In terrestrial isopods, large amounts of Ca2+ are transported across anterior sternal epithelial cells during moult-related deposition and resorption of CaCO3 deposits. Because of its toxicity and function as a second messenger, resting cytosolic Ca2+ levels must be maintained below critical concentrations during epithelial Ca2+transport, raising the possibility that organelles play a role during Ca2+ transit. We therefore studied the uptake of Ca2+into Ca2+-sequestering organelles by monitoring the formation of birefringent calcium oxalate crystals in permeabilised anterior and posterior sternal epithelium cells of Porcellio scaber during Ca2+-transporting and non-transporting stages of the moulting cycle using polarised-light microscopy. The results indicate ATP-dependent uptake of Ca2+ into organelles. Half-maximal crystal growth at a Ca2+ activity, aCa, of 0.4 μmol l-1 and blockade by cyclopiazonic acid suggest Ca2+uptake into the smooth endoplasmic reticulum by the smooth endoplasmic reticulum Ca2+-ATPase. Analytical electron microscopical techniques support this interpretation by revealing the accumulation of Ca2+-containing crystals in smooth membranous intracellular compartments. A comparison of different moulting stages demonstrated a virtual lack of crystal formation in the early premoult stage and a significant fivefold increase between mid premoult and the Ca2+-transporting stages of late premoult and intramoult. These results suggest a contribution of the smooth endoplasmic reticulum as a transient Ca2+ store during intracellular Ca2+ transit.
The rapid bidirectional Ca2+ dynamics and high flux rates of Ca2+ during the moulting cycle(Neufeld and Cameron, 1993)make crustacean epithelia excellent model systems for the analysis of epithelial Ca2+ transport(Wheatly, 1997). The crustacean exoskeleton is mineralised mainly by CaCO3(Greenaway, 1985) and, in order to grow, crustaceans have to moult their cuticle periodically. The terrestrial isopod Porcellio scaber moults every 6 weeks and develops large sternal reservoirs containing amorphous, possibly hydrated,CaCO3 (Drobne and Štrus,1996; Ziegler,1994) before every moult. The deposition and resorption of these reservoirs are linked to the unique biphasic moulting cycle of isopods in which first the posterior and then the anterior cuticle are shed. During premoult, calcium is resorbed from the posterior cuticle into the haemolymph and transported across the anterior sternal epithelium (ASE) to form CaCO3 deposits between the epithelium of the integument and the old cuticle (Messner, 1965; Steel, 1993). Between the posterior and anterior halfbody moults (intramoult), these CaCO3deposits are resorbed and used to mineralise the new posterior cuticle. Ultrastructural studies have shown that the ASE differentiates for epithelial ion transport during the formation and resorption of these CaCO3deposits (Glötzner and Ziegler,2000; Ziegler,1996).
Ca2+ transport is generally either paracellular, between the epithelial cells, or transcellular. Transcellular Ca2+ transport can be divided into three phases: the passive influx of Ca2+ at one side of the epithelial cell, Ca2+ transport from one side of the cell to the other, and the energy-consuming extrusion of Ca2+. In crustaceans, entry probably occurs through Ca2+ channels or by a Ca2+/H+ exchanger(Ahearn and Franco, 1990; Ahearn and Zhuang, 1996). Active extrusion of Ca2+ probably involves a Ca2+-ATPase(Flik et al., 1994; Greenaway et al., 1995; Roer, 1980) or a Na+/Ca2+ exchanger(Ahearn and Franco, 1993). In the ASE of P. scaber, electronprobe microanalysis(Ziegler, 2002), expression analysis of the plasma membrane Ca2+-ATPase and the Na+/Ca2+ exchanger(Ziegler et al., 2001) and the abundance of Na+/K+-ATPase in the basolateral membrane(Ziegler, 1997) suggest that the transcellular pathway dominates.
How Ca2+ is transported within the epithelial cells is still unknown. Because of the multiple physiological functions of Ca2+,the mean free cytosolic Ca2+ concentration in cells is maintained at approximately 0.1 μmol l-1. Transient rises in Ca2+ concentration are tolerated, but high sustained Ca2+ concentrations in the cytoplasm are toxic and can lead to cell death (Berridge, 1993). A cytosolic route with Ca2+ bound to proteins(Feher et al., 1989) and an organellar route (Nemere,1992; Simkiss,1996) have therefore been proposed for Ca2+ transit. Simkiss (1996) proposed a model in which the smooth endoplasmic reticulum (SER) could function as a transient Ca2+ store, leading only to micromolar gradients in the cytoplasm. Electron-probe microanalysis on sternal epithelial cells of P. scaber showed a significant increase in the total (free plus bound)cytoplasmic Ca2+ concentration during the Ca2+-transporting stages(Ziegler, 2002). It is of particular interest that this increase is due to an increase in the number of areas with Ca2+ concentrations of up to 50 mmol l-1kg-1 dry mass because this is similar to the concentrations measured in the SER of bee photoreceptors(Baumann et al., 1991) and vertebrate skeletal muscle (Somlyo et al.,1981).
In an attempt to test the hypothesis that the SER contributes to epithelial Ca2+ transport, we investigated the smooth endoplasmic reticulum Ca2+-ATPase (SERCA)-dependent uptake of Ca2+ during four different moulting stages in P. scaber employing an in situcalcium oxalate assay (Walz and Baumann,1989). The results indicate an increase in SERCA-dependent Ca2+ uptake into the SER between the nontransporting and the Ca2+-transporting moulting stages.
Materials and methods
Porcellio scaber Latreille were maintained in plastic containers filled with soil and bark and fed on pieces of fresh potatoes and carrots. Adult non-breeding animals (body length 8.5-12 mm) at four different stages of the moulting cycle (early premoult, mid premoult, late premoult and intramoult) were used for the experiments. The animals in early and mid premoult were used 6 days and 2 weeks after the moult of the anterior part of the body, respectively. Well-developed sternal CaCO3 deposits defined the late premoult stage. For intramoult (the period between posterior and anterior moult) animals, specimens with partly degraded CaCO3deposits were selected.
Animals were dissected in nominally Ca2+-free (no calcium added)physiological saline containing 248 mmol l-1 NaCl, 8 mmol l-1 KCl, 10 mmol l-1 MgCl2, 5 mmol l-1 glucose and 10 mmol l-1 Tris, pH 7.4. Anterior and posterior sternal epithelia were dissected, and fatty tissue was carefully removed if necessary. Epithelia of animals in late premoult or intramoult already carried the first layers of unmineralised cuticle. Clean sternal epithelial cell layers were transferred onto nickel grids (200 mesh, thin bar,Plano Corporation) with the apical side facing away from the grid. The specimen and grid were then mounted in a perfusion chamber with the apical side facing up and with ready access of the solution to the basal side of the epithelium. A coverslip was placed on top of the perfusion chamber leaving only a narrow gap above the mounted specimen(Fig. 1A,B). The chamber was then placed on the stage of a light microscope (Zeiss, Axiophot) equipped with polarisation filters. Whenever possible, anterior and posterior sternal epithelia were analysed simultaneously.
Calcium oxalate assay
The principle of the calcium oxalate assay for measurement of relative Ca2+ uptake rates into the Ca2+-sequestering SER was described by Walz and Baumann(1989). In permeabilised cells, oxalate moves from a loading medium into the SER by an unknown mechanism. In the presence of ATP, the SERCA pumps Ca2+ into the lumen of the SER. When the oxalate and Ca2+ concentrations exceed the solubility product, birefringent calcium oxalate precipitates form within the SER lumen, and crystal growth can be monitored in a polarisation microscope as long as Ca2+ is transported into the lumen of the endoplasmic reticulum. After a calcium oxalate loading experiment, the birefringent calcium oxalate crystals appear bright against a dark background(Fig. 2).
For calcium oxalate loading experiments, the sternal epithelial cells were first permeabilised with 20 μg ml-1 saponin in 2 mmol l-1 K2EGTA, 125 mmol l-1 KCl, 5 mmol l-1 MgCl2, 5 mmol l-1 Na2ATP, 20 mmol l-1 Hepes, pH 7.0 (adjusted with KOH), for 20 min. After permeabilisation, the tissue was incubated under constant stirring in standard loading medium containing 1 mmol l-1 K2EGTA, 125 mmol l-1 KCl, 5 mmol l-1 MgCl2, 5 mmol l-1 Na2ATP, 25 mmol l-1 potassium oxalate, 4 mmol l-1 CaEGTA, 2.5 μg ml-1 oligomycin, 5 μmol l-1 carbonyl cyanide m-chlorophenylhydrazone (CCCP) and 20 mmol l-1 Hepes, pH 7.0 (adjusted with KOH). Oligomycin and CCCP were added to the loading medium to prevent calcium oxalate forming in the mitochondria. The Ca2+ activity (aCa) of the loading medium (0.68 μmol l-1) was measured using Ca2+-sensitive mini-electrodes (ETH129) prepared as described previously (Ziegler and Scholz,1997) and calibrated using the solutions of Tsien and Rink(1980).
To measure the Ca2+-dependency of calcium oxalate formation, we varied the aCa of the loading medium between 0.15 and 2.05μmol l-1 by changing the ratio of CaEGTA to K2EGTA,keeping the total EGTA concentration constant at 5 mmol l-1. ATP-dependency was demonstrated by omitting Na2ATP from the loading medium. The effects of cyclopiazonic acid (CPA) (1 μmol l-1),ryanodine (10 μmol l-1 and 0.5 mmol l-1) and caffeine(25 and 50 mmol l-1) were investigated by adding the reagents to the loading medium at various values of aca. For experiments with inositol trisphosphate (InsP3) (3 and 5 μmol l-1), we reduced the MgCl2 concentration of the loading medium to 2 mmol l-1 and repeated the experiments at various values of aCa.
All chemicals were purchased from Merck Corporation except Na2ATP, saponin, CPA, CCCP and oligomycin, which were obtained from Sigma Corporation, and EGTA, which was obtained from Fluka Chemika Corporation.
A CCD camera (Visitron, Spot) was used to take sequences of grey-scale images with a 20×/0.50 objective (Zeiss) at 1 or 2 min intervals. We used TINA software (Raytest) for the digital analysis of transmitted light through the polariser and analyser (aligned in the crossed position). Mean grey-scale values in areas of 1.25×103 μm2 were quantified for each image and are presented as the intensity change from that of the first image (Δintensity units). Linear regression was used to calculate the relative amount of calcium oxalate formed from each series of images. Rates are given as intensity units min-1, and values are presented as means ± S.E.M. One-way analysis of variance (ANOVA)followed by the Tukey—Kramer multiple-comparison test was used for statistical analysis. Calculations were performed using GraphPad Prism 3.0 software.
Electron energy-loss spectroscopy and electron energy-loss imaging
Sternal epithelia were prepared, permeabilised and incubated in loading medium as described above. After loading the epithelial cells with calcium oxalate for 50 min, single sternites were high-pressure frozen at 2.3×108 N m-2 (Leica, EMHPF) and freeze-substituted in acetone containing 1 % H2O and 1 %OsO4 (P. Walther and A. Ziegler, unpublished data). Freeze-substitution was performed over a 22 h period using a custom-built computer-controlled device with the temperature rising exponentially from -90 to 0 °C. After washing the specimens with pure acetone at room temperature(22 °C), the samples were embedded in Epon resin. Ultrathin sections (20 nm) were cut on a Leica Ultracut microtome with the sections floating on glycerol to prevent loss of water-soluble calcium oxalate precipitates. The sections were viewed unstained in an energy-filtering transmission electron microscope (Zeiss CEM 902) at 80 kV using a 30 μm diameter objective aperture. Electron energy-loss imaging (ESI) micrographs were taken below and above the specific element energy loss edge, L2,3 (346 eV) of calcium, at ΔE=340±5 eV andΔ E=360±5 eV, where E is the energy of the electrons. Electron energy-loss (EEL) spectra were recorded in serial mode with a scintillator-PMT detector over a range of ΔE from 300 to 400 eV using a 60 μm objective aperture, a 100 μm spectrometer entrance aperture and an energy resolution of 5 eV.
Calcium oxalate formation in epithelial cells
Calcium oxalate formation in sternal epithelial cells shows a biphasic time course (Fig. 3A). After a lag phase of approximately 15 min, the signal rises first slowly and then linearly for up to 1 h. After perfusion with ATP-free loading medium, the increase in optical signal was blocked immediately(Fig. 3B); it resumed after exposure to standard loading medium containing 5 mmol l-1 ATP(N=7). Exposure to 1 μmol l-1 CPA, a specific blocker of the SERCA, caused an immediate inhibition of calcium oxalate formation(Fig. 3C). The block by CPA was not affected by changing aCa to 0.38 μmol l-1 (N=2), 0.68 μmol l-1 (N=7) and 1.3 μmol l-1 (N=2). InsP3 (N=7),ryanodine (N=5) and caffeine (N=4) had no effect on the rate of calcium oxalate formation. The rate of calcium oxalate formation depended on aCa within the range 0.15-2.05 μmol l-1(Fig. 3D), which shows the relative Ca2+ uptake rate to have a half-maximum at 0.4 μmol l-1, indicating that active Ca2+ uptake into the SER was stimulated at physiological Ca2+ concentrations. The Hill coefficient was 1.8.
Transmission electron microscopy confirmed that the precipitates were formed in the sternal epithelial cells(Fig. 4). Although the cells were permeabilised and incubated in loading medium for a relatively long time,most specimens showed good structural preservation. The oxalate crystals appeared either as large elongated structures(Fig. 4A) or as small needle-like crystals (Fig. 4B). Electron-dense, calcium-containing precipitates were confined to the cytoplasm and were surrounded by smooth membranes(Fig. 4C,D). It is possible that the smaller crystals resulted from vesiculation of the membranous structures in some cells as a result of the permeabilisation and calcium oxalate loading procedures. Calcium oxalate appeared only occasionally in mitochondria (Fig. 4A,B). Electron energy-loss spectroscopic images of precipitates produced bright signals when the energy loss was switched from 320 to 360 eV, indicating the presence of calcium (Fig. 5A,B). Electron energy-loss spectroscopy of precipitates confirmed this result because of the large signal at the CaL2,3 edge at 346 eV (Fig. 5C).
Comparison of Ca2+ uptake rate in different moulting stages
Ca2+ uptake rates in sternal epithelial cells changed significantly during the moulting cycle(Fig. 6). Calcium oxalate formation increased from undetectable rates in early premoult to considerable rates in mid premoult. A significant increase in calcium oxalate formation occurred between mid premoult (0.045±0.027 intensity units min-1) and late premoult (0.23±0.016 intensity units min-1, P<0.001) and between mid premoult and intramoult(0.19±0.03 intensity units min-1, P<0.01) in the anterior sternal epithelium. No significant differences were observed between uptake rates in the anterior and posterior sternal epithelia during the late premoult and intramoult stages. In the posterior sternal epithelium of the early premoult and mid premoult stages, we obtained no conclusive results since we could not separate the cuticle from the epithelium. In these stages,birefingent structures formed within the mineralised cuticle, probably as a result of crystallisation of amorphous CaCO3. Generally, a slight and insignificant (P>0.05) decrease in uptake rate was measured between late premoult and intramoult. During intramoult, large birefringent crystals appeared in the new cuticle of most of the posterior sternal epithelium, again probably as a result of crystallisation of amorphous CaCO3 within the partly calcified cuticle. These specimens were omitted from the analysis.
The calcium oxalate method is a valuable tool for identifying and characterising Ca2+-sequestering organelles in situ. It meets all the requirements, such as specificity and linearity, to provide a good approach for monitoring the kinetics of Ca2+ uptake into the SER (Walz, 1982). A comprehensive discussion of this and related methods is given in the review by Walz and Baumann (1989). Here,we use this method to compare SERCA activity in the sternal integument of the terrestrial isopod Porcellio scaber among four different moulting stages. For the first time, we report a correlation between the increase in SERCA-dependent Ca2+ uptake into the SER and the moult-related increase in epithelial Ca2+ transport in Crustacea.
The dependency of calcium oxalate formation on ATP in the sternal epithelial cells of P. scaber and its almost total block by cyclopiazonic acid (CPA) indicate a SERCA-mediated sequestration of cytoplasmic Ca2+ into membranous compartments. This is supported by the dependency on the Ca2+ concentration in the loading medium,with the half-maximal uptake rate at aCa being 0.4 μmol l-1. This value is similar to the submicromolar values of calcium oxalate formation found in SER cisternae of photoreceptors in insects(Baumann and Walz, 1989; Walz, 1982) and crayfish(Frixione and Ruiz, 1988). Electron microscopy, electron energy-loss imaging and electron energy-loss spectroscopy verified that the calcium-containing precipitates formed within the epithelial cells of P. scaber during a calcium oxalate loading experiment are formed in smooth membranous cisternae, most probably the smooth endoplasmic reticulum.
Within cells, Ca2+ functions as a ubiquitous second messenger regulating a vast variety of physiological processes. Ca2+ signals are mediated by an influx of extracellular Ca2+ across Ca2+ channels and/or by release of Ca2+ from the SER(Berridge, 1993). The SERCA replenishes Ca2+ stores by re-uptake of Ca2+ into the SER and restores low cytosolic free Ca2+ concentrations. In Ca2+-transporting epithelia, in which cytoplasmic Ca2+loads are high, regulation of cytosolic Ca2+ concentrations by the SERCA may be of particular importance. Recently, Simkiss(1996) reviewed the conflict between bulk cytosolic transport of Ca2+ in epithelial Ca2+ transport, its function as a second messenger and the toxicity of sustained high Ca2+ concentrations. Simple diffusion of free Ca2+ through the cytosol is impeded by its relatively slow diffusion rate, and increasing the Ca2+ gradient from one side of the epithelial cell to the other would result in high and toxic Ca2+ concentrations. Therefore, mechanisms including facilitated diffusion by Ca2+-binding proteins(Feher et al., 1989) and compartmentalisation (Simkiss,1996) have been proposed for transcellular Ca2+transit.
A recent electron-probe X-ray-microanalysis (EPMA) of freeze-dried cryosections of the sternal integument of shock-frozen P. scaberrevealed high concentrations of in situ total (bound plus free)calcium, [Ca]t, of between 4.5 and 5.7 mmol kg-1 dry mass, suggesting the presence of high concentrations of Ca2+-binding proteins (Ziegler,2002). However, comparison of [Ca]t in the sternal epithelium of P. scaber between the early premoult, late premoult and intramoult stages indicates that the concentration of Ca2+-binding proteins does not change throughout the moulting cycle, arguing against a direct role of cytosolic Ca2+-binding proteins in epithelial Ca2+ transit. In contrast, the EPMA study demonstrated an in situ increase in the number of areas with high [Ca]t (15-50 mmoll-1 kg-1 dry mass) between early premoult and intramoult resulting from the contribution of Ca2+ `hot spots' to the analysed area (Ziegler,2002). The highest values of approximately 50 mmol kg-1dry mass are similar to those measured in the SER of vertebrate(Jorgensen et al., 1988; Somlyo and Walz, 1985) and invertebrate (Baumann et al.,1991) cells. Comparison of the rates of calcium oxalate formation reported here indicates that the SERCA activity increases from undetectable values in the non-Ca2+-transporting early premoult stage to measurable values in the mid premoult stage, and increases further by a factor of up to five between mid premoult and the Ca2+-transporting late premoult and intramoult stages. This suggests that, in the ASE and the PSE of P. scaber, the SER plays a direct role in epithelial Ca2+transit and supports the proposal that the Ca2+ `hot spots'revealed by EPMA represent SER cisternae. A role for the SER in epithelial Ca2+ transit was recently suggested in rat dental ameloblasts, in which SERCA activity and SER Ca2+-binding proteins are upregulated during the calcification process (Franklin et al., 2001; Hubbard,1996).
It is of interest that, in rat dental enameloblasts, the cytoplasmic 28 kDa Ca2+-binding protein calbindin is expressed in high concentrations,although the temporal expression pattern is not consistent with a primary role in Ca2+ transport (Hubbard,1996). This situation seems to be similar to that in the sternal epithelium of P. scaber, in which EPMA also suggests a high, but invariable, concentration of Ca2+-binding proteins in the cytosol. Ameloblasts, like the sternal epithelium of P. scaber, are involved in mineralisation processes that require massive transport of Ca2+in a very short time. Ca2+ flux rates through those epithelia are expected to be much higher than in the kidney and intestine. It seems possible that organellar routes evolved in mineralising tissues since high Ca2+ transit rates generally exceed the capacity of a cytosolic route.
Simkiss (1996) suggested that the loading and discharge of membranous compartments would lead to a vectorial translocation of Ca2+. Another possibility would be that Ca2+ diffuses through the lumen of the SER, possibly facilitated by low-affinity Ca2+-binding proteins. Alternatively, the SER could function as a Ca2+ buffer to prevent the formation of high cytosolic Ca2+ concentrations during SER-independent Ca2+ transit. It is important to note that a route through the SER would be in conflict with the SER's role in Ca2+ signalling. This conflict could be avoided if functions related to Ca2+ signalling and epithelial Ca2+ transport were regulated via different SERCA isoforms, possibly in different SER subcompartments. Recent investigations have revealed at least four different SERCA isoforms in crayfish tissues (Zhang et al.,2000) and two isoforms in whole brine shrimps(Escalante and Sastre,1993).
At this stage of the investigation, other routes for Ca2+transit, such as vesicular transport or co-secretion of Ca2+,cannot be ruled out for the sternal epithelium of P. scaber. In fact,electron microscopy, electron energy-loss spectroscopy and electron-probe X-ray-microanalysis demonstrated secretion of calcium-, phosphorus- and nitrogen-containing granules at the lateral plasma membranes of the ASE during resorption of the sternal CaCO3 deposits during intramoult(Glötzner and Ziegler,2000; Ziegler, 1996, 2002), suggesting co-secretion of protein and Ca2+. A contribution of mitochondria to Ca2+ storage and/or transport during epithelial Ca2+transit has also been discussed for several crustacean Ca2+-transporting epithelia(Rogers and Wheatly, 1997; Ueno, 1980). However,electron-probe X-ray-microanalysis of ASE cells of P. scaberdemonstrated a decrease in the total mitochondrial calcium concentration rather than an increase between early premoult and late premoult and between early premoult and intramoult (Ziegler,2002), excluding the possibility that mitochondria could store or transport Ca2+ during Ca2+ transit.
As a working hypothesis, we propose that in Porcellio scaber the SER actively contributes to Ca2+ transit through the ASE and PSE cells during the formation and resorption of the CaCO3 deposits. Future investigations should attempt to develop methods for the direct monitoring of epithelial Ca2+ transport and the use of pharmacological tools to analyse the role of the SER in epithelial Ca2+ transport.
We thank Dr Tom Carefoot for critically reading the manuscript and Oliver Schmid for his help in optimising the calcium oxalate assay. This work was supported by the Deutsche Forschungsgemeinschaft Zi 368/3-3.