2008: the genome of the first mycorrhizal fungus

In March 2008 Nature published the genome of the bicoloured deceiver, Laccaria bicolor, produced by Francis Martin's team. It was the first sequenced genome of an ectomycorrhizal fungus — the kind of organism on which boreal, temperate and montane forests depend. The assembly runs to about 65 megabases and contains roughly 20,000 predicted protein-coding genes together with a very large number of transposons and repeated sequences.

The most important finding concerned a battery of small secreted proteins, abbreviated SSPs, of unknown function and resembling pathogen effectors. Several of them are expressed only in symbiotic tissues, and the most highly expressed one accumulates in the hyphae colonising the host root. Everything indicates that these proteins have a decisive role in establishing the symbiosis.

An absence proved just as telling. The Laccaria genome lacks carbohydrate-active enzymes for degrading plant cell walls, while retaining the ability to break down other polysaccharides. The symbiont therefore cannot eat its partner even if it wanted to, and can at the same time live in soil as a saprotroph. The dual lifestyle, previously inferred indirectly, was given a basis in genes.

The result changed the questions people ask about mycorrhiza. Instead of describing it as a general exchange of carbon for mineral nutrients, researchers could start studying the specific molecules through which fungus and root communicate. Later mycorrhizal genomes showed that the loss of plant-cell-wall-degrading enzymes has happened repeatedly and independently in evolution.

What the genome suggested about root partnership

The haploid Laccaria bicolor strain S238N-H82 was sequenced. Its scaffold assembly captured 64.9 Mb, and the JGI pipeline predicted 20,614 protein-coding gene models. Transposable elements accounted for 21% of the genome. Among 2,931 predicted secreted proteins, the team selected 278 small cysteine-rich proteins; five of the 20 most strongly upregulated fungal transcripts in ectomycorrhizal root tips encoded such proteins. This produced a focused set of candidates for communication with the host.

Absence was equally informative. The repertoire of enzymes for degrading plant cell walls was sharply reduced: the genome contained one candidate GH5 cellulase with a CBM1 module and no GH6 or GH7 cellulases. At the same time, 116 secreted proteases were predicted, consistent with mobilizing nitrogen from organic matter. The paper kept prediction separate from demonstrated function: the genome generated testable hypotheses about symbiosis, but did not automatically prove the role of every predicted protein.