Dynamics of Chromatin Opening across Larval Development in the Urochordate Ascidian Ciona savignyi

Ascidian larvae undergo tail elongation and notochord lumenogenesis, making them an ideal model for investigating tissue morphogenesis in embryogenesis. The cellular and mechanical mechanisms of these processes have been studied; however, the underlying molecular regulatory mechanism remains to be elucidated. In this study, assays for transposase-accessible chromatin using sequencing (ATAC-seq) and RNA sequencing (RNA-seq) were applied to investigate potential regulators of the development of ascidian Ciona savignyi larvae. Our results revealed 351 and 138 differentially accessible region genes through comparisons of ATAC-seq data between stages 21 and 24 and between stages 24 and 25, respectively. A joint analysis of RNA-seq and ATAC-seq data revealed a correlation between chromatin accessibility and gene transcription. We further verified the tissue expression patterns of 12 different genes. Among them, Cs-matrix metalloproteinase 24 (MMP24) and Cs-krüppel-like factor 5 (KLF5) were highly expressed in notochord cells. Functional assay results demonstrated that both genes are necessary for notochord lumen formation and expansion. Finally, we performed motif enrichment analysis of the differentially accessible regions in different tailbud stages and summarized the potential roles of these motif-bearing transcription factors in larval development. Overall, our study found a correlation between gene expression and chromatin accessibility and provided a vital resource for understanding the mechanisms of the development of ascidian embryos.


Introduction
The ascidian Ciona savignyi belongs to the urochordate subphylum [1][2][3], which comprises the closest relatives of vertebrates [4][5][6][7].Ciona is an ideal model organism for studying tissue formation during chordate embryonic development [8].Its early embryonic processes are relatively simple, its transgenic technology has been developed [9], and its genome has been sequenced [10,11] and deposited in the EDomics database [12].Distinct morphological changes occur in various tissues of the ascidian from the early to the late tailbud stages, including tail bending [13] and epidermal and muscle cell extension, and notochord cells also undergo a series of complex morphological events, including elongation [14], migration, lumen formation [15][16][17][18], and lumen fusion [14,19,20].The cytological process of ascidian tailbud development has been extensively studied, but the molecular mechanism of its regulation is poorly understood.
Assay for transposase-accessible chromatin with high-throughput sequencing (ATACseq) is a highly efficient technology for detecting accessible open chromatin [21].In the last decade, 65 species have been subjected to ATAC-seq, and more than 2000 datasets have been generated, including data for protozoa, fungi, plants, and animals [22].Moreover, ATACseq is useful for analyzing marine invertebrate embryos [23].Ascidian chromosomal loci exhibit time-dependent dynamic changes in chromatin accessibility and conserved open chromatin regulatory elements around developmental regulatory genes [24], which are similar to those observed in Drosophila [25], Caenorhabditis elegans [26], and vertebrates [27].However, the key regulatory factors in tailbud development remain unclear.
In our current work, we investigated accessible chromatin and the expression profiles of genes at three different developmental stages of C. savignyi, namely stages 21 (tail is 1-2-fold longer than the trunk and curves ventrally, complete with the intercalation of notochord cells), 24 (notochord vacuolation begins, palps start to be visible at the front end of the embryo, tail straightens), and 25 (ocellus melanization, all notochord cells have vacuoles), using ATAC-seq and RNA sequencing (RNA-seq).An integrated analysis of ATAC-seq and RNA-seq data was performed to identify changes in chromatin accessibility that influence gene expression as well as key regulators in the development of ascidian tailbuds, such as Krüppel-like factor 5 (KLF5) and matrix metalloproteinase 24 (MMP24), which regulate notochord lumen formation and expansion.We further

ATAC-Seq and RNA-Seq of Ciona Tailbud Embryos
We collected embryos at three stages (St.21, St. 24, and St. 25; Figure 1A) to perform ATAC-seq and RNA-seq with two biological replicates.To construct the ATAC-seq library, cell isolation and nuclei extraction were performed after embryo collection (Figure S1A).The embryos were then isolated as single cells.A portion of the single-cell suspension was used for activity detection.Furthermore, 0.4% trypan blue staining revealed high cell activity, and the cell viability rate exceeded 80%.After nucleus extraction, blue dot structures stained by DAPI were identified, indicating that intact nuclei were successfully extracted.Six ATAC-seq libraries were constructed, and sequencing data were obtained using an Illumina Nova Seq sequencer (Tables S1 and S2).Clean reads were aligned with the reference genome Ciona_savignyi_51511_TX_20220104_seqkit.fasta.
Pearson's correlation analysis was performed for ATAC-seq data, and the results revealed high correlation coefficients between the biological repeats of St. 21 (0.96),St. 24 (0.99), and St. 25 (0.98; Figure S1B).Pearson's correlation analysis for RNA-seq data also revealed high correlation coefficients of 0.98, 0.99, and 0.97, respectively, for the aforementioned stages (Figure S1C).A large number of peaks were located near the transcriptional start site, which is a regulatory area (Figure S1D).This suggests that chromatin regions are involved in transcriptional regulation.Surprisingly, we also observed strong ATAC-seq signals at transcription end sites (TESs).TES open chromatin might reflect the binding of factors engaged in transcription termination [28].All ATAC-seq libraries yielded fragment lengths with expected distributions [21,29], including most small fragments corresponding to nucleosome-free regions (≤100 bp) and mononucleosome-and dinucleosome-associated fragments (200 and 400 bp, respectively; Figure S1E), confirming the good quality of the ATAC-seq data.

Chromatin Accessibility at Different Developmental Stages
We mapped the genomic coordinates from the peak data to the 13 chromosomes of the C. savignyi genome (Figure S2A), and most regions of each chromosome were covered.We counted the number of peaks in different developmental stages, and the average number was 13,650 in St. 21, 31,369 in St. 24, and 19,645 in St. 25.Statistical analysis (t-test) revealed that the number of accessible peaks was significantly higher in St. 24 (p < 0.05; Figure S2B), indicating that gene expression is more active during this stage.ATAC-seq peaks were enriched in 3 ′ -UTRs, 5 ′ -UTRs, exons, introns, promoters, and intergenic regions (Figure 1B).The promoter region accounted for the largest proportion, confirming that chromatin opening is related to gene transcriptional regulation.
The open regions of these three stages were annotated, and multiple open chromatin regions might correspond to the same gene.A total of 351 and 138 genes adjacent to differentially accessible regions (DARs) were found to contain upregulated and downregulated regions between St. 21 and St. 24 (Figure 1C, Table S3) and between St. 24 and St. 25, respectively (Figure 1D, Table S4).These results indicate that the chromosomal loci of ascidians exhibit time-dependent dynamics concerning chromatin accessibility.These DARs were further compared with genomic sequences, and genes adjacent to these differentially open regions were analyzed using clusterProfiler for GO enrichment (Table S5).The results revealed that calcium ion binding, ion binding, homophilic cell adhesion via plasma membrane adhesion molecules, and transmembrane signaling receptor activity were the enriched pathways in St. 24 compared to St. 21 (Figure 1E).Compared to St. 25, the DAR-associated genes of St. 24 were enriched in the cell adhesion and exocytosis pathways (Figure 1F).Seven genes were randomly selected from the GO terms in the comparison of St. 21  Although some of the GO terms were not significantly enriched, these genes belonged to genes adjacent to open regions of chromatin with significant differences.The promoter constructs of these genes were electroplated into the embryos and most of the signaling was observed in epidermal and/or muscle cells (Figure S2C).These results indicate the existence of substantial dynamic regulation and function in muscle and epidermal cells during larval development.

Transcriptome Profiling of C. savignyi at Different Developmental Stages
Differentially expressed genes (DEGs) were obtained from RNA-seq for the three stages, and the results identified 1142 upregulated genes and 547 downregulated genes in St. 24 compared to St. 21 (Figure 2A, Table S6).In total, 763 upregulated and 465 downregulated genes were identified in St. 25 compared to St. 24 (Figure 2C, Table S7).Furthermore, all DEGs were mapped to GO database terms, enabling their annotation (Table S8).DEGs were mostly enriched in the membrane transport and ion transport-related pathways (Figure 2B,D).These results suggest that in addition to basic gene expression, transcription, and translation, a wide range of processes such as ion transport, membrane transport, and cell attachment are involved in ascidian tailbud development.

Correlation of ATAC-Seq and RNA-Seq Data
Combined analysis of ATAC-seq and RNA-seq data can provide more biological information that can be used to investigate the relationship between changes in open chromatin regions identified through ATAC-seq analysis and corresponding changes in gene expression.In total, 70 DEGs were identified in St. 24 compared to St. 21, and 11 DEGs were identified in St. 25 compared to St. 24 (Figure 3A,B, Table S9).In the region characterized by low chromatin openness in St. 24, 30 genes were downregulated, whereas 40 genes were upregulated (Figure S3A).Meanwhile, the region with high chromatin openness in St. 25 featured two downregulated and nine upregulated genes (Figure S3B).Eight genes were randomly selected from the GO terms for the comparison between St. 21 and St. 24 (Figure S3C, Table S10), and all eleven genes for St. 24 versus St. 25 were analyzed by qRT-PCR.Our RNA-seq and qRT-PCR analyses revealed comparable gene expression   S9).In the region characterized by low chromatin openness in St. 24, 30 genes were downregulated, whereas 40 genes were upregulated (Figure S3A).Meanwhile, the region with high chromatin openness in St. 25 featured two downregulated and nine upregulated genes (Figure S3B).Eight genes were randomly selected from the GO terms for the comparison between St. 21 and St. 24 (Figure S3C, Table S10), and all eleven genes for St. 24 versus St. 25 were analyzed by qRT-PCR.Our RNA-seq and qRT-PCR analyses revealed comparable gene expression patterns (Figures 3C and S3D and Table S11), which validated the accuracy of our RNA-seq data in all three stages.S11), which validated the accuracy of our RNA-seq data in all three stages.4A).FHDC1, FOXF1, and CTGF were expressed in epidermal cells, and GAPDH was expressed in muscle cells.The expression of these genes might be associated with muscle and epidermal cell extension during early embryonic development.RBMS1 was expressed in epidermal cells and cells near the eye spot, CTDSP1 and MYC were expressed in muscle cells, FBN1 was expressed in epidermal cells, and MSI2 and ACTA2 in mesenchymal cells.MMP24 and KLF5 were expressed in the notochord during the lumen formation stage.

patterns (Figures 3C and S3D, Table
Lumen formation is an obvious cytological process during the development of the ascidian tailbud.To investigate the role of Cs-MMP24 in lumen formation, we used ilomastat (GM6001), a potent and broad-spectrum MMP inhibitor [30], to inhibit Cs-MMP24 in Ciona tailbud embryos.At 12 h post fertilization (hpf), embryos were treated with 40 µM GM6001 or DMSO and then collected for phalloidin staining to mark the cell boundary (Figure 4B).We measured the A-P and D-V diameters of the extracellular lumen and calculated lumen volume (Figure S4, Table S12).The results illustrated that the volume of the notochord lumen in the drug-treated group was significantly smaller than that in the control groups in St. 25 (Figure 4C).We constructed an N-terminal (activating domain) deletion construct of the Cs-KLF5 fusion protein (Cs-KLF5 DN), which retained the nuclear localization signal and the zinc finger domain, to determine the roles of Cs-KLF5 in notochord lumen formation [31,32].Cs-KLF5 DN expression in the notochord resulted in abnormalities in lumen formation (Figure 4D).Taken together, these results indicate that Cs-KLF5 and Cs-MMP24 are necessary for notochord lumen formation and expansion.

Motif Enrichment Analysis of DARs in the Tailbud Stage
Transcription factors (TFs) can bind accessible regions to regulate gene expression.We analyzed the enriched motifs in the different peaks in different developmental stages.To identify potential TFs responsible for the observed differences in chromatin accessibility, the enrichment of TF motifs was screened in accessible regions in St. 21 versus St. 24 and in St. 24 versus St. 25.The enrichment of the motifs is listed in Figure 5A,B.We identified 690 TFs belonging to 143 TF families (Figure S5A, Table S13).To investigate the mRNA expression of TFs in tailbud stages, we determined the FPKM value of each TF in St. 21, St. 24, and St. 25 by RNA-seq.The TF genes were grouped into six clusters (Figure S5B, Table S14).The results illustrated that Tbx15/18/22 was upregulated from St. 21 to St. 24, and FoxD-b, Six3/6, FoxE, FoxI-a, FoxK, GATA-a, Fos-B, and FoxL2 were upregulated from St. 24 to St. 25 (Figure 5A,B).Tbx15/18/22 has been reported to maintain the expression of muscle structural genes.In Tbx15/18/22 morphant embryos, many muscle-structural genes, such as muscle actin, myosin heavy chain, troponin t, and troponin c, are downregulated [33].Previous studies reported that NvSix3/6 plays a key regulatory role in the development of a broad aboral territory in Nematostella [34], and deregulation of the Six3/6 pathway led to evolution of the pinhole eye in nautili [35].The forkhead TF family is involved in various developmental processes of the embryo.FoxD activates vegetal hemisphere-specific genes (Lhx3/4 and Zic-r.b) and represses animal hemisphere-specific genes in ascidians (Dmrt.aand Dlx.b) [36].FoxD is also essential for notochord differentiation or induction [37].Ci-FoxE is specifically expressed at the endostyle of adults, being prominent in the thyroidequivalent region of zone 7 [38].FoxI is essential for formation of the otic placode, survival of the branchial arch neural crest, and developmental remodeling of the branchial arch ectoderm [39].FoxK activity is necessary for the formation of midgut constrictions and the development of midgut vesicles [40].FoxL2 is an evolutionarily conserved gene involved in the germinal differentiation of gonadal sex [41,42].GATA2 enhances the expression of MMP2 in microvascular endothelial cells cultured within a 3D type I collagen matrix [43].We speculated that MMP2 is involved in notochord formation.Fos-B overexpression resulted in aberrant axon branching in zebrafish [44].In conclusion, we identify 15 TFs responsible for the observed differences in chromatin accessibility.The expression profiles of these TFs provided a possible clue for their functions (Figure 5C).

Discussion
Gene expression is regulated by changes in gene regulatory programs through the integration of Cis-regulatory elements (CREs), which include core promoters, promoter proximal elements [45], and TF binding sites.The binding of TFs to CREs promotes the recruitment of RNA polymerase II to the promoter of the target gene, thereby regulating the gene expression [46].Identification of the sequences of CREs and clarification of their localization on chromosomes and binding TFs will be helpful in understanding the epigenetic information of chromatin and its molecular mechanisms in a specific biological context [46,47].In addition, transposable elements (TEs) may also have a significant role in organizing chromatin structure and gene regulation, which contribute to controlling the expression of genes locally and even far away at the transcriptional and post-transcriptional levels [48].In a study of Zootoca vivipara, TE was found to be highly conserved and played a positive role in scaly animal genome evolution [49].In this study, we screened key regulatory genes in the complex regulatory network through high-throughput sequencing and revealed the molecular regulatory mechanisms of the development of ascidian larvae.
Our ATAC-seq results identified vital genes associated with notochord tubulogenesis and lumen expansion, including DLK2, BMP5, BMP2, JAG, PCDH18, FBN1, and SLC37A1.DLK2 belongs to the Notch/Delta/Serrata family, and it has been reported to regulate metastatic melanoma cell proliferation in a dose-dependent manner [50].JAG1 is the ligand for Notch1, and upregulated JAG1 promotes cell proliferation in adrenocortical carcinoma [51].BMP2 [52] and BMP5 [53] have been reported to regulate cell proliferation.They target epidermal and muscle cells, and they might be involved in regulating the proliferation of epidermal and muscle cells used for tail extension.It has been reported that BMP signaling regulates anterior neurectoderm specification and differentiation in C. intestinalis [54].PCDH18 expression was observed near the eyespot and CNS, which is similar to its localization in zebrafish [55].Diversified PCDH molecules contribute to neural circuit development.In in vivo and in vitro mouse hippocampal neurons, Pcdha knockout resulted in the decreased dendrite arborization, and this arborization defect was dependent on a PYK2/Rho/Rac pathway [56][57][58].From St. 24 to St. 25, FBN1 was expressed in epidermal cells, and SLC37A1 was expressed in notochord cells.They might be involved in epidermal cell extension and notochordal lumen formation and expansion.The RNA-seq results identified DEGs enriched in membrane transport and ion transportrelated pathways, consistent with the formation of the notochordal lumen through the establishment of polarity, cytoskeletal rearrangement, and expansion of the notochordal lumen with ion flow (SLC26A), proteoglycan secretion, vesicle trafficking (ELMOD3), and replenishment of the apical membrane (14-3-3εa and caveolin) [17, [59][60][61][62].
A comprehensive analysis of ATAC-seq and RNA-seq data demonstrated that all 12 DEGs had high chromatin openness in St. 25.The analysis revealed that increased chromatin accessibility was correlated with increased transcription, although there was a clear increase in regional accessibility corresponding to decreased transcription, which might be attributable to transcriptional repressors binding to open chromatin regions [63].Among them, MMP24 and KLF5 targeted notochord cells in the lumen formation stage.The dominant-negative effect of Cs-KLF5 and drug treatment targeting Cs-MMP24 resulted in abnormal lumen formation and expansion.KLF5 induces ECM degradation by promoting MMP expression and secretion [64].KLF5 has been identified as a crucial factor in cardiovascular diseases [65].We hypothesize that Cs-KLF5 and Cs-MMP24 coregulate the formation of the notochord lumen.
Overall, we identified 12 candidate genes through a combination analysis of ATAC-seq and RNA-seq data.Cs-KLF5 and Cs-MMP24 were proven necessary for the formation and expansion of the notochord lumen.Our approach for identifying open chromatin regions and predicting TFs involved in the regulation of ascidian development offers new resources for further mechanistic exploration in ascidian embryonic development.

Animals
Adult specimens of C. savignyi were collected from Huangdao, Qingdao, China.The animals were kept in the laboratory under controlled conditions, in seawater with a salinity of 32‰.We obtained eggs and sperms through dissecting adult animals.After fertilization and dechorionation, embryos were cultured at 16 • C. We collected embryos at three different stages for ATAC-seq and RNA-sequencing.

ATAC-Seq Analysis
Raw sequencing reads were initially processed and adapter sequences were removed using Fastp [66].Clean reads were obtained from the raw data by quality control and then aligned to the reference genome Ciona_savignyi_51511_TX_20220104_seqkit.fasta (unpublished data) with Bowtie2 [67] using default settings.MACS2 software (v2.2.7.1) [68] was used to detect peaks.The parameters of peak calling are --nomodel, --shift -100, --extsize 200.According to the genomic location and gene annotation information of peaks, peakrelated genes can be annotated using ChIPseeker [69].Moreover, the distribution of peaks in different genomic regions (such as exon, intron, promoter, downstream regions, intergenic regions, 3'-UTRs, 5'-UTRs) was estimated.Promoters are defined as 2 kb upstream and downstream of TSS.Differentially accessible chromatin regions (DARs) were identified by DiffBind software (v3.8.4) [70].Peaks with FDR under 0.05 were considered differentially accessible peaks.And the R-package ChIPseeker was used to associate differentially accessible peaks with their putative target genes [69].Transcription factor enrichment of sequences from the known urochordate motif database from JASPAR was performed using SEA (Simple Enrichment Analysis, evalue ≤ 10) [71].Downstream gene prediction was performed by motif scanning of open chromatin regions using FIMO.

RNA-Seq Analysis
RNA-seq reads that passed the filter were trimmed to eliminate low-quality reads and adapters by Fastp [66].Reads were aligned by STAR [72].The edgeR method has high sensitivity and accuracy for medium-scale RNA sequencing data, so differential gene expression was analyzed using edgeR [73].Significantly differentially expressed genes (DEGs) were identified based on the following criteria: |log2 fold change| ≥ 1 and padj < 0.05.Two biological replicates were used.The R package ggplot2 was used to plot heatmap and volcano plots [74].

GO Enrichment Analysis
GO enrichment analysis of DEGs and genes adjacent to DARs was implemented using the clusterprofiler (v4.6.2) [75] R package.GO annotations came from annotations on C. savignyi genomes (unpublished data) using interproscan.GO terms with a corrected p-value < 0.05 were considered significantly enriched.

Integration Analysis of ATAC-Seq and RNA-Seq
We compared the expression profiles of differentially accessible peak-related genes in ATAC-seq with DEGs in RNA-sequencing.Genes associated with closed-to-open peaks in ATAC-seq were compared with DEGs upregulated in RNA-seq.Genes associated with open-to-closed peaks in ATAC-seq were compared with DEGs downregulated in RNA-seq [76].Genes with changes in both gene expression and chromatin accessibility were shortlisted, followed by Spearman's correlation analysis.

Identification and Classification of TFs
The methods used here follow a previously described protocol [77].TFs were identified and categorized according to the conserved DNA binding domains (DBDs).The protein domains were analyzed using Hidden Markov Model (HMM) profiles from the Pfam database [78], applying the hmmscan program in the HMMER package.Pfam ID of DBDs were identified from the REGULATOR database [79].Genes containing DBDs were screened and classified as TFs with an e-value threshold of 10 −4 .Pfam IDs are shown in Table S13.TFs were classified into different families according to the DBDs.

Quantitative Real-Time PCR (qPCR)
These methods are described in detail in [80].RNAiso plus (TAKARA, Kusatsu, Shiga Japan) was used to isolate total RNA from frozen embryos at different stages.For reverse transcription, HiScript II Q RT SuperMix for qPCR (+gDNA wiper) was utilized.The expression levels of DEG/DAR-associated genes were detected by 2-step qPCR (Vazyme, Nanjing, China).We calculated data using 2 −∆∆Ct .Statistical analysis was conducted using paired Student's t-tests, with a significance level of p < 0.05.Prism 9 software (v 8.0.2) was used to plot the graph of qPCR results.The primers used in qPCR are listed in Table S15.

Plasmid Constructions
Tissue expression patterns were performed by gene promoter analysis.DNA sequences upstream of DEG/DAR-associated genes were amplified and subcloned into the pEGFP-1 vector for promoter analysis.And the promoter sequence was compared to the genes encoded in the genome through BLAST [81] to ensure that the promoter did not contain other genes.Then, these constructs were confirmed by sequencing.The primers used in plasmid construction are listed in Table S15.

Electroporation
We performed electroporation experiments according to a previously described method with some modifications.Specifically, the parameter of the exponential protocol was used with 50 V and 1500 µF [82].The electroporation system consisted of a plasmid of 60-80 µg, 420 µL 0.77 M D-mannitol, and ddH 2 O. Dechorionated fertilized eggs (300 µL) were mixed with the electroliquid system in 0.4 cm cuvettes of a Gene PulserX cell System (BIO-RAD, Hercules, CA, USA).

Drug Treatment and Immunostaining
Embryos were treated with 40 µM of the inhibitors ilomastat (GM6001, HY-15768, MedChemExpres, Monmouth Junction, NJ, USA) [83] and DMSO at 12 h post fertilization (hpf).Up to 22 hpf, embryos were collected and fixed with 4% paraformaldehyde (PFA) at room temperature for 2 h.Fixed embryos were washed three times with PBS containing 0.1% Triton X-100 (PBST), and then stained with phalloidin in PBST (1:300) overnight at 4 • C to mark the cell boundary of embryos.Embryos were then washed three times with PBST.DAPI staining was performed in the dark to label the cell nucleus.

Statistics
Lumen volume was measured according to the method described previously [84].We also measured the anteroposterior (A-P) diameter (d1) and dorsoventral (D-V) diameter (d2) using ImageJ and calculated lumen volume using the spheroid formula (V = 1 /6πd1d2 2 ) (Figure S4).Then, we counted the lumen volume in the DMSO and inhibitor treatment group separately.All data are presented as mean ± SEM.We performed statistical analysis using Student's t-test.p < 0.05 (*) and p < 0.01 (**) indicate statistical significance.

Figure 2 .
Figure 2. Visualization of DEGs.(A,C) Visualization of gene expression changes.padj is represented by −log10(padj) in the ordinate, and gene expression is represented by log2 (fold change) in the abscissa.Blue, red, and gray dots indicate downregulated DEGs, upregulated DEGs, and genes with no significant expression change, respectively.(B,D) GO enrichment analysis of DEGs at different stages.B: St. 21 vs.St. 24, D: St. 24 vs.St.25.

Figure 2 .
Figure 2. Visualization of DEGs.(A,C) Visualization of gene expression changes.padj is represented by −log 10 (padj) in the ordinate, and gene expression is represented by log 2 (fold change) in the abscissa.Blue, red, and gray dots indicate downregulated DEGs, upregulated DEGs, and genes with no significant expression change, respectively.(B,D) GO enrichment analysis of DEGs at different stages.B: St. 21 vs.St. 24, D: St. 24 vs.St.25.

2. 4 .
Correlation of ATAC-Seq and RNA-Seq Data Combined analysis of ATAC-seq and RNA-seq data can provide more biological information that can be used to investigate the relationship between changes in open chromatin regions identified through ATAC-seq analysis and corresponding changes in gene expression.In total, 70 DEGs were identified in St. 24 compared to St. 21, and 11 DEGs were identified in St. 25 compared to St. 24 (Figure 3A,B, Table

Figure 3 .
Figure 3. Integrated analysis of ATAC-seq and RNA-seq data.(A,B) Intersection of different mRNAs and target genes of different peaks.DAR-associated genes (green and purple indicate St. 21 vs.St.24 and St. 24 vs.St. 25, respectively) and DEGs (blue and orange indicate St. 21 vs.St.24 and St. 24 vs.St. 25, respectively).(C) Chromatin opening and expression changes of associated genes.IGV plots present ATAC-seq signals in selected gene loci.Gene expression was determined by quantitative fluorescence PCR and transcriptomic analysis.Blue, red, and green represent chromatin fragments of St. 21, St. 24, and St. 25, respectively.Black and gray bars represent transcriptome data and quantitative fluorescence PCR results of genes, respectively.

Figure 4 .
Figure 4. Expression and functional assay of key genes in larval development.(A) Tissue expression patterns of the selected genes as determined by promoter assay.Scale bar, 50 µm.(B) Treatment with an MMP24 inhibitor (GM6001, 40 µM) arrested lumen expansion.Scale bar, 10 µm.The asterisk marks the lumen.(C) Quantification of lumen volume in DMSO-(n = 191) and GM6001-treated

Figure 4 .
Figure 4. Expression and functional assay of key genes in larval development.(A) Tissue expression patterns of the selected genes as determined by promoter assay.Scale bar, 50 µm.(B) Treatment with an MMP24 inhibitor (GM6001, 40 µM) arrested lumen expansion.Scale bar, 10 µm.The asterisk marks the lumen.(C) Quantification of lumen volume in DMSO-(n = 191) and GM6001-treated embryos (n = 165) embryos at 22 hpf.Statistical analysis was performed using Student's t-test, p < 0.001 (***).(D) Overexpression of dominant-negative Cs-KLF5 led to failure of lumen formation.Schematic images located on the right side of confocal images present the phenotypes of Cs-KLF5 knockdown and control.The green, blue, and red areas represent the notochord cells, nuclei, and nuclei in cells expressing Bra>Cs-KLF5DN::tdtomato/Bra> tdtomato (red), respectively.Scale bar, 10 µm.

Figure 5 .
Figure 5. TF binding motif enrichment analysis.(A) Nine TF binding motifs enriched in accessible regions from closed to open in St. 21 versus St. 24.(B) Eight TF binding motifs enriched in accessible regions from closed to open in St. 24 versus St. 25.Red highlights the upregulation of TFs.(A,B).The X-axis represents −log10 (p-value) in the motif enrichment analysis.The Y-axis represents the

Figure 5 .
Figure 5. TF binding motif enrichment analysis.(A) Nine TF binding motifs enriched in accessible regions from closed to open in St. 21 versus St. 24.(B) Eight TF binding motifs enriched in accessible regions from closed to open in St. 24 versus St. 25.Red highlights the upregulation of TFs.(A,B).The X-axis represents −log10 (p-value) in the motif enrichment analysis.The Y-axis represents the predicted TF binding motifs.(C) Summary of the potential roles of TFs in Ciona larval development.Different colored arrows represent downstream genes that may be regulated by different transcription factors.