ABSTRACT
Digestive systems are complex organs that allow organisms to absorb energy from their environment to fuel vital processes such as growth, development and the maintenance of homeostasis. A comprehensive understanding of digestive physiology is therefore essential to fully understand the energetics of an organism. The digestion of proteins is of particular importance because most heterotrophic organisms are not able to synthesize all essential amino acids. While Echinoderms are basal deuterostomes that share a large genetic similarity with vertebrates, their digestion physiology remains largely unexplored. Using a genetic approach, this work demonstrated that several protease genes including an enteropeptidase, aminopeptidase, carboxypeptidase and trypsin involved in mammalian digestive networks are also found in sea urchin larvae. Through characterization including perturbation experiments with different food treatments and pharmacological inhibition of proteases using specific inhibitors, as well as transcriptomic analysis, we conclude that the trypsin-2 gene codes for a crucial enzyme for protein digestion in Strongylocentrotus purpuratus. Measurements of in vivo digestion rates in the transparent sea urchin larva were not altered by pharmacological inhibition of trypsin (using soybean trypsin inhibitor) or serine proteases (aprotinin), suggesting that proteases are not critically involved in the initial step of microalgal breakdown. This work provides new insights into the digestive physiology of a basal deuterostome and allows comparisons from the molecular to the functional level in the digestive systems of vertebrates and mammals. This knowledge will contribute to a better understanding for conserved digestive mechanisms that evolved in close interaction with their biotic and abiotic environment.
INTRODUCTION
A basic concept of life lies in the fact that an organism must constantly expend energy to maintain a disequilibrium between itself and its environment. Unlike plants and some protozoa, heterotrophic organisms must absorb energy in the form of organic molecules, e.g. carbohydrates, amino acids and lipids. While basal metazoan clades such as Porifera and Placozoa absorb energy in the form of dissolved or particulate matter through pinocytosis and phagocytosis (Grell and Benwitz, 1981; Lunger, 1963; Schulze, 1892), Bilateria developed specific organs for energy absorption and use extracellular digestion for a more efficient breakdown of ingested food. The development of extracellular digestion is considered a crucial step during animal evolution (Yonge, 1937; Nielsen, 2008) to exploit food more efficiently and thus to grow to larger sizes (Arendt, 2007). A multipartite gastrointestinal tract, which forms in Bilateria from the endoderm during gastrulation, allows energy uptake that is far beyond that from intracellular digestion. Higher animals have specialized cell types in their digestive system that are responsible for the production and secretion of digestive enzymes. The mammalian pancreas is a highly specialized organ that controls not only the exocrine secretion of digestive enzymes into the gut but also the endocrine production of signaling molecules (Heller, 2010). Even though invertebrates lack a true pancreas, exocrine pancreatic-like cells are described for many phyla, including tunicates and cephalochordates (Lecroisey et al., 2015; Olinski et al., 2006). In sea urchin larvae, single-cell transcriptomics and localization of pancreatic-like gene transcripts have identified exocrine pancreatic-like cells to form a ring-shaped structure directly behind the cardiac sphincter (Perillo et al., 2016). Paganos et al. (2022) proposed, based on gene transcript patterns, that sea urchin larvae have certain cell types homologous to mammalian pancreatic acinar cells. Comparing the transcription factor patterning of the midgut cells, a strong similarity between echinoderms and vertebrates has been observed, indicating that pancreatic-like cell types may have evolved in a common ancestor of deuterostomes (Annunziata et al., 2019). Single-cell transcriptomes of the sea urchin larva demonstrated that acinar cell-like cell types display a range of digestive enzyme transcripts that may be involved in extracellular digestion including carbohydrases, lipases and proteinases. However, their physiological function in the larval digestive system remains hypothetical.
In invertebrates, trypsin and chymotrypsin have been described as the major digestive proteases (Barker and Gibson, 1979; Yang and Davies, 1971). Trypsin belongs to the group of serine proteases characterized by a catalytic triad consisting of a His, Asp and Ser residues (Neurath, 1984; Pancer et al., 1996). As an endopeptidase, trypsin cleaves peptide bonds after basic amino acids such as Arg and Lys. There are several trypsin genes found in the sea urchin genome, but trypsin-2 is prominently expressed in the acinar cell-like cell type in the sea urchin larvae (Paganos et al., 2022). However, the physiological function of those genes in larval digestion has not been well characterized.
Echinoderms are used as model systems for evolutionary developmental and cell biology because of the abundance of excellent genomic resources, their relatively simple body plan and their phylogenetic position as a sister group to chordates (Arnone et al., 2015). The larval tripartite digestive tract consists of an esophagus, a midgut (stomach) and a hindgut (intestine) with cardiac and pyloric sphincters between the esophagus and midgut, and the midgut and hindgut, respectively (Burke, 1981; Nezlin and Yushin, 1994). During larval development, until the end of gastrulation (around 48 h post-fertilization, hpf), no specific segments (e.g. esophagus, midgut, hindgut) of the larval digestive tract are recognizable. The differentiation takes place between 60 and 72 hpf, when the larvae have reached the pluteus stage and the various parts of the digestive tract become clearly distinguishable (Annunziata et al., 2014). Energy uptake in the planktotrophic larvae occurs through feeding on phytoplankton and other particles by creating a current with ciliated bands. This carries food particles into the esophagus to form a bolus, which is subsequently swallowed into the midgut (Strathmann, 1975; Feehan et al., 2018). Sea urchin larvae can be used for a variety of molecular analyses as well as physiological experiments, including the determination of in vivo digestion rates (Stumpp et al., 2013). The aim of this work was to identify the major digestive proteases, such as endopeptidase, aminopeptidase, carboxypeptidase and trypsin, in the sea urchin Strongylocentrotus purpuratus. For physiological assays, we focused on trypsin as the prominent digestive protease to verify its role in larval food breakdown using in vivo digestion, larval perturbation experiments with different food treatments as well as pharmacological inhibition using protease inhibitors.
MATERIALS AND METHODS
Adult and larval cultures
Adult Strongylocentrotus purpuratus (Stimpson 1857) were collected on the Californian coast (La Jolla, CA, USA) and transferred to the Helmholtz Center for Ocean Research Kiel (GEOMAR). They were maintained in circulating systems at 10°C with sea water from the Kiel Fjord adjusted with artificial sea salt Pro Reef (Tropic Marin, Hünenberg, Swiss) to a salinity of 31.5 psu. Water was changed and sea urchins were fed ad libitum with kelp (Macrocystis porifera) 3 times a week.
For larval cultures, spawning was induced by gently shaking adult sea urchins. Eggs were collected in 0.2 µm filtered seawater (FSW, 31.5 psu) and washed briefly with FSW before fertilization. Dry sperm was directly collected and fertilization was conducted by mixing diluted sperm with eggs. Healthy zygotes with fertilization rates >99% were placed into 2 l culture flasks with an initial concentration of 15 larvae ml−1. Cultures were kept in low light conditions on a 12 h:12 h dark:light cycle at 15°C and were mixed via a light stream of pressurized air bubbles. Larvae were checked every 2 days for health via visual inspection of morphology and culture density using two 10 ml samples per replicate. Water was changed every second day and larvae were fed every other day after the water change starting at 3 dpf after gastrulation. Four different food concentrations of 0, 500, 2000 and 8000 cells ml−1Rhodomonas sp. were fed in each of the five experimental larval cultures (Table 1; Fig. S1). When water in 8000 cells ml−1 cultures was changed, remaining algae cells were clearly visible, indicating ad libitum feeding conditions. Larval growth rates were determined by measuring total body length on a fluorescence microscope (Axio Imager Z, Carl Zeiss, Jena, Germany) and analyzed with AxioVision software (Carl Zeiss) (Stumpp, 2011).
Sampling for enzyme, in vitro inhibitor and RT-qPCR assays
Experiment 1 samples were taken from day 2 to 13 dpf for ontogeny analysis. Experiment 2 samples were taken at 2, 4, 5, 6, 7, 8, 9 and 10 dpf for qPCR analysis. Experiment 3 and 4 samples were taken at 7 dpf for the analysis of protease pH sensitivity and in vitro protease inhibition (Fig. S1). The seawater chemistry including salinity, temperature and pH was determined every second day for experiment 1 and 2 samples and every day for experiment 3 and 4 samples (Table 1) using handheld WTW pH-Meter 3110 Set 2 and Conductometer WTW Cond 3110 probes (Xylem Analytics, Weilheim, Germany). Approximately 5000 (experiments 1 and 2) and 10,000 (experiments 3 and 4) larvae were pooled for each sample. Excess water was discarded and samples were frozen in liquid nitrogen and then stored at −80°C until further processing.
RT-qPCR and single-cell transcriptomic analyses
RT-qPCR was performed as previously described (Hu et al., 2018). Briefly, RNA samples from eight different developmental stages and three different food concentrations (experiment 2) were extracted using a Direct-zol RNA Micro Prep Kit (Zymo Research, Freiburg, Germany) following the instruction manual. mRNA was reverse transcribed into cDNA using the SuperScript IV Kit (Thermo Fisher Scientific, Waltham, MA, USA). qPCR analysis was carried out using the 2× qPCRBIO SyGreen Mix Hi-ROX (PCR Biosystems Ltd, London, UK) via two-step real-time qPCR. Four primer pairs (Table 2) were designed using the NCBI primer blast targeting exon junctions of each gene for which the mRNA sequence is available from the S. purpuratus genome database (Arshinoff et al., 2022). Specificity of primers was tested by regular PCR demonstrating amplification of one specific product in the predicted size range. qPCR measurements were conducted on a QuantStudio 1 Real Time PCR Instrument (Applied Biosystems, Waltham, MA, USA) with initial denaturation for 10 min (95°C) followed by 40 cycles of denaturation for 15 s (95°C), annealing for 20 s (58°C) and elongation for 35 s (72°C) with subsequent melting curve analysis. For quantification, the ΔΔCT method was used and gene expression levels were normalized to those of the housekeeping gene EF1a, which has been demonstrated to be stable over ontogeny (Hu et al., 2020).
Single-cell data were kindly provided by Jonas Brandenburg and David Garfield (Integrative Research Institute for the Life Sciences, Humboldt-Universität zu Berlin, Germany). They had been gathered as previously described (Chang et al., 2021). Briefly, pluteus stage larvae were incubated for 5 min in hyalin extraction medium (0.3 mol l−1 glycine, 0.3 mol l−1 NaCl, 0.01 mol l−1 KCl, 0.01 mol l−1 MgSO4, 0.01 mol l−1 Tris pH 8.0, and 0.002 mol l−1 EGTA, pH 8.2) on ice. Cells were dissociated by gently pipetting until reaching >95% single cell suspension. Dissociated single cells were washed twice in calcium-free seawater (CFSW) (454 mmol l−1 NaCl, 9 mmol l−1 KCl, 48 mmol l−1 MgSO4, 6 mmol l−1 NaHCO3, pH 8.2) and checked for viability (>95%) by propidium iodine staining and fixed in 90% ice-cold methanol. Fixed cells were rehydrated in CFSW+BSA 0.1% in the presence of the RNase inhibitor (SUPERaseIN, Thermo Fisher Scientific). Rehydrated cells were processed according to 10xGenomics Single cell 3′ version 3 user guide. Sequencing of libraries was performed on an Illumina Hiseq system (Illumina, San Diego, CA, USA), resulting in an average of 23.015 raw reads per cell barcode. Raw reads were demultiplexed, aligned to the S. purpuratus genome (Arshinoff et al., 2022) and counted using the CellRanger pipeline from 10xGenomics (version 3.0.2). Barcodes with fewer than 1000 reads, fewer than 500 expressed genes, or more than 25% mitochondrial reads were removed to only keep high-quality cells. Basic uniform manifold approximation and projection (UMAP) dimensional reduction and clustering was performed using the Seurat version 3.2 workflow (Butler et al., 2018).
Whole-mount in situ hybridization
Whole-mount in situ hybridization was performed according to the protocol described by Walton et al. (2006) with some minor modifications. A hybridization probe of the candidate gene trypsin-2 (LOC753460) was designed and cloned (ID 6, Table 2). Probe DNA was constructed with a pGEM-t-easy-vector (Promega, Madison, WI, USA) and in vitro transcribed into a DIG-labeled probe by T7/SP6 polymerases (Thermo Fisher Scientific). Unfed samples for whole-mount in situ hybridization were taken at day 2, 3 and 4 dpf. Larvae were fixed with 4% paraformaldehyde/FSW, dehydrated with methanol and stored at −20°C until further use. For hybridization, larvae were gradually rehydrated with diethyl pyrocarbonate (DEPC-PBST) at room temperature. Pre-hybridization was carried out for 3 h at 60°C in hybridization buffer. Afterwards, 1 ng µl−1 RNA probe was added and hybridization took place at 65°C overnight. Samples were washed for 10 min using washing buffer (50% formamide, 5× SSC, 0.1% Tween-20) and gradually transferred to 2× SSC buffer as described above at 65°C. In addition, larvae were incubated twice in 2× SSC buffer at 37°C for 10 min and washed twice in 2× SSC buffer at room temperature for 10 min before the 2× SSC buffer was gradually replaced with MABT (500 mmol l−1 maleic acid, 750 mmol l−1 NaCl, NaOH to pH 7.5, 0.1% Tween-20) at room temperature as described above. Larvae were then incubated for 2 h at room temperature in blocking buffer [1× MAB, 2% block reagent (Roche, Basel, Switzerland), 10% sheep serum (Jackson Immunoresearch, West Grove, PA, USA), 0.1% Tween-20]. Larvae were incubated at 4°C overnight in the dark after adding 1:2000 sheep Anti-Digoxigenin AP Fab Fragments (Roche, Basel, Switzerland) antibody. Then, larvae were washed 5 times for 5 min at room temperature in MABT before being incubated in BM Purple AP Substrate (Roche) for 90 min at room temperature. Staining was stopped by washing 5 times for 5 min with MABT. All incubation steps were carried out with gentle shaking.
In vitro protease assay
Total protease activity was determined as previously described with some minor modifications (Stumpp et al., 2013). Briefly, frozen larvae (5000–10,000 individuals per sample) were re-suspended in 550 µl nuclease-free water and homogenized on ice with an ultrasonic cell disruptor (Bandelin Sonoplus HD 2200, Berlin, Germany) 6 times for 15 s with 15 s pause in between. Homogenates were centrifuged for 15 min at 16,000 g at 4°C. Supernatants were transferred into new reaction tubes and kept on ice until further use. The protease assay was conducted using 20 µl crude enzyme extract in 200 µl buffer (Ellis, 1961; 0.1 mol l−1 sodium carbonate, 0.1 mol l−1 2-amino-2-methyl-1,3-propanediol, 0.1 mol l−1 sodium hydrogen phosphate and 0.1 mol l−1 citric acid) and 50 µl 2% (20 mg ml−1) azocasein (Merck, Darmstadt, Germany) as a protease substrate. Incubation was carried out at 37°C while shaking (300 rpm) in duplicates (technical replicates) overnight. Assays were stopped by adding 500 µl 8% ice-cold trichloroacetic acid (TCA) and samples were centrifuged for 10 min at 16,000 g and 4°C. Crude enzyme extract was added to negative controls after TCA addition and before centrifugation. Protease activity was determined by the absorption (A) of the supernatants from the samples and the controls measured in a BioSpectrometer (Eppendorf, Hamburg, Germany) at 366 nm. To investigate pH dependency (experiment 3), buffer with pH values ranging from 4.0 to 12.0 was used (Ellis, 1961). To determine protease activity over larval ontogeny, samples were taken at 2, 4–11 and 13 dpf (experiment 1) and measured with an assay pH of 9.5, resembling the natural pH conditions in the alkaline larval midgut (Stumpp et al., 2015). For in vitro inhibitor assays, 20 µl of either aprotinin (0.0001–10 µmol l−1) or soybean trypsin inhibitor (0.1–100 µg ml−1, both Thermo Fisher Scientific) was added to each sample. Crude enzyme extract was pre-incubated for 15 or 45 min with inhibitor in buffer before adding azocasein to start the reaction. Protein concentration was measured using a Micro BCA Protein Assay Kit (Thermo Fisher Scientific) with bovine serum albumin as the standard according to the instruction manual. Finally, enzyme activity was expressed as the change of absorption per milligram of protein per hour (ΔA mg−1 protein h−1).
In vivo protease assay
To ensure that protease inhibitors do not lose inhibitory capacity in seawater and the gastrointestinal tract of sea urchin larvae, they were both tested in artificial seawater (pH 8.1, salinity 31.5, 16°C) and simulated gastrointestinal tract conditions of sea urchin larvae (pH 9.1, salinity 31.5, 16°C). For this, 20 µl bovine trypsin (1 mg ml−1), was pre-incubated in 200 µl FSW (pH 8.1) or FSW, which was mixed in a 1:4 ratio with H2O at pH 9.1 corresponding to the alkaline and low osmotic conditions found in the larval midgut (Petersen et al., 2021). Bovine trypsin was incubated with 20 µl of either aprotinin (100 µmol l−1) or soybean trypsin inhibitor (1 mg ml−1) for 45 min before 50 µl 2% (20 mg ml−1) azocasein was added to start the reaction. Incubation was carried out at 16°C while shaking (300 rpm) in duplicates (technical replicates) for 2 h. Assays were stopped adding 500 µl 8% ice-cold TCA and samples were centrifuged for 10 min at 16,000 g and 4°C. Trypsin was added to negative controls after TCA addition and before centrifugation. Activity was determined by the absorption of the supernatants from the samples and the controls measured in a BioSpectrometer (Eppendorf) at 366 nm.
To determine the influence of protease inhibitors on the in vivo digestive activity, 5–7 dpf larvae (experiment 5, not fed) were incubated for 3 h in aprotinin (10 and 100 µmol l−1) and soybean trypsin inhibitor (100 µg ml−1 and 1 mg ml−1) at 15°C. Unfed larvae were used in order to achieve the highest possible feeding pressure on the larvae, leading to very quick feeding responses. Even though feeding history may have an influence on feeding rate and enzyme expression levels, fast responses were necessary in the in vivo assays to avoid larvae being in the small water volumes (0.5 ml) of the microscopic set up for prolonged periods and to avoid an interfering fluorescence signal in the midgut tissue caused by previously eaten algae. For in vivo digestion rate measurements, larvae were transferred to a temperature-controlled perfusion chamber (0.5 ml total volume) mounted on a Axiovert 135 inverse microscope (Carl Zeiss). Larvae were then fed with Rhodomonas sp. that were also pre-incubated with corresponding concentrations of inhibitor. After larvae had collected algal cells in their esophagus, they were quickly caught using a micropipette to which a slight vacuum was applied. Larvae were then held in place and after microalgae had been swallowed into the midgut, the fluorescence signal was monitored at 505 nm every 1.5 s over a period of 600 s. Fluorescence signal traces were analyzed using VisiVIEW Imaging Software by subtracting the background fluorescence. In vivo digestion rates were expressed as the decrease in fluorescence over 180 s after a swallowing event. As the raw data differed nominally because of the larval position and the number of swallowed cells, we normalized them by dividing each data point by the maximum signal intensity of the measurement (n=8).
Additionally, to test the effect of serine proteases on Rhodomonas sp., microalgae (1.7×106 cells) were exposed to different concentrations (50, 10, 1 and 0.1 mg ml−1) of bovine trypsin (Merck). Rhodomonas sp. were incubated with the respective trypsin concentration for 1 h at room temperature in FSW, which was mixed in a 1:4 ratio with H2O at pH 9.1 to resemble larval midgut conditions. Cells were examined optically for motility.
Statistical analyses
Figures and statistical analysis were performed using GraphPad Prism 8.4.3. Datasets were tested for normal distribution and homogeneity prior to statistical analyses and results were analyzed using one-way ANOVA followed by Tukey (gene expression and protease ontogeny) or Dunnett (effect of trypsin on Rhodomonas sp. and pharmacological inhibition of in vitro total protease activity) multiple comparisons post hoc tests. For comparison of larval growth, a multiple t-test was conducted (all Table S1).
RESULTS
Larval cultures
Abiotic parameters of all cultures were within a narrow range (Table 1). The salinity ranged from 31.2±0.4 to 31.7±0.1 psu. Temperature ranged from 15.5±0.5 to 15.7±0.2°C and pHNBS ranged from 8.24±0.03 to 8.33±0.01. For experiments 1 and 2, parameters were determined every other day (n=3). For experiments 3 and 4, parameters were determined every day (n=12). Abiotic parameters for experiment 5 samples were measured once when fertilized eggs were put into the cultures (salinity: 31.5 psu; temperature: 15.4°C; pHNBS: 8.31). Culture density did not decrease beyond 69%, indicating low mortality (Fig. S2B,C). Larval size was determined every second day for food conditions (n=28) and at 5, 6 and 7 dpf for no-food conditions (n=8) as described above. Food concentration significantly influenced larval size. At 8 dpf, larvae fed with 8000 cells ml−1 were significantly larger (309±2.4 µm body length) than larvae fed with 2000 cells ml−1 (299±4.6 µm) or 500 cells ml−1 (294±4.6 µm). A significant difference was also present at all later measurements. Significant differences in size were observed between larvae grown under 500 and 2000 cells ml−1 conditions on several days. However, these did not follow a clear trend. (Table S1, Fig. S2A).
Gene screening of digestive proteases in Strongylocentrotus purpuratus larvae
Genes coding for digestive protease of two major classes (endopeptidases: trypsins and enteropeptidases, as well as exopeptidases: carboxypeptidases and aminopeptidases) from mammals were used to identify their homologs in sea urchin larvae. Using published transcriptomic data (Arshinoff et al., 2022) the proteases encoded by trypsin-2 (LOC753460), enteropeptidase (LOC584950), carboxypeptidase B (LOC585090) and aminopeptidase N (LOC578352) were identified to be potential candidates involved in digestive processes in the larvae, as these genes have their peak expression at 72 hpf when the digestive system becomes functional (Fig. S3). Single-cell transcriptomic analyses demonstrated highest expression of trypsin-2 and aminopeptidase N in the sea urchin gut and were most prominent in pancreatic-like cells with a differential expression pattern visible in midgut and foregut cells. While trypsin-2 was strictly localized to pancreatic-like cells, the expression of aminopeptidase N was also detected in esophagus, midgut and hindgut cells as well as primary mesenchyme cells (PMCs) (Fig. S4A,B). enteropeptidase and carboxypeptidase B expression levels were relatively low, with the dominant reads in pancreatic-like cells for carboxypeptidase B and in ectodermal band cells for enteropeptidase (Fig. S4C,D).
RT-qPCR was used to detect the influence of larval age (ontogeny) and feeding status (8000, 2000, 500 Rhodomonas sp. cells ml−1) on expression of the above protease candidate genes. Expression of trypsin-2 increased to values between 0.040±0.014 and 0.033±0.014 relative to EF1a at 6 dpf. Thereafter, expression decreased continuously. Under 8000 cells ml−1 conditions, expression also decreased after 6 dpf, but increased up to 0.043±0.015 relative to EF1a at 10 dpf. The expression of trypsin-2 was significantly (P≤0.0487) higher with 8000 than with 500 Rhodomonas sp. cells ml−1 at 4, 8, 9 and 10 dpf (Fig. 1A; see Table S1 for statistics). The expression of aminopeptidase N increased from 2 to 4 dpf; until 7 dpf, expression was nearly constant at 0.035±0.008 relative to that of EF1a and decreased afterwards. With 500 Rhodomonas sp. cells ml−1, at 5 dpf and 7 dpf, expression was significantly higher than under 8000 cells ml−1 conditions (P≤0.0169) (Fig. 1B; see Table S1 for statistics). The expression of enteropeptidase increased continuously up to 0.051±0.01 relative to that of EF1a at 9 dpf. No significant difference in enteropeptidase expression was detected between feeding conditions (Fig. 1C; see Table S1 for statistics). The expression of carboxypeptidase B increased up to 0.032±0.01 relative to that of EF1a at 5 dpf, and then decreased. There was also no significant difference regarding feeding conditions (Fig. 1D; see Table S1 for statistics).
trypsin-2 proved to be the most specific candidate gene, with expression induced in response to food treatments, suggesting a prominent role in larval digestive processes. Whole-mount in situ hybridization demonstrated that trypsin-2 transcript signal was not detectable in 2 dpf gastrula stages. In 3 and 4 dpf pluteus stages, when larvae started feeding, the transcript signal was detectable in pancreatic-like cells in a ring-shaped arrangement (Fig. 2). No non-specific staining was detected in sense probe controls.
Total protease activity in response to larval age and food treatment
To characterize larval digestive proteases, total protease activity was determined using azocasein as a universal substrate over the course of development (2–13 dpf) and in response to food treatment. During larval development, total protease activity increased up to 4000% from 2 dpf to 7 dpf, irrespective of food treatment. The highest protease activity was reached in the 8000 cells ml−1 treatment, while the activity in cultures with 2000 and 500 Rhodomonas sp. cells ml−1 was lower and did not show a significant difference. From 7 dpf on, protease activity remained nearly constant in all food treatments (Fig. 3A; see Table S1 for statistics).
Characterization of total protease activity: pH dependency profiles and pharmacological identification of serine proteases
Total protease activity increased with increasing assay pH up to a maximum at pH 9.5, reaching values of 0.38±0.06 ΔA mg−1 protein h−1. More alkaline conditions induced a steady decrease in activity with little activity (0.02±0.01 ΔA mg−1 protein h−1) remaining at pH 11.5 (Fig. 3B). Two protease inhibitors (aprotinin and SBTI) were used to specify the role of serine proteases/trypsin in total protease activity. Aprotinin inhibits various intracellular serine proteases, among others plasmin, trypsin and chymotrypsin (Hewlett, 1990). SBTI inhibits mostly trypsin as well as chymotrypsin and plasmin to a lesser extent (Kunitz, 1947). In order to determine the necessary pre-incubation of crude enzyme extract with the inhibitors, we conducted time–response curves. After 45 and 15 min pre-incubation time, aprotinin and SBTI significantly (P≤0.0001) reduced the total protease activity of the crude enzyme extract by 83% and 63%, respectively (Fig. S5A,B). Using a pre-incubation time of 45 min, we conducted dose–response curves of both inhibitors. The inhibitory concentration of 50% maximum inhibition (IC50) was 0.005 µmol l−1 for aprotinin (Fig. 3C) and 0.03 µg ml−1 for SBTI (Fig. 3D).
The initial step of in vivo microalgae digestion is protease independent
To determine the role of trypsin as the dominant protease in the digestive system, aprotinin and SBTI were used in an in vivo digestion assay. To ensure that the inhibitors used were not restricted in their functionality under the prevailing seawater conditions, they were tested in vitro under seawater conditions (FSW, pH 8.1, 16°C) and simulated midgut conditions (diluted FSW, pH 9.1, 16°C) using bovine trypsin as control. Under both conditions examined, aprotinin and SBTI significantly (P≤0.0001) reduced trypsin activity (Fig. S5C). For the in vivo assay, we used the decay of the autofluorescence as an indicator for the digestion of microalgae. In vivo digestion rates were calculated from the decrease in autofluorescence of Rhodomonas sp. cells after larvae swallowed (Fig. 4C,D–K). In vivo digestion rates were constant between 0.0025±0.0009 and 0.0035±0.0009 normalized digestion s−1 at days 5–7 dpf of the experiment (Fig. 5A). To ensure that inhibitors reach sufficiently high concentrations in larval midguts through swallowing, inhibitor concentrations that had maximum inhibition capacity in the dose–response curves were used. In addition, a 10× concentration of the maximum inhibition concentration was applied to rule out the possibility of insufficient inhibitor application into the larval midgut. Both inhibitor concentrations used for larval incubations did not affect larval survival and we did not observe any effect of inhibitor concentration on swimming behavior or gross morphology. Finally, none of the inhibitor concentrations had a significant effect on the digestion rate of feeding larvae (Fig. 4A,B). To verify that trypsin is not directly involved in the initial steps of Rhodomonas sp. cell breakdown, microalgae were incubated with different concentrations of bovine trypsin using assay conditions that mimic the abiotic conditions in larval midgut (e.g. low [Na+] of ∼120 mmol l−1 and pH 9.1). The results demonstrated that pure bovine trypsin did lead to a significant (P≤0.0001) decrease in motility when used in very high concentrations of 50 mg ml−1, but not under lower trypsin concentrations of 10, 1 and 0.1 mg ml−1 (Fig. 5B; see Table S1 for statistics). Interestingly, the pH 9.1 and low [Na+] conditions in the absence of the enzyme decreased Rhodomonas sp. motility significantly (P=0.0459) compared with that at pH 8.1. Although Rhodomonas sp. did not move anymore because of the high pH and high trypsin concentration, disintegration of cells as observed during the course of larval in vivo digestion was not observed in any in vitro treatment.
DISCUSSION
Digestive proteases in sea urchin larvae
Digestive proteases from marine invertebrates are of particular importance to science and promise a wide range of applications in both biotechnological and biomedical fields (Barzkar et al., 2021). In mammalian digestion, various proteases form cascade-like structures in which enteropeptidases and trypsins play a crucial role (Kitamoto et al., 1994). One aim of this work was to characterize a total of four potential digestive proteases in sea urchin larvae. The results of this work demonstrated that gene transcripts of all four proteases investigated, trypsin-2, aminopeptidase N, enteropeptidase and carboxypeptidase B, are present in larvae of S. purpuratus. The expression of the two protease genes enteropeptidase and carboxypeptidase B is not restricted to the midgut, suggesting a minor role of these proteases in larval digestive function. In contrast to these two proteases, those encoded by trypsin-2 and aminopeptidase N seem to be of particular importance for larval digestion, as the expression of both genes was affected by the feeding regime. However, only typsin-2 transcript abundance was significantly increased in response to 8000 Rhodomonas sp. cells ml−1 conditions in comparison to 2000 or 500 cells ml−1. Given that the larval density of up to 15 larvae ml−1 in the experiment was rather high, there is the possibility that larvae in the 500 and 2000 cells ml−1 treatments were food limited, leading to the almost complete lack of significant differences in gene expression and protease activity between these two conditions. However, in invertebrates, the influence of food intake on trypsin gene expression has been described in several species. Although it was demonstrated that digestive enzymes are constitutively transcribed and the level of transcription does not drop to zero (Lehane et al., 1998) in both red palm weevil and mosquitoes, transcription of trypsin genes was directly influenced by both the quality and quantity of the ingested food (Alarcón et al., 2002; Noriega et al., 1994). Contrary to expectations, for marine invertebrates it has been demonstrated that a short period of starvation can temporarily enhance not only trypsin expression but also the activity of its product (Muhlia-Almazán and García-Carreño, 2002; Sánchez-Paz et al., 2003). This could also apply to the expression of aminopeptidase N demonstrated in this work, as gene expression levels between 5 and 7 dpf under 500 Rhodomonas sp. cells ml−1 conditions were above those under 8000 cells ml−1 conditions. As the expression of aminopeptidase N was not food inducible, no definite conclusions can be drawn from this work about the influence of aminopeptidase N on the digestion rate. We hypothesize that aminopeptidases catalyze several functions in peptide metabolism in sea urchin larvae. This is in agreement with the findings from other marine invertebrates (Alonso et al., 2020). For enteropeptidase and carboxypeptidase B, the lack of food-induced expression could be due to the rather weak expression in gut-related cells compared with the strong expression in the ecto-ciliary band masking potential minor changes in expression. In mammals, carboxypeptidases are involved in a wide range of reactions from intracellular hormone processing, vesicle transport and secretion to bone remodeling and reproduction (Cawley et al., 2012). We cannot rule out that this is not also the case for carboxypeptidase B in sea urchin larvae. The fact that gene expression of enteropeptidase in this work was not food inducible could be explained by the fact that enteropeptidases have only been described in vertebrates to date (Pavlov, 1910). Regarding enzyme activation, we hypothesize that in sea urchin larvae, trypsinogen is expressed by pancreatic-like cells and is activated by primarily pre-existing trypsin. As has already been shown, trypsinogen also exhibits autocatalytic properties, but at a much lower rate compared with trypsin (Kay and Kassell, 1971). The fact that sea urchin larvae have a much less compartmented digestive tract compared with that of vertebrates supports this theory and would underline a fundamental difference in the digestive systems of vertebrates and sea urchin larvae.
Potential regulation of total protease activity by midgut pH
The results of this work demonstrated that independently from the supply of food, total protease activity increases distinctly between 3 and 6 dpf in sea urchin larvae. Despite high variability, the protease activity of larvae grown under 8000 Rhodomonas sp. cells ml−1 conditions was significantly higher at 9 and 10 dpf than that under the 2000 and 500 cells ml−1 conditions. Again, the high larval density may contribute to the food limitation in the two low-food treatments, potentially leading to these effects. The general increase during larval development is in line with results already described for other marine invertebrates (Saborowski et al., 2006) and is paralleled by the functional development of the gastrointestinal tract. After 7 dpf, when the larvae enter an obligatory feeding state, total protease activity remained stable under all diet conditions, indicating that a functional steady state is reached. Similar observations have been made for other marine organisms (Chong et al., 2002), demonstrating that total protease activity does not increase linearly during larval development, but rather reaches a plateau once the larval digestive system is fully functional. The finding that the maximum total protease activity is at pH 9.5 and the mean pH of the S. purpuratus midgut is pH 9.1 suggests that midgut pH could serve as a regulatory vehicle for protease activity (Stumpp et al., 2015). Although information on the capacity for extracellular pH change in the gastrointestinal tract with respect to enzyme regulation in invertebrates is scarce, findings on intracellular pH change may provide insight into possible mechanisms (Schulz and Münzel, 2011). As it has been discovered in both vertebrates (Deguara et al., 2003) and invertebrates (Bärlocher and Porter, 1986) that the pH optima of digestive enzymes are not equal to the pH of the surrounding environment, it is tempting to speculate that sea urchin larvae control the activity of gastric enzymes by creating specific pH environments within their digestive tract. Given that sea urchin larvae have a digestive system consisting primarily of two compartments (e.g. midgut and hindgut), they could serve as a powerful model system to study the regulation of digestive enzyme activity by the modulation of gut pH. The fact that sea urchin larvae with pharmacologically suppressed midgut pH regulation were associated with a higher susceptibility to bacterial infections (Stumpp et al., 2020) further supports the hypothesis that regulation of midgut pH could have more implications for larval physiology than just facilitating food digestion. Thus, understanding the mechanisms by which sea urchin larvae control the abiotic conditions in specific gut segments could provide important insights into not only larval digestion but also a largely unexplored physiological system.
Trypsin-2 as a major digestive protease in sea urchin larvae
In accordance with earlier studies (Perillo et al., 2016), we confirm that in sea urchin larvae, trypsin-2 is expressed in pancreatic-like cells in the upper part of the midgut. Through comparisons between the gene expression pattern and the total protease activity, the results of this work allow us to suggest that trypsin-2 serves as one of the main digestive proteases in S. purpuratus, just as in other terrestrial and marine invertebrates (Muhlia-Almazán et al., 2008). At 3 dpf, total protease activity increased distinctly and the expression of trypsin-2 also increased. Both were food inducible in that the 8000 Rhodomonas sp. cells ml−1 condition led to higher total protease activity and stronger trypsin-2 expression compared with the 500 and 2000 cells ml−1 conditions, which were not distinguishable. However, trypsin cannot be the only serine protease in sea urchin larvae, because the in vitro experiments performed in this work with the serine protease inhibitor aprotinin had a stronger inhibitory effect than SBTI, which primarily inhibits trypsins. In addition to enteropeptidase, which, as a serine protease, if present in the sea urchin gut may also have been inhibited by the use of aprotinin, chymotrypsin is another potential candidate enzyme, which has a prominent role in the digestion of vertebrates (Blow, 1976). However, it had no distinct expression based on single-cell transcriptomic data of 3 dpf sea urchin larvae and no expression of the gene was detectable up to this time (Arshinoff et al., 2022). Protease inhibitors have been used to characterize the digestion physiology of insects (Jongsma and Bolter, 1997) and fish (Alarcón et al., 1999), but findings from studies on sea urchins are scarce. For decades, (serine) protease inhibitors have been used primarily to elucidate the mechanisms that prevent polyspermy in a variety of sea urchin species (Harding, 1951; Wicklund, 1954; Hagström, 1956). This work represents one of few exceptions in which protease inhibitors have been used to study the digestive physiology of sea urchin larvae.
We further aimed at testing the role of trypsin in larval in vivo digestion of Rhodomonas sp. cells. As it has already been demonstrated that in vivo digestion is prone to environmental change, leading to reduced in vivo digestion rates when midgut pH is reduced (Stumpp et al., 2013), we expected that this observed correlation between pH-dependent in vivo digestion rates and pH-dependent total protease activity is probably due to the function of proteases in the gut. This assumption was supported by findings from Brett et al. (1994) suggesting that the inner and surface periplast of cryptomonad algae (i.e. Rhodomonas sp.) include a proteinaceous component, as demonstrated by trypsin digestion of the periplast components. Although this work showed that total protease activity can be significantly reduced in vitro by the addition of inhibitors, no effect on the digestion of microalgae could be observed in vivo. Thus, we concluded that either trypsin plays no role in the initial step of Rhodomonas sp. digestion or the inhibitors used were ineffective in the in vivo system. To exclude the latter we used 10 times the maximum inhibitory dose of the inhibitors, an incubation time of 3 h and the pre-incubation of Rhodomonas sp. in the corresponding inhibitors. Accordingly, we are confident that the inhibitors would have reached the larval midgut in sufficient concentrations in the in vivo experiments for the following reasons: even if the larvae had not taken up any inhibitor at all during the 3 h incubation and assuming that the bolus swallowed corresponds to only 1% of the midgut volume, this would still leave concentrations of 1 µmol l−1 aprotinin and 10 µg ml−1 SBTI, which both caused a distinct reduction in total protease activity in the in vitro dose–response curves performed. As we could demonstrate for aprotinin that a reduction in total protease activity of about 40% occurs instantaneously, we assume that the lack of reduction in in vivo digestion rate is not due to a lack of inhibitor abundance. Therefore, we assume that the lack of an effect on in vivo algal breakdown is due to the trypsin independence of algal breakdown and the initial degradation of fluorescent-active components during larval digestion (summarized in Fig. 6). Given that similar digestion rates were determined at different time points and in different animals, we assume that our method to investigate larval digestion is robust. To further verify the direct impact of trypsin on the breakdown of Rhodomonas sp., we exposed microalgae to different enzyme concentrations under conditions comparable to those in the larval midgut. Interestingly, the abiotic conditions alone, including ∼120 mmol l−1 Na+ and an alkaline pH of 9.1, resulted in significantly decreased motility compared with standard oceanic conditions of 450 mmol l−1 Na+ and pH 8.1. Incubation with pure porcine trypsin under the same conditions had no significant effect on the microalgae when applied in physiologically relevant concentrations that ranged from 50 to 850 µg ml−1 (Borgström et al., 1957). However, motility significantly decreased in the presence of trypsin concentrations that were much higher (i.e. 50 mg ml−1). Furthermore, in these experiments, we could not observe any sign of breakdown of the microalgae by trypsin, supporting the in vivo digestion results that the rapid disintegration of microalgae in the larval digestive system may be attributed to other components or digestive enzymes and/or abiotic midgut conditions.
Conclusion
Digestion is a key physiological process for heterotrophic organism survival and is the foundation for nutrient and energy uptake sustaining organismic fitness and life. This work is one of the first studies focusing on the role of digestive proteases, and in particular trypsin, in food breakdown of the sea urchin larva. Sea urchin larvae and their digestive systems were demonstrated to be susceptible to ocean change such as ocean acidification because of a delicate balance between energy supply through digestive processes and energy consumption due to homeostatic regulation of the midgut. While it is obvious that the digestive system is the major organ in marine larvae that controls the animal's energy budget, it remained unknown which digestive enzymes are particularly involved in digestive function. The present study highlights the importance of trypsin in digestion-related processes but, interestingly, not in the initial breakdown of microalgae in vivo. This stands in contrast to the impacts observed by simulated ocean acidification itself: initially, Stumpp et al. (2013) proposed that low midgut pH induced by ocean acidification reduces digestion rate through decreasing proteolytic activity (pH-dependent activity) of digestive proteases in the larval midgut. As inhibitors specifically targeting proteolytic activity in this study did not induce the same response, i.e. a reduction in digestive efficiency, elucidation of the true mechanisms by which ocean acidification decreases observed digestion rates, or more specifically algal breakdown, is an important task for future research. This research track may be able to disentangle the apparently more complex interactions of environmental change and digestive processes. Importantly, it remains to be investigated which other digestive enzymes (lipases, hydrolases, etc.) are displaying similar patterns in response to pH and may account for the rapid initial digestion of ingested microalgae in echinoderm larval stages.
Acknowledgements
We thank D. Garfield and J. Brandenburg (Integrative Research Institute for the Life Sciences, Humboldt-Universität zu Berlin, Germany) for providing the single-cell transcriptome database, F. Thoben (Christian-Albrechts-Universität zu Kiel, Germany) for maintenance of the sea urchin culture system and H. Ließegang (Christian-Albrechts-Universität zu Kiel, Germany) for culturing Rhodomonas species.
Footnotes
Author contributions
Conceptualization: J.H., M.Y.H., M.S.; Methodology: J.H., W.W.C., M.S.; Validation: J.H., M.Y.H., M.S.; Formal analysis: J.H.; Investigation: J.H., W.W.C., M.Y.H., M.S.; Resources: M.S.; Data curation: J.H.; Writing - original draft: J.H.; Writing - review & editing: J.H., M.Y.H., M.S.; Visualization: J.H.; Supervision: M.S.; Funding acquisition: M.S.
Funding
M.Y.H. and W.W.C. were funded by the Emmy Noether Program (403529967) of the Deutsche Forschungsgemeinschaft. M.S. and J.H. were funded by the Emmy Noether Program (441084746) of the Deutsche Forschungsgemeinschaft.
Data availability
Data are available from the authors upon request.
References
Competing interests
The authors declare no competing or financial interests.