Abstract/Summary

Zinc oxide provides a positive effect on the intestinal microflora of pigs in controlling or preventing diarrhoea caused by both enterotoxigenic and Shiga-toxin-producing E. coli pathotypes during weaning, but it remains unclear how zinc controls or prevents gut pathogen infection. This study investigated the effects of both therapeutic and prophylactic zinc levels on the gut microbiota of pre-weaned piglets to understand how zinc treatment overcomes post-weaning diarrhoea (PWD) in pigs. An in vitro gut model approach was employed where faecal specimens from piglets were inoculated into batch gut culture vessels (n=36) in five runs. Moreover, cultivations were performed to determine the specific effects of zinc on E. coli laboratory F4 and F18 strains – the most common causal agents of postweaning diarrhoea in swine populations. For each inoculum in the case of a therapeutic zinc trial, three conditions were maintained: the control, and zinc treatment (13.2 mM) using piglet donors receiving dietary iron (FeSO4 at 150 mg/kg body weight) or piglet donors receiving intramuscular iron regime (Iron Dextran® at 200 mg iron/piglet) in addition to dietary iron. For each inoculum in the case a of sub-therapeutic zinc trial, four conditions were employed: a non-treatment (control) and three zinc treatments (0.26, 0.52 and 0.78 mM) reflecting prophylactic zinc levels (piglet donors received identical iron regimes). Samples were harvested at 0, 8, 24 and 48 hours for analysis by flow-cytometry fluorescence in situ hybridisation (Flow-FISH), short chain fatty acids (SCFAs) analysis, elemental assay and microbiota community profiling. Interestingly, the Flow-FISH results revealed that the therapeutic zinc level caused a minimum 100-fold reduction in gut bacterial population levels with no significant difference between high and low iron treatment groups at a 5% significance level. In contrast, the prophylactic zinc levels (0.52 and 0.78 mM), at 24 h, caused a significant reduction (~60-fold) in gut bacterial population levels without a significant difference between 0.52 and 0.78 mM at a 5% significance level. However, 0.26 mM zinc resulted in only ~8-fold decline. The Flow-FISH results are supported by the findings of SCFAs data revealing no production at 0.52-13 mM zinc. However, zinc addition resulted in a 20-fold lower production of SCFAs than that noticed in the absence of added Zn. The variations in total SCFA production induced by Zn treatment were highly significant (Student’s t test, p = 0.001) at 24 h in the cases of both high and low iron regimes. The ICP-OES metal analysis indicated the likelihood of Zn-mediated interference with the distribution of transition metals such as Fe, Cu, Ni, Co and Mn within the gut model between the pellet and supernatant fractions, in the case of therapeutic zinc, while little effect was exerted by the prophylactic zinc. Regarding application of therapeutic zinc, significant differences in Fe levels (as compared to the corresponding 0 h) were found in pellets of the Zn-treated cultures for the -Fe group at 24 and 48 h (increase from 5.1 to 7.7 µM, p <0.01), supernatants of Zn-free cultures in the +Fe group at both 8 and 24 h (reductions of ~0.62 and ~0.59 fold, p <0.01 and <0.05, respectively) and pellets of the Zn-free cultures in the -Fe group at 24 h (fall from 120 to 31.4 µM, p <0.05). Also, significant variations in Cu contents (with respect to the corresponding 0 h) were noticed in supernatants of Zn-treated cultures in the -Fe group at 48 h (2.5-fold rise; p <0.05), the supernatants of Zn-free cultures in the +Fe group at the 24 h (1.7-fold fall, p <0.05), supernatants of Zn-free cultures in -Fe group at both 8 h (decease of ~1.7-fold, p <0.05) and pellets of Zn-free cultures in -Fe group at 24 h (elevation of ~1.8-fold, p <0.001). Again, significant differences in Ni contents (in regards to the corresponding 0 h) were found in pellets of Zn-treated cultures of the +Fe group at 24 h (~1.12-fold rise, p ≤0.05), supernatants of Zn-treated cultures for the +Fe group at 8 h (~60- fold reduction, p =0.05) and supernatants of Zn-free cultures in the -Fe group at 24 h (~1.13- fold increase, p =0.01). Further, significant differences of Co concentrations (with regards to the corresponding 0 h) were found in pellets of Zn-treated cultures of the +Fe group at 24 h (~1.06-fold e, p <0.05) and supernatants of Zn-treated cultures of the -Fe group at both 24 and 48 h (1.3- and 1.4-fold increases, p <0.05 and p ≤0.05, respectively). Furthermore, Mn contents (as compared to the corresponding 0 h) showed significant differences in the pellets and supernatants of Zn-free cultures in the +Fe group at 48 h (elevation of ~1.34-fold at p <0.05 and reduction of ~1.6-fold at p =0.01) as well as pellets of Zn-free cultures in the - Fe group at 24 h (increase of ~1.94-fold, p <0.05). From the next generation sequencing (NGS) data analysis, it was evident that at 24 h, therapeutic Zn resulted in reduced proportions of Bacteroidetes (~1.7-fold) and Firmicutes (~1.4-fold), while a raised level for Proteobacteria (~3.3-fold). On the contrary, in the case of prophylactic Zn, 0.26 mM caused a decrease in Bacteroidetes (~3-fold, FDR p <0.05) and Proteobacteria (~2.5-fold, FDR p >0.05) but a significant (FDR p <0.01) increase in Firmicutes (~4.6-fold) at 24 h, whereas 0.52 mM causes a significantly (FDR p <0.001) lower level of Proteobacteria (~48.5-fold) and insignificantly lower levels of Bacteroidetes (~1.15-fold) but insignificantly higher Firmicutes (~6-fold) at the same time point; a similar trend was seen in the case of 0.78 mM Zn. These results were reflected at the genus level. In laboratory pure cultures of the F4 and F18 E. coli strains using either 96-well microtitre plates or shake flasks, almost complete growth inhibition was observed for 0.52 and 0.78 mM zinc. In addition, it was endeavoured to determine the effects of Zn on the ETEC pathogens of PPWD in an in vitro gut model co-culture. However, the designed Flow-FISH probes failed to detect the ETEC strains and therefore, a Taqman qPCR approach was designed (but not executed due to time constraints). To conclude, both therapeutic (13.2 mM) and prophylactic doses of zinc exert major effects in reducing the total number of the gut bacterial microbiota of pre-weaned piglets as well as F4 and F18 E. coli strains in laboratory cultivations which highlight the antibacterial impact of zinc. Therefore, this PhD study has offered novel insights into understanding the effects of Zn on the PPWD-causing ETEC pathogens and suggests that the nutritional levels of Zn can be applied as prophylactic doses to prevent PWD in porcine populations. Also, ZnCl2 appears to be more potent as an antibacterial, while ZnO seems to be less potent, possibly because it is less soluble. This suggests that future experiments should be performed to compare ZnO and ZnCl2, with/without gastric predigestion.