tions soon followed (Inohara et al. 2001). The
common NBS-LRR architecture of plant and
animal immune receptors could either be
explained by parallel evolution or result from
convergent evolution. Several authors have
attempted to resolve this issue by phylogenetic
analyses in particular by tracing the phylogenies of NBS-SSFR proteins (proteins with
NACHT or NB-ARC nucleotide-binding
domains followed by superstructure-forming
repeats) in different phyla. A recent study supports the convergent evolution model for emergence of the NBS-LRR architecture (Urbach
and Ausubel 2017). The conclusion is based
on the observation that proteins ancestral to
plant R-proteins and animal NLRs did not
have an NBS-LRR architecture. The NBS-LRR
architecture is in fact rare and occurs almost
exclusively in plants and metazoans. It is proposed that the last common ancestor of Rproteins and animal NLRs had a NBS-TPR
structure and that the path to the NBS-LRR
architecture probably involved intermediates
of a NBS-WD architecture (for R-proteins)
and non-repeat-associated NBS in the case of
animal NLRs. The authors conclude that “it
follows logically that the NBS-LRR architecture
of plant R-proteins and metazoans NLRs
evolved in independent events” (Urbach and
Ausubel 2017). However, when the authors ask
“what is so special about the NBS-LRR combination in the context of immune receptors” and
discuss “what makes LRR-repeats any more
suited to ligand-binding roles in immune
receptors relative to other repeat domains,”
we feel that their perspective does not take
into account the mounting and yet still partial
evidence that NBS-SSFR proteins might have an
immune-related function in fungi (and possibly
also in multicellular bacteria). The central point
here is that the existence of fungal proteins with
NBS-TPR and NBS-WD architectures which
induce programmed cell death in response to
non-self makes it reasonable to assume that
proposed more ancestral architectures (like
NBS-TPR or NBS-WD) already displayed
immune-related functions in a common ancestor. Urbach and Ausubel claim that NBS-TPR
and NBS-WD proteins in eukaryotes are poorly
characterized and that it is not known if they
have an immune function or play a role in
programmed cell death (Urbach and Ausubel
2017). We feel that the work on fungal NLRs
challenges this view: fungal NACHT and NBARC STAND proteins controlling programmed
cell death and non-self recognition are found in
fungi. Similarly in the case of prokaryotic
STAND proteins, Koonin has noted that the
abundance of genes with NBS-TPR and NBSWD architecture specifically in bacterial
lineages with complex multicellular structures
(Actinobacteria, Cyanobacteria, and some Proteobacteria) might be explained by the fact that
some of these proteins have roles in host
defense and programmed cell death (Koonin
and Aravind 2002). Likewise, NBS-TPR proteins were proposed to act as immune receptors
in brown algae (Zambounis et al. 2012). Urbach
et al. show that acquisition of the NBS-LRR
architecture in plants and animals is a relatively
late event and occurred independently in plants
and animal lineages (Urbach and Ausubel
2017). However if proteins with the other, possibly more ancestral E-NBS-SSFR architectures
(effector/nucleotide-binding site/superstructure-forming repeats) function as immune
receptors in fungi, multicellular bacteria, basal
metazoans, basal plants, stramenopiles, and red
algae (Chondrus crispus), then we are more
likely dealing with parallel evolution of immune
receptors with E-NBS-SSFR architectures (even
though the late event of acquisition of LRR
occurred as a convergent event). The existence
of NBS-TPR and NBS-WD with immunerelated functions represents—we feel—a
strong argument in favor of a global parallel
evolutionary history for all E-NBS-SSFR architectures, characterized here and there by convergent domain architecture re-inventions,
lineage-specific losses or expansions, and possibly horizontal gene transfer events. The perspective adopted on NLR evolution comes
down in some respect to a matter of semantics
and depends on how NLRs are defined. We
suggested, based on the functional studies in
fungi and the genomic information on NLRrelated proteins in different branches, that the
NLR designation can be safely extended to
6 NLR Function in Fungi as Revealed by the Study of Self/Non-self Recognition Systems
135
common NBS-LRR architecture of plant and
animal immune receptors could either be
explained by parallel evolution or result from
convergent evolution. Several authors have
attempted to resolve this issue by phylogenetic
analyses in particular by tracing the phylogenies of NBS-SSFR proteins (proteins with
NACHT or NB-ARC nucleotide-binding
domains followed by superstructure-forming
repeats) in different phyla. A recent study supports the convergent evolution model for emergence of the NBS-LRR architecture (Urbach
and Ausubel 2017). The conclusion is based
on the observation that proteins ancestral to
plant R-proteins and animal NLRs did not
have an NBS-LRR architecture. The NBS-LRR
architecture is in fact rare and occurs almost
exclusively in plants and metazoans. It is proposed that the last common ancestor of Rproteins and animal NLRs had a NBS-TPR
structure and that the path to the NBS-LRR
architecture probably involved intermediates
of a NBS-WD architecture (for R-proteins)
and non-repeat-associated NBS in the case of
animal NLRs. The authors conclude that “it
follows logically that the NBS-LRR architecture
of plant R-proteins and metazoans NLRs
evolved in independent events” (Urbach and
Ausubel 2017). However, when the authors ask
“what is so special about the NBS-LRR combination in the context of immune receptors” and
discuss “what makes LRR-repeats any more
suited to ligand-binding roles in immune
receptors relative to other repeat domains,”
we feel that their perspective does not take
into account the mounting and yet still partial
evidence that NBS-SSFR proteins might have an
immune-related function in fungi (and possibly
also in multicellular bacteria). The central point
here is that the existence of fungal proteins with
NBS-TPR and NBS-WD architectures which
induce programmed cell death in response to
non-self makes it reasonable to assume that
proposed more ancestral architectures (like
NBS-TPR or NBS-WD) already displayed
immune-related functions in a common ancestor. Urbach and Ausubel claim that NBS-TPR
and NBS-WD proteins in eukaryotes are poorly
characterized and that it is not known if they
have an immune function or play a role in
programmed cell death (Urbach and Ausubel
2017). We feel that the work on fungal NLRs
challenges this view: fungal NACHT and NBARC STAND proteins controlling programmed
cell death and non-self recognition are found in
fungi. Similarly in the case of prokaryotic
STAND proteins, Koonin has noted that the
abundance of genes with NBS-TPR and NBSWD architecture specifically in bacterial
lineages with complex multicellular structures
(Actinobacteria, Cyanobacteria, and some Proteobacteria) might be explained by the fact that
some of these proteins have roles in host
defense and programmed cell death (Koonin
and Aravind 2002). Likewise, NBS-TPR proteins were proposed to act as immune receptors
in brown algae (Zambounis et al. 2012). Urbach
et al. show that acquisition of the NBS-LRR
architecture in plants and animals is a relatively
late event and occurred independently in plants
and animal lineages (Urbach and Ausubel
2017). However if proteins with the other, possibly more ancestral E-NBS-SSFR architectures
(effector/nucleotide-binding site/superstructure-forming repeats) function as immune
receptors in fungi, multicellular bacteria, basal
metazoans, basal plants, stramenopiles, and red
algae (Chondrus crispus), then we are more
likely dealing with parallel evolution of immune
receptors with E-NBS-SSFR architectures (even
though the late event of acquisition of LRR
occurred as a convergent event). The existence
of NBS-TPR and NBS-WD with immunerelated functions represents—we feel—a
strong argument in favor of a global parallel
evolutionary history for all E-NBS-SSFR architectures, characterized here and there by convergent domain architecture re-inventions,
lineage-specific losses or expansions, and possibly horizontal gene transfer events. The perspective adopted on NLR evolution comes
down in some respect to a matter of semantics
and depends on how NLRs are defined. We
suggested, based on the functional studies in
fungi and the genomic information on NLRrelated proteins in different branches, that the
NLR designation can be safely extended to
6 NLR Function in Fungi as Revealed by the Study of Self/Non-self Recognition Systems
135