Ethylene as a potent inducer of sexual development


Author to whom all correspondence should be addressed.


A novel and critical function of ethylene, a potent plant hormone, has been well documented in Dictyostelium, because it leads cells to the sexual development (macrocyst formation) by inducing zygote formation. Zygote formation (sexual cell fusion) and the subsequent nuclear fusion are the characteristic events occurring during macrocyst formation. A novel gene, zyg1 was found to be predominantly expressed during the sexual development, and its enforced expression actually induces zygote formation. As expected, the zygote inducer, ethylene enhances the expression of zyg1. Thus the function of ethylene has been verified at all of individual (macrocyst formation), cellular (zygote formation), and molecular levels (zyg1 expression). Based on our recent studies concerning the behavior and function of the zyg1 product (ZYG1 protein), the signal transduction pathways involved in zygote formation are proposed in this review.


The cellular slime mold exhibits dimorphism in development: sorocarp formation as an asexual development and macrocyst formation as a sexual development. These two developmental forms are regulated by the environmental conditions, such as light and water. Dm-7, one of the strains of Dictyostelium mucoroides forms sorocarps in the light, while they form macrocysts in the dark or water (Filosa 1979). In the asexual development, amoeboid cells grow and multiply feeding on bacteria as a food supply. Upon exhaustion of the bacterial food supply (starvation), starving cells stop growing and start the differentiation process. They gather together to form cell aggregates. A tip is formed on the top of each cell aggregate, which then migrates as a slug-shaped mass. After the migration, the slug dramatically changes its shape to form a sorocarp consisting of a stalk with an apical mass of spores.

In the sexual development, macrocyst formation occurs. Macrocyst formation is characterized by the formation of large aggregates, which are subdivided into smaller masses (precysts), each of which is surrounded by a fibrillar sheath. At the center of each precyst there arises a cytophagic cell (giant cell), which in turn engulfs all the other cells in the precyst. The engulfed cells (endocytes) are eventually broken down into granular remnants. The enlarged and cytophagic cell finally becomes surrounded by a thick wall to form the mature macrocyst (Filosa & Dengler 1972). After a resting period, the macrocyst germinates to release several amoeboid cells and initiates a new life cycle (Nickerson & Raper 1973). It was found that the macrocyst formation was induced by a potent plant hormone, ethylene (Amagai 1984).

In the life cycles of plants, ethylene is well known to regulate many aspects, including seed germination, root initiation, flower development, fruit ripening, senescence, and responses to several stresses (Abeles et al. 1992). Recently, a lot of new genes involved in ethylene biosynthesis, signal transduction, and response pathways have been isolated and identified (Lin et al. 2009). In general, ethylene is synthesized from methionine through s-adenosyl-l-methionine (SAM) and 1-aminocyclopropane-1-carboxylic acid (ACC) (Adams & Yang 1979). It has also been suggested in Dictyostelium cells that ethylene may be biosynthesized through the same pathway as that in plants (Amagai & Maeda 1992). In fact, Dictyostelium homologue genes of ACC synthase and ACC oxidase have been isolated by the Japanese Dictyostelium cDNA project and the genome project of Dictyostelium (Eichinger et al. 2005). In addition, the existence of ethylene receptor(s) in Dictyostelium cells has been indicated, though its identification remains to be clarified (Amagai et al. 2007). Since Dictyostelum is an excellent model organism for investigation of various developmental aspects, studies using this organism must be promising to precisely estimate various novel functions of ethylene.

In this review, I describe how the two developmental forms are regulated, especially focusing on the function of ethylene as a potent regulator at the individual, cellular, and molecular levels. The molecular mechanism for induction of zygote formation is also discussed, correlating to the function of a novel protein, ZYG1 whose expression is augmented by ethylene.

Macrocyst formation is the real sexual process in the cellular slime molds

Brefeld (1869) first reported the macrocyst formation: Its process was regarded as a sexual process in the cellular slime molds. Indeed, the shape of a matured macrocyst looks like a sea urchin egg (Fig. 1). There are two kinds of mating systems in the macrocyst formation: homothallic and heterothallic mating types (Clark et al. 1973; Erdos et al. 1973). Dm-7 cells described mainly in this review form zygotes homothallically, while D. discoideum belongs to the heterothallic mating type. There is a lot of evidence concerning the sexuality of macrocysts (Okada et al. 1986; O’Day et al. 1987a; Amagai 1989). Although the period of meiosis still remains to be specified, the appearance of synaptonemal complex, which is formed at the stage of late leptotene during meiosis is reported during macrocyst formation in some strains of cellular slime molds, such as Polysphondylium violaceum (Erdos et al. 1972) and D. discoideum (Okada et al. 1986). Furthermore, from various crossbreeding experiments, the occurrence of recombinants during macrocyst formation has been realized in Dm-7, its mutant (MacInnes & Francis 1974), D. giganteum (Erdos et al. 1975), P. pallidum and D. discoideum (Francis 1975; Wallace & Raper 1979). These results strongly suggest that meiosis may occur during macrocyst formation.

Figure 1.

 A cluster of macrocysts formed from Dictyostelium mucoroides-7 (Dm-7) cells. Dm-7 cells were starved and cultured for 6 days under the submerged conditions. Each of the macrocysts is surrounded with a thick wall. Bar, 100 μm.

Recently, giant cells formed during macrocyst formation in Dm-7 have been proved to be zygotes that are produced by cell fusion and the subsequent nuclear fusion (Amagai 1989). That is, when Dm-7 cells stained vitally with 4´6´-diamidino-2-phenylindole dihydrochloride (DAPI) were mixed with cells stained vitally with fluorescein isothiocyanate (FITC), a significant number of giant cells containing both DAPI- and FITC-stains were formed, thus indicating that the giant cells are formed by the cell fusion between the DAPI- and FITC-stained cells. Also, some of the giant cells were found to have an enlarged and brightly DAPI-stained nucleus instead of bi-nuclei. Since an enlarged and brightly DAPI-stained nuclei contained twice the amount of DNA, it is evident that nuclear fusion occurs between two nuclei existing in a giant cell. Evidence showing sexual cell fusion and the subsequent nuclear fusion occurring during macrocyst formation has also been reported in heterothallic strains. Fusion of two gametes has been photographed by a time-lapse video recorder in D. discoideum (O’Day et al. 1987b). The appearance of nuclei containing a twofold DNA content has been also noticed in D. discoideum (Okada et al. 1986). Taken together these results indicate that the giant cell is a true zygote produced by cell fusion and the subsequent nuclear fusion, in both kinds of mating behavior, homothallic and heterothallic.

What are chemical regulators for determining the developmental pathway toward macrocyst or sorocarp formation?

Two developmental pathways, sorocarp and macrocyst formation, are regulated by several environmental conditions, such as light and water. These environmental conditions affect the synthesis of chemical regulators within the cells. Ethylene and cyclic adenosine monophosphate (cAMP) have been demonstrated as chemical regulators for the choice of developmental pathways in Dm-7 (Amagai 1984; Amagai & Filosa 1984). Ethylene was found to function as a potent inducer of macrocyst formation, while cAMP as an inhibitor. These findings have been minutely described in my recently published review (Amagai 2009).

The tight relationship between the amount of ethylene and the induction of macrocyst formation was confirmed directly, using two kinds of transformants over- and under-producing ethylene (Amagai et al. 2007). Dd-aco, an 1-aminocyclopropane-1-carboxylic acid (ACC) oxidase homologue gene, isolated from D. discoideum (DDBJ, EMBL and GenBank databases with the accession no. AB105858) was introduced into Dm-7 cells to obtain transformants overproducing ethylene. As ACC oxidase catalyzes the last step in the biosynthesis of ethylene, transformants overexpressing Dd-aco were expected to increase the production of ethylene. Actually, when the production of ethylene was determined by gas chromatography, transformants overexpressing Dd-aco (ACOOE) were found to produce a larger amount of ethylene as compared with wild type, Dm-7. On the other hand, underexpression of Dd-aco (ACO-RNAi) by means of the RNAi method caused a smaller amount of ethylene production, as compared with Dm-7. Here it is of importance to note that ACOOE cells exhibit active macrocyst formation through enhanced zygote formation, while ACO-RNAi cells fail to form macrocysts even under the conditions favorable to macrocyst formation. Therefore, it is most likely that the choice of developmental form is determined by the amounts of ethylene and cAMP produced at the aggregate stage when the developmental fate is determined. In this connection, when the amounts of ethylene and cAMP were estimated at the aggregation stage, it was found that the amount of ethylene is higher under macrocyst-forming conditions, but that the amount of cAMP is lower in macrocyst formation than in sorocarp formation (Amagai 1987). Thus, the developmental fate toward macrocyst formation is dictated by the higher amount of ethylene and the lower amount of cAMP at the aggregation stage. In contrast, the sorocarp formation is directed by the lower amount of ethylene and the higher amount of cAMP (Amagai 1989; Suzuki et al. 1992).

Volatile substance(s) released from NC-4 cells of D. discoideum is also known to induce macrocyst formation in V12M2 without the help of its mating type NC-4 (Lewis & O’Day 1977). Since the ethylene production by NC-4 cells was confirmed by gas chromatography (Bonner 1973; Amagai & Maeda 1992), it is quite possible that the volatile substance produced by NC-4 cells may be ethylene by which the macrocyst formation is induced in V12M2 cells.

How do the developmental regulators such as ethylene and cAMP act at the cellular level?

When the nucleotide sequence of ACC oxidase homologue gene isolated from Dm-7 (Dm-aco) (accession number, AB291210) was compared with that of Dd-aco derived from Dd, only one nucleotide, thymidine at the 672nd in Dd-aco was found to be replaced with cytosine in Dm-aco. Since the deduced amino acid sequences encoded by the two genes are completely identical (Amagai 2007), it must be a small matter for Dd-aco to be introduced in Dm-7 cells. However, to further make sure that Dd-aco acts correctly in Dm-7 cells, Dm-aco was introduced into Dm-7 cells. As expected, a transformant (Dm-ACOOE) overexpressing Dm-aco formed giant cells with many nuclei as a result of cell fusion (Fig. 2A), coupling with the enforced expression of Dm-aco mRNA (Fig. 2B) and the larger amount of ethylene production compared to parental Dm-7 cells (Fig. 2C).

Figure 2.

 (A) Giant multinucleate cells induced by overexpression of 1-aminocyclopropane-1-carboxylic acid (ACC) oxidase homologue in Dm7 cells. Dm-aco, isolated from the genomic DNA of Dm-7 by polymerase chain reaction (PCR) (Amagai 2007), was inserted into pDneo67 vector (da Silva & Klein 1990). Dm-7 and its transformant (Dm-ACOOE) overexpressing Dm-aco were developed by shaking for 1.5 h after starvation, fixed with methanol, and stained with 4′6′-diamidino-2-phenylindole dihydrochloride (DAPI). Although a fair number of binucleate cells are observed in parental Dm-7 cells, giant cells with numerous nuclei are formed in Dm-ACOOE (arrows). Bar, 50 μm. (B) Expressions of Dm-aco in Dm-7 and its transformant (Dm-ACOOE). Total RNAs were separately extracted from Dm 7 and Dm-ACOOE (Dm-ACOOE) cells starved for 2 h by shake culture, followed by Northern blotting (upper panel). It is evident that the Dm-aco mRNA is expressed in Dm-ACOOE cells, but not in Dm-7 cells. Et-Br stainings of total RNAs extracted from Dm7 and Dm-ACOOE cells are shown in the lower panel, respectively. (C) Enhanced production of ethylene in Dm-ACOOE cells. Dm-ACOOE and parental Dm-7 cells were separately grown with Escherichia coli (OD = 14 at 450 nm), collected and suspended in BSS (Bonner’s salt solution; Bonner 1947) at 2 × 108 cells/mL after washing by repeated centrifugation. One milliliter of cell suspension (2 × 108 cells/mL) was put into a 20 mL Erlenmeyer flask. The flasks were sealed tightly with silicone stoppers on which silicone grease had been coated, wrapped with aluminum foil and shaken rapidly at 175 rpm. After 24 h of incubation at 22°C, 2 mL of the air/gas mixture were collected by a syringe through the silicone stopper and assayed by gas chromatography (Shimazu, GC18A) using a Porapak N column operated at 40°C with nitrogen as a carrier gas. The quantity of ethylene peaks was determined by a printing integrator attached to the apparatus. The amount of ethylene released from Dm-ACOOE cells is shown as relative values to that from Dm7 cells which is designated as 100%. Three samples of two independent experiments were determined. Bars, standard deviations.

In heterothallic strains, zygote formation is also regulated positively by ethylene and negatively by cAMP (O’Day & Lydan 1989; Amagai 1992). Accordingly, ethylene and cAMP may regulate zygote formation in homothallic and heterothallic cells. The results shown in Figure 2 indicate that enforced ethylene production actually augments sexual cell fusion during the process of zygote formation. However, it is presently unknown whether or not ethylene also controls other cellular events occurring in zygote formation, such as gamete formation and nuclear fusion.

Signal transduction pathways involved in zygote formation

It is well known that calcium ion (Ca2+) plays an important role in cell fusion in many organisms (Shainberg et al. 1969; Ishihara et al. 1984). Ca2+ has also been proposed to be a critical factor for zygote formation including cell fusion in Dictyostelium. The percentage of zygotes is elevated by the presence of increased extracellular Ca2+ in D. discoideum and Dm-7 (Chagla et al. 1980; Szabo et al. 1982; Suzuki et al. 1992). In this connection, phorbol esters such as 12O-tetradecanoylphorbol-13-acetate (TPA), potent activators of protein kinase C (PKC), enhance the formation of zygotes in D. discoideum (Gunther et al. 1995). In contrast, staurosporine, an inhibitor of protein kinases including PKC, inhibites zygote formation in D. discoideum (Gunther et al. 1995) and macrocyst formation in Dm-7 (Kawai et al. 1993). Since the involvement of PKC like kinase activity in zygote formation has been mainly claimed in D. discoideum, the effect of PKC like kinase on zygote formation in Dm-7 was examined. As a result, it was found that the zygote formation is promoted by application of TPA, but is inhibited by staurosporine, as in the case of D. discoideum (Fig. 3), thus suggesting that the signaling pathway involving Ca2+ and PKC like protein may regulate zygote formation in both of D. discoideum and Dm-7.

Figure 3.

 Effects of 12O-tetradecanoylphorbol-13-acetate (TPA) and staurosporine on zygote formation of Dm-7 cells. Dm-7 cells (5 × 106 cells/mL) were shaken in BSS (Bonner’s salt solution) with or without either 5 nmol/L TPA or 2.2 μmol/L staurosporine. After 9 h of shake culture, the starved cells were collected, fixed with methanol, and stained with 4´6´-diamidino-2-phenylindole dihydrochloride (DAPI). The percentages of fused cells were calculated as a fusion index from the following formula: inline image Fusion index (%) = (the cell number of multi-nucleate cells/ the total cell number) ×100. Three samples in two independent experiments were counted. Bars, standard deviations. From this figure, it is clear that the zygote formation is significantly enhanced by TPA, but markedly suppressed by staurosporine.

To examine possible involvements of other protein kinases in zygote formation, effects of some kinase inhibitors on macrocyst formation were tested. The results obtained have demonstrated that calmodulin- and cAMP-dependent kinase (PKA) inhibit zygote formation, especially the process of gamete formation, through the signaling pathway triggered by cAMP (Kawai et al. 1993). Lydan & O’Day (1988) have reported the evidence showing roles of calmodulin as both a negative (for gamete formation) and a positive (for cell fusion) regulator of sexual cellular events in D. discoideum. Calmodulin-dependent phosphorylation and dephosphorylation involved in the zygote formation have been also reported by Lydan & O’Day (1993a). They have summarized the roles of Ca2+ and calmodulin in the signal transduction pathway in their book (Lydan & O’Day 1993b). Although proteins phosphorylated or dephosphorylated in a calmodulin-dependent manner as well as the target proteins for PKC remain to be identified, it has been recently proposed by our recent studies that a novel protein, ZYG1, is a promising target protein for PKC.

Using the differential screening of genes, the zyg1 gene was isolated from Dm-7 cells as a novel gene expressed predominantly during the macrocyst formation (DDBJ/EMBL/GenBank, accession no. AB006956) (Amagai 2002). The predicted protein, ZYG1, consists of 268 amino acids with a molecular mass of 29.4 kDa. After BLAST (Altschul et al. 1990) and FASTA (Pearson 1990) searches, the amino acid sequence as a whole shows no convincing similarity to the known proteins. As the expression pattern of zyg1 is quite similar to the developmental kinetics of zygote formation with about 1 h of precedence, zyg1 was expected to be closely involved in zygote formation (Kawai et al. 1993; Amagai 2002). Incidentally, transformants overexpressing the zyg1 gene formed the macrocysts on agar even in the light condition, under which parental Dm-7 cells were destined to develop toward sorocarp formation. The transformants also formed a number of giant cells in addition to macrocysts. Accordingly, it is evident that ZYG1, the product of zyg1, is tightly implicated for zygote formation. Using transformants overexpressing Dm-aco (ACO-OE), we have demonstrated that ethylene induces the expression of zyg1 (Amagai 2007). Since ZYG1 protein has several predicted sites phosphorylated by PKC, it is quite possible that ZYG1 protein may be one of the major substrates for PKC, and that phosphorylated ZYG1 may induce zygote formation.

Several genes involved in sexual development have been isolated from D. discoideum (Urushihara & Muramoto 2006). GmsA, one of the genes isolated, has been shown to positively relate to sexual cell fusion (Muramoto et al. 2003), though its precise function remains to be elucidated.


I would like to propose that ZYG1 protein induced by ethylene is a critical substrate for PKC in the Ca2+- and PKC-mediated signal transduction pathway involved in zygote formation. Our preliminary experiments have demonstrated that ZYG1 is actually phosphorylated by PKC at the region where sexual cell fusion occurs (unpubl. data). Based on these findings, a possible network of signals working during zygote formation is schematically shown in Figure 4.

Figure 4.

 A likely model of the signal transduction pathways involved in zygote formation. From several pharmacological studies, cyclic adenosine monophosphate (cAMP) and protein kinase A (PKA) have been shown to function as an inhibitory factor for zygote formation (Suzuki et al. 1992; Kawai et al. 1993). On the other hand, Lydan & O’Day (1988, 1993a,b) have reported evidence showing the roles of calmodulin as both a negative regulator of gamete formation and a positive regulator of sexual cell fusion (zygote formation) in Dictyostelium. We have recently demonstrated that ZYG1 is phosphorylated by Ca2+-activated PKC, thus resulting in induction of zygote formation. Induction, inline image; Inhibition, inline image.

It has been reported that the Ca2+- and PKC-mediated signaling pathway is involved in cell fusion during myogenesis (Shainberg et al. 1969; Paterson & Strohman 1972; David et al. 1990). The cell fusion as observed in the fertilization and myogenesis may share the Ca2+- and PKC-mediated signaling pathway. Importantly, we have recently found that the cell fusion of mouse myoblasts to form myotubes is induced by introduction of the Dictyostelium zyg1 gene into myoblasts (unpubl. data). From an evolutional standpoint, these data give us an idea that the Ca2+ -and PKC-mediated signal transduction pathway is generally involved in the cell fusion and well conserved beyond the difference of species. Cellular and molecular events induced by ethylene are of particular importance for understanding the mechanism of sexual development in Dictyostelium. Such an analysis will also contribute to the comprehension of the basal mechanism of cell fusion.


I thank Dr Michael F. Filosa, Dr Yasuo Maeda (Tohoku University), and all of my colleagues for their encouragement during the works presented here. I also thank Dr Jean-Paul Rieu (Lyon University) for his critical reading and insightful comments on the manuscript.