In the early 20th century, coffee wilt disease devastated coffee farms across Africa, with
catastrophic impacts on production and farmers: millions of dollars were lost, the major
coffee variety Coffea excelsa failed commercially, and widespread farm closures occurred.
Improved sanitation measures and plant breeding programmes in the 1950s briefly overcame
the effects of this first disease outbreak, but just two decades later a second prolific
epidemic attacked robusta coffee (C. canephora robusta) in Uganda, Tanzania and the
Democratic Republic of Congo. Separately, a third epidemic affected arabica coffee (C.
arabica) in Ethiopia. Coffee wilt disease is caused by Fusarium xylarioides, a soil-borne
fungal pathogen that induces vascular wilt on coffee trees. Little is understood about
coffee wilt disease and F. xylarioides compared with other Fusarium plant pathogens although
they all have a similar propensity to cause devastating crop and economic losses.
Over the past century, coffee wilt disease is the most significant disease to have affected
coffee production, and has caused over US$1 billion in losses to national economies in
central Africa. This thesis explores the use of living fungal strains preserved in culture
collections to elucidate the evolutionary past of a major disease outbreak. The history
of coffee wilt disease highlights the risk of disease re-emergence, and the importance of
rapid strategic action in response to outbreaks. The knowledge held in culture collections
around the world could help prepare us better for managing future outbreaks.
The first experimental chapter of this thesis (chapter two) used genome sequencing
of six historical culture collection strains spanning 52 years to identify the evolutionary
processes behind repeated outbreaks of coffee wilt disease. Phylogenomic reconstructions
and a screen for putative effector genes showed that the host-specific F. xylarioides
groups have diverged in gene content and sequence mainly by vertical processes within
lineages, and through expansion of certain carbohydrate-active enzyme families. A subset
of putative effector genes, however, showed evidence for horizontal acquisition and close
homology to genes from F. oxysporum.
Subsequently, the next chapter (chapter three) added seven more historic F. xylar-
ioides genomes and a long-read sequenced reference genome to test the F. xylarioides
host-specific groups found in chapter two and the nature of the putative horizontal transfer.
Four genetically differentiated populations were present: arabica, robusta and two
populations amongst the strains isolated in the 1950s from the initial 1920s-1950s outbreak.
In addition, six other Fusarium strains were sequenced from amongst species that
are also found on coffee wilt diseased trees: five F. oxysporum strains and one F. solani.
Analyses showed evidence for multiple horizontal transfers of over 20 kilobase long regions
between the F. oxysporum mobile chromosomes and the F. xylarioides populations. In
addition, F. xylarioides is enriched with miniature impala transposons that are found
in F. oxysporum and linked with the horizontal transfer of pathogenicity. These results
indicate that horizontal gene transfer occurred repeatedly in the F. xylarioides clade and
could be one of the drivers behind the repeated emergence of coffee wilt disease during
the last century via the transfer of gene- and repeat-rich genomic regions.
The final experimental chapter (chapter four) investigates whether any genes highlighted
in the previous chapters were expressed in arabica F. xylarioides strains infecting
arabica coffee host plants. RNA sequence data reveals major up-regulation of pectin lyase
enzymes, suggesting a key role for them in the onset of vascular wilt infection. It also
revealed up-regulation of three F. oxysporum secreted in xylem (SIX ) effectors, which
are all less than 1 kilobase from miniature impala transposons. This, combined with the
close phylogenetic relationship with F. oxysporum for a subset of several hundred up-regulated
genes, highlights the potential role of horizontal gene transfers in the spread of
pathogenicity between F. oxysporum and F. xylarioides in shared vascular wilt infections.
These results demonstrate how comparative genomics and transcriptomics of historic
strains can reveal mechanisms that allow fungal pathogens to keep pace with our efforts
to control them. Knowledge of horizontal transfer mechanisms and putative donor taxa
could help to design future intercropping strategies that minimize the risk of new and
emerging pathogens via effector gene transfers between closely related Fusarium taxa.