We have previously shown that GATA4 specifically regulates jejuno-ileal differences in absorptive enterocyte gene expression and function.16, 22
Here, we show that in ileum, where Gata6
is expressed but Gata4
is not, conditional deletion of Gata6
results in a decrease in cellular proliferation in crypts, a decrease in enteroendocrine cell allocation, a conversion of Paneth cells into goblet-like cells at the base of crypts, and an alteration in the expression of specific absorptive enterocyte genes that are distinct from GATA4-specific targets. In jejunum and duodenum, where Gata6
are co-expressed, conditional deletion of Gata6
did not produce this phenotype. Instead, Paneth cells were increased, perhaps as a compensatory response to the loss of Paneth cells in ileum. When both Gata6
were conditionally deleted, the proximal intestine exhibited all the changes in proliferation, differentiation and gene expression found in the ileum of single Gata6
conditional knockout mice. These data indicate that while GATA4, but not GATA6, specifically controls jejunal vs. ileal identities in absorptive enterocyte gene expression,16, 21, 22
GATA6 and GATA4 share common functions in regulating crypt cell proliferation and secretory cell differentiation. GATA6 and GATA4 have common and distinct functions in regulating the expression of specific absorptive enterocyte genes. We believe that GATA6 and GATA4 mediate their effects on proliferation and lineage differentiation by acting within stem and/or progenitor cells in crypts, and on absorptive enterocyte gene expression by mechanisms that occur within these cells on villi.
Cellular proliferation in intestinal crypts is necessary for the continuous renewal of the intestinal epithelium. Our data show that GATA6 or GATA4 is necessary for the maintenance of villus height and epithelial cell number, and of crypt cell proliferation throughout the small intestine. Because apoptosis on villi (or in crypts) was not increased, we believe that the observed decrease in villus height and cell number is a direct result of the decrease in crypt cell proliferation in which the rate of epithelial cell renewal is reduced. Although GATA factors have been shown to regulate cellular proliferation in other non-intestinal systems,35-39
our data are the first to show a role for this family of transcription factors in cellular proliferation in the small intestine. The decrease in villus height and cell number likely results in a reduction in intestinal surface area, which could adversely affect nutrient absorption, though we did not observe differences in body weight over the 4 weeks of Gata6
deletion. Taken together, these data indicate that GATA6 or GATA4 is necessary to maintain the normal regeneration of the intestinal epithelium by maintaining its proliferative capacity.
The precise allocation and orderly differentiation of specialized cell types along the crypt-villus axis are essential for establishing the functional landscape of the intestinal epithelium. Here, we also show that GATA6 or GATA4 is necessary for secretory cell differentiation. In G6G4del mice, we demonstrate normal goblet cell differentiation and migration onto villi, but a decrease in enteroendocrine and Paneth cells, and a gain of goblet-like cells in crypts. The loss of enteroendocrine cells, indicated by an overall decrease in the number of CHGA-positive cells and Ngn3 gene expression, suggests a decreased commitment to this lineage. Thus, GATA4 or GATA6 is necessary for enteroendocrine cell commitment in secretory progenitors.
The normal differentiation of goblet cells onto villi, and the loss of Paneth cells and gain of goblet-like cells in crypts where Paneth cells normally reside, suggests a defect in the Paneth cell differentiation program. Paneth cell loss and goblet-like cell accumulation did not occur until at least 2 weeks after the induction of Gata6
deletion (Supplementary Figure 5
), consistent with the slower turnover rate of Paneth cells. The goblet-like cells that accumulate in crypts display a nearly complete spectrum of goblet cell characteristics, but are different from fully differentiated goblet cells on villi in that they do not express TFF3 (). The goblet-like cells also express genes that promote Paneth cell differentiation and their crypt localization, including SOX9 and EPHB3 (), respectively, which are not normally expressed in mature goblet cells. A small percentage of cells also have mixed goblet-Paneth features (). We believe that the goblet-like cells that accumulate in the crypts are committed Paneth cells that, in the absence of GATA6 and GATA4, fail to differentiate further, and by default differentiate to a goblet-like cell-type. Together, these data indicate that GATA4 or GATA6 promote the differentiation of Paneth cells in committed (SOX9) and targeted (EPHB3) Paneth progenitors by preventing their default to a goblet-like cell type.
Wnt signaling is required for crypt maintenance and expression of specific Paneth cell genes.4, 5, 27, 40
Although nuclear β-catenin is undetectable in the goblet-like cells in the crypts and the mRNA abundances of Paneth-specific Wnt targets are reduced in the ileum of Gata6del
mice, other (non Paneth-specific) crypt targets are either unaffected (c
-Myc, Lgr5, Ephb2, Ephb3
) or up-regulated (Cd44, Sox9
) by Gata6
deletion (). We therefore conclude that overall intestinal Wnt signaling is not compromised by Gata6/Gata4
deletion. The apparent loss of Paneth-specific Wnt signaling is likely due to the loss of fully differentiated Paneth cells, in which expression of the gene encoding FZD5, the Paneth-specific Wnt receptor,27
is reduced ().
Analysis of Notch signaling reveals a shift toward a goblet cell differentiation program, as indicated by an up-regulation of Spdef,
and down-regulation of both Ngn3
(). Over-expression of Spdef
causes an increase in goblet cells and a decrease in enteroendocrine and Paneth cells, as well as a decrease in crypt cell proliferation,33
a phenotype that is strikingly similar to that in our models. Conditional deletion of Spdef
impairs goblet and Paneth cell maturation, and causes an accumulation of secretory progenitors, as indicated by more crypt cells expressing DLL1.34
Deletion of Dll1
leads to an increase in goblet cells in zebrafish intestine.8
It is thought that DLL1 laterally inhibits the secretory program in adjacent cells by activating Notch signaling and the absorptive enterocyte differentiation program. These data suggest that SPDEF promotes enteroendocrine cell allocation and Paneth cell differentiation, and inhibits goblet cell differentiation, possibly by down-stream regulation of Ngn3
. Our data are consistent with the hypothesis that GATA6 or GATA4 maintains the balance of secretory cells by repressing Spdef
gene expression in secretory progenitors. Since Gata6
mRNA abundance was unaffected by either Gfi1
deletion or Spdef
over-expression (), we conclude that GATA6 or GATA4 act independently of GFI1 and upstream of SPDEF to regulate secretory cell differentiation.
Previously, we showed that conditional deletion of Gata4
results in a jejunum-to-ileum transformation in absorptive enterocyte gene expression.16, 22
These changes occur in the presence of GATA6,16
indicating that GATA6 cannot replace GATA4 to regulate specific GATA4 gene targets. Here, we show that in the ileum, GATA6 or GATA4 activate and repress specific absorptive enterocyte genes that are different from the GATA4-specific targets (). Many of the genes down-regulated in G6del
ileum encode lipid transporters and apolipoproteins, suggesting that Gata6
regulates ileal lipid metabolism (Supplementary Figure 4C
), though we did not detect alterations in serum cholesterol or triglyceride levels (data not shown). Many of the genes up-regulated by conditional Gata6
deletion are normally more highly expressed in colon than small intestine. These data suggest that GATA factors regulate intestinal lipid metabolism, and maintain the proximal-distal transcriptome in the small intestine by distinguishing jejunal vs. ileal (GATA4-specific) and intestinal vs. colonic (GATA4 or GATA6) gene expression. These data also suggest that the expression of specific absorptive enterocyte genes is regulated by mechanisms that involve differential recruitment of specific GATA factors.
Renewal of the intestinal epithelium with precise cell distribution and gene expression patterns is tightly regulated by multiple factors and pathways.41
In this study, we add to this list of regulators by showing that GATA6 or GATA4 are required for intestinal proliferation, secretory cell differentiation and absorptive enterocyte gene expression. GATA factors are thus critical for intestinal regeneration, and could play a role in damage repair and the adaptive response after loss of functional intestinal surface, or in defects in proliferation, as in intestinal neoplasms.