Characterization of a strain-specific inhibitor of cell division from Amoeba cytoplasm

Incompatibility following nuclear transfers between different strains and species of the large free-living amoebae was observed first between Amoeba proteus (_T1P) and A. discoides (_T1D₁₃) by Lorch & Danielli (1950), and has been shown since to be of widespread occurrence. These incompatible nucleo-cytoplasmic combinations were characterized by a failure to form clones, although cell division continued from 1 to 4 times. A second type of incompatibility existing among many amoeba strains and species was first noticed by Hawkins & Cole (1965). Injection of cytoplasm from A. discoides inhibited division immediately in about 60% of cells of A. proteus, but the remaining 40% of the injected cells were uninhibited and divided normally to form viable clones.

Incompatibility occurs commonly among other organisms. In many strains of Physarum polycephalum plasmodia fail to fuse, or else show a lethal reaction following fusion (Carlile & Dee, 1967; Border & Carlile, 1974). Tartar (1953) made grafts between Stentor coerulus and S. polymorphus and found that the resulting cells would not divide, while Theodor (1970) showed that failure of grafts between strains in the Gorgonacea was due to a lethal reaction which led to cellular destruction. Amongst the fungi, for example Neurospora crassa (Garnjobst & Wilson, 1956) or Aspergillus nidulans (Caten, 1972), incompatibility reactions often have a genetic basis.

In amoebae the cytoplasmic inhibitor of division has been localized in the post-microsomal supernatant fraction prepared from A. discoides which caused 90% inhibition of division when injected into A. proteus (Hawkins, 1969). The specificity of this inhibition was demonstrated by the injection of a variety of materials such as nucleic acids, proteins, amino acids and homologous cytoplasm into amoebae without any significant effects on the division of recipient cells (Cameron, 1973). Investigation of the cellular effects of the inhibitor by incorporation of [³H]uridine followed by autoradiography led to the conclusion that the synthesis of large RNA species was decreased by 1 day after injection in inhibited A.proteus, and had virtually ceased by 2 days. However protein synthesis as demonstrated by the incorporation of a variety of amino acids into acid-precipitable compounds, continued (at a lower level) for at least 9 days after injection (Cameron & Hawkins, 1976). Little is known of the nature of this cytoplasmic inhibitor of division, although Jeon & Lorch (1970) concluded it was a protein of very high molecular weight, up to io⁶ Daltons.

We have investigated further the inhibitor of division present in A. discoides (_rlD₁₃) cytoplasm which was active against A. proteus (_T1P) using ultracentrifugation, enzyme treatments, chromatography and electrophoresis. A preliminary investigation has been made also of the inhibitor present in the cytoplasm of another strain of Amoeba, Dawson’s A. proteus (_DP).

Clones of A. proteus (_T1P), A. discoides (TID₁₃) and Dawson’s A. proteus (_DP) were maintained in ‘wheat grain’ cultures (Lorch & Danielli, 1953) in modified Chalkley’s medium at 17 °C. Mass cultures of A. discoides and Dawson’s A. proteus, used to prepare subcellular fractions, were fed regularly on washed Tetrahymena pyriformis after the method of Griffin (1960).

Amoebae were labelled directly by addition of the radioisotope to the culture medium rather than indirectly by the feeding of radioactive Tetrahymena. Mass cultures of amoebae were cleaned from food organisms and placed in Chalkley’s medium containing 6·5 μCi/ml L-[4,5-³H]leucine (60 Ci/mmol; Amersham) for 48 h. Some mass cultures were placed in Chalkley’s medium containing either 0·25 μCi/ml L-[U-¹⁴C]lysine monochloride (287 mCi/mmol) or 0·5 μCi/ml L-[U-¹⁴C]glutamic acid (225 mCi/mmol).

Post-microsomal supernatant and post-microsomal pellet fractions were prepared from mass cultures of A. discoides or Dawson’s A. proteus which had been exposed to radioactive isotopes for 48 h or which had been fed Tetrahymena 48 h previously. Cells were bulked and lightly homogenized in TKM-sucrose buffer (50 mM Tris-HCl, pH 7·4, 25 mM KC1, 5 mM MgCl₂, 0 24 M sucrose) at 4 °C. Nuclear and mitochondrial fractions were removed by low-speed centrifugation, and a post-microsomal supernatant fraction was obtained after centrifugation for 2 h at 105000 g. This supernatant fraction was centrifuged for a further 6 h at 105000 g yielding a small post-microsomal pellet which was resuspended in 400/d 50 mM Tris-HCl buffer, pH 7·4, and stored in small aliquots at −20 °C.

A. discoides post-microsomal pellet was incubated with trypsin (10 and 50 μg/ml) at 22 °C for 5 min and with ribonuclease (25 μg/ml) at 17 °C for 18 h before testing for inhibitory activity by injection into A. proteus.

Amoebae were microinjected using a de Fonbrune micromanipulator according to the method of Hawkins & Cole (1965). The amount of fluid injected was approximately 10% of the volume of the cell. Injected cells were kept singly in solid watchglasses and fed on Colpidium sp. until they divided, or for 14 days.

Radioactively-labelled amoeba post-microsomal supernatant and pellet fractions were chromatographed on a 1·5×30 cm Sephadex G-200 column (Pharmacia) and protein was eluted with 42 ml 50 mM Tris-HCl, pH 7·4, 0·6-ml fractions being collected. Aliquots of these fractions were precipitated with an equal volume of ice-cold 10% trichloroacetic acid (TCA), filtered through Whatman GF/C disks and washed with 5 % TCA, 3:1 ethanol: ether and finally ether. Disks were placed in vials with toluene-PPO-POPOP and the radioactivity determined in a Beckman LS-230 liquid scintillation spectrometer. Selected Sephadex G-200 fractions were concentrated by dialysis against 40% sucrose-50 mM Tris-HCl, pH 7·4 buffer for 16 h at 4 °C before either injection into A. proteus or further characterization of the proteins.

Protein samples were electrophoresed in 3 5% polyacrylamide gels set in 0·5 × 7·5 cm glass tubes according to the method of Davis (1964). Gels containing radioactive samples were sliced into 1-mm sections, 150 μI Nuclear Chicago Solubiliser (NCS) added and incubated at 35 °C for 2 h. Toluene-PPO-POPOP (3 ml) was added and the radioactivity per slice determined using a Beckman LS-230 spectrometer. A modification of the Periodic acid-Schiff technique was used to detect glycoproteins (Kapitamy & Zebrowski, 1973).

The SDS method of Shapiro, Vinuela & Maizel (1967) using phosphate buffer was used. Reference proteins were stained according to the method of Weber & Osborn (1969) while gels containing radioactive amoeba proteins were prepared for scintillation counting as above. Molecular weights were determined by reference to standard proteins since the logarithm of the molecular weight is directly related to protein mobility in this gel system.

5 % polyacrylamide gels containing 4% Ampholine, pH range 3·5–10 (LKB) were prepared in 0·5 × 12 cm glass tubes. Protein samples were either applied on the top of the gel, protected by a solution of 4% ampholytes, or incorporated in the gel mixture before polymerization. The anode solution was 40 mM H₂SO₄ and the cathode solution 80 mM NaOH. A constant current of 2 mA/gel was applied for 2 h, and then a constant voltage of 350 V for a further 2 5 h. Gels of reference proteins were stained according to the method of Reisner, Nemes & Bucholtz (1975).

The injection of A. discoides post-microsomal supernatant into A. proteus typically inhibited cell division in approximately 90 % of injected cells. Inhibited cells appeared normal for the first few days, but later became dark in colour, detached from the substrate and much enlarged. Cytolysis occurred about 14 days after injection. Approximately 95% of control cells injected with TKM-sucrose buffer survived the operation and divided normally. Extensive high-speed centrifugation of the A. discoides post-microsomal supernatant yielded a small translucent pellet. When this pellet was resuspended and injected into A. proteus, the percentage of cells dividing was very low, less than 5% while some pellet preparations showed total inhibition of division. The ‘post-pellet’ supernatant showed no inhibitory activity when injected into A. proteus. The results of reciprocal injections of resuspended post-microsomal pellet between A. proteus (_T1P), A. discoides (_T1D₁₃) and in addition another strain, Dawson’s A. proteus (_DP) are shown in Table 1. A. discoides and A. proteus and A. proteus and Dawson’s A. proteus were incompatible, while A. discoides and Dawson’s A. proteus were compatible.

Optical density measurements at 260 and 280 nm of the post-microsomal pellet from A. discoides showed that the pellet contained approximately 3% RNA. The results of trypsin and ribonuclease treatments on the inhibitory activity of the pellet are shown in Table 2. The protein nature of the inhibitor was indicated by the complete elimination of activity by 10μg/ml trypsin. Injections of an equivalent concentration of trypsin had no effect on the division of amoebae. At 25 μg/ml, ribonuclease partially reduced the inhibitory activity of the pellet, 56% of the injected cells were inhibited compared to 95% with the untreated pellet. The effect of higher enzyme concentrations could not be investigated as above 25 μg/ml ribonuclease itself had some effect on the division of injected cells.

Passage of a [³H]leucine-labelled pellet from A. discoides through Sephadex G-200 followed by acid-precipitation of the protein gave the profile shown in Fig. 1. The most prominent protein peak eluted out in the void volume (fractions 18-23) and only small amounts of protein eluted out later. This void volume protein peak was found to be present in the post-microsomal supernatant, but absent from the post-pellet supernatant. Thus its presence was correlated with inhibitory activity. Use of either [¹⁴C]lysine or [¹⁴C]glutamic acid rather than [³H]leucine as radioactive precursor also gave a prominent void volume peak after fractionation of the post-microsomal pellet on Sephadex G-200.

Selected fractions after Sephadex G-200 chromatography including the void volume peak were concentrated 4-fold by dialysis and injected into A. proteus. The results, as the % division of injected cells at 4 and 14 days post-operation are given in Table 3. The most active fraction was 21, which inhibited division in approximately half of the injected cells. Although the final % division in A. proteus injected with other void volume fractions was quite high (above 64%), the very low % division at 4 days after injection appeared significant. Normally some slight delay in division was experienced after microinjection, perhaps due to mechanical shock, but in general most cells had divided by 3 days after the injection of buffered sucrose solutions. Delayed division was seen also in some preparations of post-microsomal supernatant material which possessed a lower than usual inhibitory activity, possibly due to low concentration of inhibitor molecules. Therefore it was considered likely that the inhibitor was present in void volume fractions 19–22, but at a relatively low concentration. In an attempt to concentrate the inhibitor the 3 void volume fractions 19–21 were lyophilized as were 3 fractions taken well after the void volume, fractions 50–52, and the lyophilizates were taken up into a final volume of 400 μl. At this concentration, injections from the protein-containing void volume fractions were not tolerated by the amoebae, but undiluted fractions 50–52 were and gave 94% division of injected cells (68 cells). Dilution of the void volume lyophilizate gave material which was tolerated by the amoebae and resulted in 34% division in injected cells (74 cells), only 24% having divided after 4 days.

Sephadex G-200 chromatography of a post-microsomal pellet prepared from Dawson’s A. proteus showed a profile of protein elution very similar to that obtained with A. discoides.

The results obtained after injecting material from the void volume of G-200 columns suggested the presence of inhibitor, but G-200 resolves only poorly protein mixtures since it has a wide molecular weight fractionation range, hence the peak almost certainly consisted of a number of proteins with molecular weights in excess of 800000. The void volume peak of a [³H]leucine-labelled A. discoides post-microsomal pellet was placed on a 3·5% Tris-glycine polyacrylamide gel and subjected to electrophoresis for 1·25 h (Fig. 2). Most of the proteins migrated very little in the gel due either to their large size or their charge, as their native state should be preserved in this non-denaturing polyacrylamide system. A similar profile was obtained if the void volume peak from Dawson’s A. proteus was run on a 3-5% Tris-glycine gel. When [³H]leucine-labelled void volume fractions from A. discoides were placed on a denaturing SDS polyacrylamide gel (Fig. 3), a wide range of proteins with molecular weights from approximately 10000 to 500000 were apparent, the SDS-denatured derivatives of the native proteins seen after polyacrylamide gel electrophoresis using Tris-glycine. Again, a similar wide range of proteins were seen if the void volume peak of Dawson’s A. proteus pellet was examined. One advantage of our experimental protocol was that we could test each step of the purification by subsequent injection of the material into living cells. An attempt to inject protein eluted from the protein peaks seen on Tris-glycine gels resulted in failure. Nearly all the injected cells cytolysed after injection. The toxic material was identified as 0·76 M Tris present in the buffer used. Another problem which arose from testing fractions after chromatography and electrophoresis was the length of time involved since after 4·5 days at 4 °C, the preparation of inhibitor had lost its activity.

Gel isoelectric focusing of [³H]leucine-labelled A. discoides void volume proteins in a linear pH gradient from 3·0 to 9·5 resulted in a very heterogeneous profile with many proteins over the range of isoelectric points. Since the Tris-glycine gels seemed to indicate the presence of relatively few proteins, this heterogeneity was surprising. One possible cause of this heterogeneity could be a variable degree of glycosylation of the proteins. Some indication of the presence of glycoproteins was obtained after staining a 3-5% Tris-glycine gel of the void volume fractions by the Periodic acid-Schiff technique. Therefore the void volume fractions from a PH]-leucine-labelled A. discoides pellet were incubated with 250 units/ml neuraminidase (from Vibrio cholera) for 1 h at 37 °C and then subjected to isoelectric focusing. Although some reduction in heterogeneity was noticed, the results could not be accounted for on the basis of extensive glycosylation.

We have shown that the cytoplasmic inhibitor of division of A. discoides active against A. proteus is localized in a subcellular fraction obtained after extensive high-speed centrifugation of the post-microsomal supernatant material. The protein nature of the inhibitor was confirmed by trypsin digestion of the post-microsomal ‘pellet’. The partial loss of inhibitory activity after digestion with ribonuclease suggested that RNA might be involved, possibly combined with protein in a complex.

Jeon & Lorch (1970) showed a slight effect of ribonuclease on the inhibitory activity of an amoeba supernatant, although they did not attribute this to the enzymic action of ribonuclease.

Gel filtration with Sephadex G-200 was used as a preliminary step in isolating the inhibitor from the A. discoides post-microsomal pellet. The void volume protein peak has been tentatively identified as the inhibitor of division since these fractions, concentrated by dialysis, resulted in a noticeable delay in division and a final % division below that of control cells when injected into A. proteus. The molecular weight or the inhibitory molecules must be over 800000 Daltons, since these molecules were excluded from Sephadex G-200. The comparatively high final % division could be due to insufficient concentration of the inhibitor in the void volume fractions. However, one problem in the purification of the inhibitor was stability, since 4-5 days at 4 °C was the maximum time before loss of inhibitory activity occurred. This suggested that the pellet itself contained protease and/or nuclease activities. The length of time required for the Sephadex chromatography combined with this limited stability made extensive assays of the void volume material by electrophoresis followed by microinjection very difficult.

Further fractionation of the A. discoides void volume protein peak showed the presence of several proteins. Polyacrylamide gel electrophoresis using a Tris-glycine buffer system revealed slow migration of these proteins, due to large size and/or charge of these proteins when in their native state. When reduced and denatured and subjected to electrophoresis in an SDS-polyacrylamide gel, many smaller proteins were seen. The combined effects of SDS and 2-mercaptoethanol are the dissociation of complex proteins into single SDS-coated polypeptide chains. Our results suggested that at least some of the void volume proteins were polymeric. The use of gel isoelectric focusing to separate these proteins on the basis of charge rather than size showed a surprising heterogeneity. One possible explanation was the existence of several forms of some proteins, each with a slightly different isoelectric point. The presence of isozymes has been demonstrated in Chaos carolinensis (Rothschild, 1967) and in A. proteus, A. discoides and A. dubia (Kates & Goldstein, 1964). Alternatively, glycosylation could account for the heterogeneity seen (Dorner, Scriba & Weil, 1973) but the use of neuraminidase did not reduce the heterogeneity markedly. Any further purification after gel filtration needs to be non-denaturing, since when the post-microsomal pellet was placed on a SDS-polyacrylamide gel before rather than after chromatography, no inhibitor could be detected if material was eluted from the gel and injected into A. proteus. However experiments using Tris-glycine polyacrylamide gel electrophoresis after gel filtration showed that the concentration of Tris in the eluates was toxic to the injected amoebae. Other nondenaturing methods of protein fractionation need to be used.

Preliminary experiments carried out using a post-microsomal pellet prepared from Dawson’s A. proteus suggested that its division inhibitor might consist of similar types of molecules as in A. discoides. This provides some evidence for the idea of Cameron (1973) that the division inhibitor consists of homologous molecules in all incompatible strains and species of amoebae. Perhaps molecules from one strain are sufficiently similar to bind to cellular site(s) in another strain, but not sufficiently similar to function normally. Further characterization of the inhibitor present in amoeba cytoplasm, in conjunction with the effects on cellular activities in inhibited cells, should provide a means of investigating the control of cell division in amoebae.

We would like to thank the Cancer Research Campaign for financing this work.