Helicobacter pylori DNA decreases pro-inflammatory cytokine production by dendritic cells and attenuates dextran sodium sulphate-induced colitis

Abstract

Background and aims: Epidemiological data have recently emerged to suggest Helicobacter pylori may protect against certain chronic inflammatory diseases such as inflammatory bowel disease (IBD). However, the mechanism for the observed inverse association between H pylori and IBD has not been described.

Methods: The frequency of immunoregulatory (IRS) to immunostimulatory (ISS) sequences within the genome of various bacteria was calculated using MacVector software. The induction of type I IFN and IL-12 responses by DNA-pulsed murine bone marrow-derived dendritic cells (BMDC) and human plasmacytoid dendritic cells (DC) was analysed by cytokine production. The effect of H pylori DNA on Escherichia coli DNA production of type I IFN and IL-12 was assessed. The in-vivo significance of H pylori DNA suppression was assessed in a dextran sodium sulphate (DSS) model of colitis. The systemic levels of type I IFN were assessed in H pylori-colonised and non-colonised patients.

Results: H pylori DNA has a significantly elevated IRS:ISS ratio. In-vitro experiments revealed the inability of H pylori DNA to stimulate type I IFN or IL-12 production from mouse BMDC or human plasmacytoid DC. H pylori DNA was also able to suppress E coli DNA production of type I IFN and IL-12. The administration of H pylori DNA before the induction of DSS colitis significantly ameliorated the severity of colitis compared with E coli DNA or vehicle control in both an acute and chronic model. Finally, the systemic levels of type I IFN were found to be lower in H pylori-colonised patients than non-colonised controls.

Conclusions: This study indicates that H pylori DNA has the ability to downregulate pro-inflammatory responses from DC and this may partly explain the inverse association between H pylori and IBD.

Figure 1. An elevated IRS:ISS ratio within the genome of H. pylori

Using previously published data describing IRS (immunoregulatory sequences) and ISS (immunostimulatory sequences), the ratio of the frequencies of IRS to ISS were calculated for various bacteria. The highest IRS:ISS can be found in all six strains of H. pylori. H. hepaticus has a IRS:ISS ratio of greater than one which corresponds with its lack of colonic pathogenic ability in the gut of immunocompetent C57/BL6 mice (44). Although C. difficle has an IRS:ISS ratio of greater than one, its ability to cause infection is predicated on toxin formation. Interestingly, probiotic bacteria have low IRS:ISS ratios suggesting their ability to modulate the immune response is mediated through a different mechanism.

Figure 2. H. pylori DNA does not induce a pro-inflammatory response by DCs and suppresses E. coli-induced pro-inflammation

Mouse mDC were cultured with 2.5ug/mL of E. coli DNA or H. pylori DNA. Additionally, mDCs were cultured with 2.5ug/mL of E. coli DNA and 1ug/mL, 3ug/mL, or 5ug/mL of H. pylori DNA for 8 hours. The same experiment was performed using human pDCs, although cultures were performed overnight.. Averaged data from three experiments from mouse data are shown. The human data were obtained from two separate donors. (a) E. coli DNA stimulates significant production of type I IFN from mouse myeloid dendritic cells (mDCs) as compared to H. pylori DNA (37.21 pg/mL vs 1.86 pg/mL, ** = p<0.01). Further, H. pylori-DNA displayed the ability to significantly suppress E. coli-induced type I IFN production in a dose-dependent manner from mDCs (37.2 pg/mL to 11.37 pg/mL, * = p <0.05)). (b) H. pylori-DNA displayed the ability to significantly suppress E. coli-induced type I IFN production in a dose-dependent manner from human plasmacytoid dendritic cells (pDCs) (263.42 pg/mL to 30.63 pg/mL, ** = p<0.01). (c) Last, E. coli DNA stimulated significantly more IL-12 production from mDCs as compared to H. pylori DNA (588.83 pg/mL versus 23.46 pg/mL, *** = p<0.001). Further, H. pylori-DNA displayed the ability to significantly suppress E. coli-induced type I IFN production in a dose-dependent manner from pDCs (588.83 pg/mL to 303.84 pg/mL, *** = p<0.001).

Figure 2. H. pylori DNA does not induce a pro-inflammatory response by DCs and suppresses E. coli-induced pro-inflammation

Mouse mDC were cultured with 2.5ug/mL of E. coli DNA or H. pylori DNA. Additionally, mDCs were cultured with 2.5ug/mL of E. coli DNA and 1ug/mL, 3ug/mL, or 5ug/mL of H. pylori DNA for 8 hours. The same experiment was performed using human pDCs, although cultures were performed overnight.. Averaged data from three experiments from mouse data are shown. The human data were obtained from two separate donors. (a) E. coli DNA stimulates significant production of type I IFN from mouse myeloid dendritic cells (mDCs) as compared to H. pylori DNA (37.21 pg/mL vs 1.86 pg/mL, ** = p<0.01). Further, H. pylori-DNA displayed the ability to significantly suppress E. coli-induced type I IFN production in a dose-dependent manner from mDCs (37.2 pg/mL to 11.37 pg/mL, * = p <0.05)). (b) H. pylori-DNA displayed the ability to significantly suppress E. coli-induced type I IFN production in a dose-dependent manner from human plasmacytoid dendritic cells (pDCs) (263.42 pg/mL to 30.63 pg/mL, ** = p<0.01). (c) Last, E. coli DNA stimulated significantly more IL-12 production from mDCs as compared to H. pylori DNA (588.83 pg/mL versus 23.46 pg/mL, *** = p<0.001). Further, H. pylori-DNA displayed the ability to significantly suppress E. coli-induced type I IFN production in a dose-dependent manner from pDCs (588.83 pg/mL to 303.84 pg/mL, *** = p<0.001).

Figure 3. H. pylori DNA attenuated the severity of acute DSS-induced colitis

Mice were pre-treated with oral gavage of 50ug of H. pylori DNA, E. coli DNA, or vehicle control prior to DSS administration for 7 days. This was followed by a 7 day recovery period. Mice receiving H. pylori DNA displayed (a) less weight loss, less bleeding measured with Hemoccult, and greater stool consistency compared to the E. coli or vehicle control groups. The differences in each parameter widened during the 7 day recovery phase. Consequently, H. pylori DNA treated mice had lower disease activity scores, most pronounced at day 14 (H. pylori DNA: 0.75, E. coli DNA: 1.22, TE: 1.21, ** = p<0.01). Data were taken from all 8 mice per group. Furthermore, (b) H. pylori DNA-treated mice had significantly longer colons than E. coli DNA mice (* = p<0.05). and (c) less severe histological scores compared to E. coli- or vehicle-treated mice (* = p<0.05). Notice the loss of glandular architecture and significant lymphocyte infiltration in the histological sections from the TE and E. coli-DNA treated mice. This is less severe in the H. pylori DNA-treated mouse section. All 8 mice per group were scored. One section, representative of that group, is displayed. The black arrows highlight the epithelium in each histological section. As illustrated, there is significant more lymphocytic infiltration and loss of glandular architecture in the E. coli and TE groups as compared to the H. pylori DNA group.

Figure 3. H. pylori DNA attenuated the severity of acute DSS-induced colitis

Mice were pre-treated with oral gavage of 50ug of H. pylori DNA, E. coli DNA, or vehicle control prior to DSS administration for 7 days. This was followed by a 7 day recovery period. Mice receiving H. pylori DNA displayed (a) less weight loss, less bleeding measured with Hemoccult, and greater stool consistency compared to the E. coli or vehicle control groups. The differences in each parameter widened during the 7 day recovery phase. Consequently, H. pylori DNA treated mice had lower disease activity scores, most pronounced at day 14 (H. pylori DNA: 0.75, E. coli DNA: 1.22, TE: 1.21, ** = p<0.01). Data were taken from all 8 mice per group. Furthermore, (b) H. pylori DNA-treated mice had significantly longer colons than E. coli DNA mice (* = p<0.05). and (c) less severe histological scores compared to E. coli- or vehicle-treated mice (* = p<0.05). Notice the loss of glandular architecture and significant lymphocyte infiltration in the histological sections from the TE and E. coli-DNA treated mice. This is less severe in the H. pylori DNA-treated mouse section. All 8 mice per group were scored. One section, representative of that group, is displayed. The black arrows highlight the epithelium in each histological section. As illustrated, there is significant more lymphocytic infiltration and loss of glandular architecture in the E. coli and TE groups as compared to the H. pylori DNA group.

Figure 3. H. pylori DNA attenuated the severity of acute DSS-induced colitis

Mice were pre-treated with oral gavage of 50ug of H. pylori DNA, E. coli DNA, or vehicle control prior to DSS administration for 7 days. This was followed by a 7 day recovery period. Mice receiving H. pylori DNA displayed (a) less weight loss, less bleeding measured with Hemoccult, and greater stool consistency compared to the E. coli or vehicle control groups. The differences in each parameter widened during the 7 day recovery phase. Consequently, H. pylori DNA treated mice had lower disease activity scores, most pronounced at day 14 (H. pylori DNA: 0.75, E. coli DNA: 1.22, TE: 1.21, ** = p<0.01). Data were taken from all 8 mice per group. Furthermore, (b) H. pylori DNA-treated mice had significantly longer colons than E. coli DNA mice (* = p<0.05). and (c) less severe histological scores compared to E. coli- or vehicle-treated mice (* = p<0.05). Notice the loss of glandular architecture and significant lymphocyte infiltration in the histological sections from the TE and E. coli-DNA treated mice. This is less severe in the H. pylori DNA-treated mouse section. All 8 mice per group were scored. One section, representative of that group, is displayed. The black arrows highlight the epithelium in each histological section. As illustrated, there is significant more lymphocytic infiltration and loss of glandular architecture in the E. coli and TE groups as compared to the H. pylori DNA group.

Figure 4. H. pylori DNA attenuated the severity of chronic DSS-induced colitis

Oral administration of 20ưg of endotoxin-free DNA 2 hours prior to induction of colitis with 2% DSS was performed. DSS was given in the drinking water for 5 days, followed by substitution of regular water for the next 6 days (days 6–11). This cycle was repeated two more times, at the end of which the mice were sacrificed. Data are shown as an average from all 5 mice per group As seen in (a), mice treated with H. pylori DNA experienced less bleeding and diarrhea as compared to mice treated with E. coli DNA or TE. Further, there was less severe inflammation histologically in mice treated with H. pylori DNA (b). The black arrows highlight the epithelium in which there is more lymphocytic infiltration and loss of glandular architecture in the E. coli DNA and TE treated groups. As seen in (c), there was no statistical difference in the level of cecal mRNA expression of type I IFN between groups, however the level of systemic type I IFN in the H. pylori DNA-treated group was significantly less as compared to the E. coli DNA or TE-treated groups (d).

Figure 4. H. pylori DNA attenuated the severity of chronic DSS-induced colitis

Oral administration of 20ưg of endotoxin-free DNA 2 hours prior to induction of colitis with 2% DSS was performed. DSS was given in the drinking water for 5 days, followed by substitution of regular water for the next 6 days (days 6–11). This cycle was repeated two more times, at the end of which the mice were sacrificed. Data are shown as an average from all 5 mice per group As seen in (a), mice treated with H. pylori DNA experienced less bleeding and diarrhea as compared to mice treated with E. coli DNA or TE. Further, there was less severe inflammation histologically in mice treated with H. pylori DNA (b). The black arrows highlight the epithelium in which there is more lymphocytic infiltration and loss of glandular architecture in the E. coli DNA and TE treated groups. As seen in (c), there was no statistical difference in the level of cecal mRNA expression of type I IFN between groups, however the level of systemic type I IFN in the H. pylori DNA-treated group was significantly less as compared to the E. coli DNA or TE-treated groups (d).

Figure 4. H. pylori DNA attenuated the severity of chronic DSS-induced colitis

Oral administration of 20ưg of endotoxin-free DNA 2 hours prior to induction of colitis with 2% DSS was performed. DSS was given in the drinking water for 5 days, followed by substitution of regular water for the next 6 days (days 6–11). This cycle was repeated two more times, at the end of which the mice were sacrificed. Data are shown as an average from all 5 mice per group As seen in (a), mice treated with H. pylori DNA experienced less bleeding and diarrhea as compared to mice treated with E. coli DNA or TE. Further, there was less severe inflammation histologically in mice treated with H. pylori DNA (b). The black arrows highlight the epithelium in which there is more lymphocytic infiltration and loss of glandular architecture in the E. coli DNA and TE treated groups. As seen in (c), there was no statistical difference in the level of cecal mRNA expression of type I IFN between groups, however the level of systemic type I IFN in the H. pylori DNA-treated group was significantly less as compared to the E. coli DNA or TE-treated groups (d).

Figure 5. H. pylori-colonized patients have lower systemic levels of type I IFN

Serum samples from volunteers presenting for routine screening colonoscopy were collected. The serum was then tested for H. pylori IgG antibody as well as for levels of type I IFN. As seen in (a), patients colonized with H. pylori had significantly lower levels of type I IFN in their serum compared to non-colonized controls (* = p<0.05). Additionally, the number of patients with detectable type I IFN levels was higher in non-colonized patients versus H. pylori-colonized patients (b).