Long-chain fatty acids alter transcription of Helicobacter pylori virulence and regulatory genes

Infection with Helicobacter pylori is one of the most important risk factors for developing gastric cancer (GC). The type IV secretion system (T4SS) encoded in the cag pathogenicity island is the main virulence factor of H. pylori associated with GC. Additionally, other virulence factors have been shown to play a role in the H. pylori virulence, such as vacuolizing cytotoxin (VacA), urease, flagella, and adhesins. Long-chain fatty acids (LCFAs) are signaling molecules that affect the transcription of virulence genes in several pathogenic bacteria such as Salmonella enterica, Vibrio cholerae, Pseudomonas aeruginosa and Mycobacterium tuberculosis. However, the effect of LCFAs on the transcription of H. pylori virulence and regulatory genes remains unknown. Here we analyzed whether the transcription of virulence genes that encode T4SS and cellular envelope components, flagellins, adhesins, toxins, urease, as well as the transcription of different regulatory genes of the H. pylori strain 26695, are altered by the presence of five distinct LCFAs: palmitic, stearic, oleic, linoleic, and linolenic acids. Palmitic and oleic acids up-regulated the transcription of most of the virulence genes tested, including cagL, cagM, flaB, sabA, mraY and vacA, as well as that of the genes encoding the transcriptional regulators NikR, Fur, CheY, ArsR, FlgR, HspR, HsrA, Hup, and CrdR. In contrast, the other LCFAs differentially affected the transcription of the virulence and regulatory genes assessed. Our data show that LCFAs can act as signaling molecules that control the transcription of the H. pylori virulome.


INTRODUCTION
Helicobacter pylori is a Gram-negative microaerophilic bacterium that colonizes the stomach mucosa during the entire life of many humans. H. pylori has a spiral shape and signaling molecules in both eukaryotic and prokaryotic cells. In this sense, the pathogenic bacteria have developed signal transduction mechanisms able to sense LCFAs to control the expression of their virulence genes. In Salmonella enterica serotype Typhimurium, oleic, myristic and palmitic acids repressed the expression of the genes located in the pathogenicity island 1 (SPI-1), which encodes a Type III secretion system (T3SS) that is necessary for the invasion of intestinal epithelial cells (Golubeva et al., 2016). Arachidonic, oleic and linoleic acids suppress the production of cholera toxin of Vibrio cholerae (Chatterjee, Dutta & Chowdhury, 2007). LCFAs are important for the replication of Pseudomonas aeruginosa in a murine model of lung infection (Kang et al., 2010). In the case of Mycobacterium tuberculosis, the LCFAs affect the expression of some virulence genes, specifically those related with the granuloma formation (Ehrt & Schnappinger, 2007;Rodriguez et al., 2014). In addition to palmitic (16:0), stearic (18:0), oleic (18:1), and linoleic (18:2) acids, we analyzed the role of linolenic (18:3) acid, which is an essential fatty acid for humans (together to linoleic acid) and presents three double C-C bonds. Nevertheless, the effect of these LCFAs on the H. pylori virulence and/or regulatory genes has not been reported.
In this study, the transcription of H. pylori virulence genes encoding T4SS components, adhesins, urease, flagella, peptidoglycan, and LPS, as well as that of several regulatory genes, was quantified in the presence of palmitic, stearic, oleic, linoleic, and linolenic acids. Our results showed that both palmitic and oleic acids increase the transcription of most virulence genes tested, which seems to be an effect of the upregulation of different transcription factors by these LCFAs. In contrast, linoleic acid diminished the transcription of some virulence genes such as flaB, cagL, ureB and babA. Our work demonstrated that LCFAs can act as signaling molecules that control the expression of H. pylori virulence factors.

Bacterial strains and culture conditions
H. pylori strain 26695 was grown as was previously described (De la Cruz et al., 2017). Cultures were incubated for 1 h in the presence or not of individual LCFAs at a final concentration of 250 µM. 0.05% tyloxapol was added as surfactant. Fold-changes in transcription was determined by calculating the relative expression of virulence genes in the presence or not of the distinct LCFAs. On three different days, assays were performed in triplicate and the data shown are the mean of the results obtained.

RNA isolation and quantitative RT-PCR
Extraction of total RNA was performed as previously described (Jahn, Charkowski & Willis, 2008;De la Cruz et al., 2017). cDNA was synthesized using 1 µg of RNA, 0.2 µg/µL of random hexamer primers and 2 U/µL of M-MulV-RT (Reverse transcriptase of Moloney Murine Leukemia Virus; Thermo Scientific). Quantitative real-time PCR was performed in a LightCycler 480 instrument (Roche) to quantify the gene expression levels. For LightCycler reactions, a master mix of the following components was prepared: 3.0 µL of PCR-grade water, 1.0 µL (10 µM) of forward primer, 1.0 µL (10 µM) of reverse primer, 10 µL of 2x Master Mix and 5.0 µL of cDNA (50-100 ng). A multiwell plate containing all samples was loaded into the LightCycler 480 instrument. Amplification was performed in triplicate wells for each sample analyzed from three independent experiments. In each set of reactions, 16S rRNA (rrnA16S) was used as a reference gene for normalization of cDNA amount. Real-time PCR analysis was performed using the following optimized assay conditions: (1) denaturation program (95 • C for 10 min); amplification and quantification program repeated for 45 cycles (95 • C for 10 s, 58 • C/59 • C (depending Tm; see Table 1) for 20 s, 72 • C for 30 s with a single fluorescence measurement); (2) melting curve program (95 • C for 10 s, 65 • C for 1 min with continuous fluorescence measurement at 97 • C) and finally; (3) a cooling step at 40 • C for 10 s. Contamination with DNA was verified in all reactions by running an RT-minus (after 45 qPCR cycles). 2 − Ct method was used to calculate the relative gene expression (Livak & Schmittgen, 2001).

Quantification of IL-8 secretion
Quantification of IL-8 was performed as previously described (Vazquez-Jimenez et al., 2016). AGS gastric epithelial cells were infected at a multiplicity of infection (MOI) of 100 with H. pylori grown at exponential phase in BB + 10% FBS (OD 600nm = 0.8) under the presence of individual LCFAs (250 µM). After 6 h of infection at 37 • C under microaerophilic conditions, the culture supernatant was collected and stored at −80 • C until tested. IL-8 concentration was quantified in each supernatant using a commercial enzyme-linked immunosorbent assay for human IL-8 (BD Biosciences, San Diego, CA, USA), following the manufacturer's instructions. The supernatant of uninfected AGS cells was tested as negative control. Quantifications of IL-8 were performed in triplicate on three different days

Cell adherence assays
Cultured AGS gastric epithelial cells were used in adherence assays as described previously (De la Cruz et al., 2017). As inocula, cultures of H. pylori at exponential phase in BB + 10% FBS (OD 600nm = 0.8) under the presence of individual LCFAs (250 µM) were used. The cells were infected at a multiplicity of infection (MOI) of 100 for 6 h at 37 • C under microaerophilic conditions, washed thrice with PBS to remove unbound bacteria, and subsequently treated with 1 mL of 0.1% TritonX-100 for 15 min. Following the lysis of AGS cells, CFUs were quantified by plating out 10-fold dilutions of the bacterial suspensions. On three different days, quantifications were performed in triplicate, and the mean results were expressed as adhering CFUs/mL.

Biofilm formation
Biofilm formation on abiotic surface (polystyrene) was analyzed using 96-well polystyrene plates as described previously (Hathroubi, Zerebinski & Ottemann, 2018). H. pylori cultures were diluted to an OD 600 of 0.15 in Brucella broth supplemented with 2% decomplemented fetal bovine serum in the presence of individual LCFA (250 µM). The plates were incubated for three days at 37 • C under microaerophilic conditions. After incubation, the supernatant was discarded and the wells were rinsed thrice with PBS. The biofilm was stained with 0.1% Crystal Violet (Merck). After washing with PBS, the adsorbed dye was recovered with 70% ethanol and quantified at an OD 590 with a spectrophotometer (Multiskan Ascent; Thermo Scientific). On three different days, quantifications of the biofilm formation were performed in triplicate.

Heatmap construction
Heatmapper web server (http://www.heatmapper.ca/expression/) was used to illustrate the gene expression changes between the untreated control and different LCFA treatment groups.

Statistical analysis
For statistical differences, one-way ANOVA followed by the Tukey's comparison test was performed using GraphPad Prism (version 6.0c for Mac OS X), La Jolla, California, USA. p < 0.05 was considered statistically significant.

Effect of LCFAs on the expression of genes encoding T4SS components
To determine the effect of LCFAs on the expression of the H. pylori T4SS, the transcription of the cagA, cagL, and cagM genes was tested by RT-qPCR in the presence of the palmitic, stearic, oleic, linoleic, and linolenic acids. The transcription of both cagL and cagM genes, which encode the pilus structure and a transmembrane protein of the T4SS, respectively, significantly induced in the presence of palmitic, oleic and linolenic acids (Fig. 1A). Interestingly, linoleic acid exerted a positive and negative effect on the transcription of cagM and cagL, respectively (Fig. 1A). In contrast, the transcription of the cagA gene, which encodes the oncogenic cytotoxin translocated into host cells through the T4SS, was not significantly affected by any of the LCFA tested (Fig. 1A). Interestingly, the secretion of IL-8 from AGS cells, which requires the T4SS, was significantly boosted only in response to H. pylori grown in the presence of palmitic acid, but not the other LCFAs tested (Fig. 1B). Our data showed that palmitic, oleic and linolenic acids enhance the transcription of T4SS structural components of H. pylori. Intriguingly, we found that only the palmitic acid induces the IL-8 secretion from gastric epithelial cells mediated by H. pylori.

Effect of LCFAs on the expression of genes encoding flagellins and adhesins
To determine whether LCFAs affect the motility and adherence of H. pylori, we analyzed the transcription of flagellar and adhesins genes in the presence of the palmitic, stearic, oleic, linoleic, and linolenic acids. Both flaA and flaB flagellar genes, which encode the major and minor flagellin, respectively, as well as the sabA, babA and hopQ adhesin genes were assessed. The transcription of flaB was increased 24 and 3-fold in the presence of palmitic and oleic acids, respectively, and it was not affected by the other LCFAs tested; in contrast, none of the LCFAs assessed modified the transcription of flaA ( Fig. 2A). In consistent with its highly inductive effect on flaB gene expression, the palmitic acid also significantly (p < 0.001) enhanced the H. pylori motility on soft agar (Fig. 2B). On another hand, the transcription of sabA was significantly increased in the presence of most LCFAs tested; palmitic and oleic acids induced up to 15-fold the transcription level of this gene (Fig. 2C). In agreement with these expression data, both palmitic and oleic acids increased 3-fold the adherence capacity of H. pylori to AGS cells (Fig. 2D); LCFAs evaluated did not affect the transcription of both babA and hopQ genes, with the exception of linoleic acid that reduced 2.5-fold the transcription level of babA (Fig. 2C). In addition, none of the LCFA tested affected the biofilm formation by H. pylori (Fig. 2E), showing that palmitic and oleic acids did not affect the adherence of H. pylori to an abiotic surface. Hence, our results support that palmitic and oleic acids enhance the motility and adherence of H. pylori by upregulating the transcription of the flaB and sabA genes, respectively.

Palmitic and oleic acids stimulate the transcription of the mraY gene for peptidoglycan synthesis
In order to determine the effect of LCFAs on the expression of H. pylori cell envelope components, we quantified the transcription level of mraY, pbp1, and waaL genes, which encode phospho-N-acetylmuramoyl-pentapeptide-transferase, penicillin-binding protein 1A, and lipid A core-O-antigen ligase, respectively. The transcription level of mraY gene was increased 13-and 6-fold in the presence of the palmitic and oleic acids, respectively, whereas it was not affected by the other LCFAs tested (Fig. 3). In contrast, none of the LCFAs evaluated significantly affected the transcription of pbp1 and waaL (Fig. 3). Thus, these results indicate that the palmitic and oleic acids increase the transcription level of the mraY gene required for the peptidoglycan synthesis.

Effect of LCFAs on the transcription of the vacA and ureAB genes
In H. pylori, VacA cytotoxin and urease enzyme are required for the eukaryotic cell vacuolization and bacterial survival under acidic pH, respectively (Mobley, 1996;Palframan, Kwok & Gabriel, 2012). To determine whether LCFAs affect these virulence phenotypes of H. pylori, we quantified the transcription level of the vacA, ureA, and ureB genes. The transcription of vacA was significantly boosted by two-fold (p < 0.01) in the presence of palmitic and oleic acids but not affected by the other LCFAs tested (Fig. 4A). While the transcription level of ureA was unaffected by any of the LCFAs tested, a significant two-fold down-regulated gene expression was observed in ureB when H. pylori was treated with linoleic acid (Fig. 4B).

LCFAs differentially regulate the transcription of H. pylori regulatory genes
H. pylori genome encodes a limited number of transcriptional regulators, which control the expression of genes involved in bacterial survival, metabolism and pathogenicity. We determined the effect of the palmitic, stearic, oleic, linoleic, and linolenic acids on the transcription of genes encoding several of these transcriptional factors. With the exception of hrcA, the rest of the regulatory genes assessed (nikR, fur, cheY, arsR, flgR, hspR, hsrA, hup and crdR) showed increased transcription levels in the presence of the palmitic and oleic acids, while the linoleic and linolenic acids up or down-regulated the transcription of several of these genes such as fur, hrcA, arsR, flgR, hspR, hup, and crdR (Fig. 5). Therefore, these results reveal that the LCFAs tested exert a differential regulatory effect on genes encoding several transcriptional regulators of H. pylori.

DISCUSSION
H. pylori is highly adapted to one specific niche: the human gastric mucosa. Our group has described that environmental cues such as acidic pH, iron, nickel or copper are crucial to control the transcription of H. pylori virulence genes (Cardenas-Mondragon et al., 2016;De la Cruz et al., 2017). In this context, some reports indicate that LCFAs can act as signaling molecules, altering the transcription of genes that encode virulence factors in different pathogenic bacteria (Chatterjee, Dutta & Chowdhury, 2007;Ehrt & Schnappinger, 2007;Golubeva et al., 2016;Kang et al., 2010). LCFAs are mainly found in both prokaryotic and eukaryotic membrane phospholipids, LPS lipid A, bile salts and the daily diet. Palmitic, stearic, oleic, linoleic, and arachidonic acids are the main LCFAs that form the eukaryotic membrane phospholipids, which can be synthesized de novo or modified by deacylation/reacylation reactions (Hulbert, Kelly & Abbott, 2014;Schmid, Deli & Schmid, 1995). In this work we evaluated the effect of the palmitic, stearic, oleic, linoleic, linolenic acids on the transcription of genes encoding major virulence factors and different transcriptional regulators of H. pylori. The palmitic and oleic acids were the two main LCFAs that mostly up-regulated the virulence genes assessed, such as flaB, sabA, cagM, cagL, vacA and mraY (Fig. 6A), which are involved in motility, adherence, and cell wall synthesis. The palmitic and oleic acids also stimulated the transcription of almost all the genes encoding transcriptional factors tested (Fig. 6B). On another hand, the linoleic and linolenic acids similarly affected two clusters of regulatory genes: nikR, fur, cheY, and hrcA were down-regulated, whereas arsR, flgR, hspR, hsrA, hup, and crdR were up-regulated by these LCFAs. Some LCFAs have been shown to affect the function of transcriptional factors of different families. HilD and ToxT are two AraC-type transcription factors that act as master regulators of virulence genes in Salmonella and Vibrio, respectively (Ellermeier & Slauch, 2007;Weber & Klose, 2011). The palmitic and oleic acids repress the expression of hilD and toxT ; moreover, the oleic acid directly binds to and thus block the DNA-binding activity of HilD and ToxT (Cruite et al., 2019;Golubeva et al., 2016). Besides, the palmitic and oleic acids reduce the DNA-binding affinity of TetR-type regulators Rv3249c/Rv1816 and TfmR of M. tuberculosis and Xanthomonas citri, respectively (Delmar et al., 2015;Teper, Zhang & Wang, 2019). Therefore, it is reasonable to propose that the palmitic and oleic acids also interact with some positive or negative transcriptional regulators of H. pylori and thus affect the expression of virulence genes in this bacterium.
With respect to the effect of LCFAs on the expression of cag PAI, the palmitic, oleic and linolenic acids activated the transcription of the cagL and cagM genes, which encode the T4SS structural components, whilst the transcription of the cagA cytotoxin gene was not significantly affected by any of the LCFAs tested. In addition, IL-8 secretion from AGS cells was stimulated when H. pylori grew in the presence of the palmitic acid. Our results suggest that the palmitic acid could be a signaling molecule involved in CagA-independent stimulation of gastric mucosa inflammation. In agreement with this notion, it has been shown that CagL is able to stimulate the secretion of IL-8 independently of CagA translocation (Gorrell et al., 2013). This CagA-independent inflammation could be supported by the low association between dyspepsia and CagA+ H. pylori strains observed in some worldwide patient cohorts (Andreolla et al., 2016;Bommelaer et al., 2001;Heikkinen et al., 1998;Nordenstedt et al., 2007;Vasiliou et al., 2014). The positive effects of the palmitic acid on the sabA transcription and on the adherence of H. pylori to AGS cells would enhance the inflammatory process, since SabA is an adhesin required for a strong interaction between the bacterium and gastric epithelial cells, which promotes the inflammation of the gastric mucosa (Mahdavi et al., 2002).
Although flagella and peptidoglycan have been reported as structures implicated in the biofilm formation by H. pylori (Hathroubi, Zerebinski & Ottemann, 2018), the presence of the palmitic acid, which up-regulates the transcription of flaB and mraY genes, encoding the FlaB flagellin and MraY enzyme related to peptidoglycan biosynthesis, respectively, did not affected the biofilm formation by H. pylori. These observations suggest that multiple factors are involved in the biofilm formation by H. pylori in response to different environmental cues.
Contrary to palmitic and oleic acids, the linoleic acid down-regulated the transcription of virulence genes such as flaB, ureB, babA, and cagL, and also that of the fur and hrcA regulatory genes. In V. cholerae, the oleic, linoleic and arachidonic acids repressed the expression of the ctxAB and tcpA genes, which encode cholera toxin and TCP type IV pilus, respectively (Chatterjee, Dutta & Chowdhury, 2007). Interestingly, the linoleic acid also increased the motility of V. cholerae, stimulating the expression of flrA gene, which encodes the top transcriptional regulator that controls the motility in this bacterium (Chatterjee, Dutta & Chowdhury, 2007). This dual effect on virulence or regulatory genes was also observed for some LCFAs in our study. For instance, the linolenic acid enhanced the transcription of the sabA, arsR, hsrA, hup and crdR regulatory genes, but decreased the transcription of the hrcA and fur regulatory genes.
Our data support regulatory pathways previously reported for some of the genes and phenotypes that we assessed in this study. For instance, the FlgR transcriptional regulator has been shown to induce the expression of flaB but not that of flaA, the genes encoding the flagellins of H. pylori, which are expressed as independent transcriptional units, and FlgR was demonstrated to be required for H. pylori motility (Spohn & Scarlato, 1999). Consistently with these reported findings, our results indicated that the palmitic and oleic acids up-regulated the transcription of both flgR and flaB genes, but not that of the flaA gene. Although stearic and linolenic acids induced the expression of flgR gene, these LCFAs did not affect the flaB expression suggesting others regulatory mechanisms. Furthermore, we found that the palmitic acid enhanced the H. pylori motility. As described above, the positive effect of a LCFA on the motility was observed for V. cholerae, since the linoleic acid increased the bacterial swimming by up-regulation of FlrA flagellar regulator (Chatterjee, Dutta & Chowdhury, 2007). Thus, our results suggest that mainly the palmitic acid acts as a signaling molecule that favors the H. pylori motility by inducing the expression of the flaB gene through the FlgR regulator. In this way, the transcription of fur and ureB genes were repressed in the presence of linoleic acid, supporting the observations that ureB is negatively controlled by 5 ureB-sRNA, which is repressed by ArsRS two-component system (Wen et al., 2011), and this bicisctronic gene is repressed by Fur regulator (Roncarati et al., 2016). In terms of sabA, it was previously reported that this adhesin gene was repressed by the ArsRS two-component system in the H. pylori strain J99 (Goodwin et al., 2008). However, our data showed that all LCFAs, except stearic acid, were able to stimulate the expression of both arsR and sabA genes, suggesting that the ArsRS two-component system positively regulates the sabA transcription in the H. pylori strain 26695 in the presence of these LCFAs. These differences in the sabA regulation between both strains could be associated with the poly-CT tract length within the 5 portion of the coding region of each strain (Goodwin et al., 2008;Åberg et al., 2014), where ArsR DNA-binding activity could be modified by the local DNA structure and/or direct interaction with any LCFA.
Our study shows by the first time that LCFAs affect the expression of H. pylori virulence and regulatory genes. The elucidation of the mechanisms by which LCFAs alter the expression of these genes could help to develop anti-infective strategies to counteract the infections by H. pylori.

CONCLUSIONS
The palmitic and oleic acids were the two main LCFAs that increased the transcription of virulence genes involved in motility, adherence, and cell wall synthesis of H. pylori. These same fatty acids enhanced the expression of almost all the regulatory genes tested. While the linoleic acid repressed the transcription of virulence and regulatory genes, the linolenic acid showed a dual effect on the expression of both classes of genes in H. pylori.