The role of the F-Box protein Frp1 in pathogenicity of fusarium oxysporum

Previously, FRP1 has been identified in a mutant screen as a pathogenicity gene of the
plant pathogenic fungus Fusarium oxysporum f.sp. lycopersici. Deletion of the gene
confirmed the requirement of FRP1 for pathogenicity. FRP1 encodes an F-box protein and
in this thesis we set out to elucidate its molecular function further. Literature on fungal Fbox proteins in general is reviewed in Chapter II and in Chapter III the F-box protein
arsenal of four Fusarium species is analyzed by mining their genome sequences.
Furthermore, the function of the Frp1 F-box domain is studied (Chapter IV), the phenotype
of the frp1 mutant strain examined (Chapter V) and the involvement of CreA in Frp1
function investigated (Chapter VI).
Many F-box proteins play a key role in the degradation of other proteins. They recruit such
proteins to a ubiquitination machine called the SCF complex where they become
ubiquitinated, which commonly leads to recognition and degradation by the proteasome.
Binding to Skp1, a component of the SCF complex, is a strong indication that Frp1 recruits
targets to this complex. Identification of such target proteins of Frp1 could give clues about
its molecular function(s) and, therefore, several screens were designed to identify a Frp1-
target. Unfortunately, no target protein was found. This raised the question whether Frp1
actually recruits target proteins. To investigate this, mutants of Frp1 were created with
mutations in the F-box domain that abolished the binding of Frp1 to Skp1. If Frp1 would
function through binding to Skp1, then it would be expected that these mutations would
affect the function of Frp1 and, hence, pathogenicity. However, this appeared not to be the
case as the Skp1-nonbinding versions of Frp1 could restore pathogenicity of the frp1
mutant strain. This suggests that the main function of Frp1 may not be recruitment of other
proteins to an SCF complex (Chapter IV).
To find out why the frp1 mutant is unable to cause disease, its phenotype was
investigated in more detail. We found that the mutant is impaired in root colonization as
well as in root invasion. These defects are accompanied by impaired growth on a broad
array of alternative (non-sugar) carbon sources such as organic acids, amino acids and
polysaccharides. We also found that the production of cell wall degrading enzymes
(CWDE) is seriously affected in the frp1 mutant, which is likely related to the reduced
growth on polysaccharides. The colonization defects on tomato roots could be restored by
addition of sugars like glucose or sucrose. However, addition of sugars could not restore
the root invasion defects as shown by the lack of expression of the GFP-reporter gene
behind the SIX1-promoter (the SIX1 gene is specifically expressed inside roots and is
therefore a marker for root invasion). To find out whether reduced growth on C2-carbon
sources like ethanol was (partly) responsible for the non-pathogenic phenotype of the frp1
mutant, we created an ICL1 deletion mutant. Icl1 is a key enzyme of the glyoxylate cycle,
which is required for the assimilation of C2-carbon sources. We found that this mutant -like
the frp1 mutant,- is impaired in growth on organic acids, ethanol and fatty acids. However,
growth on polysaccharides was normal and this mutant was still able to colonize roots and
fully cause disease on tomato plants. Therefore we concluded that the lack of CWDE gene
expression is sufficient to explain the loss of pathogenicity of the frp1 gene, although
other factors, such as utilization of amino acid as a carbon source, could not be ruled out
(Chapter V).
The genes required for assimilation of the carbon sources on which the frp1 mutant
shows impaired growth are generally under carbon catabolite repression. In fungi, this
process is regulated by the transcriptional repressor CreA. We considered the possibility
that Frp1 is required for the derepression of these genes by directly or indirectly regulating
CreA. The relation between CreA and Frp1 was therefore investigated and we found that
the two proteins together control the expression of CreA-repressed genes (Chapter VII).
Whereas CreA is mainly (but not exclusively) responsible for repression, Frp1 is required
for activation of such genes. Although knock-out of CREA failed, we were able to make
mutants in which GST-tagged CREA, had replaced the native CREA gene. Apparrently, the
GST-tag on CreA affects CreA function since replacement of native CREA by GST::CREA
in the wild type as well as in the frp1 mutant background caused a general growth
reduction and the inability to repress gene expression during growth on glucose.
Remarkably, GST::CREA restored the ability to express CWDE genes in the frp1
mutations, as well as growth on alternative carbon sources and infection of tomato. How
Frp1 influences CreA and activates gene expression remains unknown, but it probably
does not so via SCF-mediated protein ubiquitination, as explained in Chapter IV. We also
could not demonstrate a direct interaction between Frp1 and CreA. Further, accumulation
of GFP-tagged CreA in the nucleus occurred both on glucose and on ethanol, a
derepressing carbon source, and was not affected by deletion of FRP1.
In conclusion, analysis of the function of the F-box protein Frp1 in the plant pathogenic
fungus Fusarium oxysporum revealed that it functions independently from an SCF
complex, it is involved in carbon catabolite derepression and, together with CreA, affects
expression of CWDE genes during infection.