The Impact of ymoA Gene Expression on Enterotoxin YstA Production by Yersinia Enterocolitica Strains With Different Enterotoxic Properties

Yersinia enterocolitica is one of the main causative agents of human diarrhoea and the reservoir and source of infection for humans are pigs. Strains isolated from humans with clinical yersiniosis and diarrhoea are able to produce Yersinia stable toxins – Yst. However, enterotoxin-producing capabilities have been attributed to the ymoA gene which encodes the production of the Yersinia modulator protein – YmoA. The aim of this study was to analyse ystA and ymoA genes expression in Y. enterocolitica strains with different enterotoxic properties, isolated from humans and pigs. The experiment involved two groups of Y. enterocolitica strains producing and not producing enterotoxin YstA, which were isolated from humans and pigs. All these strains were ystA- and ymoA-positive. The relative expression level of the ystA gene was signicantly higher than the expression level of the ymoA gene in Y. enterocolitica strains isolated from humans with clinical signs characteristic for yersiniosis. In others, a signicant decrease in ystA gene transcription was observed, and the relative expression level of the ymoA gene was signicantly higher than the expression level of the ystA gene. Statistically signicant differences were not observed in either group of strains isolated from pigs. The results of our study revealed a correlation between the mRNA expression levels of ystA and ymoA genes in Y. enterocolitica strains isolated from humans. The expression of ystA and ymoA mRNA in Y. enterocolitica strains with different bioserotypes, isolated from humans. The results of PCR for ystA and genes were normalized against of


Background
Yersinia enterocolitica is one of the main causative agents of human diarrhoea with growing epidemiological importance 1 . Various authors have postulated a correlation between strains isolated from healthy pigs and yersiniosis in humans [2][3][4][5] . These ndings suggest that pigs are the main reservoir and source of infection for humans. Strains isolated from humans with clinical yersiniosis and diarrhoea are able to produce enterotoxins Yersinia stable toxins (Yst), which indicates that Yst play a signi cant role in the aetiology of diarrhoea that accompanies the disease and is one of the key virulence factors of Y. enterocolitica. Two main groups of Y. enterocolitica enterotoxins have been identi ed: enterotoxin YstI which includes variants YstA, YstB, and YstC, and enterotoxin YstII whose mechanism of action probably differs from that of YstI 6 .
The best-known Y. enterocolitica enterotoxin, YstA, is a 30-amino-acid peptide whose mechanism of action is based on activation of guanylate cyclase. This mechanism of action is highly similar to that found in enterotoxin STI produced by Escherichiacoli, and it is responsible for an increase in cGMP level in intestinal epithelial cells and extracellular accumulation of liquid [7][8] . The opponents of the hypothesis postulating that YstA is the main cause of diarrhea during yersiniosis have pointed out that YstA is not produced at temperatures higher than 30°C. They have argued that enterotoxin YstA is unlikely to induce diarrhoea since the temperature in the intestines approximates 37°C. However, Mikulskis et al. 9 demonstrated that ystA transcription can be induced at 37°C by providing the pH of the culturing similar to that in the ileum at pH 7.5. They showed that under such conditions, YstA production is identical to that noted at temperatures below 30°C. Enterotoxins YstB and YstC are produced by Y. enterocolitica strains belonging to biotype 1A. These strains are generally considered as non-pathogenic, however recent research indicates that they could play a role in diarrhoea induction 10 .
YstA is encoded by the ystA gene, but not all ystA-positive strains produce enterotoxin. Enterotoxinproducing capabilities have been attributed to the ymoA gene which encodes the production of the Yersinia modulator (YmoA) protein. YmoA belongs to the family of nucleoid-associated proteins and its sequence is in 82% identical with the regulator of high haemolysin activity (Hha) proteins in E. coli and Salmonella 11 . YmoA in uences DNA supercoiling and forms heterodimers with histone-like nucleoid structuring (H-NS) proteins [12][13][14][15] . H-NS play important roles as structural proteins and gene expression modulators [16][17] . Peruzy et al. 18 recently showed that 161 Y. enterocolitica strains of differed origins, tested for the ymoA gene presence, were positive. It means that to evaluate the role of YmoA as another genes modulator it's expression should be examined. The expression of genes responsible for the pathogenicity of Y. enterocolitica has been broadly investigated. Research results indicate that YmoA is one of the main modulators of gene expression in response to environmental factors [19][20] and that it participates in the negative regulation of virulence marker transcription [21][22] .
The possible in uence of ymoA on yst genes was rst postulated by Cornelis et al. 21 who insinuated that the ymoA mutation unblocks the silencing of the yst gene and stimulates enterotoxin production.
However, the results of our previous study 23 show that two point mutations in the nucleotide sequence of the ymoA gene, which were detected with the use of the High-Resolution Melting (HRM) method, did not in uence on enterotoxic properties of the examined strains. Mikulskis et al. 9 in 1994 presented the mechanism modifying the expression of yst to a silent state. According to the cited authors, gene silencing was caused by modi cations in the status of bacterial host factors, and YmoA participated in both yst silencing and temperature regulation. YmoA was identi ed as one of the factors necessary for growth-phase regulation of yst. In 1998, Grant et al. 24 have also suggested that the lack of enterotoxic properties in selected Y. enterocolitica strains could result from the inhibitory in uence of the ymoA gene on ystA gene expression in in vitro cultures. To date, this possibility has been investigated only by Starke and Fuchs 25 who identi ed YmoA as a silencing factor for all toxic complex (tc) genes of Y. enterocolitica strain W22703 (biotype 2, serotype O:9).
The purpose of the study was to analyse ymoA gene expression in Y. enterocolitica strains with different enterotoxic properties and to evaluate the inhibitory effect of the ymoA gene on the production of enterotoxin YstA by the strains isolated from humans and pigs.

Materials
This study was performed retrospectively, based only on bacterial strains and did not require ethical approval. Y. enterocolitica strains were previously isolated from samples routinely submitted to the diagnostic laboratories and obtained from infected humans. Y. enterocolitica strains isolated from pigs were obtained from a previous study 26 . The experimental material consisted of 74 Y. enterocolitica strains isolated from humans and 51 Y. enterocolitica strains isolated from pigs. In this study, only Y. enterocolitica strains from infected humans and animals were analysed, therefore, control groups (uninfected) could not be established.

Y. enterocolitica strains isolated from humans
Group I was composed of 34 Y. enterocolitica strains isolated from clinical cases of yersiniosis (anonymous data from medical history delivered by laboratories). Group II consisted of 40 Y. enterocolitica strains isolated from humans with unknown clinical diagnosis. All Y. enterocolitica strains were before biotyped, serotyped and molecularly examined (ystA, ystB, ystC, ymoA). Primer sequences and PCR conditions were described previously 26 . Group I consisted of 13 Y. enterocolitica strains belonging to a rare in Poland bioserotype 1B/O:8 and 21 strains belonging to bioserotype 4/O:3. Group II consisted of Y. enterocolitica strains belonging only to bioserotype 4/O:3 because highly pathogenic bioserotype 1B/O:8 strains have never been isolated from humans without clinical yersiniosis. All Y. enterocolitica strains used in this study were ystAand ymoA-positive.

Y. enterocolitica strains isolated from pigs
Fifty-one Y. enterocolitica strains isolated from fattening pigs without clinical signs of yersiniosis were examined. The enterotoxic properties of these strains were determined previously, using suckling mouse bioassay 26 . Enterotoxin production was evaluated by measuring the ratio of intestinal mass to the remaining body mass in three examined sucklings. According to Gianella 27 , a ratio of ≤ 0.074 indicates a negative result, a ratio of 0.075-0.082 denotes a doubtful result, and a ratio of ≥ 0.083 represents a positive result. In this study, 13 Y. enterocolitica strains producing enterotoxin YstA in the suckling mouse bioassay formed Group I, and 38 Y. enterocolitica strains not producing enterotoxins in the suckling mouse bioassay formed Group II. All examined strains belonged to bioserotype 4/O:3 and were ystAand ymoA-positive.

RNA preparation and reverse transcription
Bacteria were grown in tryptic soy broth (TSB) at 28°C, and the inoculated medium was incubated with shaking (250 r.p.m.) by 24h. The cells were harvested by centrifugation in an Eppendorf Centrifuge 5804 R for 5 min. at a speed of 3100xg and the supernatant was discarded. Total RNA extraction was done with cell pellets containing 1 × 10 7 cells using RLT Buffer, being a part of RNeasy Protect Bacteria Mini Kit (Qiagen, Hilden, Germany). This kit includes the RNAprotect Bacteria Reagent for stabilizing RNA in bacterial samples and RNeasy spin columns for purifying up to 100 µg of high-quality RNA using the silica-membrane technology. Next steps of RNA extraction were done according to the manufacturer's instructions. RNA integrity was assessed by agarose gel electrophoresis. RNA concentration and quality were measured with the NanoDrop 2000 spectrophotometer (Thermo Fisher Scienti c Inc., Waltham, Massachusetts, USA). An A 260 /A 280 ratio of 2.0 (in the range of 2.06-2.13) was considered pure RNA.
Reverse transcription (RT) into cDNA was carried out with the QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. cDNA was stored at -20°C until further use.
Gene expression analysis using qPCR Selected genes were analysed by quantitative real-time PCR (qPCR) with the Rotor-Gene6000™ real-time analyser (Corbett Life Science, Sydney, Australia). The expression of ystA and ymoA was normalized to that of the gapA and polA reference genes, encoding D-glyceraldehyde-3-phosphate dehydrogenase, and production of DNA polymerase I, respectively 28 . The forward and reverse primers used in this study are shown in Table 1. Every sample for ystA, ymoA, and gapA, polA mRNA analysis contained cDNA (70 ng), forward and reverse primers ( nal concentration of 0.7 µM/l each) and the QuantiTect SYBR Green RT-PCR master mix (Qiagen, Hilden, Germany), according to the manufacturer's instructions. Standard curves of serial dilutions of the appropriate puri ed cDNA were used for quanti cation. Each PCR reaction (25 µl) was performed in duplicate in a 36-well rotor under the following conditions: initial denaturation at 95°C for 5 min, followed by 40 cycles of denaturation at 95°C for 10 s and annealing at 52°C for 30 s, followed by elongation at 72°C for 45 s. Final elongation at 72°C for 10 min was carried out for each PCR reaction. Melting curves were obtained based on stepwise increase in the temperature ramp from 65°C to 90°C to ensure the ampli cation of a single product for each reaction. Table 1 Sequences of the primers used in the study

Gene
Forward primer Reverse primer Reference To test the expression of ystA and ymoA mRNA in different groups of Y. enterocolitica strains, two-way ANOVA was performed. All numerical data are expressed as means ± SEM at a signi cance level of p < 0.05, p < 0.01 and p < 0.001.

Results
In vitro and in vivo expression of bacterial virulence factors can be examined with the use of reference strains which are subject to observable change when genes of interest are expressed 33 . To understand in vitro expression and the possible role of the ymoA gene during diarrhoea induction, we monitored the expression of ystA and ymoA genes in Y. enterocolitica strains with known enterotoxic properties, observed in vivo. All examined strains were ystA-, and ymoA-positive, irrespective of their ability to produce enterotoxin in vivo. Y. enterocolitica strains isolated from humans with clinical yersiniosis and Y. enterocolitica strains isolated from pigs and capable of producing enterotoxin YstA in the suckling mouse bioassay were used in this experiment. To compare, Y. enterocolitica strains isolated from infected humans with unknown clinical diagnosis and Y. enterocolitica strains isolated from pigs and not producing the enterotoxin in the suckling mice bioassay, were used.
The relative expression level of the ystA gene was signi cantly higher (p < 0.001) than the expression level of the ymoA gene in Y. enterocolitica strains isolated from humans with clinical yersiniosis (Group I).
The reverse was noted in the group of Y. enterocolitica strains isolated from humans with unknown clinical diagnosis (Group II) -a signi cant decrease in ystA gene transcription was observed in all these strains, and the relative expression level of the ymoA gene was signi cantly higher (p < 0.05) than expression level of the ystA gene (Fig. 1a). Therefore, correlation was found between the relative levels of ystA and ymoA mRNA.
Statistically signi cant differences were not observed in either group of strains isolated from pigs. Signi cant differences in the relative expression levels of ystA and ymoA genes were not noted in Y. enterocolitica strains isolated from pigs and capable of producing enterotoxin YstA in the suckling mouse bioassay (Group I). In the group of Y. enterocolitica strains isolated from pigs and not capable of producing enterotoxin YstA (Group II), the relative expression level of the ymoA gene was higher than the expression level of the ystA gene, but the observed differences were not statistically signi cant (Fig. 1b). Therefore, no correlation was found between the relative levels of ystA and ymoA mRNA, as was observed in humans.
The results of a statistical analysis of ystA and ymoA mRNA levels in Y. enterocolitica strains isolated from humans with clinical yersiniosis revealed minor differences between bioserotypes. The signi cance of the observed differences was determined at p < 0.001 in Y. enterocolitica strains belonging to the highly pathogenic bioserotype 1B/O:8 (Fig. 2a) and at p < 0.05 in Y. enterocolitica strains belonging to bioserotype 4/O:3 (Fig. 2b). Since highly pathogenic strains of bioserotype 1B/O:8 were not isolated from humans without clinical yersiniosis, Group II strains of Y. enterocolitica belonging only to bioserotype 4/O:3 were used for comparison. Similar statistical trends in the mRNA expression levels of ystA and ymoA genes in Y. enterocolitica strains were noted in Group II in comparison to both Groups I.
To summarize, the same ystA gene was expressed differently in the tested groups of Y. enterocolitica strains isolated from humans. Based on the relative expression levels of ystA and ymoA genes the strains represented two different groups. The strains obtained from patients with clinical yersiniosis expressed the ystA gene, whereas in Y. enterocolitica strains isolated from humans with unknown clinical diagnosis the expression of the ystA gene was lower than ymoA. In both groups of Y. enterocolitica strains, the expression level of ystA mRNA was correlated with the expression level of ymoA mRNA. Statistically signi cant differences were not observed in either group of strains isolated from pigs.

Discussion
It has been long suggested that YmoA is an important determinant of the production of enterotoxin Yst by Y. enterocolitica strains 9,21,24 . Research aiming to con rm or rule out the above hypothesis has not been undertaken since the above observation had been made. Although, YmoA has been con rmed as a negative regulator of the transcription of other virulence markers, such as inv, which encodes invasinthe essential factor of internalization, responsible for the transport of Y. enterocolitica across M cells [21][22] . YmoA was also shown to participate in production of Yersinia outer proteins (Yops) and Yersinia adhesin (YadA), dependent on temperature 22 . More recently, Böhme et al. 11 described YmoA as a thermosensitive virulence modulator protein which optimizes temperature apperception and ne-tunes virulence gene expression during infection. To better understand the regulatory factors that contribute to enterotoxin production by Y. enterocolitica, we have examined the molecular mechanism that switches ystA expression to a silent state. Our recent study revealed that two point mutations in the coding region of the ymoA gene nucleotide sequence do not affect the enterotoxic properties of the examined strains 23 .
Our ndings did not con rm the postulated in uence of ymoA mutations on ystA gene silencing 9,21 . However, analyses of genes encoding H-NS proteins in Yersinia spp. are hampered by the fact that their mutations are harmful for cells 34,35 . Our study was prompted by the above observation as well as the hypothesis that decreased expression of the ystA gene in in vitro cultures could be responsible for the absence of enterotoxic properties in selected Y. enterocolitica strains.
Results obtained in this study broaden the knowledge about interactions between ystA and ymoA, including their involvement in the pathogenicity of Y. enterocolitica, and the spectrum of virulence genes that are controlled by YmoA. We observed a signi cant reduction of ystA gene transcription in strains isolated from humans with unknown clinical diagnosis. The relative expression level of the ymoA gene was signi cantly higher than the expression level of the ystA gene. In patients diagnosed with yersiniosis, the relative expression level of the ymoA gene was signi cantly lower than the expression level of the ystA gene. The above was particularly evident in Y. enterocolitica strains belonged to the highly pathogenic bioserotype 1B/O:8, responsible for the most severe cases of the disease. Differences were also observed in the mRNA expression of the ystA gene in Y. enterocolitica strains isolated from humans with yersiniosis and belonged to bioserotype 4/O:3, but they were less signi cant than those noted in bioserotype 1B/O:8 strains.
According to our knowledge, this study is the second research attempt to investigate the in uence of YmoA on the production of enterotoxins by Y. enterocolitica. The other study was conducted by Starke and Fuchs 25 who demonstrated that YmoA silenced all tc genes of Y. enterocolitica strain W22703 (biotype 2, serotype O:9). Using fusions of promoter with the luciferase reporter, they detected that the deletion of ymoA increased the transcription of tcaR1, tcaR2, tcaA, tcaB, tcaC, tccC1 and tccC2 at 15°C and 37°C temperatures. They also observed that at low temperatures, the amount of thermostable YmoA in cells was not reduced, but the repressor was less functional. In the cited study, supplementation by episomal ymoA greatly reduced tc gene expression, thus con rming the inhibitory in uence of YmoA on the production of insecticidal proteins. According to Starke and Fuchs 25 , YmoA facilitates H-NS binding to tc promoters by creating a compound with this nucleoid-associated protein. The resulting compound not only binds to the upstream regions of all tc genes, but also to intragenic sites of tcaA and tcaB; therefore, it plays a signi cant role due to control the expression of both genes. Those observations are in line with our ndings, which indicate correlation between ystA and ymoA expression levels.
However, further research involving ymoA mutants seems to be essential to validate this observation as the similar differences were not found in Y. enterocolitica strains isolated from pigs. Interestingly, a decrease in the mRNA expression of the ymoA gene was not observed in Y. enterocolitica strains isolated from pigs and producing enterotoxin YstA in the suckling mouse bioassay. The above could be attributed to the fact that we disposed only 13 such strains. The relative expression level of the ymoA gene was higher than the expression level of the ystA gene in the group of Y. enterocolitica strains which were isolated from pigs and not able to produce enterotoxins, but the noted differences were not statistically signi cant. The use of more strains with proven ability to YstA enterotoxin production may allow the statistically signi cant results obtain. An interesting aspect would also be an examination of Y. enterocolitica strains isolated from other animal species and from food. If the correlation between ystA and ymoA genes expression levels would not be con rm in a larger number of toxin-producing strains, this could indicate existence of the factors that co-operate with YmoA in Y. enterocolitica strains isolated from humans.

Conclusions
The results of our study revealed a correlation between the mRNA expression levels of ystA and ymoA genes in Y. enterocolitica strains isolated from humans. However, given the lack of statistically signi cant differences between Y. enterocolitica strains isolated from pigs and characterized by different enterotoxic properties, further analyses involving a larger number of Y. enterocolitica strains and ymoA mutants are needed to con rm this observation.
Declarations Figure 1 The expression of ystA and ymoA mRNA in Y. enterocolitica strains with different enterotoxic properties, isolated from humans and pigs. The results of real-time PCR for ystA and ymoA genes were normalized against the expression of gapA and polA genes. Data are expressed as the mean ±SEM, and asterisks indicate differences between groups (*p<0.05, ***p<0.001). (a) Y. enterocolitica strains isolated from humans; Group I consists of strains isolated from humans with clinical yersiniosis; Group II consists of strains isolated from humans with unknown clinical diagnosis (b) Y. enterocolitica strains isolated from pigs; Group I consists of strains capable of producing enterotoxin YstA in the suckling mouse bioassay; Group II consists of strains unable to produce enterotoxin. Figure 2