The possibility of direct integumental absorption of the amino acid glycine from a solution in seawater was investigated in 250–260 day old (16.9–50.0 mg wet mass) phyllosoma larvae of the Japanese spiny lobster Panulirus japonicus Von Siebold 1824. The uptake of the amino acid was assessed by autoradiography and liquid scintillation counting (LSC) of larvae incubated with [2-3H]glycine and the net uptake was estimated by a time course high-performance liquid chromatography (HPLC) analysis of the concentration of glycine in the incubation medium. Autoradiography revealed the presence of labelled glycine in the cuticle, epidermis and internal tissues (digestive system, muscle, haemocytes) within 30 min of the onset of incubation. Absorption through the integument was confirmed by autoradiography and LSC as glycine uptake was observed even in larvae whose mouths were artificially sealed with cyanoacrylate bond prior to incubation. Scanning electron microscopic examination of the body surface revealed no bacterial population that could have mediated the uptake. HPLC revealed a consistent net uptake (0.29–0.39 μmol g−1 body mass h−1) of glycine in larvae incubated in 6 μmol l−1 glycine and high individual variation (e.g. absorption or release) in larvae incubated at higher concentrations (30 and 60 μmol l−1). Thus, the results of this study provide clear confirmation that, in addition to the known mode of oral feeding on macroscopic food masses, P. japonicus phyllosoma larvae are also able to absorb nutrients directly from the surrounding medium.
Utilization of particulate organic mater (POM) and dissolved organic mater (DOM) has been reported in adults and larvae of a range of marine invertebrates (for reviews, see Stephens and Schinske, 1961; Stephens, 1981; Manahan, 1983; Preston, 1993; Ben-David-Zaslow and Benayahu, 2000; Baines et al., 2005; Grover et al., 2008) and fish larvae (Otake et al., 1993; Otake et al., 1995). In many marine invertebrates, the external body surface is considered to be a very important route and, in some cases, the primary route for the absorption of exogenous amino acids (Preston, 1993). However, the capacity to absorb dissolved substances directly from the surrounding water has never been demonstrated in any species of arthropods and the impermeable exoskeleton is considered to be the major impediment to the development of this capacity (Stephens and Schinske, 1961; Anderson and Stephens, 1969; Castille and Lawrence, 1979; Preston, 1993). Indeed, the occasional utilization of dissolved nutrients in arthropods has often been attributed to the ingestion of exogenous bacteria attached to the body surface (Castille and Lawrence, 1979; Siebers, 1979) and more recently to the ingestion of water (Tonheim et al., 2000). On the other hand, Rodriguez Souza and colleagues provided preliminary evidence that the phyllosoma larvae of the Japanese spiny lobster Panulirus japonicus Von Siebold 1824 can take up POM and DOM directly from the surrounding medium (Rodriguez Souza, 1997; Rodriguez Souza et al., 1999).
In those studies, phyllosoma larvae showed improvement in histological and cellular characteristics after incubation in solutions or dispersions of representative substances from the three main classes (protein, carbohydrate, lipid). More importantly, the results of the studies indicated that the absorption of POM and DOM by the Japanese spiny lobster larvae occurred not only through the gut, from orally ingested water, but also through the integument (Rodriguez Souza et al., 1999). This possibility was advanced based on the presence of mucus-like inclusions in the integument, of pore channels in the cuticle connecting the epidermis to the external medium, and of tegumental glands in the epidermis along the cuticle. In addition, P. japonicus have a flat, leaf-like larval stage, with a large surface area to body volume ratio, and a body that is covered with a soft, uncalcified cuticle, characteristics that provide particularly favourable conditions for integumental absorption. In fact, these characteristics are commonly found in aquatic invertebrates that have the ability to absorb nutrients through the integument (Manahan, 1983).
The feeding ecology of Japanese spiny lobster phyllosoma larvae is still not understood and, in spite of a century of trials on larval rearing and development of feeding protocols for this species, survival and growth rates during the larval stage are still insufficient for mass seed production (Matsuda, 2006; Matsuda and Takenouchi, 2006). Thus, POM and DOM could provide a nutritional supplement and be the key to the improvement of performance during larval rearing. In this context, the present study was designed to explore further the absorption capacity of the Japanese spiny lobster phyllosoma larvae. Experiments were conducted to verify the uptake of the amino acid glycine and, in particular, to clarify its route of absorption and internal distribution using histology and autoradiography as well as to quantify its uptake by high performance liquid chromatography (HPLC) and liquid scintillation counting (LSC).
MATERIALS AND METHODS
Source of materials and general experimental conditions
Phyllosoma larvae were obtained through natural spawning of broodstock maintained at the Minamiizu Station, National Center for Stock Enhancement, Fisheries Research Agency, and the Mie Prefecture Fisheries Research Institute, Japan. A total of 65 larvae were used in this study. The specimens were 250–260 days old (13.0–20.7 mm carapace length, 16.9–50.0 mg wet mass) and were reared to this age using standard rearing practices as described previously (Rodriguez Souza et al., 1996). Among these, 11 larvae had their mouths sealed with cyanoacrylate bond to permit distinction between the oral and integumental uptake of glycine. Sealing was carried out under a stereomicroscope by gently touching the mouth-parts with the tip of a bond-coated copper wire. Animals were rinsed 3 times in 0.45 μm pore-filtered, ultraviolet light- or autoclave-sterilized artificial sea water (Aquamarin, Yashima Chemicals Inc., Asagi, Japan) containing 100 μg ml−1 streptomycin and 100 U ml−1 penicillin, and maintained in similar medium without food for 24 h prior to the experiments. For the experiments, larvae were incubated individually or in groups of 3–4 in 100–300 ml of artificial sea water or glycine–sea water media (see description below) at temperatures between 23 and 25°C. All rearing media were prepared with sterilized artificial sea water as described above; glassware and tools used to handle and rear the animals were autoclaved for 20 min at 120°C.
Histology and autoradiography
The route of uptake and the distribution of exogenous glycine in the larval body as well as the histological characteristics of glycine-treated larvae were examined in two experiments using a total of 28 larvae, including 4 larvae in the first experiment whose mouths were sealed with the cyanoacrylate bond as described above. Groups of 2 larvae were incubated in 300 ml of 0.04 μmol l−1 [2-3H]glycine with a specific activity of 18.6 Ci mmol l−1 in the first experiment and in 200 ml of media containing 250 μCi of [2-3H]glycine (specific activity 14.9 Ci mmol l−1) in combination with inactive glycine to produce a final concentration of 6 μmol l−1 in the second experiment. Larvae incubated in glycine were sampled at 0.5, 2, 5, 12 and 24 h. Control larvae incubated in clean sea water were sampled at 0, 5 and 12 h.
After each incubation period, the larvae were rinsed 3 times in artificial seawater alone (first experiment) or seawater containing inactive glycine (second experiment) and then fixed for 24 h in Bouin's solution. Specimens were then dehydrated in an ethyl alcohol series, cleared in xylene, and embedded in tissue embedding media (Paraplast Plus; Sigma-Aldrich Co., St Louis, MO, USA). Saggital histological sections of the cephalic shield and thorax were cut (5–6 μm) and mounted on glass slides. The slides were deparaffinized, re-hydrated, coated with Konica NR-M2 autoradiographic emulsion in a dark room, air dried, and exposed for 6 months in a light-tight desiccator at 4°C. Autoradiographs were developed after exposure and the tissue was stained with Mayer's haematoxylin and mounted for microscopic observation. Adjacent portions of the stained tissue sections were also observed without staining to identify possible artifacts from the staining procedure (Peter and Ashley, 1967). The number of silver grains per unit area (mm2) was determined in selected structures (cuticle, epidermis, muscle, haemolymph, haemocytes, midgut gland cells and lumen) of glycine-treated larvae from the second experiment. Counts were made by digital image analysis of about 20 fields of 10 μm2 for each type of structure per larvae using the ‘count object’ feature of ImagePro Plus (Media Cybernetics Inc., Bethesda, MD, USA). Correction for background labelling artifacts (based on control larvae) was not necessary as they were negligible.
High performance liquid chromatography
The net uptake of glycine by the larvae was determined by HPLC analysis of the changes in glycine concentration in the medium with incubation time. For this purpose, 13 larvae were incubated individually in 100 ml of 6, 30 and 60 μmol l−1 solutions of inactive glycine, including 2 larvae with sealed mouths at 6 μmol l−1 and 2 at 60 μmol l−1. Bottles without larvae served as controls for possible natural (bacterial) decay of glycine. Aliquots of 1 ml were collected in triplicate from each of the bottles at 0, 2, 5 and 10 h of incubation, immediately filtered through a 0.2 μm Millipore filter, and stored at −80°C. HPLC analysis was performed using a post-column derivatization and fluorescence detection method developed by Shimadzu Corporation (manufacturer's instructions; Shimadzu Corporation, Kyoto, Japan) that enables selective detection of amino acids with high sensitivity (detection limit of the order of fmol). The method uses O-phthalaldehyde (OPA) as the deriving reagent for amino acid detection. Analyses were performed in a Shimadzu LC-10AS amino acid analysis system (Shimadzu Corporation). Glycine concentrations in the medium are shown as the percentage remaining in the medium compared with the initial concentration. Values represent the means of triplicate measurements for each individual per observation period. The rate of net uptake for each individual was calculated using the first order depletion constant K, derived from the slope of a least-squares linear regression of ln-transformed concentration values (Segel, 1976; Jaeckle and Manahan, 1989), and the initial concentration. Net uptake rates were expressed as μmol of glycine g−1 body mass h−1.
Liquid scintillation counting
The uptake of glycine from the medium was analysed by LSC using 19 larvae incubated under the conditions described above for the second autoradiography experiment, including 3 larvae whose mouths were sealed. Samples were collected and initially processed as for autoradiography. Specimens were subsequently digested in 0.5 ml of Soluene (Packard Instrument Company, Meriden, CT, USA) tissue solubilizer for 72 h at 60°C. After solubilization, 10 ml of scintillation cocktail was added and scintillation counting was carried out in a Packard 2550 TR/LL liquid scintillation analyser for 10 min. Values measured by LSC were corrected for the change in the specific activity of glycine following addition of the inactive carrier for representation as μmol per larvae. The uptake rates, expressed as μmol g−1 body mass h−1, were calculated using mass data of individual larvae measured immediately after the experiment (nearest 0.1 mg).
Scanning electron microscopy
Probing of the body surface for attached bacteria by scanning electron microscopy (SEM) was performed in 5 larvae incubated in 6 μmol l−1 inactive glycine for 10–12 h. Larvae were fixed in 2.5% glutaraldehyde in cacodylate buffer (pH 7.2) for 2 h, washed overnight in the same buffer, and post-fixed in 1.3% osmium tetraoxide for 2 h. Specimens were dehydrated in an ascending ethyl alcohol series and transferred to isoamyl acetate. Samples were then critical point-dried, mounted on aluminium plates, coated with platinum/palladium (about 10 nm) with an ion sputter (E102, Hitachi, Tokyo, Japan), and observed with a Hitachi S4000 scanning electron microscope.
Histology and autoradiography
A number of structural variations were evident in the larvae immersed in glycine solutions compared with the controls. After 12 h, relative to controls, glycine-treated larvae had thicker epidermis and muscles, larger tegumental glands, numerous and large haemocytes in the haemolymph space and, except for the E-cells at the distal portion, presented larger midgut gland cells (Fig. 1). Autoradiography showed that after 30 min the labelled amino acid had already been absorbed and retained in many structures such as the cuticle, epidermis, muscle, haemocytes and tegumental glands (Figs 2, 3 and 4). Portions of the integument showed label forming dense lines across the cuticle (Fig. 2A). Substantial accumulation of labelled glycine was observed in the semi-enclosed chamber formed by the labrum and paragnaths; label was found attached to the cuticle and denticles, and expanding into the inner layers of the cuticle and tissues adjacent to the mouth of intact larvae (Fig. 3). In the same animals, faint labelling was detected in the alimentary tract (e.g. proventriculus, foregut, midgut, hindgut and midgut gland). In contrast, only weak or no labelling was found around the mouth and in the alimentary tract of larvae whose mouths were sealed, although in these animals label could be seen in areas adjacent to those covered by the bond (Fig. 4). The temporal changes in the density of label in each structure during incubation are summarized in Fig. 5. Labelled glycine showed higher accumulation early (0.5–2 h) and then decreased with time in the epidermis, muscle and haemolymph. In contrast, accumulation increased with incubation time in the cuticle, haemocytes and midgut gland cells. The density of labelling in the lumen of the midgut gland was relatively constant between the first (0.5 h) and last (24 h) samplings.
High performance liquid chromatography
Measurement by HPLC of the glycine concentration in the medium for each individual showed that the response of the larvae was highly uniform at the lowest concentration (6 μmol l−1) but markedly irregular at the higher concentrations (30 and 60 μmol l−1; Fig. 6). At 6 μmol l−1 (larval mass range 40.8–49.4 mg), all larvae showed consistent net uptake throughout incubation, as can be inferred from the continuous decrease in the concentration of glycine in the medium. At 30 μmol l−1 (mass range 40.6–50.0 mg), although all larvae eventually showed net uptake after 5 h, two of them showed temporary net losses of glycine. At 60 μmol l−1 (mass range 16.9–48.6 mg), the results varied greatly with larval body size: the 48.6 mg larva showed steady net uptake of glycine whereas larvae ranging from 16.9 to 22.5 mg had continued net losses. The intermediate size larva (30.4 mg) showed first net uptake until 5 h and then loss at 10 h. The net uptake rates for individual larvae at 6 μmol l−1 varied between 0.29 and 0.39 μmol g−1 h−1; the rates apparently were not affected by the presence of cyanoacrylate bond covering the mouth but were markedly affected by body mass (Fig. 7A). At 30 μmol l−1, larvae showed net uptake rates between 0.12 and 1.31 μmol g−1 h−1 and these rates were also positively correlated with body mass. At 60 μmol l−1, the largest larva showed a net uptake of 0.42 μmol g−1 h−1 whereas the smaller larvae showed losses ranging from 0.07 to 1.81 μmol g−1 h−1.
Liquid scintillation counting
The uptake rates for individual larvae determined by LSC analysis varied between 0.04 and 0.29 μmol g−1 h−1 and were positively correlated with body size (Fig. 7B). After 12 h of incubation, larvae accumulated an average of 2.72 μmol g−1 (N=3; s.d.=0.64) of glycine (results not shown). As for the determination using HPLC, there was no obvious difference in the accumulation of labelled glycine and the uptake rates between intact larvae and those that had their mouths sealed with bond (Fig. 7B).
Scanning electron microscopy
Observation by SEM revealed that bacteria attached to the body surface of the individuals after glycine treatment were few and were therefore unlikely to significantly affect the uptake of glycine or to cause its depletion from the rearing water (Fig. 8).
This study provides evidence from multiple sources that Panulirus japonicus phyllosoma larvae are able to absorb nutrients directly from the surrounding water. The results confirm and support our previous histological findings, which pointed to the possibility of oral and integumental absorption of nutrients (Rodriguez Souza et al., 1999). Strong labelling was present in the cuticle, epidermis and muscle, and labelling was comparatively weaker in the various parts of the digestive tract. As revealed by the results of label counting in autoradiographs, the most significant accumulation up to 2 h was observed in the epidermis and muscles. At 5 h an apparent reduction in the density of label was observed in these structures, presumably caused by utilization of amino acids in response to metabolic needs. A reduction in the density of label with time was also observed in the haemolymph space. In the cuticle, haemocytes and midgut gland cells, the accumulation seemed to be steady. The decrease in label density in the epidermis and muscles concurrently with its increase in inner structures suggests that the amino acid was being transferred from the integument to the internal organs. That this transfer is not primarily dependent on the digestive tube seems to be borne out by the low and almost invariable density of label in the midgut gland lumen.
Large accumulations of labelled material seemed to be externally retained in the mouthparts as well, probably adhering to mucus secreted by the tegumental glands in the paragnaths. The adjacent inner sections of the cuticle and contiguous tissues in this region were also strongly labelled. This observation and the low concentration of label found in subsequent parts of the digestive system indicate that a significant portion of the orally ingested amino acids may be absorbed through the cuticle in the semi-enclosed chamber formed by the labrum and paragnaths. In larvae with a sealed mouth no label was observed in the semi-enclosed chamber and subsequent portions of the digestive tract; however, labelling in other structures presented similar characteristics to those observed in regular larvae, indicating that the integumental mode of absorption may account for the presence of label in these structures. Moreover, the absence of bacterial populations over the body surface of the examined larvae indicates that the contribution to the uptake of glycine by heterotrophic organisms is negligible. Rodriguez Souza and colleagues reported the presence of pore channels in extensive areas of the cuticle providing apparent connection between the external medium and the epidermis, and suggested that these structures could be implicated in the absorptive process with the aid of mucus secreted by the tegumental glands in the epidermis (Rodriguez Souza et al., 1999). The possibility of absorption through the pores was supported in this study by the observation of label forming dense lines across the cuticle in areas corresponding to the location of the pores.
The net uptake obtained by HPLC and uptake obtained by LSC strongly supported the autoradiographic observations. Both analytical methods also showed that sealing the mouth of the larvae did not affect the response of the larvae to the treatment, a fact which further points to the integument as a primary site for amino acid uptake. The concentration of free amino acids in the cells of invertebrate tissues ranges from 200 to 500 mmol l−1, exceeding 200 mmol l−1 for a single amino acid in some species (Preston, 1993). On the other hand, the concentration of glycine in natural marine environments is between 0.6 μmol l−1 in surface waters and 6 μmol l−1 near the sediment–seawater interface (Henrichs and Farrington, 1979; Manahan, 1983). Therefore, the integumental mode of absorption in individuals exposed to environmentally relevant (that is, low) concentrations probably occurs against a concentration gradient. The transport mechanism of free amino acids through the body surface in marine invertebrates has been reported to be a sodium-dependent co-transport (Preston and Stevens, 1982; Preston, 1993) that follows Michaelis–Menten kinetics for enzyme transport. Although a comprehensive analysis of the kinetics of glycine transport was not undertaken in this study, the characteristics of the uptake strongly indicate that a similar mechanism may account for the absorption of glycine in the phyllosoma larvae. Thus, glycine at the lowest concentration of (6 μmol l−1) was efficiently absorbed and retained by the phyllosoma larvae whereas exposure to increasing concentrations of glycine produced transient (e.g. at 30 μmol l−1) and/or marked and steady (e.g. at 60 μmol l−1) effluxes of the amino acid from the larvae to the medium. Similar responses have been observed in larval stages of other marine invertebrates such as the sea urchin (Davis and Stephens, 1984) and oyster (Manahan et al., 1989). For example, Davis and Stephens described losses of serine when the larvae were pre-incubated in high concentrations of the same amino acid (Davis and Stephens, 1984). They concluded that this would be an adaptive mechanism by the larvae to keep the internal pool of free amino acids. Thus, it is likely that spiny lobster phyllosoma larvae would also show increasing glycine uptake rates with increasing concentration until a limit is reached, whereby further increases could lead to subsequent saturation and efflux of the amino acid.
The results of this study also suggested that the capacity of amino acid retention by phyllosoma larvae is related to larval size (mass). Thus, at both concentrations that produced net uptake in all larvae (6 and 30 μmol l−1), uptake rates increased with increasing larval mass. Interestingly, only the largest larva incubated at the highest glycine concentration (60 μmol l−1) showed net uptake whereas all other (smaller) larvae at the same concentration showed net losses of the amino acid. Manahan and colleagues studied the ontogenic changes in the rates of integumental amino acid transport in Echinoderm larvae and observed that the metabolic rates increased in direct proportion to larval mass in this group (Manahan et al., 1989). They also noted that all species examined showed an increase in the capacity of amino acid transport with development and increased metabolic demands. However, it must be noted that all spiny lobster phyllosoma larvae used in the current experiments were of approximately the same age and within a relatively narrow range of body sizes. Therefore, and given the long duration of the larval stage of P. japonicus (i.e. about 1 year) (Matsuda, 2006), it would be premature to conclude that this relationship between body size and uptake rate would also be valid for animals of other ages.
The contribution of amino acids absorbed through the integument to the nutritional budget of marine invertebrates has been examined by earlier investigators (Stephens, 1972; Stephens, 1981; Stephens, 1988; Jorgensen, 1976; Wright, 1985; Wright, 1988; Wright and Manahan, 1989; Preston, 1993; Baines et al., 2005; Grover et al., 2008). They noted that this mode of nutrient acquisition may account not only for the nutritional requirements of the epithelial tissues but also in some species for a significant fraction of the requirements of the entire organism. They also emphasized the particular significance it may have for planktonic larvae and other organisms under nutritional stress. The P. japonicus phyllosoma larvae have a long pelagic larval stage spent predominantly in offshore waters (Sekiguchi, 1985), limited nutrient reserve tissues, and rather limited swimming and predatory capabilities to depend exclusively on macroscopic food masses for survival. Thus, as with many other organisms, it is reasonable to expect that phyllosoma also undergo periods of nutritional stress and could take advantage of even small contributions in the form of dissolved organic matter. It is noteworthy that Stevenson suggested that the cuticle of crustaceans and other arthropods may be used as food storage sites (Stevenson, 1985).
The results of this and other studies indicate that a re-evaluation of the common belief that P. japonicus phyllosoma larvae feed strictly on macroscopic diets (Nishida et al., 1990; Mikami and Takashima, 1994; Macmillan et al., 1997) may be warranted. Thus, based on this and previous studies (Rodriguez Souza et al., 1996; Rodriguez Souza et al., 1999; Rodriguez Souza, 1997) it seems fairly secure to say that spiny lobster larvae are able to obtain nutrients in three different ways: by capture and ingestion of large food masses, by oral ingestion of particulate compounds suspended or dispersed in the ambient water, and by direct absorption of dissolved materials through the integument. In this regard, it must be noted that the low survival rates typically obtained during larval rearing of this species are often attributed to unsatisfactory diet quality (Mikami et al., 1993). However, with few exceptions [e.g. the ‘ecosystem culture method’ of Kittaka (Kittaka, 1997)], experiments have always been performed by feeding larvae exclusively with large food masses. It may be that exclusive utilization of one mode of feeding may not be enough to meet the nutritional requirements of the larvae, at least during critical stages. Essential nutrients such as fatty acids and vitamins are also found dissolved in seawater, and even in small quantities may contribute to the nutrition of marine invertebrate larvae (Jorgensen, 1966). Therefore, it is suggested that the absorption of dissolved and particulate nutrients from the surrounding seawater may be concurrent and complementary to feeding on large food masses, and that it may contribute to the normal development of P. japonicus phyllosoma larvae. Ongoing studies should clarify the extent of this contribution in reared larvae, their nutritional requirements, and the occurrence of this mode of feeding in wild specimens.
The authors would like to thank the staff of Minamiizu, National Center for Stock Enhancement, Fisheries Research Agency and the Mie Prefecture Fisheries Research Institute, Japan, for broodstock care, hospitality, and access to facilities. This study was supported by a scholarship from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan to the senior author and by a grant from the Tokyo University of Marine Science and Technology to C.A.S.