Cathepsin D in prawn reproductive system: its localization and function in actin degradation

Prawns and sperm collection

Three-month old, blue-claw giant freshwater prawns with an average weight of 125 g were purchased from the Central Market for the Promotion of Agriculture in Ayutthaya province, Thailand. The acquired prawns were transported and reared prior to sample extraction according to the guidelines of the National Research Council of Thailand, the National Aquaculture Council of Australia (Johnston & Jungalwalla, 2005), Seafish UK (Jacklin & Combes, 2007) and the Animal Care Committee, Mahidol University, Thailand (MUSC-IACUC, protocol # 2016/014). Immediately prior to sperm and tissue extraction, all prawns were anesthetized in ice, and sperm collection and their processes were followed the method described by Watthammawut et al. (2015) and Surinlert et al. (2016). Briefly, the testes and vas deferens were carefully dissected out from the prawn specimens and minced with a razor blade to obtain one mm-sized pieces. To release sperm, the tissues were agitated in phosphate-buffered saline (PBS) on an orbital shaker and then the suspension filtered through a 45-µm metal sieve (Endercotts, London, UK). The sperm samples were concentrated, repeatedly washed (500 ×g, 4 °C, 5 min), resuspended in PBS and placed on ice for use in subsequent experiments. For consistency in all following experiments, sperm were counted with a hemacytometer to obtain an experimental sample each with a concentration of  1 × 10⁷ cells/ml.

Sequence comparison of CAT-D and phylogenic tree analysis

Since full-length of MrCAT-D was already deposited in the Genbank database (accession number KP262355.1), we thus conducted amino acid sequence comparison with the other 10 animal species using a Clustal Omega software (Thompson, Higgins & Gibson, 1994). The MrCAT-D amino acid sequence was submitted to predict the signal sequence by SignalP 4.1 server (http://www.cbs.dtu.dk/services/SignalP-4.1/) (Petersen et al., 2011) and N-linked glycosylation by NetNGlyc 1.0 server (Technical University of Denmark, Lyngby, Denmark). CAT-D amino acid sequences were obtained from NCBI through these following accession numbers: Homo sapiens (NP_001900.1), Bos taurus (NP_001159993.1), Canis familiaris (NP_001020792.1), Xenopus laevis (BAC57431.1), Salmo salar (ACH70630.1), Pinctada maxima (AEI58896.1), Bombus impatiens (XP_003489428.1), Homarus americanus (ACV53024.1), Penaeus vannamei (ROT71313.1), Penaeus japonicus (AIF27797.1), Macrobrachium rosenbergii (AMQ98967.1). The phylogenetic tree was performed using a Maximum likelihood method in a Molecular Evolutionary Genetic Analysis (MEGA) software version 7.0. The reliability was based on 1,000 replications where the percent bootstrap confidence value was indicated at each branch node (Kumar, Stecher & Tamura, 2016).

Reverse transcriptase polymerase chain reaction (RT-PCR)

Primers used for both RT-PCR and in situ hybridization were designed from the existing MrCAT-D full sequence (Table 1 and Fig. S1). Various tissues used for RNA extraction including testis (Tes), vas deferens (Vas), terminal ampoule (Ter), Muscle (Mus), hepatopancreas (Hep), stomach (Sto), heart (Hea), gill (Gil), testicular sperm (Tsp), vas deferens sperm (Vsp) and spermatophore (Ssp) were placed in the tubes and immediately homogenized in Trizol reagent (Invitrogen, Carlsbad, CA) for RNA extraction following the manufacturer’s protocol (Direct Zol-RNA miniprep, Zymoresearch, CA). Briefly, tissue homogenate was centrifuged and the supernatant was transferred into the RNase-free tube. Equal volume of ethanol (95–100%) was added and gently mixed and transferred to the column. The DNA digestion buffer which contained DNase-I was added and incubated (30 min). The obtaining RNA was washed with washing buffer and loaded into the column. The RNA was eluted from the column by DNase/RNase-free water and its concentration was measured by Nano drop-2000C spectrophotometer (Thermo Scientific, Massachusetts). Thereafter, 2 µg of total RNA was converted to first-strand cDNA by a RevertAid First Strand cDNA synthesis kit (ThermoFisher Scientific Inc. Massachusetts) following the manufacturer’s protocol using oligo-dT-adaptor primer and Multi Serobe™ reverse transcriptase (200 U/ µL). The mixtures were amplified at 25 °C for 5 min, 42 °C for 60 min, 70 °C for 5 min and 4 °C, respectively. To amplify target DNA sequence, the first-stranded cDNA was synthesized from mRNA isolated from prawn tissues and used as the template for PCR amplification. The oligonucleotide primers (Pacific Science, Bangkok, Thailand) used in this study were shown in Table 1. β-actin of M. rosenbergii was used as reference gene (Priyadarshi et al., 2015). These primer sequences were designed and selected by using NCBI/primer blast and Clustal Omega/multiple sequence alignment. The PCR reaction was initiated with a cycle of denaturation at 94 °C (5 min), followed by 35 cycles of denaturation at 94 °C (30 s), annealing at 57 °C (30 s) and extension at 72 °C (30 s) and further extension at 72 °C (10 min). The PCR products (predicted product length = 323 bp) were checked by 1.2% agarose electrophoresis and stained with ethidium bromide followed by visualization under a GelDoc 1,000 (BioRad, Hercules, CA). The relative density of these PCR bands from the triplicated experiments were calculated by ImageJ software and plotted as mean band density. Occasionally, these PCR products were further verified their sequences by Macrogen (Seoul, Republic of Korea).

Immunohistochemistry with anti-CAT-D and phalloidin

The paraffin sections of testes collected from mature blue claw male giant freshwater prawns (aged >6 months, weighed 100-150 gm) were deparaffinized with series of xylene and rehydrated in a decreasing grade ethanol (100% to 70%). The sections were treated with 1% hydrogen peroxide (H₂O₂) to quench endogenous peroxidase and with 4% BSA in 0.1 M glycine in PBS to block free aldehyde and non-specific antibody binding. The tissues were further treated 0.1 M citrate buffer, pH 6.0 at 60 °C and 0.5% Triton-X 100 in PBS for antigen retrieval and tissue permeabilization, respectively. The sections were exposed to rabbit anti-human CAT-D polyclonal antibody (EMD Merck, Darmstadt, Germany) at a dilution of 1:200 (4 °C, overnight) followed by exposing the sections to horseradish peroxidase (HRP) conjugated goat anti-rabbit IgG at dilution of 1:400 (4 °C, 1 hr). It should be noted that the sequence similarity between heavy chain of human CAT-D (used as antigen to generate antibody) and MrCAT-D is about 58% (Suppl Fig. 2). Enzymatic reaction was developed using 3, 3′- diaminobenzidine (DAB) and the reaction was stopped by an excessive rinsing in the distilled water. The sections were mounted and visualized under an Olympus BX53 microscope.

To staining the filamentous actin, 5-µm thick-paraffin sections of M. rosenbergii testes were processed under similar conditions mentioned above. The sections were permeabilized (room temperature, 10 min) with 0.5% TritonX-100 in PBS and blocked non-specific staining with 4% BSA. They were then incubated with Alexa 594 conjugated phalloidin (ThermoFisher Scientific, Waltham, MA) in 1% BSA at the dilution of 1:200 (30 min, room temperature). Fluorescent signals were acquired by an Olympus FV10i confocal microscope.

Immunofluorescent staining of isolated sperm

For loose sperm staining, the sperm were processed either live or fixed with 2% paraformaldehyde (30 min, room temperature). After washing in PBS, the cells were treated with 2% BSA to block non-specific antibody binding and incubated (4 °C, 1 hr) with 1:200 rabbit anti-CAT-D polyclonal antibody and centrifuged (500 × g, 5 min) to wash away the unbound antibody. The sperm were further incubated (4 °C, 30 min) with 1:400 goat anti-rabbit IgG conjugated with Alexa-488 (Abcam, Cambridge, UK). The cells were washed twice with PBS (500 × g, 5 min), mounted with 50% glycerol (v/v) and visualized by an Olympus FV10i confocal microscope.

In situ hybridization

The specific-DNA probe for Macrobrachium rosenbergii CAT-D was constructed by PCR-DIG labelling kit (Promega, Madison, WI). The reaction mixture contained 1 × PCR buffer, 1 × dNTP mixture (DIG-dUTP), Taq DNA polymerase and CAT-D forward and reverse primers (Table 1). The probe concentration was measured by Nanodrop-200C spectrophotometer and analyzed with 1.5% agarose gel. Testis sections were dewaxed, rehydrated in graded series ethanol (100%–70%) and incubated with TNE buffer (50 mM Tris–HCl, 10 mM NaCl, 1 mM EDTA, pH 7.4) containing 10 µg/ml RNase-free proteinase K (37 °C, 15 min). After treating with DEPC-PBS containing 4% paraformaldehyde, the sections were incubated (37 °C, 2 hr) with prehybridization buffer (4 × SSC: 600 mM NaCl, 60 mM Na-Citrate in DEPC-H₂O, 50% deionized formamide). They were then treated with hybridization buffer (50% deionized formamide, 50% dextran sulfate, 50 ×Denhardt’s solution, 20 × SSC, 10 mg/ml denatured salmon sperm DNA) containing 100 ng of DIG-labeled DNA probe (42 °C, overnight, humid chamber). The probes for MrCAT-D were designed from primers CatD-F: 5′-CCT GTT TTC TAC AAT ATG GTT A -3′ and CatD-R: 5′- GGC TTA GCA CCA ATC TTC TTG TTG-3′ (Table 1) with the expected probe length of about 320 nucleotides. Negative control sections were incubated without DIG labelled MrCAT-D probe and processed under the same conditions as described above. The slides were washed and blocked a non-specific staining with 10% normal goat serum in PBS and incubated (2 hr, room temperature) with the anti-DIG antibody conjugated with alkaline phosphatase (1:500) in a humid chamber. Enzymatic products were developed with NBT/BCIP substrate (Promega, Madison, WI) and the reaction was stopped with TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.1). The sections were mounted with 50% glycerol in PBS (v/v), observed and photographed under an Olympus BX53 microscope.

Fluorogenic enzyme assay of MrCAT-D and its hydrolysis of filamentous actin

We tested the function of MrCAT-D in hydrolyzing the specific fluorogenic CAT-D substrate and filamentous actin extracted from SH-SY5Y neuronal stem cells (a kind gift from Assoc. Prof. Permphan Dharmasaroja, NeuroScience laboratory, Department of Anatomy, Faculty of Science, Mahidol University). MrCAT-D enzymatic activity was measured from testicular extracts (TE) and Tsp lysates using a fluorogenic enzyme assays. CAT-D specific substrate was fluorogenic-4-methylcoumarin-7-amide (MCA)-Gly-Lys-Pro-Ile-Leu-Phe-Phe-Arg-Leu-Lys (Dnp)-D-Arg-NH₂) (Enzo Life Sciences, Lausen, Switzerland). The reaction mixtures consisted of 10 µl (10 µg) of protein samples, 80 µl cathepsin assay buffer (0.1 M NaoAc, 0.2 NaCl, pH 3.5) and 10 µl of 20 µM fluorogenic substrates. The activities were monitored at 5-60 min at the excitation and emission wavelengths of 340 nm and 420 nm, using a fluorescent microplate reader (Tecan Spark, Männedorf, Switzerland). The enzyme activity was calculated based on AMC standard curve. One unit of enzyme activity was defined as micromoles of AMC released per min at 25 °C. The specific activity of enzyme was defined as units per mg protein. Inhibition assay of the enzyme was performed using 1 mM of CAT-D inhibitor, pepstatin A (a competitive type inhibitor that is selective for aspartic proteases; EMD Merck, Darmstadt, Germany). The inhibitor was pre-incubated with the protein samples for 30 min at room temperature before performing the enzyme assay. The remaining enzyme activities were measured in the same conditions as mentioned above.

To determine the function of MrCAT-D in hydrolyzing filamentous actin, 10 µg of SH-SY5Y proteins were incubated with 1) 100 µg of Tsp lysates, 2) 100 µg of TE extracts, 3) 0.2 µg of purified human liver CAT- D and 4) samples 1 and 2 pre-incubated 30 min with 1mM pepstatin-A inhibitor prior to the reaction. The mixture of reaction was allowed to stand at 37 °C (24 hr) in an incubation buffer (50 mM sodium acetate buffer, pH 5.5 containing 12.5 mM NaCl, 1.5 mM dithiothreitol and 7.7 mM sodium azide). The treated samples were subjected to 12.5% SDS-PAGE and either stained with Coomassie blue staining (for protein profiling) or Western blotting. Briefly, the membrane was treated with 5% skimmed milk and probed (2 hr, room temperature) with 1 µg/ml anti-actin or anti-GAPDH monoclonal antibody (Cell Signaling Inc., Beavery, MA) followed by HRP-conjugated corresponding secondary antibody (1:2,000 dilution, 1 hr, room temperature). The antigen-antibody complex was visualized by an enhanced chemiluminescent kit (Amersham Pharmacia, Buckinghamshire, UK). Analysis of the band densities in all samples was performed in the same pixel areas (490 ×765 pixels) using an ImageJ software (http://www.imagej.nih.gov/ij).

Sequence analysis of MrCAT-D and its phylogenetic analysis

As mentioned above that full-length of MrCAT-D sequence was already available in Genbank database (accession number KP262355.1), therefore, only sequence comparison with 10 different species, from crustaceans to mammals, was reported herein (Fig. 1). Generally, MrCAT-D consisted of 1,361 nucleotides which was translated into 385 deduced amino acid residues. The first 15 amino acids (1–15) at N-terminal sequence represented putative signal peptide (italicized, bold, red letters) and there was a cleavage site of putative pro-peptide (17–37, underlined) that made up of a single chain of pro-CAT-D protein, known to be an intermediate form of CAT-D. Following proteolytic processing, two cleavable chains of MrCAT-D was expected including an expected light chain with the amino acids between Trp46-Tyr116 (71 amino acids, calculated mass of 8.9 kDa) and expected heavy chain between Asp118-Ala385 (268 amino acids, calculated mass of 28.4 kDa). The putative catalytic motif was localized at Asp82, Thr83 and Gly84 (empty boxed amino acids). MrCAT-D had six cysteine residues to form three putative disulfide bonds which were Cys94-Cys102, Cys261-Cys265 and Cys304-Cys341 (gray box). Finally, the enzyme had one N-glycosylation site which was predicted at Asn119 (bold).

Phylogenic tree analysis based on a Maximum Likelihood method was used to compared CAT-D amino acid sequences from different species (Fig. 2). MrCAT-D showed high amino acid similarity to P. japonicus and H. americanus with 88% and 87% identities, respectively. The neighbor-joining tree of CAT-D sequences showed the four distinct subgroups: (1) vertebrate CAT-D, (2) insecta CAT-D, (3) mollusca CAT-D and (4) crustacean CAT-D. The result clearly revealed high conservation of CAT-D within three prawn/shrimp species, namely, they clustered within a single clade of the phylogenetic tree. It should be noted that CAT-D shared its ancestral origin towards many mammals including humans (H. sapiens) where the similarity of 57.58% was noted.

Distribution of MrCAT-D gene in prawn tissues

RT-PCR amplification was performed to show distribution of MrCAT-D in various sources of tissues, including reproductive related tissues - testis, vas deferens, terminal ampule, and other somatic tissues - muscle, hepatopancreas, stomach, heart and gill. The results in Figs. 3A and 3B revealed an intense, single PCR band at the expected 323 bp in all tissues studied, particularly those of reproductive related tissues (lanes 1–3) with the band density of about 2.0–2. 7 × 10⁴ arbitrary units. Of particular interest, we further confirmed that MrCAT-D transcripts were also present in sperm cells collected from testis (Tsp), vas deferens (Vsp) and spermatophore sperm (Ssp) (Figs. 3C, 3D) with the band density of about 1.7–2. 7 × 10³ arbitrary units suggesting self-biogenesis of MrCAT-D transcripts by developing germ cells in the testis which was rather different from mouse and human cases reported earlier (Saewu et al., 2012; Asuvapongpatana et al., 2013). The single 212-bp β-actin band, as internal controls, was evenly distributed in all tissues studied confirming an equal amount of RNA loaded in each tissue.

Localization of MrCAT-D transcript and protein in prawn testis

Using in situ hybridization, the transcripts of MrCAT-D were detected using the specific DIG-labelling oligonucleotide probes designed from MrCAT-D sequence. Specifically, an intense staining of MrCAT-D probe was detected within the supporting cells (Sc, elongated nuclei) and spermatogonia (Sg, round nuclei) (Ruiz et al., 2020) that lied against basal laminar of testicular tubules (Figs. 4B and 4C). Developing germ cells in the adluminal compartment were also intensely labelled with MrCAT-D probe, suggesting that both the somatic epithelial cells and germ cells actively synthesized MrCAT-D mRNA. Negative control of testicular section showed only a background staining (Fig. 4A).

To confirm the expression of MrCAT-D at the translational level, we used anti-CAT-D polyclonal antibody to visualize CAT-D protein expression in the prawn tissues. The markedly intense anti-CAT-D reactivity (as a dark brown enzymatic product) was found in the supporting cells (Sc) lining testicular tubules, clump of spermatogonia (Sg) and in Tsp cells with different degrees of staining (Figs. 4E and 4F). To further verify CAT-D localization in the sperm cells, we carried out indirect immunofluorescence (IIF) of live sperm (whose membranes remain intact and impermeable to antibody) and aldehyde fixed sperm (highly permeable to antibody). The IIF results (Figs. 4H and 4K) clearly demonstrated an intense staining of anti-CAT-D at the anterior spike (S) and at the crescent base (*) over the nuclei either in live or fixed sperm, confirming immunohistochemical results mentioned above (Figs. 4E and 4F). Negative controls of both live and fixed isolated cell staining exhibited minimal staining of antibody (Figs. 4I and 4L). The results thus suggest that CAT-D is a sperm inherent enzyme that is targeted to the membrane upon its biosynthesis during germ cell development.

Cathepsin activity in testicular extracts and sperm and their hydrolysis over filamentous actin

We further checked MrCAT-D enzyme activity in Tsp and TE extracts using the specific fluorogenic CAT-D substrate (MCA-Gly-Lys-Pro-Ile-Leu-Phe-Phe-Arg-Leu-Lys (Dnp)-D-Arg-NH₂) with or without the presence of 1 mM pepstatin, a competitive aspartic enzyme inhibitor. The results revealed a considerable amount of the emitted AMC fluorescent intensity at many time points studied reflecting CAT-D activities in TE and Tsp extracts (Fig. 5A). The specific enzyme activity (determined when the substrate used was >5 fold of the K_m of CAT-D (Yasuda et al., 1999)) was about 57.8 ± 2.2 and 34.9 ± 5.8 units/mg protein, respectively (Fig. 5B). These activities could be drastically abolished (<4.00 units/mg protein) when pepstatin was present in the incubates, verifying the existence of MrCAT-D in the epithelial cells of testicular tubules as well as sperm cells.

To further tested the function of MrCAT-D, we exploited protein extracts from SH-SY5Y neuronal cells (due to their enrichment in cytoskeletal proteins including actin filaments) and further exposed them to either TE or Tsp extracts or purified human CAT-D (positive control). As shown in Fig. 5C, there was a minimal change in the protein patterns of all sample studied, even when the protein extracts were exposed to purified CAT-D enzyme (engaging a rather strong hydrolytic activity). Therefore, digestion of filamentous actin was further checked by the reactivity of anti-actin antibody with the filamentous actin left after MrCAT-D hydrolysis (Fig. 5D). It should be first noted that monoclonal anti-actin that we used was rather specific to mammalian (SH-SY5Y) actin while it showed no reactivity with shrimp actin that we tested at all (Fig. 5D, lanes TE and Tsp). The intensity of an immunoreactive band at 42 kDa (a molecular mass of filamentous actin) was markedly reduced upon hydrolyzed by TE or Tsp extracts in the similar trend as that of purified human CAT-D, although not as much as the latter group. Upon pre-treating Tsp, TE extracts and purified CAT-D with pepstatin A, a higher intensity of the 42-kDa anti-actin reactive bands was seen (lanes +Pep-A), suggesting a less digestion of filamentous actin by MrCAT-D in the tissues under the influence of its inhibitor. Densitometric analysis further confirmed reduction of anti-actin reactive band intensities in SH-SY5Y samples treated with TE, Tsp and purified CAT-D compared with either those inhibited by pepstatin-A or non-treated SH-SY5Y protein. The densities ranged from 1.8-2. 0 ×10⁴ arbitrary units in those treated samples where they were 2.6-3. 5 ×10⁴ arbitrary units in pepstatin-A inhibited samples and 3. 9 ×10⁴ arbitrary units in non-treated SH-SY5Y sample (Fig. 5D, lower panel). It should also be noted that the intensity of the 42-kDa band inhibition by pepstatin-A (Fig. 5D, lanes 2, 4, 6) was relatively lower than the starting concentration of the actin band (lane 7), thus a complete inhibition could not be claimed.

Presence of actin filamentous meshwork in between sperm and supporting cells

We investigated whether actin meshwork existed in the junction between apical aspect of supporting cells and testicular sperm, by probing with phalloidin (a well-known marker for filamentous actin). The results in Fig. 6 showed an intense phalloidin staining as the patch-like structure in between spermatogonia, supporting cells and testicular sperm (Tsp; panel C, arrow), suggesting the presence of filamentous actin meshwork in this area. In addition, the intense staining patches of phalloidin disappeared at the sites where Tsp were detached from supporting epithelium and released into the lumen, leaving only a faint actin staining within the cells (panel D, arrowhead). This result thus provided a good clue on the presence of actin-rich meshwork at the junction between supporting and germ cells (similar to mammalian ES-like complex) (Cheng & Mruk, 2011; Pelletier, 2011). Interestingly, this result also suggested that actin meshwork must be disrupted or disappeared which should be in conjunction with the sperm release into the lumen of testicular tubules. As mentioned, presence of actin as the faint staining rings within epithelial cells and Tsp was also noted (Figs. 6C and 6D), which should represent basic cytoskeletal element that supports cellular architecture.