Ectomycorrhizal development and function – transcriptome analysis
Differential gene expression during ectomycorrhizal development has been investigated intensely during the past decade, revealing a first glance of events physiologically important for the development and function of this symbiosis (Martin, 2001). These investigations have enabled the formulation of different working hypotheses that could explain structural and functional adaptations observed during plant–fungal interactions (Martin et al., 1999; Ditengou & Lapeyrie, 2000; Nehls et al., 2001). Nevertheless, all hypotheses are based on the rather small number of genes that have been investigated. To overcome this limitation, a number of EST-projects have been initiated using a small set of ectomycorrhizal plant and fungal systems (Colebatch et al., 2002; Tunlid, 2003). In this issue (pp. 117–129), Peter et al. describe the generation and analysis of large EST sets obtained from two different ectomycorrhizal fungi, Laccaria bicolor and Pisolithus microcarpus (Cairney, 2002; Martin et al., 2002).
‘To complete the picture of fungal gene expression during ectomycorrhizal development and function a genome project for at least one of the well-established model fungi is essential’
Models – two trees, many fungi
Poplar and birch are the two angiosperm partners mainly used for ectomycorrhizal plant–fungal EST projects. In addition to their value for studying ectomycorrhizas, both plants are handled as model trees to address questions that are difficult or impossible to investigate in Arabidopsis (e.g. wood formation, seasonal nutrient cycling and storage; see Chaffey, 2002). Thus, for both trees, transformation is possible and large EST collections from different organs (including roots –Kohler et al., 2003) are already available. In addition, a genome project has been initiated for poplar, which will be finished by the end of this year (News in brief, 2002) allowing genome-wide expression analysis. On completion, and then with ectomycorrhizas obtained from both natural sites and well defined lab conditions, ectomycorrhizal research with these angiosperms will speed up. Compared with angiosperms, model gymnosperms (such as Picea or Pinus spp.) lag behind – they are more difficult to handle, there is no reliable method of transformation and genome projects are yet to be initiated.
While only a limited number of tree genera (though with large economical and ecological importance) are able to form ectomycorrhizas, several thousand different fungal species, mainly basidiomycetes, can undergo symbiosis. To reflect this enormous biodiversity found in forest ecosystems, a larger number of model fungi has been established, containing pioneer as well as intermediate- and late-stage fungi (Amanita muscaria, Hebeloma cylindrosporum, Laccaria bicolor, Paxillus involutus, Pisolithus microcarpus, Suillus bovinus, Tuber borchii). To get an initial set of genes for broad-range investigation of gene expression during development and function of ectomycorrhizas, several EST-projects have recently been initiated (T. Johansson et al. pers. comm.; Lacourt et al., 2002; U. Nehls et al. unpublished; Podila et al., 2002; Peter et al., 2003; C. Plassard et al. pers. comm.; Voiblet et al., 2001) using nonmycorrhized hyphae as well as mycorrhizas of different ectomycorrhizal model fungi as their sources.
Laccaria and Pisolithus
In the research by Peter et al. (pp. 117–129 in this issue), the sequences of L. bicolor were obtained from nonmycorrhizal hyphae cultivated under two different carbon regimes, and those of P. microcarpus from nonmycorrhizal hyphae grown in a rich carbon source as well as from ectomycorrhizas. A total of 905 and 806 unique transcripts, respectively, were analyzed, representing about 10% of the expected genes of these fungi.
A very large number of these genes (60% of the Laccaria transcripts and 50% of those of Pisolithus) did not reveal any sequence homology in existing databases. This rather large proportion of new genes is typical for EST projects of filamentous fungi, where usually 50–65% of the sequences obtained represent unknown genes (Skinner et al., 2001) compared with only 20–25% for plant ESTs (Ronning et al., 2003).
When the 20 most abundant ESTs of each fungus, representing 19.8% of the L. bicolor and 31.7% of the P. microcarpus dataset, were compared, only one homologous EST of each set could be found in the dataset of the other fungus. The total percentage of TCs (tentative consensus sequences) shared between L. bicolor and P. microcarpus tissues was about 5% (based on nucleotide sequence similarity). A similar low percentage of shared TCs was also found when the EST sets of either L. bicolor or P. microcarpus were compared with those of two saprophytic fungi, Agaricus bisporus and Pleurotus ostreatus.
The high percentage of unknown genes as well as the low percentage of homology between different datasets of basidiomycetes illustrates the limitations for ectomycorrhizal research. By contrast to ascomycetes, where a number of well investigated models and thus large EST datasets as well as genome sequences exist (Saccharomyces, Neurospora, Aspergillus, Magnaporthe), only a very few basidiomycetes (e.g. Ustilago, Cryptococcus) have been studied in closer detail. Thus, compared to the large biodiversity found in the fungal kingdom, only a relatively small number of fungal DNA sequences in general and basidiomycete sequences in particular are available in databases and could be used for homology analysis. In addition, fungal genes are much more variable in their DNA sequence than plant genes, frequently making it difficult to identify proteins of homologous function in different species.
In silico transcript profiling
For large EST sets, digital analysis of gene expression (in silico transcript profiling) can be performed by comparing the number of ESTs for a given gene within EST populations obtained from cells after different treatments (Ewing et al., 1999). Peter et al. compared in silico transcript-profiling for the 20 most abundant ESTs of both datasets with expression studies obtained by macro-array hybridization. Five of the 20 most abundant ESTs of L. bicolor and eight of P. microcarpus were also detected as highly expressed by the array technique. However, several transcripts identified by microarray analysis as highly expressed in both samples (different C-regimes for L. bicolor and nonmycorrhizal vs mycorrhizal hyphae for P. microcarpus) were found only in one of the two EST libraries. These data indicate that in silico profiling is robust enough to detect the most abundant transcripts; however, they also show that results of digital profiling should be very carefully interpreted, as they are often subject to technical limitations (e.g. a limited number of ESTs). Therefore, unless verified by other techniques (microarray analysis, Northern blot, RT-PCR), the data obtained by in silico transcript profiling can only provide an initial glimpse about gene expression.
Even though several independent EST projects are now in progress, only a limited number of fungal genes can be discovered by this strategy due to redundancy (which increases with sample number) in the DNA sequences obtained. In conclusion therefore for the establishment of a detailed picture of fungal gene expression during ectomycorrhizal development and function, which will be available in the near future for one of the angiosperm models (poplar), a genome project for at least one of the well-established model fungi is essential.