Dictyostelium Myb Transcription Factors Function at Culmination as Activators of Ancillary Stalk Differentiation

Abstract

ecmB and mrrA are expressed in the cups that cradle Dictyostelium spore heads, and MybE is necessary for their expression in lower but not upper cup cells. A Myb site within the mrrA promoter is necessary for expression in both cups. Thus, multiple Myb proteins are required for ancillary stalk differentiation.

Although Dictyostelium development provides one of the simplest models for cellular differentiation, the combinatorial promoter codes that direct cell type-specific gene expression are not well understood. One complicating factor is the existence of cell type heterogeneity. The anterior one-fifth of the slug is composed of two kinds of prestalk cells; pstA cells make up the front half of the prestalk region and pstO cells lie behind them (1). The two cell types are normally identified by their ability to utilize different regions of the ecmA promoter but can also be distinguished using noncompound promoters (5). PstO differentiation is induced by DIF-1, a polyketide synthesized by the prespore cells, but the pstA inducer is unknown (4, 6). There are anterior-like cells (ALC) in the prespore region; at culmination, some of the ALC and some of the pstO cells differentiate further to form the lower and upper cups, ancillary stalk cell structures that cradle the developing spore head (3).

MybE is a single Myb domain protein of a kind found in plants but not animals (2). It binds to a 22-nucleotide (nt) region of the ecmA promoter that directs prestalk-specific, DIF-inducible gene expression when multimerized (2). In a mybE null (mybE⁻) strain the ecmA promoter is inactive in the pstO cells and the ALC, and it is not DIF inducible. We show that mrrA, a novel gene of unknown function, and ecmB, a gene encoding an extracellular matrix protein, are both spatially regulated by MybE during culmination.

The mrrA gene is strongly expressed in upper and lower cup cells at culmination.

MrrA (dictyBase reference no. DDB0168555) was identified in a small-scale microarray screen searching for genes that are aberrantly expressed in the MybE null strain. It was somewhat underexpressed in the null strain. It is predicted to be a single-pass, integral membrane protein (with transmembrane helices at amino acids 4 to 22, presumably a signal peptide, and 196 to 218). In the predicted extracytoplasmic domain it shows homology to the repeats that make up most of the extracytoplasmic domain of the cation-independent mannose 6-phosphate receptor (pfam00878). Hence, we term it MrrA for Mannose 6-phosphate receptor-repeat A.

The complete mrrA upstream promoter region (see Fig. 1A legend) was fused to lacZ, and the construct, mrrA:lacZ, was introduced into both Ax2 cells and mybE⁻ cells. In parental structures during slug formation, staining cells are concentrated at the rear and there is a weak band of staining in the pstO region. At the equivalent stage, the rear region of the mybE⁻ structures is also stained but there is less pstO staining. Thus, just as for ecmA, expression in pstO cells requires MybE (2). The posterior staining of both strains and the pstO staining in the parental strain disappear after a period of slug migration, probably through deposition into the slime trail, where large clumps of stained material are visible (Fig. 1A). During culmination of the parental strain the upper cup and lower cup become strongly stained. The fact that mrrA is activated only in stalk ancillary structures at culmination distinguishes it from ecmB, the paradigmatic marker of stalk cell differentiation, which is activated in ancillary structures and at the entrance to the stalk tube (3) (Fig. 1B).

FIG. 1.

LacZ staining of parental and mybE⁻ cells. (A) Analysis of mrrA expression. A promoter fragment of mrrA, encompassing nt −647 to +116, numbered relative to the ATG initiation codon, was fused to lacZ and transformed into parental and mybE⁻ cells. The next gene upstream (DDB0216939) terminates at −525, relative to the mrrA initiation codon, so this fragment should contain all the upstream regulatory information, and there are no predicted introns in mrrA. Pools of transformants were selected with 20 μg/ml G418 and developing structures at the indicated stages analyzed for β-galactosidase activity as described previously (2). The stained structures were from cells prestained with neutral red before development (7), and the arrow on the MybE null structure indicates the position of the lower cup. The arrows in the top right panel indicate the positions of stained cellular aggregates that seem to be shed from the back as newly formed slugs leave their sites of formation. (B) Analysis of ecmB expression. Expression of ecmB:lacZ (2) was analyzed in late culminants as described for Fig. 1A.

In the mybE⁻ strain mrrA:lacZ expression in the upper cup appears to be normal but there are no expressing cells in the position of the lower cup (Fig. 1A). Because there is a rearward loss of mrrA:lacZ-expressing cells during slug migration, it was important to check whether there is a physical structure corresponding to the lower cup in the mybE⁻ strain. When mutant cells are labeled with neutral red, a vital dye that stains all prestalk cells, a red-stained lower cup is observable (Fig. 1A). Thus, the mrrA reporter construct requires MybE for its expression in lower cup cells but not in upper cup cells.

In a previous study analyzing ecmB expression in the mybE null strain, we quoted unpublished data that showed stained cells in the position of the lower cup (2). The mrrA results described above caused us to revisit this issue. We now find that during culmination of the mybE⁻ strain ecmB:lacZ is expressed normally in the upper cup but is not expressed in the lower cup (Fig. 1). We are unsure of the reason for this disparity, but the MybE null strain is developmentally highly aberrant. Hence, rearward accumulation of cells in “stalked” migrating slugs may have been mistaken for a lower cup.

The mrrA promoter contains a functional MybE binding site.

Just over 250 nt upstream of the mrrA initiation codon there is a 9-nt identity with the region of the ecmO promoter element that contains a Myb dyad (2). When the region −292 to −252 is used in a band-shift assay with recombinant MybE protein a retarded band is observed (Fig. 2A). A multiply point-mutated form was created by replacing the Myb dyad with randomly selected nucleotides (Fig. 2A). This is a significantly less effective competitor. Thus, MybE binds specifically to the Myb dyad.

FIG. 2.

Analysis of the mrrA Myb site. (A) Gel retardation analysis of MybE binding to the mrrA Myb site. A region of MybE encompassing the Myb domain, amino acids 371 to 712, was expressed in Escherichia coli with a GST tag. Purified protein was used in gel retardation analysis, using oligonucleotide mrrA-299/-252 as a probe, with the indicated amounts of mrrA-299/-252 and mrrA-299/-252m used as a competitor as described previously (2). The central lane is without competitor. (B) LacZ staining of cells expressing mrrA(-299/-252):lacZ and mrrA(-299/-252m):lacZ. A tandemly duplicated version of the -299/-252 promoter fragment of mrrA and of its point mutant form (Fig. 2A) were fused to lacZ via the basal promoter elements of the Actin 15 gene as described previously (2). The constructs were transformed into parental cells, and pools of transformants, selected at 200 μg/ml G418, were analyzed for β-galactosidase activity.

A lacZ fusion, containing two tandemly arrayed copies of region −292 to −252 of the mrrA promoter upstream of basal promoter elements, was constructed (Fig. 2B). When transformed into Ax2 cells, expression of lacZ is first detectable during culmination in scattered cells in the positions of the upper and lower cups (Fig. 2B). The mutations in the MybE site that prevent MybE binding in vitro were inserted into a tandemly duplicated form of the −292 to −252 region. Again this sequence was cloned upstream of basal promoter elements and transformed into parental cells. The mutant construct shows no detectable level of transcription. There is also a perfect Myb dyad (AACTGTT) at nt −362 in the ecmB promoter, but the ecmB promoter is relatively complex, with multiple activator domains, and we have not analyzed it further.

Thus, the Myb dyad within the mrrA promoter binds MybE in vitro and, when dimerized, a region containing the dyad directs upper and lower cup-specific expression. The expression is weak, but it is spatially specific, and point mutations that ablate Myb binding in vitro prevent expression. Therefore, a Myb site is necessary for upper cup expression but, as we have shown, MybE itself is not necessary. Presumably, one of the other Dictyostelium Myb family proteins serves this function. In conjunction, these results imply a major role for Myb family proteins in ancillary stalk cell differentiation.