.G. Watson, L. Wang, G.D. Westrop, G.H. Coombs, T. Zhang, Evaluation of mobile

CE

phase characteristics on three zwitterionic columns in hydrophilic interaction liquid

AC

chromatography mode for liquid chromatography-high resolution mass spectrometry based untargeted metabolite profiling of Leishmania parasites, J. Chromatogr. A 1362 (2014) 168-179. [24] J.C. Schoeman, I. du Preez, D.T. Loots, A comparison of four sputum pre-extraction preparation methods for identifying and characterising Mycobacterium tuberculosis using GCxGC-TOFMS metabolomics, J. Microbiol. Meth. 91 (2012) 301-311. [25] L. Venter, P.J. van Rensburg, D.T. Loots, A. Vosloo, J.Z. Lindeque, Untargeted metabolite profiling of abalone using gas chromatography mass spectrometry, Food Anal. Method. 9 (2016) 1254-1261.

15

ACCEPTED MANUSCRIPT

Figure captions Fig 1. Optimization of extractants and chromatographic separation. Number of detected features extracted with different extractants (a) and different gradients (b) in RPLC, with different extractants (c) and different gradients (d) in HILIC. The optimal protocol was represented by asterisk.

PT

Fig 2. Representative multivariate statistical analysis results of CJ-11168 and CJ-ermB by reversed

RI

phase separation. OPLS-DA scores plots of positive (a) and negative (b) ionization datasets for

SC

CJ-11168 (red) and CJ-ermB (green). OPLS-DA S-plots illustrating detected features in positive (c) and negative (d) mode.

NU

Fig 3. Confirmation of putative metabolite by reference standard. Chromatogram of putative metabolite

MA

(RT 8.61 min) in the extract of C. jejuni (a) and reference standard myricanone (b, RT 8.64 min). MS spectrum of putative metabolite in the extract of C. jejuni (c) and reference standard myricanone (d).

D

Fig 4. Biofunctions (a) and cellular locations (b) of differentially expressed metabolites.

PT E

Fig 5. Capability of biofilm formation of CJ-11168 and CJ-ermB. Absorbance at 570 nm was

AC

CE

determined for assessing the biofilm-forming capability at 5 time points.

16

ACCEPTED MANUSCRIPT

Table 1. Putative identities of differential metabolites in C. jejuni with ermB Variation

ID

RT

Ions (Da)

Putative identification

Detection

T P

(min) up-regulated

01 02 03 04 05 06 07 08

9.18 9.61 7.92 8.66 9.04 9.11 9.19 3.17

338.3420 713.5191 327.1265 381.1726 395.1877 443.2642 311.1676 640.3115

I R

1-(1-Pyrrolidinyl)-1-octadecanone DG(22:6n3/0:0/22:6n3)

8,8-Dimethyl-2-oxo-7,8-dihydro-2H,6H-pyrano[3,2-g]chromen-7-yl (2Z)-2-methyl-2-butenoate

SC

2,2'-[3-(2-Methoxyphenyl)-1,2-propanediyl]bis(1H-benzimidazole)

U N

7-[(6-Hydroxy-2,5,5,8a-tetramethyl-4-oxo-1,4,4a,5,6,7,8,8a-octahydro-1-naphthalenyl)methoxy]-2H-chromen-2-one 2,4,6-Trihydroxy-5-[1-(3-hydroxy-1,1,5-trimethyldecahydrocyclopropa[e]inden-5-yl)-3-methylbutyl]isophthalaldehyde 4-Undecylbenzenesulfonic acid

A M

(7Z,10S,11S,12R,13S,14R,15R,16S,17S,18E,20Z)-4,10,12,14,16-Pentahydroxy-9-(hydroxymethyl)-3,7,11,13,15,17,21-hep

mode

Fold change (n=6) b mean

C18-ESI+

CV (%)

10.65

16

C18-ESI

+

2.76

9

C18-ESI

-

5.59

23

C18-ESI

-

3.67

21

C18-ESI

-

3.61

30

C18-ESI

-

2.80

28

C18-ESI

-

2.12

19

+

5.29

21

HILIC-ESI+

4.25

11

HILIC-ESI

+

2.10

17

+

HILIC-ESI

tamethyl-23-azatricyclo[22.3.1.0~5,27~]octacosa-1(27),2,4,7,18,20,24-heptaene-6,22,26,28-tetrone 09 10

5.22

123.0899 138.0533

D E

2-Ethyl-5-methylpyrazine a 2-Aminobenzoic acid

PT

11

5.15

268.1658

(4E)-3-(Benzylamino)-5-phenyl-4-penten-1-ol

HILIC-ESI

1.50

36

12

3.62

276.9881

2,4-cyclodiphosphate

HILIC-ESI-

2.92

25

13

4.70

678.0945

UDP-N-acetylmuramate

HILIC-ESI-

1.81

10

-

14 Down-regulated

5.14

15 16 17 18

5.22 3.93 4.90 8.79 7.32

322.0439

E C

Cytidine 5′-monophosphate

C A

389.2175 414.2169 256.2639 227.2004

a

HILIC-ESI

Methyl9-hydroperoxy-9-{5'-[(1E)-1-propen-1-yl]-3,3'-bi-1,2-dioxol-5-yl}nonanoate Delanzomib

a

Palmitamide (5E)-5-Tetradecenoic acid

1.56

21

C18-ESI

+

12.70

26

C18-ESI

+

6.94

20

C18-ESI

+

2.47

19

C18-ESI

+

2.09

27

+

1.67

15

1.54

10

19

4.91

339.2316

(2Z)-5-(2,3-Dimethyltricyclo[2.2.1.0~2,6~]hept-3-yl)-2-methyl-2-penten-1-yl phenylacetate

C18-ESI

20

7.13

589.3942

4-((4,6-Bis(octylthio)-1,3,5-triazin-2-yl)amino)-2,6-di-tert-butylphenol a

C18-ESI+

17

ACCEPTED MANUSCRIPT

21

8.65

379.1787

2,6,9-Trihydroxy-11-(1-hydroxy-2-propanyl)-1,5,10-trimethyl-8-oxatetracyclo[7.4.1.1~7,10~.0~2,7~]pentadec-11-ene-1

C18-ESI-

6.03

25

C18-ESI-

5.52

34

C18-ESI-

2.46

25

C18-ESI

-

2.19

7

C18-ESI

-

3,15-dione 22

7.88

474.193

23

5.75

455.2425

(2S,3S,5R,6R)-3-{[2-({N-[(2R)-2,4-Dihydroxy-3,3-dimethylbutanoyl]-beta-alanyl}amino)ethyl]sulfanyl}-6-(1-hydroxyethyl)

T P

-7-oxo-1-azabicyclo[3.2.0]heptane-2-carboxylic acid

24 25 26 27 28 29 30 31

6.48 8.59 5.15 5.16 5.14 2.02 5.15 5.14

509.2879 355.1579 286.1766 330.1657 352.149 193.171 302.1712 266.1517

GPGro(14:0/0:0)[U]

I R

GPGro(18:1(9E)/0:0)[U] Myricanone

a

C S U

L-Leucylglycyl-L-proline Reticuline

Ethanesulfonic acid - 1-isobutyl-1H-imidazo[4,5-c][1,5]naphthyridin-4-amine (1:1)

N A

3-Pentyl-4,5,6,7-tetrahydro-1H-indazole

(2E,6E)-7-(1,3-Benzodioxol-5-yl)-N-isobutyl-2,6-heptadienamide Nordoxepin

M

1.51

12

HILIC-ESI

+

7.86

23

HILIC-ESI

+

5.44

9

HILIC-ESI

+

3.85

12

HILIC-ESI

+

3.22

19

HILIC-ESI

+

2.44

17

HILIC-ESI

+

2.36

23

+

32

5.15

312.1572

3-Hydroxy-6-methoxy-1-methyl-4,5-diphenyl-2-piperidinone

HILIC-ESI

2.27

19

33

5.31

316.1499

(S)-3-Hydroxy-N-methylcoclaurine

HILIC-ESI+

2.13

34

HILIC-ESI

+

1.81

7

HILIC-ESI

-

4.92

27

HILIC-ESI

-

1.98

13

34 35 36

3.57 3.7 3.72

159.0753 344.1342 179.0562

a

metabolites validated by standards.

b

determined by LC-MS/MS.

5-hydroxyectoine

PT

Gevotroline hydrochloride Inositol

D E

E C

C A

18

ACCEPTED MANUSCRIPT Highlights (1) Optimized workflows were developed for metabolomic profiling of C. jejuni. (2) 36 differential metabolites were identified between CJ-11168 and CJ-ermB. (3) Identified metabolites were associated with signaling, membrane, energy and

AC

CE

PT E

D

MA

NU

SC

RI

PT

nutrient.

19

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5