Molecular and functional characterization of a Rho GDP dissociation inhibitor in the filamentous fungus Tuber borchii

Abstract

Background: Small GTPases of the Rho family function as tightly regulated molecular switches that govern important cellular functions in eukaryotes. Several families of regulatory proteins control their activation cycle and subcellular localization. Members of the guanine nucleotide dissociation inhibitor (GDI) family sequester Rho GTPases from the plasma membrane and keep them in an inactive form.

Results: We report on the characterization the RhoGDI homolog of Tuber borchii Vittad., an ascomycetous ectomycorrhizal fungus. The Tbgdi gene is present in two copies in the T. borchii genome. The predicted amino acid sequence shows high similarity to other known RhoGDIs. Real time PCR analyses revealed an increased expression of Tbgdi during the phase preparative to the symbiosis instauration, in particular after stimulation with root exudates extracts, that correlates with expression of Tbcdc42. In a translocation assay TbRhoGDI was able to solubilize TbCdc42 from membranes. Surprisingly, TbRhoGDI appeared not to interact with S. cerevisiae Cdc42, precluding the use of yeast as a surrogate model for functional studies. To study the role of TbRhoGDI we performed complementation experiments using a RhoGDI null strain of Dictyostelium discoideum, a model organism where the roles of Rho signaling pathways are well established. For comparison, complementation with mammalian RhoGDI1 and LyGDI was also studied in the null strain. Although interacting with Rac1 isoforms, TbRhoGDI was not able to revert the defects of the D. discoideum RhoGDI null strain, but displayed an additional negative effect on the cAMP-stimulated actin polymerization response.

Conclusion: T. borchii expresses a functional RhoGDI homolog that appears as an important modulator of cytoskeleton reorganization during polarized apical growth that antecedes symbiosis instauration. The specificity of TbRhoGDI actions was underscored by its inability to elicit a growth defect in S. cerevisiae or to compensate the loss of a D. discoideum RhoGDI. Knowledge of the cell signaling at the basis of cytoskeleton reorganization of ectomycorrhizal fungi is essential for improvements in the production of mycorrhized plant seedlings used in timberland extension programs and fruit body production.

Figure 1

Nucleotide and deduced amino acid sequences of the Tuber borchiigdi gene. Deduced amino acid sequences are indicated on top of the nucleotide sequences. Introns are in lower case gray italics. The stop codon is marked by an asterisk. Upstream of the ATG starting codon is an adenine at position -3, in agreement with the Kozak consensus sequence. GenBank accession number EU044761.

Figure 2

Southern blot analysis using a Tbgdi specific probe. T. borchii genomic DNA was digested with EcoRI, ScaI and BamHI restriction enzymes and blotted onto a nylon membrane. The blot was probed with a 207 bp long radioactively labeled probe. Tbgdi appears to be a two-copy gene.

Figure 3

Tuber RhoGDI and RhoGDI proteins from other organisms. A. Multiple alignment of RhoGDI proteins from T. borchii and representative species. Sequences were aligned with ClustalX and the output file was subsequently edited manually. Dashes indicate gaps introduced for optimal alignment. Residues identical or similar in at least four sequences are boxed in black or grey, respectively. Secondary structure elements are indicated on top of the aligned sequences and are based on the structure of human LyGDI [61]. Residues involved in the formation of the isoprenyl-binding pocket, as determined for bovine RhoGDI [32], are indicated by open or blue circles under the aligned sequences. Blue circles indicate residues of the "hydrophobic triad" critical for binding of the distal isoprene unit. Cyan circles indicate residues involved in the formation of an acidic patch in the isoprenyl-binding pocket. Important residues involved in interactions with the Rho GTPase are indicated by green boxes and are compiled for bovine RhoGDI [32] and human LyGDI [61]. B. Phylogenetic tree showing the relationship of TbRhoGDI with RhoGDI proteins of selected species from several eukaryotic phyla. Sequences were aligned using the ClustalX program with a BLOSUM62 matrix and default settings, followed by manual edition. Only the GDI core, devoid of hypervariable amino-terminal sequences, was considered. Phylogenetic trees were constructed using the neighbor-joining algorithms of the ClustalX program. Construction of the tree was done with TreeView. The position of TbRhoGDI is highlighted. Note grouping of TbRhoGDI with other fungi, in particular ascomycetes. The scale bar indicates 10% divergence. Accession numbers of the sequences retrieved for the phylogenetic analysis are as follows. T. borchii, EU044761. S. cerevisiae, Z74183. S. pombe, Z98533. C. neoformans, EAL19587. U. maydis, EAK86096. K. lactis, CAG98029. P. guillermondi, EDK38281. C. albicans, EAL04316. A. niger, CAK43261. G. zeae, XP_385458. H. sapiens RhoGDI1, X69550; RhoGDI2 (LyGDI), L20688; RhoGDI-3, U82532. M. musculus RhoGDI1, AU080000; RhoGDI2, U73198; RhoGDI3, Q62160. B. taurus RhoGDI1, X52689; RhoGDI2, AF182001. C. elegans, U36431. E. histolytica, AF080396. D. melanogaster, AE003515. A. japonica, C24513. B. malayi, AW159949. S. japonicum, AAW27341. H. roretzi, AV383364. I. punctatus, BE468333. N. tabacum, CAB77025. A. thaliana RhoGDI1, AAF70843; RhoGDI2, AAC17610; RhoGDI3, AAF21198. For S. scrofa, G. gallus, D. rerio, X. laevis, B. mori, G. max, Z. mays, T. aestivum, L. esculentum, L. pennellii and G. arboreum sequences were reconstructed from diverse EST sequences.

Figure 4

Expression of the Tbgdi gene. Real time PCR quantification of Tbgdi and of Tbcdc42 in Tuber borchii mycelia grown in the presence of the host plant (Tester) or root exudates (TSA) compared to untreated mycelia (Driver). The ΔΔCT method was used as described in Material and Methods. Data are average ± standard deviation of at least four independent experiments each performed in triplicate.

Figure 5

Interaction of RhoGDIs with Rho GTPases. A. Two-hybrid interactions between TbRhoGDI or LyGDI and the indicated Rho GTPases. After co-transformation of the corresponding plasmids into the Y190 yeast strain, colonies were allowed to grow on -Trp/-Leu plates and interactions verified and in colony-lift β-galactosidase assays. DdRac1b and DdRac1c behaved like DdRac1a (not shown). B. Translocation of TbCdc42 by TbRhoGDI. A membrane fraction of HeLa cells expressing GFP-tagged TbCdc42 was incubated in the absence (-) or presence (+) of 40 μM purified bacterially expressed His-tagged TbRhoGDI. The membranes were sedimented by centrifugation and aliquots of the supernatants were subjected to SDS-PAGE. The aliquot of membrane fraction (M) corresponds to the percentage of the analyzed supernatant (SN). TbCdc42 was detected with an antibody reactive against GFP. In these experiments His-tagged TbRhoGDI was always recovered in the supernatant. C. Translocation of D. discoideum Rho GTPases by bovine RhoGDI1. Membrane fractions of insect cells infected with baculoviruses encoding the indicated Dictyostelium GST-tagged Rho GTPases were incubated in the absence (-) or presence (+) of 40 μM purified bacterially expressed His-tagged RhoGDI1. The membranes were processed as in B. GTPases were detected with an antibody reactive against GST. In these experiments His-tagged RhoGDI1 was always recovered in the supernatant.

Figure 6

TbRhoGDI does not alter the growth rate of S. cerevisiae. Yeast cells were transformed with pYES2-Tbgdi for expression of TbRhoGDI or with the empty plasmid as control. The OD₆₀₀ was measured at the indicated times and normalized over the starting OD₆₀₀ of the respective culture. Data are average ± standard deviation of three independent determinations, each done in duplicate. TbRhoGDI does not seem to interfere with the Rho signaling pathway of yeast.

Figure 7

Complementation of a Dictyostelium RhoGDI1 null mutant with T. borchii and mammalian RhoGDIs. GDI1 null cells were transfected with plasmids that allow expression (indicated by an R, for rescue) of GFP fusions of Dictyostelium RhoGDI1(GDI1), TbRhoGDI, bovine RhoGDI1 and human RhoGDI2 (LyGDI). The wild-type Dictyostelium strain AX2 was used as reference. A. Total cell homogenates of 4 × 10⁵ cells were resolved in 12% polyacrylamide gels and blotted onto nitrocellulose. The blot of the upper panel was incubated with a GFP-specific mAb K3-184-2. The blot of the lower panel was incubated with Dictyostelium GDI1-specific mAb K8-322-2. All GFP fusions are expressed at levels comparable to those of the endogenous Dictyostelium GDI1. B. Growth of GDI1^- and complementation mutants in shaking suspension. GDI1^- mutant cells have a reduced growth rate and reach lower cell densities than the wild type. This defect was restored after expression of bovine RhoGDI1, but not by TbRhoGDI or human LyGDI. Curves are representative of at least three independent determinations, each done in duplicate. C. Distribution of the number of nuclei in GDI1^- and complementation mutants. Cells were allowed to grow on coverslips, then fixed with cold methanol and stained with DAPI. 300 cells of each population were scored. Photographs show DAPI staining (left panels) and phase contrast (right panels). Arrows point at examples of giant multinucleate cells. Scale bar, 50 μm. In the GDI1^- mutant, cells with four or more nuclei account for about 10% of the population. This defect was restored completely after expression of bovine RhoGDI1 and partially after expression of TbRhoGDI, whereas human LyGDI had no effect.

Figure 8

Actin polymerization response in Dictyostelium upon cAMP stimulation of aggregation competent cells. The relative F-actin content was determined by TRITC-phalloidin staining of cells fixed at the indicated time points after stimulation with 1 μM cAMP. Each data point represents the average of at least three independent measurements. For the sake of clarity, error bars are not shown and the results are presented in two graphs. In GDI1^-cells the initial response of actin polymerization and depolymerization was reduced, and was restored after expression (indicated by an R, for rescue) of bovine RhoGDI1, but not human LyGDI or TbGDI. Expression of TbRhoGDI resulted in a further decreased first peak and an abolished second peak of actin polymerization.