Digestive tract morphology and metallocarboxypeptidase M14 gene family divergence associated with trophic niche differentiation in Odontodactylus japonicus, Lysiosquilla sulcirostris and Oratosquilla oratoria

Abstract

Background

Stomatopods may have evolved divergent trophic niches during adaptive radiation, yet the morphological and molecular bases underlying this diversification remain poorly understood. Here, we integrated multi-level microscopic observations with full-length (FL) transcriptome data to compare digestive-tract morphology and metallocarboxypeptidase M14 (MCP M14) characteristics among Odontodactylus japonicus, Lysiosquilla sulcirostris and Oratosquilla oratoria. The FL transcriptome of L. sulcirostris was newly generated in this study, whereas those of O. japonicus and O. oratoria were obtained from previously published datasets.

Results

Pronounced interspecific differences were observed in digestive morphology. O. oratoria exhibited the highest density of cardiac-stomach setae and the narrowest gland-filter pores, whereas L. sulcirostris had the thickest cardiac-stomach wall and chitin layer and the most developed pyloric muscular layer. In contrast, O. japonicus showed the greatest hindgut villus height and hepatopancreas index. Using the newly generated FL transcriptome of L. sulcirostris and previously published datasets for O. japonicus and O. oratoria, we identified 45 MCP M14 sequences in O. japonicus, 26 in L. sulcirostris and 15 in O. oratoria. MCP M14 sequences from O. japonicus and L. sulcirostris were assigned to the M14A, M14B and M14D subfamilies, whereas only M14A sequences were detected in O. oratoria. Conserved-motif and sequence analyses indicated overall structural conservation but substantial variation in subfamily composition and canonical catalytic residues. Integrative analysis further revealed distinct morphological–molecular profiles, with O. japonicus showing greater hindgut villus development and a broader detected MCP M14 repertoire, L. sulcirostris displaying greater specialization in mechanical food processing, and O. oratoria exhibiting more pronounced filtration-related traits.

Conclusion

These findings indicate that digestive adaptation among the three stomatopods is associated with distinct combinations of digestive morphology and MCP M14 characteristics rather than uniform changes across the digestive system. Such contrasting morphological–molecular configurations provide an integrated basis for understanding trophic niche differentiation in stomatopods.

Background

The trophic niche reflects the realized and potential capacity of animals to exploit food resources. It provides a central framework for resolving patterns of resource use and the flow of matter and energy between animals and their habitats, and is therefore of broad relevance to both community ecology and adaptive evolution [59]. Previous studies have shown that animals evolve optimal foraging strategies in response to divergent energetic demands, preferentially selecting food resources that are easy and low-risk to capture, energetically profitable and efficiently digested [7, 16, 20, 33]. As animals adapt to habitat-specific food resources, adaptive radiation can shape specialized modes of food acquisition and utilization. These processes, in turn, can drive the coordinated evolution of functional traits and trophic niches through natural selection and niche construction mechanisms [42], [52]

The digestive tract is the primary organ system responsible for mechanical processing, chemical digestion and nutrient absorption [13]. Its morphology often shows pronounced diet-associated divergence, making it one of the most direct phenotypic indicators of trophic niche differentiation [53]. Indeed, length, diameter, muscle type and extensibility, goblet-cell abundance and fold number of the digestive-tract can directly affect the rate of food passage, retention time, and the extent of mechanical damage to the digestive epithelium, and ultimately determined the efficiency of nutrient absorption [13]. In addition, enzyme-mediated chemical digestion is a critical stage in nutrient utilization [54]. Protein digestion in crustaceans involves several major protease systems, including trypsin, chymotrypsin, cathepsins, aminopeptidases and carboxypeptidases, which act at different stages of protein degradation [11, 46]. In the present study, we focused on the metallocarboxypeptidase M14 (MCP M14) family not because it is the only major digestive protease family, but because it represents a diverse group of exopeptidases that participate in the terminal stages of protein digestion by removing C-terminal amino acid residues and complementing the activity of endopeptidases [25, 41]. In addition, MCP M14 proteins comprise several functionally differentiated subfamilies with variation in substrate preference, catalytic residues and structural features, making this family suitable for comparative analysis of molecular divergence among species [14, 34]. Moreover, MCP M14 sequences in O. oratoria have been characterized previously [55], providing a direct comparative basis for extending the analysis to O. japonicus and L. sulcirostris. Therefore, MCP M14 was selected here as a representative digestive-enzyme family to explore molecular differences among the three stomatopod species, rather than to imply that other digestive proteases are less important.

Stomatopods, belonging to Arthropoda, Crustacea, Malacostraca and Hoplocarida, exhibit body-segmentation patterns and life histories that differ markedly from those of closely related crustacean lineages [1]. More than 450 extant stomatopod species have been described, belonging to the Erythrosquilloidea, Bathysquilloidea, Gonodactyloidea, Lysiosquilloidea, Squilloidea, Eurysquilloidea and Parasquilloidea [38]. A defining innovation of stomatopods is the highly specialized raptorial appendage, which possesses a distinctive mechanical system composed of a saddle-like elastic structure and two rotational axes. During predation, deformation of the saddle stores elastic energy [18, 21], which is then released within milliseconds to generate exceptionally high instantaneous impact forces [31, 32]. The rapid strike also produces cavitation bubbles in the surrounding seawater,their collective collapse imposes secondary damage on prey [23, 39]. By virtue of this remarkable biomechanical innovation, stomatopods occupy high trophic positions in nearshore benthic ecosystems [2]. On the basis of terminal raptorial-appendage morphology, stomatopods are generally classified into smashers and spearers and these groups exhibit clear trophic niche partitioning. Specifically, smasher species possess robust, impact-resistant appendages and mainly prey on hard-shelled organisms such as bivalves and brachyurans, whereas spearer species have sharp, piercing appendages and primarily capture soft-bodied prey, including teleost fishes and shell-less molluscs. Importantly, raptorial-appendage morphology also varies among species within the same functional group, suggesting that stomatopods may have undergone interspecific differentiation in predatory strategy and diet, with corresponding diversification of trophic niches [56]. However, the mechanisms underlying this diversification remain poorly understood.

In this study, we selected three stomatopod species, including the hard-prey smasher Odontodactylus japonicus and two soft-prey spearers, Lysiosquilla sulcirostris and Oratosquilla oratoria. By integrating conventional morphological measurements, microstructural observations and full-length (FL) transcriptome data, we compared digestive-tract phenotypes and MCP M14 characteristics among the three species. We hypothesized that the smasher would exhibit greater development of gastric structures associated with mechanical food processing and a broader MCP M14 subfamily repertoire, whereas the two spearers would show relatively less-developed gastric mechanical structures and a simpler MCP M14 subfamily composition. This comparative framework was used to test whether differences in prey-processing strategy are reflected jointly in digestive morphology and MCP M14 composition.

Materials and methods

Sample collection

Specimens of O. japonicus, L. sulcirostris and O. oratoria were collected in November 2024 from coastal waters of the Taiwan Strait (approximately 25°30′ N, 119°58′ E). The collection depth was approximately 26 m, and the in situ seawater temperature and salinity were 26 °C and 35, respectively. Thirty individuals of each species were measured for basic biometric characteristics. In L. sulcirostris, carapace length, total body length and wet weight were 46.61 ± 2.42 mm, 229.73 ± 11.38 mm and 150.72 ± 15.75 g, respectively. The corresponding values were 30.55 ± 2.02 mm, 123.02 ± 7.72 mm and 40.56 ± 7.57 g for O. japonicus, and 27.30 ± 2.43 mm, 111.92 ± 8.58 mm and 21.96 ± 5.42 g for O. oratoria. Values are presented as mean ± SD (n = 30 per species). All animals were rapidly transported in incubator filled with seawater to the Laboratory of Fishery Resources and Ecological Environment (School of Ocean, Yantai University). Individuals were maintained under laboratory conditions (Temperature: 24 ℃, salinity: 28, pH: 8.1) for 7 days to minimize transport-induced stress. After acclimation, vigorous, uninjured individuals were selected and subjected to gradient cooling treatment (a 4 °C seawater bath for 20 min followed by exposure to an ice bed for 10 min) to induce anesthesia before histological examination of the digestive system and FL transcriptome sampling.

Histological observation of the digestive system

The digestive systems of O. japonicus, L. sulcirostris and O. oratoria were dissected to obtain the cardiac stomach, pyloric stomach, hindgut and hepatopancreas. Each tissue was cut into approximately 5-mm pieces and immediately fixed in Bouin’s fixative for 24–48 h. Fixed samples were rinsed with physiological saline and dehydrated through a graded ethanol series (70%, 80%, 90%, 95% and absolute ethanol) using an automated tissue processor (Leica Pearl). The dehydrated tissues were cleared in xylene and then embedded using a paraffin embedding system (Leica HistoCore). Serial sections of 3–5 μm thickness were prepared using a rotary microtome (Leica HistoCore), and the sections were then placed on glass slides and dried. Sections were stained with haematoxylin and eosin, sealed with neutral balsam and examined under a trinocular biological microscope (OLYMPUS CX33). Images were captured using Cap Studio software and processed and analysed using ImageJ. For quantitative morphometric analysis, three individuals were examined for each species. For each individual, three non-consecutive sections were selected from each digestive organ, and three non-overlapping microscopic fields were examined per section. Setal cluster density was quantified within a standardized area of 2,500 μm². Cardiac-stomach wall thickness, chitin-layer thickness, pyloric muscle thickness, hindgut villus height and villus width were measured using ImageJ. Measurements from different fields and sections were averaged for each individual, and the individual means were used for subsequent statistical analyses.

Scanning electron microscopy (SEM) was used to observe the ultrastructure of the cardiac stomach, pyloric stomach, hindgut and hepatopancreas. Fresh samples from each digestive organ were cut into blocks of < 3 mm³. After the tissues were washed with PBS, they were fixed in 2.5% glutaraldehyde at room temperature in a light-protected environment for 2 h. After being washed three times with phosphate buffered saline (0.1 mol/L, pH 7.4), the tissue was successively dehydrated in 30%, 50%, 70%, 80%, 90%, 95%, and 100% ethanol, and then dried using a critical point dryer (Quorum K850), and finally coated with a platinum film on an ion sputter coater (HITACHI MC1000). Finally, the processed samples were placed under a scanning electron microscope (HITACHI SU8100) with a working voltage of 3 kV to capture images.

Although the mid-gut is also an important part of the digestive system, its microstructure has already been reported in our previous study, so it will not be elaborated on here.

FL transcriptome sequencing

FL transcriptome data of O. oratoria and O. japonicus were obtained from previously published datasets generated by our group [28]. Therefore, FL transcriptome sequencing in the present study was performed only for L. sulcirostris. Given the limited genomic resources currently available for stomatopods, the FL transcriptome of L. sulcirostris was constructed as a multi-tissue transcriptomic resource to increase transcript diversity and obtain broader coverage of its expressed gene repertoire. Therefore, muscle, hepatopancreas and gonad tissues were collected, snap-frozen in liquid nitrogen and pooled in equal amounts for library construction. A single pooled FL transcriptome library was constructed, without independent biological replicate libraries. Accordingly, the FL transcriptome was used for transcript discovery and sequence characterization rather than for quantitative differential expression analysis. Subsequently, total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. RNA concentration, purity and integrity were assessed using the Qubit fluorometer (Thermo Fisher Scientific, Waltham, MA, USA), NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific) and Qsep analyser (BIOptic, Taipei, Taiwan, China), respectively. RNA samples with a concentration > 50 ng/μl, an A260/A280 ratio > 2.0 and an RNA integrity number > 8 were considered suitable for library construction. Then, mRNA was enriched using VAHTS mRNA Capture Beads (Vazyme, Nanjing, China). FL cDNA was synthesized and amplified using the Clontech SMARTer PCR cDNA Synthesis Kit (Takara Bio, Kusatsu, Shiga, Japan) and a fraction of cDNA fragments shorter than 1 kb were removed. Further end repair and addition of SMRT adapters were carried out for the FL cDNA to complete the construction of the cDNA library. Finally, the constructed library was subjected to real-time FL transcriptome sequencing on the PacBio RSII (Pacific Biosciences, Menlo Park, CA, USA) sequencing platform.

Processing of FL transcriptome data

Raw polymerase reads generated by the sequencing platform were first filtered using SMRTLink (v.10.1; [17] software to remove low-quality and invalid reads, thereby generating circular consensus sequences (CCSs). Adapter sequences, subreads and polymerase reads shorter than 50 bp, and polymerase sequences with an accuracy below 0.75 were discarded. These CCSs containing both 5’ and 3’ cDNA primers, poly(A) tail and a correctly positioned insert were defined as full-length non-chimeric reads (FLNCs). Meanwhile, the Lima and Refine modules embedded in Iso-Seq3 (v.3.4.0, [19] were used to remove 5′ and 3′ cDNA primers and poly(A) tails from the FLNCs. Subsequently, redundant FLNCs were filtered using Iso-Seq3 (v.3.4.0, [19] to generate consensus FL transcripts. These consensus FL transcripts with an accuracy > 99% and support from at least two FL reads were classified as high-quality sequences. To ensure consistent clustering of transcripts derived from homologous genes, high-quality consensus FL transcripts were further clustered using CD-HIT (v.4.7, [27] software. The resulting non-redundant FL transcripts were used for subsequent analyses.

Identification and characterization of the MCP M14 gene family

The MCP M14 characteristics of O. japonicus and L. sulcirostris were analyzed in the present study, whereas the 15 MCP M14 sequences of O. oratoria had been reported previously [55]. In the present study, these published O. oratoria sequences were used only as a reference dataset for comparative analyses, and no previously published figures were reused. The hidden markov model (HMM) profile of the MCP M14 domain was downloaded from the Pfam database (http://pfam.sanger.ac.uk/) with accession PF00246. Based on the HMM model, we used the HMMER (v.3.0, [51]) software to search for members of the MCP M14 gene family from the FL transcripts of O. japonicus and L. sulcirostris, and the parameter was E-value < 1.2e-28. Subsequently, the candidate sequences were verified against the NCBI non-redundant protein database using BLASTP, with an E-value threshold of 1 × 10⁻⁵. Meanwhile, we used the SMART online server (https://smart.embl.de/) to predict the conserved domains of these sequences. Because MCP M14 sequences were identified from FL transcriptomes rather than genome assemblies, the numbers reported here represent the transcript-derived MCP M14 repertoire detected under the present sampling conditions and should not be interpreted as absolute genomic gene copy numbers.

The physicochemical properties of the MCP M14 family were analysed using the ExPASy ProtParam server (https://web.expasy.org/protparam/). Subsequently, the amino acid sequences of MCP M14 family members from the three species were aligned using ClustalW [5]. The best-fitting amino acid substitution model was selected in MEGA-X using the Find Best Protein Models (ML) function according to the Bayesian Information Criterion (BIC). A maximum-likelihood phylogenetic tree was then constructed in MEGA-X under the LG + G model, with 1,000 bootstrap replicates. The resulting tree was visualized and refined using Evolview (v.3.0,https://www.evolgenius.info/evolview/). Conserved motifs in the amino acid sequences of MCP M14 family members were identified using MEME (https://meme-suite.org/meme/tools/meme), with the maximum number of motifs set to 20, and visualized using TBtools (v.2.084). Secondary structures of MCP M14 proteins were predicted using the SOPMA online server (https://npsaprabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=npsa_sopma.html), which estimates the relative proportions of α-helices, β-turns, extended strands and random coils. Three-dimensional structures of the MCP M14 proteins were predicted by homology modelling using SWISS-MODEL (https://swissmodel.expasy.org/). Finally, the amino acid sequence of Bos taurus BtCPA (accession number P00730) was downloaded from GenBank as a reference sequence, and protein sequences of all MCP M14 family members were aligned with the reference sequence using ClustalW and GeneDoc [30].

Integrative analysis of morphology and MCP M14 characteristics

To assess the association between digestive morphology and MCP M14 characteristics, nine morphological traits were integrated with six molecular variables, including the total number of detected MCP M14 sequences, the numbers of detected M14A, M14B and M14D sequences, subfamily richness, and the proportion of proteins retaining canonical Zn²⁺-binding residues. All variables were Z-score standardized to enable direct comparison among species. Spearman’s rank correlation coefficients (ρ) were calculated to characterize cross-species covariation between morphological and molecular traits. Given the limited number of species (n = 3), the correlations were used for exploratory comparison rather than statistical inference, and the standardized trait profiles and correlation matrices were visualized as heatmaps.

Results

Histological variation in the cardiac stomach

The cardiac stomach showed similar gross morphology among the three stomatopod species, forming a triangular sac in the central cephalothorax, with a narrow anterior region and a broader posterior region. In all three species, the cardiac-stomach wall is composed of chitin layer, mucosal epithelium, submucosal layer, muscular layer and outer membrane from the inside to the outside (Fig. 1A-F). The chitin-layer surface was irregular and differentiated into setal clusters (Fig. 1G-I). Setal branching varied among species, with 2–5 branches in O. japonicus, 5–9 in L. sulcirostris and 5–11 in O. oratoria. Seta were longest in O. oratoria, followed by O. japonicus and L. sulcirostris. The number of setal clusters per unit area (2,500 μm²) differed significantly among species (one-way ANOVA, F₂,₆ = 97.34, P < 0.001), with the highest value in O. oratoria (16.00 ± 1.00 clusters), followed by O. japonicus (13.33 ± 0.58 clusters) and L. sulcirostris (7.67 ± 0.58 clusters). Tukey’s HSD test showed significant differences among all three species (P < 0.05). The mucosal epithelium was thin and consisted of simple columnar epithelium and a basal membrane, whereas the submucosal layer was well developed and composed of loose connective tissue. The muscular layer was dominated by longitudinal muscle fibres, with fewer circular fibres, and lacked a clear boundary with the submucosa. Chitin-layer thickness differed significantly among species (one-way ANOVA, F₂,₆ = 288.64, P < 0.001), being greatest in L. sulcirostris (65.77 ± 0.81 μm), followed by O. japonicus (46.20 ± 3.74 μm) and O. oratoria (22.43 ± 0.21 μm). Tukey’s HSD test showed significant differences among all three species (all P < 0.001). Total cardiac-stomach wall thickness also differed significantly among species (one-way ANOVA, F₂,₆ = 673.21, P < 0.001), being greatest in L. sulcirostris (1,565.70 ± 25.96 μm), followed by O. oratoria (1,253.57 ± 10.13 μm) and O. japonicus (1,055.63 ± 10.37 μm). Tukey’s HSD test showed significant differences among all three species (all P < 0.001).

Fig. 1.

The histological structure of the cardiac stomach of O. japonicus (A, D, G), L. sulcirostris (B, E, H), and O. oratoria (C, F, I). Note: CL: chitin layer; CLL: chitin-layer thickness; BM: basal membrane; SCO: simple columnar epithelium; SUL: submucosal layer; CM: circular muscle; LM: longitudinal muscle; OUM: outer membrane; SE: setae

Histological variation in the pyloric stomach

The pyloric stomach had broadly similar histological organisation among the three species, but relative pyloric-stomach length differed significantly (one-way ANOVA, F₂,₆ = 7.61, P = 0.023). It was greatest in O. japonicus (0.0402 ± 0.0034), followed by O. oratoria (0.0379 ± 0.0048) and L. sulcirostris (0.0290 ± 0.0026). Tukey’s HSD test showed that O. japonicus had a significantly greater relative pyloric-stomach length than L. sulcirostris (P = 0.024), whereas the other pairwise comparisons were not significant (P > 0.05). The pyloric-stomach wall was structurally similar to that of the cardiac stomach and consisted of a chitin layer, mucosal epithelium, muscular layer and submucosal layer (Fig. 2A-F). The ventral projection narrowed towards the lumen and divided it into paired ventral chambers, forming the inter-ampullary ridge. The luminal chitin layer of this ridge differentiated into regularly arranged robust setae. Lateral projections formed the superior-ampullary ridge, whose chitin-layer surface bore dense, slender setae. Together, the setae between the inter-ampullary and superior-ampullary ridges formed the gland filter (Fig. 2G-L). The unilateral gland filter contained 39 pores in O. japonicus and 56 pores in both L. sulcirostris and O. oratoria. Gland-filter pore width differed significantly among species (one-way ANOVA, F₂,₆ = 33.99, P < 0.001), with values of 43.07 ± 0.53 μm in O. japonicus, 44.80 ± 0.36 μm in L. sulcirostris and 38.33 ± 1.60 μm in O. oratoria. Tukey’s HSD test showed that O. oratoria differed significantly from both O. japonicus (P = 0.003) and L. sulcirostris (P < 0.001), whereas O. japonicus and L. sulcirostris did not differ significantly (P = 0.164). The muscularis consisted mainly of circular muscle fibres in all species, and its thickness differed significantly among species (one-way ANOVA, F₂,₆ = 1257.03, P < 0.001), being greatest in L. sulcirostris (187.53 ± 5.10 μm), followed by O. oratoria (102.80 ± 0.44 μm) and O. japonicus (69.19 ± 0.65 μm). Tukey’s HSD test showed significant differences among all three species (all P < 0.001).

Fig. 2.

The histological structure of the pyloric stomach of O. japonicus (A, D, G, J), L. sulcirostris (B, E, H, K), and O. oratoria (C, F, I, L). Note: AC: ampullary channel; VC: ventral chamber; DC: dorsal chamber; IAR: inter-ampullary ridge; SAR: superior-ampullary ridge; GF: gland filter; SUL: submucosal layer; EPS: epithelial cell; SEP: setal panel; GF: gland filter; SE: setae

Histological variation in the hindgut

The hindgut showed similar gross morphology among the three species. It was tubular, narrower anteriorly and wider posteriorly, and opened to the anus on the ventral side. Relative hindgut length differed significantly among species (one-way ANOVA, F₂,₆ = 19.24, P = 0.002), being greatest in O. oratoria (0.1352 ± 0.0136), followed by L. sulcirostris (0.1274 ± 0.0088) and O. japonicus (0.0878 ± 0.0063). Tukey’s HSD test showed that O. japonicus had a significantly lower relative hindgut length than L. sulcirostris (P = 0.007) and O. oratoria (P = 0.003), whereas the latter two species did not differ significantly (P = 0.631). In all species, the hindgut wall is composed of mucosal epithelium, submucosal layer, muscular layer and adventitia from the lumen outwards (Fig. 3A-L). The mucosal epithelium projected into the lumen to form large longitudinal ridges, referred to as hindgut villi. These villi were mainly composed of tall columnar cells with round or oval nuclei, and their apical surfaces bore microvilli, increasing the epithelial surface area exposed to digesta (Fig. 3M-R). The muscularis was weakly developed and contained sparse longitudinal and circular muscle fibres. Goblet cells were present between columnar epithelial cells. Hindgut villus height differed significantly among species (one-way ANOVA, F₂,₆ = 61.99, P < 0.001) and was greatest in O. japonicus (559.37 ± 31.84 μm), followed by L. sulcirostris (360.03 ± 14.52 μm) and O. oratoria (357.73 ± 26.85 μm). Tukey’s HSD test showed that villus height was significantly greater in O. japonicus than in L. sulcirostris and O. oratoria (both P < 0.001), whereas L. sulcirostris and O. oratoria did not differ significantly (P = 0.993). Columnar-cell density did not differ significantly among species (one-way ANOVA, F₂,₆ = 2.46, P = 0.166), and no significant difference was detected in goblet-cell density (F₂,₆ = 2.81, P = 0.138). Columnar- and goblet-cell densities were slightly higher in O. japonicus (12.00 ± 1.72 columnar cells and 8.00 ± 1.00 goblet cells per 10,000 μm²) than in L. sulcirostris (9.67 ± 1.53 columnar cells and 6.67 ± 1.15 goblet cells per 10,000 μm²) and O. oratoria (9.67 ± 1.15 columnar cells and 6.00 ± 1.00 goblet cells per 10,000 μm²).

Fig. 3.

The histological structure of the hindgut of O. japonicus (A, B, C, D, M, P), L. sulcirostris (E, F, G, H, N, Q), and O. oratoria (I, J, K, L, O, R). Note: HV: hindgut villi; DC: hindgut cavity; VL: length of hindgut villi; VW: width of hindgut villi; MEL: mucosal epithelium layer; CM: circular muscle; LM: longitudinal muscle; OUM: outer membrane; GC: goblet cells; CC: columnar cells

Histological variation in the hepatopancreas

The hepatopancreas surrounded the digestive tract in all three species. Hepatopancreas length did not differ significantly between O. japonicus and O. oratoria, but both were significantly shorter than that of L. sulcirostris. In all species, the hepatopancreas consisted of two central glands and multiple glandular tubules. The tubules were composed of a basal lamina and a single layer of columnar epithelial cells, and their lumina showed variable shapes. The central glands comprised mucosal epithelium, submucosal layer and adventitia. Four cell types were identified in the hepatopancreas, including absorptive cells, secretory cells, fibrillar cells and embryonic cells. Among them, absorptive cells were the predominant cell type in all three species (Fig. 4A-F). SEM showed that the hepatopancreatic surface of L. sulcirostris and O. oratoria had numerous folds but was relatively smooth, whereas that of O. japonicus had fewer folds and a rougher surface (Fig. 4G-L). The hepatopancreas index did not differ significantly among species (one-way ANOVA, F₂,₆ = 2.76, P = 0.142), although the mean value was highest in O. japonicus (2.20 ± 0.65%), followed by O. oratoria (2.04 ± 0.29%) and L. sulcirostris (1.34 ± 0.42%).

Fig. 4.

The histological structure of the hepatopancreas of O. japonicus (A, D, G, J), L. sulcirostris (B, E, H, K), and O. oratoria (C, F, I, L). Note: CG: central gland; MG: mid-gut; HT: glandular tubule; LU: glandular tubule lumen; R: absorptive cell; B: secretory cell; F: fibrillar cell; E: embryonic cell

FL transcriptome information of L. sulcirostris

The present study generated the first FL transcriptome resource for L. sulcirostris. After removal of low-quality raw SMRT reads, 3,417,996 sequences were identified as FLNCs, corresponding to 4,697,832,384 bp with an average length of 1,374.75 bp. These FLNCs were further clustered into 74,697 consensus sequences, with a total length of 121,122,451 bp and an average length of 1,621.91 bp. After redundancy removal, 69,197 non-redundancy FL transcripts were eventually retained, with an average length of 1,618.08 bp.

Physicochemical properties of MCP M14 proteins in O. japonicus and L. sulcirostris

Using the HMM profile and the FL transcriptomes, we identified 45 and 26 candidate MCP M14 transcript sequences in O. japonicus and L. sulcirostris, respectively, compared with the 15 MCP M14 transcript sequences previously reported in O. oratoria. All predicted members contained the characteristic Zn_pept domain. We further analyzed the protein physicochemical properties of the MCP M14 gene family in O. japonicus and L. sulcirostris (Table 1). Results showed that the amino acid length, molecular weight (MW), isoelectric point (pI), instability index, aliphatic index and grand average of hydropathicity (GRAVY) of MCP M14 gene family in O. japonicus ranged from 94 to 1,315 aa, 9.84 to 142.67 kDa, 3.95 to 9.31, 16.00 to 53.90, 38.55 to 85.14 and −0.829 to −0.115, respectively, whereas the corresponding ranges in L. sulcirostris were 104 to 1,241 aa, 11.10 to 134.73 kDa, 4.37 to 9.19, 10.13 to 52.99, 56.90 to 91.21 and −0.689 to −0.041, respectively. Together with our previously reported data for O. oratoria, these results showed that 93.02% of the MCP M14 proteins in the three stomatopods were acidic (pI < 7), 83.72% were predicted to be stable (instability index < 40), and 97.67% were hydrophilic (GRAVY < 0).

Table 1.

Physicochemical properties of proteins belonging to the MCP M14 gene family in O. japonicus and L. sulcirostris

Species	Transcript_ID	Coding sequence (bp)	Amino acid (aa)	Molecular weight (kDa)	Isoelectric point	Instability index	Aliphatic index	Grand average of hydropathicity
O. japonicus	Oj44137	909	302	33.86	4.59	23.08	67.81	−0.375
	Oj36690	1131	376	41.94	5.10	24.48	81.41	−0.153
	Oj25159	1284	427	47.73	5.40	26.45	82.22	−0.226
	Oj16194	864	287	31.93	4.65	20.37	67.28	−0.362
	Oj43634	897	298	33.29	4.67	23.51	67.42	−0.370
	Oj34801	1089	362	40.50	5.08	26.46	72.49	−0.354
	Oj28066	1293	430	48.01	5.49	26.61	80.74	−0.223
	Oj27120	1260	419	46.83	5.57	27.89	80.07	−0.273
	Oj51378	1158	385	43.26	4.78	31.10	70.44	−0.397
	Oj57472	921	306	34.71	4.65	30.68	63.73	−0.443
	Oj40950	1314	437	48.81	5.35	30.76	76.32	−0.291
	Oj67764	882	293	32.89	4.62	33.30	62.56	−0.448
	Oj27208	1281	426	47.46	5.00	22.92	78.59	−0.241
	Oj42109	891	296	32.77	4.80	25.37	72.23	−0.282
	Oj51088	960	319	35.44	4.85	27.81	70.09	−0.331
	Oj34649	1092	363	40.35	4.97	26.10	74.21	−0.281
	Oj33788	1281	426	47.43	5.00	23.39	78.59	−0.229
	Oj31555	1275	424	47.12	5.30	27.09	81.89	−0.192
	Oj54010	744	247	27.71	4.56	23.42	65.55	−0.386
	Oj35087	1119	372	42.03	6.19	27.32	81.02	−0.250
	Oj75869	726	241	27.46	4.41	33.65	61.08	−0.498
	Oj50046	714	237	26.42	4.47	23.12	69.24	−0.306
	Oj33920	1212	403	45.64	5.89	36.99	79.58	−0.265
	Oj48120	1032	343	38.57	6.01	34.70	75.31	−0.320
	Oj39459	1077	358	40.03	5.45	24.99	81.45	−0.257
	Oj28180	1539	512	57.85	5.10	45.47	80.18	−0.465
	Oj67209	633	210	23.32	4.62	27.45	71.19	−0.257
	Oj86710	537	178	20.05	4.54	21.81	65.73	−0.393
	Oj61498	579	192	21.30	4.67	25.94	67.19	−0.28
	Oj39121	1068	355	39.92	9.06	42.38	77.75	−0.264
	Oj35201	1497	498	56.42	5.25	46.94	85.14	−0.409
	Oj48740	1212	403	45.68	5.13	49.10	77.47	−0.570
	Oj33167	375	124	14.41	6.49	27.23	38.55	−0.829
	Oj78473	465	154	16.89	4.55	26.73	78.05	−0.099
	Oj54914	1104	367	41.47	4.90	50.15	73.11	−0.629
	Oj50605	411	136	14.99	4.34	21.42	78.16	−0.240
	Oj76253	381	126	13.77	4.38	19.85	78.97	−0.200
	Oj47948	606	201	22.72	6.34	40.92	84.33	−0.327
	Oj61703	984	327	36.91	4.96	52.07	76.09	−0.563
	Oj74181	348	115	12.51	4.1	18.46	83.3	0.115
	Oj86406	324	107	11.55	3.95	19.10	84.95	0.044
	Oj81727	285	94	9.85	4.16	16.00	85.11	−0.141
	Oj18479	1803	600	66.12	6.24	43.56	67.25	−0.454
	Oj1729	3948	1315	142.67	9.31	53.9	66.43	−0.548
	Oj76963	720	239	26.73	4.81	52.05	76.12	−0.548
L. sulcirostris	Ls23747	1077	358	40.32	5.87	21.97	71.12	−0.413
	Ls26670	1134	377	42.44	6.10	24.59	56.9	−0.689
	Ls24559	1137	378	42.70	5.64	22.27	71.48	−0.420
	Ls27228	909	302	34.07	5.00	17.75	65.53	−0.457
	Ls27746	1014	337	37.95	5.52	19.3	68.61	−0.461
	Ls28741	1182	393	44.37	5.79	23.40	72.24	−0.399
	Ls29109	1296	431	48.25	6.13	24.83	80.35	−0.260
	Ls25943	1299	432	48.38	6.29	24.79	80.16	−0.268
	Ls22370	1419	472	52.97	6.45	27.48	81.4	−0.245
	Ls39545	861	286	32.04	5.07	17.34	66.12	−0.444
	Ls42539	1089	362	40.90	5.20	28.03	77.32	−0.329
	Ls26973	1293	430	48.07	5.58	25.90	82.09	−0.207
	Ls30182	1314	437	48.97	5.22	30.15	83.91	−0.198
	Ls56385	867	288	32.51	4.97	28.19	71.08	−0.359
	Ls22321	1281	426	47.35	5.02	21.49	79.48	−0.228
	Ls43678	759	252	28.74	5.90	23.76	61.51	−0.574
	Ls28591	1194	397	44.92	8.87	34.98	76.70	−0.357
	Ls24944	1644	547	61.93	5.37	39.80	78.45	−0.572
	Ls67760	585	194	21.95	4.76	22.74	71.34	−0.398
	Ls53403	558	185	20.63	5.02	15.69	69.03	−0.476
	Ls49572	1017	338	37.22	5.30	34.68	91.21	−0.041
	Ls50196	561	186	20.50	4.85	32.52	73.55	−0.311
	Ls71028	477	158	17.66	4.39	28.76	76.52	−0.253
	Ls58146	438	145	16.12	4.93	19.81	79.38	−0.297
	Ls62848	315	104	11.10	4.37	10.13	84.42	−0.198
	Ls2267	3726	1241	134.73	9.19	52.99	65.5	−0.533

Phylogenetic relationships of the MCP M14 gene family in three stomatopod species

Phylogenetic analysis of 86 transcript-derived MCP M14 sequences from the three stomatopod species showed that the sequences identified in O. japonicus and L. sulcirostris clustered into three subfamilies: M14A (O. japonicus: 36; L. sulcirostris: 24), M14B (O. japonicus: 7; L. sulcirostris: 1) and M14D (O. japonicus: 2; L. sulcirostris: 1), whereas all 15 MCP M14 sequences detected in O. oratoria clustered within the M14A subfamily (Fig. 5). Meanwhile, members of the same subfamily clustered within conserved clades, even when they originated from different species. The phylogenetic topology further suggested that the MCP M14 family members of O. oratoria have a closer homology with those of the L. sulcirostris compared to those of the O. japonicus.

Fig. 5.

Phylogenetic tree of the MCP M14 gene family in O. japonicus, L. sulcirostris, and O. oratoria

Conserved motifs of the MCP M14 gene family

We further characterized conserved motifs among MCP M14 gene family members of three stomatopod species and found that there were clear differences in the composition of conserved motifs among the three MCP M14 subfamilies (Fig. 6), whereas members of the same subfamily showed relatively conserved motif numbers and types. The M14A, M14B and M14D subfamilies contained up to 13, 8 and 3 conserved motifs, respectively, indicating that the M14A subfamily has the highest structural complexity. Notably, Motif 6 was present in nearly all MCP M14 gene family members, suggesting that it represents a core conserved element and may contribute to digestive function. Meanwhile, Motif 2, Motif 5 and Motif 8 were detected only in the M14A subfamily, whereas Motif 15 was restricted to the M14B subfamily, these motif patterns may be associated with functional diversification and subfamily-specific adaptation. Overall, members within the same MCP M14 subfamily displayed similar conserved motif architectures, further supporting their functional conservation.

Fig. 6.

Conserved motifs of the MCP M14 gene family in O. japonicus, L. sulcirostris, and O. oratoria

Multiple sequence alignment of the MCP M14 gene family

Using the B. taurus BtCPA sequence as a reference, we examined the canonical conserved features of MCP M14 gene family members in O. japonicus and L. sulcirostris (Fig. 7). Highly conserved Zn²⁺-binding residues (including His69, Glu72 and His196) were detected in 22 and 16 MCP M14 gene family members from O. japonicus and L. sulcirostris, respectively. These numbers were substantially higher than the number of O. oratoria MCP M14 gene family members retaining complete Zn²⁺-binding residues (3 members). In the remaining MCP M14 gene family members, the Zn²⁺-binding sites showed varying degrees of substitution or deletion. In addition, 21 and 16 MCP M14 gene family members from O. japonicus and L. sulcirostris retained the substrate-binding and catalytic residues Arg124, Arg127 and Arg145, respectively. Furthermore, 30 and 19 MCP M14 gene family members from O. japonicus and L. sulcirostris retained the conserved Cys138 and Cys161 residues, respectively, which form a disulfide bond associated with enzyme stability. These indicators are clear higher than those of the O. oratoria.

Fig. 7.

Multiple amino acid sequence alignment of MCP M14 family members in O. japonicus (Oj) and L. sulcirostris (Ls)

Prediction of secondary and tertiary protein structures of MCP M14 proteins

We predicted the secondary protein structures of MCP M14 gene family in O. japonicus and L. sulcirostris, and results showed that random coils (28.72–77.19% in O. japonicus and 34.62–75.50% in L. sulcirostris) constituted the dominant protein structural component. Meanwhile, α-helices represented the second most abundant component, accounting for 13.23–42.55% and 14.10–38.46% in O. japonicus and L. sulcirostris, respectively, whereas β-folding showed the lowest proportions, accounting for 1.37–10.32% and 1.05–9.62%, respectively. This overall secondary-structure protein composition was consistent with that previously observed in O. oratoria (Table 2).

Table 2.

Secondary protein structures of MCPM14 gene family in O.japonicus and L. sulcirostris

Species	Transcript_ID	α-helice (%)	Extended strand (%)	β-folding (%)	Random coil (%)
O. japonicus	Oj44137	32.12	21.19	5.63	41.06
	Oj36690	35.37	21.01	5.05	38.56
	Oj25159	37.94	17.80	5.15	39.11
	Oj16194	29.97	21.95	6.27	41.81
	Oj43634	31.21	21.14	5.03	42.62
	Oj34801	35.91	19.06	5.52	39.50
	Oj28066	36.98	17.44	4.19	41.40
	Oj27120	36.75	18.85	5.01	39.38
	Oj51378	34.03	19.48	4.68	41.82
	Oj57472	31.37	20.92	4.9	42.81
	Oj40950	38.67	17.85	4.12	39.36
	Oj67764	30.72	21.16	5.46	42.66
	Oj27208	36.62	17.61	4.23	41.55
	Oj42109	31.08	20.61	5.41	42.91
	Oj51088	33.23	18.81	5.33	42.63
	Oj34649	33.88	19.01	4.41	42.70
	Oj33788	36.62	18.08	4.69	40.61
	Oj31555	37.74	17.45	4.48	40.33
	Oj54010	32.39	19.84	4.86	42.91
	Oj35087	37.10	18.28	4.03	40.59
	Oj75869	31.95	18.67	5.81	43.57
	Oj50046	30.38	19.41	5.06	45.15
	Oj33920	34.00	18.61	3.97	43.42
	Oj48120	34.99	18.95	3.79	42.27
	Oj39459	34.92	19.83	4.75	40.50
	Oj28180	21.48	17.77	5.08	55.66
	Oj67209	27.14	20.00	6.19	46.67
	Oj86710	30.34	19.10	6.74	43.82
	Oj61498	29.17	18.75	5.21	46.88
	Oj39121	35.49	18.31	4.51	41.69
	Oj35201	21.29	20.28	5.42	53.01
	Oj48740	19.11	17.12	6.45	57.32
	Oj33167	21.77	19.35	4.84	54.03
	Oj78473	33.12	17.53	6.49	42.86
	Oj54914	16.62	17.44	6.81	59.13
	Oj50605	38.24	19.85	8.82	33.09
	Oj76253	34.13	21.43	10.32	34.13
	Oj47948	30.35	16.42	4.98	48.26
	Oj61703	18.35	17.74	6.12	57.80
	Oj74181	32.17	24.35	6.96	36.52
	Oj86406	35.51	19.63	7.48	37.38
	Oj81727	42.55	19.15	9.57	28.72
	Oj18479	24.67	16.17	2.50	56.67
	Oj1729	13.23	8.21	1.37	77.19
	Oj76963	14.35	20.25	7.59	57.81
L. sulcirostris	Ls23747	35.47	18.44	4.75	41.34
	Ls26670	35.01	19.89	4.77	40.32
	Ls24559	34.92	19.31	5.03	40.74
	Ls27228	31.13	21.19	6.62	41.06
	Ls27746	36.80	18.69	5.04	39.47
	Ls28741	35.88	19.08	4.33	40.71
	Ls29109	37.82	18.56	4.87	38.75
	Ls25943	37.50	18.29	4.63	39.58
	Ls22370	35.59	19.70	5.3	39.41
	Ls39545	28.23	22.38	6.29	43.01
	Ls42539	34.81	18.78	4.70	41.71
	Ls26973	37.44	17.91	5.12	39.53
	Ls30182	36.16	18.08	4.58	41.19
	Ls56385	27.78	22.57	5.56	44.10
	Ls22321	37.09	18.08	4.23	40.61
	Ls43678	31.75	18.65	5.16	44.44
	Ls28591	33.75	18.89	4.28	43.07
	Ls24944	20.48	15.72	5.12	58.68
	Ls67760	29.90	18.56	5.67	45.88
	Ls53403	29.73	19.46	5.95	44.86
	Ls49572	38.46	13.91	2.96	44.67
	Ls50196	27.96	19.35	6.45	46.24
	Ls71028	34.18	17.72	5.70	42.41
	Ls58146	36.55	18.62	6.21	38.62
	Ls62848	38.46	17.31	9.62	34.62
	Ls2267	14.10	9.35	1.05	75.50

We further constructed tertiary protein structure models (Fig. 8), and the predicted models showed high agreement with their corresponding templates, supporting the overall reliability of the models. Tertiary protein structure prediction for O. japonicus and L. sulcirostris was based on 14 (1aye.1.A, 1dtd.1.A, 1kwm.2.A, 1m4l.1.A, 1pca.1.A, 1qmu.1.A, 1uwy.1.A, 2boa.1.A, 2nsm.1.A, 5mrv.1.A, 5om9.1.A, 7eqz.1.A, 8v3m.1.A, 8v4k.1.E) and 10 (1aye.1.A, 1kwm.2.A, 1pca.1.A, 2boa.1.A, 2nsm.1.A, 3jlg.1.A, 5mrv.1.A, 5om9.1.A, 7eqz.1.A, 8v4k.1.E) templates, respectively, and these template numbers were higher than that used for O. oratoria (5 templates). For O. japonicus and L. sulcirostris, GMQE values ranged from 0.27 to 0.86 and from 0.32 to 0.83, respectively, while QMEAND values ranged from 0.54 ± 0.05 to 0.82 ± 0.05 and from 0.56 ± 0.05 to 0.79 ± 0.05, respectively (Table 3). These results were comparable to those obtained for O. oratoria, indicating that the predicted models were generally reliable, although structural variation was evident among members. Furthermore, our results also confirmed that the tertiary protein structure of the MCP M14 gene family exhibited a structural distribution that was consistent with the predicted secondary protein structure composition.

Fig. 8.

Three-dimensional protein structure prediction models for the MCP M14 gene family members of O. japonicus (Oj) and L. sulcirostris (Ls)

Table 3.

Detailed information of the predicted three-dimensional of MCP M14 gene family in O.japonicus and L. sulcirostris

Species	Transcript_ID	Template	Seq_Identity (%)	Oligo-State	GMQE	QMEAND
O. japonicus	Oj44137	1m4l.1.A	42.91	Monomer	0.82	0.78 ± 0.05
	Oj36690	5om9.1.A	37.46	Monomer	0.75	0.73 ± 0.05
	Oj25159	1aye.1.A	36.99	Monomer	0.73	0.73 ± 0.05
	Oj16194	7eqz.1.A	45.94	Monomer	0.82	0.77 ± 0.05
	Oj43634	7eqz.1.A	45.24	Monomer	0.82	0.78 ± 0.05
	Oj34801	2boa.1.A	39.94	Monomer	0.78	0.74 ± 0.05
	Oj28066	1pca.1.A	35.70	Monomer	0.73	0.73 ± 0.05
	Oj27120	1aye.1.A	36.98	Monomer	0.73	0.73 ± 0.05
	Oj51378	1aye.1.A	39.45	Monomer	0.74	0.72 ± 0.05
	Oj57472	1m4l.1.A	43.24	Monomer	0.80	0.77 ± 0.05
	Oj40950	1pca.1.A	37.22	Monomer	0.71	0.73 ± 0.05
	Oj67764	7eqz.1.A	44.52	Monomer	0.80	0.76 ± 0.05
	Oj27208	1aye.1.A	44.90	Monomer	0.77	0.77 ± 0.05
	Oj42109	1dtd.1.A	50.68	Monomer	0.86	0.82 ± 0.05
	Oj51088	1aye.1.A	49.68	Monomer	0.83	0.8 ± 0.05
	Oj34649	1aye.1.A	48.03	Monomer	0.81	0.76 ± 0.05
	Oj33788	1aye.1.A	45.15	Monomer	0.77	0.78 ± 0.05
	Oj31555	1aye.1.A	35.66	Monomer	0.72	0.73 ± 0.05
	Oj54010	7eqz.1.A	47.54	Monomer	0.82	0.77 ± 0.05
	Oj35087	1pca.1.A	33.33	Monomer	0.71	0.71 ± 0.05
	Oj75869	7eqz.1.A	45.92	Monomer	0.79	0.73 ± 0.05
	Oj50046	1dtd.1.A	49.57	Monomer	0.84	0.78 ± 0.05
	Oj33920	1aye.1.A	31.99	Monomer	0.61	0.68 ± 0.05
	Oj48120	1pca.1.A	35.43	Monomer	0.68	0.69 ± 0.05
	Oj39459	1aye.1.A	36.45	Monomer	0.69	0.69 ± 0.05
	Oj28180	2nsm.1.A	45.90	Monomer	0.61	0.72 ± 0.05
	Oj67209	1aye.1.A	50.97	Monomer	0.80	0.73 ± 0.06
	Oj86710	1kwm.2.A	47.40	Monomer	0.79	0.75 ± 0.07
	Oj61498	1aye.1.A	50.27	Monomer	0.82	0.77 ± 0.06
	Oj39121	1pca.1.A	30.92	Monomer	0.55	0.66 ± 0.05
	Oj35201	2nsm.1.A	45.3	Monomer	0.55	0.7 ± 0.05
	Oj48740	2nsm.1.A	45.45	Monomer	0.62	0.7 ± 0.05
	Oj33167	7eqz.1.A	50.00	Monomer	0.63	0.63 ± 0.08
	Oj78473	1dtd.1.A	47.68	Monomer	0.82	0.76 ± 0.07
	Oj54914	2nsm.1.A	44.52	Monomer	0.59	0.67 ± 0.05
	Oj50605	5mrv.1.A	42.22	Monomer	0.82	0.78 ± 0.07
	Oj76253	5mrv.1.A	46.34	Monomer	0.82	0.79 ± 0.07
	Oj47948	1qmu.1.A	38.36	Monomer	0.56	0.71 ± 0.07
	Oj61703	1qmu.1.A	44.72	Monomer	0.60	0.7 ± 0.05
	Oj74181	1aye.1.A	48.25	Monomer	0.83	0.78 ± 0.08
	Oj86406	5mrv.1.A	42.45	Monomer	0.83	0.79 ± 0.08
	Oj81727	5mrv.1.A	47.83	Monomer	0.79	0.74 ± 0.09
	Oj18479	8v3m.1.A	46.91	Monomer	0.53	0.66 ± 0.05
	Oj1729	8v4k.1.E	47.60	Monomer	0.27	0.54 ± 0.05
	Oj76963	1uwy.1.A	37.75	Monomer	0.43	0.58 ± 0.07
L. sulcirostris	Ls23747	2boa.1.A	40.97	Monomer	0.78	0.74 ± 0.05
	Ls26670	2boa.1.A	39.06	Monomer	0.75	0.72 ± 0.05
	Ls24559	1aye.1.A	39.29	Monomer	0.77	0.74 ± 0.05
	Ls27228	7eqz.1.A	44.75	Monomer	0.82	0.78 ± 0.05
	Ls27746	2boa.1.A	42.81	Monomer	0.79	0.75 ± 0.05
	Ls28741	1aye.1.A	38.16	Monomer	0.76	0.73 ± 0.05
	Ls29109	1aye.1.A	38.01	Monomer	0.72	0.73 ± 0.05
	Ls25943	1pca.1.A	36.80	Monomer	0.73	0.74 ± 0.05
	Ls22370	1aye.1.A	38.11	Monomer	0.66	0.72 ± 0.05
	Ls39545	7eqz.1.A	45.71	Monomer	0.83	0.79 ± 0.05
	Ls42539	2boa.1.A	40.40	Monomer	0.76	0.73 ± 0.05
	Ls26973	1aye.1.A	36.99	Monomer	0.73	0.73 ± 0.05
	Ls30182	1aye.1.A	40.36	Monomer	0.70	0.71 ± 0.05
	Ls56385	7eqz.1.A	45.71	Monomer	0.82	0.78 ± 0.05
	Ls22321	1aye.1.A	43.11	Monomer	0.76	0.76 ± 0.05
	Ls43678	7eqz.1.A	44.35	Monomer	0.81	0.76 ± 0.05
	Ls28591	5om9.1.A	35.78	Monomer	0.61	0.66 ± 0.05
	Ls24944	2nsm.1.A	47.50	Monomer	0.54	0.64 ± 0.05
	Ls67760	3jlg.1.A	48.13	Monomer	0.79	0.74 ± 0.06
	Ls53403	1kwm.2.A	46.70	Monomer	0.82	0.77 ± 0.06
	Ls49572	2boa.1.A	38.33	Monomer	0.65	0.66 ± 0.05
	Ls50196	1aye.1.A	48.90	Monomer	0.83	0.77 ± 0.06
	Ls71028	1kwm.2.A	46.75	Monomer	0.79	0.75 ± 0.07
	Ls58146	5mrv.1.A	40.97	Monomer	0.81	0.77 ± 0.07
	Ls62848	5om9.1.A	48.04	Monomer	0.76	0.71 ± 0.08
	Ls2267	8v4k.1.E	47.86	Monomer	0.32	0.56 ± 0.05

Integrated patterns of digestive adaptation

Integrated analysis revealed distinct morphological–molecular profiles among the three stomatopod species (Fig. 9A). O. japonicus exhibited the greatest hindgut villus height and hepatopancreas index, together with the largest and most diverse detected MCP M14 repertoire. In contrast, L. sulcirostris was characterized by greater cardiac-stomach wall and chitin-layer thickness, a thicker pyloric muscular layer, and the highest proportion of MCP M14 proteins retaining canonical Zn²⁺-binding residues. O. oratoria displayed the highest setal density, the narrowest gland-filter pores and the greatest relative hindgut length, but the smallest detected MCP M14 repertoire. Exploratory correlation analysis further indicated that morphological and molecular traits did not vary uniformly across species (Fig. 9B). Hindgut villus height covaried positively with the total number of detected MCP M14 sequences and the numbers of detected M14A, M14B and M14D sequences (ρ = 1.00), whereas relative hindgut length exhibited the opposite pattern (ρ = − 1.00).

Fig. 9.

Integrated comparison of digestive morphology and MCP M14 characteristics among three stomatopod species. A Z-score-standardized heatmap of nine morphological traits and six transcript-derived MCP M14 variables in O. japonicus (OJ), L. sulcirostris (LS) and O. oratoria (OO). B Exploratory Spearman correlation heatmap between morphological and MCP M14 variables. Cell values represent Z scores in (A) and Spearman’s ρ in (B)

Discussion

This study compared digestive-system morphology and MCP M14 characteristics among O. japonicus, L. sulcirostris and O. oratoria to characterize interspecific differences and examine whether these differences followed the expected smasher–spearer pattern associated with trophic niche differentiation. The results provided only partial support for this expectation. O. japonicus showed a broader detected MCP M14 repertoire, whereas the most developed gastric structures associated with mechanical food processing were found in L. sulcirostris, one of the two spearers. This contrast suggests that digestive differentiation among the three species is more species-specific and cannot be explained simply by the smasher–spearer classification.

Morphological differentiation of the digestive tract is associated with trophic niche and feeding adaptation among three stomatopods

Trait divergence is generally regarded as an important outcome of species adaptation to distinct habitats and resource-use strategies [24]. The morphology and histological characteristics of the animal digestive tract are closely associated with diet [53]. In stomatopods, the cardiac stomach primarily contributes to preliminary food filtration and mechanical trituration [48]. In the present study, the number of setal clusters per unit area and setal length in O. oratoria reached the highest values, followed by O. japonicus, while L. sulcirostris showed the lowest values. These observations indicate greater development of setal filtration structures in O. oratoria. Cardiac-stomach wall thickness also differed significantly among species, being greatest in L. sulcirostris, followed by O. oratoria and O. japonicus. Notably, the substantial difference between the two spearers, L. sulcirostris and O. oratoria, indicates that digestive morphology is not determined solely by the broad smasher–spearer classification. Instead, these differences may reflect species-specific digestive specialization. Their relationship with prey diversity or food-particle composition remains a hypothesis because diet was not measured directly in the present study. Chitin-layer thickness differed significantly among species, with the greatest value observed in L. sulcirostris, followed by O. japonicus and O. oratoria. The thicker chitin layer in L. sulcirostris may provide greater structural reinforcement during mechanical food processing. However, because prey hardness was not directly measured, its relationship with dietary properties remains to be tested directly.

The pyloric stomach further grinds and filters food through rhythmic contractions of the muscular layer, which drive the gland filter [43]. Pyloric muscle thickness differed significantly among species, following the order L. sulcirostris > O. oratoria > O. japonicus. The pronounced difference between the two spearers further supports species-specific differentiation in digestive architecture rather than a simple morphological division between smashers and spearers. In fact, the small particles of food that enter the pyloric stomach will sink into the ventral chamber and be filtered by the gland filters within it, the particles that are small enough will then be transported to the hepatopancreas for digestion and absorption, while the larger particles will travel along the dorsal chamber and enter the midgut [9]. Gland-filter pore width also differed significantly among species. O. oratoria had significantly narrower pores than O. japonicus and L. sulcirostris, whereas no significant difference was detected between the latter two species. These differences may indicate interspecific variation in particle-processing characteristics, although the size and composition of naturally ingested food particles were not directly examined in this study.

The hepatopancreas is a major digestive gland in crustacean and serves as the principal site for digestive enzyme secretion and nutrient absorption [11, 48]. No obvious interspecific difference was observed in hepatopancreatic cell types among the three stomatopods, all were composed of absorptive cell, secretory cell, fibrillar cell, embryonic cell, which is consistent with the observations of other crustaceans [4]. The hepatopancreas index did not differ significantly among species, although O. japonicus showed the highest mean value. Therefore, the present results do not provide statistical evidence for interspecific differences in relative hepatopancreas size, and differences in digestive or absorptive performance cannot be inferred from this index alone. Overall, although the three stomatopods share a broadly conserved histological organization of the digestive system, significant interspecific differences were detected in cardiac-stomach setal density, chitin-layer thickness, pyloric muscle thickness, gland-filter pore width, relative hindgut length and hindgut villus height. In contrast, columnar- and goblet-cell densities and the hepatopancreas index did not differ significantly among species. These morphological differences demonstrate clear interspecific divergence in digestive architecture. Their associations with food-resource use and trophic niche differentiation are plausible but require direct dietary evidence for further validation.

The interspecific specificity and functional differentiation MCP M14 of the gene family in three stomatopod species

Carboxypeptidases, particularly metallocarboxypeptidase, have been confirmed to be critical exopeptidases involved in food digestion, protein degradation and zymogen activation in animals [15, 22, 25, 41]. Given the differences in feeding strategies and trophic niches among O. japonicus, L. sulcirostris and O. oratoria [3], it is reasonable to expect species-specific features in their metallocarboxypeptidase gene repertoires. In the absence of high-quality reference genomes, FL transcriptomes provide an important resource for resolving the compositional differences and expression potential of the MCP M14 gene family in stomatopods [40]. In this study, we generated the first FL transcriptome resources for L. sulcirostris and integrated them with our previously published FL transcriptome of O. oratoria and O. japonicus to systematically identify and compare the MCP M14 gene family in three stomatopod species. Based on the available FL transcriptomic resources, 45, 26 and 15 MCP M14 transcript sequences were identified in O. japonicus, L. sulcirostris and O. oratoria, respectively. Given the limited genomic resources available for stomatopods, the use of multi-tissue FL transcriptomes enabled broader recovery of expressed transcripts and provided a more comprehensive basis for MCP M14 identification. However, because the tissues were pooled, the tissue origin of individual MCP M14 transcripts could not be specifically determined. These numbers therefore reflect the MCP M14 transcript repertoire detected under the present sampling conditions rather than absolute genomic gene copy numbers. Moreover, because the FL transcriptomes were not generated with biological replicates, they were not used for quantitative interspecific expression comparisons. These interspecific differences suggest that long-term differences in predatory strategy and trophic niche differentiation may be associated with potential differences in peptide and protein digestion among these stomatopods [3, 6, 38]. In fact, similar differences have also been observed in other animals with close genetic relationships, such as 14 members of the Antheraea pernyi and 20 members of the A. yamamai [58]. Further phylogenetic classification showed that all identified MCP M14 gene family members could be assigned to M14A, M14B and M14D subfamilies, broadly consistent with the previously published clustering results of the MCP M14 gene family [36]. A large proportion of MCP M14 proteins in the three stomatopod species were characterized by acidic isoelectric points, high structural stability, hydrophilicity and long half-life period, this may be to maintain the integrity and functional persistence of the protein structure in the acidic intestinal environment, and ultimately improve the efficiency of food digestion [57].

Notably, M14A transcripts were detected in all three stomatopod species, whereas no M14B or M14D transcripts were detected in O. oratoria. Because these sequences were identified from FL transcriptomes rather than whole-genome data, the lack of detected M14B and M14D transcripts should not be interpreted as evidence of their genomic absence. It may instead reflect differences in gene expression at the time of sampling, tissue composition or pooling, and sequencing depth. Among the detected MCP M14 sequences, M14A was the most abundant subfamily. This suggests that the M14A subfamily may play a more central role in dietary protein digestion in stomatopods [8, 47]. Additionally, the M14A subfamily has also been proposed to participate in endodermal digestion and absorption and epidermal renewal during arthropod moulting [29, 44]. Therefore, interspecific member differences of M14A subfamily may be associated not only with variation in dietary protein digestion, but also with species-specific differences in growth cycles and moulting processes. We also found that the decomposition capabilities of three MCP M14 subfamilies also vary. Specifically, members of the M14A subfamily preferentially cleave aromatic or aliphatic amino acid residues, whereas members of the M14B subfamily preferentially act on basic amino acids and can also hydrolyse some non-basic residues [35]. Although M14A, M14B and M14D subfamilies share broadly similar three-dimensional structures, their Zn²⁺-binding amino acid motifs differ, being represented by H₆₉A(S)RE₇₂…H₁₉₆, H₆₉GNE₇₂…H₁₉₆ and H₆₉PGE₇₂…H₁₉₆, respectively. These differences may lead to significant variations in the activation and catalytic activities of enzymes, and ultimately affect the degradation efficiency of proteins [34]. The M14A subfamily contained the largest number of conserved motifs (13 motifs), suggesting that it may perform more complex or diverse physiological functions, including interference with viral maturation, neuropeptide hormone processing and regulation of cell growth [37, 45]. By contrast, the smaller number of conserved motifs in the M14B and M14D subfamilies suggests more specialized functions. Among all motifs, motif 6 was almost universally present, indicating that it may encode a core structural domain required for the catalytic activity of metallocarboxypeptidases and may play an essential role in the basal catalytic function of MCP M14 gene family. Meanwhile, motif 2, motif 5, motif 8 and motif 15 were completely retained only in the M14A or M14B subfamilies, and these motifs may provide the structural basis for subfamily-specific substrate recognition, catalytic specificity and functional adaptation [14]. Overall, the MCP M14 gene family in the three stomatopod species is highly conserved in its general structural organization, but variation in Zn²⁺-binding residues and conserved motifs still points to potential functional differentiation among species and subfamilies.

Alternative digestive strategies underlying trophic niche differentiation

Our integrative analysis indicates that digestive adaptation in the three stomatopods is characterized by modular rather than uniform morphological–molecular differentiation. Digestive performance can evolve through coordinated changes in food-processing structures, absorptive surfaces and digestive-enzyme repertoires, without requiring parallel enhancement of all components of the digestive system [24],Štrus et al., 2019; Navarrete [11, 46]. Comparative studies across animal lineages further show that diversification of digestive-enzyme gene repertoires can accompany dietary transitions and ecological specialization [26, 49, 50, 60]. Within this framework, the combination of greater hindgut villus development and a broader detected MCP M14 repertoire in O. japonicus, enhanced mechanical-processing structures in L. sulcirostris, and more pronounced filtration-related traits in O. oratoria reveals contrasting morphological–molecular configurations among the three species. These patterns are consistent with different forms of digestive specialization, but they should not be interpreted as direct evidence of diet-driven adaptation because diet composition was not empirically examined in this study. Previous comparative studies have shown that digestive morphology can be influenced by both trophic ecology and phylogenetic history, emphasizing the need to consider shared ancestry when interpreting interspecific morphological differences [10]. Lineage-specific evolutionary changes in digestive-enzyme genes may likewise accompany dietary and ecological divergence [26, 49]. Habitat may also contribute to variation in digestive traits, as comparative analyses have shown that both diet and habitat are associated with differences in intestinal morphology across species [12]. Therefore, the morphological and MCP M14 differences observed here may reflect a combination of dietary, phylogenetic and habitat-related effects. Because these factors were not independently quantified in the present three-species comparison, their relative contributions cannot be separated. Future studies combining direct diet analysis, broader phylogenetic sampling and habitat characterization will be required to test these relationships more rigorously.

Conclusion

This study reveals distinct interspecific differences in digestive-tract morphology and MCP M14 characteristics among three stomatopod species. Integrative analysis further revealed contrasting morphological–molecular profiles, with O. japonicus characterized by greater hindgut villus development and a broader detected MCP M14 repertoire, L. sulcirostris by enhanced mechanical-processing structures, and O. oratoria by more pronounced filtration-related traits. These patterns are consistent with species-specific digestive specialization and possible trophic niche differentiation, but their direct relationship with diet requires further empirical testing.

Authors’ contributions

Conceptualization, Fangrui Lou; methodology, Jiantong Zhang; software, Jiantong Zhang and Xiaowen Duan; validation, Fangrui Lou; formal analysis, Jiantong Zhang; investigation, Jiantong Zhang and Xiaowen Duan; resources, Fangrui Lou; data curation, Jiantong Zhang and Xiaowen Duan; writing-original draft preparation, Fangrui Lou and Qi Huang; writing-review and editing, Fangrui Lou and Qi Huang; visualization, Fangrui Lou and Qi Huang; supervision, Fangrui Lou; project administration, Fangrui Lou; funding acquisition, Fangrui Lou.

Funding

This work was supported by the Natural Science Foundation of Shandong Province (ZR2024MD076) and the Development Plan of Youth Innovation Team in Colleges and Universities of Shandong Province (2025KJI006).

Data availability

The data underlying this article are stored in the SRA database of NCBI under BioProject PRJNA1190238.

Declarations

Ethics approval and consent to participate

Odontodactylus japonicus, Lysiosquilla sulcirostris and Oratosquilla oratoria are not protected species in any country. All the animal experiment procedures comply with the ethical standards and have received approval from the Experimental Animal Ethics Committee of Yantai University.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.