Label-free quantitative proteomic analysis of the inhibition effect of Lactobacillus rhamnosus GG on Escherich coli biofilm formation in coculture

Background: Escherich coli ( E.coli ) is the principal pathogen that causes biofilm formation; the latter is associated with infectious diseases and antibiotic resistance. In our previous work, we demonstrated that probiotic microcapsules have superior biofilm inhibition capacity compared to probiotic sterile culture supernatant. Herein, the mechanism of the inhibition effects was investigated using label-free quantitative proteomics analysis. Results: The proteomic analysis characterized a total of 1655 proteins in E.coli K12MG1655 and 1431 proteins in Lactobacillus rhamnosus GG (LGG). Among them, after coculture treatment, there were 262 and differentially expressed proteins that were specific for E.coli and 291 for LGG. The differentially expressed proteins after coculture were related to cellular metabolism, the stress response, transcription, and the cell membrane. In addition, we identified five strain-specific genes in E.coli and LGG, respectively, which were consistent with the proteomics results. Conclusions: These findings indicate that LGG microcapsules may inhibit E.coli biofilm inhibition by disrupting metabolic processes, particularly in relation to energy metabolism and stimulus responses, both of which are critical for the growth of LGG. Together, these findings increase our understanding of the interactions between bacteria under coculture conditions. decreased the expression of secondary metabolism genes, energy metabolism genes, and protein biosynthesis genes in E.coli and LGG strains. These findings further our understanding of the possible molecular action of LGG microcapsules against E.coli biofilm formation. Our future studies will focus on the analysis of posttranslational modifications of differentially expressed proteins as well as endogenous protein complexes and protein-protein interactions.


Applied & Industrial Microbiology
General Microbiology KEYWORDS label-free quantitative proteomics, coculture, Lactobacillus rhamnosus GG microcapsules, mechanism microcapsules showed superior biofilm inhibition capacity compared to probiotic sterile culture supernatant. As a result, we established a Lactobacillus rhamnosus GG (LGG) microcapsule-planktonic Escherich coli (E.coli) coculture model and evaluated the biofilm inhibition effect [5].
Proteomics analysis combined with bioinformatics tools has emerged as an important approach to extract detailed information on cellular regulatory mechanisms at the protein level. In this study, we aimed to investigate the modulatory effects of LGG microcapsules on the E.coli proteome. To this end, a label-free quantitative proteomic approach was used to identify proteins with significantly changed expression profiles, in order to confirm their functions in LGG-treated E.coli. Ultimately, these findings will further our understanding of the possible molecular action of probiotic microcapsules against pathogen biofilm formation and provide a powerful platform for future mechanistic studies of bacterial interactions.

Results
The inhibition effect of LGG microcapsules on E.coli biofilm formation The biofilms formed by coculture and pure culture were stained with SYTO-9 (20 mM) for 15 min at room temperature and were analyzed by confocal microscopy (Leica SP8, Germany) at 480 nm (excitation) (Fig. 1). The observed reduction in the intensity of fluorescence provides evidence of biofilm inhibition.
Global proteomic survey of LGG and E.coli before and after coculture treatment In this study, we evaluated E.coli in the biofilm functional genome at the proteome level in response to LGG microcapsules, before and after coculture. The experiments were performed in biological triplicates. Proteins from total bacterial lysates were extracted and digested in solution. Then, the resulting peptides were analyzed using the LC/MS E approach [6]. A total of 1655 and 1431 bacterial proteins were identified in E.coli and LGG, respectively. Differential expression was considered only for proteins that were significantly different with an ANOVA p-value < 0.05 and with at least a 2-fold change (cut-off value). Based on these parameters, 19 proteins were identified before coculture in E.coli, but not identified after coculture, and 68 proteins were not identified before coculture in E.coli., but identified after coculture. For LGG, these values were 35 and 15, respectively (Fig. 2a).
Hierarchical cluster analysis was performed for the coculture group and the control group (Fig. 2b). As shown in Fig. 2b, the differentially expressed proteins in the coculture group did not cluster together with the ones in the control group.
Functional categorization of proteomes of E.coli and LGG before and after coculture Following the label-free quantification analysis, the functions of the regulated bacterial proteins were enriched according to Gene Ontology (GO) terms, with their redundant GO terms summarized and unified. For E.coli, 262 proteins were differentially expressed after co-culture, among which 31 proteins were up-regulated in the presence of LGG microcapsules and 231 were down-regulated. The number of up-regulated and down-regulated proteins for LGG was 104 and 187, respectively (Table 1). Among the GO terms for cellular component, the most common up-regulated proteins belonged to "cytosol" (33.86%), while the most common down-regulated proteins belonged to "plasma membrane" (12.58%). For molecular function, 13.51% of up-and 11.69% of down-regulated proteins were related to "4 iron, 4 sulfur cluster binding" and "ATPase activity", respectively (Fig. 3).
Functional categorization of LGG proteome before and after coculture A total of 211, 138, and 20 GO terms for molecular function, biological process, and cellular component, respectively, were generated. Furthermore, 85, 50, and 10 GO terms, respectively, belonged to up-regulated proteins in LGG after coculture, while 150, 102, and 15 GO terms, respectively, belonged to down-regulated proteins. Among the GO terms for biological process, "carbohydrate metabolic process" (22.75%) was the most common function in up-regulated proteins in the coculture group. Among the GO terms for cellular component, the most common up-regulated proteins belonged to "cytoplasm" (33.86%), while the most common down-regulated proteins belonged to "ATP binding" (12.58%). For molecular function, 17.92% of up-and 16.55% of downregulated proteins were related to "oxidoreductase activity" and "structural constituent of ribosome", respectively (Fig. 3).
E.coli and LGG affect the mRNA levels of each other's target proteins in the coculture model The proteomics analysis identified differential regulation of several proteins in E.coli and LGG microcapsule coculture (Supplementary Table S1 and S2). We then performed real-time RT-PCR analysis of select targets to validate the observed differential protein levels ( Fig. 4). In line with the findings from the global proteomics analysis, we observed increased bioD2, panD, and ygiW mRNA levels in E.coli. We also quantified the mRNA levels of bamE and dnaK, which encode down-regulated expression at the transcription level. Additionally, given the extensive effect of LGG microcapsules on biofilm inhibition, we determined that the mRNA expression levels of purD and purM were upregulated 3.1-fold and 7.3-fold, respectively. The mRNA levels of murB, murF, and ackA were also decreased after coculture with E.coli for 48 h.

Discussion
Biofilm formation is associated with resistance to antibiotic therapy, and thus, continues to be a major health threat in both hospital and community settings [7] [8]. We previously reported that probiotic LGG microcapsules could inhibit E.coli biofilm formation without causing antibiotic resistance.
Probiotics have many benefits for human health and are used both therapeutically and in the food industry [9]. These findings prompted us to evaluate the mechanism of action underlying the inhibitory effect of LGG microcapsules on antibacterial biofilm formation.
Effects of LGG microcapsules on the growth and metabolism of E.coli in coculture Our proteomic analysis detected a difference between the differentially expressed proteins of E.coli before and after coculture; these differences were related to cellular response to DNA damage stimulus and cell wall organization. Thus, these findings indicate that the proteins involved in responses to the environment have changed during coculture. From this point of view, we focus on the stress response of E.coli in the coculture model. First, we observed increased mRNA levels of the bioD2 gene. The bioD2 protein is an ATP-dependent dethiobiotin synthetase that encodes a homolog of dethiobiotin synthetase, which is a penultimate enzyme in the biotin synthesis pathway. Thus, it is likely that such up-regulated expression in the presence of LGG microcapsules enhanced the degraders requirement for biotin, which is synthesized de novo under the conditions of acid, osmotic, and oxidative stresses with the involvement of different isozymes. This explanation is supported by the up-regulated expression in both the proteomics and RT-PCR analyses, and is further validated by the inhibition effect of LGG microcapsules on E.coli biofilm formation.
Aspartate 1-decarboxylase (PanD) is the only enzyme capable of β-alanine synthesis in E. coli. In bacteria, fungi, and plants, β-alanine is a precursor to pantothenate, which in turn is a required metabolite for the synthesis of coenzyme A (CoA) in all organisms [10]. Research indicates that chloroplast engineering of the beta-alanine pathway by over-expression of E. coli panD enhances thermotolerance of photosynthesis and biomass production following high temperature stress [11].
During the coculture process, E.coli strains are frequently confronted by acid stress produced by LGG metabolism. The ygiW protein is reported to be involved in the stress response associated with exposure to H 2 O 2 , cadmium, and acid [12]. An earlier study also reported that expression of functional YgiW and QseC proteins is necessary for optimal biofilm growth of A.
actinomycetemcomitans [13]. When we compared the expression levels of the ygiW gene between LGG microcapsules and coculture conditions, we observed 4.0-fold change ratios.
The proteomics analysis revealed that several virulence-related proteins, such as bamE and dnaK, were down-regulated when E.coli was treated with LGG microcapsules. The most down-regulated protein, bamE (MHC class II analog protein, Log2FC=-9.2), is an integral outer membrane β-barrel protein (OMP) that is assembled by the beta-barrel assembly machine (Bam) complex in Gramnegative bacteria [14]. Another down-regulated protein, DanK, is an important factor in all three antibiotic-related persister formation pathways. The decreased persistence phenotype, as well as the growth defect of dnaK, seems to depend on functional (p)ppGpp [15]. Heterogeneous expression of the dnaKgene from A. acidoterrestris can significantly enhance the resistance of host bacteria E.
Effects of E.coli on the growth and metabolism of LGG microcapsules in coculture Proteomics analysis indicated that E.coli coculture with LGG microcapsules elicited a cellular response in LGG and E.coli strains which is related to a certain intracellular mechanism. Coculture with LGG microcapsules places environmental stress on E.coli and this in turn raises a cellular response in LGG as well [18]. For LGG, the possible responses to the E.coli coculture include physiological and developmental changes, reprogramming of the resistance gene or proteins, and alterations to the way in which energy is supplemented. Proteomics analysis revealed that coculture with E.coli significantly up-regulated two nucleotide metabolism-related genes, purD and purM. The pur-operon (purEKCSQLFMNHD) is responsible for the catalysis of de novo synthesis of inosine monophosphate (IMP) from phosphoribosyl pyrophosphate [19]. In S. aureus, purine biosynthesis enzymes have been closely implicated in virulence, persistence, and tolerance to stresses such as antibiotic resistance [20] [21]. Such extensive effects could be attributed to the potential modulation of the transcription of the operon by bacteria secreted extracellular compounds. In another study, purD and purF mutants were constructed in macrophage-like RAW264.7 and HeLa cells. The purD and purF mutants showed significantly decreased intracellular survival, and complementation of these mutants with intact copies of purD or purF genes of Brucella abortus strain RB51 restored these defects. These findings suggest that genes encoding the early stages of purine biosynthesis (purD and purF) are required for intracellular survival and virulence of the RB51 strain [22]. Thus, it may be that LGG strains maintain intracellular survival and homeostasis by up-regulating purD and purM genes.
Coculture of LGG microcapsules with E.coli appeared to down-regulate murB gene expression and to completely abolish expression of the murF gene. Bacteria generally synthesize their own active form of N-acetylmuramic acid, UDP-N-acetylmuramic acid [23], and the MurB enzyme (UDP-N-acetylglucosamine pyruvate enol ether reductase) plays an important role in the biosynthesis of this substance [24]. The MurB enzyme converts UDP-N-acetylglucosamine pyruvate enol ether to UDP-Nacetylmuramic acid by reducing the double bond in the enol ether structure [25]. Inhibition of the MurB enzyme reduces or blocks the synthesis of peptidoglycan, resulting in incomplete structure of the bacterial cell wall; this eventually leads to the production of lytic bacteria under the pressure of permeation [26]. Thus, down-regulation of the murB and murF genes implies suppressed LGG cell membrane biosynthesis, to some extent, when cocultured with E.coli.
Metabolism refers to the basic physiological processes that maintain a living organism. Coculture of LGG microcapsules with E.coli was associated with down-regulation of metabolism-related genes.
Acetate presumably provides a relevant nutrient for Enterobacteria was well as other bacteria [27] [28]. In E. coli, the primary pathway of acetate production involves two enzymes that are intimately connected to central metabolism, phosphotransacetylase (Pta) and acetate kinase (AckA) [29]. During exponential growth, acetyl-CoA, the product of glycolysis and the consumable substrate for the tricarboxylic acid (TCA) cycle, can be converted into acetylphosphate (AcP) by Pta and then into acetate by AckA. E. coli also takes up acetate, using the Pta-AckA pathway in reverse, resulting in synthesis of acetyl-CoA. This pathway typically operates at high extracellular acetate concentrations (≥ 8 mM) [30]. Disruption of the Pta-AckA pathway during overflow metabolism causes a significant reduction in the growth rate and viability of the bacteria; although, this is not due to intracellular ATP depletion [31] [32]. Hence, down-regulation of the ackA gene will affect LGG metabolism.

Conclusion
To the best of our knowledge, the present study is the first published attempt to determine protein expression differences associated with probiotics-E.coli in situ coculture. In this work, we applied the label-free proteomics approach to quantify the E.coli proteome in response to LGG microcapsules in coculture. The proteomic analysis revealed quantitative changes in a total of 1655 and 1431 proteins in E.coli and LGG, respectively. Among them, 19 proteins were not identified and 68 proteins were specifically identified after coculture; for LGG, these numbers were 35 and 15, respectively.
Functional characterization of these proteins by GO and KEGG enrichment analysis suggested that coculture of LGG microcapsules with E.coli increased the expression of stress-related genes and decreased the expression of secondary metabolism genes, energy metabolism genes, and protein biosynthesis genes in E.coli and LGG strains. These findings further our understanding of the possible molecular action of LGG microcapsules against E.coli biofilm formation. Our future studies will focus on the analysis of posttranslational modifications of differentially expressed proteins as well as endogenous protein complexes and protein-protein interactions.

Bacterial strains and materials
The Lactobacillus rhamnosus GG and Escherich coli K12MG165 were obtained from American Type Culture Collection (ATCC 53103 and ATCC 47076).
LGG was cultured in a modified MRS broth in which glucose was replaced by galactose with anaerobic at 37℃. E.coli strains were cultured at 37℃ in Luriae Bertani broth [33]. The cell suspension was subsequently used as described below.
Sodium alginate was purchased from the Qingdao Crystal Salt Bioscience and Technology Corporation (Qingdao, Shandong, China). Chitosan was degraded from raw chitosan using the chemical method (Yuhuan Ocean Biomaterials Corporation, China). All other reagents and solvents were analytical reagents and were used without further purification.

Preparation of LGG microcapsules
Preparation of LGG alginate beads using the emulsification/internal gelation technique The LGG microcapsules were prepared using the emulsification/internal gelation method described previously [34]. In general, sodium alginate powder was dissolved in 0.9% (w/v) NaCl solution to obtain the final concentration of 1.5% (w/v). (v/v) Tween 80 solution and distilled water, and were stored in water at 4℃.

Preparation of LGG alginate-chitosan microcapsules
Chitosan solution was dissolved in 0.1 mol/L acetate buffer. The cell entrapped calcium alginate beads were immersed in 0.5% (w/v) chitosan solution by gently shaking, at a bead/solution ratio of 1:5 (v/v).
After rinsing and liquefication for 6 min using 0.055 mol/L sodium citrate, the cell entrapped alginatechitosan microcapsules were formed.

Preparation of cell samples for proteomic analysis
Briefly, LGG microcapsules were cocultured with planktonic E.coli for 48 h for biofilm inhibition, as reported previously [5]. Then, the biofilms were collected by sonication using high intensity focused In order to identify proteins with the abundance being changed significantly after heat treatment, a two-way t-test using multiple hypothesis tests using Benjamini Hochberg correction was applied. Unless otherwise stated, the statistical significance was determined by using a 1% FDR threshold. A second data set is generated by filtering to include proteins detected in at least three replicates of any biological sample in order to detect proteins that are only present at moderately high abundance at one or two specific growth temperatures. By filtering to include proteins detected in at least three replicates of any biological sample, a second dataset was produced, so as to enable detection of proteins existing at moderate or high abundance only at one or two specific growth temperatures. UniProt [39] was used to identify homologs of unidentified proteins and, where appropriate, using BLASTN's protein sequence analysis method, using the Kyoto Encyclopedia of Genes and Genomes (KEGG) [40] and National Biotechnology Information Center (NCBI) to determine putative identity [41].

Bioinformatic analysis of identified proteins and prediction of promoters in intergenic regions
Functional profiling of the identified proteins was performed according to their molecular functions GO category based on the functional annotations of E.coli and LGG [42].
The regulated trends of secreted proteins were visualized with the use of a heat-map constructed

Ethics approval and consent to participate
Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The datasets generated and/or analysed during the current study are available in the [figshare] repository, [10.6084/m9.figshare.11923542]

Competing interests
The authors declare that they have no competing interests.

Funding
The materials used in this manuscript were supported by the National Natural Science

Authors' contributions
Song HY and Shang D designed and planned the study. Lou N and Liu JianJ conducted experiments.
Xiang H contributed analytical tools and analyzed the data. Song HY wrote the manuscript. All authors were involved in writing the manuscript and all made substantial contributions to the content and approved the final manuscript. Figure 1 Confocal sections of biofilms of E.coli before and after coculture.  Subcategories with GO terms less than 1% are classi-fied as "other".

Figure 4
Relative mRNA expression of select targets from global proteomics analysis. Total RNA isolated from coculture treatment or pure culture of the E.coli and LGG was reversed transcribed and cDNAs were quantified by qRT-PCR using target-specific primers. The data represents the mean ± SD of triplicate experiments normalized with 16S RNA. Statistically significant differences between coculture treatment and pure culture-treatment as determined by Student's t-test analysis (unpaired, two-tailed) is represented as *p≤ 0.5, **p≤ 0.1 and ***p≤ 0.01.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.