iPSC-Derived Pancreatic Progenitors Lacking FOXA2 Reveal Alterations in miRNA Expression Targeting Key Pancreatic Genes

Recently, we reported that forkhead box A2 (FOXA2) is required for the development of human pancreatic α- and β-cells. However, whether miRNAs play a role in regulating pancreatic genes during pancreatic development in the absence of FOXA2 expression is largely unknown. Here, we aimed to capture the dysregulated miRNAs and to identify their pancreatic-specific gene targets in pancreatic progenitors (PPs) derived from wild-type induced pluripotent stem cells (WT-iPSCs) and from iPSCs lacking FOXA2 (FOXA2–/–iPSCs). To identify differentially expressed miRNAs (DEmiRs), and genes (DEGs), two different FOXA2–/–iPSC lines were differentiated into PPs. FOXA2–/– PPs showed a significant reduction in the expression of the main PP transcription factors (TFs) in comparison to WT-PPs. RNA sequencing analysis demonstrated significant reduction in the mRNA expression of genes involved in the development and function of exocrine and endocrine pancreas. Furthermore, miRNA profiling identified 107 downregulated and 111 upregulated DEmiRs in FOXA2–/– PPs compared to WT-PPs. Target prediction analysis between DEmiRs and DEGs identified 92 upregulated miRNAs, predicted to target 1498 downregulated genes in FOXA2–/– PPs. Several important pancreatic TFs essential for pancreatic development were targeted by multiple DEmiRs. Selected DEmiRs and DEGs were further validated using RT-qPCR. Our findings revealed that FOXA2 expression is crucial for pancreatic development through regulating the expression of pancreatic endocrine and exocrine genes targeted by a set of miRNAs at the pancreatic progenitor stage. These data provide novel insights of the effect of FOXA2 deficiency on miRNA-mRNA regulatory networks controlling pancreatic development and differentiation. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1007/s12015-023-10515-3.


Introduction
Forkhead Box A2 (FOXA2) is one of the earliest transcription factors (TFs) expressed during pancreatic development and remains to be expressed in all types of pancreatic cells [1]. During human pancreatic organogenesis, FOXA2 starts to be expressed at 4 weeks of gestation and continues onwards [2][3][4]. Previous studies demonstrated that FOXA2 controls the expression of several TFs and genes involved in pancreatic endocrine cell fate and β-cell functionality [5,6]. Using human pluripotent stem cells (hPSCs), we and others reported that FOXA2 plays very important roles during human pancreatic and hepatic development [7][8][9]. A recent study reported that heterozygous missense variants in FOXA2 can lead to monogenic diabetes [10]. Another study showed that in humans, risk alleles of type 2 diabetes (T2D) are associated with FOXA2-bound enhancers [11]. These findings indicate the contribution of FOXA2 defects in diabetes development and its important role during pancreatic endocrine differentiation.
Recent progress in human induced PSC (hiPSC) technology has paved the way for many essential applications that could be used for disease modeling, targeted therapy, drug screening, and precision medicine. Therefore, here, we take advantage of our recently established FOXA2 knockout hiPSC (FOXA2 -/-iPSC) model to identify the alterations in the miRNA and mRNA profiles in PPs lacking FOXA2 to understand the miRNA-mRNA regulatory networks regulating pancreatic development. Our results showed that loss of FOXA2 leads to the upregulation of numerous miRNAs targeting key PP genes involved in pancreatic exocrine and endocrine development.

Immunocytochemistry
Immunostaining was performed on differentiated iPSCs as previously reported [32,36]. Cells were washed once with PBS then 4% paraformaldehyde (PFA) was added on the cells for 20 min and placed on a shaker at room temperature. The cells were then washed with tris-buffered saline + 0.5% Tween 20 (TBST) thrice in a 10-minute interval on a shaker. Cells were then permeabilized for 15 min at room temperature using phosphate buffered saline (PBS) + 0.5% Triton X-100 (PBST) twice, later blocked overnight with 6% Bovine Serum Albumin (BSA) in PBST at 4 o C. Afterwards, guinea pig anti-PDX1 (1:500, ab47308, Abcam) and mouse anti-NKX6.1 (1:2000, F55A12-C, DSHB) primary antibodies diluted in 3% BSA in PBST were added to the cells and incubated overnight at 4 o C. Cells were washed three times with TBST and then Alexa Fluor secondary antibodies (ThermoFisher Scientific) diluted in PBS (1:500) were added for 1 h at room temperature then washed again three times using TBST. Cell nuclei were stained for two minutes with Hoechst 33,258 (DAPI) diluted 1:5000 in PBS (Life Technologies, USA). After washing three times with PBS, images were captured using inverted fluorescence microscope (Olympus).

Western Blotting
Total protein was extracted from one well of a 6-well plate using RIPA lysis buffer with protease inhibitor (ThermoFisher Scientific). Measurement of protein concentration was done using Pierce BCA kit (ThermoFisher Scientific). 20 µg of total protein were separated by SDS-PAGE and transferred onto PVDF membranes. Membranes were blocked with 10% skimmed milk in TBST then incubated with rabbit anti-FOXA2 (1:4000, #3143, Cell Signaling) and mouse anti-β-actin (1:10,000, sc-47,778, Santa Cruz) primary antibodies overnight at 4 o C. Membranes were washed with TBST then horseradish peroxidase-conjugated secondary antibodies (Jackson Immunoresearch) diluted in TBST (1:10,000) were added for 1 h at room temperature then washed again using TBST. Membranes were developed using SuperSignal West Pico Chemiluminescent substrate (Pierce, Loughborough, UK) then visualized using iBright™ CL 1000 Imaging System (Invitrogen).

RNA Extraction and RT-qPCR Analysis
1 × 10 6 cells were collected using 700 µL of TRIzol Reagent (Life Technologies) then total RNA extraction was performed using Direct-zol™ RNA Miniprep (Zymo Research, USA). For mRNA, cDNA was synthesized from 1 µg of RNA using SuperScript™ IV First-Strand Synthesis System following manufacturer protocol (ThermoFisher Scientific, USA). RT-qPCR was performed using GoTaq qPCR SYBR Green Master Mix (Promega, USA) run 1 3 in triplicates. Average Ct values were normalized to the WT samples for each gene tested. GAPDH was used as an endogenous control (primer details are listed in Supplementary Table 2).

Differential Gene Expression Analysis
Following manufacturer's protocol, NEBNext Poly(A) mRNA Magnetic Isolation Kit (NEB, E7490) was used for capturing mRNA using 1 µg of total RNA. Generation of RNA-sequencing (RNA-seq) libraries was done using NEBNext Ultra Directional RNA Library Prep Kit (NEB, E7420L) and libraries were sequenced using Illumina Hiseq 4000 system. Raw data were converted to FASTQ files using Illumina BCL2Fastq Conversion Software v2.20 while running quality controls in parallel. Pair-end FASTQ files were subsequently aligned to the GRCh38 reference genome using built-in module and default settings in CLC genomics workbench v21.0.5. Normalized expression data (TPM (transcripts per million)) mapped reads were sequentially imported into the AltAnalyze v.2.1.3 software for differential expression analysis as we described before. (https:// doi. org/ 10. 3390/ cance rs132 15350) For identifying DEGs, genes with log2 fold change (FC) > 1 and <-1 with a P-value < 0.05 were considered. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analyses were performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) [37].

Differential miRNA Expression and Potential Target Analysis
miRNA expression profiling was conducted on differentiated and collected PP total RNA samples from WT and FOXA2 -/-iPSCs. From the extracted total RNA, ~ 100 ng was used for miRNA library preparation following the manufacturer's instructions of the library kit (E7560S, New England BioLabs Inc., USA). The amplified cDNA constructs were purified using the Monarch PCR purification kit (Biolabs, New England). MicroRNA analysis was carried out in CLC genomics workbench 20.0 using built-in small RNA analysis workflow. miRNA count reads were normalized using the TMM (trimmed mean of M values) normalization method and log2 CPM (Counts per Million) values were subsequently subjected to differential analysis. A log2 FC > 1 with a P-value < 0.05 was used as a cutoff to determine the differentially expressed miRNA in FOXA2 -/-iPSCs versus WT-iPSCs. Pathway analysis and the microRNA target filter were employed to identify potential miRNA-mRNA networks using Ingenuity Pathway Analysis (IPA) software (QIAGEN, Germany).

Statistical Analysis
At least three biological replicates were used in most of the experiments, otherwise technical replicates were used for statistical analyses. Statistical analysis was performed using unpaired two-tailed student's t-test by Prism 8 software. Data are represented as mean ± standard deviation (SD).

Identification of Differentially Expressed Genes in iPSC-Derived Pancreatic Progenitors Lacking FOXA2
To investigate the effects of FOXA2 loss on mRNA and miRNA expression in PPs, we used two CRISPR/Cas9generated FOXA2 -/-iPSC lines with their isogenic controls (WT-iPSCs) as we recently reported [7]. Generated iPSCs were differentiated into PPs that co-express PDX1 and NKX6.1 using our established differentiation protocol ( Fig. 1A) [33]. The expression levels of FOXA2 in WT-and FOXA2 -/-PPs were validated at protein level using Western blotting where there was a clear absence of FOXA2 band in FOXA2 -/-PPs ( Fig. 1B). At PP stage, FOXA2 -/-PPs showed a significant decrease in the expression levels of the two key progenitor TFs, PDX1 and NKX6.1, as indicated by immunocytochemistry and RT-qPCR (Fig. 1C, D) which is concordant with our previously reported data [7].

Discussion
FOXA2 is an important TF that starts to be expressed at a very early stage of pancreatic development, where the first expression is detected at the definitive endoderm stage and continues in all stages. Our recent study showed that loss of FOXA2 during pancreatic differentiation of iPSCs prevents the formation of α-and β-cells [7]. However, there are currently no data available on the effects of FOXA2 deficiency on the expression pattern of miRNAs and their specific targets in PPs. Here, we provide evidence that FOXA2 deficiency is associated with significant alterations in the expression levels of miRNAs targeting key pancreatic genes at PP stage. The alterations in miRNA expression in PP derived from iPSCs lacking FOXA2 may reflect an impairment in pancreatic differentiation. A direct role for FOXA2 in regulating the expression of selected miRNAs warrants further investigation. PPs are characterized by the expression of several TFs and genes involved in directing the PPs into different types of pancreatic cells (endocrine, exocrine, and ductal cells). Our recent report showed that loss of one FOXA2 allele in iPSCs generated from a patient with FOXA2 haploinsufficiency significantly reduced the expression of pancreatic TFs involved in the development of endocrine pancreas [7]. In agreement with these findings, our RNA-Seq and RT-qPCR results showed loss of FOXA2 to significantly downregulate the expression of key endocrine-associated genes, such as PDX1, NKX6.1, NEUROG3, NEUROD1, NKX2.2, RFX6, GLIS3, HES6, ARX, PAX4, PAX6, MNX1, GATA6, FEV, INSM1, TCF7L2, GP2, and CHGA, which were targeted by several upregulated miRNAs in PPs lacking FOXA2. ARX and PAX4 are known to be essential for the formation of pancreatic α-cells and β-cells, respectively [38,39]. Furthermore, the downregulated genes associated with exocrine and ductal cell specification such as PTF1A, CPA1, CPA2, SOX9, GATA4, and ONECUT1 were also targeted by several upregulated miRNAs, indicating that FOXA2 is not only essential for pancreatic endocrine development, but also plays an important role in pancreatic exocrine and ductal development. Many of those downregulated genes are associated with diabetes and pancreatic development. These findings indicate that lack of FOXA2 negatively impacted the iPSC differentiation into exocrine and endocrine pancreas through downregulating the expression of essential pancreatic developmental genes.
miRNAs are known to play essential roles in post-transcriptional regulation through targeting mRNAs [40,41]. The role of miRNAs in regulating pancreatic β-cell development and function has been previously reported [12]. However, limited studies have tackled the role of miRNAs in regulating PP development. In the current study, we noticed that most DEmiRs had several predicted targets in PPs. On the other hand, most of key pancreatic targets were predicted targets for at least two DEmiRs. For example, we found that miR-184 expression level was upregulated and among its predicted targets is NKX6.1, which is a key TF in pancreatic endocrine development and later becomes restricted to pancreatic β-cells [42,43]. It has been reported that miR-184 participates in regulating β-cell expansion and negatively correlates with insulin biosynthesis and secretion [44][45][46].
We have also identified an upregulation of miR-9-5p in our FOXA2 -/-PPs, targeting PDX1, ONECUT1, and ARX, which were significantly downregulated. Previous studies have linked the upregulation of miR-9 cluster of miRNAs with glucose-stimulated insulin secretion impairments [26]. This cluster has also been identified as a regulator of insulin exocytosis and secretion machinery through modulating Sirt1 expression [52]. A previous study showed that miR-9 targets Onecut2 and decreases its mRNA expression in pancreatic β-cells, which subsequently leads to an increase in Onecut2 downstream target, granuphilin (a negative regulator of the insulin exocytosis) [26]. miR-124a (i.e., a precursor for miR-124-3p/5p) is expressed in human islets and has been reported to be associated with T2D. It has been found that miR-124a represses important target genes involved in pancreatic β-cell function and insulin secretion [53], including Foxa2 and Pdx1 [54]. Although our analysis did not show NKX6.1 as a predicted target for miR-124-5p, it has recently been reported that it induces pancreatic β-cell differentiation by regulating NKX6.1 expression [55]. In this study, miR-124-3p from the same miRNA cluster, was significantly increased in FOXA2 -/-PPs and its predicted targets were the downregulated pancreatic TFs, NEUROG3, PROX1, RFX6, GATA6 as well as NEUROD1, SOX9, and FOXA2, which were experimentally validated in previous research [53,56]. Our results showed miR-92a-2-5p among the upregulated DEmiRs that targets eight downregulated key pancreatic TFs including FOXA2, NKX6-1, FEV, NKX2-2, ONECUT1, SOX9, HNF1B, and TCF7L2. A recent study found that miR-92a-2-5p regulates insulin production and pancreatic β-cell apoptosis [57,58]. Our results showed increased expression of miR-577 upon FOXA2 loss. Previous studies showed that  4). Data are represented as mean ± SD; ***p < 0.001 ◂ miR-577 inhibits pancreatic β-cell activity and survival by targeting FGF21, which promotes β-cell function and survival through AKT signaling pathway [59,60]. miR-204 was found to be associated with the endocrine part of pancreatic islets and insulin regulation [61,62]. miR-15a-5p was also found to regulate insulin production by suppressing UCP-2 gene expression, a mitochondrial anion carrier that reduces oxidative stress [63], resulting in more insulin biosynthesis [64]. On the other hand, upregulation of miR-146a/b has been found to increase cytokine-induced β-cell apoptosis [65]. From these results, we speculate that the lack of FOXA2 at PP stage can cause alterations of several miRNAs important for pancreatic β-cell development and function from an early stage of pancreas development before reaching the mature β-cell stage, causing the cells to follow a different trajectory  miRNAs have been also identified as epigenetic modifiers that regulate gene expression levels without targeting its mRNA sequence but by targeting important enzymes including DNA methyltransferases (DNMTs), histone methyltransferases (HMT), and histone deacetylases (HDACs) [66,67]. In addition, miRNAs are posed to epigenetic modification and regulation such as DNA methylation and RNA/histone modifications. The interchangeable relationship between miRNAs and epigenetic modifications forms the bases of miRNA-epigenetic feedback loop that can affect cellular processes [68], physiological functions and disease conditions [69]. Recently, a study has discovered that FOXA2 physically interacts with ten-eleven-translocation methylcytosine dioxygenase 1 (TET1) in which β-cell specification is significantly hindered upon TET1 loss [70]. This lays a good example of TF crosstalk with epigenetic regulators in regulating pancreatic β-cell differentiation and specification. Our miRNA-seq data identified several DEmiRs which have been previously associated with epigenetic modifications in different tissue samples [71,72]. We predict that the lack of FOXA2 does not only affect miRNAs regulating other genes, but also affects miRNAs regulating epigenetic modifications that can directly affect histone modification for accessing DNA for transcription. It was previously found that alterations in circulating miRNA expression occur in diabetic patients in which they can be even be used as biomarkers for diabetes prediction and progression [73,74]. miRNAs can also be used as biomarkers for pancreatic cancer progression and prognosis [75]. Furthermore, alterations in gene regulation by miRNA can be the cause some forms of pancreatic cancers as some miRNAs can act as oncogenes and are associated with poor disease prognosis [76]. Another important aspect of miRNAs is that they can serve as potential therapeutic agents for regenerative medicine [77,78]. Recent advances in research have led to the development of miRNA delivery systems to regulate gene expression [77]. Therefore, our identified DEmiRs may serve as potential novel biomarkers or therapeutic modulators for diabetes or pancreatic cancer diseases. However, further functional validation is required to provide a proof-of-concept for the link between identified miRNAs and diseases.
In conclusion, we showed that FOXA2 loss led to dysregulation of several miRNAs and mRNAs expressed in iPSC-derived PPs. Our findings demonstrated that FOXA2 is not only crucial for endocrine islet development, but also it is essential for exocrine pancreas development. Integrating miRNA and mRNA profiling results revealed that the potential targets of DEmiRs identified in this study are known to play an essential role in pancreatic development and function. These data provide proof of the regulatory relationship between pancreatic TFs and miRNAs in controlling the expression of main pancreatic differentiation drivers during pancreatic islet differentiation. Also, the data presented here would serve as the platform for future studies focusing on understanding the function of identified DEmiRs. In addition, further understanding of miRNA-mRNA and miRNAepigenetic feedback loop would help in identifying potential novel therapeutic strategies and targets that are not limited to FOXA2 mutations but include cancer and regenerative medicine.
Funding Open Access funding provided by the Qatar National Library. This work was funded by grants from Qatar Biomedical Research Institute (QBRI) (Grant No. IGP3 and QBRI-HSCI Project 1).

Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations
Ethical Approval Not applicable.

Competing Interests
The authors declare no competing interests in this manuscript.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.