Sea cucumbers and their symbiotic microbiome have evolved to feed on seabed sediments

Abstract

Sea cucumbers are predominant deposit feeders in benthic ecosystems, providing protective benefits to coral reefs by reducing disease prevalence. However, how they receive sufficient nutrition from seabed sediments remains poorly understood. Here, we investigate Holothuria leucospilota, an ecologically significant tropical sea cucumber, to elucidate digestive mechanisms underlying marine deposit-feeding. Genomic analysis reveals intriguing evolutionary adaptation characterized by an expansion of digestive carbohydrase genes and a contraction of digestive protease genes, suggesting specialization in digesting microalgae. Developmentally, two pivotal dietary shifts, namely, from endogenous nutrition to planktonic feeding, and from planktonic feeding to deposit feeding, induce changes in digestive tract enzyme profiles, with adults mainly expressing carbohydrases and lipases. A nuanced symbiotic relationship exists between gut microbiota and the host, namely, specific resident bacteria supply crucial enzymes for food digestion, while other bacteria are digested and provide assimilable nutrients. Our study further identifies Holothuroidea lineage-specific lysozymes that are restrictedly expressed in the intestines to support bacterial digestion. Overall, this work advances our knowledge of the evolutionary innovations in the sea cucumber digestive system which enable them to efficiently utilize nutrients from seabed sediments and promote food recycling within marine ecosystems.

Sea cucumbers are predominant deposit feeders in benthic ecosystems. This study elucidates the mechanisms within the sea cucumber digestive system and their symbiotic microbiome which enable them to efficiently utilize nutrients from seabed sediments.

Introduction

Sea cucumbers, classified as Holothuroidea within the phylum Echinodermata, are large and abundant members in marine benthic communities¹. Sea cucumbers greatly impact the health and integrity of the marine ecosystem through bioturbation, organic matter processing, nutrient recycling, seawater chemistry balancing, biodiversity supporting, energy transfer in food chains^1,2, and provide significant protective benefits to coral reefs by reducing their diseases³. Most sea cucumbers are deposit feeders, acquiring nutrients through the consumption of bacteria, microalgae, decayed plants, and meiofauna, along with other organic detritus^4–6. By ingesting and defecating a large amount of seabed sediment, sea cucumbers facilitate bacterial decomposition and reduce the organic load^4,7. To date, the specific mechanisms by which sea cucumbers obtain sufficient nutrition from the marine benthic deposit remain unclear.

The digestive tract is the primary site for food digestion and nutrient absorption in bilateral animals⁸, where a diverse array of digestive enzymes are secreted and participate in food decomposition⁹. To adapt to changes in habitats or food sources, animals may alter their diets accordingly. During this process, the adaptive evolution of digestive enzymes can also occur at the genetic level¹⁰. In beetles, as their diet shifted towards plants, an expansion of lineage-specific detoxification enzyme genes for countering plant-derived toxic chemicals has been observed¹¹. In modern cetaceans, positive selections in protease and lipase genes have evolved to enhance their capacity for the digestion of proteins and lipids, enabling dramatically change from herbivory to carnivory¹². Starch has emerged as a prominent dietary component in agricultural societies, leading to genetic variation in the amylase gene copy numbers in human¹⁰ and domesticated animals like dogs¹³. However, the lack of genomic information in the past limited comprehensive insights into the genetic-level changes of digestive enzymes in marine deposit feeders. Recent sequencing of sea cucumber genomes, such as Apostichopus japonicus¹⁴, Holothuria leucospilota¹⁵ and Holothuria glaberrima¹⁶, has made these studies possible.

Gut microbiota is an integral component of the digestive system as they provide supplementary digestive enzymes¹⁷. For instance, herbivorous animals like cattle harbor bacteria in their rumen to produce cellulase for cellulose digestion¹⁸. Similar microbiota-derived cellulases have also been found in other animals that rely on cellulose as a nutritional source, including marsupials¹⁹, termites²⁰, and wood-feeding beetles²¹. The symbiotic gut microbiota can also carry out other specific digestive functions for the host²², such as providing pectin-degrading enzymes for breaking down pollen walls in honeybees²³, and offering dietary nitrogen supplementation in herbivorous ants²⁴. In addition, gut microbiota can detoxify plant defensive secondary metabolites for frugivorous birds²⁵. On the other hand, gut microbiota may become a source of pathogenicity and toxicity to the host, thereby facilitating the development of host-specific intestinal immune and detoxification systems²⁶. In the burying beetles, the gut is strictly compartmentalized into parts for digestion, immune defense and detoxification, permitting rapid digestion of food and suppression of microbial growth²⁷. For marine deposit feeders, their food sources are rich in bacteria. Yet, how they control and utilize bacteria derived from the food remains mechanistically unclear.

H. leucospilota is an ecologically significant tropical sea cucumber species widely distributed in the West Pacific and Indian Oceans¹⁵. H. leucospilota lives as an active seabed deposit feeder (Fig. 1a), capable of ingesting and defecating seabed sediment up to 88.8 g·ind⁻¹·d⁻¹²⁸. In this study, through a combined analysis of genome, transcriptome and microbiome, we investigated evolutionary innovations of the sea cucumber digestive system regarding marine deposit feeding, which encompass genetic alterations, digestive tract region-specific enzyme features, developmental enzyme profile changes, and digestive contributions from gut microbiota. Additionally, we identified Holothuroidea lineage-specific intestine-expressed lysozymes that would control bacterial proliferation and convert them into a food source. Our study thus provides insights into the evolutionary and adaptative mechanisms underlying how sea cucumbers efficiently utilize nutrients from seabed sediments, enabling them to play crucial roles in food recycling within marine ecosystems.

Fig. 1. The landscape of feeding habits and digestive enzymes of the tropical sea cucumber H. leucospilota.

a Scene of H. leucospilota feeding on seabed sediments. b Relative abundance of the 10 most abundant eukaryotic phyla in the digestive tract contents determined by 18S sequencing (Supplementary Data 1). The digestive tract regions include foregut (Fg, n = 5), midgut (Mg, n = 5) and hindgut (Hg, n = 5) collected from different individuals. Each color represents one specific eukaryotic phylum. Source data are provided as a Source Data file. c Alpha-diversity index of microbiota in the environment (En, n = 10 in total), feces (Fc, n = 5) and digestive tract (DT, n = 19 in total) determined by 16S sequencing (Supplementary Data 2), displayed by the Chao1 estimator and Shannon diversity index. The environment samples include seawater (Sw, n = 5) and seabed sediments (SS, n = 5). The first quartile forms the bottom and the third quartile forms the top of the box, in which the line represents the median value. The whiskers range from 2.5^th to 97.5^th percentile, and points below and above the whiskers are drawn as individual dots. P values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test, where only the P values between SS and Fc groups are shown (****P < 0.001). d The numbers of digestive enzyme amylase (AMY), maltase-glucoamylase (MGA), sucrase-isomaltase (SUIS), lactase (LPH), trehalase (TREA), chymotrypsins (CTR), chymotrypsin-like elastase (CLE), pancreatic triacylglycerol lipase (LIPP), gastric triacylglycerol lipase (LIPG), hepatic triacylglycerol lipase (LIPC), chitinase (CHIA) and chitinase domain-containing protein (CHIP) genes in 23 Deuterostomia species. The size of the circles represents the number of digestive enzyme genes of a particular category (Supplementary Data 3). The squares indicate the feeding habits of different species, while the clusters indicate their evolutionary status. Arrows and bubbles on the right indicate whether a certain digestive enzyme group has undergone gene expansion or contraction within a certain taxonomic group.

Results

Food composition and the correlated expansion/contraction of digestive enzyme genes

The eukaryotic and prokaryotic food compositions of wild H. leucospilota were analyzed by 18S (Fig. 1b; Supplementary Fig. 1a, d) and 16S (Fig. 1c; Supplementary Fig. 1b, e) amplicon sequencing, respectively. By 18S sequencing, a total of 373 amplicon sequence variants (ASVs) were obtained from the eukaryotic composition of gut contents. Dinophyceae was the dominant eukaryotic food source for H. leucospilota, exhibiting the highest abundance across all digestive tract regions, whereas Ciliophora and Basidiomycota were primarily detected in the foregut, and Arthropoda and Ascomycota were predominantly found in the midgut (Fig. 1b; Supplementary Data 1). By 16S sequencing, the microbiota distributions with 13056 ASVs in seawater, seabed sediments, sea cucumber feces, and different regions of the digestive tract, including the esophagus, foregut, midgut, hindgut and rectum were demonstrated (Fig. 1c; Supplementary Data 2). The β-diversity, as presented by PCoA, indicated that the microbiota in the gut contents of H. leucospilota closely resembled those in the environmental seabed sediments (Supplementary Fig. 2a), particularly in the contents of the esophagus (Supplementary Fig. 2b). The α-diversity of the microbiota, measured by the Chao1 richness estimator and the Shannon-Wiener diversity index, exhibited a significant decrease in the excreted feces when compared to the ingested seabed sediments (Fig. 1c; Supplementary Fig. 3), suggesting that a considerable portion of the microbiota served as the sea cucumber’s food and underwent digestion within the digestive tract.

The H. leucospilota genome contains various digestive enzyme genes, specifically, 8 amylases (AMY), 6 maltase-glucoamylases (MGA), 3 sucrase-isomaltases (SUIS), 5 lactases (LPH), 1 trehalase (TREA), 3 chymotrypsins (CTR), 1 chymotrypsin-like elastase (CLE), 8 pancreatic triacylglycerol lipases (LIPP), 2 gastric triacylglycerol lipases (LIPG) and 2 chitinase domain-containing proteins (CHIP). However, it does not possess genes for pepsin (PEP), trypsin (TRY), hepatic triacylglycerol lipase (LIPC), chitinase (CHIA) or cellulase (CEL) (Supplementary Data 3). A cross-genomic analysis was conducted to illustrate the expansion/contraction of those digestive enzyme genes across 23 deuterostomia species (Fig. 1d; Supplementary Data 3). Compared to vertebrates, large-scale expansions of digestive carbohydrase genes and contractions of digestive protease genes are observed in the genomes of Ambulacraria, which include echinoderms and hemichordates.

Developmental dietary shifts and corresponding digestive enzyme expression

During development from embryo to larva, juvenile and adult stages, H. leucospilota experiences motilities of non-swimming, swimming, attaching and bottom-crawling lifestyles, and changes nutritional sources from endogenous nutrition to planktonic feeding during the larval stage, and to deposit feeding after attachment (Fig. 2a). Through transcriptomic analysis, 11 of the 39 digestive enzyme genes in the H. leucospilota genome were shown to have no expression across all developmental stages, while 2 genes were exclusively expressed during the embryonic and larval stages, and 8 genes began to exhibit expression only till the juvenile and adult stages (Fig. 2b). Expression level analysis showed a sequential expression pattern of three sets of digestive enzymes, corresponding to the transition from endogenous nutrition to planktotrophic feeding, and eventually to deposit feeding (Fig. 2c; Supplementary Data 4). Specifically, lipases and chitinase homologs were expressed during the endogenous nutritional stage; carbohydrases were predominantly expressed during the planktotrophic feeding stage; carbohydrases, proteases, and lipases were expressed during the deposit-feeding stage, indicating changes in the expression pattern of digestive enzymes which accommodates dietary shifts during development.

Fig. 2. The dietary shift and digestive enzyme expression patterns during the embryonic and larval development of H. leucospilota.

a Developmental stages include the fertilized egg (FE), 2-cells (2C), 4-cells (4C), 8-cells (8C), 16-cells (16C), morula (Mr), blastula (Bs), rotated-blastula (RB), early-gastrula (EG), late-gastrula (LG), early-auricularia (EA), mid-auricularia (MA), auricularia (Ar), doliolaria (Dl), pentactula (Pt), 1-mm juvenile (J1), 20-mm juvenile (J20) and adult (A). The orange text represents different developmental stages, including embryos, larvae, juveniles, and adults. The blue text represents the different locomotion models, including non-swimming, swimming, attaching and bottom-crawling. The green text represents the sources of nutrition, including endogenous nutrition, planktotrophic feeding, and deposit feeding. b The expressional presence and absence of digestive enzyme amylase (AMY), maltase-glucoamylase (MGA), sucrase-isomaltase (SUIS), lactase (LPH), trehalase (TREA), chymotrypsins (CTR), chymotrypsin-like elastase (CLE), pancreatic triacylglycerol lipase (LIPP), gastric triacylglycerol lipase (LIPG) and chitinase domain-containing protein (CHIP) genes during the fertilized egg (FE), embryonic (E), larval (L), juvenile (J) and adult (A) stages (Supplementary Data 4). Absence or presence in expression of a certain digestive enzyme gene during a certain developmental stage is marked in yellow or blue. c Heatmap illustrating digestive enzyme expression among different developmental stages. The red box indicates three sets of digestive enzyme expression patterns corresponding to the dietary shift (Supplementary Data 4). Blue and red colors represent relatively low and high expression levels, respectively, as scaled by the rows. Clusters in different colors represent different expression patterns for digestive enzyme genes during development. The gene expression level at each stage is derived from the average of samples (n = 3) taken from different individuals. Source data are provided as a Source Data file. d Spatial distribution of CHIP (Hl-20219), SUIS (Hl-27148), and CTR (Hl-19115) mRNA in the fertilized egg (FE), embryos (E), larvae (L) detected by WMISH, and in the juveniles (J) detected by FISH. The intestines (In) in the juvenile sections are indicated. The black and white scale bars for WMISH and FISH are 100 μm and 200 μm, respectively. Each experiment was performed for one time.

The results of whole-mount in situ hybridization (WMISH) and fluorescence in situ hybridization (FISH) showed that CHIP Hl-20219, a chitinase homolog expressed during the endogenous nutritional stage, was distributed in non-intestinal tissues of the embryos and larvae; SUIS Hl-27148, a carbohydrase expressed during the planktotrophic feeding stage, was distributed in the newly formed intestine of the larvae; CTR Hl-19115, a protease expressed during the deposit-feeding stage, was distributed in the fully developed intestine of the juveniles (Fig. 2d). In summary, specific digestive enzymes are expressed in non-intestinal tissues during the endogenous nutritional stage, and in the immature and mature intestines during the planktotrophic and deposit feeding stages, respectively.

Digestive enzyme gene expression and activity in different gut regions

Based on the morphological and histological characteristics, the digestive tract of H. leucospilota was divided into five regions, namely, a short esophagus, followed by a straight foregut, a midgut connected with the rete mirabile, and a long hindgut that terminated in an expanded rectum (Fig. 3a). From the anterior to the posterior, there is a gradual increase in the thickness of the wall of the digestive tract, accompanied by a progressive thinning of the folded brush border (Fig. 3b).

Fig. 3. The structure and digestive enzyme system of the H. leucospilota digestive tract.

a Anatomical structure of the digestive tract, which is further divided into regions including the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc). Foregut, midgut and hindgut make up the intestine (In). Other tissues shown include the body wall (Bw), muscle (Ms), mouth (Mt), anus (An) and rete mirabile (RM). b Histological structures of different regions of the digestive tract, including the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg), and rectum (Rc), as revealed by HE staining. The black scale bars are 100 μm. c Heatmap illustrating the expression of digestive enzyme genes amylase (AMY), maltase-glucoamylase (MGA), sucrase-isomaltase (SUIS), lactase (LPH), trehalase (TREA), chymotrypsins (CTR), chymotrypsin-like elastase (CLE), pancreatic triacylglycerol lipase (LIPP), gastric triacylglycerol lipase (LIPG) and chitinase domain-containing protein (CHIP) in different tissues, including the body wall (BW), muscle (Ms), oral tentacles (OT), Cuvierian organ (CO), respiratory tree (RT), Polian vesicle (PV), coelomocytes (Cc), ovary (Ov), testis (Ts), rete mirabile (RM), transverse vessel (TV) and intestine (In) (Supplementary Data 5). The digestive tract is further divided into five regions, including the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc) (Supplementary Data 6). Blue and red colors represent relatively low and high expression levels, respectively, as scaled by digestive enzyme classifications. The gene expression level in each tissue is derived from the average of tissue samples (n = 3 or 4) taken from different individuals. Source data are provided as a Source Data file. d FISH of MGA (Hl-27400), CTR (Hl-19115), LIPG (Hl-25219), and CHIP (Hl-20219) mRNA in the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc) of the digestive tract. The white scale bars are 400 μm. Each experiment was performed for one time. e Enzyme activity assay of amylase, protease, lipase, chitinase, and cellulase in the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc) of the digestive tract. Data presented here are expressed as mean±SEM (n = 5 from 5 individuals). Source data are provided as a Source Data file. f Diagram illustrating the functional compartmentalization of the H. leucospilota digestive tract, showing different expression and activity patterns for digestive enzymes across the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc).

Tissue transcriptomic analysis showed that the expression of AMY, MGA, SUIS, LPH, TREA and LIPP was predominantly located in the intestine, while CTR, CLE, LIPG and CHIP exhibited rather dispersed expression, indicating that the H. leucospilota digestive tract indeed possesses digestive capabilities on carbohydrates and lipids, but it may be deficient in digesting proteins or chitins (Fig. 3c; Supplementary Data 5).

Based on further analysis of specific transcriptomes for different digestive tract regions, it was found that the foregut exhibited the highest expression levels of digestive enzyme genes on average, followed by the midgut, esophagus, and hindgut (Fig. 3c; Supplementary Data 6). The results of FISH indicated that cells expressing different digestive enzymes were dispersed throughout the brush border, mucosa and muscle layer of the esophagus, but the stronger expression was only found in the brush border of the foregut, midgut, and hindgut (Fig. 3d). The highest activities of amylase, protease and lipase were exhibited in the foregut, with amylase also showing relatively high activity in the esophagus and midgut, and protease and lipase activities were also relatively elevated in the hindgut (Fig. 3e). Chitinase activity was not detected in any digestive tract region (Fig. 3e), suggesting that this enzyme is not involved in intestinal food digestion in adult H. leucospilota. On the contrary, cellulase activity was observed in the esophagus, foregut and midgut (Fig. 3e), despite the absence of cellulase gene in the H. leucospilota genome. Taken together, within the digestive tract, the foregut appears to be the primary site of food digestion, the midgut, connected to the rete mirabile, focuses on nutrient absorption, the hindgut further processes and absorbs proteins and lipids, and the esophagus and rectum respectively manage to sediment ingestion and faces excretion (Fig. 3f). The digestive tract provides partial enzyme activity for food digestion, while the remaining portion may be contributed by symbiotic microorganisms.

Gut resident bacteria provide enzyme activities for food digestion

The microorganisms in the digestive tract contents, feces and surrounding environment of H. leucospilota were analyzed using 16S sequencing (Fig. 4a; Supplementary Data 2). Significant differences in microbial composition were observed among seawater, seabed sediments and sea cucumber feces from phylum to genus levels (Fig. 4b; Supplementary Fig. 4). The α-diversity of microbiota varied across different digestive tract regions of H. leucospilota (Supplementary Fig. 3). The abundance of Proteobacteria, Bacteroidetes and Cyanobacteria, which are prevalent in the ocean environment, significantly decreased after passing through the digestive tract (Fig. 4a). Notably, Bacteroidetes present only in the esophagus and absent in other digestive tract regions. On the contrary, the abundance of Firmicutes and Planctomycetes is low in the environment but higher in the digestive tract (Fig. 4a). The Firmicutes and Planctomycetes in the feces may originate from resident bacteria in the digestive tract, while the high abundance of Acidobacteria in the feces likely derives from seawater. Taken together, Bacteroidetes and Cyanobacteria are considered to be the bacteria digested through the digestive tract, while Firmicutes and Planctomycetes are considered to be the resident bacteria in the digestive tract (Fig. 4b). Proteobacteria, on the other hand, contains both digested and resident bacteria (Fig. 4b).

Fig. 4. Microbiome within the H. leucospilota digestive tract and its contribution to digestive activities.

a Relative abundance of the 10 most abundant phyla along the seawater (Sw, n = 5) and seabed sediments (SS, n = 5) in environments, feces (Fc, n = 5), and different regions of the digestive tract, including the esophagus (Es, n = 4), foregut (Fg, n = 4), midgut (Mg, n = 5), hindgut (Hg, n = 3) and rectum (Rc, n = 3) (Supplementary Data 2). Each color represents a specific microbial group. Source data are provided as a Source Data file. b Phylogenetic relationships and taxonomic classifications of the 1922 microbial species from the digestive tract with active transcription (Supplementary Data 7). c Heat plots displaying the values for mRNA expression of microbial genes involved in protein, carbohydrate, and lipid metabolism in the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc) of the digestive tract (Supplementary Data 8). The size and color of the plots represent relative expression level, with redder and larger plots indicating higher expression levels. Digestive tract region samples were collected from 3 individuals then mixed into 1 for sequencing. Source data are provided as a Source Data file. d Enzyme activity assay of amylase, protease, and lipase in the intestine after antibiotic exposure for 0- (control), 3- and 7-days. Data presented here are expressed as mean±SEM (n = 5 from 5 individuals; P values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test, where **P < 0.01, ***P < 0.001).

Active microorganisms in the H. leucospilota digestive tract were detected using metatranscriptomic sequencing (Fig. 4b; Supplementary Fig. 1c, f). In this case, Proteobacteria (n = 1001) and Firmicutes (n = 373) were found to be the dominant phyla in the H. leucospilota digestive tract, followed by Bacteroidota (n = 187), Actinobacteria (n = 77), Planctomycetota (n = 56) and Verrucomicrobia (n = 32) (Fig. 4b; Supplementary Data 7). Subsequently, the production of digestive enzymes by microorganisms in various digestive tract regions was further analyzed at the transcript level (Fig. 4c; Supplementary Fig. 5, Supplementary Data 8). Genes encoding enzymes involved in the hydrolysis of proteins, carbohydrates, and lipids were identified. Proteases were found to be distributed throughout all digestive tract regions, carbohydrases were primarily distributed in the foregut and midgut, while lipases were mainly distributed in the foregut and hindgut (Fig. 4c). Furthermore, when antibiotics were applied to interfere with the homeostasis of the microbiota, the activities of protease and lipase in the H. leucospilota digestive tract significantly decreased (Fig. 4d). In contrast, amylase activity did not show significant changes (Fig. 4d). These results indicate that the gut microbiota provides an additional portion of the digestive enzymes to sea cucumbers, particularly the proteases that are lacked in the sea cucumber genome.

Holothuroidea-specific intestinal lysozymes digest bacteria into nutrients

To investigate the mechanism for sea cucumbers applied to defend against the pathogenicity and toxicity aroused from bacteria-rich food, the immune and detoxification-related genes were screened throughout the H. leucospilota genome, and their expression in the intestine was examined with transcriptomic analysis. In this case, no specific expression of genes related to respiratory burst, detoxification, lectin, or pattern recognition receptor was observed in the H. leucospilota intestine compared to other tissues (Supplementary Fig. 6, Supplementary Data 9). Conversely, two among a total of five lysozyme genes (Hl-36988 and Hl-36992) exhibited specifically high expression in the intestine, and the other three lysozyme genes (Hl-18105, Hl-18109, and Hl-18110) were primarily expressed in coelomocytes (Fig. 5a; Supplementary Data 5). The transcript levels of the two intestine-expressed lysozymes were highly comparable, with the expression level of Hl-36988 being relatively higher than that of Hl-36992 (Fig. 5a). Within the digestive tract, Hl-36988 was more predominantly expressed in the hindgut, while Hl-36992 is more evenly distributed in the foregut, midgut and hindgut (Fig. 5a; Supplementary Data 6). The results of FISH indicated that both Hl-36988 and Hl-36992-expressed cells were localized in the brush border of the intestine (Fig. 5b). During the embryonic and larval development, expression of Hl-36988 and Hl-36992 began with the appearance of the intestine in the embryos and increased in accordance with the onset of deposit-feeding stage after larval attachment (Fig. 5a; Supplementary Data 4).

Fig. 5. The roles of Holothuroidea-specific intestinal i-type lysozyme in bacterial digestion.

a Heatmap illustrating the expression of lysozyme genes (Hl-18105, Hl-181095, Hl-18110, Hl-36988 and Hl-36992) in different tissues (Supplementary Data 5) and digestive tract regions (Supplementary Data 6), as well as embryonic and larval development stages (Supplementary Data 4). Tissue samples include the body wall (BW), muscle (Ms), oral tentacles (OT), Cuvierian organ (CO), respiratory tree (RT), Polian vesicle (PV), coelomocytes (Cc), ovary (Ov), testis (Ts), rete mirabile (RM), transverse vessel (TV) and intestine (In). Digestive tract regions include the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc). Embryonic and larval development stages include the fertilized egg (EF), 2-cells (2C), 4-cells (4C), 8-cells (8C), 16-cells (16C), morula (Mr), blastula (Bs), rotated-blastula (RB), early-gastrula (EG), late-gastrula (LG), early-auricularia (EA), mid-auricularia (MA), auricularia (Ar), doliolaria (Dl), pentactula (Pt), 1-mm juvenile (J1), 20-mm juvenile (J20) and adult (A). Blue and red colors represent relatively low and high expression levels, respectively. The gene expression level in each tissue or developmental stage is derived from the average of samples (n = 3 or 4) taken from different individuals. Source data are provided as a Source Data file. b FISH of in-iLyz (Hl-36988 and Hl-36992) mRNA in the esophagus (Es), foregut (Fg), midgut (Mg), hindgut (Hg) and rectum (Rc) of the digestive tract. The white scale bars are 400 μm. c Phylogenetic tree of lysozyme genes in typical vertebrate, ecdysozoan, lophotrochozoan, and echinoderm species (Supplementary Fig. 7). Different dot colors represent different taxonomic groups and different line colors represent different lysozyme types. d Phylogenetic analysis and heatmap comparison for tissue expression of the lysozyme genes in H. leucospilota and H. scabra (Supplementary Data 5). Selected tissues include body wall (BW), coelomocytes (Cc), rete mirabile (RM) and intestine (In). Blue and red colors represent relatively low and high expression levels, respectively, as scaled by the rows. The gene expression level at each stage is derived from the average of samples (n = 3) taken from different individuals. Source data are provided as a Source Data file. e Enzyme activity assay of lysozyme in the body wall (BW), coelomic fluid (CF), rete mirabile (RM), and intestine (In) of H. leucospilota. Data presented here are expressed as mean±SEM (n = 4 from 4 individuals). Source data are provided as a Source Data file. f The mRNA expression of cc-iLyz (Hl-18105) in the coelomocytes and in-iLyz (Hl-36988 and Hl-36992) in the intestine after V. harveyi injection for 0 (control), 24, 48 and 72 hours. Data presented here are expressed as mean±SEM (n = 10 from 10 individuals; P values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test, where * P < 0.05, ***P < 0.001). Source data are provided as a Source Data file. g The mRNA expression of cc-iLyz (Hl-18105) in the coelomocytes and in-iLyz (Hl-36988 and Hl-36992) in the intestine after starvation for 0 (control), 7, 14 and 30 days. Data presented here are expressed as mean±SEM (n = 10 from 10 individuals; P values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test, where *P < 0.05, ***P < 0.001, ****P < 0.0001). Source data are provided as a Source Data file. h The relative abundances of the resident bacteria and digested bacteria in the seabed sediments (SS, n = 5) and the feces (Fc, n = 5) and digestive tract (DT, n = 19) of H. leucospilota (Supplementary Data 2). Data presented here are expressed as mean±SEM (P values are calculated by one-way ANOVA followed by Tukey’s multiple comparisons test, where *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Source data are provided as a Source Data file. The corresponding lysoplate assay of the recombinant H. leucospilota in-iLyz (rHl-in-iLyz) protein or hen egg white lysozyme (HEWL) was performed against seedbed sediment bacteria P. marcusii (Rhodobacteraceae), O. marina (Verrucomicrobiaceae), B. cremea (Planctomycetaceae), B. aggregatus (Geodermatophilaceae), F. oceanosedimentum (Flavobacteriaceae), P. aurantiacus (Halomonadaceae), probiotic bacteria B. subtilis (Bacillaceae), pathogenic bacteria V. harveyi (Vibrionaceae) and standard bacteria M. lysodeik. The PBS group and the HEWL group were used as the negative and positive control, respectively. Each experiment was repeated four times.

The phylogenetic analysis revealed that echinoderms possess only i- (invertebrate) type lysozymes, while c- (chicken or conventional) and g- (goose) type lysozymes are absent (Fig. 5c; Supplementary Fig. 7). The echinoderm i-type lysozymes can be further divided into two branches, one of which is shared among all echinoderms, while the other is specific to the class Holothuroidea (Fig. 5c). Based on the expression analysis of lysozyme genes in the sea cucumbers H. leucospilota and Holothuria scabra, the lysozyme type shared among echinoderms was predominantly expressed in the coelomocytes, while the lysozyme specific to the class Holothuroidea was primarily expressed in the intestine (Fig. 5d). Therefore, these two types of lysozymes can be defined as the intestinal i-type lysozyme (in-iLyz) and the coelomocyte i-type lysozyme (cc-iLyz) in echinoderms. Analysis of the enzyme activity in different tissues of H. leucospilota further supported this classification, as lysozyme activities were observed in both the coelomic fluid and intestine (Fig. 5e).

Under challenge by the pathogenic Vibrio harveyi, transcription of the cc-iLyz (Hl-18105) gene in the coelomocytes increased, while the expression levels of in-iLyz (Hl-36988 and Hl-36992) in the intestine remained stable (Fig. 5f). These expression responses of sea cucumber cc-iLyz can also be mimicked by challenged of lipopolysaccharide (LPS) but not Polyinosinic:polycytidylic acid [poly(I:C)] (Supplementary Fig. 8). Conversely, during starvation, gene expression of in-iLyz in the intestine decreased, while that of the cc-iLyz in the coelomocytes remained unaffected (Fig. 5g). These results indicate that cc-iLyz is associated with immunity, while in-iLyz is correlated with nutritional status.

As resident bacteria, the families Rhodobacteraceae, Verrucomicrobiaceae and Planctomycetaceae are abundant in the digestive tract and feces of H. leucospilota but are less abundant in the environment (Fig. 5h). In contrast, as digested bacteria, the families Geodermatophilaceae, Flavobacteriaceae and Halomonadaceae are more abundant in the environment and less abundant in the digestive tract and feces (Fig. 5h). Based on the lysoplate assay, the bacteriolytic activities of the recombinant H. leucospilota intestinal i-type lysozyme (rHl-in-iLyz) protein (Supplementary Fig. 9) were weaker against Paracoccus marcusii, Oceaniferula marina and Blastopirellula cremea, which correspond to the resident bacteria in families Rhodobacteraceae, Verrucomicrobiaceae and Planctomycetaceae, respectively, and stronger against Blastococcus aggregatus, Flavobacterium oceanosedimentum and Pistricoccus aurantiacus, which correspond to the digested bacteria in families Geodermatophilaceae, Flavobacteriaceae and Halomonadaceae, respectively (Fig. 5h). Additionally, the bacteriolytic ability of rHl-in-iLyz was relatively strong against the standard bacterium Micrococcus lysodeikticus and the pathogenic V. harveyi, while weaker against the probiotic Bacillus subtilis. These results indicate that in-iLyz is involved in the digestion of bacteria ingested by sea cucumbers from the seabed deposits.

Discussion

Marine animals acquire nutrition through different feeding modes, including herbivory (e.g. sea urchin)²⁹, carnivory (e.g. starfish)³⁰, filter feeding (e.g. oyster)³¹, symbiotic nutrition (e.g. coral and giant clam)³² and deposit feeding. The process of deposit feeding involves ingesting large amounts of sediment to obtain organic matter. Deposit feeders are not common in the marine benthic community, except for polychaete annelids (e.g. Nereis succinea)³³, sipunculids (e.g. Sipunculus nudus) and echiuroids (e.g. Urechis unicinctus)³⁴, while sea cucumbers are among the most significant groups. Organic matters in the shallow seabed sediments mainly consist of the remains of dead animals and plants, as well as sediment-attached microalgae, bacteria and fungi³⁵. However, the organic matter content in the seabed substrate is relatively low, therefore, sea cucumbers need to ingest large quantities of sediment to obtain sufficient nutrients for survival, growth and reproduction¹.

The dietary compositions of sea cucumbers were analyzed using specimens of H. leucospilota collected from Daya Bay, a typical shallow habitat for this species. The 18S sequencing results revealed that algae constitute the primary eukaryotic food source, while animal-based food is rapidly decomposed in the anterior part of the digestive tract (Fig. 1b). On the other hand, the 16S sequencing data indicated that the primary prokaryotic food sources include Proteobacteria, Bacteroidetes and Cyanobacteria, which are of significantly reduced abundance in the feces of H. leucospilota (Fig. 4a). Similar food compositions can also be found in other sea cucumbers (e.g. A. japonicus, Stichopus monotuberculatus, Stichopus chloronotus and Holothuria atra)^36,37, as well as in other marine deposit feeders (e.g. S. nudus and U. unicinctus)³⁴, which primarily consume microalgae. Corresponding to their food sources, expansion and contraction of digestive carbohydrase and protease genes were observed in the sea cucumber genomes, respectively (Fig. 1d). The importance of digestive carbohydrase in food digestion of sea cucumbers, such as amylase³⁸ and trehalase³⁹, has been described previously. Highly mobile vertebrates have more protease genes and fewer carbohydrase genes, while benthic marine invertebrates have fewer protease genes and more carbohydrase genes (Fig. 1d). On the other hand, mammals with different feeding habits (e.g. humans, sheep and cats) share similar amounts of digestive enzyme genes (Fig. 1d), suggesting that the extensive expansion/contraction of digestive enzyme genes is a long-term evolutionary event.

In animals, the digestive tract normally exhibits functional compartmentalization, resulting in distinct digestive capacities among regions²⁷. In this study, the anatomical feature of the H. leucospilota digestive tract showed a classification of five regions, namely, the esophagus, foregut, midgut, hindgut and rectum (Fig. 3a, b). The foregut exhibited the highest digestive capacity, which was evidenced by the highest activities of amylase, protease and lipase (Fig. 3e, f). Additionally, the expression of carbohydrase, protease and lipase predominantly took place in the foregut region, while the hindgut region showed lower levels of protease and lipase expression (Fig. 3c, f). However, relatively high enzyme activities of protease and lipase were also observed in the hindgut (Fig. 3e, f), indicating a potential contribution of gut microbes to the digestion process. Furthermore, despite the absence of cellulase genes in the H. leucospilota genome (Fig. 1d), cellulase activity was detected in the digestive tract (Fig. 3c), suggesting that the activity might also be facilitated by gut microbiota, evidenced by the presence of cellulose-decomposing bacteria from the genus Clostridium, Ruminococcus, Cellvibrio, Bacillus and Pseudomonas in the digestive tract (Supplementary Data 7).

Symbiotic relationships between hosts and gut microbiotas are commonly documented⁴⁰. The H. leucospilota digestive tract harbors a diverse bacterial community, primarily composed of Proteobacteria, Planctomycetes, Firmicutes, and Verrucomicrobia (Fig. 4b), which remained stable throughout the individual development⁴¹. The gut microbial composition of the sea cucumber exhibits greater similarity to that of the sediment rather than the seawater⁴². Proteobacteria, Bacteroidetes, and Cyanobacteria, which are abundant in the sediment, experienced a significant reduction in abundance upon traversing the digestive tract (Figs. 1c, 4a), implying their potential role as a food source that is subject to digestion. The high abundance of Proteobacteria found in the digestive tract (Fig. 4a) indicates that it is also a type of resident bacteria. Planctomycetes, a bacterial group known for its ability to convert nitrate nitrogen into ammonia nitrogen in anoxic environments⁴³, displayed a remarkably high abundance in the digestive tract of H. leucospilota (Fig. 4a), but not in other echinoderms⁴⁴. This feature of Planctomycetes might enable the ecological role of H. leucospilota in mitigating seawater acidification¹. Firmicutes are primarily associated with the metabolism of intestinal substances⁴⁵ by encoding enzymes such as 2-oxoacid ferredoxin oxidoreductase (OFOR), triosephosphate isomerase (TPI), and secretory phospholipase A2 (sPLA2), which involved in the decomposition of carbohydrates and lipids (Fig. 4c). In addition, when antibiotics were applied to disrupt the homeostasis of gut microorganisms, a significant decrease in digestive protease and lipase activities was observed (Fig. 4d), indicating that the gut microbiota contribute to the sea cucumber’s food digestion by providing certain enzyme activities.

To control the proliferation of pathogenic bacteria entering the digestive tract, sea cucumbers need to counteract their pathogenicity. It was found that among all immune genes, the intestinal lysozyme genes displayed the most significant difference at the transcript level (Fig. 5a). Echinoderms possess only i-type lysozymes, which divide into two branches, one of which is shared among all echinoderms and specifically expressed in the coelomocytes, named as cc-iLyz, while the other branch is unique to the class Holothuroidea and specifically expressed in the intestine, named as in-iLyz (Fig. 5c). The Holothuroidea-specific in-iLyz genes (Hl-36988 and Hl-36992) are expressed in the brush border of the intestine (Fig. 5b), similar in location to those digestive enzymes (Fig. 3d). Upon pathogenic challenge with vibrio, the transcript of cc-iLyz was up-regulated (Fig. 5f), while the expression of in-iLyz was drastically decreased under food deprivation (Fig. 5g), indicating a link between the sea cucumbers’ in-iLyz but not cc-iLyz expression and their nutrition status. H. leucospilota in-iLyz strongly lysed the seabed sediment bacteria B. aggregatus, F. oceanosedimentum and P. aurantiacus, but showed lesser bacteriolytic activities on P. marcusii, O. marina and B. cremea (Fig. 5h). As an enzyme that kills bacteria by lysing their cell wall peptidoglycan, lysozymes have been found in the digestive systems of various animals, such as, earthworms (Eisenia andrei)⁴⁶, oysters (Crassostrea virginica)⁴⁷ and the larvae of insects Musca domestica⁴⁸, Drosophila melanogaster⁴⁹ and Lutzomyia longipalpis⁵⁰. Although these animals use different food sources, a common characteristic of their diets is rich in bacteria. As a result, the lysozymes in their digestive systems may have undergone convergent evolution to control and digest ingested bacteria. This may explain why there have been no reports of intestinal lysozymes in other benthic animals, such as the round goby Neogobius melanostomus⁵¹, which feed by filtering sediment through their mouthparts to collect food particles, rather than directly swallowing the sediment.

During the whole developmental process, sea cucumbers undergo a dietary shift from endogenous nutrition to planktotrophic feeding, and finally to deposit feeding after settling to the seabed (Fig. 2a). Correspondingly, three sets of digestive enzyme systems are expressed sequentially (Fig. 2b, c). During the endogenous nutrition period, H. leucospilota primarily relies on the lipase LIPG and the chitinase homolog CHIP to hydrolyze maternal nutrients from the yolk. During the planktotrophic feeding period, H. leucospilota primarily relies on carbohydrases SUIS, LPH, and MGA to digest and obtain nutrition from planktonic algae. After attachment, H. leucospilota digests seabed sediments mainly through the carbohydrases AMY, MGA, SUIS and LPH, the protease CTR, and the lipase LIPP. Simultaneously, in-iLyz begins to express during this period (Fig. 5a), and it works together with the digestive enzymes to digest the microalgae and bacteria rich in the seabed sediments. It’s noteworthy that although chitinase-like genes are observed within the genome of H. leucospilota (Fig. 1d), they exhibit predominant expression during embryonic development rather than in the adult intestine (Figs. 2c, 3c). Consequently, no chitinase activity was identified within the digestive tract (Fig. 3e).

Although sea cucumbers inhabit a broad geographic expanse globally and their living environments vary with the climate, most are shallow-water deposit feeders. The repertoire of digestive enzyme genes in their genomes remains evolutionarily conserved, whereas their diets and gut microbial communities may exhibit variation due to environmental differences. This study reveals the evolutionary innovations of sea cucumbers for adapting to the deposit-feeding style, including the expansion of the digestive carbohydrase gene repertoire, contraction of digestive protease gene repertoire, and the occurrence of intestinal i-type lysozyme genes in their genomes. The changes in the expression patterns of digestive enzymes accompanying the dietary shifts during the sea cucumber development are also elucidated. This study further demonstrates relationships between the gut microbiota and the digestion process in sea cucumbers, namely, some resident bacteria provide digestive enzymes, while others are digested by intestinal lysozymes to serve as a food source. Our study may provide insights into the mechanisms by which sea cucumbers effectively utilize nutrients from seabed sediments in marine ecosystems.

Methods

Compliance statement

All experiments on sea cucumbers were conducted in accordance with ethical regulations and research guidelines set up by the research ethics committee of the South China Sea Institute of Oceanology, Chinese Academy of Sciences.

Animals and reagents

Wild H. leucospilota specimens (~200 g) were collected from Daya Bay (114°53′E, 22°55′N), China. Following ice anesthesia, sea cucumbers were carefully dissected to obtain the different digestive tract regions (n = 4) and other tissue (n = 3) samples for RNA sequencing. Embryonic, larval and juvenile samples (n = 3) were acquired using a 200-mesh filter during each developmental stage after artificial spawning⁵². For barcode sequencing, tissue samples were collected from the different digestive tract regions (n = 3−5). Water samples (n = 5) were gathered and processed following the protocol from a previous study⁵³. Feces and seabed surface sediments (0−2 cm depth, n = 5) were collected individually from the surroundings (<20 cm) of the sea cucumbers, employing a 50 mL benthic sampler. In addition, information regarding the kits, reagents and bacterial strains used in this study is listed in Supplementary Data 10.

Cross-genomic analysis

For identification of digestion-related enzyme genes through cross-genomic analysis, 23 representative species of deuterostomes were selected. Among these organisms, Homo sapiens, Ovis aries, Felis catus, Gallus gallus, Xenopus tropicalis and Lepisosteus oculatus represent vertebrates; Ciona intestinalis represents urochordates; Branchiostoma lanceolatum represents cephalochordates; Saccoglossus kowalevskii and Ptychodera flava represent hemichordates; Anneissia japonica, Acanthaster planci, Asterias rubens, Patiria miniata, Strongylocentrotus purpuratus, Lytechinus variegatus and Lytechinus pictus represent other echinoderms; while Chiridota heheva, A. japonicus, S. monotuberculatus, H. glaberrima, H. scabra and H. leucospilota represent class Holothuroidea in echinoderms (Supplementary Data 11). Genes from 15 gene families, including AMY, MGA, SUIS, LPH, TREA, PEP, TRY, CTR, CLE, LIPP, LIPG, LIPC, CHIA, CHIP and LYZ, were identified with reference to the annotations by the SwissProt database. The corresponding phylogenetic relationships were derived and visualized using MEGA 6.0.

RNA sequencing and gene expression analyses

For RNA sequencing, the selected tissues comprised the body wall, muscle, oral tentacles, Cuvierian organ, respiratory tree, Polian vesicles, coelomocytes, ovaries, testes, rete mirabile, transverse vessel and intestine¹⁵. The coelomocytes were harvested from coelomic fluids that were filtered by 100-μm sterile nylon mesh and centrifuged immediately at 4 °C and 1000×g for 10 min. In the parallel experiment, corresponding tissues from H. scabra were also obtained. The digestive tract was further classified into the esophagus, foregut, midgut, hindgut and rectum. Developmental stages under study included fertilized eggs, 2-cells, 4-cells, 8-cells, 16-cells, morula, blastula, rotated-blastula, early-gastrula, late-gastrula, early-auricularia, mid-auricularia, auricularia, doliolaria, pentactula, 1-mm juvenile and 20-mm juvenile, as previously described⁵². The RNA libraries were sequenced using HiSeq X Ten and NovaSeq 6000 platforms (Illumina), where 150 bp paired-end reads were generated (Supplementary Data 12). Clean reads were obtained by using SOAPnuke v1.5.6. Paired-end clean reads were aligned to the reference genome with HISAT2 v2.1.0. Transcripts were assembled and read counts of each gene were calculated by using StringTie v1.3.5. Counts per million mapped reads (CPM) were calculated, and cross-sample normalization was performed using DESeq2. The heatmaps were generated using the TBtools tool.

Histology and fluorescence in situ hybridization

The histological features of different digestive tract regions in H. leucospilota were observed on 4 μm-transverse sections that were cut from paraffin-embedded samples and stained with hematoxylin and eosin (H/E) for visualization. For FISH, digoxin (DIG)-labeled antisense cRNA probes (Supplementary Data 13) generated by DIG RNA labeling mixture (Roche) and diluted 1:100 in PBS were used. The sections were subjected to overnight hybridization at 42 °C, followed by incubation with biotin-conjugated AffiniPure mouse anti-DIG IgG (Boster Bio) diluted 1:50 in PBS for 1 h. To amplify the signal, HRP-conjugated streptavidin (Invitrogen) diluted 1:100 in PBS was added dropwise and incubated for 15 min. Subsequently, the Alexa Fluor™ 555 tyramide SuperBoost™ Kit (Invitrogen) was employed for signal amplification and DAPI reagent (Roche) was used for counter-staining of cell nuclei. The FISH sections were then viewed and imaged by an LSM800 Confocal Laser Scanning Microcopy (Zeiss).

Whole-mount in situ hybridization for embryos and larvae

Embryos and larvae were fixed in 4% paraformaldehyde in high-salt MOPS fixing buffer at 4 °C overnight, then washed with 25%, 75%, and 100% ice-cold ethanol, respectively. WMISH was conducted with modifications as previously described⁵⁴. DIG-labeled antisense cRNA probes were hybridized with the samples at a final concentration of 0.2 ng/mL at 55 °C for three days. After hybridization, the samples were blocked for 30 min using a Blocking reagent (Roche), followed by incubation with a 1:2000 dilution of alkaline phosphatase (AP)-conjugated anti-Dig antibody (Roche) at 4 °C for 12 h. The samples were washed in MABT and PBST, and signals were detected using NBT/BCIP reagent (Roche). Imaging was performed using an EX31 microscope (SHUNNY) with Light Tools software (ORA).

Enzyme activity analysis

The tissue samples, which were frozen by liquid nitrogen, were thawed, weighed, homogenized and then centrifuged at 4 °C and 3000×g for 10 min. The coelomic fluids samples were directly centrifuged at 4 °C and 1000×g for 10 min. The resulting supernatant was used to measure digestive enzyme activity. The total protein content in the supernatant was determined using the BCA method. The enzyme activities of amylase, protease, lipase, chitinase, cellulase and lysozyme in the extracts were determined using specific detection kits (Nanjing Jiancheng Bioengineering Institute).

18S and 16S sequencing

DNA was extracted from the contents of the esophagus, foregut, midgut, hindgut, and rectum of sea cucumbers, as well as from their feces, and the surrounding seawater and seabed sediments, using the Mag-Bind® Soil DNA Kit (Omega Bio-Tek). The V4 region of the eukaryotic 18S rRNA gene was amplified using primers 547 F and V4R (Supplementary Data 14), while the V4 region of the bacterial and archaeal 16S rRNA gene was amplified using primers 515 F and 806 R (Supplementary Data 14). The amplified products were used for library construction using the TruSeq Nano DNA LT Library Prep Kit (Illumina) and sequenced by NovaSeq 6000 System (Illumina). The obtained sequences were further subjected to quality filtering, denoising, merging, and chimera removal using the DADA2 plugin⁵⁵. The NCBI-nt database and Greengenes database were used for taxonomy assignment of 18S and 16S rRNA genes, respectively. Species annotation for each ASV was performed in QIIME2 using the classify-sklearn algorithm with a pre-trained Naive Bayes classifier and default parameters⁵⁶. An ASV abundance table was generated, and relative abundances were obtained using rarefaction. Various α-diversity indices were derived using QIIME2’s diversity alpha function.

Metatranscriptomic sequencing

For metatranscriptomic analysis, the digestive tract was dissected into five regions including the esophagus, foregut, midgut, hindgut and rectum. The contents were routinely removed, and the samples were rinsed five times using PBS. Total RNA extraction was extracted from the contents of different digestive tract regions, and the double-strand cDNA was synthesized using purified mRNA as a template. The double-strand cDNA was randomly fragmented by the Whole Genome Shotgun (WGS) strategy, and libraries of inserts of appropriate length were constructed. Sequencing was performed on these libraries in Paired-end mode using 2 × 150 bp reads. Clean reads were obtained by removing low-quality sequences and reads with kmer depth less than 2 using BBCMS. Assembly was conducted using MEGAHIT, and the sequences matched to the host sequences were discarded to minimize host contamination.

Species annotation was executed using Kraken2 against the NCBI-nt and GTDB databases⁵⁷. The lineage information of the selected bacterial species was obtained through TaxonKit. The bacteria kingdom was chosen, and only species with clear definitions were selected. The lineage information was used to generate a Newick formatted tree through Taxonomizr and the tree was visualized using iTOL. The sense RNA sequences were identified using TransGeneScan. Protein sequences were aligned against the NCBI-nr and Swiss-Prot database, and those annotated as Metazoa and Viridiplantae were removed from the annotation results. Salmon’s quant command was used to map high-quality sequences to the predicted gene sequences to obtain transcripts per kilobase million (TPM) values for each sample⁵⁸.

Antibiotic expose experiments

A total of 15 healthy adult H. leucospilota were randomly divided into three groups (n = 5) and exposed to antibiotics at final concentrations of ampicillin (1 g/L, Macklin), vancomycin (500 mg/L, Macklin), metronidazole (1 g/L, Macklin) and neomycin (1 g/L, Macklin) in artificial seawater in 30-L tanks. The sea cucumbers were harvested at 0- (control), 3- and 7-day after antibiotic exposure. The intestine samples were collected for subsequent analysis of digestive enzyme activities.

Pathogenic challenge experiment

A total of 120 healthy adult H. leucospilota were randomly divided into three groups and injected with 100 μL of V. harveyi (E385, approximately 5×10⁷ cells), LPS (2 μg/μL, Sigma) or Poly (I:C) (2 μg/μL, Sigma) diluted in PBS. The sea cucumbers were harvested at 0- (control), 24-, 48- and 72-h after injection (n = 10). After the experiment, intestine and coelomocyte samples were collected for the detection of the transcript expression of in-iLyz (Hl-36988 and Hl-36992) and cc-iLyz (Hl-18105), respectively.

Starvation experiment

The adult H. leucospilota were raised in cement pools for 9 months, with regular feeding of artificial compound algae powder⁵². The pool bottoms were covered with sand that were full of microalgae and bacteria, while the sea cucumbers exhibited normal feeding and defecation. For starvation, a total of 40 healthy sea cucumbers were transferred to cement pools covered with clean sand without microalgae and bacteria, and feeding was stopped accordingly³⁹. The individuals were randomly assigned into four groups (n = 10), namely, 0- (control), 7-, 14-, and 30-days of starvation. After the experiment, intestine, and coelomocyte samples were collected for the detection of the transcript expression of in-iLyz (Hl-36988 and Hl-36992) and cc-iLyz (Hl-18105), respectively.

Realtime PCR

The mRNA levels of lysozyme genes were determined by quantitative PCR (qPCR). Total RNA was extracted with TRIzol reagent (Invitrogen) and digested with gDNA Eraser (Takara) and reverse transcription was conducted with the PrimeScript™ RT reagent Kit (Takara). qPCR reactions were performed by using SYBR Premix Ex Taq™ II (Takara) and EF-1α (Hl-40064) was used as an internal control to verify qPCR results (Supplementary Data 15). The target gene/ EF-1α mRNA ratio was calculated using the formula 2^˗ΔΔCt, and the raw data were simply transformed into the percentage of the mean values over the internal control for statistical analysis purposes.

Antimicrobial assay

A cDNA fragment for a H. leucospilota in-iLyz (Hl-36992) mature peptide was cloned into the PET28a vector (Novagen). The resulting construct was then transformed into Rosetta™(DE3) Competent Cells (Novagen). The recombinant H. leucospilota in-iLyz (rHl-in-iLyz) protein was expressed by IPTG induction and purified using His-Bind Kits (Novagen). Finally, desalination was carried out using PD-10 Desalting Columns (GE Healthcare).

The antimicrobial activity of the rHl-in-iLyz protein was evaluated using the lysoplate assay with night bacterial strains. The seabed sediment bacteria P. marcusii, O. marina, B. cremea, B. aggregatus, F. oceanosedimentum and P. aurantiacus were provided by the Marine Culture Collection of China (MCCC), and B. subtilis, V. harveyi and M. lysodeik were used as the probiotic, pathogenic and standard bacteria, respectively. Nutrient agar medium or 2216E medium was used to prepare 90 mm gel plates containing bacteria at a concentration of 10⁷ CFU/mL. After the plates solidified., 12 wells in diameter of 0.4 cm divided into three groups were punched on the culture. Next, 30 μL drops of rHl-in-iLyz protein (0.67 μg/μL) were added into these wells, and hen egg white lysozyme (HEWL, 0.8 μg/μL, Sangon), as well as PBS, were used as positive and negative controls, respectively. The plates were then incubated for 24 hours at 28 °C and the diameter of the transparent zone surrounding each well was measured to determine antimicrobial activity.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 7.0 (Graph-Pad Software). All data are presented as the mean ± standard error of the mean (SEM). Statistical differences were estimated via one-way ANOVA followed by Tukey’s multiple comparisons test.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

C.R., C.H. and T.C. conceived the study, C.R., X.J., S.G., C.H. and T.C. designed the scientific objectives and coordinated the project. W.P., Xuan Wang, C.R., N.W, C.H. and T.C. led the manuscript preparation and writing. W.P., Z.X., P.L. and H.L. collected and cultured the sea cucumbers. J.H., Y.Y., Xiaofen Wu and D.H. collected and handled adult samples. W.P., Xiaomin Li, B.M. and Y.L. collected and handled embryonic and larval samples. W.P., X.J., Z.X., J.H. and Z.E. and Jianlong Li collected and handled environmental samples. W.P., Xuan Wang, C.R., Xiaomin Li, J.H., Y.Y., Xiaofen Wu and T.C. performed transcriptomic sequencing and analysis. W.P., X.J., S.G., D.F. and T.C. performed barcode sequencing and analysis. W.P., S.G. and T.C. performed metatranscriptomic sequencing and analysis. W.P., Xiaofen Wu, D.H., B.M., Y.L, Xujia Liu, C.C. and T.C. performed in vivo experiments. W.P., Xuan Wang, J.H., X.Z. performed histological analysis. W.P., Xuan Wang, and Jingxuan Liang performed enzyme activity analysis. Xuan Wang, X.R., S.Y. performed lysozyme experiments. Xuan Wang, D.F., X.R., Z.Q. and T.C. performed the bioinformatics analysis. C.R., X.J., Z.X., P.L., H.S., L.Y., Y.W., C.H. and T.C. contributed reagents/analytic tools. W.P., Xuan Wang, C.R., X.J., N.W, C.H. and T.C. participated in the final data analysis and presentation. All authors have read and approved the submitted version of the manuscript.

Peer review

Peer review information

Nature Communications thanks Pedro Martinez and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The 18S sequencing, 16S sequencing, and metatranscriptome data have been deposited in the GenBank under accession codes PRJNA1071841. The RNA-seq data for larvae development have been deposited in the GenBank under accession code PRJNA1071214. The RNA-seq data for tissues have been deposited in the GenBank under accession code PRJNA747844 and PRJNA1074116. The enzyme activity assays data are provided in the Source Data file Fig. 3e, Fig. 4d and Fig. 5e. The qPCR data for lysozyme mRNA expression are provided in the Source Data file Fig. 5f, Fig. 5g, S Fig. 8a and S Fig. 8b. The accession information for RNA, 18S and 16S sequencing data are also provided in Supplementary Data 12. Expression of all genes in different tissues is also provided in Supplementary Data 16. Expression of all genes in different developmental stages is also provided in Supplementary Data 17. All Source data are provided as a Source Data file. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-53205-5.