The genome sequences of three Aspergillus fungi are reported in this issue: Aspergillus oryzae1, used in making the Japanese drink sake; the human pathogen Aspergillus fumigatus2; and the genetic model species Aspergillus nidulans3. The 185 known species of Aspergillus include 20 human pathogens, numerous plant pathogens and a variety of species that we use to produce foods, chemicals and industrial enzymes. The genomes provide a wealth of information about the evolution of this fascinating group of organisms, and about the beneficial or detrimental characteristics of each species.

The sequences, published by teams from Japan, the United States and Europe, cover an average of nearly 95% of each genome. In total across the three species, more than 95 megabases have been sequenced, crammed with over 33,500 protein-coding genes contained on 24 chromosomes (eight chromosomes per species). By comparison, the human genome has about 30,000 protein-coding genes in 3,000 megabases.

Aspergillus oryzae has been used for nearly a thousand years to produce traditional Japanese fermented foods and drinks. Its genome1 has about seven to nine megabases more DNA than A. fumigatus and A. nidulans. To account for this, the authors propose that some genes were transferred to A. oryzae from other species during evolution. The extra DNA stretches are dispersed throughout the genome and are enriched in genes involved in the synthesis and the transport of numerous secondary metabolites — the chemical compounds in an organism that are not directly involved in normal growth, development or reproduction. Secondary metabolites are often specific to one or a few species, so they provide a window on the particular biology of the species.

Species closely related to A. oryzae, such as Aspergillus flavus and Aspergillus niger, have similar gene acquisitions. For instance, the toxic A. flavus has 25 genes encoding proteins involved in the pathway that produces the poisonous ‘aflatoxin’. These genes are present in A. oryzae but are not expressed. It is likely that an ancestor of A. flavus passed these genes to A. oryzae, and that they were then inactivated during the subsequent evolution of A. oryzae.

Aspergillus fumigatus is a potentially deadly human pathogen and a major allergen. The A. fumigatus sequence2 pinpoints nine previously unknown allergens, numerous genes involved in the production of specific secondary metabolites, and a set of essential genes that may be potential targets for drugs. However, the factors that underlie the pathogenicity of this species are complex, and their identification required other approaches to complement the genome analysis. For instance, for A. fumigatus to thrive inside warm-blooded creatures such as ourselves, it must be able to tolerate our high body temperature (compared with that of the external environment). Using DNA microarray analysis, a set of ‘thermotolerance genes’ whose activity increases at 37 °C has been identified. But it seems that warming up to 37 °C is insufficient to turn on many genes that are associated with virulence in this species.

Aspergillus nidulans has long been a model organism used to study the genetics of fungi. Its genome3 was crucial for the comparative analysis of the three aspergilli, but it also had some features of its own to reveal. For instance, the regulation of several of its genes was clarified, with the identification of putative binding sites for gene regulatory factors and control elements, as well as many short open reading frames that lie upstream of genes; these short sequences may stall the expression of neighbouring genes. In addition, the sequence disclosed many previously unknown genes involved in peculiar metabolic (fatty-acid oxidation), developmental (polarized growth) and DNA-repair pathways.

The three species diverged several hundred million years ago, and their genomes differ considerably3. There are almost 3,000 proteins that are closely related, or ‘homologous’, among the genomes. On average, these proteins have only 68% of their constituent amino acids present in all three genomes — a value comparable to that of proteins homologous between mammals and fish, which diverged around 450 million years ago. Nevertheless, the order of the homologous proteins along chromosomes (synteny) is conserved in the three species, indicating that no whole-genome duplication occurred during evolution. However, large regions lack any synteny because of small tandem repeats, gene rearrangements in the chromosome extremities, and considerable random breakage and rearrangements of syntenic blocks. Such genome reorganization is seen to a greater extent in A. oryzae than in A. fumigatus. The rates of amino-acid evolution within homologous genes are similar for all three species, so the evolution of large structural rearrangements does not parallel the rate of individual amino-acid changes.

The chief revelation from the three-genome comparison is the mating systems in A. fumigatus and A. oryzae. Sexual reproduction in yeast can take place only between individuals of opposite mating type, or ‘sex’, as determined by mating-type genes. Aspergillus nidulans has two mating-type genes: one contains an alpha box and the other a high-mobility-group (HMG) domain. So each cell can have two sexes at once, and A. nidulans is self-fertile. Aspergillus nidulans can also reproduce asexually by ‘mitotic reproduction’, creating spores that are sprinkled by structures known as conidiophores.

Aspergillus fumigatus and A. oryzae were believed to reproduce only through the asexual mitotic process. Unexpectedly, however, the A. oryzae and A. fumigatus genomes each have a mating-type gene: the A. oryzae sequence contains an alpha mating-type gene, whereas the A. fumigatus sequence has an HMG mating-type gene. These genes occupy nearly identical positions in their respective genomes, with conserved synteny for 1.7 megabases on either side. In addition, 215 genes implicated in different phases of the A. nidulans mating process occur in A. oryzae and A. fumigatus. These and other recent data4 raise the possibility that A. fumigatus and A. oryzae are heterosexual, and that conversion of bisexuality to heterosexuality occurred during the evolution of the Aspergillus genus (Fig. 1).

Figure 1: Evolutionary model of the Aspergillus mating genes.
figure 1

In a bisexual common ancestor, the two mating-type genes, alpha (red) and HMG (blue), were fused head to tail on the same chromosome and share a similar flanking, regulatory region (green). In the bisexual A. nidulans, the chromosome is broken and the two mating-type genes end up on different chromosomes (with their flanking regions). In the ancestor of A. oryzae and A. fumigatus, the two mating-type genes dissociate in different strains, but remain flanked by similar genes. After speciation, both A. fumigatus and A. oryzae become fully heterosexual, with some isolates having only the alpha mating type and others only the HMG mating type, both being in similar chromosomal environments.

Full size image

These reports describe only the initial examination of the genomes, of course, and the sequences provide much scope for further analyses. The sequencing of other Aspergillus genomes is under way and will provide an even broader perspective on the biology and evolution of these fungi. The most keenly anticipated Aspergillus sequence is that of A. niger, which has long been used in the industrial production of citric acid5. The commercial significance of several Aspergillus species has meant that their genome sequences, including that of A. niger, have been kept behind the closed doors of biotechnology companies for some time. However, this practice seems to be changing: a consortium of Japanese companies has agreed to release its A. oryzae sequence, and Monsanto provided access to its A. nidulans genome sequence, so that they could be added to the publicly funded sequences now published. And, fortunately, the US Department of Energy has undertaken to complete one of the industrial A. niger sequences (which is currently of low coverage) to make public a useful version of this genome. Perhaps the time when genome sequences belong exclusively to industry is over.