Atoh8 was revealed to be one of the 6 transcription factors upregulated by the PRC2 in M cell development. When comparing M cell differentiated organoids with crypt stem cell organoids and differentiated organoids Atoh8 turned out to be significantly
100
expressed in M cell differentiated state (log2 fold changes −3.15 RankL vs WENRC and -2.58 RankL vs ENRI).
Atoh8 (Atonal BHLH Transcription Factor 8) is a member of the basic Helix-Loop-Helix bHLH transcription factor family. All the members of this family have 2 highly conserved binding domains that together makeup 60 amino acid residues. bHLH transcription factors have been documented to be involved in the regulation of the cell cycle, cardiovascular development, hematopoiesis, stem cell maintenance, and skeletal muscle development. Atoh8 is a member of the group A of the bHLH transcription factor family. Atonal superfamily members control numerous aspects of differentiation for vertebrate organ development (R. T. Yan 1998; Tomita et al.
2000; Hutcheson and Vetter 2001; Durand et al. 2012). In the intestinal Atonal BHLH Transcription Factor 1 or Atoh1 plays a critical role in regulating Paneth cell, these cells provide stem cell niche to Lgr5+ cells in the intestinal crypt and is also required for the differentiation of functional secretory lineages through lateral inhibition (Durand et al. 2012). Atoh8 being the sole mammalian member of the bHLH factor that is a part of the NET family (Rawnsley et al. 2013), its role in M cell differentiation was unknown.
Atoh8 expression was found to be localized in the follicle associate epithelium of the Peyer’s patches in wildtype mice. To investigate Atoh8’s role in M cell differentiation and development, we used an Atoh8 intestinal-specific knockout to characterize its role. We observed a higher number of Gp2+ mature M cell expression in the isolated follicle-associated epithelium when compared to its VilCre counterpart. Our whole-mount immunofluorescence of Peyer’s patch isolated from Atoh8 lox/VilCre and VilCre confirmed our expression studies as we observed a 1.5-fold increase of M cells in Atoh8 knockout mice. Early developmental markers such as MarcksL1 and TNFAIP2 were observed to be significantly upregulated indicating that Atoh8’s activity could be upstream of the canonical RelA/p50. Critical transcription factors that act as master regulators of M cell development, Spi-B, Esrrg and Sox8 were revealed to be upregulated as well including the expression of Sox8. To understand if this phenotype observed was related to epithelial deletion of Atoh8 or lymphocytes in the Peyer’s patches, we isolated organoids from the Atoh8 lox/VilCre mice and VilCre mice. RANKL treated Atoh8 null organoids exhibited a significant increase in Gp2 expression along with Spi-B, Esrrg, and Sox8 indicating that epithelial intrinsic Atoh8 was responsible for the increase in M cell numbers. Next, we checked to see if the increase in M cells in Atoh8 knockout mice meant an increase in
101
functional transcytosis capacity. Atoh8 lox/VilCre and Atoh8 VilCre mice were orally administered with equal volumes of 200nm fluorescent latex beads, Atoh8 lox/VilCre mice were able to uptake these beads significantly higher than VilCre mice indicating that M cells in the Atoh8 null mice were functional.
OPG is a receptor that binds to RANKL and acts as a decoy receptor to RANK receptor, lack of OPG increased RANKL expression leading to an increase in M cells higher transcytosis capabilities, and enhanced immune response (Shunsuke Kimura et al. 2020). We explored the possibility if the lack of Atoh8 led to an increase in RankL production in our mice however, our flow cytometry analysis showed no increase in RankL+ T cells, indicating that Atoh8 acts independent of RankL signaling in the GALT. Interestingly, even though OPG deficient mice exhibited a higher number of M cells and enhanced immune responses, they were highly susceptible to infection by pathogenic bacteria. Botulinum toxins and scrape prion protein have been known to exploit M cells to gain entry and induce inflammation and infection (Matsumura et al. 2015). Therefore, given that Atoh8 null mice exhibited an increase in M cells, it is logical to reason that Atoh8 plays a role in regulating M cell numbers to limit the transcytosis of pathogenic agents. This delicate equilibrium of maintaining M cell density in the Peyer’s patch via Atoh8 and OPG may have been established over time as an evolutionary mechanism to control intestinal immune homeostasis with context to invasive antigens and mucosal immune responses. Atoh1, a family member of Atoh8, has been documented to regulate Paneth cell differentiation in the intestinal crypt via notch signaling and lateral inhibition (T. H. Kim et al. 2014). Atoh1 and notch signaling prevents the adjacent cell next to the Paneth cell from expressing Atoh1 and thereby differentiating into another Paneth cells; it can be speculated that Atoh8 employs a similar mechanism in maintaining the population. However, further investigative studies are needed to understand the mechanistic pathway with which the epithelial intrinsic Atoh8 regulates the population of M cells in the Peyer’s patches.
Our attempts to identify if canonical (RelA/p50) or non-canonical (RelB/p52) NF-κB signaling was responsible for Atoh8 expression revealed no changes in Atoh8 expression. Overexpression of RelA/p50 and RelB/p52 failed to induce or significantly upregulate Atoh8 expression without RANKL indicating the existence of another RANKL mediated pathway in M cell differentiation. Previous research studied Atoh8’s role in the differentiation of osteoblasts; in this context, Atoh8 was revealed to be induced by BMP signaling and was essential to regulate the
102
RANKL/OPG ratio indirectly via Runx2 to regulate osteoclast number and to maintain appropriate bone volume in mice (Yahiro et al. 2020). To confirm the presence of BMP signaling in M cells, Intestinal organoids isolated from wildtype mice and grown in RANKL exhibited significant upregulation of BMP2 and BMP6 signaling compared to organoids grown without RANKL. Next, we explored if Atoh8 expression was induced by BMP2 or BMP6, we grew intestinal organoids in maintenance condition (Egf, noggin, and R-spondin), M cell conditions (RANKL), in the presence of BMP2 and BMP6 (noggin was removed as they act an antagonist against BMP signaling) and organoids in EGF and R-spondin for control purposes.
We observed a significant increase in Atoh8 expression in both the organoid cultures that were grown with just BMP2 recombinant protein and BMP6 recombinant protein. The expression of Atoh8 in these organoids was significantly higher than organoids grown in RANKL alone. These combinations of experiments prove that besides RANKL induced NF-κB signaling in M cell differentiation, there also exists a RANKL-induced BMP signaling which is essential for activation of Atoh8. We further investigated to see if Runx2 played a role in BMP induced Atoh8 signaling, we did not observe any significant changes to Runx2 expression indicating that further investigation is needed to find if RANKL induced BMP signaling directly activates Atoh8 expression.