Calcium-transport function of the chick embryonic chorioallantoic membrane ii. functional involvement of calcium-binding protein, Ca²⁺ -atpase and carbonic anhydrase

Calcium translocation by the chick embryonic chorioallantoic membrane (CAM) is a developmentally regulated, uni-directional and calcium-specific active-transport function (Garrison & Terepka, 1972a,b; Terepka et al. 1976; Tuan & Zrike, 1978; Dunn, Graves & Fitzharris, 1981; also see accompanying paper). Previous studies carried out in this laboratory have led to the identification of three biochemical, macromolecular components that appear to play functional roles in CAM calcium transport. These are: (1) a specific, high-molecular-weight calcium-binding protein (CaBP) (Tuan & Scott, 1977; Tuan et al. 1978a,b,c; Tuan, 1980a); (2) a Ca²⁺-activated, Mg²⁺-dependent ATPase (Tuan & Knowles, 1984); and (3) carbonic anhydrase (Tuan & Zrike, 1978; Tuan, 1984). All three proteins are localized in the ectoderm of the CAM, which lies adjacent to the shell membrane and eggshell (the calcium source), and constitutes the calcium-transporting cell layer of the CAM (Coleman & Terepka, 1972b). Specifically, the CaBP is associated with the cell surface of the ectoderm (Tuan et al. 1978b); the Ca²⁺-ATPase is an integral membrane component, which appears to be a near neighbour of the CaBP (Tuan & Knowles, 1984), and the carbonic anhydrase is most probably a cytosolic component (Anderson et al. 1981 ; Tuan, 1984). The involvement of these components in CAM calcium transport is suggested by a close temporal and spatial correlation between their appearance and the onset of CAM calcium transport in the developing chick embryo.

To define the roles of these putative transport components requires: (1) the availability of specific reagents for targeted perturbations of these components in situ in the CAM, and (2) calcium-transport assay systems that would permit the application of the aforementioned reagents for directly evaluating the functional importance of these components. Reagents specific for each of these components are indeed available. These include specific antibodies raised against the CaBP (Tuan et al. 1978b), pharmacochemical inhibitors, such as quercetin, for the CAM Ca²⁺-ATPase (Tuan & Knowles, 1984) and sulphonamides, such as acetazolamide, for the CAM carbonic anhydrase (Tuan & Zrike, 1978; Tuan, 1984). Using these reagents in conjunction with the two in vitro calcium-uptake assay systems (CAM tissue disks and cell-free microsomal membranes) discussed in the preceding paper and with an in vivo system (Tuan & Zrike, 1978), we directly tested the functional involvement of the CaBP, the Ca²⁺-ATPase and carbonic anhydrase in CAM calcium transport. Results reported here strongly indicate that these components are integrally and functionally involved in CAM calcium transport. These results also provide further insight into the mechanism of the transport function and have revealed its association with the plasma membrane.

Fertilized white Leghorn chicken eggs were incubated at 37·5 °C in a humidified commercial eggincubator for the desired period of time. Whole CAM was harvested by dissecting it away from the embryo and was rinsed clear of adhering materials with cold physiological saline.

This was carried out as described in the accompanying paper. Microsomes were suspended in a buffer containing 10mM-imidazole, pH 7·0, containing 0·1 M-KC1 (buffer C) until use.

Two methods were used to fractionate whole CAM microsomes. (1) Gel filtration. Size fractionation of microsomal membranes was carried out on a Sephacryl S-1000 (Pharmacia Biochemicals) column (1 cmX28 cm) eluted with 0’6M-KC1, 10 mM-imidazole, pH 7·4, containing 0-3M-sucrose at a flow rate of 32mlh⁻¹. The sample load was routinely l-2ml. (2) Densitygradient centrifugation. Microsomes in buffer C were made up to 10% sucrose (w/w) and 1-5 ml was loaded on top of a step sucrose gradient (20% to 55% in 5% increments of 1 -44 ml each). After centrifugation in a Beckman SW41 rotor at 20 500 rev. min⁻¹ for 3h at 4°C, the gradient was fractionated by upward elution into nine equivolume fractions using a gradient fractionator (ISCO Inc.). All fractions were suspended in buffer C and collected by centrifugation at 80000 g for 80 min and were suspended in buffer C until use.

The procedure used has been described (Crooks & Simkiss, 1975 ; Tuan & Zrike, 1978; Tuan, 1980b, 1983; preceding paper). Calcium uptake activities at 25°C were expressed as mol calcium min⁻¹ cm’².

The procedure was as described in the accompanying paper. Calcium uptake activities were expressed as mol calcium min⁻¹ cm⁻².

The procedure was similar to that recently used for human placental microsomes (Tuan, 1985) and is described in the accompanying paper. Activities were calculated as mol calcium s⁻¹ mg protein⁻¹.

The procedures for the following enzyme assays all involved the measurement of released inorganic phosphate, which was determined using the Malachite Green method as described previously (Tuan & Knowles, 1984). All assays were carried out at 37 °C using preparations solubilized in 1% Triton and activities were expressed as mol phosphate released min⁻¹.

This was carried out as described previously (Tuan & Knowles, 1984) using 1 mM-ATP, and activity was defined as the difference between ATPase activity levels measured in the absence of Ca²⁺ (i.e. 2 mM-EGTA added to the assay mixture) and in the presence of additional 10mM-CaCl₂.

The assay was carried out essentially as described previously (Tuan, 1979), in 50mM-Tris, pH8’0, containing 3mM-ATP, 0·112M-NaCl, 20mM-KCl (2mM-EGTA was included to eliminate Ca²⁺-ATPase activity). The activity was defined as the difference between ATPase activity levels measured in the absence and in the presence of 1 mM-ouabain.

The assay mixture contained 20 mM-AMP as described previously (Tuan, 1979).

The assay mixture contained 80 mM-glucose 6-phosphate in a histidine buffer, pH6·5, as described previously (Tuan, 1979).

Carbonic anhydrase activity was assayed on the basis of the hydration of CO₂ as described previously (Tuan & Zrike, 1978).

Rabbit-derived, specific anti-CaBP antisera (Tuan et al. 1978b) were used to determine CaBP levels by means of the enzyme-linked immunosorbent assay (ELISA) (Hudson & Hay, 1980). Briefly, samples to be tested were serially diluted and adsorbed onto the wells of a polystyrene microtitre plate (96-well; Linbro, Hamden, CT) in a carbonate buffer, pH 9-0. After washing, specific anti-CaBP antiserum (1/250 dilution) was applied to the wells. After incubation for 2h at room temperature, the plate was washed and the enzyme-linked secondary antiserum (alkaline phosphate-conjugated goat anti-rabbit immunoglobulin G (IgG), Sigma Chemical Co., 1/1000 dilution) was added. After additional incubation for 2h and further washing, the ELISA was developed usingp-nitrophenyl phosphate (Hudson & Hay, 1980) as a substrate. All samples were compared with a stock standard consisting of an extract of CAM prepared from 18-day chick embryos.

Fractionation of IgG from anti-CaBP antisera was carried out as previously described (Tuan et al. 1978b). Monovalent Fab’ fragments of the IgG were obtained by pepsinization followed by reduction and alkylation and fractionation on a Sephadex G-100 (Pharmacia) column according to standard protocols (Stanworth & Turner, 1978).

Protein was determined by the method of Lowry et al. (1951) with bovine serum albumin (Sigma Chemicals) as a standard.

Tissue samples of CAM were processed for electron microscopy as described previously (Tuan & Chang, 1975): fixation with 2-5% glutaraldehyde in cacodylate buffer (pH 7·4), postfixation with osmium tetroxide, and embedding in Epon. Ultrathin sections were stained with lead and uranyl acetate, and examined on a JEOL 100S electron microscope.

All chemicals used were of reagent grade. Radiolabelled compounds were purchased from Amersham Corp. (Chicago, IL). Sources for the following compounds were: quercetin, ethacrynic acid, p-nitrophenyl phosphate, ATP (vanadate-free), and tetracaine (Sigma Chemicals); trifluoperazine (Boehringer-Mannheim Biochemicals); and acetazolamide (Diamox, Lederle Labs). Goat-derived anti-(rabbit IgG) antibodies were obtained from Miles Laboratories.

The availability of specific anti-CaBP antibodies (Tuan et al. 1978; Tuan, 1980b) made it possible to study directly the functional involvement of the CaBP by observing the effect of these antibodies on CAM calcium uptake. This was first studied using the in vivo method, which involved the construction of an uptake chamber on top of the CAM in ovo (Crooks & Simkiss, 1975; Tuan & Zrike, 1978). With this system, we observed that pre-incubation of CAM with anti-CaBP IgG resulted in substantial reduction of uptake activity in a dose-dependent manner, which resembled a standard immunotitration profile (Fig. 1A). On the other hand, a similar treatment with pre-immune serum resulted in no inhibition. The finding that an ‘immunoequivalent point’ existed in the antibody-mediated inhibition of uptake suggested that immobilization due to immuno-crosslinking could be the mechanism of inhibition and that, furthermore, CaBP existing in a ‘mobile’ state on the CAM cell surface was a requirement for functional CAM calcium uptake. This supposition was further supported by the finding that monovalent Fab′ fragments derived from anti-CaBP antibodies exhibited no inhibitory effect on CAM calcium uptake (Fig. 1A). In addition, it was found that inhibition could also be achieved in excess anti-CaBP (1/100 dilution) if a secondary antibody (goat-derived anti-rabbit IgG) was applied in tandem (activities: control, 100%; anti-CaBP alone, 50%; goat anti-rabbit IgG alone, 98%; anti-CaBP plus anti-rabbit IgG, 28%). This finding was therefore consistent with the supposition stated above. The inhibitory effect of anti-CaBP on calcium uptake was also observed using the tissue disk method in vitro (Fig. IB). On the other hand, calcium uptake by CAM microsomes was only slightly affected when they were pre-treated with anti-CaBP antibodies (Fig. IB). Since CAM microsomes contained CaBP (see below), the failure of the anti-CaBP antibodies to inhibit calcium uptake significantly could be due to the inaccessibility of CaBP to the antibodies, perhaps because CaBP was segregated within the relatively intact membranous vesicles (see Fig. 2A).

We have previously identified several pharmacochemical inhibitors of the CAM Ca²⁺-ATPase (Tuan & Knowles, 1984). When tested in the microsomal calcium-uptake assay, both quercetin and tetracaine were effective inhibitors of CAM calcium uptake (Table 1). On the other hand, ethacrynic acid elicited only limited inhibition of uptake (Table 1). Since these compounds, which are effective inhibitors of detergent-solubilized CAM Ca²⁺-ATPase (Tuan & Knowles, 1984), differ in their solubilities in lipids, the above findings could have resulted from differences in membrane partitioning efficiency. It was also previously observed (Tuan & Knowles, 1984) that anti-calmodulin agents, such as phenothiazine, did not inhibit CAM Ca²⁺-ATPase activity. As assayed here, trifluoperazine, a potent anticalmodulin phenothiazine (Cheung, 1982), also did not effectively inhibit CAM microsomal calcium uptake (Table 1). Finally, quercetin and ethacrynic acid were also tested in the tissue disk calcium-uptake assay (Table 1), which showed that both effectively inhibited calcium uptake, with the latter requiring a substantial period of pre-incubation of the tissue with the inhibitor. Taken together, these findings are therefore consistent with and strongly suggest the functional participation of the Ca²⁺-ATPase in CAM calcium uptake.

onsistent with our previous observation that treatment of intact CAM in situ with sulphonamides (specific carbonic anhydrase inhibitors) strongly inhibited calcium uptake (Tuan & Zrike, 1978), we found that the sulphonamide, acetazolamide, was also a potent inhibitor of calcium uptake by CAM tissue disks in vitro (Table 1). However, when tested in the cell-free microsomal system, acetazolamide was an ineffective inhibitor (Table 1). Furthermore, no detectable carbonic anhydrase activity was present in the microsomal preparations. The involvement of carbonic anhydrase was also tested using an antiserum specific for chick CAM carbonic anhydrase (Tuan, 1984). It was found that the antiserum did not inhibit calcium uptake in either the tissue disks or the microsomal system (data not shown). These findings are therefore consistent with carbonic anhydrase being a cytosolic component that is probably involved as a facilitative rather than an intrinsic component of transmembrane calcium uptake (see Discussion).

To examine further the functional association between the CaBP and the Ca²⁺-ATPase and CAM calcium-uptake activity, and to gain insight into the subcellular membrane location of uptake activity, the CAM microsomal preparations were fractionated on the basis of size (gel filtration, Fig. 2) and density (sucrose densitygradient centrifugation, Fig. 3).

Upon gel filtration on Sephacryl S-1000, CAM microsomes were separated into fractions according to their size, with most of the vesicular components (04—0-3 pm in diameter) probably eluting at the column void volume (Dickson et al. 1983; Reynolds et al. 1983). Other fractions (Fig. 2) that eluted in the resolved and salt volumes probably contained small vesicles and membrane sheets, and non-specifically adsorbed proteins. The chromatogram was then analysed with respect to calcium-uptake activity, enzyme activities and CaBP levels. The results in Fig. 2 clearly showed that CaBP, Ca²⁺-ATPase and calcium-uptake activities were largely coincident at the column void volume and that their distribution profiles differed considerably from that of glucose-6-phosphatase, a marker enzyme for endoplasmic reticulum (de Duve et al. 1962).

A similar distribution was also observed upon density-gradient fractionation of the microsomal membranes. Specifically, as shown in Fig. 3, co-distribution was observed for Ca²⁺-ATPase activity, CaBP, calcium-uptake activity, and activities of a plasma membrane marker enzyme, 5’-mononucleotidase (de Duve et al. 1962), all of which differed substantially in distribution compared with glucose-6-phosphatase (endoplasmic reticulum marker; de Duve et al. 1962) and acid phosphatase (lysosomal marker; de Duve et al. 1962;data not shown).

These findings are therefore consistent with the notion that the CaBP and the Ca²⁺-ATPase are associated with, and are probably functional components of, the CAM calcium-uptake mechanism. In addition, the enrichment of uptake activity, CaBP and Ca²⁺-ATPase in fractions that were also enriched in marker enzymes of the plasma membrane strongly suggests that the CAM calcium-uptake system is most probably a component of the plasma membrane.

Our results demonstrate the functional involvement of the previously identified, putative transport components: the CaBP, the Ca²⁺-ATPase and carbonic anhydrase in calcium uptake by the chick embryonic CAM. Furthermore, these data strongly indicate that the calcium-uptake activity is associated with the plasma membrane. Although the exact modes of action of these components remain to be elucidated, the findings reported here, taken together with previous observations (Terepka et al. 1976; Coleman & Terepka, 1972a,b;Garrison & Terepka, 1972a,b; Crooks & Simkiss, 1975; Anderson et al. 1981 ; Dunn et al. 1981; Tuan, 1980a,b, 1983, 1984; Tuan & Knowles, 1984; Tuan & Scott, 1977; Tuan & Zrike, 1978; Tuan et al. 1978a,b,c), allow certain speculations.

We have previously postulated (Tuan & Zrike, 1978; Tuan, 1984) that carbonic anhydrase may provide localized acidification in the ectodermal zone to promote in vivo the dissolution of the eggshell calcite (CaCO₂) mineral to produce ionized calcium ready for uptake and, or, to regulate the metabolic fate of the HCO₃⁻ released from the shell. Since shell dissolution was not necessary in calcium uptake measured in vitro as described here, our findings suggest that the enzyme is perhaps involved in specific, regional acidification to increase localized ionization or the accessibility of transport component(s). Moreover, the fact that active calcium uptake may take place in a subcellular membrane preparation containing no carbonic anhydrase activity strongly suggests that the enzyme is most probably not an integral part of the active transport mechanism, but instead participates in a facilitative role.

The functional involvement of the CaBP in CAM calcium transport has been suggested previously from many lines of experimental evidence (Tuan, 1980a,b, 1983; Tuan & Knowles, 1984; Tuan & Scott, 1977; Tuan et al. 1978a,b,c). Further indirect evidence is also reported in the accompanying paper, i.e. the almost identical ion specificity and affinity of CaBP and calcium-uptake activities. More importantly, the inhibition of calcium uptake by CAM tissue (in vivo and in vitro) by exogenously applied anti-CaBP antibodies reported here (Fig. 1) is the first direct demonstration of the functional necessity of the CaBP in CAM calcium transport. The inability of the anti-CaBP antibodies to inhibit microsomal calcium uptake indicates that the subcellular vesicles are probably inside-out structures, with the CaBP located inside and therefore inaccessible to the exogenously applied antibodies. In this manner, the CaBP therefore serves as an internal calcium sink for the subcellular vesicles in vitro or as a calcium sequestrator on the shell-facing ectodermal cell surface in vivo (see below for further discussion).

Our recent finding (Tuan & Knowles, 1984) of a plasma membrane Ca²⁺-activated ATPase, which is also a near neighbour of the CaBP in the CAM, has raised the possibility that the enzyme is somehow involved in the transport process, particularly since it shares some properties with the many plasma membrane Ca²⁺-pumping ATPases reported in various systems (Penniston, 1982; Schatzmann, 1982). The findings reported here are clearly consistent with this notion, since enzyme inhibitors also inhibit calcium uptake. However, since an ‘ouabain-like’, specific inhibitor for plasma membrane Ca²⁺-ATPase has not been found, a specific one-to-one correspondence is not possible and the functional importance of the enzyme in calcium uptake remains to be elucidated. It should be pointed out that calcium uptake measured here in vitro exhibits K_m values of 0·3-0·5 mM-Ca²⁺, which correspond well with the lower affinity (K_m 0·3 mM) of the enzyme (Tuan & Knowles, 1984).

For the purpose of exploring further experimental means of analysis, it is useful to postulate, on the basis of currently available information and analogy with other plasma membrane calcium pumps (Terepka et al. 1976; Penniston, 1982), how the CAM calcium translocating mechanism may be assembled. As stated above, we postulate that subcellularly fractionated, calcium-uptake-competent microsomal vesicles are inside-out structures formed from plasma membranes and contain at least the CaBP and the Ca²⁺-ATPase, the former located in the internal space of the vesicles whereas the latter is an integral membrane protein oriented so that ATP hydrolysis is coupled to inward translocation of calcium. This inside-out orientation is therefore consistent with the observed requirement of high external K⁺ (i.e. cytosol-like conditions) for functional uptake as reported in the accompanying paper. We believe that this hypothetical model adequately accounts for the characteristics of the in vitro microsomal calcium-uptake process. Onacellular level, Terepkaetaf. (1976) and we (Tuan et al. 1978b) have previously postulated a pinocytosis model to explain transcellular calcium translocation in the CAM. In line with this model, the CáBP may be postulated as serving the role of a cell surface ‘calcium-specific receptor’, which upon calcium binding triggers specific, adsorptive endocytosis leading to the formation of pinocytic vesicles (Roth & Woods, 1982) that contain the CaBP and the membranous Ca²⁺-ATPase, in a manner similar to that in the cell-free microsomal vesicles studied here. The Ca²⁺-ATPase is orientated in situ so that it is capable of continuously pumping Ca²⁺ inwards into the pinocytic vesicle and thereby safeguarding its calcium load. Finally, by an unknown mechanism, perhaps by acidification (Brown et al. 1983) or other ionic changes, the Ca²⁺-CaBP complex is later dissociated, so that the calcium load may be delivered to the serosal side of the perivascular processes of the ectodermal cells, whereas component(s) of the transport apparatus, e.g. the CaBP, may be re-cycled to the proper region of the plasma membrane facing the calcium source. It is at least partly consistent with this hypothetical mechanism that the ultrastructure of the perivascular cells of the CAM ectoderm reveals a significant number of ‘pinocytotic-like’ vesicles (Fig. 4), previously also observed by many other investigators (e.g. see Ganóte et al. 1964; Coleman & Terepka, 1972a; Dunn & Fitzharris, 1979). Our current efforts are directed towards the testing of each of these steps to evaluate the validity of the proposed mechanism.

This work was supported in part by grants from the National Institutes of Health (HD 15306, HD 15822, and HD 17887) and the National Foundation/March of Dimes Birth Defects Foundation (Basil 0’Connor Starter Research grant 5-343 and Basic Research grant 1-939).