Identification of the molecular interplay between dietary fatty acids and gut microbiota in Non Alcoholic Fatty Liver Disease

Obesity promotes Non Alcoholic Fatty Liver Diseases (NAFLD) through mechanisms involving the gut microbiota. From 2 cohorts of obese women (FLORINASH FP7), in which a multilevel omics approach was used, we identified a microbiome architecture indicating increased dietary lipids processing. This was associated with hepatic insulin signaling perturbations, a transcriptomic profile indicative of NAFLD and triglyceride accumulation. From two other cohorts (MICIMAB & ROLIVER), transcriptomic analyses from intestinal biopsies and 16S sequencing of liver samples suggested: 1. impaired intestinal defense favoring translocation of bacteria towards the liver, inflammation and lipid deposition and 2. impaired intestinal and liver lipid metabolism. The interaction between dietary lipids and the gut microbiota in the etiology of NAFLD is unknown and could impair intestinal defense and lipid handling. Using original and complementary models of genetically modified mice and germ-free mice colonized with human microbiota, we will study the impact of different lipid-enriched diets on: 1. gut microbiota and its causal role in liver disease; 2. intestinal immune- and non-immune defense systems; 3. translocation of bacteria towards the liver responsible for inflammation; 4. lipid handling processes in the liver and the intestine (bile acids/FXR and fatty acids/PPAR); 5. gender by studying the role of the estrogen receptor  (ER) in the gut microbiota/dietary lipids interplay.

 Main objectives for the time period 20-21

  • To Finalize the phenotyping of the mutant mice regarding their impact on hepatic steatosis
  • To generate more OMICS from the different animal model
  • To initiate and fulfil the database
  • To initiate the analysis of the corresponding data
  • To make decisions about the usefulness of the humanized and germ free conventionalize mice


Partner Organization Partner Country
University of Bari Italy
Gothenburg University Sweden
INRA France



  • Based on our first sets of experiments we have analyzed the host glucose and lipid metabolism including glucose- and insulin tolerance tests, MRI and gut transit time weight and plasma biochemistry. we also have performed liver and adipose tissue inflammation and metabolism: characterized using histology, biochemistry, immunohistochemistry,; and characterized the inflammation of tissue by qPCR. Briefly, Liver histology quantifying inflammation, steatosis and fibrosis, TG, Cholesterol esters and cholesterol and qPCR analysis of liver reporter genes for inflammation, lipogenesis have been run.
  • We have identified 2 fat-enriched diets out of 7 which control in the opposite way the liver lipid storage. The diets used are rich in saturated fatty acids (palm oil), saturated/mono-unsaturated fatty acids (lard), mono-unsaturated fatty acids (oleic sunflower oil), omega-6 poly-unsaturated fatty acids (linoleic sunflower oil), omega-3 poly-unsaturated fatty acids (fish oil), saturated long- and medium chain fatty acids (milk fat) or medium-length fatty acids (man-made formula), respectively (7 fat-modified diets +1 control diet).
  • Although the diets have similar amount of proteins, carbohydrate and energy content, the liver lipid are dramatically accumulating for one while not accumulating for the other. in addition, we have identified the associated gut microbiota which could be controling the differential impact of both diets. eventually, lipidomics revealed that the bile acids could be at play in the differential phenotypes. we have run the experiments in males and females.
  • We have analyzed the potential mechanisms associated with the differential liver lipid storage. The caecum has been analysed for SCFA, bile acids, and fatty acids. We have analysed diet and caecum content for soluble and insoluble fatty acid salts. We did not yet perform liver lipidomics but only TG/ChoE and Cholesterol in the liver and the gallbladder bile.
  • We have analyzed the role of the intestinal, and liver immune systems. the data are currently being analyzed.
  • We have collected liver, gut, and adipose tissue samples to analyze the corresponding microbiota and study the bacterial translocation process potentially at play on the control of liver lipid accumulation. the data should be obtained before summer.
  • In the WP3 we have initiated some experiments to demonstrate causality using PPARa/g, ERa; FXR ko mice fed the above selected diets. the results are on going. In vivo experiments with Alb-Cre is on-going while the in vivo part of the Villin-Cre is finished but downstream analyzes still remains to be performed. Metabolic in vivo characterization similar to what is outlined above (GTT, fating glucose, insulin, liver fat etc) is being performed.
  • Regarding the role of PPARg: We used High Fat Diet (HFD) feeding as a model of obesity in C57BL/6 J male Wild-Type mice (WT), in whole-body Pparα- deficient mice (Pparα−/−) and in mice lacking Pparα only in hepatocytes (Pparαhep−/−). We provide evidence that Pparα deletion in hepatocytes promotes NAFLD and liver inflammation in mice fed a HFD. This enhanced NAFLD susceptibility occurs without development of glucose intolerance. Moreover, our data reveal that non-hepatocytic PPARα activity predominantly contributes to the metabolic response to HFD.Glucose and lipid homeostasis was analyzed, as well as liver histology, transcriptomics, lipidomics and 1H-NMR based metabolomics. No study of gut microbiota was performed in these mice. In addition, we used male and female WT and mice lacking Pparα only in hepatocytes (Pparαhep−/−) treated with several dietary models of steatosis (high-fat diet, choline-deficient high-fat diet and western-diet). Extensive hepatic phenotyping was performed (histology, transcriptomics, 1H-NMR based metabolomics, lipidomics). Upon all dietary challenges, we recapitulate the male-specific hepatic alterations (triglyceride accumulation, inflammation, fibrosis…) and female-specific protection, as observed in NAFLD pathology in humans. We show that hepatic PPARa pathway is involved in these sex-specific features and confirm the involvment of PPARa in the etiology of NAFLD in mens using a human cohort of male and female NAFLD patients.
  • Regarding the role of FXR: In order to evaluate the role of nuclear receptor FXR in diet-induced NAFLD and gut microbiota composition, we performed in vivo experiments in both genders. The male and female transgenic mice with selective constitutive intestinal FXR activation (VP16FXR) and their littermate controls (VP16) have been fed with 3 selected diets. In vivo experiment with Reference and Chow diets are ended, but further analyses have to be done. For glucose and lipid metabolism, GTT and ITT have been performed, as well as serum cholesterol, TG, ALT and AST have been measured. For these mice, gut microbiota and bile acid composition are being analysed. Liver histology quantifying inflammation, steatosis and fibrosis and qPCR analysis of liver reporter genes for inflammation, lipogenesis and fibrosis have not yet been performed.
  • Regarding the role of ERa: we have generarted mice deleted for ERa in the intestinal epithelial cells. the colony is just been raised and the data should be available in the next 6 months
  • The causality of gut microbiota through conventionalization of germ free mice had to be delayed. However, we have treated conventional mice with antibiotics prior to colonization. the results are being analyzed. 

Highlights 20-21

  • Animal models
    • The objectives of this year were to study the role of different a priori genes such as ppara/g fxr/era on the impact of the identified diets on hepatic steatosis. We challenged all the animal models of KO mice with the selected diets in males and females. The phenotypes have been reported. The major observation is that the overexpression of fxr in the intestine leads to a full protection against hepatic steatosis in females only, from our reference high fat diet. The other fat-enriched diets have been tested and show intermediate phenotypes. This is in agreement with their quality fat content i.e. less saturated fat, when compared to the reference fat diet (milk saturated fat).
    • ERalpha animal models: although these animal models grow very slowly due to the mutation we succeeded in generating some more over the year. We removed ERa from the intestine in vilin-CRE ERa mice. So far we have positive insights about an increase liver fibrosis which could be a consequence of some degree of liver inflammation. Interestingly, female mice gain less weight and control better their glycemia when on a fat-enriched diet. The inflammatory status is up in both sexes studied.
    • The role of ppara/g under the different fat-enriched diet has not been done since no major changes have been observed on the regular high fat diet. Therefore we decided to remove this part of the project.
    • In all relevant animal models more metabolic and molecular and cellular phenotyping will be done over the course of the coming year to fully phenotype the models. Some blood and tissue biochemistry as well.
    •  Altogether we will complete our database with numerous molecular and biochemical variables.
  • Omics analyses
    • the objective of this WP was to generate a large database where all metabolic, molecular and cellular data from all animal models will be implemented. Our colleagues at Berlin have been centralizing all data from the partners. A first database has been set and updated with the phenotypes and physiological data obtained on the animal models. We have also obtained some of the OMICS data from the models notably the gut and the tissue microbiota; they have been updated in the database. However, more omics data are currently generated and will update the database.
  • Database:
    • Our colleagues in Toulouse and Berlin have been analyzing the database. We have initiated large multivariate analyses. The data show the impact of the different diets on the clinical data sets allowing to group the animals according to their diet. Liver transcriptomic shows similar signatures. It was however a bit less pronounced for the microbial signature of the intestine. Interestingly the signature of the fecal fatty acids showed that for some animals a very large degree of heterogeneity which was not found in all diet-groups. Some co-omics correlated analyses have been run to identify that the impact of diets are different according to the diet considered. As an example the crosstalk between fecal fatty acids and gut microbiota is driven by the A and C diets and the liver transcriptomics and the fecal fatty acids as well. Other co-omics correlations are driven by other diets.
  • Tissue microbiota analyses.
    • The hypothesis is that a leaky gut is responsible for the selective translocation from the intestine to the liver and adipose depots of bacteria (or bacterial DNA) that depends upon the diets. The reference diets seem to be the more responsible for the leaky gut.
    • The first tissue microbiota signatures showed on animal models discriminant distribution of the individuals according to the diet studied. We have analyzed as well the adipose tissue microbiota from the intestinal FXR overexpression model and show that the mutation slightly impact the signature. We will perform liver transcriptomics, and run some qPCR on the intestine to identify occludins and other proteins involved in intestinal integrity. From the transcriptomics analyses inflammatory pathways could be identified. It is proposed to sequence the liver RNA. We will discuss with the transcriptomic groups if doable.
    • We have been discussing about the most relevant means to analyze the large sets of data. Multivariate and discriminant analyzes as well as clustering strategies will help us to identify variables among the gut microbiota that are related to hepatic steatosis. We will construct multiomics pathways and identify the corresponding biochemical pathways featuring the impact of the diets and the gut microbiota on the steatosis phenotypes. We will include all omics to ascertain the molecular hypothesis to be tested then in vitro and in vivo.
    • We still had to run the NMR analyses however, such analyses usually bring low levels of information when analyzing the crosstalk between soluble molecules and liver fatty acid storage.
    • Regarding the germ free mouse analyses. We have re-developed the germ free mouse facility which was turned down at the past lockdown. The germ free facility is up and running in Toulouse. The discussion was related to the relevance of colonizing mice with the microbiota from the diet fed mice. It is argued that since the recipient mice need to be under the different diets the grafted microbiota will not resist the impact of the diet and thereby no causal answer from the grafts will be obtained if the grafted mice will be fed a fat-enriched diet. If we don’t feed the recipient mice with the diets therefore we will be out of the frame. The germ free mice will be used when some monocolonisation to validate the direct impact of a given set of identified bacteria.
    • Humanizing mice is not yet relevant: although humanized germ line could be fed the ref, A/G diets and the samples stored for reviewer questions.


Title: Hepatic expression of Lipopolysaccharide Binding Protein (Lbp) is induced by the gut microbiota through Myd88 and impairs glucose tolerance in mice independent of obesity
Author: Molinaro A, Koh A, Wu H, Schoeler M, Faggi I, Carreras A, Hallén A, Bäckhed F, Caesar R.
Link: https://doi.org/10.1016/j.molmet.2020.100997
Title: Extra-Virgin Olive Oil from Apulian Cultivars and Intestinal Inflammation
Author: Cariello M, Contursi A, Gadaleta RM, Piccinin E, De Santis S, Piglionica M, Spaziante AF, Sabbà C, Villani G, Moschetta A
Link: https://doi.org/10.3390/nu12041084
Title: Fibroblast Growth Factor 19 modulates intestinal microbiota and inflammation in presence of Farnesoid X Receptor
Author: Gadaleta RM, Garcia-Irigoyen O, Cariello M, Scialpi N, Peres C, Vetrano S, Fiorino G, Danese S, Ko B, Luo J, Porru E, Roda A, Sabbà C, Moschetta A
Link: https://doi.org/10.1016/j.ebiom.2020.102719
Title: Role of Oleic Acid in the Gut-Liver Axis: From Diet to the Regulation of Its Synthesis via Stearoyl-CoA Desaturase 1 (SCD1).
Author: Piccinin E, Cariello M, De Santis S, Ducheix S, Sabbà C, Ntambi JM, Moschetta A
Link: https://doi.org/10.3390/nu11102283
Title: PGC-1beta induces susceptibility to acetaminophen-driven acute liver failure
Author: Piccinin E, Ducheix S, Peres C, Arconzo M, Vegliante MC, Ferretta A, Bellafante E, Villani G, Moschetta A
Link: https://doi.org/10.1038/s41598-019-53015-6
Title: Extra Virgin Olive Oil: Lesson from Nutrigenomics
Author: De Santis S, Cariello M, Piccinin E, Sabbà C, Moschetta A
Link: https://doi.org/10.3390/nu11092085
Title: Gut microbiota of obese subjects with Prader-Willi syndrome is linked to metabolic health
Author: Olsson LM, Poitou C, Tremaroli V, Coupaye M, Aron-Wisnewsky J, Bäckhed F, Clément K, Caesar R.
Link: https://doi.org/10.1136/gutjnl-2019-319322


Project number:
Duration: 100%
Duration: 100 %
Project lead and secretary:
Rémy Burcelin
Responsible organisation:
Inserm, France