Time‐Series RNA‐Seq of Acropora tenuis Reveals Molecular Waves Leading to Synchronous Mass Spawning of Scleractinian Corals

ABSTRACT

Although mass, synchronised spawning of scleractinian corals is a well‐known phenomenon, its underlying molecular mechanisms, especially those that achieve synchronous release of gametes, are still unknown. In the genus Acropora, the dominant scleractinian coral in shallow reefs, spawning timing is synchronised in any given location, but often varies among locations. Here, we report gene expression cascades potentially driving synchronous mass spawning, revealed through transcriptome monitoring of Acropora tenuis , tracking both daily and monthly dynamics during a year‐long experiment that included two spawning events. We identified 236 genes in four waves of molecular events that culminated in spawning. First, receptor function and prostaglandin secretion became active 2 weeks before spawning and gradually decreased, but remained elevated until spawning, suggesting communication of maturation among colonies. Second, 1–2 weeks before spawning, TGF‐β signalling and spermiogenesis involving kinases were activated. Third, sperm capacitation and preparation of egg‐sperm bundle material commenced a week before spawning. Finally, activation of transcription factor ELF1 triggered a signal cascade that induced spawning. This moonlight‐independent system may serve to fine‐tune the timing of spawning and may explain the broad geographic distribution and ecological success of Acropora, making it the most diverse and abundant genus of scleractinian corals in reef ecosystems.

1. Introduction

Organisms have evolved in response to geophysical cycles of varying periodicities (Raible et al. 2017). Among marine organisms, synchronised reproduction linked to the lunar cycle is a widespread strategy that enhances reproductive success (Babcock et al. 1986; Hagman and Vize 2003; Hoffman et al. 2011; Ritson‐Williams et al. 2005). Due to its critical ecological and evolutionary significance, many studies have sought to identify mechanisms that drive these rhythms. If a single environmental factor, such as moonlight, serves as the primary cue for synchronised reproduction, a relatively simple molecular system involving a biological clock is assumed. For example, in marine bristle worms, which reproduce monthly during a specific lunar phase under laboratory conditions, endogenous molecular clock genes, including a cryptochrome and an opsin, ensure precise entrainment (Zurl et al. 2022). However, molecular mechanisms of marine organisms underlying lunar‐regulated rhythms have remained obscure, partly due to the complexity of interacting environmental cues.

During synchronised mass spawning of corals, the largest reproductive event on Earth (Harrison 2011), hundreds to thousands of colonies spawn simultaneously on just one to several nights around a full moon (Babcock et al. 1986). Coral mass spawning is influenced by environmental factors such as water temperature, tides, wind speed, and moonlight (Fogarty and Marhaver 2019; Sakai et al. 2020). Upregulation of blue‐light‐sensing photoreceptors, such as cryptochromes, under a full moon compared to during a new moon (Levy et al. 2007) led to the hypothesis that light‐sensing molecules mediate synchronised mass spawning (Kaniewska et al. 2015; Rosenberg et al. 2017). However, it remains unclear whether observed gene expression changes are directly involved in regulating annual spawning or these simply reflect normal circalunar rhythms. A recent study on the coral, Dipsastraea speciosa, proposed that spawning date is determined by the presence of a dark period between sunset and moonrise (Lin et al. 2021). Nevertheless, this mechanism does not apply to Acropora, the most diverse and abundant stony coral genus. In Acropora, spawning timing is complex, varying among locations, species, and years (Lin and Nozawa 2017; Sakai et al. 2024), sometimes even occurring near the new moon (Monfared et al. 2023). Actually, coral mass spawning occurs in the rainy season in Okinawa, indicating that moonlight intensity varies with weather, so the moonlight cycle cannot act as a stable cue for synchronous spawning. In addition, spawning synchronicity is high in the subtropical zone, whereas it is low in tropical and temperate zones (Gouezo et al. 2020; Mezaki et al. 2007). These facts suggest that environmental factors influence spawning timing of Acropora corals, but how Acropora corals fine‐tune spawning timing remains unresolved.

In this study, to clarify how synchronised spawning is achieved in Acropora corals, we conducted a time‐series RNA‐seq analysis during a year‐long experiment with captive Acropora tenuis . By extending our observations beyond the traditional spawning season, we captured the full spectrum of gene expression changes throughout the annual cycle. This comprehensive dataset represents the most detailed transcriptome analysis to date, identifying genes potentially involved in orchestrating synchronous mass spawning.

2. Materials and Methods

2.1. Preparation of Acropora tenuis Colonies and Sample Collection

Fragments (approx. 30 cm in diameter) of six gravid colonies of A. tenuis were collected around Ishigaki Island, Okinawa, Japan in April 2018 under permit from the Okinawa Prefectural Government (Permit No. 29–74) and were maintained in an aquarium with flow‐through seawater throughout the experiment at the Yaeyama Station, Fisheries Technology Institute (Okinawa, Japan). The aquarium was 117 cm long, 78 cm wide and 38 cm high (water depth was ~30 cm). Seawater was pumped from 200 m offshore. Semi‐transparent roof panels shielded corals from rain but allowed exposure to natural light (Figure S1). Branch tips (~3 cm long) were collected from each colony around 4–6 PM every 1–2 days, between April and May, or once a month between June and March in 2018 and 2019 (Figure 1). All fragments were preserved with ~25 mL of RNAlater stabilisation solution (Thermo Fisher Scientific) for 24 h and then stored at −80°C until RNA extraction.

FIGURE 1.

Overview of experimental design and annual transcriptome dynamics of Acropora tenuis . (A) Spawning was observed around 8 PM on May 20th, 2019. (B) Surface seawater temperature around northern Ishigaki Island in 2018 and 2019. The grey line indicates values collected by the Japan Meteorological Agency. The black line indicates the trend line generated by the ‘stat_smooth’ function in ggplot2. The experimental period is highlighted with a red dashed line. (C) Overview of moon phase (yellow: Full moon; black: New moon), the day of branch tip collection (grey box), and the day of spawning in the tank (‘S’: Mass spawning, ‘PS’: Partial spawning). (D) UpSet plot showing the number of genes exhibiting highly variable expression. Each row contains genes represented by black circles. Lines connecting black circles indicate genes shared between colonies. Orange indicates 293 highly variable genes common to all colonies. Standard deviations (SD) of gene expression in each colony are shown with bar plots (right upper). The dashed line indicates SD = 1. (E) The number of highly variable genes with maximum monthly expression levels throughout the year. Numbers of genes shared by all colonies are shown at the top. (F) Enriched biological processes of 160 genes that were synchronously upregulated in all colonies. All terms exhibited p values < 0.05.

2.2. RNA Extraction and Sequencing

All fragments were crushed into powder in liquid nitrogen with a mortar and pestle. Total RNA was extracted from the powder using an RNeasy Plant Mini Kit (Qiagen), including a DNase digestion step. A Collibri 3′ mRNA Library Prep Kit for Illumina (Thermo Fisher Scientific) was used for sequencing library preparation. Sequencing adaptors were attached by PCR amplification with 16 cycles of annealing, according to the manufacturer's protocol. Each library was sequenced on a NovaSeq 6000 (Illumina) with 50‐bp, single‐end reads.

2.3. Seasonal Expression Analyses

Low‐quality reads (quality score < 20 and length < 20‐bp) and Illumina sequence adaptors were trimmed with CUTADAPT v4.3 (Martin 2011). Then, clean reads were mapped to A. tenuis gene models v1.0, downloaded from the OIST Marine Genomics Unit browser, using BWA v2.2.1 (Vasimuddin et al. 2019) with default settings. Transcript abundances of all genes in each sample were determined using Salmon v1.5.2 (Patro et al. 2017) with default settings. Salmon output (the number of mapped reads) was normalised to reads per million (RPM) using EdgeR (Robinson et al. 2010) in R v4.3.2. Since this study employed 3′ mRNA sequencing technology, each read corresponds to a single transcript. Therefore, we use TPM (transcripts per million) instead of RPM in this study. Samples with fewer than 1 million reads were excluded from further analyses. In total, 200 samples with more than 1 million RNA‐seq reads mapped to A. tenuis gene models were used in this study (Table S1). Seasonal expression analysis was carried out by following the method of Nagano et al. (2019), with minor modifications. We compared results among individuals, ensuring that all genes identified in this study were validated across six biological replicates, thereby providing a robust dataset. Expression levels were averaged monthly for each colony. Genes with mean Log₂ (TPM + 1) > 2 in all samples in each colony were considered expressed, and genes expressed in more than five colonies were used for further analysis (Figure S2).

In terms of monthly dynamics, genes in individual colonies with standard deviations (SD) Log₂ (TPM + 1) > 1 were considered variable. Genes common to all colonies were considered ‘stable’ (SD < 1) or ‘variable’ (SD > 1) (Figure 1D; Figure S3). An SD of Log₂ (TPM + 1) > 0.5 was used to detect as many genes as possible when analysing spawning season data (Figure S4).

2.4. Embryonic Transcriptome Analyses

Publicly available RNA‐seq data of A. tenuis eggs, blastula, gastrula, planula larvae, polyps and a fragment collected from nongravid adult (Yoshioka et al. 2022) were utilised in this study. Data processing, including trimming of low‐quality reads and sequencing adaptors, read mapping, and read quantification, was performed as related above. Genes that met the following two conditions were considered embryonic genes: (1) expression levels in eggs were more than 1 TPM; (2) expression levels in eggs were larger than those in adults. A list of genes expressed in eggs is provided in Table S2.

2.5. Functional Annotation of Acropora tenuis Gene Models

Gene models of A. tenuis were annotated with eggNOG‐mapper v2.1.12 (Cantalapiedra et al. 2021), based on the eggNOG orthology database v5.0 (Huerta‐Cepas et al. 2019) and BLASTP searches (Camacho et al. 2009) against Swiss‐Prot (E‐value cut‐off: 1e‐3). If an Acropora gene was not annotated by eggNOG‐mapper, a candidate gene name was assigned based on a BLAST search with the words ‘X‐like’ appended. GO functional enrichment analysis was performed on the web platform DAVID v6.8 (Sherman et al. 2022). UniProt IDs from all A. tenuis genes were used as the background dataset in the enrichment analysis, and UniProt IDs assigned to genes of interest were analysed.

3. Results

3.1. Annual Transcriptome Dynamics of Acropora tenuis

We maintained six colonies of A. tenuis in a tank for approximately 1 year until the end of May 2019 (Figure 1A,B), when the next annual mass spawning occurred in the tank (Figure 1C). Throughout the experimental period, we collected branches from each colony at 4–6 PM, several hours before the spawning time (around 8 PM), and investigated their transcriptome dynamics using 200 RNA‐seq samples (Figure 1C; Table S1). Among 14,393 expressed genes, 3612 genes, including housekeeping genes such as ‘composition of Golgi membrane’ and ‘ribosome’, were stably expressed in A. tenuis throughout the year, regardless of the season (SD < 1, Figure S3) as observed in land plants (Dai et al. 2023; Nagano et al. 2019). When we focused on genes with highly variable expression patterns through the year, 293 genes were common to all colonies (SD > 1, Figure 1D). No upregulation of stress‐related genes, for example, immunity‐related genes, was observed, indicating that repeated sampling of branches did not induce stress in these corals. Surprisingly, 81%–86% of these genes (236–252 genes) reached their expression maximum in May or April in each colony (Figure 1E; Figure S5; Table S3). Furthermore, expression levels of 160 of those genes reached their zenith in all samples in May, the spawning month, in all colonies (Figure 1E). Genes involved in gametogenesis such as ‘spermatid development’ and ‘neuromuscular junction development’ were statistically enriched (p < 0.05) among the 160 synchronously upregulated genes in May (Figure 1F). These biological functions are involved in spermiogenesis in the coral, Fimbriaphyllia ancora (Chiu et al. 2020), suggesting that variable genes upregulated in May are mainly related to gametogenesis.

3.2. Transcriptomic Signature of A. tenuis During the Spawning Season

When we analysed transcriptome dynamics focused on the spawning season (April to May) in 2018 and 2019, we found that 1268 genes exhibited highly variable expression patterns in all colonies in the 2018 samples and 2324 genes in the 2019 samples (SD > 0.5, Figure 2A,B; Figure S4). The difference in variable gene number between 2018 and 2019 may reflect the culturing environment. Colonies spawned in 2018 were kept in tanks throughout the year until the 2019 spawning. When repertoires of variable genes identified in the 2018 and 2019 spawning were compared, 731 genes were common to both (Table S4). To eliminate contamination from gamete or bundle RNA, we removed genes expressed in A. tenuis eggs. Then, we identified 441 of those 731 genes (Figure 2C; Figure S6) that are highly likely to be involved in signal cascades for spawning in A. tenuis .

FIGURE 2.

Gene expression dynamics of A. tenuis during the spawning season. (A, B) Genes for which expression levels varied from April to May (spawning season) in 2018 and in 2019. n indicates gene number. (C) Variable genes that were common between the 2018 (1268 genes) and 2019 (2324 genes) spawning seasons. (D) NMDS plot based on gene expression levels showing the three clusters of spawning‐related genes. n indicates gene number. Ellipses showing 95% confidence intervals for individual clusters are drawn using ‘stat_ellipse’ of the ggplot package. (E) Numbers of COG categories assigned to spawning‐related genes in each cluster.

Based on gene expression levels from April to May for the 441 genes, four clusters (C1–C4) were distinguished (Figure S7). Gene expression patterns in each cluster showed that C1 genes were most highly expressed approximately 2 weeks before spawning and that expression gradually decreased as spawning approached. C2 genes reached their greatest expression levels approximately a week before spawning, and C3 genes were continuously expressed from 2 weeks before spawning, but decreased immediately thereafter (Figure S8). In contrast, genes in the remaining cluster, C4, maintained certain levels of gene expression even after spawning (Figure S8), suggesting that C4 genes may not be directly related to spawning. Thus, we focused on C1–C3 (236 genes) as ‘spawning‐related genes’ (Figure 2D).

Putative gene functions were estimated for 33 of 46 genes in C1, 55 of 104 genes in C2, and 71 of 86 genes in C3 (Table S5). Although only a few functional terms were enriched for C1 and C3 genes (Table S6), clusters of orthologous groups (COG) revealed that genes involved in ‘signal transduction’ predominated in each cluster (8 genes in C1, 24 8 in C2, and 15 8 in C3) (Figure 2E; Table S5), suggesting that certain cascades in cellular pathways are activated by signal transduction leading to spawning in A. tenuis . The TGF‐β signalling pathway is widely conserved in metazoans and is essential for a variety of cellular functions ranging from embryo development to injury repair (Massagué and Sheppard 2023; Wu and Hill 2009). The two key‐regulator proteins of the TGF‐β signalling pathway, fibrillin‐1 (FBN1) (aten_s0017.g147) and latent‐transforming growth factor beta‐binding protein (LTBP)‐like genes (aten_s0003.g20; aten_s0130.g7), found in C1 and C2, respectively, were activated approximately 7 days before spawning, suggesting involvement of TGF‐β signalling. Mutation of these genes leads to reduced TGF‐β signalling and phenotypic changes in mammals (Rifkin et al. 2022). Four testis‐specific serine/threonine‐protein kinase (Tssk)‐like genes (aten_s0068.g76; aten_s0109.g29; aten_s0150.g36; aten_s0166.g26), required in later steps of spermatogenesis and/or spermiogenesis in mice (Shang et al. 2010), were also included in C2 (Table S5), suggesting that the final step of spermatogenesis occurs around a week before spawning. Consistent with the enriched terms ‘male meiotic nuclear division’ in C1 genes and ‘protein phosphorylation’, ‘flagellated sperm motility’ and ‘sequestering of TGFbeta in extracellular matrix’ in C2 genes (Table S6), genes in C1 and C2 may be associated with spermiogenesis and TGF‐β signalling, preparatory to mass spawning.

Expression levels of C1 and C2 genes returned to baseline levels about a week before spawning, while expression of C3 genes remained constant until spawning (Figure 3A), suggesting that C3 genes may be involved in determining the spawning day. Five receptor‐like genes were detected in C3, whereas one and two receptor‐like genes were in C1 and C2, respectively (Table S5), suggesting activation of receptor functions around spawning, as observed in previous studies showing upregulation of genes encoding G‐protein‐coupled receptors (GPCRs) during bundle setting and spawning in Acropora spp. (Kaniewska et al. 2015; Rosenberg et al. 2017). Previous studies also highlighted the importance of photoreceptors as upregulated GPCRs, for example, melanopsins (Kaniewska et al. 2015; Rosenberg et al. 2017), and moonrise time is crucial for synchronised spawning in the coral, Dipsastraea (Lin et al. 2021). However, we detected no upregulation of photoreceptor‐like genes in A. tenuis (Table S5), although two GPCRs (aten_s0153.g33 and aten_s0035.g82) were in C3, supporting the hypothesis that spawning is potentially mediated by GPCR signalling cascades in Acropora, as previously suggested, but that these receptors receive signals other than moonlight. Indeed, in this study, major spawning occurred 6 days after the full moon in 2018 and the next day after the full moon in 2019 (Figure 1B).

FIGURE 3.

Expression patterns of three major clusters and a schematic of gene expression changes during the spawning season. (A) Relative gene expression was determined in comparison with expression levels on the spawning day. Trends for each gene were extracted using the ‘stat_smooth’ function in ggplot2. Blue lines indicate the spawning day. (B) A time series of gene expression changes leading to mass spawning. (1) activation of TGF‐β signalling and spermiogenesis with kinases occurs 1–2 weeks before spawning; (2) active cell division with spermatogenesis and prostaglandin secretion from 2 weeks before spawning until spawning; (3) sperm capacitation and preparation of egg‐sperm bundle material during the week before spawning; and (4) activation of transcription factor ELF1 to trigger signal cascades that induce spawning.

We further subdivided C3 genes into four groups, based on expression patterns: C3a (70 genes), in which median gene expression peaked 7 days before spawning, but continued afterward; C3b (5 genes), in which the highest gene expression was observed from a week before spawning until the spawning day; C3c (5 genes), characterised by a rapid increase in expression on the spawning day; and C3d (6 genes), in which gene expression was unstable during the spawning season (Figure S9). Given that gene expression levels continue to increase until the spawning day, genes in C3a and C3b may be involved in determining the spawning date. For C3a genes, enriched terms ‘prostaglandin secretion’, ‘regulation of DNA damage checkpoint’, and ‘cell division’ were statistically significant (p < 0.05) (Table S6).

3.3. Genes Increasing Their Expression Levels as Spawning Approached

C3b, which exhibited gradual upregulation from 1 week before until the spawning day, included 5 genes: dopamine receptor D5 (DRD5)‐like (aten_s0035.g82); voltage‐gated delayed rectifier potassium channel KCNH4 (KCNH4: aten_s0210.g5); protein phosphatase 1 regulatory subunit 27 (Ppp1r27)‐like (aten_s0115.g19); thyroxine 5‐deiodinase (DIO3: aten_s0035.g81); and vitellogenin 1 (Vg1: aten_s0102.g40) (Figure 3A; Figure S10). The dopamine receptor, DRD5, activates adenylate cyclase, leading to the production of cAMP, which subsequently triggers downstream protein kinase A (PKA) (Beaulieu and Gainetdinov 2011). The activity of PKA promotes sperm motility in corals (Speer et al. 2021). A potassium channel, KCNH4, is thought to be involved in the adjustment of pH in sperm cytosol (Christen et al. 1982). DIO3 is one of the enzymes involved in the activation of thyroid hormones, which influence many aspects of physiology (St. Germain et al. 2009), including daily sperm production (Martinez et al. 2016). Taken together, these three genes may be involved in sperm capacitation and a series of biological processes occurring during the week before spawning.

On the contrary, Vg, the major yolk protein precursor (Shikina et al. 2013; Tan et al. 2020), has been identified in A. tenuis and its expression tends to correlate with maturation of oocytes (Tan et al. 2021). The presence of multiple Vg copies negates the classical view of vitellogenin as a simple source of nourishment for developing embryos. In fishes, it serves as a multivalent pattern recognition receptor (Carducci et al. 2019). Since the expression peak of Vg2 (aten_s0086.g39) was observed in April, Vg2 has the canonical function of vitellogenin (Figure S11). Vg1 (aten_s0102.g40), which showed higher expression from a week before spawning until spawning, likely possesses a different function (Figure S11). Similar expression patterns of Vg1 and Vg2 have also been reported in A. tenuis near Sesoko Island, Okinawa, Japan (Takekata et al. 2022), suggesting that the result obtained in this study could probably apply to A. tenuis in other locations. Most broadcast spawning scleractinians, including A. tenuis , release bundles during spawning, in which eggs cohere to bundle materials (sulfated mucosubstances), with a central cavity for sperm (Okubo and Motokawa 2007). Bundle materials may be secreted by oocytes (Padilla‐Gamiño et al. 2011), and Vg1 may be involved in bundle formation in Acropora.

C3c exhibits a rapid increase in expression on the spawning day and includes five genes, ETS‐related transcription factor Elf‐1 (ELF1: aten_s0006.g122), a gene homologous to a gene of unknown function (aten_s0389.g14) and genes with no annotation (aten_s0010.g40; aten_s0295.g162; aten_s0118.g7) (Figure 3A; Figure S12). ELF1 is associated with diverse transcriptional programmes, for example, embryogenesis, antiviral activity and tumour progression, in mammals (Bassuk et al. 1998; Paczkowska et al. 2019; Seifert et al. 2019). Although further validation is needed, ELF1, specifically upregulated on the spawning day, may be one of the final switches in the signal cascade involved in synchronised mass spawning in A. tenuis .

4. Discussion

Both in abundance and diversity, Acropora corals are overwhelmingly dominant in shallow reefs in the Indo‐Pacific, with more than 100 described species (Wallace 1999). This dominance is thought to be enabled by synchronised mass spawning, which increases fertilisation success and maximises larval recruitment. Unlike other corals that use the moonrise as their main cue (Lin and Nozawa 2017; Lin et al. 2021), this synchronous mechanism is thought to be achieved through mutual signalling between neighbouring colonies via hormones (Twan et al. 2006), but it has been unclear how this is actually accomplished. The present results in A. tenuis suggest that genes with different expression patterns (C1, C2, C3a, C3b and C3c) work together to achieve synchronous spawning commencing 2 weeks before the event (Figure 3B). The main processes are as follows. First, receptor function and prostaglandin secretion become active 2 weeks before spawning and gradually decrease from their peak but remain elevated until spawning (C3a). Second, TGF‐β signalling and spermatogenesis‐related kinases are activated one to 2 weeks before spawning (C1 and C2). Third, sperm capacitation and preparation for bundle formation begin a week before spawning (C3b). Finally, molecular pathways are activated through several genes (C3c), triggering spawning on the actual night. Previous studies showed that hydrogen peroxide could be used to induce artificial spawning 16–28 h before natural spawning (Hayashibara et al. 2004; Suzuki 2020). Together with the findings, we assume that corals can fine‐tune the timing of spawning during Phase III, when sperm capacitation and bundle preparation occur. Importantly, the increase in expression of C3 that is accompanied by a wave‐like increase and decrease suggests that surrounding colonies may adjust the pace of their preparations for spawning commencing 2 weeks beforehand. We propose that regulation of maturation signals through gene expression enables Acropora corals to synchronise spawning among neighbouring conspecific colonies without relying solely on the lunar phase. In future studies, causality between expression changes of candidate genes and spawning timing should be addressed.

For most Acropora species, the spawning date is unpredictable as it varies from year to year (Monfared et al. 2023). Therefore, the proposed molecular mechanism is likely widely conserved within the genus, although some species‐specific modifications may account for differences in spawning timing among species. Divergence of the genus Acropora is estimated to have occurred during the Eocene and Oligocene (around 25–50 million years ago) (Shinzato et al. 2021), a period when the environment was different from today, with global temperatures thought to have been 5°C–8°C higher (McInerney and Wing 2011). Development of moonlight‐independent spawning may explain the broad geographic distribution and ecological success of the genus Acropora, making it the most diverse and abundant group of scleractinian corals in coral reef ecosystems. For other scleractinian corals, spawning timing also varies between years, for example, Montipora, Galaxea and Lobophylia. Thus, the proposed molecular mechanisms may apply to other coral genera for fine‐tuning the timing of synchronous spawning.

Author Contributions

G.S. designed research. Y.Y., G.S., Y.F., S.T., T.U. and C.S. performed the research. E.S. and N.S. contributed new reagents and analytical tools. Y.Y. analysed data. Y.Y., G.S. and C.S. wrote the paper.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figures S1–S12: mec70054‐sup‐0001‐Supinfo.docx.

Tables S1–S6: mec70054‐sup‐0002‐Tables.xlsx.

Data Availability Statement

Raw RNA‐sequencing data have been deposited in the DDBJ/EMBL/GenBank databases under accession numbers DRR683534–DRR683733 (BioProject ID: PRJDB20729). There are no benefits to report.