Implications of expansin-like 3 gene in Dictyostelium morphogenesis

Dictyostelium harbors multiple expansin-like genes with generally unknown functions. Thus, we analyzed the expansin-like 3 (expL3) gene and found that its expression was reduced in a null mutant for a STATa gene encoding a transcription factor. The expression of expL3 was developmentally regulated and its transcript was spliced only in the multicellular stages. The expL3 promoter was activated in the anterior prestalk region of the parental strain and downregulated in the STATa null slug, although the expL3 promoter was still expressed in the prestalk region. The expL3 overexpressing strain exhibited delayed development and occasionally formed an aberrant structure, i.e., a fruiting body-like structure with a short stalk. The ExpL3-myc protein bound cellulose. Electronic supplementary material The online version of this article (doi:10.1186/s40064-015-0964-0) contains supplementary material, which is available to authorized users.


Background
Expansins are cell wall proteins in plants that loosen the cell wall, thereby regulating cell wall enlargement in growing cells. Expansins may act by breaking the noncovalent bonds among wall polysaccharides (Cosgrove 2005). Expansin activity is also associated with morphogenesis and other developmental events, such as leaf primordium formation, fruit ripening, xylem formation, pollination, seed germination, and abscission (Sampedro and Cosgrove 2005). Plant expansins comprise a large gene superfamily of four divergent families: α-expansin (EXPA), β-expansin (EXPB), expansin-like A (EXLA), and expansin-like B (EXLB) (Sampedro and Cosgrove 2005).
The existence of expansin-like proteins is also known in bacteria and fungi, such as EXLX1 in Bacillus subtilis, which is structurally and functionally similar to plant expansins. The cellular slime mold Dictyostelium is a lower eukaryote, which produces a cellulose-based cell wall. It is a rare species that harbors multiple expansinlike genes in its genome, i.e., at least nine genes (expL1-9) (Darley et al. 2003;Ogasawara et al. 2009), although expL4 appears to be a putative pseudogene (http://dictybase.org). It is of evolutionary importance to understand the role of expansin-like molecules in Dictyostelium, which is significantly divergent from plants.
In a previous study, we demonstrated the role of an expansin-like gene, expL7, which regulates morphogenesis during Dictyostelium development (Ogasawara et al. 2009). The expression of expL7 is under the control of CudA, a putative transcription factor (Wang and Williams 2010). STATa is a transcription factor that activates the expression of cudA in the anterior prestalk tip region (Fukuzawa and Williams 2000), thus STATa indirectly controls expL7 gene expression. Slugs of both STATa and cudA null mutants tend to migrate for longer (slugger phenotype) and they fail to culminate, eventually forming an aberrant structure (Fukuzawa et al. 1997;Mohanty et al. 1999). The anterior prestalk region serves as an organizer during multicellular development in Dictyostelium, and the region where STATa is activated to express cudA is designated as the "tip-organizer." In this study, we analyzed the function of another expansin-like gene, expL3, which was positively regulated by STATa.

Cells and growth conditions
Dictyostelium discoideum Ax2 cells were axenically cultured in HL5 medium at 22°C. Cells of the STATa null strain were grown in HL5 supplemented with 10 μg/ml blasticidin S (Kaken Pharmaceutical, Japan). The expL3 null strain was grown in HL5 supplemented with 36 μg/ ml hygromycin B (Wako, Japan). Transformants with the Neo R cassette construct were selected using HL5 supplemented with 20 μg/ml G418 (geneticin; ICN Biochemicals Inc.).
Analysis of gene expression using semi-quantitative and quantitative RT-PCR Ax2 and STATa null cells were allowed to develop at 22°C on Omnipore filters (JGWP04700, Millipore), which were placed on non-nutrient agar plates. Total RNA was extracted from Ax2 and STATa null strains every 3-4 h. cDNA synthesis and RT-PCR were conducted as described previously (Shimada et al. 2004(Shimada et al. , 2005 using a pair of primers: expL3-RT-1 and expL3-RT-2 or expL3-G7-i and expL3-G8-i. The quantitative RT-PCR analysis was performed as previously described (Shimada et al. 2010). The primers used for RT-PCR are listed in Additional file 1: Table S1.
lacZ fusion construct and β-galactosidase staining The promoter fragment of the expL3 gene (the 5' end point is located 948 nucleotides upstream from the putative translation initiation site) was amplified by PCR to add an XbaI site at the 5' end and a BglII site at the 3' end. After digestion with XbaI and BglII, the fragment was gel-purified and subcloned into XbaI/BglII-cut pDdgal-17(H+) (Harwood and Drury 1990) to yield pDdNeo R [expL3/lacZ]. To detect the promoter activity, cells transformed with pDdNeo R [expL3/lacZ] were grown and developed on Omnipore filters. Fixation and staining were performed as previously described (Shimada et al. 2005).

expL3 expression vectors
The fragment corresponding to the entire open reading frame (ORF) of the expL3 gene was amplified by PCR using a plasmid DNA containing the expL3 gene to add a SalI site at the 5' end and a BamHI site at the 3' end, before subcloning into pTOPO-Blunt II (Invitrogen) to produce pTOPO[expL3-ORF(G4/G5)]. After digestion with SalI and BamHI, the ORF fragment was gel-purified and subcloned into SalI and BamHI-digested pLD1ΔBXmyc (unpublished) to yield pLD1ΔBX[act15/expL3-myc]. Plasmid DNA that contained the promoter region of the ecmF gene (Shimada et al. 2004) was purified by gel electrophoresis after digesting pLD1ΔBX[ecmF/dutA(nF)] (unpublished) with SalI and NotI. The expL3-myc fragment was purified by gel electrophoresis after digesting pLD1ΔBX[act15/expL3-myc] with SalI and NotI. Both of the purified DNA fragments were ligated to yield pLD1ΔBX[ecmF/expL3-myc]. It should be noted that each of the ExpL3 expression constructs used in this study contained an intron.

Western blot analysis of ExpL3-myc protein
Cells transformed with pLD1ΔBX[act15/expL3-myc] or pLD1ΔBX[ecmF/expL3-myc] were allowed to develop until the slug stage, before the slugs were solubilized and analyzed on 7.5% (w/v) SDS-polyacrylamide gels, followed by blotting onto Hybond-C extra filters (Amersham Biosciences, UK). The filters were blocked and detected using the Promega Proto Blot II AP System with Stabilized Substrate, according to the manufacturer's protocol (Promega). Anti-c-Myc monoclonal antibody 9E10 (1:2000 dilution; Wako) was used as a primary antibody and alkaline phosphatase (AP)-conjugated anti-mouse IgG (1:20,000 dilution; Promega) was used as a secondary antibody.

Cellulose-binding assay
The cellulose-binding ability of the ExpL3-myc fusion protein was tested according to a previously described procedure (Kunii et al. 2014) with some modifications. First, 2 × 10 7 cells that overexpressed the ExpL3-myc protein (act15:expL3 myc OE strain) were allowed to develop until the slug stage on the Omnipore filter (Millipore), before the slugs were harvested and ground with a plastic pestle in the presence of 250 μl of 1 × phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 8.1 mM Na 2 HPO 4 , 1.47 mM KH 2 PO 4 , pH 7.4) containing cOmplete, Mini, EDTA-free (Roche) as protease inhibitors. Furthermore, 250 μl of 2 × lysis buffer [50 mM KCl, 10 mM Tris-HCl, 2.5 mM MgCl 2 , 0.45% (w/v) Tween 20, pH 8.0] was added to lyse the slug. Microcrystalline cellulose beads (Avicel PH-101; Sigma) were suspended in binding buffer (100 mM Tris-HCl, pH 8.0) at a final concentration of 5% (w/v). Subsequently, 500 μl of the Avicel slurry was added to the cell lysate, which was allowed to bind with rotation at 4°C for 1 h. The Avicel was pelleted by centrifugation at 20,000 × g for 2 min and washed three times each with 1 M NaCl/50 mM phosphate buffer (pH 7.5) and with 50 mM phosphate buffer (pH 7.5). The bound protein was eluted by heating at 105°C for 8 min in 1 × SDS sample buffer. The unbound and input fractions were concentrated by ultrafiltration (Microcon Ultracel YM-10, Millipore), mixed with 1/2 volume of 3 × SDS sample buffer, and heated as described above. The fusion protein in each fraction was detected by Western blot analysis, as described above.

STATa-dependent expression of the expL3 gene
Previously, we showed that a Dictyostelium expansin-like gene, expL7, has crucial roles in morphogenesis during development (Ogasawara et al. 2009). Its expression is regulated by CudA (Wang and Williams 2010), where the expression of this gene in the tip-organizer cells is regulated by STATa (Fukuzawa and Williams 2000). Therefore, we examined the STATa-dependency of gene expression for other expansin-like family members. The majority of the genes examined in the family exhibited STATa-independent expression (see Additional file 2 and Additional file 3: Figures S1 for expL4 and expL6), whereas expL3 (dictyBase ID DDB_G0276287) exhibited STATa-dependent expression because according to quantitative RT-PCR, its transcript level was significantly reduced in the STATa null strain compared with that in the parental Ax2 strain at the corresponding stages ( Figure 1).

Expression of expL3 is developmentally regulated
To elucidate the roles of the expL3 gene, we investigated the similarity between Dictyostelium ExpL3 and ExpL7. The alignment obtained using CLUSTALW showed that the deduced amino acid sequence of ExpL3 shared 16% identity and 43% similarity with that of Dictyostelium ExpL7. Similarly, ExpL3 shared 17% identity and 59% similarity with Arabidopsis thaliana expansin, EXPA1. Similar to ExpL7, ExpL3 harbors the conserved motifs or domains that characterize the expansin found in Arabidopsis EXPA1, although the homology of some of the Figure 1 Comparison of expL3 transcript levels in Ax2 and STATa null strains. Ax2 and STATa null cells were allowed to develop until the desired developmental stages. Total RNA was extracted from Ax2 cells (shown as blue bars) at the first finger/early slug (indicated as ES), late slug (LS), and Mexican hat/early culminant (M/EC) stages, and from STATa null mutant cells (shown as pale blue bars) at the first finger/early slug (ES), migrating slug after 20 h of development (S1), and slug at 24-h development (S2) stages, and used to amplify the specific expL3 DNA fragment. The expression of expL3 in Ax2 and STATa null cells was investigated by quantitative real-time PCR. The IG7 transcript was used as a normalization control. The bar heights represent the average relative transcript level based on triplicate measurements, with the error bars. The expression of expL3 in the STATa null early slug stage was set at 1.00. Statistically differences of expression level were analyzed by Student's t-test, which showed all p-values were less than 0.01 between Ax2 and STATa null strain at comparable stages. To amplify the specific DNA fragment, we used 0.1 μg of total RNA and the primers: expL3-Fq and expL3-Rq (see Supplementary Figure 3a for their locations). The developmental stage is shown at the bottom of each column and the name of the strain is shown below each horizontal bar.
domains is weak, such as the numbers and locations of the conserved tryptophan residues in the cellulose-binding domain (see Additional file 2 and Additional file 3: Figure S2).
To further characterize the expL3 gene, we investigated the expression profiles of expL3 both temporally and spatially, where we examined the relative expL3 transcript level using semi-quantitative RT-PCR. The expL3 gene has a single intron in the coding region; thus, PCR primers were designed for both sides of the intron (see Additional file 2 and Additional file 3: Figure  S3A). The results showed that expL3 expression is developmentally regulated, with a peak at approximately 18 h, which is the late slug to Mexican hat stage (Figure 2).
The expL3 transcript was efficiently spliced only during the developmental stages between 12 h (tip stage) and 21 h (Mexican hat stage) because a smaller band that corresponded to the fragment without the intron was strongly amplified (Figure 2a). During the earlier stages until 9 h (mound stage), only a larger band that corresponded to the fragment with the intron was detected. Both bands were detectable at 24 h (fruiting body). These results indicate the posttranscriptional regulation of expL3 in addition to transcriptional regulation, which is specific to the multicellular developmental stages.

Expression of the expL3 is prestalk-specific
To clarify the spatial expression pattern, we produced an expL3/lacZ reporter construct and the promoter activity was assayed in terms of its β-galactosidase activity. A scattered signal was detected at the mound stage ( Figure. 3Aa). Furthermore, the signal was sorted to the prestalk (pst) region during tip formation ( Figure 3A-b, c). During the first finger, slug, and Mexican hat stages, a strong signal was detected in the pstA region, whereas a weak scattered signal was detected in the pstO region ( Figure 3A-d, e, f).
Later, a signal was detectable in the stalk of the culminant ( Figure 3A-g).
When the STATa null cells were transformed with the same construct, weaker staining of the pstA cells was visible during any of the multicellular stages, but almost no pstO or scattered staining was visible ( Figure 3B). This indicates that STATa is necessary for adequate expression of the expL3 gene, but it does not confer pstAspecificity.

Overexpression of the expL3 gene caused morphological aberrations
To examine the functions of the expL3 gene, a mutant strain that lacked this gene was created via homologous recombination (Additional file 2 and Additional file 3: Figure S3). Three independent targeted clones (expL3 null) were isolated, where the cells of these clones developed normally on non-nutrient agar plates and in other conditions. They formed normal looking fruiting body with calcofluor stained stalk and mature spores with normal shape and size (data not shown).
Furthermore, we created strains that overexpressed the ExpL3-myc fusion protein using two different constructs. The first contained the actin15 promoter to drive expL3myc fusion gene, which was transformed into the Ax2 strain to yield Ax2/[act15]:expL3:myc (designated as act15:expL3 myc OE ). When act15:expL3 myc OE cells were allowed to develop on non-nutrient water agar plates, the development of the slug stage was delayed by 1-2 h, and they tended to migrate longer than Ax2 slugs (data not shown). The actin15 promoter is not constitutive, and it peaks at 2-6 h before its transcription level gradually reduces as development proceeds (Cohen et al. 1986); therefore, the effect of its overexpression may not be retained during the later stages of development. To resolve this A B Figure 2 Temporal expression profiles of the expL3 gene. A: Developmental time course of expL3 splicing detected by semi-quantitative RT-PCR. Total RNA was extracted from Ax2 cells every 3 h and used as a template for the amplification of a specific expL3 to detect DNA fragments with or without an intron. The image shows the RT-PCR results for expL3 after 30 cycles of amplification. Note that after 30 cycles of amplification, the band intensity was almost saturated between 12-21 h. The band of approximately 200 bp contained an intron whereas the band of approximately 100 bp did not. The stages are loose aggregate (6 h), mound (9 h), tipped aggregate (12 h), slug (15 h), Mexican hat (18 h), culminant (21 h), and mature fruiting body (24 h), respectively. B: Expression profiles of the expL3 gene during multicellular developmental stages. The upper row shows the RT-PCR results for expL3 after 27 cycles of amplification before the PCR reaction saturated. IG7 was detected using the same cDNA as a normalization control. The IG7 DNA product (lower row) was detectable after 17 cycles of amplification.
issue, we attempted to place the fusion gene under the control of its own expL3 promoter. However, this construct was unstable and we could not create it successfully. Therefore, we attempted to use the ecmF promoter to replace the actin15 promoter, and to create a new overexpression construct ( Figure 4A) because the ecmF promoter activity is similar to that observed for expL3 both temporally and spatially, i.e., late stages and pstAspecific (Shimada et al. 2004; also see Figures 2 and 3A).
Overexpression of the fusion protein was confirmed in the resultant transformant of Ax2/[ecmF]:expL3:myc (designated as expL3 oe ) by Western blot analysis (Figures 4B and Additional file 2 and Additional file 3: Figure S4).
The expL3 oe cells developed normally on non-nutrient water agar plates until the first finger/early slug stage (approximately 15 h). However, the development of the expL3 oe started to delay at slug stage, the time when the ExpL3-myc protein became detectable (Additional file 2 A B Figure  and Additional file 3: Figure S4). The slugs tended to migrate for longer, where the majority (~60%) remained as slugs even after 24 h of development ( Figure 5). A limited number of Mexican hat-like structures were observed at 21 h and culminant-like structures (culminants) were present among the slugs at 24 h. After 3 days of development, most of them (more than 95%) became culminants; however, they exhibited aberrant terminal structures, where a short (~1 mm long or shorter) but thick stalk-like structure supported the spore mass ( Figure 5). They formed cellulose (Additional file 2 and Additional file 3: Figure S5), mature spore (data not shown) but they did not form an unequivocal basal disc. This indicates that the expL3 gene causes morphological aberrations when it is overexpressed in the anterior pstA cells.

ExpL3 protein binds to cellulose
Most plant expansins localize in the cell wall and they are considered to disrupt noncovalent bonds among polysaccharides, as well as possessing a region with a homology to cellulose-binding domain; therefore, they are reported to bind cellulose (McQueen-Mason and Cosgrove 1995). We tested this hypothesis using Dictyostelium ExpL3, where we examined the binding activity of the myc-tagged ExpL3 using the act15:expL3 myc OE strain ( Figure 6). The cells were ground and allowed to , were allowed to develop on non-nutrient water agar plates. The time of development is indicated at the left of each row and the strain is indicated above each column. After 24 h of development, although some (~40%) of the expL3 oe strain formed culminant-like structures, majority (~60%) remained as slugs (the far right column). After 3 days of development, the structure formed was apparently abnormal because they formed a thick and short stalk, although they formed a spore head. Bars denote 0.5 mm. Arrows shows slightly unclear basal disc.
bind to crystalline cellulose beads (Avicel). The ExpL3myc protein was detected in the bound fraction ( Figure 6, left panel), which indicates that ExpL3 is capable of binding to cellulose.

Discussion
It has been reported that the expL6 transcript is the only developmentally regulated member of this family of genes (Darley et al. 2003), but we showed that the expression of the expL3 gene is late stage-specific (Figure 2) in the present study. The reason for this discrepancy is unknown; however, the RT-PCR profile obtained in this study almost matched with the dictyExpress RNAsequencing (RNA-Seq) database (Parikh et al., 2010; Additional file 1: Table S2). The RNA-seq database indicates that most of the expansin-like family genes in Dictyostelium are expressed in a stage-dependent manner (Additional file 1: Table S2). Based on the current results and our previous study of expL7 (Ogasawara et al. 2009), we conclude that the expression levels of all expansin-like genes are developmentally regulated in Dictyostelium.
The expL3 gene is expressed only in prestalk cells, i.e., strongly in pstA and weakly in pstO cells (Figure 3). Again, there was a discrepancy in the tissue specificity according to previous in situ hybridization results (Maruo et al. 2004; EST clone SSI248) and the βgalactosidase staining results obtained in the present study ( Figure 3). This may have been caused by crosshybridization of the probe used for in situ hybridization, although we cannot exclude the possibility that the promoter region upstream of the 5' end point of the expL3/ lacZ construct confers expression in prespore cells. The RNA-seq database indicates prestalk (pst) or prespore (psp) enrichment for expansin-like family member transcripts, i.e., expL1, expL3, expL7, and expL9 transcripts are pst-enriched; expL2 and expL8 are psp-enriched; and expL4, expL5, and expL6 exhibit no obvious tissue enrichment (Parikh et al., 2010; Additional file 1: Table S2). Thus, we conclude that the expression of the expL3 gene is prestalk-specific.
The results of expL3/lacZ β-galactosidase staining in the STATa null mutant indicated that STATa may be involved in the strength of expL3 expression but it does not contribute to pstA specificity, although STATa is activated in pstA cells. Alternatively, the STATa null mutant lacks most of the pstA cells, and thus only the weaker expression in pstO cells was detectable. STATa-dependent expression was investigated among the family of genes, except expL9, but no apparent STATa-dependency was observed other than that in expL3 and expL7. These results suggest that the roles of these two expansin-like genes, i.e., expL3 and expL7, are of particular importance during development.
The lack of the phenotype in the expL3 null mutant might be an effect of the functional redundancy of the closely related genes, i.e., expL1-9, in the Dictyostelium genome. In contrast to the single null mutation, the overexpression of expansin or expansin-like genes in transgenic plants and Dictyostelium obtained the morphogenetic phenotype (Choi et al. 2003;Ogasawara et al. 2009). In agreement with these observations, overexpression of the ExpL3-myc protein via the pstA-specific ecmF promoter led to a developmental delay after slug formation and a morphological aberration during culmination ( Figure 5). We do not know how the Dictyostelium ExpL3 protein exerts its effect on the morphology. However, it is possible that ExpL3 exerts its effect on cellulose in the stalk tube and that it regulates stalk elongation because we found that ExpL3 binds cellulose ( Figure 6). We did not test whether the ExpL3 protein has a cell wall-loosening activity. If it possesses this activity, overexpression of the ExpL3 protein in stalk cells might weaken the stalk strength against gravity to yield the short, broad stalk.
Actually, expL3 and expL7 do not share many some features except STATa-dependency. The expL7 is reported CudA-dependent (Wang and Williams, 2010), but we have preliminary result expL3 is different (data not shown). Cell type-specificity is also different; pstA-specific for expL3 ( Figure 3) and tip-organizer cell-specific for cudA (Fukuzawa and Williams, 2000). Therefore, we think these two genes may behave independently. Indeed, double overexpressor strain of expL3 and expL7 genes (expL3 OE /expL7 OE ) displayed the phenotype as if it was like that seen in the expL3 OE until culmination ( Figure 5), after that it was like that seen in the expL7 OE (Ogasawara et al. 2009) (data not shown). Whatever the case, the phenotype of expL3 oe implies the involvement of ExpL3 in morphogenesis during slug migration and culmination in Dictyostelium.