Expression of membrane transporters in cane toad Bufo marinus oocytes

Frog oocytes have been used widely for the expression of foreign proteins (for reviews, see Colman, 1984; Sigel, 1990). Gurdon et al. (1971) demonstrated that oocytes from the South African clawed toad Xenopus laevis were able to synthesise haemoglobin when injected with rabbit haemoglobin mRNA. Since then, this system has proved to be an extremely useful and sensitive method for translating mRNAs from eukaryotes (plants, animals, yeasts) and prokaryotes (bacteria, viruses) (Colman, 1984; Sigel, 1990). This oocyte expression system has been used extensively for expression cloning (Hediger et al., 1987; Bertran et al., 1992b, 1993; Markovich et al., 1993a,b; Bissig et al., 1994; Pajor, 1995) of novel proteins, as well as for studying modes of regulation and structure/function relationships of transporters, ion channels and receptors (Colman, 1984; Dascal, 1987; Sigel, 1990).

The cane toad Bufo marinus, indigenous to Central and South America, is a widespread amphibian species that has shown considerable tolerance for reproduction and survival in all regions of the globe. Cane toads were introduced into Australia in June 1935 to combat the sugar cane beetles that were destroying sugar cane in Queensland. However, since then, they have become a biological pest because of their ability to adapt and multiply at an alarming rate (each female toad being able to produce 35 000 oocytes within a single spawn), without reducing the sugar cane beetle numbers. Fully grown oocytes of B. marinus are very similar in size to X. laevis oocytes and have also been shown to translate rabbit haemoglobin mRNA (May and Glenn, 1974; Glenn and May, 1975). However, unlike with the widely used X. laevis oocyte expression system (Sigel, 1990), there have been no subsequent studies that have used B. marinus oocytes as a protein translation system for the expression of heterologous proteins of foreign sources.

Because of the widespread abundance and easy accessibility to cane toads, it was our aim to determine whether oocytes of B. marinus could be used for the expression of foreign proteins. In this study, we demonstrate that B. marinus oocytes can efficiently translate foreign mRNA (encoding several membrane transporters) with protein expression properties identical to those of the well-characterised X. laevis oocytes, thereby providing a useful and viable alternative expression system, which could be used for studying membrane proteins of all origins.

Female Xenopus laevis toads were obtained from African Xenopus Facility C.C., Noordhoek, South Africa. Bufo marinus toads (females) were caught locally and maintained in the departmental Animal House. Small clumps of oocytes (total approximately 500–1500 oocytes) were treated for 60–90 min with collagenase Type 4 (2 mg ml⁻¹; Worthington Biochemical Corporation, New Jersey, USA; for X. laevis oocytes) or collagenase D (2 mg ml⁻¹; Boehringer Mannheim; for B. marinus oocytes) in calcium-free ORII solution (82.5 mmol l⁻¹ NaCl, 2 mmol l⁻¹ KCl, 1 mmol l⁻¹ MgCl₂, 10 mmol l⁻¹ Hepes/Tris, pH 7.5). Oocytes were then washed thoroughly five times with ORII solution followed by five washes with modified Barth’s solution [MBS: 88 mmol l⁻¹ NaCl, 1 mmol l⁻¹ KCl, 0.82 mmol l⁻¹ MgSO₄, 0.4 mmol l⁻¹ CaCl₂, 0.33 mmol l⁻¹ Ca(NO₃)₂, 2.4 mmol l⁻¹ NaHCO₃, 10 mmol l⁻¹ Hepes/Tris pH 7.4, gentamicin sulphate 20 mg l⁻¹]. The oocytes were sorted for morphologically intact, healthy-looking stage V–VI X. laevis oocytes or B. marinus oocytes of equal size. They were incubated in MBS at 17 °C and were injected with 50 nl of water (control), 1–3 ng of cRNA per oocyte or 35 ng of mRNA per oocyte using a Nanoject automatic oocyte injector (Drummond Scientific Co., Broomall, PA, USA). Oocytes were then kept at 17 °C in MBS for 1–4 days, with daily changes of MBS.

Rat NaSi-1 (Markovich et al., 1993a), rat sat-1 (Bissig et al., 1994; Markovich et al., 1994), rabbit NaDC-1 (Pajor, 1995), rabbit SGLT-1 (Hediger et al., 1987), human 4F2 hc (Teixeira et al., 1987; Bertran et al., 1992a) and human rBAT (Bertran et al., 1993) cRNAs were synthesised in vitro as described previously (Markovich et al., 1993a,b, 1994). Briefly, plasmids (1–2 μg) linearised by restriction enzyme digestion at the 3′ ends of the cloned cDNAs were subjected to the following transcription mixture: transcription buffer 1× (40 mmol l⁻¹ Tris/HCl, pH 7.9, 2 mmol l⁻¹ spermidine and 6 mmol l⁻¹ MgCl₂), 0.5 mmol l⁻¹ rATP, 0.5 mmol l⁻¹ rCTP, 0.5 mmol l⁻¹ rUTP, 0.5 mmol l⁻¹ m7G(5′)ppp(5′)G, 0.1 mmol l⁻¹ rGTP, 10 mmol l⁻¹ dithiothreitol, RNAase inhibitor (50 units) and appropriate RNA polymerase (50 units). The reaction mixture was incubated at 37 °C for 2 h, then RNAase inhibitor (50 units) and DNAase I (RNAase-free; 10 units) were added to the samples, which were incubated for a further 15 min at 37 °C. cRNA was then extracted twice with phenol:chloroform:isoamyl alcohol (25:24:1) and precipitated with 1 volume of ammonium acetate (7.5 mol l⁻¹) and 2.5 volumes of ethanol. cRNA was resuspended in 15 μl of water and used directly for injection. Total RNA from X. laevis oocytes was isolated by extraction with acid guanidinium thiocyanate/phenol/chloroform and then purified to poly(A⁺) RNA (mRNA) using oligo (dT)-cellulose, as described previously (Markovich et al., 1993a,b, 1994).

Uptakes were performed as described previously (Markovich et al., 1993a,b, 1994). In brief, oocytes (10 oocytes per individual data point) were first washed at room temperature (22–25 °C) for 1–2 min in solution A (100 mmol l⁻¹ choline chloride, 2 mmol l⁻¹ KCl, 1 mmol l⁻¹ CaCl₂, 1 mmol l⁻¹ MgCl₂, 10 mmol l⁻¹ Hepes/Tris, pH 7.5). This solution was then replaced by 100 μl of solution B (100 mmol l⁻¹ NaCl, 2 mmol l⁻¹ KCl, 1 mmol l⁻¹ CaCl₂, 1 mmol l⁻¹ MgCl₂, 10 mmol l⁻¹ Hepes/Tris, pH 7.5) supplemented with the desired concentration of cold substrate (K₂SO₄, succinate, D-glucose, L-arginine or L-leucine; see figure legends for details) and labelled substrates Na₂35SO₄, [¹⁴C]succinate, D-[¹⁴C]glucose, L-[³H]arginine or L-[³H]leucine (New England Nuclear Radiochemicals) (370–740 kBq ml⁻¹). After incubation at room temperature, the uptake solution was removed and the oocytes were washed three times with 3 ml of ice-cold stop solution (solution A). Each single oocyte was then placed into a scintillation vial, dissolved in 250 μl of 1 % SDS, followed by the addition of 2 ml of scintillation fluid (Emulsifier Safe, Canberra Packard) and counted (2 min per oocyte) using liquid scintillation spectrometry.

Values are given as means ± S.E.M. for 7–10 oocytes per condition and are representative of three similar experiments. Error bars not visible on graphs are smaller than the symbol used for that point. Statistical significance was tested using the unpaired Student’s t-test, with P<0.05 considered significant. For the transport kinetic studies, the Michaelis–Menten and generalised Hill equations were used to calculate the Michaelis constant (K_m) and maximal rate (V_max) using non-linear regression.

Initially, cDNAs encoding the following membrane proteins were transcribed in vitro: the rat renal Na⁺/sulphate cotransporter NaSi-1 (Markovich et al., 1993a), the rat liver sulphate/anion transporter sat-1 (Bissig et al., 1994; Markovich et al., 1994), the rabbit renal Na⁺/dicarboxylate cotransporter NaDC-1 (Pajor, 1995), the rabbit intestinal Na⁺/glucose cotransporter SGLT-1 (Hediger et al., 1987) and the human amino acid transporters rBAT (Bertran et al., 1993) and 4F2 hc (Teixeira et al., 1987). The cRNAs were injected into the cytoplasm of both X. laevis and B. marinus oocytes independently and were left for several days to allow the proteins to be translated, modified, sorted and trafficked to the plasma membrane. We then performed radiotracer uptakes as a measure of protein (expressed) activities.

Fully grown B. marinus oocytes (1.0–1.5 mm in diameter, approximate volume 1 μl), similar in size to X. laevis stage VI oocytes were selected for injection. It is notable that, unlike the distinct hemispherical coloration of X. laevis oocytes (a light vegetal pole and a darker animal pole), B. marinus oocytes (of all stages) showed no distinctive colour difference, but instead were a uniform jet black (opaque) colour (although some batches did appear to have a very small lighter vegetal pole; data not shown).

Oocytes of both X. laevis and B. marinus showed very similar uptake activities for all the above transport proteins (Figs 1–3), except for rBAT activity (L-leucine uptake) which was expressed only in X. laevis oocytes (Fig. 3B). It has been suggested previously that rBAT does not encode a true transporter, but instead an amino acid activator that induces an endogenous amino acid transporter in X. laevis oocytes (Bertran et al., 1992b; Taylor et al., 1996; Van Winkle, 1993). To determine whether we could induce the expression of this ‘putative’ X. laevis endogenous protein in B. marinus oocytes, we purified mRNA from X. laevis oocytes and co-injected it with rBAT cRNA into B. marinus oocytes (Fig. 3C). No significant stimulation of L-leucine transport above control rates (water-injected oocytes) was observed in response to either co-injection of X. laevis oocyte mRNA with rBAT cRNA or the injection of X. laevis oocyte mRNA alone into B. marinus oocytes (Fig. 3C). Endogenous oocyte transport activities (water-injected oocytes) for sulphate (Na⁺-dependent and Na⁺-independent), Na⁺/succinate, Na⁺/D-glucose and L-leucine uptakes were similar in both X. laevis and B. marinus oocytes (Figs 1, 2, 3B), with the exception of endogenous L-arginine uptake, which was approximately 2.5 times faster in B. marinus oocytes (0.96±0.28 pmol oocyte⁻¹ min⁻¹) than in X. laevis oocytes (0.36±0.045 pmol oocyte⁻¹ min⁻¹; Fig. 3A).

To characterise further protein expression in B. marinus oocytes, we performed time-dependence (length of uptake) and time-course (day of uptake) experiments with one of the above transporters, the NaSi-1 cotransporter (Markovich et al., 1993a). B. marinus oocytes were injected with NaSi-1 cRNA, and Na⁺/sulphate uptake was measured for various times (5–120 min; Fig. 4A). The NaSi-1-induced Na+/sulphate uptake rate increased linearly up to 120 min. Water-injected oocytes showed insignificant Na⁺/sulphate uptake, which remained low up to 120 min (Fig. 4A). Similar results were observed for X. laevis oocytes (data not shown). Next, B. marinus oocytes were injected with NaSi-1 cRNA, and Na⁺/sulphate uptake was measured on days 1–4 post-injection (Fig. 4B). NaSi-1-induced Na⁺/sulphate uptake rate increased up to day 3 and remained stable thereafter (Fig. 4B). Water-injected oocytes showed insignificant rate of Na+/sulphate uptake, which remained minimal up to day 4 (Fig. 4B). Similar results were obtained using X. laevis oocytes (data not shown). Furthermore, to determine whether B. marinus oocytes could be used for evaluating the kinetic variables (substrate affinities and transport capacity) of membrane proteins, we injected NaSi-1 cRNA into B. marinus oocytes and determined its maximal capacity (V_max) and substrate affinity (K_m) for sulphate (Fig. 5A) and Na⁺ (Fig. 5B). NaSi-1-induced transport in B. marinus oocytes showed an exponentially saturable sulphate interaction, representative of Michaelis–Menten kinetics, with the calculated variables being: V_max=45.5±1.9 pmol oocyte⁻¹ min⁻¹ and K_m=0.67±0.09 μmol l⁻¹ for sulphate (Fig. 5A). NaSi-1-induced transport showed a sigmoidal relationship with [Na⁺], and when the data were fitted to a generalised Hill equation, the following variables were calculated: V_max=14.8±0.6 pmol oocyte⁻¹ min⁻¹, K_m=22.0±1.5 mmol l⁻¹ for Na⁺ and the Hill coefficient n=2.6±0.4. These values are in very close agreement with values obtained in X. laevis oocytes (Markovich et al., 1993a; Busch et al., 1994).

Finally, since no study to date has performed a detailed characterisation of the ion specificity of the NaSi-1 cotransporter, we aimed to characterise NaSi-1 anion and cation transport specificity in oocytes (Fig. 6). NaSi-1-induced sulphate transport in B. marinus oocytes was maximal in the presence of Na⁺, suggesting a strong preference for Na⁺, whereas the other cations tested (choline, Li⁺, ammonium, K⁺) significantly reduced (greater than fourfold) sulphate transport (compared with the Na⁺-containing medium; Fig. 6A). NaSi-1-induced Na+/sulphate cotransport in B. marinus oocytes was significantly inhibited by selenate, molybdate, tungstate, oxalate and thiosulphate, but not by phosphate (each at 5 mmol l⁻¹; Fig. 6B). Identical results were observed for the cation and anion specificities of NaSi-1 transport in X. laevis oocytes (data not shown).

In the present study, our aim was to determine whether cane toad Bufo marinus oocytes could be used as a heterologous protein expression system. Our data suggest that B. marinus oocytes translate and express (plasma membrane) proteins equally as well as the widely used X. laevis oocytes. Several membrane transport proteins, NaSi-1 (Markovich et al., 1993a), sat-1 (Bissig et al., 1994; Markovich et al., 1994), NaDC-1 (Pajor, 1995), SGLT-1 (Hediger et al., 1987) and rBAT (Bertran et al., 1993), encoded by different stuctural entities and having individual substrate specificities, can indeed be expressed in B. marinus oocytes at levels comparable with those in X. laevis oocytes (Figs 1–3).

The exception is rBAT, which could not induce any amino acid uptake activity in B. marinus oocytes, whereas it was clearly expressed in X. laevis oocytes (Fig. 3B; Bertran et al., 1992a,b, 1993). The discrepancy could be because rBAT may not be a transporter itself, but is activating an endogenous X. laevis transporter that is absent from B. marinus oocytes, as we (Bertran et al., 1992b) and others (Van Winkle et al., 1993; Taylor et al., 1996) have previously postulated. Co-injection of X. laevis oocyte mRNA with rBAT cRNA into B. marinus oocytes did not further stimulate L-leucine (Fig. 3C) or L-arginine (data not shown) transport above levels in control oocytes, suggesting that co-expression of X. laevis proteins with rBAT in B. marinus oocytes was not sufficient for expression of rBAT-activated amino acid transport.

The fact that rBAT could not be expressed in B. marinus oocytes cannot verify or disprove that rBAT is an ‘activator of’ or ‘a protein-encoding’ amino acid transporter, but implies that B. marinus oocytes may lack the ability to express rBAT protein and/or lack the endogenous protein that may be activated by rBAT in X. laevis oocytes. Alternatively, rBAT may require the presence of another protein (in the oocytes) for proper function. rBAT and 4F2 hc encode structurally related glycoproteins with 52 % sequence similarity, which interact covalently with so-called ‘light chains’ (Bertran et al., 1992a). Recent results have shown that amino acid transport by these two proteins is dependent on the co-expression of their respective ‘light chains’ in X. laevis oocytes (Estevez et al., 1998; Palacin et al., 1998). This provides further evidence that rBAT requires an endogenously expressed (light chain) protein in X. laevis oocytes for proper transport function. On the basis of these observations, we speculate that B. marinus oocytes do not express this rBAT ‘light chain’ and therefore that no amino acid transport is induced in B. marinus oocytes, which would make this system ideal for studying the as yet unidentified light chain protein.

Another explanation of why rBAT activity cannot be measured in B. marinus oocytes could be that the rBAT mRNA contains consensus sequences, not present in the other transporters tested, rendering it more sensitive to endogenous RNAase H activity in B. marinus oocytes, which may be less active in X. laevis oocytes. In fact, the 3′ untranslated region of rBAT mRNA contains numerous AT(U)-rich motifs, not present in the other transporters tested, making it more susceptible to RNA degradation (Markovich et al., 1993b). Clearly, further work is needed to resolve these issues.

Previous electrophysiological studies have examined the electrical properties of B. marinus oocytes (Dascal, 1987; Iwao et al., 1981). The mean resting membrane potential of B. marinus oocytes (−50 mV: Iwao et al., 1981) was found to be identical to that of X. laevis oocytes (−50 mV: Dascal, 1987); however, in both species, it varied depending on the method used for defolliculation. Membrane resistance (R_m) in B. marinus oocytes were reported to be 10–40 KΩ cm⁻², with a membrane capacitance of 6–11 μF cm⁻², whereas values of R_m as high as 700 KΩ cm⁻² have been measured in X. laevis oocytes, with a comparable membrane capacitance of 4–7 μF cm⁻² (Iwao et al., 1981; Dascal, 1987). Intracellular ion concentrations were found to be similar in B. marinus and X. laevis oocytes (Dascal, 1987). These similarities in electrical, membrane and ionic properties suggest that B. marinus oocytes could be used for electrophysiological studies of transporters and ion channels. Previous comparative studies with X. laevis oocytes have shown that amino acid transport, urea transport and Cl⁻ channels are expressed in oocytes from the newt Cynops pyrrhogaster (Aoshima et al., 1988) and the frogs Rana esculenta (Martial et al., 1991) and Rana perezi (Ivorra and Morales, 1997).

Our radiotracer uptake studies demonstrate that B. marinus oocytes possess similar functional expression properties to those of X. laevis oocytes. We present the following lines of evidence to confirm this: (i) the levels of induction of transport in B. marinus oocytes by several transporters (NaSi-1, sat-1, NaDC-1, SGLT-1, 4F2 hc) are similar in magnitude to those in X. laevis oocytes; (ii) the levels of endogenous uptake of several substrates (sulphate, succinate, D-glucose and L-leucine) are comparable; (iii) the kinetics of the uptake are analogous; and (iv) the kinetic variables (V_max and K_m) detemined are similar in both species. We also observed that, in both X. laevis and B. marinus oocytes, the NaSi-1 cotransporter shows a strong preference for the Na⁺ cation, with insignificant transport when Na⁺ is replaced with choline, Li⁺, ammonium or K⁺. Thiosulphate and the tetra-oxyanions selenate, molybdate, tungstate and oxalate, but not phosphate, were all able to block significantly the NaSi-1-induced sulphate transport in X. laevis and B. marinus oocytes, probably by competing for the sulphate-binding site because of their structural (molecular) similarities to the sulphate anion. On the basis of these data, we believe that the B. marinus oocytes can be used as a viable alternative to the well-characterised X. laevis oocyte expression system.

In conclusion, this is the first study to show that oocytes of the cane toad B. marinus can be used successfully for the expression of membrane proteins. We propose that the widespread availability of the B. marinus species, especially in Australia, Asia and the Americas, would allow it to be used as a viable alternative and a more readily accessible system for the expression of proteins of all origins.

This work was supported by a grant from the National Health and Medical Research Council of Australia (to D.M.).