Intricate MIB1-NOTCH-GATA6 Interactions in Cardiac Valvular and Septal Development

Genome-wide association studies and experimental mouse models implicate the MIB1 and GATA6 genes in congenital heart disease (CHD). Their close physical proximity and conserved synteny suggest that these two genes might be involved in analogous cardiac developmental processes. Heterozygous Gata6 loss-of-function mutations alone or humanized Mib1 mutations in a NOTCH1-sensitized genetic background cause bicuspid aortic valve (BAV) and a membranous ventricular septal defect (VSD), consistent with MIB1 and NOTCH1 functioning in the same pathway. To determine if MIB1-NOTCH and GATA6 interact in valvular and septal development, we generated compound heterozygote mice carrying different Mib1 missense (Mib1K735R and Mib1V943F) or nonsense (Mib1R530X) mutations with the Gata6STOP/+ heterozygous null mutation. Combining Mib1R530X/+ or Mib1K735R/+ with Gata6STOP/+ does not affect Gata6STOP/+ single mutant phenotypes. In contrast, combining Mib1V943F/+ with Gata6STOP/+ decreases the incidence of BAV and VSD by 50%, suggesting a suppressive effect of Mib1V943F/+ on Gata6STOP/+. Transcriptomic and functional analyses revealed that while the EMT pathway term is depleted in the Gata6STOP/+ mutant, introducing the Mib1V943F variant robustly enriches this term, consistent with the Mib1V943F/+ phenotypic suppression of Gata6STOP/+. Interestingly, combined Notch1 and Gata6 insufficiency led to a nearly fully penetrant VSD but did not affect the BAV phenotype, underscoring the complex functional relationship between MIB1, NOTCH, and GATA6 in valvular and septal development.


Introduction
Bicuspid aortic valve (BAV) is the most prevalent congenital heart defect, affecting up to 2% of the population, and is characterized by two asymmetrical leaflets instead of the three leaflets found in normal aortic valves [1,2].BAV individuals often do not show symptoms and are usually diagnosed incidentally unless BAV is linked to significant aortic valve dysfunction, including stenosis and regurgitation [3][4][5][6].Importantly, BAV disease in younger patients is associated with aortic medial defects that result in a higher prevalence and accelerated rate of aortic dilatation, increasing the risk of lethal dissection or rupture [7,8].Aortic valve development entails the formation of endocardial cushions, their cellularization through epithelial-to-mesenchymal transition (EMT), and remodeling by mesenchymal compaction.Concomitant alignment and fusion of the endocardial cushion mesenchyme is required to ensure cardiac septation [9,10].
Leaflet orientation in BAV varies relative to the coronary arteries, resulting in different classifications, which depend on the number of raphe [11,12].In rodents, two main types of BAV have been described, involving the right and left leaflets (R-L) and, more commonly, the right and non-coronary leaflets (R-N) [13].Mechanistically, RL-BAV may be caused by defective cardiac neural crest cells (CNCCs) patterning of the outflow tract (OFT) [13].RNtype BAV may result from defective endocardial cushion formation, presumably because of impaired EMT [14].
BAV is inherited in a dominant manner, with incomplete penetrance and variable expressivity [2,15].Genome-wide association studies have identified chromosomal regions 9q, containing NOTCH1, and 18q, harboring GATA6 and MIB1, as linked to BAV [16].Mouse Gata6 and Mib1 genes are closely associated within 300 kb of DNA in a region of conserved synteny with the human locus 18q11.2region, which is frequently implicated in cardiac abnormalities.
GATA6 is a key transcription factor that plays a crucial role in various developmental processes, including initiating the EMT during mesoderm development and promoting the migration of CNCCs [17,18].GATA6 mutations have been reported in a wide spectrum of CHD, including BAV [19][20][21].Gata6 haploinsufficiency causes BAV [20,22], whereas deletion in CNCCs causes aortic arch defects, suggesting a requirement in vascular smooth muscle cell differentiation [23,24].
GATA6 and MIB1's close physical proximity on chromosome band 18q11 suggests their involvement in shared developmental mechanisms.In this study, we investigated the genetic interactions between Gata6 and Mib1 and found a complex functional relationship between GATA6 and MIB1-NOTCH, which provides valuable insights into the genetic determinants of BAV, VSD, and related cardiac anomalies.

Mouse Strains
Mouse strains used in this study are as follows: [36].Genotyping primers are shown in Table S1.

Tissue Processing for Histological Procedures
For histological procedures, whole embryos or torsos were fixed in 4% paraformaldehyde (PFA, Electron Microscopy Sciences, 50980487, Hatfield, PA, USA) overnight at 4 • C. Paraffin-embedded embryos/torsos were cut into 7 µm sections.Hematoxylin and eosin (H&E) staining was performed according to standard protocols.

Microscopy and Confocal Imaging
Brightfield imaging was performed using an Olympus BX51 Microscope (Tokyo, Japan) and Olympus cellSense software (Version number 1.7).Images were processed in Adobe Photoshop Creative Suite 5.1.

Statistical Analysis
Sample sizes, statistical tests, and p-values are specified in the corresponding figure legends and corresponding subsections of Section 2. For comparisons between two groups, a mean ± SD is represented, and an unpaired two-tailed Student's t-test was performed.For experiments comparing two groups of categorical variables, a mean ± SD per group is represented, and a Chi-Square test was performed.Differences were considered statistically significant at a p-value < 0.05 (Table S2).Statistical analysis and graphical representation were performed using GraphPad Prism, version 8.

RNA-Seq
RNA was isolated from E11.5 Mib1 V943F/+ Gata6 +/STOP and control OFT samples.Samples were distributed in three pools of four pairs of OFT per genotype.Tissue was homogenized with a pestle mechanical homogenizer, and RNA was extracted with an Arcturus PicoPure RNA Isolation kit (Thermo Fisher Scientific, Norristown, PA, USA, KIT0214).RNA libraries were prepared using the NEB Next Ultra II Directional RNA Library Prep Kit (Ipswich, MA, USA) and sequenced in a Nextseq 2000 Illumina sequencer (San Diego, CA, USA) using a 60 bp single-end elongation protocol.Sequenced reads were QC and pre-processed using cutadapt v1.18 [37] to remove adapter contaminants.The resulting reads were aligned, and gene expression was quantified using RSEM v1.2.3 [38] over mouse reference GRCm38 with Ensembl genebuild.Differential gene expression was analyzed with the EdgeR R package (v3.32.1 on R 4.0.3)[39].Genes with 1 count per million (cpm) in at least 3 samples were defined as expressed and retained for later analysis.Counts were normalized by the TMM method.Differential gene expression was tested using a generalized linear model as implemented in the EdgeR package.Genes showing altered expression with an adj.p-value < 0.05 were considered differentially expressed.The set of differentially expressed genes was used for functional analysis with Ingenuity Pathway Analysis Software (version 111725566, Qiagen-IPA, Redwood City, CA, USA) [40], where we used a Benjamani-Hochberg adjusted p-value < 0.05 for significance.Gene Set Enrichment Analysis (GSEA) was performed with GSEA on the complete set of expressed genes against the Hallmark term database.A NOM p-value < 0.1 was used to select for significantly enriched gene sets.
We performed a comparative Gene Set Enrichment Analysis (GSEA) against "HALL-MARK" gene sets, which are collections of predefined genes representing fundamental, well-defined biological processes [43].Gene sets differentially enriched between Mib1 V943F/+ Gata6 +/STOP and Gata6 STOP/+ mutants and respective controls [22] were broadly similar, although several changes were exacerbated in compound heterozygotes (Figure 2C).For example, "UV RESPONSE DOWN", "INFLAMMATORY RESPONSE," and "PROTEIN SECRETION" were more significantly enriched, and "MYOGENESIS", "ADIPOGENESIS", and "FATTY ACID METABOLISM" were more significantly depleted (Figure 2C).However, enriched pathways evoking a proliferative defect like "G2M CHECKPOINT" and "KRAS SIGNALING UP", and depleted metabolic pathways like "OXIDATIVE PHOSPHORY-LATION" and "GLYCOLYSIS" did not deviate significantly.Importantly, "EPITHELIAL MESENCHYMAL TRANSITION" ("EMT"), the key process by which the endocardial cushions become cellularized by endocardial and neural crest-derived mesenchyme, was depleted in the Gata6 STOP/+ mutant but enriched in the Mib1 V943F/+ Gata6 +/STOP mutant (Figure 2C,D).These data indicate that the Mib1 V943F/+ mutation partially suppresses the Gata6 haploinsufficient phenotype by enhancing EMT./+ Gata6 +/STOP vs. control contrast.Significantly downregulated and upregulated genes (adjusted p-value < 0.05) are labeled in blue and red, respectively.Non-differentially expressed genes are labeled in gray.(B) Circular plot highlighting selected Ingenuity Pathway Analysis diseases and function terms enriched in the Mib1 V943F/+ Gata6 +/STOP genotype.Red dots, upregulated genes in the pathway.Blue dots, downregulated genes in the pathway.The height of the inner circle section represents the Benjamani-Hochberg (B-H) p-value < 0.05 (higher is more significant), and enrichment z-score values are color-coded from positive (red) to negative (blue).(C) Bubble plots show 24 enriched statistically significant Hallmark gene sets in Gata6 STOP/+ vs. control and Mib1 V943F/+ Gata6 +/STOP vs. control contrasts by Gene Set Enrichment Analysis (GSEA).Negative logarithm of NOM p-value < 0.1 is represented by the size of the bubble (bigger is more significant).Normalized enrichment score (NES) is color-coded from positive (red) to negative (blue).Low-color density bubbles represent the Hallmark gene set categories that are not statistically significant in this specific comparison.(D) Gene enrichment profiles for the "EMT" gene set in the comparison of Gata6 STOP/+ (NOM p-value = 0.0590; NES = −1.28)and Mib1 V943F/+ Gata6 +/STOP (NOM p-value = 0.0001; NES = 2.02).The expression intensity is plotted on a scale from red (upregulated) to blue (downregulated).(E) H&E staining on aortic valve and ventricular chamber sections in controls Notch1 KO/+ , Gata6 STOP/+ , and Gata6 STOP/+ ; Notch1 KO/+ mice at E16.

Discussion
Genetic interactions analysis aims to determine if two different genes participate in the same developmental process by combining mutations in these genes and examine if the resulting phenotype diverges from that caused by individual mutations [44].In this study, we examined whether the combination of Mib1-Notch1 and Gata6 mutations had a more profound impact on valvular and septal development than the loss of Gata6 alone.Previously, we observed that Mib1 mutations affect valvular and septal development only in a sensitized, Notch-deprived genetic background [28,32].We identified three types of interactions between Mib1-Notch1 and Gata6 mutations: 1. Neutral interaction: Mib1 K735R/+ Gata6 +/STOP and Mib1 R530X/+ Gata6 +/STOP mutants show BAV and VSD frequencies similar to those of GATA6 STOP/+ heterozygotes; 2. Synergistic interaction: Notch1 KO/+ ; Gata6 STOP/+ mutants exhibit a higher than expected VSD frequency compared to Gata6 STOP/+ heterozygotes alone.3. Partially suppressive interaction: Mib1 V943F/+ Gata6 +/STOP mutations show lower BAV and VSD frequencies than Gata6 STOP/+ heterozygotes.These findings suggest that the interactions between MIB1-NOTCH1 and GATA6 are complex and may result in different biological outcomes, such as endocardial cushion formation and epithelial-mesenchymal transition (EMT), depending on the developmental context and/or temporal-spatial expression.
Our data may reflect differences due to the nature of the mutations analyzed and the NOTCH dosage, which is less stringent in the context of valvular and septal development in mice compared to humans.Specifically, the Mib1 R530X and Mib1 K735R mutations are predicted to be loss-of-function (null) and dominant-negative, respectively.These mutations contribute to the manifestation of valvular and septal phenotypes when combined with a Notch1 mutation, indicating that cardiac phenotypes are sensitive to Mib1-Notch1 dosage [28,32].This interpretation is supported by the nearly fully penetrant VSD phenotype (95%) observed in compound Notch1 KO/+ ; Gata6 STOP/+ heterozygotes, compared to the 40% VSD frequency in Notch1 KO/+ mutants.
Our finding that heteroallelic Notch1 KO/+ ; Gata6 STOP/+ haploinsufficiency does not exacerbate the BAV phenotype underscores the more significant role of GATA6 in the dosage-dependent control of aortic valve development.This is further supported by the high mutational burden of GATA6 in human CHD [45].Moreover, GATA6 is a key driver in the early differentiation of the second heart field [46] and the recruitment of CNCC lineages [24].These processes collectively shape the cardiac outflow tract (OFT), potentially acting upstream of NOTCH inhibitory functions during early cardiogenesis.
Our finding that Mib1 V943F/+ partially restores the wild-type valvular and septal phenotype when combined with Gata6 STOP is intriguing, especially considering the expected negative effects from NOTCH loss-of-function interactions.Although Mib1 V943F has previously been characterized as a dominant negative mutation [31], its phenotypic expression still requires Notch1 sensitization [28].The underlying mechanism of this suppression involves the restoration of EMT in Mib1 V943F/+ Gata6 +/STOP/+ mice, which aligns with the known roles of GATA6 and NOTCH in promoting this process.
The mechanism underlying the compensatory effect on the valvular and septal phenotypes of the Mib1 V943F/+ mutation when combined with Gata6 STOP is unclear.Quantitative PCR experiments indicate that Gata6 mRNA expression in the Mib1 V943F/+ Gata6 +/STOP mutants is the predicted 50%, suggesting a post-transcriptional mechanism of compensation.GATA6 interacts with GATA4 [47][48][49], GATA factors, and other cardiac factors, ie.TBX5 and/or NKX2.5 [50].Compensation mechanisms could be explored by quantifying the protein expression levels of GATA6 and interacting partners in the compound heterozygotes.
The limitations of our study are that Gata6 haploinsufficiency has a disproportionately large effect size on the BAV phenotype alone and that heterozygous Mib1 mutations fail to produce phenotypes in mice unless they are sensitized by a second NOTCH mutation.Additional insight might be obtained from conditional null disruption experiments affecting the tissue-specific functions of GATA6 and NOTCH.
Animal studies were approved by the CNIC Animal Experimentation Ethics Committee and by the Community of Madrid (Ref.PROEX 155.7/20).All animal procedures conformed to EU Directive 2010/63EU and Recommendation 2007/526/EC regarding the protection of animals used for experimental and other scientific purposes, enforced in Spanish law under Real Decreto 1201/2005.

Figure 2 .
Figure 2. Mib1 V943F partially restores EMT in Gata6 STOP/+ mice.(A) Volcano plot of the transcripts detected by RNA-seq of Mib1 V943F/+ Gata6 +/STOP vs. control contrast.Significantly downregulated and upregulated genes (adjusted p-value < 0.05) are labeled in blue and red, respectively.Non- 5. The asterisk (*) indicates the position of the leaflets.Arrowheads indicate VSD.rv, right ventricle.lv, left ventricle.(F) Quantification of BAV and VSD percentages, respectively.Statistical significance was determined by the Chi-Square test. V943F