In analyzing the transcriptomic data of E10.5 embryos (Liang et al., 2021), we noted upregulation of macrophage-related genes, such as C1q1, Adgre1, and Mrc1, in Sorbs2 ^−/− mutants (Figure 1A). However, Sorbs2 is not expressed in macrophages but is highly expressed in embryonic hearts (Supplementary Figure S1). We hypothesized the upregulation of macrophage-related genes occurs within the heart. Since macrophages start to populate hearts as early as E9.5 (Epelman et al., 2014), we collected E12.5, E15.5 and E18.5 ventricles to perform RNA-seq (Supplementary Tables S1–3). We selected genes significantly downregulated in E12.5 mutant hearts [log₂ (fold change) <-0.58, p < 0.05] to perform GO analysis and found these genes enriched in pathways related to the electron transport chain and mitochondrial translation elongation (Figures 1B, C), suggesting that Sorbs2 positively regulates myocardial maturation. Using the same threshold, we selected genes significantly upregulated in E12.5 mutant hearts to perform GO analysis. Interestingly, these genes are enriched in pathways regulating immune response (Figure 1B). Upon closer examination, we observed differential expression of macrophage marker genes such as Cx3cr1 and Lyz2 (Figure 1C), suggesting that macrophages may be activated in Sorbs2 ^−/− hearts. These changes in gene expression patterns persisted throughout E18.5 (Figure 1C). Taken together, our RNA-seq results revealed that Sorbs2 deficiency impairs myocardial maturation and triggers a response in macrophages.

The increased expression of macrophage marker genes in Sorbs2 ^−/− embryonic hearts prompted us to examine macrophage numbers in mutant hearts. To this end, we performed whole-mount immunostaining on E12.5 hearts, using an antibody against the pan macrophage marker F4/80. Indeed, it revealed increased macrophages in the ventricles of E12.5 Sorbs2 ^−/− hearts (Figure 2A). In section analysis, macrophages were mainly distributed under the epicardium and in the outer layer of the myocardium (Figure 2B). Consistently, sections of Sorbs2 ^−/− heart displayed increased macrophage counts (Figures 2B, C).

Next, we used macrophage lineage-tracing to further validate this observation. Cx3cr1 is a marker of embryonic heart macrophage (Leid et al., 2016). We bred Cx3cr1 ^CreERT2 and Rosa26 ^mTmG alleles into Sorbs2 ^−/− mice. Tamoxifen-induced CreERT-mediated recombination in the Rosa26 locus led to EGFP expression in Cx3cr1-lineage macrophages, confirming increased macrophages in the ventricular walls of E12.5 Sorbs2 ^−/− hearts (Figures 2D, E). These data indicate that increased expression of macrophage-related genes results from the increased number of macrophages in Sorbs2 ^−/− hearts.

Macrophages are vital residents of the developing heart. During heart development, tissue resident macrophages regulate coronary vessel formation and lymphatic network development (Cahill et al., 2021; Leid et al., 2016). They are also essential for the developmental remodeling of cardiac valves (Shigeta et al., 2019). The increased macrophages in Sorbs2 ^−/− hearts led us to question whether macrophages might play an unknown role in the abnormal embryonic hearts.

To this end, we took a cell depletion approach with the Rosa26 ^DTA alleles, which encodes cytotoxic diphtheria toxin A (DTA) after Cre-induced recombination removes the STOP element, therefore killing cells that express DTA (Ivanova et al., 2005). Tamoxifen was administered to Cx3cr1 ^CreERT2/+ ; Rosa26 ^DTA/+ mice at E9.5 and E11.5 through oral gavage, and hearts were collected at E12.5 for F4/80 immunostaining to check the efficiency of macrophage depletion (Figures 3A, B). Results showed that macrophages were decreased in Cx3cr1 ^CreERT2/+ ; Rosa26 ^DTA/+ hearts (Figure 3C). The quantification indicated that macrophage numbers significantly decreased and the reduction ratio was about 40% (Figure 3D).

We obtained Cx3cr1 ^CreERT2/+ ; Rosa26 ^DTA/+ ;Sorbs2 ^−/− mice to examine the effect of macrophage depletion on Sorbs2 ^−/− hearts. We collected E12.5 ventricles for RNA-seq (Supplementary Table S4). Results showed that macrophage marker genes, such as Cx3cr1, Cd68 and Csf1r, were significantly reduced compared with control littermate embryos (Figure 3E), verifying reduced macrophages. Pathway analysis showed downregulated genes involved in organ regeneration, lipid metabolism, muscle development, and neurotransmitter signaling, while upregulated genes were associated with innate immune response activation, DNA damage-induced senescence, diabetic cardiomyopathy, and mitochondrial biogenesis (Figure 3F). These results suggest that macrophage depletion impairs cardiac metabolism and causes cardiomyocyte damage in mid-gestation stage. By E18.5, transcriptomic changes between macrophage-depleted and non-depleted groups were minimal (Supplementary Table S5), indicating a transient effect of our macrophage depletion strategy, though we noted continued downregulation in neural development genes and upregulation in DNA damage response genes (Figure 3F).

As previously reported, about 40%–60% Sorbs2 ^−/− mice died within 1 week after birth, with about 40% presenting atrial septal defect (ASD) (Zhang et al., 2016; Liang et al., 2021). We wondered whether the macrophage increase in Sorbs2^−/− hearts is an attempt to repair the structural defect caused by Sorbs2 deficiency. To this end, we collected embryos at E18.5 to check cardiac structure defects. Genotype distribution ratios were consistent with Mendel’s law, suggesting no embryo loss during embryonic development (Supplementary Table S6). In Tamoxifen-administered hearts, all 8 macrophage-depleted WT and Sorbs2 ^−/− hearts exhibited no structural abnormalities, whereas we observed ASD in Sorbs2 ^−/− hearts as reported previously (Figure 4A). None of Sorbs2 ^−/− hearts showed conotruncal defects, but we noted membranous ventricular septal defect (VSD) in both macrophage-depleted and non-depleted Sorbs2 ^−/− hearts (Figures 4B, C). Among 21 macrophage-depleted Sorbs2 ^−/− embryos, 8 showed ASD, 1 showed membranous VSD, and 1 showed muscular VSD (Figures 4B,C). In contrast, among the 14 macrophage-non-depleted Sorbs2 ^−/− embryos, 5 showed ASD and 1 showed membranous VSD, with one embryo exhibiting both ASD and VSD (Figure 4C). Compared to macrophage-non-depleted Sorbs2 ^−/− embryos, there was a trend toward increased penetrance of cardiac defects in macrophage-depleted Sorbs2 ^−/− embryos, though the difference was not significant (Figure 4C). This non-significant increase in cardiac defect penetrance, particularly the occurrence of one muscular VSD in macrophage-depleted Sorbs2 ^−/− embryos, suggests that macrophages might play a repairing role in cardiac development.

A previous report shows that cardiac resident macrophages are required for valve formation (Shigeta et al., 2019). We also noted an increase in the number of Cx3cr1-lineage macrophages in the endocardial cushions of E12.5 hearts (Figures 5A, B). Therefore, we harvested E18.5 embryos to perform a morphological analysis of valves. In histological sections, we did not observe any obvious morphological abnormality in mitral and tricuspid valves of both macrophage-depleted and non-depleted Sorbs2 ^−/− hearts (Figure 5C). To obtain a whole view of cardiac valves, we used light sheet fluorescence microscopy to reconstruct a three-dimensional visualization of valves. Surface rendering was applied to delineate and calculate the volumes of the mitral, tricuspid, pulmonary and aortic valves. We did not detect any obvious abnormality in cardiac valves (Figure 5D). Quantification of valve volume showed no significant differences in any of cardiac valves between macrophage-depleted and non-depleted Sorbs2 ^−/− groups, and nor in comparisons between Sorbs2 ^+/− and Sorbs2 ^−/− groups (Figures 5E, F).

The mouse strains utilized in this study comprise Sorbs2 ^- (Liang et al., 2021), Cx3cr1 ^CreERT2 (Xu et al., 2020), Rosa26 ^DTA (Ivanova et al., 2005) and Rosa26 ^mTmG (Muzumdar et al., 2007). The Cx3cr1 ^CreERT2 allele was a gift from Dr. Bo Peng’s lab (Fudan University, Shanghai, China). All strains were backcrossed with C57BL/6 to ensure the consistent genetic background. Tamoxifen was administered at E9.5 and E11.5 through oral gavage. Mice were maintained under specific pathogen-free conditions in the animal facility at Shanghai Children’s Medical Center. All animal procedures adhered to the guidelines set by the Institutional Animal Care and Use Committee of the Shanghai Children’s Medical Center, affiliated with the Shanghai Jiao Tong University School of Medicine.

For the preparation of frozen sections, dissected embryonic hearts were fixed in 4% paraformaldehyde for 20 min at 4°C, followed by equilibration in a 30% sucrose PBS solution overnight at 4°C.Subsequently, the hearts were embedded in 100% OCT compound within Cryomolds. The prepared blocks were immediately frozen at −80°C. The sample blocks were sectioned into 10 μm thin slices using a Leica CM3050S cryostat. For paraffin sections, the dissected embryonic hearts were fixed in 4% paraformaldehyde for 24 h at 4°C. The fixed hearts were rinsed with PBS, then subjected to a graded ethanol series (30%, 50%, 70%, 80%, 95%, 100%) for complete dehydration. Typically, each ethanol step was maintained for 30 min, followed by the embedding process. Upon completion of dehydration, the hearts were soaked in xylene for 30 min, followed by overnight paraffin infiltration. Finally, the hearts were processed for embedding in paraffin. The sample blocks were sectioned into 5 μm thin slices.

The sections were dewaxed with xylene and subsequently rinsed with ethanol. The sections were stained with a hematoxylin dye solution for a duration varying between 5 and 20 min, followed by a rinse with running water. The sections were then subjected to a 30-second differentiation process using a differentiation solution and subsequently rinsed with running water for 5 min. The slides were then immersed in eosin dye for 2 min. This was followed by conventional procedures for dehydration, clearing, and mounting.

The frozen sections were permeabilized in 0.5% Triton X-100/phosphate-buffered saline (PBS) for 20 min, followed by blocking in 3% bovine serum albumin/PBS for 1 h. The sections were then stained with F4/80 antibodies (1:400, ab16288; Abcam). The nuclei were subsequently stained with 4′,6-diamidino-2-phenylindole (DAPI). Fluorescent images were captured using a high-resolution fluorescence microscope.

Initially, the samples are fixed in 4% paraformaldehyde/phosphate-buffered saline (PFA/PBS), followed by a sequential dehydration and rehydration process. Subsequently, the samples are treated with proteinase K and then inactivated with hydrogen peroxide (H₂O₂). Thereafter, blocking is performed, followed by the incubation with primary and secondary antibodies, and subsequent multiple washes. Afterwards, the Elite ABC reagent is prepared and incubated. Following staining with DAB reagent, the samples are transferred to PBS, fixed again in 4% PFA, and finally dehydrated and stored in 100% methanol.

Total RNA of E12.5, E15.5 and E18.5 cardiac ventricles were isolated using TRizol reagent (Thermo Fisher Scientific; 15596018). Library preparation and transcriptome sequencing on an Illumina HiSeq platform were performed by Novogene Bioinformatics Technology Co., Ltd. to generate 100-bp paired-end reads. HTSeq v0.6.0 was used to count the read numbers mapped to each gene, and fragments per kilobase of transcript per million fragments mapped (FPKM) of each gene were calculated. We used FastQC to control the quality of transcriptome sequencing data. The expression level of each gene under different treatment conditions was obtained by HTSeq-count after standardization. The differentially expressed genes were analyzed by DESeq2 package (version 1.42.0). Functional enrichment of differentially expressed genes was analyzed on Metascape website. Heatmaps were created by the Pheatmap package (version 1.0.12) in R (version 4.3.1). scRNA-seq data for embryonic hearts (GSE150817) were retrieved from the Gene Expression Omnibus (GEO) database. Seurat toolkit (version 4.3.0) was used for scRNA-seq analysis. After data integration, batch effect elimination, normalization, and scaling, different cell populations were identified based on existing references. Gene expression was plotted using normalized read counts.

For three-dimensional visualization of the embryonic heart valves, E18.5 embryos were harvested in PBS, fixed overnight in 10% formaldehyde and 2.5% glutaraldehyde, rinsed twice in PBS, and dehydrated through a graded series of alcohol (50%, 75%, 90%, and 100% twice) for 30 min per step at room temperature. The hearts were then transferred into specially designed glass tubes containing 100 µL of BABB solution (1:2 benzyl alcohol: benzyl benzoate) for complete clearing. A custom-developed device was used to mount the heart within the Zeiss Lightsheet Z.1 microscope chamber filled with 87% glycerol (RI = 1.45). 3D images were captured using the 561 nm laser line and detection optics 5x/0.16 (n = 1.45). The reconstruction of the image stacks was analyzed with Imaris 10.0 software. Surface rendering was applied to delineate and calculate the volumes of the mitral and tricuspid valves.

Statistical significance was performed using a two-tailed Student’s t test, or nested ANOVA test as appropriate. Statistical significance is indicated by *, where p < 0.05, **, where p < 0.01, and ***, where p < 0.001.