Involvement of microbiota and short-chain fatty acids on non-alcoholic steatohepatitis when induced by feeding a hypercaloric diet rich in saturated fat and fructose

Consumption of high-energy-yielding diets, rich in fructose and lipids, is a factor contributing to the current increase in non-alcoholic fatty liver disease prevalence. Gut microbiota composition and short-chain fatty acids (SCFAs) production alterations derived from unhealthy diets are considered putative underlying mechanisms. This study aimed to determine relationships between changes in gut microbiota composition and SCFA levels by comparing rats featuring diet-induced steatohepatitis with control counterparts fed a standard diet. A high-fat high-fructose (HFHF) feeding induced higher body, liver and mesenteric adipose tissue weights, increased liver triglyceride content and serum transaminase, glucose, non-HDL-c and MCP-1 levels. Greater liver malondialdehyde levels and glutathione peroxidase activity were also observed after feeding the hypercaloric diet. Regarding gut microbiota composition, a lowered diversity and increased abundances of bacteria from the Clostridium sensu stricto 1, Blautia, Eubacterium coprostanoligenes group, Flavonifractor, and UBA1819 genera were found in rats featuring diet-induced steatohepatitis, as well as higher isobutyric, valeric and isovaleric acids concentrations. These results suggest that hepatic alterations produced by a hypercaloric HFHF diet may be related to changes in overall gut microbiota composition and abundance of specific bacteria. The shift in SCFA levels produced by this unbalanced diet cannot be discarded as potential mediators of the reported hepatic and metabolic alterations.

Body weight and food intake were monitored daily. Faecal samples were collected 126 and processed as explained elsewhere (Milton-Laskibar et al., 2021). Serum was 127 obtained by blood sample centrifugation after clotting (1,000g for 10 min, at 4ºC). 6 Liver, as well as different white adipose tissue depots (subcutaneous, epididymal, 129 perirenal, and mesenteric) were dissected, weighed, and immediately frozen in liquid 130 nitrogen. Fresh faecal samples were collected at the end of the intervention period, prior 131 to the overnight fasting. To do so, the animals were taken one at a time and housed in a 132 clean, single cage to separatedly obtain faeces directly after defecation induced by a soft 133 abdominal massage. All samples were stored at -80ºC until analysis.   shaking. Then, samples were centrifuged (5 minutes, 15000 rpm at 4ºC) and 40 μL of 168 the upper organic layer was transferred and evaporated to dryness using a SPE-dryer. 169 The residual was reconstituted in 200 μL of 50 % aqueous MeOH, briefly vortexed and  The chromatographic separation was performed with a gradient, which was 0.1 173 % formic acid in water with 10 mM of ammonium formate for mobile phase A and 0.1 174 % formic acid in methanol: isopropanol (9:1 v/v) for mobile phase B. The column 175 temperature was set at 45ºC, and the injection volume was 1 μL. The source parameters 176 were optimised operating in positive electrospray sionisation (ESI) to obtain the 8 maximum response. The validation of the analytical methodology was carried out by 178 analysing a faecal sample pool by standard addition, using the internal standards 179 mentioned above. The quality parameters determined were linearity, limit of detection 180 (MLD), limit of quantification (MQL), and both intraday and interday precision 181 (repeatability and intermediate precision, respectively).    Figure 1B). Indeed, this same pattern 232 was also found regarding liver triacylglycerol content, where the values found in the 233 HFHF group were significantly higher than those in the control group (Table 1). As far 234 as the weight of different adipose deposits is concerned, significant differences were 235 only found in the case of mesenteric adipose tissue. In this case, the values observed in 236 the HFHF group were significantly higher than those found in the control group ( Figure   237 1C). In turn, no significant changes were found between the two groups regarding 238 visceral adipose tissue nor total adipose tissue weights, although non-statistically 239 significant trends were found in both cases (p < 0.1) towards increased weights in the 240 HFHF group.

241
With regard to serum parameters, the fasting serum glucose level observed in the 242 animals from the HFHF group was significantly higher in comparison to that found in 243 the control group (Table 1). Serum non-high-density lipoprotein cholesterol (non-HDL-244 c), which was calculated by subtracting serum high-density cholesterol (HDL-c) levels 245 from total serum cholesterol levels, was significantly increased in the HFHF group 246 compared to the control group. Lastly, the analysis of serum MCP-1 levels revealed that 247 this parameter was significantly increased in the HFHF group in comparison to the 248 control group (Table 1). The hepatic oxidative stress analysis revealed that the amounts of MDA found in 251 the liver samples of animals in the HFHF group were significantly greater than those 252 found in the animals in the control group (Table 1). In addition, while no changes were 253 observed between the two groups regarding CAT activity, a higher GPx activation was 254 observed in the livers of the animals in the HFHF group compared to those in the 255 control group (Table 1). Among all the studied SCFAs, significant differences were found in isobutyric, 258 isovaleric and valeric acids, whose faecal concentrations were significantly greater in 259 the HFHF group compared to the control group (Table 2).

261
To understand the underlying mechanisms by which HFHF diet contributed to 262 hepatic damage, the effects of dietary strategies on the gut microbiota composition were 263 explored. As measured by β-diversity, there was a significant difference in overall 264 microbial composition between the two experimental groups (p < 0.001, 265 PERMANOVA). Figure 2 shows PCoA using the weighted UniFrac distance matrix.

266
The microbiota of HFHF group was clearly separated from that of the control group.   In accordance with these facts, in the present study, rats fed with the high-fat, 320 high-fructose diet showed increased liver weight and triglyceride content, which were

328
In addition, the enhanced hepatic MDA content found in the rats fed the high-fat 329 high-fructose diet suggests that a greater ROS production occurred in these animals,  Traditionally, the so called "two hit" theory has been used to describe the events 337 resulting in NAFLD. According to this theory, the "first hit" is originated by insulin 338 resistance mediated excessive hepatic lipid accumulation due to enhanced de novo 339 lipogenesis and altered fatty acid transport, whereas the "second hit" accounts for 340 hepatic oxidative stress, inflammation and mitochondrial dysfunction. All these events 341 would lead to NAFLD development, as well as progression to NASH (Engin, 2017).

342
However, this theory has been considered as too simplistic, and thus the "multiple hit" 343 theory has been proposed, which besides the aforementioned mechanisms, also    In conclusion, the current study demonstrates that the alterations induced by a