Unexpected roles of a Dictyostelium homologue of eukaryotic EF-2 in ...

1 downloads 0 Views 408KB Size Report
phosphorylated by Ca2+/calmodulin-dependent protein kinase. 2647. EF-2 is believed to be indispensable for polypeptide chain elongation in protein synthesis ...
Research Article

2647

Unexpected roles of a Dictyostelium homologue of eukaryotic EF-2 in growth and differentiation Sohsuke Watanabe, Kohji Sakurai, Aiko Amagai and Yasuo Maeda* Department of Developmental Biology and Neurosciences, Graduate School of Life Sciences, Tohoku University, Sendai 980-8578, Japan *Author for correspondence (e-mail: [email protected])

Accepted 13 March 2003 Journal of Cell Science 116, 2647-2654 © 2003 The Company of Biologists Ltd doi:10.1242/jcs.00476

Summary EF-2 is believed to be indispensable for polypeptide chain elongation in protein synthesis and therefore for cell proliferation. Surprisingly, we could isolate ef2 null cells from Dictyostelium discoideum that exhibited almost normal growth and protein synthesis, which suggests that there is another molecule capable of compensating for EF2 function. The knock-out of Dictyostelium EF-2 (DdEF2H; 101 kDa phosphoprotein) impairs cytokinesis, resulting in formation of multinucleate cells. The initiation

Introduction In general, growth and differentiation are mutually exclusive and precisely regulated during development. Thus the mechanisms involved in the transition of cells from a proliferation to differentiation state are of basic interest to developmental biologists and in the field of cancer research. Amoebae of the cellular slime mould Dictyostelium discoideum (strain Ax-2) cells grow and multiply by binary fission as long as nutrients are available. Upon exhaustion of nutrients, however, starving cells differentiate to acquire aggregation competence and aggregate by means of chemotaxis to cAMP (Bonner et al., 1969) and EDTA-resistant cohesiveness (Gerisch, 1961). Subsequently, cells differentiate into two cell types in a migrating pseudoplasmodium (slug): anterior prestalk and posterior prespore cells. They eventually culminate in a fruiting body consisting of a mass of spores and a supporting cellular stalk. The growth and differentiation phases are temporally separated from each other and easily controlled by nutritional conditions. A temperature-shift method for synchronizing the cell-cycle phase of Ax-2 cells has been established (Maeda, 1986), and a particular checkpoint (referred to as a putative shift point; PS-point) from growth to differentiation phase has also been specified in the mid-late G2 phase of the cell cycle (Maeda et al., 1989). That is, Ax-2 cells at any cell-cycle phase initiate differentiation by a departure from the PS-point under starvation conditions. Therefore, Dictyostelium development offers us a particularly useful system for elucidating the cellular and molecular mechanisms of the growth/differentiation transition (GDT). We have identified several genes (car1, caf1, quit3, dia1, dia2, dia3) that are specifically or predominantly expressed in response to differentiation of starved Ax-2 cells from the PSpoint and analyzed their functions (Abe and Maeda., 1994; Abe and Maeda., 1995; Okafuji et al., 1997; Itoh et al., 1998; Chae and Maeda, 1998a; Chae and Maeda, 1998b; Chae et al., 1998;

of differentiation, including the acquisition of aggregation competence, was delayed in Dd-ef2 null cells compared with that in wild-type. By contrast, Dd-ef2 overexpression enhanced the progression of differentiation, thus indicating a positive involvement of Dd-EF2H in growth/ differentiation transition. Key words: Dictyostelium discoideum, EF-2, Cell cycle, Growth, Differentiation, Cytokinesis, Mitochondria

Inazu et al., 1999; Hirose et al., 2000). We have also demonstrated that the phosphorylation levels of 90 kDa and 101 kDa phosphoproteins are specifically reduced during early cellular differentiation from the PS-point (Akiyama and Maeda, 1992). The 90 kDa phosphoprotein is a homologue of GRP94 (glucose-regulated protein 94; the endoplasmic reticulum HSP90) in D. discoideum (Dd-GRP94) (Morita et al., 2000). The expression of grp94 is induced by a variety of stress conditions, such as glucose-depletion (Pouyssegur et al., 1977; Shiu et al., 1977) and Ca2+ depletion in the ER (Drummond et al., 1987). Differentiation and morphogenesis of Dictyostelium cells is actually impaired by the overexpression of Dd-GRP94 (Morita et al., 2000). Since another protein (the 101 kDa phosphoprotein) remained to be identified, we sequenced the protein and analyzed its function in Dictyostelium development. As presented here, a partial amino acid sequence ‘VNFTIDQIRA’ of the 101 kDa phosphoprotein purified by 2D-SDS-PAGE was found to be identical with the polypeptide chain elongation factor 2 (EF-2) in D. discoideum (Dd-EF2) that was originally reported by Toda et al. (Toda et al., 1989). Ef-2 is believed to be indispensable for the polypeptide chain elongation step in eukaryotic protein synthesis. EF-2 translocates a peptidyl-tRNA from the aminoacyl site to the peptidyl site on a ribosome (Weissbach and Ochoa, 1976), thus being essential for cell proliferation (Perentesis et al., 1992; Mendoza et al., 1999). The N-terminus of Dd-EF2 is a GDPbinding domain and the C-terminal half interacts with ribosomes. Both show homology to hamster EF-2. The amino acid sequence of the carboxy half includes the site of ADPribosylation by diphtheria toxin. In mammalian cells, the activity of mammalian EF-2 for translocation is regulated by its state of phosphorylation. The dephosphorylated state is the active form (Ryazanov et al., 1988). EF-2 is specifically phosphorylated by Ca2+/calmodulin-dependent protein kinase

2648

Journal of Cell Science 116 (13)

III, known as EF-2 kinase (Nairn and Palfrey, 1987), and dephosphorylated by PP2A (phosphatase 2A) (Michael et al., 1989). EF-2 is also known to be a GTP-binding protein (Kohno et al., 1986) and is colocalized with actin (Shestakova et al., 1991; Bektas et al., 1994). Surprisingly, the present work using Dd-EF2–null cells has revealed that the 101 kDa phosphoprotein is not required for protein synthesis and cell proliferation but it is involved in cytokinesis and the growth to differentiation transition. Materials and Methods Cell culture and developmental conditions Vegetative cells of Dictyostelium discoideum Ax-2 were grown axenically in PS-medium (1% Special Peptone (Oxoid), 0.7% yeast extract (Oxoid), 1.5% D-glucose, 0.11% KH2PO4, 0.05% Na2HPO412H2O), 40 ng/ml vitamin B12, 80 ng/ml folic acid). ef-2OE or ef-2AS cells overexpressing or underexpressing Dd-EF2H, respectively, were shaken in PS-medium containing 30 µg/ml of G418. Dd-EF2null (ef2-null) cells were selected and grown in PS-medium containing 10 µg/ml of blasticidin S. To allow cells to differentiate, cells were harvested at the exponential growth phase, washed twice in BSS (Bonner’s salt solution) (Bonner, 1947) and allowed to settle either in a 24-well titer plate (Falcon #3047) or on 1.5% non-nutrient agar. This was followed by incubation at 22°C, as previously described (Chae and Maeda, 1998a; Chae and Maeda, 1998b; Inazu et al., 1999). Transformation of cells For the overexpression of Dd-ef2h, the full-length cDNA (clone SLE406) was inserted into the original vector (pDNeo2), using SalI and BamHI. To create a vector bearing the antisense Dd-ef2h, a 815bp fragment (–99-+716) of the cDNA clone was inserted into the BamHI and BglII sites of pDNeo2 in antisense direction. The vector constructs with sence or antisense Dd-ef2 were separately introduced into Ax-2 cells by electroporation, as described (Howard et al., 1988). Transformed cells were cloned and selected in PS-medium containing 30 µg/ml of G418 in 96- or 384-well titer plates (Falcon). Five to six days after the appearance of colonies of transformed cells, the colonies were transferred to 24-well plates. Dd-ef2hoverexpressing (ef-2OE) and -underexpressing cells (ef-2AS) were incubated by shake-culture in PS-medium containing 30 µg/ml of G418. To disrupt the Dd-ef2h gene, the blasticidin S (bsr) gene cassette (1.3 kb) was inserted into the vector for Dd-ef2h overexpression in which nucleotides +494-+808 of the Dd-ef2h cDNA had been deleted using SalI and EcoR1. This plasmid was amplified and the linearized SalI-NotI fragment was introduced into Ax-2 cells by electroporation. After 15 minutes at room temperature, the cells were dispensed into three 9 cm culture dishes and growth medium (PS-medium) added. Selection at 10 µg/ml blasticidin S was started 10-20 hours later in PS-medium, and bsr-resistant cells were cloned by axenic culture. Assay of protein synthesis To examine protein synthesis in various transformants and parental Ax-2 cells, Trans 35S-label (35S-methionine-cysteine; ICN) was applied to exponentially growing cells (5×106 cells/ml in PS-medium) at a concentration of 3.7 MBq/ml and shaken for 2 hours at 150 rpm. The radio-labeled cells thus obtained were washed twice in 20 mM Na/K-phosphate buffer (PB; pH 6.5) and suspended at 1×107 cells/ml in 20 mM PB. After 10 µl of 20 µg/ml BSA was added to the same volume (10 µl) of cell suspension, aliquots (5 µl) of the mixed suspension were plated on GF/A filters (Whatman) and then dried. The filters were treated with 5% TCA twice for 10 minutes on ice, 10

minutes at 100°C, and 10 minutes at room temperature to remove free 35S-methionine-cysteine. Subsequently, the filter was washed twice in absolute ethanol, twice in diethylether, and dried. The radioactivity levels of the filters were measured by liquid scintillation counting. The kinds of proteins synthesized during the pulse (2 hours)-label of cells with 35S-methionine-cysteine were determined by isoelectric focusing (IEF), subsequent 2D-SDS-PAGE and autoradiography of the samples, as previously described (Akiyama and Maeda, 1992). Preparation of the anti-Dd-EF2H antibody and western blot analysis Chemically synthesized oligopeptide (RKRKGLAPEIPALDK; from amino acids 799-813 of Dd-EF2H) with an additional cysteine residue at the C-terminus was conjugated with KLH (keyhole limpet hemocyanin) as a carrier protein (Research Genetics, Huntsville, AL). The KLH-conjugated oligopeptide was injected 4×1 ml subcutaneously (s.c.) into the foot pads of rabbits with complete Freund’s adjuvant. The total amount of the antigen was 5 mg per animal. Five weeks later, a total amount of 1 mg KLH-conjugated oligopeptide per animal with adjuvant was injected s.c. Samples of blood (about 40 ml) were collected 10 days after the final injection, and aliquoted serum containing the polyclonal anti-Dd-EF2H antibody was stored at –80°C. Following SDS-PAGE, the gels were transferred to a PVDF membrane, and the membrane was gently shaken in TBS-T containing 5% BSA or 5% nonfat milk, overnight at 4°C. Subsequently, the membrane was probed with the primary antibody diluted 1:5000 in TBS-T containing 5% BSA or 5% nonfat milk and 0.15% Tween 20, overnight at 4°C. After washing in TBS-T for 20 minutes, the blots were probed for 1 hour with a HRP-conjugated anti-rabbit secondary antibody (Amersham Pharmacia Biotechnology) diluted 1:30,000 in TBS-T. The blots were developed with ECL detection reagents (Amersham Bioscience) for 1 minute and exposed to X-ray film (Amersham Hyperfilm-MP) for 25-120 seconds. Double staining of cells with the anti-Dd-EF2H antibody and DAPI Vegetative cells were harvested at the exponential growth phase, prefixed in ice-cold 50% methanol for 10 minutes and then fixed in absolute methanol for 10 minutes on an ice-bath. After the fixed cells were dried on cleaned coverslips, they were dipped in PBS (140 mM NaCl, 3 mM KCl, 10 mM Na2HPO4, 2 mM K2HPO4, pH 7.2) for 10 minutes. The anti-Dd-EF2H antibody or preimmune rabbit serum, both of which were diluted 1:50 in PBS containing 1% BSA, was placed as a droplet on the coverslip and incubated for 2-3 hours at room temperature. The samples were washed in three changes of PBS (10 minutes for each). Subsequently, FITC-conjugated anti-rabbit IgG (Amersham) diluted 1:50 in PBS containing 1% BSA and 10% DAPI (4′-6-diamidino-2-phenylindole) was placed as a droplet on the coverslip and incubated for 2-3 hours at room temperature. After five washes in PBS (5 minutes for each), the samples were mounted in PBS containing 20% glycerol and observed under a fluorescence microscope. The FITC- and DAPI-stains in the same optical field were visualized using BV- and UV-excitation, respectively. Staining of cells with a mitochondrion-specific dye, MitoTracker Orange Dd-ef2-null cells and Dd-ef-2AS cells that had been starved for 2 hours at 22°C were stained with the anti-Dd-EF2H antibody and 0.5 µM MitoTracker Orange CMTMRos (Molecular Probes) for 15 minutes. The stained cells were washed twice in BSS and then fixed for staining with the anti-Dd-EF2 antibody, as described above. The preparations were observed under scanning confocal fluorescence microscope.

Unexpected roles of Dictyostelium EF-2 Results Targeted 101 kDa phosphoprotein is a homologue (DdEF2H) of EF-2 in D. discoideum As previously reported (Akiyama and Maeda, 1992), the 101 kDa phosphoprotein is located in the cytoplasm of Ax-2 cells, and the phosphorylation sites are serine residues at which the phosphorylation level becomes lower in response to differentiation of starved Ax-2 cells from the growth/ differentiation checkpoint (PS-point). Protein samples extracted from vegetatively growing AX-2 cells and justdifferentiated Ax-2 cells were separately applied to 2D-SDSPAGE, transferred to PVDF membranes (Immubilon; Millipore) and stained with CBB. The CBB-stained spots of the 101 kDa phosphoprotein were then detected by western blotting using a monoclonal antibody (PSR-45, mouse; Seikagaku, Tokyo, Japan) raised against phosphoserine. Among the spots around 101 kDa, we carefully specified one spot in which the phosphorylation of serine residues almost completely disappeared, coupling with the initiation of differentiation. Sequencing of the spot gave a partial amino acid sequence (VNFTIDQIRA). Interestingly, this sequence was found to be the same as the N-terminus of the elongation factor 2 (EF-2) of D. discoideum (Dd-EF2) reported by Toda et al. (Toda et al., 1989), being devoid of only N-terminal methionine. For further analyses of the Dictyostelium EF-2 homologue (referred to as Dd-EF2H), we used a cDNA clone, SLE406, with the full-length of Dd-ef2h, which was kindly supplied by the Dictyostelium cDNA project of Japan, and also prepared an antibody specific to the 101 kDa Dd-EF2H protein, as noted in Materials and Methods. Changes of the Dd-EF2H expression during the progress of cell cycle and starvation The northern blot analyses of Dd-ef2h showed that an mRNA of about 2.6 kb was strongly expressed during the vegetative growth phase and also during the early stage of starvation, followed by a decrease at about 6 hours of starvation (Fig. 1A). This is consistent with previous results (Toda et al., 1989). The developmental kinetics of the Dd-EF2H protein were qualitatively the same as those of the Dd-ef2h mRNA, except that the decrease in the amount of the protein during 4-6 hours of starvation was scarcely recognized (Fig. 1B). Fig. 2A shows temporal changes in the amount of Dd-EF2H during the progression of the cell cycle and the starvation of synchronized cells. No significant differences in the amount were observed among T0, T1, T3, T5, T7 and T9 cells synchronized by the temperature shift method. Comparison of the Dd-EF2 levels in just-differentiating T7+2 cells with that in starved but not differentiated cells (T1+2 and T3+2 cells) also indicated no significant difference between them (Fig. 2A). 101 kDa Dd-EF2H is not required for protein synthesis and cell proliferation in Dictyostelium Since EF-2 is believed to be indispensable for protein synthesis and cell proliferation, it was a surprise that we could obtain transformants (Dd-ef2AS cells) in which the expression of Ddef2h mRNA was supressed by antisense-mediated gene inactivation. As shown in Fig. 2B, both of the Dd-ef2h mRNA

2649

Fig. 1. Expression patterns of the Dd-ef2h mRNA and the Dd-EF2H protein during early development in Ax-2 cells and various transformed cells. Cells were harvested at the exponential growth phase, washed twice in BSS and shaken for the indicated times (hours) at 22°C. Total RNAs were prepared as described (Nellen et al., 1987). Northern hybridization was performed using the RI (Amersham), as previously described (Hirose et al., 2000). As a probe for detection of the Dd-ef2h mRNA, a PCR-product (3 kb) obtained by amplification of the cDNA clone SLE406 with the fulllength of Dd-ef2h using M13-20 and M13R primers was used. Western blotting was performed as described in Materials and Methods. The expression patterns in ef-2AS cells underexpressing the Dd-ef2h mRNA (A), Dd-ef2 null cells produced by homologous recombination (B), and ef-2OE cells overexpressing the Dd-ef2h mRNA (C) are presented in comparison with those in parental Ax-2 cells. In the lower panel in B, the amount of actin in each lane (stained with CBB) is shown.

(2.6 kb) and Dd-EF2H protein (101 kDa) were scarcely detected in Dd-ef2AS cells. The Dd-ef2h gene is known to be unique in the Dictyostelium genome (Toda et al., 1989). An exhaustive search of the Dictyostelium genome databases has also demonstrated that the Dd-ef2h gene is located as a single copy on Chromosome 2, and thus far there have been no similar nucleotide sequences found in the databases. Only EF-G (Dictyostelium EF-G; Dd-EF-G), which shows little homology (22.5% similarity in aa) to Dd-EF2H, has been mapped on Chromosome 6. This raised the possibility that Dd-ef2 null cells could be isolated by homologous recombination, which was later accomplished (Fig. 1B). However, enforced expression of Dd-EF2H in Dd-ef20E cells was not as striking as shown in Fig. 1C, though the reason for such an incomplete overexpression is presently unknown. Fig. 3 shows the growth kinetics of the several transformed cells and parental Ax-2 cells in growth medium. The data indicate that Dd-ef20E cells grew normally with almost the same growth rate as that of Ax-2 cells, and that both Dd-ef2AS and Dd-ef2 null cells exhibited slightly delayed growth. Here

2650

Journal of Cell Science 116 (13) Table 1. Relationship between protein synthesis ability and Dd-ef2h expression Protein synthesis activity (%)* Cells Ax-2 ef-2OE ef-2AS ef-2-null

Exp. 1

Exp. 2

100 98 93 106

100 105 107 102

Protein synthesis was determined by incorporation of 35S-methioninecysteine into proteins, as described in Materials and Methods. *The percentage incorporation of 35S-methionine-cysteine relative to that of Ax-2 cells.

altered levels of 101 kDa Dd-EF2H in the respective cells (Table 1). Autoradiography of 2D-SDS-PAGE also showed no significant differences in quantity and quality of proteins pulselabeled with 35S-methionine and 35S-cysteine among ef-2OE, ef-2AS, ef2-null and Ax-2 cells, with the exception of the enhanced synthesis of a 43 kDa protein in ef-2AS and ef2-null cells compared with that in ef-2OE and Ax-2 cells (data not shown).

Fig. 2. Changes of the Dd-EF2H protein during progression of cell cycle and starvation of synchronized cells. (A) Growth-phase Ax-2 cells synchronized by the temperature-shift method (Maeda, 1986) were withdrawn at the indicated cell-cycle phases. Exponentially growing Ax-2 cells (1-2×106 cells/ml) at 22.0°C, with a doubling time of about 7.5 hours, were shifted to 9.5°C, shaken for 14.5 hours and then reshifted to 22.0°C. Under these conditions, the cell number doubled within about 2 hours after a lag phase of about 1 hour. Tt cells, t hours after the shift-up from 9.5°C to 22.0°C, were harvested for western blot analysis. In another experiment, T7 cells, 7 hours after the shift-up, were harvested just before the PS-point, starved by washing twice in 20 mM Na/K-phosphate buffer (pH 6.5), and shaken at 1×107 cells/ml for 2 hours at 150 rpm. This yielded T7+2 cells, i.e. newly differentiating cells from the PS-point. T1+2 and T3+2 cells were also prepared by starving T1 and T3 cells for 2 hours in the buffer, as starved but not differentiated cells. The cell pellets were dissolved in 9 volumes of SDS-sample buffer, and the samples derived from the same number of cells (5×104 cells) were applied to SDS-PAGE (10% gel), followed by transfer to PVDF membranes and western blotting. In another experiment, synchronized T1, T3 and T7 cells were starved in BSS for 2 hours to obtain T1+2, T3+2 and T7+2 cells, respectively. Their western blot analyses were carried out as described above. (B) A schematic representation of the cell-cycle of an Ax-2 cell. The checkpoint (PSpoint) of growth/differentiation is interposed just after T7.

it is noteworthy that Dd-ef2AS cells as well as Dd-ef2 null cells are larger than Ax-2 and Dd-ef20E cells. To know if the overexpression or underexpression of Ddef2h affects protein synthesis, incorporation of 35S-labelled methionine-cysteine into proteins was compared in transformed cells and parental Ax-2 cells. As a result, there was no significant difference in protein synthetic activity among ef-2OE, ef-2AS, ef2-null and Ax-2 cells, in spite of

Dd-EF2 is required for normal cytokinesis during growth Although ef-2AS and ef2-null cells during shake culture in growth medium (PS medium) exhibited almost normal growth kinetics, they were found to be considerably larger than Ax-2 cells. DAPI-staining of fixed cells has revealed that most of the large cells are multinucleate (Table 2). Seventy to eighty percent of Ax-2 cells are known to be mononucleate with one nucleus per cell, and about 20% binucleate with two nuclei per cell (Maeda et al., 1989). In contrast, it is clear that the ratio

Fig. 3. Growth kinetics of ef-2OE, ef-2AS, ef2-null and parental Ax-2 cells. The four kinds of cells were separately shaken in growth medium without G418- or blasticidin S-addition at 22.0°C at 150 rpm. The number of cells at the exponential growth phase was determined using a hemocytometer. Similar results were obtained by cell counts in three independent experiments (䊉, Ax-2; 䉫, ef-2OE; 䉱, ef-2AS; 䊐, ef2-null).

Unexpected roles of Dictyostelium EF-2

2651

Table 2. Facilitated formation of multinucleate cells by inactivation of the Dd-ef2h gene Percentage of total cells (mean±s.d.) Cells Ax-2 ef-2OE ef-2AS ef-2-null

Mononucleate cells

Binucleate cells

Cells with three or more nuclei

79.8±3.7 78.5±3.9 68.6±2.9 47.1±3.1

18.5±1.7 19.8±1.5 27.7±2.5 38.7±2.9

1.9±0.3 2.2±0.4 5.2±0.7 14.9±1.2

Cells were fixed in methanol and stained with DAPI, as previously described (Maeda, 1986). Cells with one, two or more nuclei were counted separately. A minimum of 1000 cells were counted for each sample.

of multinucleate cells (with three or more nuclei per cell; about 2% in Ax-2 cells) is significantly increased up to 5% and 15% in ef-2AS and ef2-null cells, respectively, with a decrease in the numbers of mononucleate cells (Table 2). Taken together, these results suggest that the Dd-EF2H protein might be involved in cytokinesis, also that slightly delayed cell proliferation (Fig. 3), as observed in ef-2AS and ef2-null cells, might be due to augmented formation of multinucleate cells during growth. However, when ef2-null cells that had been shake-cultured in growth medium were transferred into a 24-well plastic plate and incubated without shaking, many large (multinucleate) Fig. 5. Development of ef-2OE (B,D,F) and parental Ax-2 cells (A,C,E) under submerged conditions. ef-2OE and Ax-2 cells were processed as described in the legend of Fig. 4. At 5.0 hours of incubation, Ax-2 cells remain as round-shaped single cells (A), but ef-2OE cells are elongated in shape with aggregation competence (B). At 6.0 hours, Ax-2 cells acquire aggregation competence (C), while ef-2OE cells exhibit enhanced differentiation to form aggregation streams (D). At 7.5 hours, ef-2OE cells form a more advanced stage of aggregation streams (F), compared with Ax-2 cells (E). Similar results were obtained by observations in three independent experiments. Bars, 200 µm.

cells were found to become smaller mononucleate cells by division within 3 hours of incubation at 22°C. This seems to indicate that impaired cytokinesis in ef2-null cells may be restored by cell adhesion to the substratum.

Fig. 4. Development of ef2-null (B,D,F) and parental Ax-2 cells (A,C,E) under submerged conditions. ef2-null and Ax-2 cells were harvested during the exponential growth phase, washed twice in BSS and plated in a 24-well titer plate at a density of 5×105 cells/ml (1 ml of cell suspension/ well). This was followed by incubation at 22°C for 6.0 hours (A,B), for 7.5 hours (C,D) and 10.0 hours (E,F). At 6.0 hours of incubation, Ax-2 cells acquire aggregation competence (A), while ef2-null cells show no sign of cell aggregation (B). At 7.5 hours, aggregation streams are formed in Ax-2 cells (C), but not in ef2-null cells (D). At 10.0 hours, Ax-2 cells form tight aggregation streams (E), but ef2-null cells still remain as aggregation-competent cells (F). Similar results were obtained by observations in at least three independent experiments. Bars, 200 µm.

Involvement of Dd-EF2H in cellular differentiation and morphogenesis When ef2-null cells and parental Ax-2 cells were separately starved and incubated in BSS under submerged conditions, the former exhibited delayed differentiation compared with the latter. Ax-2 cells acquired aggregation competence at 6 hours of incubation (Fig. 4A), whereas ef2-null cells showed no sign of cell aggregation (Fig. 4B). The early morphogenesis including aggregation in ef2-null cells was also delayed compared with that in Ax-2 cells: although Ax-2 cells formed tight aggregation streams at 10 hours (Fig. 4E), whereas ef2null cells still remained as aggregation-competent cells (Fig. 4F). ef2-null cells were just able to form aggregation streams after 12 hours of incubation. Essentially the same results were obtained using Ax-2 and ef2-null cells, both of which had been grown without shaking and starved in a 24-well plastic plate. On agar, starving ef2-null cells showed no sign of cell aggregation after 5.5 hours of incubation, whereas Ax-2 cells

2652

Journal of Cell Science 116 (13) while Ax-2 cells were just able to become aggregation competent after 6 hours of incubation (Fig. 5C) and then formed early aggregation streams at 7.5 hours (Fig. 5E). A similar result was obtained on agar using starving ef-2OE cells and Ax-2 cells. Just as in submerged conditions, the progression of morphogenesis in ef-2OE cells on agar was enhanced by about 1.5 hours compared with that in parental Ax-2 cells.

Fig. 6. Localization of Dd-EF2H in ef-2AS, ef2-null and parental Ax2 cells. (A-H) Cells were harvested at the exponential growth phase, fixed in absolute methanol and double-stained with the anti-DdEF2H antibody or non-immune serum, and then with DAPI, as noted in Materials and Methods. In Ax-2 cells (A-C), the cytoplasm was immuno-stained by the anti-Dd-EF2H antibody (B). Surprisingly, in ef-2AS cells (D-F), immunostaining of cytoplasmic granules is retained. Higher magnification of ef2-null cells double-stained with the anti-Dd-EF2H antibody and DAPI (G,H) indicates that the distribution of cytopasmic granules stained with the antibody is exactly the same as that of mitochondria stained with DAPI. (IK) ef2-null cells were double-stained with the anti-Dd-EF2H antibody (I) and MitoTracker Orange (K), as described in Materials and Methods. It is clear that both the stains are completely merged (J). Bars, 10 µm.

formed aggregation streams. Subsequently, ef2-null cells formed early aggregation streams after 7.0 hours of incubation. Just as in submerged conditions, ef2-null cells delayed the initial step of differentiation, as realized by the elapsed time for cell aggregation. In spite of the delay in development, ef2null cells eventually formed fruiting bodies with normal morphology. Essentially the same result was obtained using ef2AS cells underexpressing the Dd-ef2 gene under the control of actin 6 promoter. In contrast to ef2-null cells and ef-2AS cells, ef-2OE cells displayed more rapid aggregation than Ax-2 cells under submerged conditions (Fig. 5). ef-2OE cells acquired aggregation competence after 5 hours of incubation (Fig. 5B) and formed early aggregation streams at 6 hours (Fig. 5D),

Intracellular localization of 101 kDa Dd-EF2H and the existence of a 70 kDa mitochondrial protein sharing the antigenicity with Dd-EF2H In general, EF-2 is located in the cytoplasm. Immunostaining using the anti-Dd-EF2H antibody has confirmed that in Ax-2 and ef-2OE cells the Dd-EF2H protein is located in the cytoplasm (Fig. 6B). As was expected, ef-2OE cells exhibited slightly stronger staining than that in Ax-2 cells (data not shown). Surprisingly, however, cytoplasmic granules were found to be stained in ef-2AS and ef2-null cells devoid of 101 kDa Dd-EF2H (Fig. 6E,G,I). The stained pattern in ef-2AS cells was almost the same as that in ef2-null cells. Double-staining of ef2-null cells with the anti-Dd-EF2H antibody and DAPI revealed that both of the cytoplasmic stains are completely merged with each other (Fig. 6G,H). This was also confirmed by double-staining of the cells with the anti-Dd-EF2H antibody and MitoTracker Orange (Fig. 6I-K), thus indicating that the cytoplasmic granules stained with the anti-Dd-EF2H antibody are mitochondria. Western blot analysis using the anti-Dd-EF2H antibody gave a single band at the position of 101 kDa Dd-EF2H when proteins extracted from a relatively small number of Ax-2 cells were applied to SDS-PAGE, whereas no positive band was detected in ef-2AS and ef2-null cells under this condition (Fig. 7A). However, when loaded protein concentrations were increased by 100-times, the antibody recognized several bands even in the protein samples from ef-2AS and ef2-null cells (Fig. 7B). Among these bands are 71 kDa and 41 kDa proteins. Considering the results of immunostaining, it is possible that the two proteins are predominantly localized in mitochondria. To test this possibility, cytosolic and mitochondria-rich fractions derived from Ax-2 cells were processed for western blot analysis using the anti-Dd-EF2H antibody. As a result, the antibody detected the 71 kDa protein mainly in the mitochondria-rich fraction, while it recognized the 101 kDa Dd-EF2H protein predominantly in the cytosolic fraction (Fig. 7C). Taken together, these results seemed to indicate that the 41 kDa protein of Fig. 7B might be a hydrolysed product of the 71 kDa mitochondrial protein, and that the 58 kDa and 36 kDa proteins of Fig. 7B might be hydrolysed products of 101 kDa Dd-EF2H. Again, it is of importance to note that the 71 kDa mitochondrial protein shares the antigenicity with 101 kDa Dd-EF2H. Discussion EF-2 has been highly conserved in evolution, and the Dd-EF2H has 87.4% homology to the hamster EF-2 in the amino acid sequence (Toda et al., 1989). There are three regions in these sequences where especially high homology is found: amino acids 1-187, 314-565 and 590-779. EF-2 is known to have two

Unexpected roles of Dictyostelium EF-2

Fig. 7. Western blot analyses of proteins extracted from a different number of cells and subcellular fractions, using the anti-Dd-EF2H antibody. (A,B) Cell lysates prepared from the indicated number of cells were applied to lanes of 10% SDS-PAGE, followed by western blotting. (C) Ax-2 cells were fractionated to cytosolic (Cyt) and mitochondria-rich (Mit) fractions. Vegetative cells were harvested at the mid-late exponential growth phase and pelleted by centrifugation (400 g, 1 minute). The pellet was suspended in ice-cold 50 mM TrisHCl buffer (pH 7.5) containing 2.5 M sucrose and homogenized using a glass homogenizer. The homogenate was centrifuged for 10 minutes at 900 g to remove undisrupted cells and nuclei, and the resulting supernatant was centrifuged for 20 minutes at 2000 g. The pellet thus obtained was used as a Mit fraction. The supernatant was again centrifuged for 1 hour at 16,000 g, and the supernatant was put to use as a Cyt fraction. All of the processes were carried out around 4°C. In the Mit fraction, a band of 71 kDa protein can be detected in addition to contaminated 101 kDa Dd-EF2H. Based on the band patterns observed, it is most likely that the 58 kDa and 36 kDa proteins in B are hydrolysis products of 101 kDa Dd-EF2H, while the 41 kDa protein may be a hydrolysis product of the mitochondrial 71 kDa protein.

functional domains. The first homologous region seems to correspond to the first domain, essential for GTP binding and GTPase activity, whereas the next two regions contain the domains involved in interaction with the ribosome (Nilsson and Nygard, 1985; Kohno et al., 1986). The high homology in these regions indicates a pronounced evolutionary conservation of these functional domains. EF-2 is believed to be indispensable for cell proliferation as well as eukaryotic protein synthesis, and therefore its knockout caused a lethal effect on cells (Livingston and Bodley, 1992). Surprisingly, however, the results presented here have demonstrated that the 101 kDa Dd-EF2H of Dictyostelium cells is not required for protein synthesis, and that Dd-EF2H-null cells as well as ef-2AS cells are able to grow almost normally, except for formation of multinucleate cells during growth. Protein synthesis in ef-2ASand ef2-null cells was almost the same as that in parental Ax-2 cells. This seems to indicate that in Dictyostelium cells there must be a molecule(s) other than 101 kDa Dd-EF2H that is involved in protein synthesis and therefore capable of compensating for the function of EF-2.

2653

Proteins containing the same or a similar epitope that would crossreact with the anti-Dd-EF2H antibody have not yet been identified in the Dictyostelium mitochondrial genome or the genome sequencing project. After northern analysis, no RNA transcripts, as probed by the full-length Dd-ef2h cDNA, were detected in Dd-ef2-null cells. Therefore, it is unlikely that smaller Dd-EF2H fragments may work to compensate the EF2 function. The lack of RNA transcripts and Dd-EF2H protein in Dd-ef2hAS cells also suggests that molecules structurally similar to the 101 kDa Dd-EF2H protein cannot compensate for EF-2 function. EF-G (Dd-EF-G), which is devoid of the epitope used for preparation of the anti-Dd-EF2H antibody, might be a candidate capable of compensating for EF-2 function, although its homology to Dd-EF2H is not very high (22.5% similarity in aa). Alternatively, it is possible that the 43 kDa protein and/or 71 kDa mitochondrial protein are responsible for protein synthesis instead of 101 kDa Dd-EF2H, because the former is predominantly synthesized in Dd-ef2AS and Dd-ef2-null cells compared with parental Ax-2 cells, and the latter shares the antigenicity with the 101 kDa Dd-EF2H. As noticed particularly in ef2-null cells, the knock-out of 101 kDa Dd-EF2H brought about impaired cytokinesis, thus resulting in the appearance of multinucleate cells. Positive participation of 101 kDa Dd-EF2H in the process of cytokinesis would be a novel function. In connection with this, some of eukaryotic EF-2 is known to be colocalized with actin microfilament bundles in mouse embryo fibroblasts, which suggests a possible link between the protein synthetic machinery and the cytoskeleton (Shestakova et al., 1991; Bektas et al., 1994). The 101 kDa protein was originally marked as a phosphoprotein involved in the growth/differentiation transition at the PS-point, and was now identified as Dd-EF2H. This protein is known to be strongly labeled with 32Pi in growing and starving Ax-2 cells at most cell-cycle phases other than the PS-point, and was never phosphorylated around the PS-point under conditions of nutritional deprivation (Akiyama and Maeda, 1992). This suggested that a low phosphorylation level of Dd-EF2H might favor the entry of Ax-2 cells into differentiation from the PS-point. In this connection, it has been demonstrated that the activity of EF-2 in translation is regulated by its phosphorylation levels, and that the dephosphorylated state is generally the active form (Ryazanov et al., 1988). EF-2 is specifically phosphorylated by Ca2+/ calmodulin-dependent protein kinase III, known as EF-2 kinase (Nairn and Palfrey, 1987), and dephosphorylated by PP2A (polycation-stimulated serine/threonine-specific protein phophatase; phosphatase 2A) (Michael et al., 1989). It has been shown in Ax-2 cells that the phosphorylation level of Dd-EF2H around the PS-point is low in starvation medium, even in the presence of 0.5 µM calyculin A, a potent and specific inhibitor of the PP2A and PP1 (ATP-Mg2+-dependent serine/threoninespecific protein phosphatase) that is capable of completely inhibiting entry of cells (located just before the PS-point) into differentiation in response to starvation, and that dephosphorylation of a 32 kDa protein is perfectly inhibited by 0.5 µM calyculin A (Akiyama and Maeda, 1992). Thus, in addition to low activity of the serine/threonine protein kinases including EF-2 kinase under starvation conditions around the PS-point, pronounced dephosphorylation of the 32 kDa protein might be required for transition of Dictyostelium cells from

2654

Journal of Cell Science 116 (13)

growth to differentiation. Although the overexpression of Ddef2h is not striking (Fig. 1C), more rapid aggregation was achieved in Dd-ef2h-overexpressing cells (ef-2OE) than in parental Ax-2 cells. In contrast, gene inactivation of Dd-ef2h by homologous recombination or antisense RNA considerably impaired the progression of differentiation. Again, these results seem to indicate that the presence of dephosphorylated DdEF2H may be involved in the initiation of differentiation. We thank Richard H. Kessin and Dieter Malchow for their critical reading of the manuscript and insightful comments. We are also grateful to the Dictyostelium cDNA project in Japan for support, and to the Japan Society for the Promotion of Science for the kind gift of the cDNA clone SLE406. This work was supported by a Grant-in-Aid (No. 13874114 and No. 14654170) from the Ministry of Education, Science, Sports and Culture of Japan. This work was also funded by The Mitsubishi Foundation.

References Abe, F. and Maeda, Y. (1994). Precise expression of the cAMP receptor gene, CAR1, during transition from growth to differentiation in Dictyostelium discoideum. FEBS Lett. 342, 239-241. Abe, F. and Maeda, Y. (1995). Specific expression of a gene encoding a novel calcium-binding protein, CAF-1, during transition of Dictyostelium cells from growth to differentiation. Dev. Growth Differ. 37, 39-48. Akiyama, M. and Maeda, Y. (1992). Possible involvements of 101kDa, 90kDa, and 32kDa phosphoproteins in the phase-shift of Dictyostelium cells from growth to differentiation. Differentiation 51, 79-90. Bektas, M., Nurten, R., Gurel, Z., Sayers, Z. and Bermek, E. (1994). Interactions of eukaryotic elongation factor 2 with actin: a possible link between protein synthetic machinery and cytoskeleton. FEBS Lett. 356, 8993. Bonner, J. T. (1947). Evidence for the formation of cell aggregates by chemotaxis in the development of the cellular slime mold Dictyostelium discoideum. J. Exp. Zool. 106, 1-26. Bonner, J. T., Barkley, D. S., Hall, E. M., Konijn, T. M., Mason, J. W., O’Keefe, G. and Wolfe, P. B. (1969). Acrasin, acrasinase and the sensitivity to acrasin in Dictyostelium discoideum. Dev. Biol. 20, 72-87. Chae, S.-C. and Maeda, Y. (1998a). Preferential expression of the cDNA encoding the proteasome subunit during the growth/differentiation transition of Dictyostelium cells. Biochem. Biophys. Res. Commun. 245, 231-234. Chae, S.-C. and Maeda, Y. (1998b). Cloning and sequence analysis of the cDNA encoding elongation factor-1β of Dictyostelium discoideum. Biochim. Biophys. Acta 1383, 1-3. Chae, S.-C., Inazu, Y., Amagai, A. and Maeda, Y. (1998). Underexpression of a novel gene, dia2, impairs the transition of Dictyostelium cells from growth to differentiation. Biochem. Biophys. Res. Commun. 252, 278-283. Drummond, I. A., Lee, A. S., Resendez, E., Jr and Steinhardt, R. A. (1987). Depletion of intracellular calcium stores by calcium ionophore A23187 induces the gene for glucose-regulated proteins in hamster fibroblasts. J. Biol. Chem. 262, 12801-12805. Gerisch, G. (1961). Zellfunktionen und Zellfunktionwechsel in der Entwicklung von Dictyostelium discoideum. V. Stadienspezifische Zellkontaktbildung und ihre quantitative Erfassung. Exp. Cell Res. 25, 535554. Hirose, S., Inazu, Y., Chae, S.-C. and Maeda, Y. (2000). Suppression of the growth/differentiation transition in Dictyostelium development by transient expression of a novel gene, dia1. Development 127, 3263-3270. Howard, P. K., Aber, K. G. and Firtel, R. A. (1988). Establishment of a transient expression system for Dictyostelium discoideum. Nucleic Acids Res. 16, 2613-2633. Inazu, Y., Chae, S.-C. and Maeda, Y. (1999). Transient expression of a

mitochondorial gene cluster including rps4 is essential for the phase-shift of Dictyostelium cells from growth to differentiation. Dev. Genet. 25, 339352. Itoh, M., Noguchi, M. and Maeda, Y. (1998). Overexpression of CAF1 encoding a novel Ca2+-binding protein stimulates the transition of Dictyostelium cells from growth to differentiation. Dev. Growth Differ. 40, 677-683. Kohno, K., Uchida, T., Ohkubo, H., Nakanishi, T., Fukui, T., Ohtsuka, E. Ikehara, M. and Okada, Y. (1986). Amino acid sequence of mammalian elongation factor 2 deduced from the cDNA sequence: homology with GTPbinding proteins. Proc. Natl. Acad. Sci. USA 83, 4978-4982. Livingston, D. M. and Bodley, J. W. (1992). Saccharomyces cerevisiae elongation factor 2. J. Biol. Chem. 267, 1190-1197. Maeda, Y. (1986). A new method for inducing synchronous growth of Dictyostelium discoideum cells using temperature shifts. J. Gen. Microbiol. 132, 1189-1196. Maeda, Y., Ohmori, T., Abe, T., Abe, F. and Amagai, A. (1989). Transition of starving Dictyostelium cells to differentiation phase at a particular position of the cell cycle. Differentiation 41, 169-175. Mendoza, A., Serramia, M. J., Capa, L. and Garcia-Bustos, J. F. (1999). Translation elongation factor 2 is encoded by a single essential gene in Candida albicans. Gene 229, 183-191. Michael, G., Walter, K., Gottfried, M. and Friedrich, M. (1989). A type 2A protein phosphatase dephosphorylates the elongation factor 2 and is stimulated by the phorbol ester TPA in mouse epidermis in vivo. FEBS Lett. 257, 357-360. Morita, T., Saitoh, K., Takagi, T. and Maeda, Y. (2000). Involvement of the glucose-regulated protein 94 (Dd-GRP94) in starvation response of Dictyostelium discoideum cells. Biochem. Biophys. Res. Commun. 274, 323331. Nairn, A. C. and Palfrey, H. C. (1987). Identification of the major Mr 100,000 substrate for calmodulin-dependent protein kinase III in mammalian cells as elongation factor-2. J. Biol. Chem. 262, 17299-17303. Nellen, W., Datta, S., Reymond, C., Silvertsen, A., Mann, S., Crowley, T. and Firtel, R. A. (1987). Molecular biology in Dictyostelium: tools and applications. Methods Cell Biol. 28, 67-100. Nilsson, L. and Nygard, O. (1985). Structural and functional studies of the interaction of the eukaryotic elongation factor EF-2 with GTP and ribosomes. Eur. J. Biochem. 148, 299-304. Okafuji, T., Abe, F. and Maeda, Y. (1997). Antisense-mediated regulation of Annexin VII gene expression during the transition from growth to differentiation in Dictyostelium discoideum. Gene 189, 49-56. Perentesis, J. P., Phan, L. D., Gleason, W. B., la Porte, D. C., Livingston, D. M. and Bodley, J. W. (1992). Saccharomyces cerevisiae elongation factor 2. Genetic cloning, characterization of expression, and G-domain modeling. J. Biol. Chem. 267, 1190-1197. Pouyssegur, J., Shiu, R. P. C. and Pastan, I. (1977). Induction of two transformation-sensitive membrane polypeptides in normal fibroblasts by a block in glycoprotein synthesis or glucose deprivation. Cell 11, 841847. Ryazanov, A. G., Shesakova, E. A. and Natapov, P. G. (1988). Phosphorylation of elongation factor 2 by EF-2 kinase affects rate of translation. Nature 334, 170-173. Shestakova, E. A., Motuz, L. P., Minin, A. A., Gelfand, V. I. and Gavrilova, L. P. (1991). Some of eukaryotic elongation factor 2 is colocalized with actin microfilament bundles in mouse embryo fibroblasts. Cell Biol. Int. Rep. 15, 75-84. Shiu, R. P. C., Pouyssegur, J. and Pastan, I. (1977). Glucose depletion accounts for the induction of two transformation-sensitive membrane proteins in Rous sarcoma virus-transformed chick embryo fibroblasts. Proc. Natl. Acad. Sci. USA 74, 3840-3844. Toda, K., Tasaka, M., Mashima, K., Kohno, K., Uchida, T. and Takeuchi, I. (1989). Structure and expression of elongation factor 2 gene during development of Dictyostelium discoideum. J. Biol. Chem. 264, 1548915493. Weissbach, H. and Ochoa, S. (1976). Soluble factors required for eukaryotic protein synthesis. Annu. Rev. Biochem. 45, 191-216.