scenario of convergent evolution, implying that
this locus has been repeatedly co-opted as a
non-self recognition locus (Heller et al. 2018).
In C. parasitica, the vic4 locus displays two
incompatible alleles, vic4-1 and vic4-2 (Fig. 6.1a).
vic4-2 encodes a NLR with an NACHT NOD
domain and WD-repeats and a N-terminal signaling/effector domain with an a-/b-hydrolase
fold, while vic4-1 encodes a predicted small molecule kinase of the fructosamine kinase family
(Pfam PF03881) (Zhang et al. 2014). Further
functional studies will be required to determine
whether this system shares mechanistic similarities with the other incompatibility systems
involving NLRs, that is, whether the WD-repeats
are involved in the interaction with the putative
kinase and whether the predicted hydrolase
represents a cell death effector domain.
Although not all incompatibility systems
involve NLRs (Paoletti 2016), these studies (in
three different and distantly related fungal species) reveal a frequent occurrence of members
of this protein family among the products of het
genes and related genes. All these systems
involve a protein partner that appears to be
surveyed by the NLR. Still, the genetic architecture of these systems is variable (Fig. 6.1b). In
the het-e/het-c system, unlinked genes encode
the NLR and the putative “guardee,” leading to
a classical non-allelic incompatibility system.
In the case of het-z (PaPlp1/PaSec9), interactions between the NLR and the “guardee” are
non-allelic, but since the gene pair is encoded at
the same locus as a haplotype, classical genetics
identified this system as allelic. Finally at C.
parasitica vic4 locus, interactions appear allelic,
but the NLR and the “guardee” are encoded by
idiomorphs (sequences encoded at the same
locus but totally unrelated, a situation already
described for mating-type incompatibility in N.
crassa), (Glass et al. 1988).
III. In Silico Survey of Fungal NLR
Repertoires
The functional studies reported above show
that in several fungal species, proteins with a
tripartite domain architecture (effector or signaling domain/NOD/superstructure-forming
repeats) resembling plant and animal NLRs
operate in allorecognition and the control of
programmed cell death. These observations
support the hypothesis that NLR function is
conserved from fungi to animals and plants,
prompting the description of the general occurrence of NLR-like proteins in fungal genomes.
A survey of NLR occurrence, domain architecture, and variability was reported in 2014
(Dyrka et al. 2014). Here, we have repeated
this survey with a similar methodology but
with a larger dataset of complete fungal genomes now available (identified by a NCBI BioProject ID) (882 strains of 561 species versus
198 strains in 2014) (Fig. 6.2). Fungal genomes
harbor a variable number of NLR-encoding
genes. Considering only species with at least
one hit (487 strains out of 882), the median
number of NLRs per genome is 41 in all fungi
(with 42 in Ascomycota and 49 in Basidiomycota). The mean number is 57 (52 in Ascomycota and 83 in Basidiomycota). As described for
NLRs in plants, the number of NLRs varies
greatly between species, reaching 602 in Fibularhizoctonia (the so-called cuckoo fungus producing sclerotia mimicking termite eggs)
(Matsuura et al. 2009). Table 6.1 provides a list
of the species with the highest NLR repertoire
content, and Fig. 6.2b shows the abundance of
NLR hits by phylogenetic class. The top-ranked
groups are Basidiomycota, but NLR count can
also be very high in Ascomycota in particular in
species displaying symbiotic interactions with
plants such as endophytic species like Phialocephala subalpina or forming ectomycorrhizal
associations like Cenococcum geophilum. The
NLR repertoires in fungi are generally a bit
smaller than in plants and possibly larger than
in most metazoan lineages (although large variations in repertoire size between and within
lineages are also common) (Jones et al. 2016).
NLRs were found in a large majority of the
Pezizomycotina and Agaricomycetes genomes
but were absent from Ascomycota and Basidiomycota yeast genomes (Saccharomycetes, Schizosaccharomycetes, and Tremellomycetes,
which were dominated in the dataset by Cryptococcaceae) suggesting that presence of NLRs
is associated with multicellularity (Fig. 6.2b).
6 NLR Function in Fungi as Revealed by the Study of Self/Non-self Recognition Systems
127
this locus has been repeatedly co-opted as a
non-self recognition locus (Heller et al. 2018).
In C. parasitica, the vic4 locus displays two
incompatible alleles, vic4-1 and vic4-2 (Fig. 6.1a).
vic4-2 encodes a NLR with an NACHT NOD
domain and WD-repeats and a N-terminal signaling/effector domain with an a-/b-hydrolase
fold, while vic4-1 encodes a predicted small molecule kinase of the fructosamine kinase family
(Pfam PF03881) (Zhang et al. 2014). Further
functional studies will be required to determine
whether this system shares mechanistic similarities with the other incompatibility systems
involving NLRs, that is, whether the WD-repeats
are involved in the interaction with the putative
kinase and whether the predicted hydrolase
represents a cell death effector domain.
Although not all incompatibility systems
involve NLRs (Paoletti 2016), these studies (in
three different and distantly related fungal species) reveal a frequent occurrence of members
of this protein family among the products of het
genes and related genes. All these systems
involve a protein partner that appears to be
surveyed by the NLR. Still, the genetic architecture of these systems is variable (Fig. 6.1b). In
the het-e/het-c system, unlinked genes encode
the NLR and the putative “guardee,” leading to
a classical non-allelic incompatibility system.
In the case of het-z (PaPlp1/PaSec9), interactions between the NLR and the “guardee” are
non-allelic, but since the gene pair is encoded at
the same locus as a haplotype, classical genetics
identified this system as allelic. Finally at C.
parasitica vic4 locus, interactions appear allelic,
but the NLR and the “guardee” are encoded by
idiomorphs (sequences encoded at the same
locus but totally unrelated, a situation already
described for mating-type incompatibility in N.
crassa), (Glass et al. 1988).
III. In Silico Survey of Fungal NLR
Repertoires
The functional studies reported above show
that in several fungal species, proteins with a
tripartite domain architecture (effector or signaling domain/NOD/superstructure-forming
repeats) resembling plant and animal NLRs
operate in allorecognition and the control of
programmed cell death. These observations
support the hypothesis that NLR function is
conserved from fungi to animals and plants,
prompting the description of the general occurrence of NLR-like proteins in fungal genomes.
A survey of NLR occurrence, domain architecture, and variability was reported in 2014
(Dyrka et al. 2014). Here, we have repeated
this survey with a similar methodology but
with a larger dataset of complete fungal genomes now available (identified by a NCBI BioProject ID) (882 strains of 561 species versus
198 strains in 2014) (Fig. 6.2). Fungal genomes
harbor a variable number of NLR-encoding
genes. Considering only species with at least
one hit (487 strains out of 882), the median
number of NLRs per genome is 41 in all fungi
(with 42 in Ascomycota and 49 in Basidiomycota). The mean number is 57 (52 in Ascomycota and 83 in Basidiomycota). As described for
NLRs in plants, the number of NLRs varies
greatly between species, reaching 602 in Fibularhizoctonia (the so-called cuckoo fungus producing sclerotia mimicking termite eggs)
(Matsuura et al. 2009). Table 6.1 provides a list
of the species with the highest NLR repertoire
content, and Fig. 6.2b shows the abundance of
NLR hits by phylogenetic class. The top-ranked
groups are Basidiomycota, but NLR count can
also be very high in Ascomycota in particular in
species displaying symbiotic interactions with
plants such as endophytic species like Phialocephala subalpina or forming ectomycorrhizal
associations like Cenococcum geophilum. The
NLR repertoires in fungi are generally a bit
smaller than in plants and possibly larger than
in most metazoan lineages (although large variations in repertoire size between and within
lineages are also common) (Jones et al. 2016).
NLRs were found in a large majority of the
Pezizomycotina and Agaricomycetes genomes
but were absent from Ascomycota and Basidiomycota yeast genomes (Saccharomycetes, Schizosaccharomycetes, and Tremellomycetes,
which were dominated in the dataset by Cryptococcaceae) suggesting that presence of NLRs
is associated with multicellularity (Fig. 6.2b).
6 NLR Function in Fungi as Revealed by the Study of Self/Non-self Recognition Systems
127