Hans Clevers' team from the Hubrecht Institute (with Dr. Lin Lin as the first author and co-corresponding author) published an article titled "Unbiased transcription factor CRISPR screen identifies ZNF800 as master repressor of enteroendocrine differentiation" in Science. The research team used CRISPR gene editing technology to conduct a comprehensive knockout screen of all transcription factors (TFs) in adult human small intestinal organoids, aiming to identify the core transcription factors regulating enteroendocrine cell (EEC) differentiation. This pioneering screen successfully identified ZNF800, a previously uncharacterized repressive transcription factor. ZNF800 exhibits a significant inhibitory effect on enteroendocrine cell differentiation by directly acting on an endocrine transcription factor network centered on PAX4. This discovery not only deepens our understanding of ZNF800's function but also strongly demonstrates Organoid model combined with CRISPR screening technology has tremendous potential in identifying genes that determine cell fate.

Background

Enteroendocrine cells (EECs) are hormone-secreting cells found in the epithelium of the stomach, small intestine, and colon. Like all other epithelial lineages, EECs originate from regionally specific Lgr5+ intestinal stem cells. EECs regulate various aspects of metabolic activity by sensing nutrients and releasing hormones, including insulin levels, satiety, and gastrointestinal secretion and motility. EECs can be further classified based on the hormones they produce, such as enterochromaffin cells (ECs) that secrete 5-hydroxytryptamine, L cells that secrete glucagon, I cells that secrete cholecystokinin, D cells that secrete somatostatin, N cells that secrete neurotensin, G cells that secrete gastrin, and M/X cells that secrete motilin and growth hormone-releasing peptide. The differentiation process from stem cells to different EEC subtypes is highly regulated, but the mechanisms underlying the generation of different lineages of enteroendocrine cells are not fully understood.

Many specific transcription factors (TFs) in the intestinal tissue form regulatory networks to regulate the expression of downstream genes, thereby regulating the differentiation of enteroendocrine cells. Through systematic unbiased identification of transcription factors (TFs) regulating the differentiation of human enteroendocrine cell (EEC) lineages, it was found that ZNF800 is a master transcriptional repressor that plays a leading role in the directed differentiation of endocrine lineages. ZNF800 directly inhibits the endocrine TF regulatory network (its target genes include the core TFs in this network: INSM1, SOX4, NEUROG3, etc.), thereby inhibiting the differentiation of intestinal stem cells into EECs; ZNF800 also directly inhibits PAX4, driving a biased differentiation trajectory toward enterochromaffin cells by directly controlling the endocrine TF network centered on PAX4.

Results

1. TFome-wide CRISPR knockout screen of the human SI organoid genome

The research team combined organoid technology with high-throughput CRISPR gene screening technology to unbiasedly and systematically identify transcription factors that regulate the differentiation of human EECs. The researchers used a CRISPR gene knockout library based on a lentiviral vector for screening. This library contained 7210 sgRNAs targeting 1800 human transcription factors and 100 sgRNA negative controls. In a triple fluorescent reporter system established in human intestinal organoids (CHGA-IRES-iRFP670 labeling EECs, MUC2-mNeongreen labeling goblet cells, and DEFA5-IRES-dsRed labeling Paneth cells), the authors enriched differentiated EECs, goblet cells, and other cells with negative reporter genes, and further assessed the insertion level of sgRNAs in the genome by sequencing. The authors first confirmed the traditionally known TFs that control EEC differentiation (NEUROG3, INSM1, SOX4). By enriching the expression levels of EEC cells and goblet cells after knockout of these genes (as shown in Figure 1B). Notably, the reported NEUROG3, SOX4 and INSM1 NFIC, TEF and ZHX2 affect EEC cell differentiation because knocking them out in organoids did show a slight reduction in CHGA EECs, but the team discovered new strongly regulating EEC differentiation TFs: GFI1 and ZNF800.

Figure 1 Transcriptome knockout using CRISPR knockout screening in human gastrointestinal organoids

2. ZFN800 inhibits EEC differentiation

First, knockout of the GFI1 gene in human organoids, while it can inhibit the formation of goblet cells and lead to an increase in the number of enteroendocrine cells (EECs), is not the optimal choice. Using CRISPR screening technology, the research team identified ZFN800 as the strongest inhibitor of EEC differentiation. ZNF800 −/− organoids contained increased EECs and strongly reduced the number of goblet and Paneth cells (Figures 1D and 1E). Transmission electron microscopy further confirmed the absence of goblet and Paneth cells, characterized by apical secretory vesicles, and increased EECs with basolateral secretory granules (Figure 1F).

To further confirm this result, the research team performed single-cell sequencing on cells sorted from WT organoids and ZNF800 −/− organoids, mixed at a ratio of 80% CHGA- cells and 20% CHGA+ cells. Cluster analysis revealed no significant differences between the two in terms of cloning efficiency and cell cycle progression, indicating that ZNF800 does not affect the ISC homeostasis of organoids. Meanwhile, the results showed a significant reduction in the number of goblet cells (2.3%) in ZNF800 −/− organoids compared to WT organoids. Notably, early EECs and ECs were increased in ZNF800 −/− organoids, while other EEC subtypes were depleted (Figures 2C and 2D). This indicates that ZNF800 knockout induces EEC differentiation at the expense of goblet cells.

Re-expression of WT ZNF800 protein in ZNF800 −/− organoids using doxycycline (dox) (Figure 2E) effectively rescued goblet and Paneth cell differentiation while inhibiting EEC lineage commitment (Figure 2F). Notably, the biased differentiation pattern of EECs was also reversed after ZNF800 restoration, leading to reduced serotonin secretion, a hallmark hormone of ECs (Figure 2G).

Figure 2 Phenotypic characterization of ZNF800-deficient cells

3. ZNF800 inhibits the endocrine TF regulatory network

To elucidate the regulatory mechanisms of regulatory factors in EEC cells, the research team used scRNA-seq data sets on differentially expressed genes in WT and ZNF800−/− organoid CHGA+ cell populations for single-cell regulatory network inference and clustering (SCENIC) analysis, identifying 249 activators of EEC differentiation, showing higher activity, including SOX4, NEUROD2, and PAX4 (Figure 3A). Using antibody chromatin immunoprecipitation sequencing (ChIP-seq), the team revealed the distribution of ZNF800 in wild-type (WT) organoids and rescued ZNF800-deficient organoids. In WT organoids, 11,565 significant ZNF800 binding consensus peaks (q-value <0.01) were identified. In contrast, this number was reduced to 7085 in rescued ZNF800-deficient organoids. Notably, most ZNF800 binding sites were closely located within the ±5 kb region of the transcription start site (TSS), further highlighting the potential role of ZNF800 in gene transcription regulation. Further, differential gene expression (DGE) analysis (P-value <0.05) explored the effect of ZNF800 on the expression of its target genes. The analysis revealed that after knocking out ZNF800, 461 (15%) of the ZNF800-bound genes were significantly upregulated, while only 59 (2%) genes were significantly downregulated (Figure 3A and Figure 4B). This result strongly suggests that ZNF800 mainly acts as a transcriptional repressor in gene expression regulation.

Figure 3 ZFN800 involved in endocrine differentiation

4. PAX4 directly drives EC differentiation downstream of ZNF800

Through the above research, the research team found that ZNF800 and GFI1 were identified as key inhibitors of enteroendocrine cell (EEC) differentiation in the small intestine in a CRISPR screen. To further understand its mechanism of action, the research team constructed a double knockout model of GFI1 and ZNF800 in a human small intestinal organoid model. The experimental results showed that in organoids lacking both GFI1 and ZNF800, the number of goblet cells and Paneth cells was further reduced, while the number of EEC cells was significantly increased. However, it is noteworthy that the single knockout of GFI1 did not lead to a shift in the differentiation direction of intestinal epithelial cells (EC), but instead slightly reduced the number of EC cells and slightly increased the proportion of L cells (PYY positive).

Given the expression characteristics of GFI1 in goblet cells and Paneth cells, and the significant effects of ZNF800 knockout on these cell types, we further investigated whether there is a synergistic effect between GFI1 and ZNF800. By detecting the expression of GFI1 in sorted goblet cells in wild-type and ZNF800 knockout organoids, we found that the expression level of GFI1 was not significantly different (Figure 4A and Figure 4B), indicating that ZNF800 and GFI1 act independently as inhibitors of EEC differentiation.

When using anti-ZNF800 antibody chromosome technology for gene ontology enrichment analysis, we identified NSM1, NEUROG3, and PAX4 as transcription factors (TFs) with the highest ZNF800 binding activity. To further study the target genes with stronger functional significance, the authors selected three genes, Insm1, Sox4, and Pax4, and performed knockout experiments in ZNF800 knockout organoids. The experimental results showed that the knockout of these three genes significantly inhibited the induced differentiation of EECs, but the degree of influence varied. In particular, the knockout of Insm1 almost completely reversed the EEC-biased differentiation caused by ZNF800 knockout.

Further research found that simultaneous knockout of INSM1 or SOX4 based on ZNF800 knockout resulted in a slight decrease in the number of CHGA-positive EC cells (TPH1-positive), and the differentiation of L cells (GCG and PYY-positive) and D cells (SST-positive) was also further inhibited. However, these manipulations did not reverse the phenotype after ZNF800 knockout. In contrast, PAX4 knockout showed a trend of robust differentiation of EECs into other EEC subtypes and reversed the increased EEC differentiation and EC-biased differentiation after ZNF800 knockout. By detecting the hormones secreted by L cells and M/X cells, we further confirmed this function of PAX4. Therefore, it can be considered that PAX4 is a direct target of ZNF800.

In addition, the research team also noted differences in the regulatory mechanisms of EC differentiation between humans and mice. In mice, EC differentiation is completely dependent on PAX4, while ARX is crucial for the differentiation of L, G, I, and N enteroendocrine cell types. Based on this finding, it is speculated that there may be an interaction between PAX4 and ARX. To this end, they downregulated the expression of Arx in ZNF800 knockout organoids and upregulated the expression of Arx in PAX4 knockout organoids and ZNF800/PAX4 double knockout organoids (Figure 4D). The experimental results showed that re-expression of PAX4 in ZNF800/PAX4 double knockout organoids can significantly inhibit the expression of ARX and redirect EEC differentiation to an EC-biased trajectory. However, ChIP-qPCR experiments showed that unlike ZNF800, ARX did not bind to the enhancer region upstream of PAX4. These research results provide new perspectives for in-depth understanding of the regulatory mechanisms of intestinal endocrine cell differentiation.

Figure 4 PAX4 as a direct downstream target of ZNF800 is responsible for the EC-biased differentiation trajectory

Literature Sources

Lin Lin et al., Unbiased transcription factor CRISPR screen identifies ZNF800 as master repressor of enteroendocrine differentiation. Science 382, 451-458 (2023). DOI: 10.1126/science.adi2246