Combinatorial control of cell differentiation by cAMP and DIF-1 during development of Dictyostelium discoideum

At least three distinct types of cell arise from a population of similar amoebae during Dictyostelium development: prespore, prestalk A and prestalk B cells. We report evidence suggesting that this cellular diversification can be brought about by the combinatorial action of two diffusible signals, cAMP and DIF-1. Cells at different stages of normal development were transferred to shaken suspension, challenged with various combinations of signal molecules and the expression of cell-type-specific mRNA markers measured 1-2 h later. pDd63, pDd56 and D19 mRNAs were used for prestalk A, prestalk B and prespore cells respectively. We find the following results. (1) Cells first become responsive to DIF-1 for prestalk A differentiation and to cAMP for prespore differentiation at the end of aggregation, about 2 h before these cell types normally appear. (2) At the first finger stage of development, when the rate of accumulation of the markers is maximal, the expression of each is favoured by a unique combination of effectors: prespore differentiation is stimulated by cAMP and inhibited by DIF-1; prestalk A differentiation is stimulated by both cAMP and DIF-1 and prestalk B differentiation is stimulated by DIF-1 and inhibited by cAMP. (3) Half-maximal effects are produced by 10-70 nM DIF-1, which is in the physiological range. (4) Ammonia and adenosine, which can affect cell differentiation in other circumstances, have no significant pathway-specific effect in our conditions. These results suggest that cell differentiation could be brought about in normal development by the localized action of cAMP and DIF-1.     

The role of DIF-1 signaling in Dictyostelium development

We have constructed a mutant blocked in the biosynthesis of DIF-1, a chlorinated signal molecule proposed to induce differentiation of both major prestalk cell types formed during Dictyostelium development. Surprisingly, the mutant still forms slugs retaining one prestalk cell type, the pstA cells, and can form mature stalk cells. However, the other major prestalk cell type, the pstO cells, is missing. Normal pstO cell differentiation and their patterning in the slug are restored by development on a uniform concentration of DIF-1. We conclude that pstO and pstA cells are in fact induced by separate signals and that DIF-1 is the pstO inducer. Positional information, in the form of DIF-1 gradients, is evidently not required for pstO cell induction.     

Cell-fate choice in Dictyostelium: intrinsic biases modulate sensitivity to DIF signaling

Cell fate in Dictyostelium development depends on intrinsic differences between cells, dating from their growth period, and on cell interactions occurring during development. We have sought for a mechanism linking these two influences on cell fate. First, we confirmed earlier work showing that the vegetative differences are biases, not commitments, since cells that are stalky-biased when developed with one partner are sporey with another. Then we tested the idea that these biases operate by modulating the sensitivity of cells to the signals controlling cell fate during development. Cells grown without glucose are stalky-biased when developed with cells grown with glucose. We find, using monolayer culture conditions, that they are more sensitive to each of the stalk-inducing signals, DIFs 1-3. Mixing experiments show that this bias is a cell-intrinsic property. Cells initiating development early in the cell cycle are stalky compared to those initiating development later in the cycle. Likewise, they are more sensitive to DIF-1. Assays of standard markers for prestalk and prespore cell differentiation reveal similar differences in DIF-1 sensitivity between biased cells; DIF-1 dechlorinase (an early prestalk cell marker enzyme) behaves in a consistent manner. We propose that cell-fate biases are manifest as differences in sensitivity to DIF.     

Cross-induction of cell types in Dictyostelium: evidence that DIF-1 is made by prespore cells.

To investigate how cell type proportions are regulated during Dictyostelium development, we have attempted to find out which cell type produces DIF-1, a diffusible signal molecule inducing the differentiation of prestalk-O cells. DIF-1 is a chlorinated alkyl phenone that is synthesized from a C12 polyketide precursor by chlorination and methylation, with the final step catalysed by the dmtA methyltransferase. All our evidence points to the prespore cells as the major source of DIF-1. (1) dmtA mRNA and enzyme activity are greatly enriched in prespore compared with prestalk cells. The chlorinating activity is also somewhat prespore-enriched. (2) Expression of dmtA is induced by cyclic-AMP and this induction is inhibited by DIF-1. This regulatory behaviour is characteristic of prespore products. (3) Short-term labelling experiments, using the polyketide precursor, show that purified prespore cells produce DIF-1 at more than 20 times the rate of prestalk cells. (4) Although DIF-1 has little effect on its own synthesis in short-term labelling experiments, in long-term experiments, using 36Cl(-) as label, it is strongly inhibitory (IC(50) about 5 nM), presumably because it represses expression of dmtA; this is again consistent with DIF-1 production by prespore cells. Inhibition takes about 1 hour to become effective. We propose that prespore cells cross-induce the differentiation of prestalk-O cells by making DIF-1, and that this is one of the regulatory loops that sets the proportion of prespore-to-prestalk cells in the aggregate.     

Signals controlling cell differentiation and pattern formation in Dictyostelium

The major inducers of cell differentiation in Dictyostelium appear to be cyclic AMP and DIF-1. Recently we have chemically identified DIF-1, together with the closely related DIF-2 and -3. They represent a new chemical class of potent effector molecules, based on a phenyl alkanone with chloro, hydroxy, and methoxy substitution of the benzene ring. Previous work has shown that DIF-1 can induce prestalk-specific gene expression within 15 min, whereas it suppresses prespore differentiation. Hence, DIF-1 can control the choice of pathway of cell differentiation in Dictyostelium and is therefore likely to be involved in establishing the prestalk/prespore pattern in the aggregate. In support of this, we show that DIF treatment of slugs results in an enlarged prestalk zone. Cyclic AMP seems less likely to have such a pathway-specific role, but later in development it becomes inhibitory to stalk cell differentiation. This inhibition may be important in suppressing terminal stalk cell differentiation until culmination. Spore differentiation can be induced efficiently by high levels of Br-cyclic AMP, a permeant analogue of cyclic AMP. In this, it phenocopies certain spore-maturation mutants, and we propose that during normal development spore differentiation is triggered by an elevation in intracellular cyclic AMP levels. How this elevation in cyclic AMP levels is brought about is not known. The experiments with Br-cyclic AMP also provide the first direct evidence that elevated levels of intracellular cyclic AMP induce differentiation in Dictyostelium.     

Cell differentiation and patterning in Dictyostelium.

In Dictyostelium development, prestalk cells first differentiate at scattered positions in the aggregate and then sort out, probably by chemotaxis to cAMP. They may regulate their proportions by selective depletion of the stalk cell inducer, DIF-1. Once sorted, prestalk cells form a DIF-1 sink, which can produce gradients of DIF-1 and its metabolites in the slug. Global movements of cells in the slug may be regulated by cAMP signals, as in aggregation. Terminal differentiation of stalk and spore cells requires activation of cAMP-dependent protein kinase, possibly brought about by ammonia depletion. Finally, a technique for insertional mutagenesis promises the ready isolation of developmental genes.     

A possible role for DIF-2 in the formation of stalk cells during Dictyostelium development.

The differentiation inducing factor (DIF) is essential for stalk cell formation in monolayers of Dictyostelium discoideum and is necessary for the expression of several prestalk cell-specific genes. DIF activity has been fractionated into a major species, designated DIF-1, and several minor species, including DIF-2. Although DIF-1 is an excellent inducer of stalk cell formation from vegetative cells, it is a poor inducer of stalk cell formation from prestalk cells. In contrast, DIF-2 is more active for the conversion of prestalk cells into stalk cells, than for the conversion of vegetative cells to stalk cells. The same results were obtained regardless of whether chemically synthesized or naturally occurring components were utilized. In addition, stalk cell formation was three- to fourfold higher when vegetative cells were incubated with DIF-1 for a suboptimal period and then subsequently incubated with DIF-2, than when cells were incubated with DIF-2 first and then subsequently with DIF-1. These results indicate a distinct role for DIF-2 during stalk cell formation and suggest the possibility that DIF-1 and DIF-2 act sequentially.     

Cyclic AMP is an inhibitor of stalk cell differentiation in Dictyostelium discoideum

Cyclic AMP and DIF-1 (1-(3,5-dichloro-2,6-dihydroxy-4-methoxyphenyl)-1-hexanone) together induce stalk cell differentiation in vitro in Dictyostelium discoideum strain V12M2. The induction can proceed in two stages: in the first, cyclic AMP brings cells to a DIF-responsive state; in the second, DIF-1 alone can induce stalk cell formation. We report here that during the DIF-1-dependent stage, cyclic AMP is a potent inhibitor of stalk cell differentiation. Addition of cyclic AMP at this stage to V12M2 cells appreciably delays, but does not prevent, stalk cell formation. In contrast, stalk cell differentiation in the more common strain NC4 is completely suppressed by the continued presence of cyclic AMP. This fact explains earlier failures to induce stalk cells in vitro in NC4. We now consistently obtain efficient stalk cell induction in NC4 by removing cyclic AMP in the DIF-1-dependent stage. Cyclic AMP also inhibits the production of a stalk-specific protein (ST310) in both NC4 and a V12M2 derivative. Adenosine, a known antagonist of cyclic AMP action, does not relieve this inhibition by cyclic AMP and does not itself promote stalk cell formation. Finally, stalk cell differentiation of NC4 cells at low density appears to require factors in addition to cyclic AMP and DIF-1, but their nature is not yet known. The inhibition of stalk cell differentiation by cyclic AMP may be important in establishing the prestalk/prespore pattern during normal development, and in preventing the maturation of prestalk into stalk cells until culmination.     

Evidence that a cell-type-specific efflux pump regulates cell differentiation in Dictyostelium

We have identified a cellular efflux pump, RhT, with the properties of an MDR transporter-a type of ATP-binding cassette transporter whose substrates include small hydrophobic molecules. RhT transports rhodamine 123 (Rh123) and is inhibited by low temperature, energy poisons, and several MDR transport inhibitors, such as verapamil. All vegetative cells have RhT activity, but during development prestalk cells lose RhT activity while prespore cells retain it. We also identified several RhT inhibitors. The most effective inhibitor is the stalk cell-inducing chlorinated alkyl phenone, DIF-1. The RhT inhibitors disrupted development, to varying degrees, and induced stalk cell formation in submerged culture. The inhibitors displayed the same rank order of pharmacological efficacy for stalk cell induction as they did for Rh123 transport inhibition. We also found that cerulenin, a specific inhibitor of DIF-1 biosynthesis (R. R. Kay, 1998, J. Biol. Chem. 273, 2669-2675), abolished the induction of stalk cells by each of the RhT inhibitors, and this effect could be reversed by DIF-1. Thus, DIF-1 synthesis appears to be required for the induction of stalk cells by the RhT inhibitors. Since DIF-1 is the most potent inhibitor of RhT activity, and thus a likely transport substrate itself, we propose that RhT inhibitors induce stalk cell differentiation by blocking DIF-1 export, causing DIF-1 to build up within cells. Our results provide evidence for a prespore-specific efflux pump that regulates cell fate determination, perhaps by regulating the cellular concentration of DIF-1.     

Identification of new differentiation inducing factors from Dictyostelium discoideum

Developing Dictyostelium discoideum amoebae form a stalked fruiting body in which individual cells differentiate into either stalk cells or spores. The major known inducer of stalk cell differentiation is the chlorinated polyketide DIF-1 (1-(3,5-dichloro-2,6-dihydroxy-4-methoxyphenyl)hexan-1-one); however a mutant blocked in the terminal step of DIF-1 biosynthesis still produces one of the prestalk cell subtypes – the pstA cells – as well as some mature stalk cells. We therefore searched for additional stalk cell-inducing factors in the medium supporting development of this mutant. These factors were purified by solvent extraction and HPLC and identified by mass spectroscopy and NMR. The mutant lacked detectable DIF-2 and DIF-3 (the pentanone and deschloro homologues of DIF-1) but four major stalk cell-inducing activities were detected, of which three were identified. Two compounds were predicted intermediates in DIF-1 biosynthesis: the desmethyl, and desmethyl-monochloro analogues of DIF-1 (dM-DIF-1 and Cl-THPH, respectively), supporting the previously proposed pathway of DIF-1 biosynthesis. The third compound was a novel factor and was identified as 4-methyl-5-pentylbenzene-1,3-diol (MPBD) with the structure confirmed by chemical synthesis. To investigate the potential roles of these compounds as signal molecules, their effects on morphological stalk and spore differentiation were examined in cell culture. All three induced morphological stalk cell differentiation. We found that synthetic MPBD also stimulated spore cell differentiation. Now that these factors are known to be produced and released during development, their biological roles can be pursued further.     

The effects of presumptive morphogens on prestalk and prespore cell gene expression in monolayers of Dictyostelium discoideum

Abstract. The expression of three prestalk cell-specific genes ( ecm A, ecm B and pDd26) was examined during in vitro differentiation in cell monolayers, in an attempt to explain the spatial heterogeneity of the prestalk region of migrating Dictyostelium pseudoplasmodia. Under these conditions ecm A, ecm B and pDd26 mRNAs were expressed sequentially in response to the addition of differentiation inducing factor-1 (DIF)-1, a temporal sequence similar to that observed during normal development. ecm A and ecm B mRNAs reached a maximum level 2–4 h after DIF-1 supplementation and then declined, whereas pDd26 mRNA levels increased more slowly but remained high 24 h after DIF addition. The increases in expression in response to increasing concentrations of either DIF-1 or DIF-2 were identical for the three genes, suggesting that neither alteration in DIF concentration nor species was an important determinant of spatial heterogeneity. Ammonia had the same inhibitory effect on the expression of all three prestalk cell-specific genes and stimulated the expression of the prespore cell-specific gene, D19. These results indicate that ammonia is also not responsible for the spatial heterogeneity of the prestalk cell region. In contrast, cyclic AMP had a differential effect on the expression of the prestalk cell specific genes: ecm A expression was variably stimulated, pDd26 expression was inhibited and ecm B expression was sometimes stimulated and sometimes inhibited. These results are difficult to explain in terms of a gradient of cyclic AMP in the prestalk region. We postulate that temporal responses are more important than spatial responses to cyclic AMP in regulating stalk cell differentiation.     

Identification of new differentiation inducing factors from Dictyostelium discoideum

Developing Dictyostelium discoideum amoebae form a stalked fruiting body in which individual cells differentiate into either stalk cells or spores. The major known inducer of stalk cell differentiation is the chlorinated polyketide DIF-1 (1-(3,5-dichloro-2,6-dihydroxy-4-methoxyphenyl)hexan-1-one); however a mutant blocked in the terminal step of DIF-1 biosynthesis still produces one of the prestalk cell subtypes - the pstA cells - as well as some mature stalk cells. We therefore searched for additional stalk cell-inducing factors in the medium supporting development of this mutant. These factors were purified by solvent extraction and HPLC and identified by mass spectroscopy and NMR. The mutant lacked detectable DIF-2 and DIF-3 (the pentanone and deschloro homologues of DIF-1) but four major stalk cell-inducing activities were detected, of which three were identified. Two compounds were predicted intermediates in DIF-1 biosynthesis: the desmethyl, and desmethyl-monochloro analogues of DIF-1 (dM-DIF-1 and Cl-THPH, respectively), supporting the previously proposed pathway of DIF-1 biosynthesis. The third compound was a novel factor and was identified as 4-methyl-5-pentylbenzene-1,3-diol (MPBD) with the structure confirmed by chemical synthesis. To investigate the potential roles of these compounds as signal molecules, their effects on morphological stalk and spore differentiation were examined in cell culture. All three induced morphological stalk cell differentiation. We found that synthetic MPBD also stimulated spore cell differentiation. Now that these factors are known to be produced and released during development, their biological roles can be pursued further.     

CulB,a putative ubiquitin ligase subunit,regulates prestalk cell differentiation and morphogenesis in Dictyostelium spp

Dictyostelium amoebae accomplish a starvation-induced developmental process by aggregating into a mound and forming a single fruiting body with terminally differentiated spores and stalk cells. culB was identified as the gene disrupted in a developmental mutant with an aberrant prestalk cell differentiation phenotype. The culB gene product appears to be a homolog of the cullin family of proteins that are known to be involved in ubiquitin-mediated protein degradation. The culB mutants form supernumerary prestalk tips atop each developing mound that result in the formation of multiple small fruiting bodies. The prestalk-specific gene ecmA is expressed precociously in culB mutants, suggesting that prestalk cell differentiation occurs earlier than normal. In addition, when culB mutant cells are mixed with wild-type cells, they display a cell-autonomous propensity to form stalk cells. Thus, CulB appears to ensure that the proper number of prestalk cells differentiate at the appropriate time in development. Activation of cyclic AMP-dependent protein kinase (PKA) by disruption of the regulatory subunit gene (pkaR) or by overexpression of the catalytic subunit gene (pkaC) enhances the prestalk/stalk cell differentiation phenotype of the culB mutant. For example, culB⁻ pkaR⁻ cells form stalk cells without obvious multicellular morphogenesis and are more sensitive to the prestalk O (pstO) cell inducer DIF-1. The sensitized condition of PKA activation reveals that CulB may govern prestalk cell differentiation in Dictyostelium, in part by controlling the sensitivity of cells to DIF-1, possibly by regulating the levels of one or more proteins that are rate limiting for prestalk differentiation.     

DIF signalling and cell fate

The DIFs are a family of secreted chlorinated molecules that control cell fate during development of Dictyostelium cells in culture and probably during normal development too. They induce stalk cell differentiation and suppress spore cell formation. The biosynthetic and inactivation pathways of DIF-1 (the major bioactivity) have been worked out. DIF-1 is probably synthesised in prespore cells and inactivated in prestalk cells, by dechlorination. Thus, each cell type tends to alter DIF-1 level so as to favour differentiation of the other cell type. This relationship leads to a model for cell-type proportioning during normal development.     

Zinc ions promote prestalk-to-stalk and prespore-to-stalk conversions in Dictyostelium discoideum

Abstract To clarify the mechanism of stalk cell differentiation in Dictyostelium discoideum (strain NC4), we have examined the effects of Zn²⁺ on in vitro cell differentiation of prestalk and prespore cells isolated from normally formed slugs. Prestalk cells did not differentiate into stalk cells under submerged conditions, but in the presence of the stalk-inducing factor-1 (DIF-1) at 100 nM or Zn²⁺ at 5 mM, a small number of the cells (< 15%) differentiated into stalk cells. Interestingly, Zn²⁺ in combination with DIF-1 induced the prestalk-to-stalk conversion at high efficiencies (approx. 60%). Furthermore, isolated prespore cells were also converted to stalk cells at high efficiencies (approx. 50%) in the presence of both DIF-1 and Zn²⁺, while the conversion poorly occurred in the absence of Zn²⁺. These results indicate that Zn²⁺ may mimic some cellular interaction(s) which are required for stalk cell formation in this strain.     

Changes during differentiation in requirements for cAMP for expression of cell-type-specific mRNAs in the cellular slime mold, Dictyostelium discoideum

A number of genes encoding developmentally regulated mRNAs in the cellular slime mold, Dictyostelium discoideum, have been described. Many of these are regulated by cAMP. Analysis of the earliest time at which elevated levels of cAMP can induce the expression of these mRNAs reveals a more complex pattern of regulation in which genes change in their ability to be induced in response to cAMP with developmental stage. A prestalk mRNA (C1/D11) previously thought not be regulated by elevated levels of cAMP is inducible by cAMP between aggregation and loose mound stage; later in development its expression becomes independent of elevated cAMP. The early prespore genes (prespore class I) also show two modes of regulation; early in development they are induced independently of continuous elevated levels of cAMP, while later in development their expression is dependent upon elevated cAMP. The period during development when the prestalk genes are cAMP inducible precedes by 2 hr the first time at which either the early prespore class I or late prespore class II mRNAs are inducible by continuous elevated levels of cAMP. Previous analysis of these mRNAs has been carried out using Dictyostelium cells grown axenically. In this report we have studied the developmental expression of these mRNAs in cells grown on bacteria. A substantial shutoff of the class I prestalk and early prespore (class I) mRNAs not seen in axenically grown cells is observed when bacterially grown cells are plated for development. Less than 10% of the maximal level of these mRNAs remains in the cells at the time of mature spore and stalk differentiation. Additionally, in the bacterially grown cells two distinct patterns of developmental regulation are observed for mRNAs which in axenically growing cells appear to be constitutively expressed throughout growth and development.     

Morphogenesis and differentiation of Dictyostelium cells interacting with immobilized glucosides: dependence on DIF production

Previous work has shown that multicellular morphogenesis of submerged Dictyostelium cells is inhibited when they bind to glucosides covalently linked to polyacrylamide gels. The amoebae aggregate normally, but then the aggregates repeatedly disperse and reaggregate, whereas control cells go on to form tight aggregates. We have investigated the role of the stalk cell differentiation inducing factors (DIFs) in this process. In the presence of cyclic AMP, amoebae submerged at high cell density accumulate DIF and differentiate into stalk cells. We find that stalk cell differentiation is inhibited by interaction of the cells with glucoside gels in these conditions, but can be restored by the addition of exogenous DIF-1. Since the responsiveness of cells to DIF-1 is not altered, it appears likely that the effect of the glucoside gel is to block DIF-1 production. Further, the addition of DIF-1 or DIF-2 stimulates the formation of tight aggregates by cells developing on glucoside gels in the absence of cyclic AMP, thus preventing the rounds of aggregation and disaggregation otherwise seen. This suggests a role for DIF in morphogenesis as well as in controlling cell differentiation. We propose a model in which immobilized glucosides activate a specific receptor ("food sensor") which drives the amoebae toward the vegetative state and inhibits DIF accumulation. DIF, on the other hand, induces tight aggregate formation and so locks the amoebae into the developmental program.