INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

The cellular slime mold Dictyostelium discoideum is an ideal organism for studying molecular mechanisms regulating developmental decisions. Dictyostelium cells start the developmental process and finally form fruiting bodies when they are starved. At an early stage of development, the cells undergo chemotactic aggregation and start developing into multicellular organisms without cell division, responding to relayed pulses of cAMP. This process is controlled through a series of integrated signal transduction pathways (1). A number of novel Dictyostelium genes involved in this process have already been identified by means of an efficient random tagging method called restriction enzyme-mediated integration (REMI)¹ (2).

The TPR (tetratricopeptide repeat) motif composed of 34 amino acid residues with the consensus sequence WX₂LGX₂YX₈AX₃FX₂AX₄P (X, any amino acid) is found in many proteins of a variety of organisms from bacteria to mammals (3-5). The TPR family members generally contain 3-19 TPR units, often arranged in tandem. Comparison of the TPR units in a TPR protein shows minimal homology limited to the consensus sequence. However, when the corresponding TPR units are compared among TPR proteins, there is striking homology beyond the consensus sequence. It has been proposed that the consensus sequence in the TPR unit functions as a "knob and hole" for the formation of a tightly packed helix-turn-helix domain (3, 6) and that the residues outside the consensus sequence play roles in protein-protein interactions (3, 7). Individual TPR units may interact with particular target proteins (8) and play a wide range of roles in cell cycle regulation, transcription, splicing, or protein import (8-14).

One of the well studied TPR proteins is Ssn6 (Cyc8) of Saccharomyces cerevisiae (9, 15). Genetic studies have shown that it forms a transcription complex with Tup1 (16), which then recognizes specific DNA-binding proteins and represses a set of unrelated promoters (17). A distinct set of TPR units is responsible for the repression of a distinct set of unrelated promoters (8). In this way, the Ssn6·Tup1 complex functions as a global repressor. In yeast cells, Ssn6 is required for normal growth, sporulation, mating, and other glucose-repressible phenotypes (9, 15, 18).

By means of the REMI method with the bsr marker (19-21), we identified a Dictyostelium gene, trfA, that is highly homologous to SSN6. It is likely that the gene product (TRFA), the first Dictyostelium TPR protein ever found, may be a global repressor for promoters of genes required for normal growth and early development of Dictyostelium cells.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Cell Growth, Development, and Chemotaxis-- All the strains used in this study were derived from D. discoideum AX2 cells (22). The wild-type AX2 cells and mutant cells were grown axenically in HL5 medium containing penicillin and streptomycin at 6 units/ml and 6 µg/ml, respectively. Transformants were selected with blasticidin S at 10 µg/ml in axenic medium.

Assays for Adenylyl Cyclase-- Activation of adenylyl cyclase by cAMP was examined as described (26). Briefly, the wild-type and mutant cells were shaken in 15 ml of 12 mM sodium potassium phosphate, pH 6.2, for 6 h. The density of the wild-type cells was 10⁷ cells/ml. Since many of the mutant cells were smaller than the wild-type cells, their density was adjusted to give the same turbidity as that of the wild-type cells. They were then washed and suspended in 1.5 ml of 12 mM sodium potassium phosphate, pH 6.2. The cells were aerated for 10 min and then stimulated with 2-deoxy-cAMP (10 µM) to activate adenylyl cyclase. A portion of the stimulated cells were mixed with an equal volume of 3.5% perchloric acid at the times indicated and then frozen. To determine the amount of cAMP, frozen cells were thawed, neutralized with 50% KHCO₃, and then centrifuged. Supernatants were used for assays by the cAMP assay kit from Amersham Pharmacia Biotech.

Restriction Enzyme-mediated Integration-- The Bsr-REMI procedure was described previously (20, 27). Briefly, a plasmid, pUCBsrBam, was linearized with BamHI and then introduced into the AX2 cells along with a restriction enzyme, DpnII, by electroporation. The electroporated cells were grown overnight in HL5 medium and then plated clonally on DM-agar plates including blasticidin S (20 µg/ml) in association with E. coli B/r cells. Colonies that formed smooth plaques were picked up and further selected with HL5 medium containing blasticidin S (10 µg/ml). They were recloned on DM agar plates with an E. coli B/r lawn. Finally, a Dictyostelium clone, CC01, generating smooth plaques on an E. coli lawn was isolated.

Southern and Northern Blot Analyses-- Genomic DNA was isolated from Dictyostelium cells as described (28). Digested DNA was separated by 0.6% agarose gel and then transferred to a Fine Trap NT-21 membrane (Nihon Eido Co., Japan) by electroblotting. The 0.3-kb XbaI/PstI fragment of the trfA gene (Fig. 1) or linearized pUC118 was used as a probe for the Southern blot analysis. Total RNA was prepared from either Dictyostelium cells at the vegetative state or cells that were washed free of medium and starved in 12 mM sodium potassium phosphate buffer, pH 6.2, for the times indicated, using ISOGEN (Nippon Gene, Japan). RNA was also prepared from the starved cells stimulated with repeated additions of cAMP (100 nM every 6 min) as described previously (29). Ten to 40 µg of total RNA was separated by 1% agarose, 1.1 M formaldehyde gels and transferred to a nitrocellulose membrane. A 0.4-kb fragment of the cAR1 gene (30) was obtained from its 3'-terminal coding region by polymerase chain reaction amplification of Dictyostelium genomic DNA and used as a probe for the Northern blot analysis. A 0.9-kb fragment of the ACA gene (31) was also amplified from its 3'-terminal coding region and used as another probe.

Molecular Cloning and Sequence Analysis of the trfA Gene-- Fragments of the trfA gene were cloned from the genomic DNA of the CC01 strain by the plasmid rescue method (2, 20). Dictyostelium genomic DNA (0.5 µg/100 µl) was digested with a restriction enzyme and then circularized with T4 DNA ligase. The ligated product was precipitated with ethanol and then redissolved in 5 µl of 0.1 × Tris-EDTA buffer (32). One µl of the DNA solution was used to transform E. coli SURE cells. Electroporation of the E. coli cells was carried out with a Gene Pulser (Bio-Rad) connected to a pulse controller according to the instruction manual. Transformants were selected as ampicillin-resistant colonies. An EcoRI fragment (8.6 kb) and a BglII fragment (5.8 kb) were successfully rescued. These fragments contained pUC118 (33) as well as the open reading frame of the trfA gene. The sequences of these fragments were determined with a DNA sequencer (SQ5500L, Hitachi).

Knockout of the trfA Gene-- Knockout of the trfA gene was carried out by the same method as for Bsr-REMI, except that a restriction enzyme was not included. Twenty µg of the linearized targeting plasmid (SHR6), which contained pUCBsrBam and the flanking sequences of the trfA gene (from HindIII to SpeI, Fig. 1), was introduced into AX2 cells by electroporation. The electroporated cells were plated clonally on DM agar plates including blasticidin S (20 µg/ml) in association with E. coli B/r cells. Colonies that formed smooth plaques were picked and selected further with HL5 medium containing blasticidin S (10 µg/ml). They were recloned on DM agar plates with an E. coli B/r lawn.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Molecular Cloning of a Dictyostelium Gene Involved in Aggregation upon Starvation-- -Dictyostelium cells were mutagenized by randomly inserting a BamHI-cut tagging plasmid, pUCBsrBam (Fig. 1), into the genome. The insertion of the tag was stimulated by adding a restriction enzyme, DpnII. Transformants were selected on agar plates covered with an E. coli lawn and containing blasticidin S. From these transformants, one strain, CC01, was identified as a clone that formed smooth plaques.

View larger version (6K): [in this window] [in a new window]   Fig. 1.   Restriction map of Dictyostelium genomic DNA around the trfA gene. The insertion site of the tagging plasmid is shown. The trfA coding sequence consists of four exons (thick lines). i, the genomic DNA fragment rescued in pBgclone; ii, the genomic DNA fragment rescued in pEcoclone; iii, the genomic DNA fragment used in the targeting vector, SHR6; iv, the XbaI/PstI fragment used as the probe for Southern blotting. Ba, BamHI; Bg, BglII; E, EcoRI; H, HindIII; P, PstI; S, SpeI; V, EcoRV; X, XbaI.

Homologous Recombination and Phenotype of the Resulting Mutant-- Two genomic DNA fragments flanking the tagging plasmid in the CC01 strain were ligated to pUCBsrBam to generate a targeting plasmid, SHR6 (Fig. 1). The plasmid was linearized and then introduced into AX2 cells without addition of a restriction enzyme. Among 150 transformants selected with blasticidin, 12 colonies formed smooth plaques on an E. coli lawn. Of the 12 transformants, 1 clone designated as K07 was used for the detailed analysis as below.

The Disrupted Gene Encodes a TPR Family Gene, trfA-- The sequences of the rescued genomic DNA fragments revealed an open reading frame interrupted by three introns (Fig. 1). The intron-exon boundaries were confirmed by sequencing cDNA clones prepared from mRNA at the vegetative stage (data not shown). In the original REMI mutant, the tagging plasmid was inserted into the third exon of the gene (Fig. 1). The open reading frame encoded a 160-kDa protein comprising 1390 amino acid residues (Fig. 2). We designated this gene as trfA and the protein as TRFA. Homology searches of SWISS PLOT and Protein Information Resource using BLASTP (34) revealed that TRFA showed the highest similarity scores with S. cerevisiae SSN6 (9), a member of the TPR family. Like Ssn6, TRFA contained 10 copies of the TPR unit in tandem (Fig. 3A). The amino acid sequences of the 10 TPR units within TRFA were 58% identical to those in SSN6. Like other TPR motifs analyzed by Sikorski et al. (4), the TPR units in TRFA contained small, uncharged amino acid residues at positions 8 (Gly or Ala), 20 (Ala), and 27 (Ala), aromatic residues at positions 4 (Trp), 11 (Tyr or Phe), and 24 (Tyr or Phe), a large aliphatic residue at position 7 (Leu or Ile), and a Pro residue at the border of the TPR units, particularly at position 32 (Fig. 3, B and C).

View larger version (116K): [in this window] [in a new window]   Fig. 2.   The deduced amino acid sequence of the TRFA protein. GenBank/EMBL/DDBJ accession number AB009080. The boxed region indicates the 10 consecutive TPR motifs (171-521).

View larger version (63K): [in this window] [in a new window]   Fig. 3.   A, schematic representation of the primary structures of the TRFA and SSN6 proteins. Each white box represents one TPR unit. Q, Gln-rich region; N, Asn-rich region; E.K.S.T and E.S.T, the regions containing high proportions of Glu, Lys, Ser, and Thr, and Glu, Ser, and Thr, respectively. B, alignment of the TPR units within TRFA. C, alignment of the 10 TPR units of TRFA and SSN6. Identical residues are highlighted. Residues of the TPR consensus sequence are indicated by asterisks.

Northern Blot Analysis of the trfA Gene-- The expression pattern of the trfA gene in the wild-type cells before and after initiating the development was examined by the Northern blot analysis. The gene was expressed at the vegetative stage and remained to be active at an early developmental stage, although its expression level was lower than that at the vegetative stage (data not shown).

Phenotype of trfA Cells-- The original REMI mutant (CC01 cells) as well as the homologous recombinant (K07 cells) formed smooth plaques on bacterial lawns (Fig. 4B). Consistent with this observation, these cells rarely formed large aggregates when they were washed free of medium and plated on phosphate-buffered agar plates (Fig. 4, C-F). Only a small fraction of trfA cells aggregated into small mound-like structures, which did not develop into fruiting bodies even on prolonged incubation (data not shown). These results indicate that the trfA cells have defects in an early developmental process, especially at the aggregating stage.

View larger version (134K): [in this window] [in a new window]   Fig. 4.   Phenotypes of the wild-type and mutant cells at the developmental stage. A, and B, AX2 cells (A) and CC01 cells (B) were plated onto bacterial lawns and then allowed to develop for 7 days. Under these conditions, the AX2 cells formed fruiting bodies, whereas the CC01 cells formed smooth plaques without forming mounds or fruiting bodies. C-F, AX2 cells (C and E) and CC01 cells (D and F) were plated onto phosphate-buffered agar plates and then allowed to develop. After 9 h, AX2 cells developed into mounds (C), whereas CC01 cells did not form any large aggregations (D). After 23 h, AX2 cells formed fruiting bodies (E), whereas CC01 cells did not form large aggregations (F). The bar, 1 mm.

View larger version (64K): [in this window] [in a new window]   Fig. 5.   Responses of the wild-type and trfA cells to cAMP. A, an example of small population assays. The wild-type and trfA cells were plated as drops on an agar plate, 4 mm from the edge of a well filled with cAMP (10 µM). Chemotactic responses of the wild-type cells (a) and the trfA cells (b) to cAMP were recorded after 6 h. Wells containing cAMP are located at the right hand. Magnifications, ×40. The bar, 100 µm. B, transient increase of cAMP in the wild-type and trfA cells upon stimulation with 2-deoxy-cAMP. Open triangles, wild-type cells. Closed circles, trfA cells.

View larger version (60K): [in this window] [in a new window]   Fig. 6.   Phenotypes of the wild-type and mutant cells at the vegetative stage. A, growth of Dictyostelium cells in suspension culture (22 °C, 150 rpm). Turbidity changes of AX2, CC01, and K07 cells were followed. Turbidity was measured as the absorbance at 660 nm. B, size and shape of axenically grown Dictyostelium cells. Cells were grown in suspension and then transferred onto a slide glass. Immediately after the transfer, the cells were observed under a phase contrast microscope. a, AX2. b, CC01. c, K07. The bar, 50 µm.

Expression of cAR1 and ACA Genes in trfA Cells upon Starvation-- The fact that trfA cells had defects in cell aggregation at the early developmental stage suggests two possibilities. One is that the cells did not enter the developmental stage because they failed to express genes such as cAR1 (the cAMP receptor gene) (30) and ACA (the adenylyl cyclase A gene) (31), which were crucial for generating the cAMP pulses. The other is that the cells entered in the developmental stage but failed to aggregate because they had defects in their signal transduction pathway to induce the chemotactic response to the cAMP signal. To determine which is the case, expression of the two critical genes, cAR1 and ACA, in trfA cells upon starvation was examined. As shown in Fig. 7, expression of these genes was rarely detected in the wild-type and trfA cells at the vegetative stage. When the cells were starved by washing free of the medium, the cAR1 and ACA genes were expressed in the wild-type cells as observed before (30, 31). When the starved wild-type cells were stimulated by repeated addition of cAMP (100 nM cAMP every 6 min) (29), their expression pattern did not change much because the starved cells generated their own cAMP pulses. Expression of the cAR1 and ACA gene in trfA cells followed the similar pattern upon starvation, and repeated addition of cAMP did not change this pattern, indicating that the developmental stage was actually initiated in trfA cells upon starvation.

View larger version (45K): [in this window] [in a new window]   Fig. 7.   Northern blot analysis of the wild-type cells and the trfA cells (trfA Null). Total RNA was isolated from the wild-type and trfA cells at the vegetative stage (0 h) and at the early developmental stages (2, 4, and 6 h after initiating the development). RNA was also isolated from the wild-type and trfA cells, which were starved and repeatedly pulsed with cAMP (+cAMP). Total RNA (10 µg) was separated and probed with the cAR1 gene or the ACA gene.

Transient Increase of the cAMP Level in trfA Cells upon Stimulation with cAMP-- When the wild-type cells were starved, the newly induced ACA became ready to respond to the cAMP signal. As shown in Fig. 5B, the cAMP concentration in the wild-type cells increased transiently upon stimulation with 2-deoxy-cAMP and then quickly returned to its original level, a typical response to the cAMP signal. The cAMP level in the starved trfA cells also exhibited a similar response to the stimulation with 2-deoxy-cAMP, although the maximal level was higher.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

Yeast strains bearing mutations in the SSN6 gene have multiple mutant phenotypes, including constitutive expression of glucose-repressible genes, calcium-dependent flocculation, mating defects of MAT cells, inability of homozygous diploids to sporulate, and poor growth (9, 15, 18, 44). These diverse phenotypes result from a failure to repress different kinds of genes (17). Genetic analysis has revealed multiple functional interactions between distinct subsets of TPR units in Ssn6 with target proteins in this process (45). When the TPR units in TRFA were compared with those in Ssn6, striking homology was detected not only in the TPR consensus sequence but also in the region outside this sequence. In particular, the third, fourth and ninth TPR units of TRFA and Ssn6 shared high proportions of identical residues. The TPR consensus sequence may be required to maintain the basic structural architecture of the TPR domain, whereas residues outside the consensus sequence would be required as a surface for the interaction between TPR units and target proteins (8). Thus, the fact that the TPR units of TRFA and Ssn6 are highly homologous beyond the consensus sequence implies that these proteins interact with similar target proteins.

It must be also noted that the sequence similarity between TRFA and Ssn6 extends even outside the homologous TPR units. Both TRFA and Ssn6 have poly(Gln) stretches, which also contain interspersed Pro residues. Consequently, this region contains many PXXXQ repeats, where X is an uncharged residue (9). Such a Gln- and Pro-rich segment is present in many transcription factors (38-40). Moreover, in the C-terminal regions of TRFA and Ssn6, there are EKST-rich or EST-rich sequences containing putative phosphorylation sites for cAMP- and cGMP-dependent kinases as well as for casein kinase II.

As in yeast cells bearing mutations in SSN6, trfA cells exhibited multiple defects. First, the mutant cells grew poorly, and cells with abnormal sizes were generated in suspension culture. Second, the mutant cells showed a defect in an early developmental process, especially at the aggregation stage. The cells rarely associated into large aggregates to form mounds and slugs of a normal size. Consistent with this defect in the early development, individual trfA cells exhibited a defect in the autonomous chemotactic response toward cAMP, independent of the cAMP relay between cells.

Northern blot analysis showed that upon starvation, trfA cells started expressing two critical genes for the developmental stage, i.e. cAR1 and ACA, an indication that the cells entered in the early developmental stage. Adenylyl cyclase (ACA) of trfA cells thus induced was transiently activated by cAMP or its analog with a similar time course to that of the wild-type cells, suggesting that the cAMP relay in trfA cells functioned normally. Taken together, it is likely that the major defect in trfA cells at the early developmental stage was not in the system for generating the cAMP signal but in the system required for correctly responding to it.

The striking structural similarity between TRFA and Ssn6 as well as the pleiotropic phenotypes of trfA cells indicate that TRFA may function as a multiple regulator for multiple cellular functions like Ssn6. Several transcription factors regulating the developmental process of Dictyostelium cells such as the G-box-binding protein and STAT protein have been already cloned (46, 47). The coordination among these factors would be essential for normal growth and development of Dictyostelium cells.