GATA4 regionalizes intestinal metabolism and barrier function to prevent immunopathology  

Earley, Z. et al. (BioRxiv) doi: 10.1101/2021.10.28.466194


  • Intestinal Epithelial Cells

  • Microbiota

  • GATA4

  • Retinoic Acid

  • Celiac Disease


Main Findings

The intestine represents an important barrier site which integrates nutrient and metabolic signals, thereby always requiring to control microbiotal composition through balancing mucosal immunity. Along the intestine, tissue sites represent highly compartmentalized regions with distinct metabolic and immune functions. It has previously been shown that intestinal epithelial cells in the jejunum (part of the small intestine) have high expression of the transcription factor GATA4. It is known that B cell-deficiency can downregulate GATA4 in these cells, likely through a microbiota-dependent mechanism. However, it remains still largely unclear how GATA4 deficiency can itself affect intestinal immunity and whether it may be relevant to intestinal homeostasis during immunopathologies.


In the preprint by Earley et al., the authors use an intestinal epithelial cell (IEC) genetic GATA4-KO models (Villin-Cre driven, GATA4^IEC) to address this question. They find that deletion of GATA4 shifts the jejunum transcriptomic profile towards a more ileum-like phenotype. Among the altered genes, the authors identify a subset of genes which is differentially regulated by IL-17 or IFN-g or both. In line with an IL-17 and IFN-g induced transcriptional profile, intestinal lymphocytes are responsible for the expression of these cytokines in this model. These changes proof to a large extend to be dependent on microbiota as shown by germ-free studies. The loss of GATA4 in IECs also leads to a slight shift in microbiota composition with the expansion of the segmented filamentous bacteria (SFB) Cadidatus arthromitus specifically in the jejunum (but not changes in the ileum). The authors demonstrate that SFB can better adhere and colonize the jejunum of GATA4^IEC. Using a series of fecal transplantation, the authors demonstrate nicely that it is the increased colonization of SFB which increases IL-17 and IFN-g producing immune cells. 

Next the authors addressed how GATA4 loss leads to the increased SFB colonization, showing that GATA4IEC have reduced IgA production and IgA-producing B cells. External supplementation of IgA prevents increased SFB colonization in GATA4^IEC mice. This is ascribed to findings in their transcriptomic dataset in retinoic acid metabolism (among other pathways) and to the fact that many genes in this pathway are direct targets of GATA4. The authors confirm the importance of retinoic acid through vitamin A-free diet (which expands SFB colonization in WT mice) and RA administration (which increases IgA production again). 

Here, a more consistent readout of the rescue, including T cell phenotype, would add power to the findings (see limitations) as well as to know how retinoic acid confers changes to IgA production (directly or indirectly).

Next the authors demonstrate that C. rodentium can colonize the jejunum in GATA4^IEC mice, but not WT mice. This colonization (despite sparking or because of it) can promote systemic dissemination and decrease overall survival.

Lastly, the authors analyze celiac disease patients which seem to be predominantly motivated through its localization (predominantly in jejunum) and phenotype (atrophy of villi). The authors demonstrate that active celiac disease patients have reduced GATA4 expression in their epithelium, although GATA4 expression can be recovered by removal of the main antigen gluten. The authors highlight similar transcriptional changes as seen in the mouse model, when grouping the patient cohort into the top 30% of GATA4 expressers and the lower 30% of GATA4 expressers. Interestingly, the authors identify that Actinobacillus correlates with GATA4 expression, IL-17 downstream pathways and retinoic acid. Although hese bacteria are not very often found in active celiac disease patients, splitting the active disease patients on Actinobacillus presence appears to reduce GATA4even further. These results indicate that Actinobacillus can have an impact on GATA4 expression, but it is not the only factor controlling GATA4 (likely inflammation itself would do too).

Overall, this study is well-performed, and the authors have a well-crafted manuscript in their hands. The limitations below would in my eyes help elevating the overall messages and allow for a slightly better understanding of the crosstalk in the gut surrounding GATA4 in IECs.


  • The authors could exploit their rescue experiments more, in particularly through reading out the supposing downstream effects on T cell immunity. In general (since the paper layout focusses at the end on the correlation of GATA4, retinoic acid pathway and IL-17)

  • It remains largely unclear how GATA4 loss affects ILCs in the intestine and whether these cells are contributing to the IL-17 phenotype observed in their transcriptomic analysis.

  • Generally, the paper would flow better if the authors could link the initial mouse data better to their celiac disease patients. It remains a bit unclear how the IL-17 phenotype is connected here, since IL-17 does from current literature does not seem to be a major contributor to celiac disease (to the best of my knowledge). Currently, one could interpret the change in IL-17 immunity to be a by-stander effect to the changes in microbiotal composition of the jejunum.

  • For the  part, it would support their model if they could show a “therapeutical” purpose of their study by showing a rescue of this phenotype using RA or IgA supplementation.

  • In addition, from the human celiac data presented it is also clear that GATA4 expression can be recovered by removing gluten from the diet, thus suggesting that inflammation can alter intestinal GATA4 levels as well. While this does not need to be further dissected in the paper, it should however be discussed and allows for a reciprocal influence of immunity on epithelial GATA4 and .

  • Minor detail: But (in my eyes) it would be advisable to place figure 3J after the IgA part of the figure, such as after figure 3E. In that why the paper would flow a little better.  



The presented data is an important finding as it expands our knowledge on the complex cellular cross-communication at barrier sites and thereby highlighting that the GATA4-driven identity of epithelial cells in the jejunum are critical to prevent colonization of potentially pathogenic bacteria.


The authors identify that active celiac disease patients that have Actinobacillus present have an even further reduced level of GATA4 and an even higher IL-17 downstream pathway enrichment, which may allow for further subgrouping of celiac disease patients and could be of relevance for future therapy options.


Reviewed by Felix Clemens Richter as part of the cross-institutional journal club of the Immunology Institute of the Icahn School of Medicine, Mount Sinai, the Kennedy Institute of Rheumatology and the Oxford Centre for Immuno-Oncology  (OXCIO) (University of Oxford, GB) and Karolinska Institute’s Center for Infectious Medicine (CIM) & Center for Molecular Medicine (CMM). Follow him on Twitter.