Mutant RAS gene induces elevated levels of inositol tris-and hexakisphosphates in Dictyostelium

Dictyostelium discoideum has been shown to possess a single ras gene that is highly homologous to the three human ras genes (Reymond et al. 1984; Pawson et al. 1985; Weeks & Pawson, 1987). This gene has been cloned and has been re-introduced into D. discoideum and shown to be expressed (Reymond et al. 1985). Using site-directed mutagenesis a missense mutation that codes for threonine at position 12 of the gene product instead of the wild-type glycine has been constructed and introduced into Dictyostelium amoebae (Reymond et al. 1986). In human ras genes amino acid substitutions at homologous positions produce neoplastic transformation and decreased GTPase activity. A study of the phenotypes of the D. discoideum amoebae transformed with the wild-type ras-Glyl2 gene has revealed no measurable difference from the untransformed control, despite the fact that cells possess approximately fourfold higher levels of the ras gene product. In contrast, however, amoebae transformed with the mutant ras-Thrl2 gene showed aberrant development with defective chemotactic aggregation in low-density cultures. At high densities aggregates were formed but were aberrant in developing multiple organizing tips. When various components of the signal relay system connected with adenylate cyclase were measured, no aberration could be detected in the ras-Thr12-transformed cells that could explain the mutant phenotype (Reymond et al. 1986; Van Haastert et al. 1987). However, the formation of cyclic GMP in response to pulses of cyclic AMP was found to be abnormally low, due apparently to an effect on the signal adaptation mechanism. These results suggested that the role of the ras gene product was connected with the signalling pathway concerned with chemotaxis via guanylate cyclase rather than with signal relay that operates via the adenylate cyclase pathway.

Evidence has previously been presented indicating that the increased activity of guanylate cyclase occurs by stimulation of InsP₃* formation after binding of the chemotactic signal cyclic AMP to specific cell surface receptors; the InsP₃ formed then releases Ca²⁺ from internal (non-mitochondrial) stores that is able to trigger cyclic GMP formation (Europe-Finner & Newell, 1985, 1986a,b, 1987a,b; Small et al. 1986; Newell et al. 1987). Evidence for the formation of the other product of the InsP₃-forming reaction (1,2-diacyl glycerol) and its activation of protein kinase C is so far only indirect (Ludérusei al. personal communication). The involvement of a G-protein in the inositol phosphate pathway has been deduced from the stimulatory effects of GTP and non-hydrolysable analogues on inositol phosphate formation in saponin-permeabilized amoebae (Europe-Finner & Newell, 1987b).

In mammalian systems, recent evidence has implicated ras genes in coding for GTP-binding proteins involved in the inositol phosphate pathway of mam-malian cells (Fleischman et al. 1986; Wakelam et al. 1986, 1987; Preisset al. 1986). It seemed appropriate, therefore, to investigate the effects of normal and oncogene-like mutant ras genes on inositol phosphate formation in D. discoideum.

L-wyo-[l,2-³H]inositol (sp. act. 58·5 Ci mmol⁻¹) was obtained from New England Nuclear and passed through a

^*InsP, InsP₂, InsP₃, lnsP₄, InsP₅ and InsP₆ represent inositol mono-, bis-, tris-, tetrakis-, pentakis-and hexakis-phosphates, respectively, with assignment of phosphate position where appropriate, e.g. Ins(1,4,5)P₃; InsP_x represents the fraction of inositol polyphosphates (InsP₃, InsP₄, InsP₅and InsP₆) eluted from Dowex anion-exchange resins using 1 M-ammonium formate.

Dowex 1-X8 column (formate form) prior to use to remove trace impurities. [³H]inositol 1,4,5-trisphosphate (potassium salt, sp. act. 1 Ci mmol⁻¹) was obtained from Amersham Int. PLC. Hepes (N-2-hydroxyethyl-piperazine-N′-2-ethanesul-phonic acid), Geneticin and EDTA were from Sigma. Dowex X8 (100–200 mesh chloride form) anion-exchange resin was obtained from Bio-Rad. The chloride form of the Dowex resin was converted to the formate form before use by passage of 3 M-ammonium formate through the resin in a 40 mm × 450 mm column, until acidified silver nitrate gave no reaction, followed by deionized water to neutrality.

The axenic strain A×3 and the Dictyostelium ras (Dd-ras)-transformed derivatives D. discoideum were grown axenically as described by Watts & Ashworth (1970) in medium supplemented with 250 μgml⁻¹ dihydrostreptomycin sulphate. Amoebae were harvested during the exponential growth phase between 2 × 10⁶ and 5 × 10⁶ml⁻¹ and the axenic medium was removed by washing in P buffer (17mM-Na/K phosphate buffer, pH 6·1) and centrifugation at 190 for 2min. After three such washes the cells were resuspended in P buffer at 2×10⁷cellsml⁻¹. To maintain the high copy-number of the plasmid bearing the Dd-ras genes the axenic medium was supplemented with 20μg ml⁻¹ Geneticin (also called G418) (Reymond et al. 1985).

Amoebae suspended in P buffer at 2× 10⁷ cells ml⁻¹ were incubated in the presence of 170 nM-[l,2-³H]inositol at a specific activity of 10μCi ml⁻¹. Dihydro-streptomycin sulphate and CaC₂ were added to give final concentrations of 0·25 mgml⁻¹ and 1 mM, respectively. Cells were incubated at 22°C with aeration in a rotary incubator at 170 revs min⁻¹. After 3h cells were pulsed for 1h with cyclic AMP (50 nM final concentration) at 15-min intervals so as to synchronize the amoebae. After a total incubation period of 4h the amoebae were washed four times with 20 ml of P buffer to wash out [³H]inositol, and resuspended at 10⁸ cells ml⁻¹ in P buffer.

Samples (1ml) of labelled washed amoebae (1abelled by method 1) were incubated in the presence of 1 mM-CaC1₂ in plastic vials and shaken on an IKA-Vibrax platform shaker at 1400 revs min⁻¹ at 22°C or 4°C. Cells were stimulated with 20 μl of 2·6 μM cyclic AMP (to give a final concentration of 50 nM) and stopped at the appropriate time with 100 μl of 10% (v/v) HC1O₄. The acid extracts were then left on ice for at least 30min, neutralized with 135 μl of 1·53M-KOH and then buffered to pH 7·2 with 40 μl of 75mM-Hepes. KC1O₄ was precipitated at 0°C for 30 min on ice and removed by brief centrifugation.

Neutralized extracts, as prepared above, were diluted to 5 ml with 5 mM-sodium tetraborate/0·5 mM-EDTA and applied to columns (40mm × 6mm) of Dowex 1-X8 (100–200mesh) formate (Bone et al. 1984; Berridge et al. 1983) as described by Europe-Finner & Newell (1987a). The peak eluting with 100 mM-formic acid/1 M-ammonium formate was verified to contain InsP₃ by co-elution with an authentic sample of [³H]Ins(1,4,5)P₃ (ICimmor⁻¹).

Amoebae suspended in P buffer at 2–10⁸cells ml⁻¹ were dispensed as 0·5 ml samples onto 47 mm diameter Whatman 50 filter discs supported on 47 mm diameter absorbant pads (Millipore AP1004700) saturated in P buffer (Newell & Sussman, 1969). After the excess liquid that soaked through the filters had been removed, the filters and pads were incubated at 22°C in 90 mm Petri dishes in the dark. The filters were prevented from drying out (and any ammonia produced was absorbed) by 70 mm Whatman no. 17 filter circles soaked in P buffer attached to the undersides of the lids of the Petri dishes. After the required period of incubation (2 or 4h) each filter was removed from its supporting pad, the pad discarded, and the filter and cells carefully placed on a 100μ1 drop containing 100 μCi of L-WVO-[1,2-³H]inositol (sp. act. 58·5 Ci mmol^-1) placed in the centre of a 90 mm Petri dish so that the liquid spread evenly under the filter without excess. No supporting pads were used. The filters were incubated for 2 h at 22°C in the dark as before and then the cells were removed from the filters into 20 ml of P buffer and washed four times with 10 ml of P buffer to remove [³H]inositol, before being finally resuspended at 2× 10⁷ cells ml⁻¹ in P buffer.

Samples (5 ml) of 2× 10⁷ cells ml⁻¹ labelled by method 2 were shaken on an I KA-Vibrax platform shaker at 1000 revs min⁻¹ at 22°C (or at 4°C) in plastic scintillation vials. After 10 min incubation extracts were made by addition of 500 μl of 10% (v/v) HCIO₄, and were then left on ice for at least 30 min. Cell debris was removed by brief centrifugation (200 g for 2 min) and the supernatants were neutralized with 730μl of 1·53 M-KOH, with 200μ1 of 75 mM-Hepes buffer added to adjust the pH to 7·2. The precipitated KC10₄ was removed by brief centrifugation and the supernatant stored overnight at − 20°C. Any further precipitate was removed by further brief centrifugation. Supernatants were then transferred to separate tubes containing 500 μl of 10mM-EDTA (Downes et al. 1986).

Samples (1 ml) of supernatants from cell extracts prepared as described above were filtered through HV 0·45 μm Millipore filters (SJHV004NS) and 0·8 ml was injected into a 2 ml injection loop of a Beckman System Gold HPLC apparatus fitted with a 0·46 cm × 25 cm Partisil Sax 10 high-pressure anion-exchange column (packed by Technicol, Brook Street, Stockport, Cheshire SKI 3HS, UK). The elution conditions were a modification of the technique of Irvine el al. (1985) described by Batty el al. (1985) and were as follows. The elutant (ammonium formate buffered to pH 3·7 with ortho-phosphoric acid) at a flow rate of 1·25 ml min^-1 was increased in concentration linearly from 0 to 0·75 M over a period of 5 min to elute lower inositol phosphates. The eluant was then held at this concentration for 2 min followed by a linear increase over 6min to 10M to elute InsP₃. After a further 5 min at this concentration, the eluant was increased over 10 min to a final concentration of 1·7M and maintained for 6 min before being returned linearly to water. Fractions were collected at 18-s intervals and dissolved in 1 ml water/methanol (1:1, v/v), followed by 5 ml of scintillant (2 vol. of xylene containing 0·8% (w/v) 2,5-diphenyloxazole and 0·02% (w/v) 1,4-bis(5-phenyloxazole-2-yl) benzene with 1 vol. of Triton X-100). The InsP₃ peak was verified by coelution with authentic ³H-labelled Ins(1,4,5)P₃ added to test samples.

Separation of inositol phosphates was essentially as described by Heslopet al. (1985). The flow rate was set at 2·5 ml min⁻¹ and fractions were collected at 30-s intervals using two linear ammonium formate gradients. For the first 7 min after sample injection, water was passed through the column. The eluant concentration was then increased linearly from 0 to 0·85 M ammonium formate (buffered to pH 3·7 with ortho-phosphoric acid) over a period of 23·5 min to elute InsP₃ and lower inositol phosphates. The eluant concentration was then increased linearly over the next 11·5 min to 3·4 M to elute InsP₄, InsP₅ and InsP₅,. The eluant was held at this concentration for 10 min before returning to water. The identities of the peaks were verified by addition to the samples of authentic ³²P-labelled Ins(1,4,5)P₃, Ins(1,3,4,5)P₄ and Ins(1,3,4,5,6)P₅ generously supplied by Dr Philip J. Hawkins (Smith Kline & French Research Ltd, UK). InsP₆, was identified by Dr R. F. Irvine (AFRC Institute of Animal Physiology, Babraham, Cambridge, U.K.) using an electro-phoretic technique (Seiffert & Agranoff, 1965).

Previous studies concerned with the formation of inositol phosphates in D. discoideum employed the wild-type NC4 strain grown in association with Kleb-siella aerogenes (Europe-Finner & Newell, 1987a, b). However, the /as-transformed strains used in this study (Reymond et al. 1984) were derived from the axenic mutant strain Ax3 and it was therefore essential to examine first the formation of inositol phosphates of this parental strain after growth in axenic medium. In the initial series of experiments described below the inositol phosphates were determined by separation using Dowex anion-exchange resin chromatography. Under the conditions we employ, the fraction eluted from these columns by 1 M-ammonium formate (which was reported by Europe-Finner & Newell (1987a) to include InsP₃ and lnsP₄) has been found by more recent work using h.p.l.c. separation (see below) to include InsP₃, InsP₄, InsP₅ and part of the lnsP₆, fraction (which is present in large amounts in D, discoideum). As a consequence, it will be referred to in this paper as IP_x. The measurement of IP_x is used in the experiments described below to reveal rapid changes in the flux through the inositol phosphate pathway prior to determination of individual inositol phosphates by h.p.l.c.

The results of experiments with strain Ax3 (Fig. 1) indicate that at both 22°C and 4°C the effect of cyclic AMP stimulation on InsP_x. formation was found to resemble closely the previously reported results for NC4. The timing of the peaks was similar to NC4, except that the axenic strain showed a somewhat slower first peak: at 22°C it could be reproducibly observed at 5 s (rather than being almost over at this time as observed for NC4) and at 4°C it appeared at 10 s (rather than at 5 s).

When A×3 amoebae that had been transformed with the normal D. discoideum ras homologue, ras-Glyl2, were stimulated with cyclic AMP at 22°C, the results strongly resembled those obtained with Ax3, showing the same basal and cyclic AMP-stimulated levels (Fig. 2A). However, when the same experiment was carried out using A×3 amoebae transformed with the mutant ras-Thrl2 gene, the basal level of InsP_x was found to be greatly increased (Fig. 2B). Stimulation with cyclic AMP always produced oscillations but with only a small increase over the controls stimulated with water. The troughs of the oscillations were, however, noticeably below the controls, possibly suggesting the presence of a very active phosphatase such that brief cessation of InsP._Y synthesis led to its rapid degradation. Presumably these oscillations are sufficient to induce the weak cyclic GMP responses that have been observed in the ras-Thrl2-transformed cells at 22°C.

Previous experiments had shown that at 22°C, rapidly formed InsP_x peaks could be missed if they occurred within 5 s of stimulation, but that such peaks could be observed if the events were slowed down by performing the experiments at 4°C (Europe-Finner & Newell, 1987a). To ascertain whether such rapid effects were being missed in the experiments described above with the ras-Thrl2 strain, the InsP_x assays were repeated on cells incubated at 4°C. The results shown in Fig. 3 were obtained. As at 22°C the ras-Glyl2-transformed cells showed basal InsP._x levels similar to the parental Ax3 strain and could be stimulated by a pulse of cyclic AMP. The ra.s-Thrl2-transformed amoebae, however, showed approximately a threefold higher basal level of InsP._x and cyclic AMP did not show a significant increase or any evidence for oscillations.

When the mean values for the basal InsP_Y levels were calculated from all of the water controls shown in Figs 1–3, the results shown in Fig. 4 were obtained. It is evident that the ras-Glyl2-transformed cells showed only very minor differences from the Ax3 parental cells but the elevation of InsP_x levels in the ras-Thrl2-transformed cells was highly significant, particularly in experiments performed at 4°C.

In order to ascertain which of the inositol phosphates were elevated in the mutant ras-transformed cells, cell extracts from [³H] inositol-labelled cells were separated by h.p.l.c. using 25 cm Partisil Sax 10 anion-exchange columns (Irvine et al. 1985; Batty et al. 1985). Inorder to get sufficient ³H label into this fraction for accurate determinations, the labelling conditions were altered from the conditions described for lnsP._x determination (see Materials and methods) with the cells being deposited for development on filters rather than in starving suspension culture. Using these techniques the amounts of labelled InsP, InsP₂, InsP₃ and InsP₄ were determined in the ras-Glyl2 strain. It was found that the InsP and I11SP₂ levels corresponded well with determinations made using Dowex anion-exchange columns on the same samples and were very similar to values reported previously for strain NC4 (Europe-Finner & Newell, 1987). It was found, however, that the amounts of InsP ₃ and InsP ₄ were much less than the lnsP_x peak seen with Dowex chromatography, suggesting that a significant amount of a higher inositol phosphate was present. When the eluant conditions that were used with the h.p.l.c. column were altered to those reported to elute higher inositol phosphates such as InsP ₅ and InsP₆, (Heslop et al. 1985) a small amount of a ³H-labelled compound was found, which was identified as InsP₅ on the basis of its co-elution with authentic ³²P-labelled Ins(1,3,4,5,6)P5 added to the sample. Further elution revealed an additional large ³H-labelled peak that corresponded in its elution time to that of InsP ₆ reported by Heslop et al. (1985). This identification was verified by Dr R. F. Irvine (AFRC Institute of Animal Research, Babraham, Cambridge, U.K.) using the electrophoretic method of Seiffert Agranoff (1965).

The most important inositol phosphate peak in terms of those with known functions in D. discoideum is Ins(1,4,5)P₃, which releases Ca²⁺ from non-mitochondrial stores (Europe-Finner & Newell, 1986b). When this peak was analysed in cell extracts from the /as-transformed strains the results shown in Fig. 5 were obtained, indicating that at both 22°C and at 4°C the ras-Thr 12-transformed amoebae had a three-to five-fold higher level of Ins(1,4,5)P ₃ compared to amoebae transformed with ras-Glyl2.

It was also of interest to determine whether the InsP ₆ fraction, which was found to be the major ³H-labelled inositol polyphosphate in these cells (approximately 50-to 100-fold greater than the InsP ₃ fraction), also showed the same pattern. The results (Fig. 6) indicate that, while considerable variation in the amount of InsP₆ was observed for different batches of cells (shown by the error bars) the rax-Thrl2-transformed amoebae showed between three-and fivefold elevated levels compared to those of the r<2s-Glyl2-transformed cells at both 22°C and 4°C.

Analysis of other peaks seen by h.p.l.c. indicated that a peak provisionally identified as glycerophosphoinositol was very similar in magnitude in both the ras-Glyl2-and ras-Thrl2-transformed strains. While the levels of InsP, InsP₂, InsP ₄ and InsP₅ showed too much variation between batches of cells to permit meaningful assessment of differences, an unidentified peak that eluted after InsP ₆ showed a significant two-to threefold elevation in the ras-Thrl2-transformed amoebae at 22°C and 4°C.

As a control to ensure that the differences in the InsP3 and InsPg levels described above were not due to any subtle differences in the rate of development of the two ras-transformed strains, these two peaks were analysed in amoebae of both strains after labelling with [³H]inositol for different periods (either 2–4 h or 4–6 h) during development on filters. It was found that for both labelling times the pattern of differences between the two strains was identical and was as described above.

The data presented in this paper indicate that transformants bearing an oncogene-like mutation in the D. discoideum ms gene leading to substitution of threo-nine for glycine at position 12 of the ms gene product show a three-to fivefold elevated steady state level of Ins(1,4,5)P ₃, and a similarly raised level of InsP ₆, which is the major inositol polyphosphate of D. discoid-eum. These findings implicate the ms gene product in the signal transduction system leading from the cell surface cyclic AMP receptors to formation of InsP3 and polyphosphates. The observed effects are not due simply to the fourfold increase in the level of ms protein in the transformants as cells transformed with the normal (Gly) ms gene under similar conditions behaved indistinguishably from the untransformed Ax3 strain.

In mammalian systems, similar studies using trans-formation with mutant ras genes have been found to lead to elevation of inositol phosphates (Fleischman et al. 1986; Wakelamet al. 1987; Preiss, 1986) although in some systems only elevation of diacyl glycerol has been reported (Wolfman & Macara, 1987). From the results presented here and from previous studies it is evident that the Dictyostelium signal-transduction pathway closely resembles that of mammalian cells both in its components and in at least some aspects of its regulation. It is a striking (and from an evolutionary point of view instructive) finding that such a strong similarity exists between both the inositol phosphate and the adenylate cyclase signalling systems of higher multicellular organisms and those of a ‘primitive’ eukaryote.

The nature of the regulatory action of the ms gene product is, however, unknown. It seems, from its size, to be unlikely to be identical with any of the known subunits of the G-proteins of signal-transduction pathways, and further understanding of its role must await the identification of protein(s) with which it interacts.

We thank Frank Caddick for drawing the figures and Michael Berridge, John Heath and Julian Gross for helpful discussions. We gratefully acknowledge financial support from the Science and Engineering Research Council and the Cancer Research Campaign. M.E.E.L. was supported by an EMBO Short Term Fellowship. We also thank Philip Haw-kins for his generous gift of the ³²P-labelled inositol polyphos-phate standards and Robin Irvine for his help in the identification of InsP ₆.