Combined analysis of chromatin accessibility and gene expression profiles provide insight into Fucoxanthin biosynthesis in Isochrysis galbana under green light

Isochrysis galbana, as a potential accumulator of fucoxanthin, has become a valuable material to develop functional foods for humans. Our previous research revealed that green light effectively promotes the accumulation of fucoxanthin in I. galbana, but there is little research on chromatin accessibility in the process of transcriptional regulation. This study was conducted to reveal the mechanism of fucoxanthin biosynthesis in I. galbana under green light by analyzing promoter accessibility and gene expression profiles. Differentially accessible chromatin regions (DARs)-associated genes were enriched in carotenoid biosynthesis and photosynthesis-antenna protein formation, including IgLHCA1, IgLHCA4, IgPDS, IgZ-ISO, IglcyB, IgZEP, and IgVDE. The motifs for the MYB family were also identified as candidates controlling metabolic regulation responses to green light culture of I. galbana, including IgMYB1, IgMYB2, IgMYB33, IgMYB42, IgMYB98, IgMYB118, and IgMYB119. The results of differential expression analysis and WGCNA showed that several genes or transcription factors (TFs) related to carotenoid metabolism and photosynthesis exhibited a higher expression level and were significantly upregulated in A-G5d compared with A-0d and A-W5d, including IgMYB98, IgLHCA1, IgLHCX2, IgLHCB4, and IgLHCB5. This suggests that upregulation of these genes by green light may be the key factor leading to fucoxanthin accumulation by regulating the photosynthesis-antenna protein pathway. An integrated analysis of ATAC-seq and RNA-seq showed that 3 (IgphoA, IgPKN1, IgOTC) of 34 DARs-associated genes displayed obvious changes in their chromatin regions in ATAC-seq data, suggesting that these genes specific for green light may play a key role in fucoxanthin biosynthesis in I. galbana through a complex regulatory network of multiple metabolic pathways interacting with each other. These findings will facilitate in-depth understanding the molecular regulation mechanisms of fucoxanthin in I. galbana and its role in response to green light regulation, providing technical support for the construction of high fucoxanthin content strains.

Isochrysis galbana, as a potential accumulator of fucoxanthin, has become a valuable material to develop functional foods for humans. Our previous research revealed that green light effectively promotes the accumulation of fucoxanthin in I. galbana, but there is little research on chromatin accessibility in the process of transcriptional regulation. This study was conducted to reveal the mechanism of fucoxanthin biosynthesis in I. galbana under green light by analyzing promoter accessibility and gene expression profiles. Differentially accessible chromatin regions (DARs)associated genes were enriched in carotenoid biosynthesis and photosynthesisantenna protein formation, including IgLHCA1, IgLHCA4, IgPDS, IgZ-ISO, IglcyB, IgZEP, and IgVDE. The motifs for the MYB family were also identified as candidates controlling metabolic regulation responses to green light culture of I. galbana,including IgMYB1,IgMYB2,IgMYB33,IgMYB42,IgMYB98,IgMYB118, and IgMYB119. The results of differential expression analysis and WGCNA showed that several genes or transcription factors (TFs) related to carotenoid metabolism and photosynthesis exhibited a higher expression level and were significantly upregulated in A-G5d compared with A-0d and A-W5d, including IgMYB98, IgLHCA1,IgLHCX2,IgLHCB4,and IgLHCB5. This suggests that upregulation of these genes by green light may be the key factor leading to fucoxanthin accumulation by regulating the photosynthesisantenna protein pathway. An integrated analysis of ATAC-seq and RNA-seq showed that 3 (IgphoA, IgPKN1, IgOTC) of 34 DARs-associated genes displayed obvious changes in their chromatin regions in ATAC-seq data, suggesting that these genes specific for green light may play a key role in fucoxanthin biosynthesis in I. galbana through a complex regulatory network of multiple metabolic pathways interacting with each other. These findings will facilitate in-depth understanding the molecular regulation mechanisms of fucoxanthin in I. galbana and its role in response to green light regulation, providing technical support for the construction of high fucoxanthin content strains.

Introduction
Aquatic algae have acquired a range of adaptive mechanisms to cope with the harsh features of aquatic environments, such as poor light quality, low temperature, high salt, and low nutrition. In the long history of evolution, algae have lived in environments ranging from shallow to deep seas, where light is the most important environmental factor for the growth of microalgae. The marine microalgae Isochrysis galbana is a group of oxygenic photosynthetic organisms that possess fucoxanthin Chl a/c-binding proteins as light-harvesting antennae, which has exceptional blue-green light harvesting and photoprotection capabilities (Pi et al., 2019;Wang et al., 2020). Isochrysis galbana, as a potential accumulator of fucoxanthin, is recognized for being a rich source of fucoxanthin (more than 10% of dry weight biomass) and lipids (7.0-20.0% dry weight biomass), and has become an valuable material to develop functional foods for humans due to its small size, fast growth rate, high fucoxanthin content, and large-scale artificial cultivation (Kim, 2012;Zhu et al., 2018). Stress conditions could induced microalgae to synthesize carotenoids. Weighted gene co-expression network analysis (WGCNA) was used to investigate the underlying molecular mechanisms of Dunaliella salina response to salinity stress, identifying several salinity specific hub genes (Panahi and Hejazi, 2021). In order to overcome the low biomass of Auxenochlorella protothecoides and increase the productivity of secondary metabolites by using heterotrophic growth model, WGCNA was used to identify that some hubs, such as serine hydroxymethyltransferase (SHMT1), was the best candidate genes for the development of metabolites accumulating strains in microalgae (Panahi et al., 2020). Additionally, this marine microalga is much easier to process compared with other kind of algae because of its lack of a cell wall and is considered an ideal bait for the development of aquatic animals . Studies have shown that fucoxanthin has remarkable biological characteristics, including antioxidant, antitumor, antibacterial, antiviral, antiobesity, neuroprotective, and other pharmacological effects, which are in increasing demand in the biopharmaceutical and cosmetic fields (Maeda et al., 2015;Kim et al., 2018).
An assay for transposase accessible chromatin sequencing (ATACseq) is an efficient technology used to study genome-wide open chromatin region accessibility by investigating chromatin status and identifying transcription factor binding sites. This technique is popular because of its simple operation, time-saving, strong repeatability, and low material input compared to ChIP-seq, DNAse-seq, and FAIRE-seq (Wei et al., 2018;Sen et al., 2021Sen et al., , 2022. Although ATAC-seq has been used successfully in Arabidopsis thaliana, Vitis amurensis, embryonic cells, and tumors, there is little research on chromatin accessibility for gene expression in algae (Corces et al., 2018;Del Priore et al., 2021;Farmer et al., 2021;Ren et al., 2021). In this study, ATAC-seq was used to identify the key DARs, motifs, and TFs that caused the fucoxanthin accumulation in I. galbana under green light treatment. Additionally, RNA-seq can truly reflect the gene expression profile at specific time points, providing an important reference for gene regulation mechanism of fucoxanthin in I. galbana. Integrating ATAC-seq and RNA-seq analyses can complement each other to analyze the regulation and expression of differentially expressed genes (DEGs) and their related chromatin open regions, further identifying the key regulatory elements responsible for fucoxanthin in I. galbana. These results can help to explore the molecular regulation mechanism of fucoxanthin in I. galbana and its role in response to green light regulation, providing technical support for the construction of high fucoxanthin content strains.

Materials and methods
Samples and treatment conditions I. galbana LG007 used in this study was separated from the near sea area of Chuanshi Island in Fujian and deposited at Fujian Normal University, China. The algae was inoculated in f/2 culture medium in conical flasks at a density of 10 6 cells per liter. Incubation conditions were as follows: the light intensity was 100 μmol·m −2 ·s −1 , the temperature was 23 ± 1°C, the light cycle was 24 h, and the flasks were shaken three times a day. In our previous study, the fucoxanthin yield of I. galbana reached its maximum at 5 days but started to decrease significantly at 7 days, indicating that Day 5 might be a key time point in fucoxanthin accumulation. When the cells reached the end of logarithm (~5 × 10 6 ), the cells were divided into two groups with three biological replicates: one group remained cultured with white light of 100 μmol photons m −2 s −1 for 5 days as a control group (W5d), and the other group was cultured with green light of 100 μmol photons m −2 s −1 for 5 days as the treatment group (G5d) ( Figure 1A).

ATAC sequencing and analysis
Nine samples were used for ATAC sequencing, namely A-0d-1, A-0d-2, and A-0d-3 (samples for initial culture at Day 0 under white or green light), A-W5d-1, A-W5d-2, and A-W5d-3 (samples cultured for 5 days under white light), and A-G5d-1, A-G5d-2, and A-G5d-3 (samples cultured for 5 days under green light). Cells (~5 × 10 4 ) were collected by centrifugation at 7,200 rpm for 5 min and then washed two times with cold phosphate buffer solution (pH 7.4). The obtained cells were suspended with cold lysis buffer and centrifuged again with the same parameters to remove the supernatant. The isolated nuclei were resuspended with the Tn5 transposase for transposing reaction system and incubated at 37°C for 30 min, then immediately purified using a Qiagen MiniElute Kit (Qiagen, CA, United States). The PCR amplification reaction of the purified DNA was carried out as previously reported and sequenced on the Novaseq 6,000 platform with PE150. Clean reads were obtained from raw reads of Illumina by Cutadapt (version 1.8.3) with default parameters (Martin, 2011). The qualityfiltered data were mapped onto the I. galbana LG007 genome using Bowtie2 (version 2.2.4) (Langmead and Salzberg, 2012). Density distribution of sequencing reads within the 3 kb interval up-and downstream of the TSS for each gene was performed by using DeepTools (version 2.07) (Ramírez et al., 2016). Peak calling was carried out using MACS2 (version 2.1.1) with a threshold of FDR < 0.05, then functionally annotated for genome-wide peaks by using ChIPseeker (version 1.2.6) (Zhang et al., 2008;Yu et al., 2015). MEME-ChIP (version 4.11.2) was used to identify and annotate the Motif (Machanick and Bailey, 2011). DARs were identified using DiffBind (version 2.2.11) with a threshold of fold change > = 1.2 and FDR < 0.05 (Stark and Brown, 2011). Associated genes in the DARs were compared with NR, Swiss-prot, GO, KEGG, COG, KOG, eggNOG, and Pfam databases to obtain the annotation information (Ashburner et al., 2000;Tatusov et al., 2000;Apweiler et al., 2004;Kanehisa et al., 2004;Deng et al., 2006;Finn et al., 2014;Huerta-Cepas et al., 2016). GO and KEGG enrichment analyses of DARs-associated genes in the promoter region were performed using the R package clusterProfiler (version 4.2) (Yu et al., 2012). We finally generated a total of ~1.16 Gb clean reads with Q30 ≥ 96.71%; over 86.19% of total reads were mapped to the I. galbana LG007 genome, indicating that the quality of data generated by sequencing is good (Supplementary Tables S1, S2).

mRNA sequencing and analysis
Nine samples were used for mRNA sequencing, namely R-0d-1, R-0d-2, and R-0d-3 (samples for initial culture at Day 0 under white or green light), R-W5d-1, R-W5d-2, and R-W5d-3 (samples cultured for 5 days under white light), and R-G5d-1, R-G5d-2, and R-G5d-3 (samples cultured for 5 days under green light). Total RNA was extracted with TRIzol reagent (Invitrogen Life Technologies, CA, United States). A NanoPhotometer (IMPLEN, CA, United States) and a 2,100 Bioanalyzer (Agilent Technologies, United States) were used to check the purity and quality of RNA. RNA libraries were prepared as previously described and sequenced on the Novaseq 6,000 platform with PE150. Clean reads All experiments were performed in triplicate. Each value presents the mean ± SD. "I" represents error bars for the various determinations (n = 3). "*" and "**" indicate that significance is at 0.05 and 0.01, respectively.
Frontiers in Microbiology 04 frontiersin.org were obtained from raw reads of Illumina by Trimmomatic (version 0.36), and then mapped to the genome using HISAT2 (version 2.2.1) (Anthony et al., 2014;Pertea et al., 2016). We finally generated a total of ~116 Gb data with Q30 ≥ 91.42%, and the mapping ratio of each sample to the genome assembly ranged from 92.60 to 93.83% (Supplementary Tables S3, S4). Differentially expressed genes (DEGs) were identified by using EdgeR (version 4.2) with the criteria of |log2 Fold change| ≥1.5 and value of p <0.05 (Nikolayeva and Robinson, 2014). Enrichment analysis of GO and KEGG was carried out by using topGO R package (version 3.8) and KOBAS (version 3.0), respectively (Ashburner et al., 2000;Xie et al., 2011). We finally generated a total of ~116 Gb raw data with Q30 ≥ 91.42%, and the mapping ratio of each sample to the genome assembly ranged from 92.60 to 93.83% (Supplementary Tables S3, S4).

Weighted gene coexpression network analysis
The WGCNA software package in R program was used to construct a coexpression network involved in fucoxanthin accumulation in I. galbana under different lights (Zhao et al., 2018). The weight values were calculated using a standardized gene expression matrix by pickSoftThreashold in the WGCNA package. We chose a soft power value β (β = 13) to approximate a scale-free network topology to generate a network, guided by a convenient 1-step network construction and module detection function in the R Tutorial (version 1.1). Then, the Module Eigengene (ME) was calculated, which represents the expression profile of each module. Next, based on the correlation between the ME and trait, we estimated the module-trait relationships to identify highly correlated modules. The module is considered to be associated with traits, where the moduletrait relationship value is ≥0.6 and p ≤ 0.05. The softConnectivity function was used to calculate the connectivity degree of genes, and the top 10 genes in the module were selected as the core genes of the module.

Correlation analysis of mRNA and ATAC sequencing
Expression profiles of differentially peak-related genes in ATAC were combined with DEGs in RNA-seq. When multiple differential peaks were associated with the same gene, the highest peak proximal to the gene was selected. Upregulated DEGs from RNA-seq data were compared to the related genes with up-egulated peaks in ATAC. Downregulated DEGs from RNA-seq data were compared to the related genes with downregulated peaks in ATAC. Selected candidate genes were analyzed for GO and KEGG enrichment analysis by using topGO R package (version 3.8) and KOBAS (version 3.0), respectively (Ashburner et al., 2000;Xie et al., 2011).

Fucoxanthin accumulation of Isochrysis galbana under different lights
To explore the differences in fucoxanthin accumulation in I. galbana under different lights, we detected the fucoxanthin content by HPLC under white and green light at 0 and 5 days. The dry cell weight of I. galbana under white light after 5 days (0.18 g/l) was not significantly increased compared to that under green light (0.19 g/l), suggesting that light quality had little effect on biomass accumulation ( Figure 1B). The fucoxanthin yield of I. galbana in culture, unit dry weight, and unit cell under green light after 5 days (0.71 mg/l, 4.15 mg/g, and 2.04 ug/10 7 cell) increased by 1.25, 1.33, and 1.67 times as those under white light (0.57 mg/l, 3.11 mg/g, and 1.22 ug/10 7 cell), respectively ( Figures 1C-E).
These results indicate that the fucoxanthin content of I. galbana can be significantly increased under green light conditions.

Landscape of accessible chromatin regions in Isochrysis galbana genome
A density distribution heatmap showed that genes were enriched around 3 kb upstream and downstream of the transcription start sites (TSSs), suggesting that enriched TSSs at both Day 0 and Day 5 under white or green light exhibited a relatively good signal area (Figure 2A). Pearson correlation analysis showed a high reliability (correlation coefficients, 0.96-0.99) of sampling and data between the biological repeats of A-0d, A-W5d, and A-G5d (Supplementary Figure S1). By peak calling, we obtained 1,330,155 peaks from all samples with the average length of 1,620 bp, and the average percentage of promoter (≤1 kb) and exon obtained after mapping these TSSs of each sample to I. galbana genome were 41.33 and 12.17%, respectively (Figure 2B and  Supplementary Tables S6, S7). GO and KEGG enrichment analysis of peak-associated genes in promoter regions showed that most of these genes were distributed in metabolic process (GO:0008152), carbohydrate metabolic process (GO:0005975), regulation of transcription (GO:0006355), stimulus response (GO:0050896), carbon metabolism (ko01200), fatty acid metabolism (ko01212), carotenoid biosynthesis (ko00906), and phosphatidylinositol signaling (ko04070) (Supplementary Figures S2, S3). By searching the plant TFs motifs database, we found that a number of TFs belong to ZNF (ZNP148, ZNF263,and ZNF281), ERF (ERF2, ERF10, and ERF104), and BPC (BPC1, BPC5, and BPC6) families within the 20 known motifs, suggesting that they may play a key role in green light culture quality of I. galbana ( Figure 2C). Additionally, the motifs for ABR, MYB, KLF, RAP, WRKY, and bZIP families were also identified as candidates controlling metabolic regulation responses to green light culture quality of I. galbana.

Integration analysis of ATAC-seq and RNA-seq
Association analysis of DARs of ATAC-seq predicted that the open chromatin region mediated by green light caused the change of the ability of the promoter regulatory region, and finally mediated the upregulation of downstream gene expression. Venn diagram analysis revealed that 34 genes were shared by DEGs and DARs-associated genes, whereas 1,855 and 212 genes were unique to DEGs obtained from RNA-seq and DARs-associated genes obtained from ATAC-seq, respectively ( Figure 6A). Some of these 34 shared genes exhibited significant expression differences and were involved in cellular processes, amino acid metabolism, MAPK signaling pathway, and signal transduction, including IgphoA (IZ001124), IgPKN1 (IZ000809), IgOTC (IZ001378), IgCOPB2 (IZ002053), IgABP1 (IZ007807), and IgRLM1 (IZ010897) ( Figure 6B). The 34 DARs-associated genes were analyzed and compared in the open region of chromatin; three genes (IgphoA, IgPKN1, IgOTC) showed obvious changes in the chromatin region in ATAC-seq data. For example, the open chromatin and the differential genes of IgphoA and IgPKN1 were suppressed by green light mediated culture relative to the white light group (Figures 6C,D). OTC is a key enzyme gene involved in urea cycle and arginine biosynthesis, participating in the formation of secondary metabolites and amino acids in plants (Zhao et al., 2018). IgOTC gene expression increased under green light treatment compared to white light, whereas chromatin opening was affected under both culture cycle and green light treatments ( Figure 6E). It is likely that these genes specific for green light play a key role in the process of fucoxanthin biosynthesis in I. galbana through a complex regulatory network of multiple metabolic pathways interacting with each other.

Discussion
A number of studies have preliminarily explored the regulatory mechanisms of fucoxanthin biosynthesis for the construction of high fucoxanthin content strains Chen et al., 2022;Truong et al., 2022). Recent studies investigated the effects of light treatment on fucoxanthin biosynthesis pathways in I. galbana (Chen et al., 2022). However, how the molecular mechanism of fucoxanthin biosynthesis responds to green light-mediated culture for different culture stages remains to be investigated. Although our previous research revealed that green light effectively promoted the accumulation of fucoxanthin, there is little research on chromatin accessibility in the process of transcriptional regulation. In the present study, ATAC-seq was used to analyze the chromatin opening profiles under different light treatments (white and green lights). Integrating ATAC-seq and RNA-seq analyses were performed to identify the key TFs or genes affecting fucoxanthin biosynthesis after green light treatment.
KEGG and GO analyses showed that peak-associated genes involved in carotenoid biosynthesis and photosynthesis-antenna protein formation were significantly enriched, including IgLHCA1, IgLHCA4,IgcrtB,IgPDS,IglcyB,IgCCD8,IgZEP,and IgVDE (Chen et al., 2022). In kiwifruit, AcMYB7 activated the promoter of the AdLCY-β gene, regulating the biosynthesis of carotenoids and chlorophyll (Ampomah-Dwamena et al., 2019). We further proved that the members of MYB gene family are important regulators for fucoxanthin biosynthesis. DARs-associated motif analysis showed that BPC5 and BPC6 had a strong binding effect on the promoter regions of IgDES4, IgCYP51A1, IgELO3, and IgPRKG1, which were closely related to fatty acid biosynthesis, steroid biosynthesis, and signaling pathways. Studies reported that AtBPC6 was involved in photosynthesis, photoreactions, photosynthetic membranes, and reproductive development processes in petunias (Yu et al., 2022). AtBPC1, AtBPC2, AtBPC4, and AtBPC6 were involved in the regulation of plant seedling growth, development, and stress responses in Arabidopsis thaliana (Mu, 2016). A linear relationship between the synthesis of pigment and fatty acids has been reported, and the upregulation of genes related to fatty acid metabolism was partly responsible for the large accumulation of pigment in Haematococcus pluvialis (Zhekisheva et al., 2002;Saha et al., 2013;Cheng et al., 2017). We speculated that BPC6 may regulate IgDES4 and IgELO3 to participate in lipid metabolism, resulting in a higher fucoxanthin content in I. galbana under green light treatment.
To better understand the regulatory mechanisms of fucoxanthin accumulation, WGCNA was performed to reveal the candidate genes involved in the fucoxanthin biosynthesis. IgcrtB, IgVDE, IgZDS, IgPDS, and IgZEP genes from green and yellow modules were significantly correlated with fucoxanthin content. For example, the high expression level of PtVDE gene played a critical regulatory role in fucoxanthin biosynthesis of Phaeodactylum tricornutum (Yang and Wei, 2020). The upregulation of the PtZDS gene resulted in the strongest synthesis ability of fucoxanthin in P. tricornutum under the treatment of 100 μmol/l MeJA . We suggested that the expression of genes upstream of the carotenoid biosynthetic pathway plays a crucial role in the downstream of fucoxanthin accumulation. LHCA1 is a chlorophyll a/b-binding protein that plays a key role in plant photosynthesis (Zou et al., 2014). LHCB5 and LHCB6 play unique roles in plant development, thylakoid organization, and PSII photo-protection (Gao et al., 2009). Interestingly, we found that the hub genes IgLHCA1 and IgLHCB5 showed higher expression levels and were significantly upregulated in A-G5d compared with A-0d and A-G5d, indicating that the high expression or upregulation of these genes can affect the fucoxanthin biosynthesis by regulating the photosynthesis-antenna protein pathway. An integrated analysis of ATAC-seq and RNA-seq showed the importance of IgphoA, IgPKN1, IgOTC, and fucoxanthin biosynthesis in I. galbana under green light treatment. The phoA gene participated in the encoding and synthesis of alkaline phosphatase, which has a signal peptide, playing a key role in cofactor biosynthesis (Kikuchi et al., 1981). Algae alkaline phosphatase (AP) belonging to Frontiers in Microbiology 10 frontiersin.org an atypical type AP (PhoA(aty)) could utilize marine phytoplankton to scavenge phosphorus from dissolved organic phosphorus (Lin et al., 2015). In addition to utilizing phosphorus nutrients, AP also inhibits pigment synthesis, photosynthesis, fatty acid synthesis, and cell division, maintains the metabolic homeostasis, and prevents premature cell division (Zhang et al., 2021). We speculated that alkaline phosphatase is the most important organophosphorus hydrolase by releasing inorganic phosphorus for the growth and metabolism of algae, which may promote the synthesis of secondary metabolites to a certain extent. PKN1 is an important regulator of growth and metabolism, which participates in signal transduction, the cell cycle, cell growth, and gene expression . The PKNs genes in Arabidopsis and Pharbitis had been found to play a role in cell division, regulating the differentiation of shoots, leaves and roots (Kobayashi et al., 2000). The downregulation of IgPKN1 under green light may partially inhibit other physiological metabolic processes, promoting the accumulation of fucoxanthin in I. galbana cells. These findings will facilitate in-depth understanding the molecular regulation mechanisms of fucoxanthin in I. galbana and its role in response to green light regulation, providing technical support for the construction of high fucoxanthin content strains.
Data availability statement mRNA and ATAC sequencing data have been deposited at the National Genomics Data Center, Beijing Institute of Genomics, Chinese Academy of Sciences, under BioProject accession number CRA008916 and CRA008876 (https://ngdc.cncb.ac.cn/gsa/). The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

Author contributions
YC and TX designed and coordinated the entire project. YH, XZ, JingC, YX, and TC performed the collection and processing of samples. TX, DC, HL, XC, and JiannanC performed the omics analysis. TX, HL,