Interferon hyperactivity impairs cardiogenesis in Down syndrome via downregulation of canonical Wnt signaling

Summary Congenital heart defects (CHDs) are frequent in children with Down syndrome (DS), caused by trisomy of chromosome 21. However, the underlying mechanisms are poorly understood. Here, using a human-induced pluripotent stem cell (iPSC)-based model and the Dp(16)1Yey/+ (Dp16) mouse model of DS, we identified downregulation of canonical Wnt signaling downstream of increased dosage of interferon (IFN) receptors (IFNRs) genes on chromosome 21 as a causative factor of cardiogenic dysregulation in DS. We differentiated human iPSCs derived from individuals with DS and CHDs, and healthy euploid controls into cardiac cells. We observed that T21 upregulates IFN signaling, downregulates the canonical WNT pathway, and impairs cardiac differentiation. Furthermore, genetic and pharmacological normalization of IFN signaling restored canonical WNT signaling and rescued defects in cardiogenesis in DS in vitro and in vivo. Our findings provide insights into mechanisms underlying abnormal cardiogenesis in DS, ultimately aiding the development of therapeutic strategies.


INTRODUCTION
Down syndrome (DS), caused by trisomy of human chromosome 21 (HSA21) (T21), is a leading cause of intellectual and developmental disability, affecting 1 in 700 newborns. [1][2][3] Individuals with DS are predisposed to various medical conditions including congenital heart defects (CHDs), autoimmune disorders, autism spectrum disorders, and leukemia. [1][2][3][4] Approximately 50% of newborns with DS are affected by CHDs including ventricular septal defects (VSD), atrial septal defects (ASD), and atrioventricular septal defects. 5 CHDs are a primary and significant risk factor for mortality in people with DS through age twenty. [6][7][8] Despite this extremely high incidence of heart defects in DS, the underlying mechanisms are poorly understood. 3 Intensive effort has been focused on identifying chromosomal regions of HSA21 responsible for CHDs. [9][10][11][12] HSA21 is homologous to portions of at least three murine chromosomes: MMU16, MMU17, and MMU10 and various mouse models of DS have been created with triplication of these regions and smaller subregions. [13][14][15][16][17][18][19] Trisomy for MMU10 or MMU17 does not cause CHDs in these mouse models. 16 However, 50% of mice with trisomy of MMU16, such as Dp(16)1Yey/+ (hereafter referred to as Dp16), display CHDs, 17 resembling cardiac defects in people with DS. Using mouse strains with trisomy of MMU16, Liu et al. suggested a 3.7 Mb genomic region from Tiam1 to Kcnj6 was sufficient to cause heart defects in mice. 12 This 3.7 Mb genomic region contains a cluster of four interferon (IFN) receptor (IFNR) genes: Ifnar, Ifnar2, Ifngr2, and Il10rb. In addition, genetic variants in human IFNGR2 and IL10RB have been associated with CHDs in DS from two cohort studies composed of 198 individuals and 702 individuals, respectively. 20,21 In the Dp16 mouse model, we recently showed that normalization of copy number of the Ifnr cluster prevented heart malformations, 22 but the mechanisms by which hyperactive IFN signaling may disrupt normal heart development await elucidation.
Studies in mouse and human pluripotent stem cells have established a hierarchical model of cardiac lineage differentiation and specification of cardiac progenitors precisely regulated by specific signaling Since genes contained on HSA21 are spread over three syntenic regions on MMU16, MMU17, and MMU10, it is difficult to create a mouse model to fully recapitulate the effects of T21. Recently, human models using induced pluripotent stem cells (iPSCs) derived from individuals with DS have emerged to study pathogenesis of various medical conditions caused by T21. [38][39][40][41][42][43] Bosman et al. generated one line of human embryonic stem cells (hESC) from a donated embryo with T21, and two control hESC lines from siblings, and another hESC line from an unrelated embryo with T21. 43 These four hESC lines were differentiated into embryoid bodies (EBs) and gene expression in EBs was measured at various differentiation stages by real-time PCR (qPCR) and bulk RNA-sequencing (RNA-seq). This study identified hundreds of genes that were dysregulated in cardiomyocyte clusters differentiated from T21 hESC lines. However, the mechanisms driving such dysregulated cardiac expression programs were not elucidated in this report. 43 Here, using a combination of a human iPSC-based model and Dp16 mice, we demonstrate that hyperactivation of IFN signaling downstream of increased dosage of IFNR genes downregulates canonical Wnt signaling and leads to cardiogenic dysregulation in DS. We differentiated human iPSCs derived from individuals with DS and co-occurring CHDs (DS/CHD iPSCs), as well as from healthy donors (control iPSCs) into cardiac cells. Increased IFN signaling, decreased canonical Wnt pathway, and impaired cardiac differentiation were observed during differentiation of DS/CHD iPSCs. Binding to IFN ligands to IFNRs induces JAK/ STAT signaling, and markers of elevated JAK/STAT signaling were observed in DS/CHD iPSCs at the protein and transcriptome levels. Normalization of IFN signaling with a JAK inhibitor (JAKi) restored canonical Wnt signaling and rescued defects in cardiac differentiation of DS/CHD iPSCs. JAKi treatment normalized the canonical Wnt pathway and prevented heart malformations in Dp16 embryos. These data from both in vitro and in vivo studies define a role for a signaling axis involving hyperactive IFN signaling and decreased Wnt signaling in cardiogenic defects in DS, while also revealing avenues for potential therapeutics.

Trisomy 21 impairs cardiac differentiation of iPSCs
To characterize cardiogenesis in DS, we derived iPSCs from three individuals with DS and co-occurring CHDs (DS/CHD iPSCs), and from three healthy donors (control iPSCs) using non-integrating modified mRNAs and miRNAs 44 (Table S1). T21 in individual iPSC lines was confirmed by karyotyping (Table S1 and Figure S1A). Pluripotency of these human iPSC lines was validated by assessing the expression of pluripotent stem cell markers ( Figure S1B). To prevent the reversion of T21 to disomy of chromosome 21 (D21), we performed all experiments using iPSCs undergoing less than 20 passages. Karyotypic analysis of individual iPSC lines with the highest passage numbers in this study demonstrated correct copy numbers of chromosome 21 in these iPSC lines ( Figure S1A and Table S1).
Next, we investigated whether human D21 iPSCs could model cardiac lineage differentiation by analyzing expression of lineage-specific markers using a monolayer-based protocol as shown previously [45][46][47] (Figure S2A). RNA-seq analysis of control iPSC lines (C62 and C68) indicated that all cardiac lineage markers were absent in undifferentiated iPSCs. In contrast, MESP1, a pre-cardiac marker, reached its highest expression at differentiation day 3. ISL1 and NKX2-5, markers for cardiac progenitors, reached peak expression at days 5-7. Markers for cardiac lineages including endothelial cells, smooth muscle cells, and cardiomyocytes were activated beginning on day 7 ( Figure S2B). We next investigated whether DS/ CHD iPSCs exhibit defects during cardiac differentiation. We induced cardiac differentiation on two control lines (C62 and C68) and two DS/CHD lines (D7 and D49) (Table S1), and monitored differentiation efficiency by qPCR analysis of NKX2-5 and ISL1 expression at day 7. Expression of these cardiac progenitor markers was significantly downregulated in DS/CHD cells, compared to control cells ( Figure S2C). We specifically examined cardiac differentiation by observing the initiation of spontaneous beating. Spontaneous beating clusters in control iPSC cultures appeared on day 9 G 1 post-induction. However, spontaneous beating clusters in DS/CHD iPSC cultures did not appear until day 12 G 1 post-induction. Weaker and slower calcium transients were also observed in beating clusters differentiated from DS/CHD iPSCs, compared to controls ( Figures 1A and S3), consistent with the finding that beating clusters in DS/CHD iPSC cultures exhibited weak and slow contraction (Videos S1 and S2  iScience Article cultures was distinct from DS/CHD cultures ( Figure 1B). We then performed Gene Ontology (GO) analysis on differentially expressed genes (DEGs) between control and DS/CHD cells. Top GO terms associated with downregulated DEGs in DS/CHD cells involved cardiac differentiation and heart development (Figure 1C). In contrast, upregulated genes in DS/CHD cells were generally associated with non-cardiac development, such as nervous system development ( Figure 1C). Notably, we observed maximal numbers of contractile clusters in control and DS/CHD cultures on differentiation day 13. Therefore, we analyzed late cardiogenesis in iPSC-derived cells on day 13. Similar to transcriptomic results on day 7, RNA-seq demonstrated that patterns of gene expression in control cultures (C62 and C68) were distinct from DS/CHD cultures (D7 and D49) on differentiation day 13 ( Figure 1D). GO analysis of the DEGs demonstrated that downregulated DEGs in DS/CHD cells are predominantly associated with heart development events ( Figure 1E). Upregulated DEGs in DS/CHD cells are predominantly associated with non-cardiac development, such as neurogenesis ( Figure 1E). We also observed some upregulated DEGs associated with cardiac development ( Figure 1E). Taken together, these data indicate that T21 dysregulates cardiogenic gene expression along with defects in cardiac differentiation of DS/CHD iPSCs.

T21 induces hyperactive IFN signaling during cardiac cell differentiation
We then sought to identify specific genes on HSA21 and signaling pathways that could contribute to impaired cardiogenesis in the iPSC model. Given that impaired cardiac differentiation of DS/CHD iPSCs was observed on day 7 ( Figures 1B and 1C), we hypothesized that dysregulation of signaling pathways that cause defects in cardiac differentiation must occur at earlier time points. We performed gene set enrichment analysis (GSEA) 48,49 of RNA-seq data generated on day 0 and day 3. GSEA demonstrated that genes upregulated in DS/CHD cells were enriched in IFN-inducible genes (ISGs) (Table S2, Figures 1F, S4A, and S4B). qPCR analysis showed that IFNRs encoded on HSA21, especially the type I receptors IFNAR1 and IFNAR2, were significantly upregulated in DS/CHD cells during early cardiac differentiation of iPSCs ( Figure S4C). Recently, hyperactivation of IFN signaling was consistently observed in multiple immune and non-immune cell types from people with DS. [50][51][52][53] Additionally, genomic sequencing suggested that genetic variants in HSA21 genes encoding IFNRs may be associated with CHDs in DS. 20,21 Furthermore, in the Dp16 mouse model, normalization of IFNR copy number prevented heart malformations. 22 Importantly, Dp16 embryonic hearts display transcriptome changes indicative of IFN hyperactivity that are dependent on IFNR triplication. 22 However, it is unclear whether hyperactivation of IFN signaling also occurs during human cardiogenesis in DS. The IFN signaling cascade is comprised of three combinations of receptors and downstream effectors. 54 Binding of diverse IFNs to specific receptors activates tyrosine kinase 2, Janus kinase 1 (JAK1), and/or JAK2, which results in phosphorylation, dimerization, and nuclear translocation of signal transducer and activator of transcription (STAT) factors. Immunoblotting analysis demonstrated that phosphorylation of STAT1 (p-STAT1) was significantly upregulated on day 3 when DS/CHD iPSCs lost pluripotency and began to differentiate into MESP1 + cardiovascular progenitor cells ( Figure S2B and Figures 1G and 1H). Taken together, these results demonstrate that IFN signaling is hyperactivated during cardiac differentiation in DS/CHD iPSCs.

Inhibition of IFN signaling improves cardiac differentiation of DS/CHD iPSCs
Next, we sought to determine whether inhibition of IFN signaling was sufficient to rescue defects in human cardiogenesis in vitro. First, we normalized IFN receptor levels using shRNAs to knockdown specific receptors. We first focused on IFNAR1 and IFNAR2 because their expression in DS/CHD iPSC-derived cells increased more than IFNGR2 and IL10RB, compared to healthy euploid controls ( Figure S4C). After we transduced iPSC lines with either scramble shRNA or shRNAs targeting IFNAR1 and IFNAR2, we selected stable iPSC lines wherein the expression of IFNAR1 and IFNAR2 was comparable to control cells (Figure S5A). Normalization of IFNAR1 and IFNAR2 levels in DS/CHD iPSCs decreased the activity of IFN signaling by decreasing p-STAT1 levels ( Figure S5B) and rescued defects in cardiac differentiation by increasing cardiac gene expression ( Figures S5B and S5C). We next normalized IFN-JAK signaling by treating differentiating DS/CHD iPSCs with 1 mM of tofacitinib, a JAK1/3 inhibitor (hereafter referred to as JAKi). This dose of JAKi normalized the activity of the IFN-JAK-STAT pathway in DS/CHD iPSC-derived cells, as assessed by immunoblotting analysis of p-STAT1 levels ( Figures S6A, 2A, and 2B). Treatment with JAKi also rescued defects in cardiac differentiation of DS/CHD iPSCs by increasing cardiac gene expression by qPCR ( Figure 2C), resulting in increasing numbers and strength of contractile clusters (Videos S3, S4, and S5). Spontaneously beating clusters appeared on differentiation day 9.5 in control iPSC cultures, but until differentiation day 11.5 in DS/CHD iPSC cultures. However, spontaneously beating clusters appeared in DS/ CHD iPSC cultures treated with JAKi on differentiation day 9.8, comparable to control cultures ( Figure S6B).   Figure 2E). GO analysis of these overlapping DEGs revealed an enrichment of GO terms associated with cardiac development majorly in downregulated genes in DS/CHD cells, and an enrichment of GO terms associated with nervous development in upregulated genes in DS/ CHD cells. This may indicate that T21 might affect cell adhesion, cell communication, and cell fate specification during heart development ( Figures 1E and 2F). Importantly, multiple cardiac gene sets altered in DS/CHD differentiating cells were significantly rescued by JAKi treatment (Figures 2F and S6C), indicating that inhibition of IFN signaling effectively reversed abnormalities in cardiac differentiation of DS/CHD iPSCs in vitro.

IFN hyperactivity impairs early cardiac differentiation via downregulation of Wnt signaling
Downregulated genes in DS/CHD cells on differentiation day 7 are significantly associated with cardiac development terms ( Figure 1C). To further interrogate the mechanisms responsible for abnormal cardiac development in DS, we investigated signaling pathways that are dysregulated in cardiac differentiation of DS/CHD iPSCs, using REACTOME analysis of genes significantly downregulated by > 1.5-fold in DS/CHD cells on differentiation day 3. This analysis demonstrated that genes related to Wnt signaling were profoundly downregulated in DS/CHD cultures ( Figure 3A). Activation of canonical Wnt/b-catenin signaling (hereafter referred to as Wnt signaling) is essential for early cardiogenesis. [29][30][31][32][33][34][35][36][37] Therefore, we examined Wnt signaling by analyzing b-catenin activation through dephosphorylation at Ser37/Thr41. Active b-catenin was substantially reduced in DS/CHD iPSC-derived cells on differentiation day 3 ( Figures 3B and 3C), indicating reduced Wnt signaling. Next, we examined whether increasing Wnt signaling activity could rescue the phenotype of cardiac differentiation of DS/CHD iPSCs. CHIR99021, a GSK-3 inhibitor, has been used to induce Wnt signaling in monolayer-based cardiac differentiation of human iPSCs. [45][46][47] Our standard differentiation protocol uses 6 mM of CHIR99021 from day 0 to day 1 for monolayer-based iPSC differentiation. Treatment with 10 mM of CHIR99021 from day 0 to day 1 did not significantly alter the ratio of active b-catenin/total b-catenin in control cells, but significantly increased active b-catenin levels as well as the ratio of active b-catenin/total b-catenin in DS/CHD cells ( Figures 3B and 3C).
To exclude potential effects of CHIR99021 on the Wnt pathway in DS cells, we differentiated iPSCs without adding any Wnt manipulators utilizing an EB-based differentiation protocol wherein the activity of Wnt signaling reached the peak at day 4. 55 Immunoblotting revealed that Wnt signaling activity and expression of CPC marker ISL1 decreased in DS/CHD cells at day 4 compared to control cells ( Figures 3D and 3E). CHIR99021 treatment increased the activity of the canonical Wnt pathway and expression of cardiac gene ISL1 in DS/CHD cells ( Figures 3D and 3E). Moreover, monolayer-based cardiac differentiation of DS/CHD iPSCs in the presence of 10 mM of CHIR99021 is comparable to control iPSCs, as shown by appearance time of beating clusters ( Figure 3F), amplitude and frequency of Ca 2+ transients ( Figures 3G, S3A, and S3B), and contractile frequency and strength (Videos S6 and S7). Interestingly, treatment with 10 mM CHIR99021 did not dramatically increase the activity of Wnt signaling and cardiogenesis in control iPSCs ( Figures 3B-3F). Taken together, these studies using both monolayer-based differentiation system and EB-based differentiation system demonstrate that normalization of the Wnt pathway, which is significantly downregulated during early differentiation of DS/CHD iPSCs, reverses defects in cardiac differentiation.
To further investigate molecular mechanisms by which activation of Wnt signaling rescued cardiac differentiation of DS/CHD iPSCs, we performed RNA-seq analysis to profile global gene expression in cells differentiated from control and DS/CHD iPSCs (C62 and C68 versus D7 and D49) with lower or higher activity of canonical Wnt signaling. We performed transcriptomic analysis on differentiation day 7.  Figures 3H and S7). These data indicate that reduced Wnt signaling contributes to defects in cardiac differentiation in DS in vitro.
Next, we investigated the interplay between IFN hyperactivity and reduced Wnt signaling during cardiogenesis in DS/CHD iPSCs. IFNa treatment of control cells significantly increased the activity of IFN signaling and dramatically decreased the activity of Wnt signaling during the differentiation of iPSCs ( Figures S8A-S8D). Hyperactivated IFN signaling in control cells significantly decreased the frequency and strength of Ca 2+ transients of beating clusters to the level of DS/CHD (Figures S3A-S3B, and S8E), indicating the detrimental effect of IFN hyperactivity on cardiac differentiation of iPSCs. These data indicate that IFN hyperactivity is sufficient to cause T21 phenotypes. Strikingly, the normalization of IFN signaling by JAKi not only ameliorated cardiac differentiation of DS/CHD iPSCs but also restored Wnt signaling ( Figures 3I and 3J). In contrast, normalization of the Wnt signaling pathway ameliorated cardiac differentiation of DS/CHD iPSCs ( Figures 3B-3H) but did not significantly affect IFN signaling ( Figures 3K and 3L). In other words, hyperactive IFN signaling did not impair cardiac differentiation of DS/CHD iPSCs once the Wnt pathway was normalized in these cells.
Altogether, these data indicate that IFN hyperactivity impairs in vitro cardiogenesis in DS by downregulation of Wnt signaling.

JAK inhibition restores Wnt signaling and prevents cardiogenic dysregulation in Dp16 mice
Next, we examined the potential of JAKi to normalize cardiogenesis in DS in vivo using the Dp16 mouse model. CHDs are observed in 50% of people with DS, as well as 50% of Dp16 mice. 17 We hypothesized that the activity of IFN signaling could be highly variable in Dp16 embryonic hearts, potentially contributing to CHD occurrence. We performed immunoblotting for p-STAT1 in developing hearts from wild type (WT) and Dp16 at E15.5. Interestingly, p-STAT1 levels were higher but variable in Dp16 versus WT ( Figures 4A  and 4B), which could potentially explain the variable penetrance of CHDs in Dp16 mice. Next, we sought iScience Article to determine whether normalization of the IFN pathways by JAKi could prevent heart defects in Dp16 mice. We crossed Dp16 male mice to WT females. Pregnant females were treated with JAKi daily at a dose of 10 mg/kg/day beginning on 6.5 days after conception ( Figure S9). This JAKi dosage was determined based on previous studies, 56,57 and normalized levels of p-STAT1 in Dp16 hearts at E15.5 ( Figures 4C and 4D).
The mouse heart achieves its prenatal morphogenesis with the development of 4-chamber and septation by E15.5. 23,58 Therefore, we examined heart morphology at E15.5 by H&E staining of serial sections (Figure S10). We observed heart defects including ASD, VSD, and ASD/VSD, in 72% of Dp16 embryos, compared to only 15% of WT embryos at E15.5 ( Figures 4E and 4F), similar to a previous report wherein the occurrence of heart defects in Dp1Tyb with trisomy of chromosome 16 and WT mice at E14.5 is 62% and 27%, respectively. 59 Strikingly, JAKi treatment significantly decreased the occurrence of heart abnormalities in Dp16 mice at E15.5 from 72% to 29%, without affecting WT embryos ( Figure 4F), indicating that pharmacological attenuation of IFN signaling prevents abnormalities during heart development in Dp16 mice. To investigate the underlying mechanisms by which JAKi rescues defects in Dp16 heart development, we profiled global gene expression in WT and Dp16 hearts at E15.5 treated with or without JAKi using RNAseq. DEGs between Dp16 hearts with high p-STAT1 levels but not Dp16 hearts with low p-STAT1 levels as shown in Figure 4B and WT hearts are significantly associated with terms of IFN signaling, cardiac development, and metabolism ( Figure 4G), suggesting a crucial role of IFN signaling hyperactivation in the development of heart defects in Dp16 mice. These terms altered in the Dp16 heart were dramatically rescued by JAKi treatment (Figure 4G), indicating that normalization of IFN signaling effectively reversed abnormal gene expression programs associated with heart malformations in DS in vivo.
The Wnt pathway is activated during early heart development in mice (e.g. E8.5-E10.5). 34,37 Immunofluorescence assays demonstrated that the activity of Wnt signaling was significantly downregulated in Dp16 mice at E9.5 ( Figure 4H). These in vivo data are consistent with our previous findings that the Wnt pathway is downregulated during early cardiac differentiation of DS/CHD iPSCs (Figures 3A-3H). Treatment with JAKi, which prevented abnormal heart development in Dp16 mice ( Figures 4E-4G), also normalized the activity of Wnt signaling in Dp16 developing hearts at E9.5 ( Figures 4H and 4I).
Therefore, our in vitro and in vivo data strongly support that IFN hyperactivity impairs cardiogenesis in DS by dysregulating the Wnt pathway ( Figure 4J).

DISCUSSION
CHDs are very frequent in children with DS, with a prevalence of 50% compared to less than 1% in typical children. The mechanisms by which T21 causes cardiogenic dysregulation remain elusive. Using a combination of the human iPSC-based model and a mouse model for DS, Dp16, we illuminated a pathway involving increased dosage of IFNRs encoded on chromosome 21 and downstream downregulation of iScience Article Wnt signaling as a driver of defects in cardiogenesis in DS. Triplication of chromosome 21 causes IFNR overexpression concurrent with hyperactive IFN signaling and downregulated Wnt signaling during cardiogenesis in DS. Normalization of IFN signaling restored Wnt signaling and rescued defects in cardiogenesis in DS in vitro and in vivo. Normalization of the Wnt pathway rescued cardiogenesis, while not affecting the activity of IFN pathways. Therefore, our studies demonstrate that hyperactivation of IFN signaling causes abnormal cardiogenesis in DS by dysregulation of the Wnt pathway ( Figure 4J).
CHDs are the most common developmental abnormality in newborns, with a prevalence of 0.9%. 60 It is estimated that more than 2 million people are living with CHDs in the USA. 61 Increased levels of inflammatory cytokines have been observed in children with CHDs, 62,63 indicating that dysregulation of inflammatory pathways may impact fetal heart development. Analysis of genomic sequences in several hundred individuals with DS identified a few genetic variants in the IFNR genes located on HSA21. 20,21 These studies indicate crucial roles of inflammatory pathways in the risk of CHDs. However, the underlying mechanisms are not well known. We found that hyperactivation of IFN signaling is a causative factor of impaired cardiac differentiation of DS/CHD iPSCs (Figures 2 and S5) and cardiogenic dysregulation in Dp16 mice ( Figure 4). Furthermore, treatment of pregnant mice with JAKi improved heart development in Dp16 embryos ( Figures 4E-4G). Our findings that treatment with a JAKi rescued defects in human iPSC cardiogenesis (Figure 2) and Dp16 mouse heart development ( Figures 4E-4G) serve as a proof of concept that pharmacological intervention could treat heart malformations in DS. The JAKi used in this study is able to suppress all three types of IFN signaling, as they all employ JAK1 for signal transduction. IFN-mediated immune responses are central to host defense against infection 64 and attenuating type I IFN pathway may increase the risk of maternal infections during pregnancy. 65 Therefore, it is necessary to conduct further studies focusing on distinguishing targets of IFN signaling that are drivers of heart defects from those that are essential for host defense against microbial infection. Building upon these studies, we would be able to selectively suppress detrimental aspects of IFN signaling in DS.
CHD survivors are at greater risk of cardiovascular disease such as cardiac arrhythmia and heart failure, 23,66,67 but the underlying mechanisms remain elusive. Circulating pro-inflammatory cytokines are significantly increased in patients with heart failure, [68][69][70] as well as in people with DS. 51 IFN signaling pathways play distinctive roles in cardiovascular disease. Knockout of Ifng exaggerated cardiac hypertrophy in systolic and diastolic heart failure, 71,72 indicating a protective role of the IFNg-mediated type II IFN signaling in cardiac pathogenesis. However, treatment with an IFNAR-neutralizing antibody improved cardiac function and survival post-myocardial infarction, 73 suggesting a deleterious effect of the type I IFN pathway in response to myocardial infarction. Knockdown of IFNAR1/2 attenuated defects in cardiac differentiation of DS/CHD iPSCs ( Figure S5). Further studies focusing on determining the association of individual IFN pathways with the risk of CHDs and cardiomyopathy/heart failure in people with DS could identify targets for future therapy.
The Wnt signaling pathway is highly conserved among species. Temporal regulation Wnt signaling is required for cardiac development. Activation of the Wnt pathway is required for the expansion of CPCs during early development stages. [29][30][31][32]34 Decreased activity of the canonical Wnt pathway occurred during early differentiation of DS/CHD iPSCs (Figures 3A-3E) and early heart development in Dp16 embryos ( Figures 4H and 4I). Treatment with CHIR99021, a GSK-3 inhibitor that enhances the activity of canonical Wnt signaling, improved cardiac differentiation of DS/CHD iPSCs ( Figures 3B-3H and Videos S6 and S7). Highly potent and selective GSK-3 inhibitors have been investigated to treat neurodegenerative disorders in preclinical and clinical studies. 74 The Wnt pathway is essential for cardiovascular, neuronal, craniofacial, and musculoskeletal development. The development of cardiovascular, neurological, and musculoskeletal systems is particularly affected in individuals with DS. [1][2][3] The Wnt pathway was dramatically downregulated in the entire Dp16 embryo at E9.5, restored by the JAKi treatment ( Figure 4H), indicating that the downregulation of Wnt signaling could contribute to DS hallmarks, across multiple organ systems. Therefore, our findings hold the potential for multidimensional therapeutic endpoints in DS by modulation of Wnt signaling.
Crosstalk between the Wnt pathway and IFN signaling has been shown in previous studies. 75 iScience Article However, increasing the activity of Wnt signaling using a GSK3 inhibitor did not affect the activity of IFN signaling ( Figures 3K and 3L). These findings reveal that hyperactivation of IFN signaling causes defects in cardiogenesis in DS by downregulating the activity of the Wnt pathway. Recently, Lana-Elola et al. 59 identified a 4.9 Mb genome region (from Mir802 to Zbtb21) that when present in 3 copies, was sufficient to cause cardiac abnormalities. This 4.9 Mb genomic region overlapping with the previously identified 3.7 Mb region 12 contains the Dyrk1a gene, which functions as an inhibitor of Wnt signaling in some contexts. 77 Future studies focusing on defining mechanisms by which IFN signaling hyperactivation downregulates Wnt signaling would greatly aid in discovering effective therapies to treat DS.

Limitations of the study
Using human iPSCs derived from healthy donors and individuals with DS and CHDs, we reveal that hyperactivation of the IFN pathways causes cardiogenic defects in DS by downregulating the Wnt signaling pathway. Studies using the Dp16 mouse model also demonstrate that IFN hyperactivity downregulates the Wnt pathway and disrupts cardiogenesis in DS in vivo. The iPSC platform provides a tool to study molecular and cellular mechanisms to control human cardiogenesis. However, the monolayer-based (Figure S2) or EB-based ( Figures 3D, 3E, and 3L) cardiac differentiation of iPSCs cannot completely mimic human heart development in vivo. More powerful and faithful models like the human heart organoid platform would better understand heart defects in people with DS.
Our current study focuses on signaling pathways that dysregulate cardiogenesis in DS. It is likely that defects in cardiogenesis eventually lead to CHDs at birth. However, this study cannot directly address these questions: (1) whether IFN hyperactivity causes CHDs by downregulating the Wnt pathway, (2) how IFN hyperactivity causes many types of CHDs in DS, [1][2][3] and (3) whether JAKi can be used to prevent CHDs in DS. Our future studies on examining heart defects in multiple mouse models of DS (e.g., Ts65Dn, 78 Tc1, 79 and Dp16) around and after birth will be performed to address these questions.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:   Data and code availability d RNA-seq data related to this study have been deposited at GEO and are publicly available as of the date of publication. Accession numbers are listed in the key resources table.
d Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
d This paper does not report original codes.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
All animal research complied with the protocols (#00319) approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Colorado Anschutz Medical Campus. Dp(16Lipi-Zbtb21) 1Yey/J (hereafter ''Dp16'') mice, a mouse model of DS/CHD, were purchased from the Jackson Laboratory (Cat# JAX:013530, RRID:IMSR_JAX013530). Both male and female mice at E15.5 were randomly assigned to experimental groups. The phenotype described in the study is gender independent.
Human iPSC lines were obtained from the Crnic Institute Human Trisome Project (HTP), under a study protocol approved by the Colorado Multiple Institutional Review Board (COMIRB 15-2170, NCT02864108, see also www.trisome.org). Detailed information, including gender and age on human iPSC lines in this study is presented in Table S1.

Culture and characterization of human iPSCs
Human iPSCs (hiPSCs) were cultured in mTeSR1 in a 12-well plate coated with MatriGel (Corning). mTeSR1 was refreshed daily. The pluripotency of the hiPSCs was examined by examination of the expression of pluripotency markers Nanog, SSEA-4 and TRA-2-49. Karyotyping was performed by WiCell Research Institute and Molecular Pathology Shared Resource Cytogenetics at the University of Colorado for G-banding assay and Thermo Fisher Scientific for KaryoStat+ TM Assay.
Monolayer cardiac differentiation of hiPSCs into CMs hiPSCs were dissociated with Accumax (STEMCELL Technologies) and replated in a 24-well plate precoated with MatriGel. Cells were incubated in 1 mL of mTeSR1 per well for 4 days before induction. GSK3 Inhibitor CHIR99021 (Cayman) was applied at induction day 0 to day 1 in RPMI1640 (Life Technologies) with B27 supplement minus insulin (Life Technologies). Cells were incubated with IWP2 (Thermo Fisher Scientific) from day 3 to day 5. On day 5, cells were incubated in fresh RPMI1640 with B27 supplement minus insulin. Cells were then cultured in RPMI1640 with B27 (Life Technologies). For JAKi treatment, 1 mM Tofacitinib (LC Laboratories) was added to the culture medium from differentiation day 0 to day 3. For IFNa treatment, 10 ng/mL IFNa (R&D Systems) was added to the culture medium from differentiation day 0 to the end of the experiment.

Embryoid body differentiation of hiPSCs
Embryoid bodies (EBs) were generated from hiPSCs by adapting the methods reported in a previous study 55 with modifications. Briefly, hiPSCs were seeded on AggreWell 800 Plates (STEMCELL) with UM/ F-medium supplied with 10 mM Y-27632 (Enzo Life Sciences). UM/F-medium composed of DMEM/F12 (Thermo Fisher Scientific) containing 20% Knockout Serum Replacement (Thermo Fisher Scientific), 13 MEM nonessential amino acids (Thermo Fisher Scientific), 1x GlutaMAX (Thermo Fisher Scientific) and 0.1 mM b-mercaptoethanol (Thermo Fisher Scientific). hiPSCs were cultured with UM/F-medium supplied with 10 mM Y-27632 for 5 days (medium refreshed every two days) to allow the formation of spheroids. On day 6, EBs were resuspended in fresh UM/F-medium supplied with 10 mM Y-27632 and transferred to regular cell culture plates. On the next day (differentiation day 1), EBs were cultured with EB20 medium composed of DMEM/F12 containing 20% FBS (Thermo Fisher Scientific), 13 MEM nonessential amino acids, 1x GlutaMAX and 0.1 mM b-mercaptoethanol. For CHIR99021 treatment, 6 mM CHIR99021 was supplemented to EB20 medium from differentiation day 2 to day 4.