A Fungal Sesquiterpene Biosynthesis Gene Cluster Critical For Mutualist-Pathogen Transition in Colletotrichum Tofieldiae

Article https://doi.org/10.
1038/s41467-023-40867-w
A fungal sesquiterpene biosynthesis gene

cluster critical for mutualist-pathogen
transition in Colletotrichum tofieldiae
Received: 5 August 2022 Kei Hiruma 1,2 , Seishiro Aoki 3, Junya Takino4, Takeshi Higa1,
Yuniar Devi Utami 1, Akito Shiina1, Masanori Okamoto 5, Masami Nakamura1,
Accepted: 11 August 2023
Nanami Kawamura 2, Yoshihiro Ohmori6, Ryohei Sugita 7, Keitaro Tanoi 6,
Toyozo Sato8, Hideaki Oikawa 9, Atsushi Minami 4, Wataru Iwasaki 3 &
Yusuke Saijo 2
Check for updates
1234567890():,;
1234567890():,;
Plant-associated fungi show diverse lifestyles from pathogenic to mutualistic

to the host; however, the principles and mechanisms through which they shift
the lifestyles require elucidation. The root fungus Colletotrichum tofieldiae (Ct)
promotes Arabidopsis thaliana growth under phosphate limiting conditions.
Here we describe a Ct strain, designated Ct3, that severely inhibits plant
growth. Ct3 pathogenesis occurs through activation of host abscisic acid
pathways via a fungal secondary metabolism gene cluster related to the bio-
synthesis of sesquiterpene metabolites, including botrydial. Cluster activation
during root infection suppresses host nutrient uptake-related genes and
changes mineral contents, suggesting a role in manipulating host nutrition
state. Conversely, disruption or environmental suppression of the cluster
renders Ct3 beneficial for plant growth, in a manner dependent on host
phosphate starvation response regulators. Our findings indicate that a fungal
metabolism cluster provides a means by which infectious fungi modulate
lifestyles along the parasitic–mutualistic continuum in fluctuating
environments.
Plants associate intimately with diverse microbes, including pathogens, conditions, without changing the microbial genomic sequences3,4.
nonpathogenic commensals, and beneficial (mutualistic) microbes that These findings imply that different lifestyles of plant-associated
promote plant growth. Plant–microbe interactions are typically con- microbes are continuous within the same plant species and even
text-dependent, as these microbes dynamically change their infection coexist within the plant individual. Moreover, these microbes possess
modes according to the environment and host conditions1,2. One the capacity to refine infection strategies according to the given host
microbe can switch between pathogenic and beneficial infection modes environments5,6. However, the mechanisms by which infectious
even in the same host, depending on the host and environmental microbes transit between their contrasting lifestyles remain elusive.
1
Department of Life Sciences, Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1, Komaba, Meguro-ku, Tokyo 153-8902, Japan. 2Department
of Science and Technology, Nara Institute of Science and Technology, Nara 630-0192, Japan. 3Department of Integrated Biosciences, Graduate School of
Frontier Sciences, The University of Tokyo, Chiba 277-0882, Japan. 4Department of Chemistry, Faculty of Science, Hokkaido University, Kita 10, Nishi 8, Kita-ku,
Sapporo 060-0810, Japan. 5Center for Bioscience Research and Education, Utsunomiya University, 350 Mine-cho, Utsunomiya, Tochigi 321-8505, Japan.
6
Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113–8657, Japan. 7Radioisotope Research Center,
Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan. 8Genetic Resources Center, National Agriculture and Food Research Organization, Ibaraki
305-8602, Japan. 9Innovation Center of Marine Biotechnology and Pharmaceuticals, School of Biotechnology and Health Sciences, Wuyi University, Jiang-
men, Guangdong 529020, China. e-mail: hiruma@g.ecc.u-tokyo.ac.jp
Nature Communications | (2023)14:5288 1

Article https://doi.org/10.1038/s41467-023-40867-w
Plants have evolved an elaborate system called phosphate star- six Ct strains belong to the same species (Supplementary Fig. 1).
vation response (PSR) to cope with low inorganic phosphate (Pi) Notably, however, these five strains exhibited different growth mor-
conditions7. Under low Pi, A. thaliana plants induce extensive tran- phology during in vitro growth (Fig. 1a and Supplementary Fig. 2a).
scriptome reprogramming, mainly through the R2R3-MYB family This prompted us to investigate the possible intraspecies varia-
transcription factors PHOSPHATE RESPONSE1 (AtPHR1) and related tions of Ct in plant infection effects and strategies. We compared six Ct
PHR1-LIKE1 (AtPHL1)8,9. Plant adaptation to Pi deficit involves the acti- strains in their inoculation effects on A. thaliana Col-0 under gnoto-
vation of genes that promote phosphate absorption, allocation, and biotic low Pi conditions (50 μM KH2PO4). Five of the tested Ct strains,
usage. A. thaliana plants also accommodate beneficial root-associated including Ct61, significantly promoted plant growth, indicated by
endophytic fungus Colletotrichum tofieldiae (Ct) that helps nutrient primary root length and shoot fresh weight (Fig. 1b, c and Supple-
acquisition. Their hyphae acquire and transfer phosphorus to the host, mentary Fig. 2b, c). The results suggest that plant growth-promoting
providing an extension of the plant root system3. AtPHR1 and AtPHL1 (PGP) function under low Pi, at least discernible in A. thaliana, is
positively regulate phosphate transporter genes during Ct coloniza- extensively shared by Ct strains from various host and geographical
tion and are required for Ct-mediated plant growth promotion under niches. Notably, in contrast to these PGP strains, Ct3 severely inhibited
low Pi3. AtPHR1/AtPHL1 restricts fungal overgrowth and potential shoot growth under low Pi (Fig. 1b, c and Supplementary Fig. 2b, c). Ct3
virulence during beneficial interactions with Ct3, whereas negatively inhibited plant growth in additional 15 A. thaliana accessions (Sup-
regulating plant immunity against bacteria10,11. However, our knowl- plementary Fig. 2d) and also in Brassica rapa var. perviridis (Supple-
edge on the mechanisms by which AtPHR1/AtPHL1 play varied roles in mentary Fig. 2f, Left), whose growth was promoted by beneficial Ct4
different plant-microbe associations or host PSR influences microbial (Supplementary Fig. 2f, Right), on unsterilized low nutrient soil. These
lifestyles is limited. results suggest that Ct3 pathogenic lifestyle is common in a broad
Microbes have evolved an enormous repository of secondary diversity of Brassicaceae species, at least under the tested conditions.
metabolites, including plant hormone mimics for auxin and gibberellic Since Ct61 PGP function is specific to low Pi3, we tested whether
and abscisic acid (ABA). Whereas more than half of the genes are phosphate availability influences Ct3 and Ct4 infection phenotypes.
organized in operons in bacteria, functionally related genes are typi- Under normal Pi conditions (625 µM), Ct3 caused plant growth inhi-
cally distributed across the genome in eukaryotic fungi12. However, bition and leaf chlorosis (Fig. 1d and Supplementary Fig. 2c, g),
fungal secondary metabolite biosynthetic genes, as well as their reg- whereas Ct4 did not cause plant growth inhibition or leaf chlorosis
ulatory genes, are often clustered in genetic loci13,14. The necrotrophic (Fig. 1c, d). Under low Pi, Ct3 compromised plant growth but did not
fungal pathogen, Botrytis cinerea, harbors a cluster of biosynthetic cause leaf chlorosis or reduce leaf chlorophyll contents (Fig. 1c, d;
genes for ABA and produces ABA in culture15–17. There is a good cor- Supplementary Fig. 2c, g). In contrast to Ct3, Ct4 promoted plant
relation between the fungal ABA production and the host ABA sig- growth (Fig. 1d). Since sucrose increased and accelerated Ct3 growth
naling activation during B. cinerea infection, which facilitates the but not Ct4 growth in culture (Supplementary Fig. 2h), we hypothe-
suppression of immune-related genes in A. thaliana18,19. However, the size that Ct3 pathogenesis is dependent on host photosynthate
expression patterns of fungal ABA biosynthesis genes during plant supply and is alleviated when its supply is likely reduced under
infection remain elusive. The possible significance also remains to be phosphate deficiency.
explored for these ABA and/or other secondary metabolite biosynth- To trace fungal colonization dynamics, we generated transgenic
esis genes in the activation of host ABA pathways and fungal virulence. fungal strains expressing cytoplasmic GFP (Ct3-GFP or Ct4-GFP) under
Furthermore, these secondary metabolism genes are often not constitutive GPDA regulatory DNA sequences3. Conidia of the trans-
expressed even during host interactions in conventional laboratory genic Ct strains were then inoculated onto A. thaliana roots, expres-
settings20,21. Consistently, disruptions of individual genes in secondary sing an aquaporin PIP2A fused with mCherry, a plasma membrane/ER
metabolism clusters typically do not alter fungal phenotypes in- marker in living host cells3. Up to 5 days post-inoculation (dpi), both
planta22. These limitations have hampered the precise determination Ct3 and Ct4 hyphae both reached the cortex cell layer in the root
of the roles of secondary metabolism clusters in plant-infecting fungi. (Fig. 1e, f, Supplementary Movie 1, Supplementary Movie 2). The
In the present study, we describe a pathogenic strain, Ct3, causing intracellular hyphae of Ct3 and Ct4 were weakly labeled with the
severe plant growth inhibition, in contrast to beneficial strains pre- mCherry-expressing plant plasma membrane, indicating that these Ct
vailing in Ct. Comparative genomics and functional analyses between strains colonize inside the roots of A. thaliana, as previously described
these Ct strains indicate that, following root colonization, Ct3 displays for Ct613. The results indicate that Ct3 and Ct4 are either pathogenic or
transcriptional activation of biosynthesis gene clusters, putatively for beneficial fungi colonizing A. thaliana, respectively, under the tested
ABA and the associated sesquiterpene metabolite botrydial (desig- conditions.
nated ABA-BOT), thereby promoting root infection and pathogenesis. To grasp the genomic basis for the intraspecies variations
Conversely, Ct3 colonization gives the host benefits when this fungal between pathogenic Ct3 and beneficial Ct4, we determined their
cluster is genetically disrupted or transcriptionally suppressed at whole-genome sequences using a long-read generating PacBio
mildly elevated temperatures in a manner dependent on sequencing platform. PacBio long reads provided high-quality genome
AtPHR1/AtPHL1. These findings highlight the key role of fungal sec- assemblies for all strains, ranging from 53 to 55 Mb, with similar gene
ondary metabolites in the infection-mode transition of plant- numbers for candidate effector proteins, carbohydrate-active
associated fungi. enzymes, and secondary metabolism clusters (Supplementary
Table 2). Whole-genome alignment between Ct3 and Ct4 showed a
Results high degree (Median: >98.6%) of nucleotide identity between the two
A Ct strain severely inhibits plant growth in a nutrient- genomes, despite their contrast infection strategies. In contrast, a
dependent manner comparison between the beneficial strains Ct61 and Ct4 showed
A Ct strain, Ct61, isolated from a wild A. thaliana population in Spain, numerous genomic rearrangements and a lesser degree of nucleotide
promotes plant growth under low Pi conditions by transferring phos- identity (88.9%), implying a genome-wide divergence between Ct61
phorus to the host3. In addition to Ct61, five different Ct strains have and Ct4, despite their similarities in host benefits (Fig. 2a). These
been isolated from various plant species and geographical locations results suggest that nucleotide identity scores largely reflect the geo-
(Supplementary Table 1)23,24. Molecular phylogenetic analysis using six graphical distances of their origins. However, a molecular phyloge-
molecular markers conventionally utilized for the identification of netic analysis using the conserved 1509 single-copy genes among the
Colletotrichum species21,23–26 is consistent with the view that the tested tested 71 fungal species suggests that beneficial Ct strains have evolved

d
p < 0.0001
1000
b c 1000
Chlorophyll a + b (μg/g SFW)

Chlorophyll a + b (μg/g SFW)
* * * * * *
750 *
750
0.03 500
Shoot Fresh Weight (g)
500
250 250
0.02 0 0
Mock Ct61 Ct4 Ct3
+ Pi - Pi + Pi - Pi
0.01
e f
Invading hyphae Invading hyphae
Ct3 Ct4
g Intra-cellular hyphae Inter-cellular hyphae h Intra-cellular hyphae Inter-cellular hyphae

Ct3 Ct3 Ct4 Ct4
from their pathogenic relatives, such as C. incanum (Fig. 2b; Supple- effects of beneficial Ct strains were not yet apparent (Supplementary
mentary Data 1). It seems likely that beneficial lifestyles have evolved in Fig. 3a). At 24 dpi, the degree of plant growth inhibition by Ct3 was
pathogenic ancestors of Ct species. lowered under Pi deficiency compared with Pi sufficiency (Supple-
mentary Fig. 3b), whereas PGP by beneficial Ct61 and Ct4 became
Pathogenic Ct3 activates host PSRs and ABA signaling pathways discernible only under Pi deficiency (Fig. 1b and Supplementary
during early root infection Fig. 3b). In contrast to these Ct strains, the KHC strain (previously
We next investigated plant responses following root inoculation with classified as C. tofieldiae), closely related to C. higginsianum (Fig. 2b
beneficial and pathogenic Ct strains via RNA-seq analysis at two-time and Supplementary Fig. 1), did not alter plant growth under low Pi
points, 10 dpi and 24 dpi (Fig. 3a). At 10 dpi, Ct3 caused plant growth condition (Fig. 3a and Supplementary Fig. 3b), displaying a non-
inhibition under both normal and low Pi conditions, whereas PGP pathogenic lifestyle.

Fig. 1 | Root infection of a Colletotrichum tofieldiae strain (Ct3) severely inhibits different means between mock and Ct3-inoculated plants (±SD, p < 0.0001, two-
plant growth in A. thaliana. a Growth of Ct strains identified from various geo- tailed t-test). e–h Confocal microscopic images of Ct3 (e, g) or Ct4 (f, h) expressing
graphical locations in Mathur’s nutrient media after 3 days of incubation. Bar = 1 cm. cytoplasmic GFP (green) and root cells of A. thaliana expressing PIP2A-mCherry
b Determination of A. thaliana shoots fresh weight following fungal inoculation (magenta). These experiments were repeated three times with similar results.
under low Pi, 24 days after seed germination in the presence of the indicated fungal Maximum projection of z-stack images shows Ct3 (e) or Ct4 (f) hypha penetrating a
strains. Boxplot represents combined results from three independent experiments. root epidermal cell surrounded by PIP2A-mCherry-labeled host membranes
Each dot represents individual plant samples (Mock: n = 51, Ct61: n = 42, Ct49: (arrowheads). Representatives of hyphal penetrations are shown as enlarged pro-
n = 43, Ct127: n = 39, Ct130: n = 44, Ct4: n = 44, Ct3: n = 41 biologically independent jected images of different optical sections and orthogonal views made from areas
samples). The median values are described within each boxplot. Asterisks indicate indicated by white dotted line. Maximum projection of z-stack images shows Ct3
significantly different means between mock and fungal-inoculated plants (p < 0.01, (g) or Ct4 (h) hypha elongating intra- or inter-cellular spaces. Intra- or inter-cellular
two-tailed t-test). c A. thaliana plants grown in normal or low Pi conditions with or hyphae were indicated by white arrowheads and orthogonal views were made from
without fungal inoculation (24 dpi). d Shoot chlorophyll a and b contents with or areas indicated by white dotted line. Cells allowing elongation of intra-cellular
without Ct3 or Ct4 inoculation under normal and low Pi (n = 3 for Ct3_Low Pi or hypha or contacting with inter-cellular hypha were surrounded by a yellow dotted
n = 4 for others biologically independent samples). Asterisks indicate significantly line. Bar = 10 μm.
We first focused on the divergence in plant responses between Upregulation of PSR genes was associated with an increase in shoot P
these Ct strains at 10 dpi, when Ct3 already caused plant growth concentrations following Ct3 inoculation under low Pi (Supplementary
inhibition (Supplementary Fig. 3a). Multidimensional scaling analysis Fig. 3f), coincident with alleviated pathogenesis. Notably, Ct3 inocu-
indicated a clear separation of plant transcriptome in response to Ct3 lation upregulated a subset of AtPHR1/AtPHL1 regulons, which were
compared to beneficial Ct and nonpathogenic KHC strains, suggesting otherwise not induced, under Pi sufficient conditions (Fig. 3c
distinct plant responses to Ct3 at this stage (Fig. 3b). The results (Cluster 3) and Supplementary Fig. 3e, 69 of 193 genes, FDR < 0.05).
revealed 758 A. thaliana genes that were differentially upregulated in The results suggested that Ct3 infection results in AtPHR1/AtPHL1
Ct3-infected roots compared with beneficial Ct strains, under normal regulon activation, at least during an early infection phase, despite the
or low Pi conditions (log2FC> 1, FDR < 0.05, Supplementary Data 2). eventual negative effects on plant growth.
K-means clustering classified these differentially expressed genes We then examined how AtPHR1/AtPHL1-dependent PSR influences
(DEGs) into three different clusters (Fig. 3c and Supplementary Data 3). pathogenic Ct3 lifestyles by testing for Ct3 inoculation phenotypes in
In Clusters 1 and 2, Gene Ontology (GO) categories related to ABA phr1 phl1 plants. Compared to WT, phr1 phl1 displayed severe growth
signaling were overrepresented (Fig. 3c and Supplementary Data 3), inhibition following Ct3 inoculation under low Pi (Supplementary
suggesting that Ct3 specifically induces host ABA responses. ABA Fig. 3g). The results indicate a critical role for AtPHR1/AtPHL1 in the
promotes plant adaptation to water deficit27 and influences immunity alleviation of plant growth inhibition under low Pi. Together with the
or susceptibility-related pathways28. qRT-PCR analyses validated the transcriptome analyses above (Supplementary Fig. 3), we infer from
increased expression of ABA-responsive plant genes, AtRD29A29 and the results that activation of AtPHR1/AtPHL1 regulons serves to restrict
AtMAPKKK1830, during pathogenic Ct3 infection (Fig. 3d, e). ABA Ct3 pathogenesis.
accumulation also increased in roots following Ct3, but not Ct4,
colonization (Fig. 3f). The results indicate that Ct3, but not Ct4, Putative ABA and BOT biosynthesis gene clusters are distributed
strongly induces ABA responses in roots. We then tested whether and across plant-associated fungi, possibly along with fungal
how host ABA contributes to Ct3-mediated plant growth inhibition, virulence
with plant mutants disrupted in ABA signaling, abi1-1C; ABA bio- To assess the basis for transcriptional activation of host ABA respon-
synthesis, aba1 and aba2-12; and ABA perception, pyr1-1 pyl1 pyl2 pyl4 ses, we examined fungal transcriptome profiles during pathogenic Ct3
pyl5. Ct3-mediated plant growth inhibition was alleviated in these ABA- infection (Supplementary Data 4). As Ct3-mediated ABA response
defective mutants under low Pi (Fig. 3g). In contrast, plant growth activation and plant growth inhibition were pronounced at 10 dpi
promotion via beneficial Ct4 under low Pi was not affected in abi1-1C (Fig. 3c and Supplementary Fig. 2g), we assembled Ct3 genes that were
(Supplementary Fig. 3c). These analyses suggest that pathogenic but differentially upregulated at 10 dpi compared to 24 dpi. This resulted
not beneficial lifestyles of Ct3 are dependent on the core ABA bio- in 304 fungal genes upregulated in an early phase under normal or low
synthetic and response pathways of the host. Pi conditions (Fig. 4a, log2FC > 1, FDR < 0.05). The number of these
During beneficial interactions under low Pi, Ct61 induced a subset genes was greater under normal Pi than low Pi (Fig. 4a), consistent with
of PSR-related genes in plants, including phosphate transporter AtPHT enhanced Ct3 impacts on plant growth under normal Pi (Fig. 1).
genes, at 24 dpi3. Since PSR-related transcriptional reprogramming Interestingly, 92 among 304 Ct3 genes were only expressed at 10
largely relies on AtPHR1 and AtPHL19, we next assessed the possible dpi under both Pi conditions, but their expression was below detect-
impact of beneficial Ct strains on AtPHR1/AtPHL1-regulated PSR- able levels at 24 dpi, coincident with fungal pathogenesis that was
related genes under low Pi at 24 dpi when PGP by beneficial Ct61 pronounced at 10 dpi (Supplementary Data 5). We noticed that puta-
and Ct4 became discernible. Cross-referencing our data with pre- tive ABA biosynthesis genes (Ct3ABA1, Ct3ABA2, and Ct3ABA3) were
viously described 193 AtPHR1/AtPHL1 regulons9 indicated increased included in the 92 pathogenesis-associated genes. They show high
expression for approximately half of them [88 of 193 genes (Ct61), or sequence similarity to ABA biosynthesis genes (BcABA1-BcABA3) in B.
99 of 193 genes (Ct4), FDR < 0.05] during beneficial Ct interactions cinerea (Supplementary Data 4). BcABA1 and BcABA2 encode cyto-
compared with the mock controls, specifically under low Pi (Supple- chrome P450, and BcABA3 encodes sesquiterpene synthase catalyzing
mentary Fig. 3d). In contrast, PSR gene activation was not increased the initial step from farnesyl diphosphate (FPP), in a fungal ABA bio-
during pathogenic or nonpathogenic interaction with Ct3 or KHC, synthesis pathway that is entirely different from that of plants16,17,31.
respectively, at 24 dpi (Supplementary Fig. 3d). At 10 dpi, however, Interestingly, in Ct3, these putative ABA biosynthesis genes are clus-
Ct3 strongly upregulated a subset of PSR-related genes under low Pi tered together with those highly related to the biosynthesis of botry-
compared with the mock controls (Supplementary Fig. 3e, 146 of 193 dial (BOT), an ABA-related sesquiterpene metabolite in B. cinerea
genes, FDR < 0.05), indicated by the induction of Clusters 2 and 3 (Fig. 4b, Supplementary Table 3)32,33. Five BOT biosynthesis genes
genes related to PSRs, in which GOs overrepresented “Cellular (named BOT1-BOT5) were co-activated with the three putative ABA
response to starvation (Cluster 2)” and “cellular response to phosphate biosynthesis genes during root colonization in Ct3 but not in other
starvation (Cluster 3)” (Fig. 3c and Supplementary Data 3). beneficial Ct strains (Fig. 4c, Supplementary Table 4), despite the high

a Median: Median:
Ct4_27
Ct4_25 Ct3_01
Ct4_23Ct4_ 24 0Mb Ct4_26 0Mb 0Mb
98.6%
0Mb
88. 9%
4Mb
0Mb
21 _22 0Mb
Ct4_ Ct40Mb Ct61_20
Ct61_18
0Mb
Ct61_21 Ct4_01
20 Ct61_19
0Mb
6
Ct4_ Ct3_
Ct61_1
0Mb 0Mb
0Mb _17 0Mb 0Mb 0Mb
Ct610Mb
4Mb
19 02
Ct4_ 4Mb _14 1_15 0Mb 0Mb
Ct61 Ct6 0Mb
0Mb
8 Ct4_
_1 02
Ct4
0M
b
3 0Mb
7 _1 4Mb
_1 Ct61
0Mb
Ct4 b 0M
0M b
6 12
1_
_1
0M
C0Mt6b
b
Ct40Mb Ct4
C
11 _0
t3
5
_0
1_ 3
_1
3 4M
Ct6b
Ct4
b
4M
0M
b
0M b
4
_1
Ct4
10
b
0M
0M
1_
b
Ct6
3
0M
Ct4
_1
b
0M
Ct4
_0
b
0M
4
12
Ct3_
09
4M
Ct4_
b
1_
044Mb
Ct6
0Mb
0Mb
0Mb
Ct4_
_11
Ct4
05
0Mb
_0 8
4Mb
0Mb
Ct61
Ct4_10
0Mb
0Mb
0Mb
Ct4_06
Ct3_05 4Mb
Ct4_09
Ct61_07
0Mb
0Mb
0Mb
Ct4_07
Ct4_08
0 Mb
Ct61_06
0Mb
Ct3_06
0Mb
0Mb
Ct4_08
4Mb
Ct4_07
4Mb
Ct61_0
0Mb
0Mb
Ct4_090Mb
5
07
0Mb
Ct3_
Ct4_
Ct4_ 0Mb
06
4Mb
10
4Mb
0Mb
Ct61
b
0M
4M
b
_0
Ct4 0Mb
4
1
_1
8
Ct4
_0
0M
Ct3
b
_0
5
_1 0Mb
2
b
0M
4M
b
Ct4
b
4M
4M
0M
b
b
Ct6
_1 0Mb
3
Ct4 9
_0
1_
Ct4
Ct3
03
_0
4 b
0M 4
_1
0M
b 0M
b b
Ct4
0M
4M 0 0M
b
5
b _1 _1
Ct3
Ct4_ 4M
b Ct4
03 0Mb
11 Ct6b 0M 6
1_ _1
0Mb
0Mb Ct3_ 02 Ct4
0Mb
3_12
0Mb
7
_1
Ct3_13Ct
4Mb 0Mb
Ct4_02 0Mb
0Mb
0Mb Ct4
18
Ct3_14
0Mb 4Mb
Ct4_01 0Mb
19 Ct4_
Ct4_
4Mb
0Mb
Ct61_0
1 0Mb
0Mb
Ct4_20
Median:
0Mb
0Mb 0Mb 0Mb 0Mb 0Mb
22
Ct4_26 Ct4_24 Ct4_ Ct4_21
Ct4_27 Ct4_25 Ct4_23
84. 9% Ct61
_14 Ct61_
6 Ct61
Ct61_1Ct61_1
0Mb
15 0Mb
0Mb
_18 Ct61_20
7 0MbCt61_1
0Mb
9 0MbCt61_21
0Mb 0Mb Ct3_01
4Mb
0Mb
C0Mt6b
1_
12
Ct61
0Mb
_1
3 0Mb Ct3_
02
4M
b
LCB identity:
11
0M
b
>97%
1_
C0Mt6b
Ct3
_0
10
3 4Mb
1_
Ct6
>90%
b
0M
0M
b
9
_0
Ct61
Ct3_
0Mb
>80%
04 4Mb
8
Ct61_0
<80%
0Mb
0Mb
Ct61_07
Ct3_05 4Mb
0Mb
Strain:
Ct61_06
0Mb
Ct3_06
0Mb
Ct3
4Mb
4Mb
Ct61
_05
0Mb
Ct4
0Mb
7
_0
4M
Ct3
b
Ct6
b
4M
1_
Ct61
04
b
0M
0M
b
8
_0
Ct3
4M
b b
Ct6 4M
1_ b
03 0M
9
_0
Ct3
0M
b
0Mb
4Mb
Ct61 10
_02 Ct3_
0Mb
0Mb
0Mb
0Mb
Ct3_11
Ct3_13Ct3_12
4Mb 0Mb
Ct61_01
b
Ct3 (Pathogenic)
100 97 Ct4 (Beneficial)
100 Ct61 (Beneficial)
100
C. liriopes_MAFF_242679 (Pathogenic)
100
100 C. incanum_MAFF238712 (Pathogenic)
C. incanum_MAFF_238704
100
C. spaethianum_MAFF_239500
C. sublineola_TX430BB
100 C. graminicola_M1.001
100
100 C. higginsianum_KHC
C. higginsianum_MAFF_245053
100
C. higginsianum_IMI_349063
100 100 C. higginsianum_MAFF305635-RFP
C. shisoi_PG-2018a
100 C. tanaceti
100 C. scovillei_Coll-365
100 100 100 C. scovillei_Coll-153
100 C. scovillei_Coll-524
100 C. scovillei_TJNH1
C. nymphaeae_SA-01
100 C. simmondsii_CBS122122
100 100 100 C. abscissum_Ca142
C. lupini_IMI_504893
100 100 C. filicis_CBS_101611
C. fioriniae_PJ7
C. salicis_CBS_607.94
100 C. orchidophilum_IMI_309357
C. chlorophyti_NTL11
43 Colletotrichum strains
Botrytis_cinerea_B05.10
0.05
Fig. 2 | Genomic analyses of beneficial and pathogenic Ct strains. a Syntenies of concatenated alignment of 1509 single-copy orthologous protein sequences from
Ct3, Ct4, and Ct61 genomes. The length of each contig was described as 0.8 Mb/ 70 Colletotrichum strains using IQ-TREE. Ultrafast bootstrap values (1000 repli-
scale. The percentage in the figure represents the median % of the nucleotide cates) are shown on the branches. B. cinerea B5.10 was used as an outgroup. The
similarity against the whole locally collinear block (LCB). Pairwise genome com- lifestyles of some Colletotrichum fungi in A. thaliana roots under low Pi were
parisons point to the highest similarity between Ct3 and Ct4 and the lowest investigated and are described in the tree (beneficial or pathogenic).
between Ct61 and Ct3. b Maximum-likelihood phylogenomic tree generated from a

a b c
1 2 3
3.5
Effects on plant shoot growth a p_Ct3e_1
4
Normalized expression value

Ct3_P_1 a P_Ct3e_1
Promotion Inhibition a p_Ct3e_2
Ct4_P_1 4
0.02 Ct3_P_2
KHC_P_1
KHC_P_2 a P_Ct3e_2 3.0
KHC_P_0Ct4_P_0 a p_Ct4e_0
Ct3_P_0
KHC_m_1 Ct4_P_2 a P_Ct4e_0
Ct4_m_2 a p_Ct4e_1
KHC_m_0 Ct4_m_1
3 2.5 cluster
Ct3_m_2 a P_Ct4e_1
= > > Ct3_m_0 Ct61_m_2 Ct61_P_0 a p_Ct4e_2 3

1
Ct3_m_1 KHC_m_2 Ct61_P_1 a P_Ct4e_2
0.00 Ct4_m_0 2
Ct61_P_2 a p_Ct61e_0
2.0 3
Ct61_m_1 a 0
P_Ct61e_0
M2
Ct4 Ct61 KHC Ct3 a p_Ct61e_1
2
Ct61_m_0 a P_Ct61e_1 2
Beneficial Nonpathogenic Pathogenic a p_Ct61e_2
a 2
P_Ct61e_2
1.5
-0.02
a _
p_CtKHCe_0
Condition: Normal Pi (625 μM KH2PO4) a _0
P_CtKHCe_0
a _1
p_CtKHCe_1
Low Pi (50 μM) a _1
P_CtKHCe_1
1 1 1.0
M_P_0 M_P_1
Timepoint: Early: 10 dpi (early colonization) a p_CtKHCe_2
Mock
Ct61
Ct3
Ct4
KHC
Mock
Ct61
Ct3
Ct4
KHC
Mock
Ct61
Ct3
Ct4
KHC
Mock
Ct61
Ct3
Ct4
KHC
Mock
Ct61
Ct3
Ct4
KHC
Mock
Ct61
Ct3
Ct4
KHC
M_m_1 a P_CtKHCe_2
M_P_2
Late: 24 dpi (late colonization) -0.04 a p_Me_0
M_m_0 M_m_2 a P_Me_0 Pi + - + - + -
-0.050 -0.025 0.000 0.025 a Pi
p_Me_1 +
M1 a P_Me_1
a GO
M 2
・reponse to ・reponse to ・cellular
term abscisic acid abscisic acid response
・reponse to ・cellular to phosphate
salt stress response starvation
・reponse to to starvation
water ・toxin catabolic
deprivation process
・reponse to
chitin
d e f g
* * * Ct3/Mock 0.57 0.77** 0.72 * 0.82** 0.76 **
0.100 120
0.006
AtMAPKKK18 / Plant ACT
0.008
AtRD29A / Plant ACT
ABA pg/mg DW
0.075
Shoot Fresh Weight (g)
80
0.004
0.006
0.050
40
0.002
0.025
0.004
0.000 0.000 0
k
t3
t4
k
k
t3
t4
t3
t4
oc
oc
oc
C
C
M
M
0.002
k
t3
k
t3
k
t3
k
t3
k
t3
oc
oc
oc
oc
oc
C
C
M
M
Col-0 abi1-1C aba1 aba2 pyr1
pyl1 2 4 5
Fig. 3 | Activation of the host ABA responses during Ct3 pathogenesis. biologically independent samples). f Root ABA content following fungal inocula-
a Experimental setup for dual RNA-seq analysis of A. thaliana and Ct tran- tion under low Pi (Mock: n = 4, Ct3: n = 3, Ct4: n = 3 biologically independent sam-
scriptomes. b Multidimensional scaling (MDS) analysis chart generated from plant ples). Asterisks indicate significantly different means compared to the mock
transcriptome profile during association with the indicated Ct and KHC strains controls (±SD, p = 0.002651, two tailed t-test). g Shoot fresh weight at 24 dpi with
under normal (P) and low Pi (m) conditions, with three biological replicates. Ct3 under low Pi conditions in the indicated plant genotypes (Col-0_Mock: n = 17,
c Expression patterns of A. thaliana genes differentially expressed during patho- Col-0_Ct3: n = 17, abi1-1C_Mock: n = 17, abi1-1C_Ct3: n = 17, aba1_Mock: n = 16,
genic Ct3 association compared with beneficial Ct associations (( | log2FC | >1, aba1_Ct3: n = 14, aba2_Mock: n = 18, aba2_Ct3: n = 17, pyr1 pyl1 2 4 5_Mock: n = 18,
FDR < 0.05), indicated by cluster analysis divided by kmean (K = 3, PAM clustering pyr1 pyl1 2 4 5_Ct3: n = 18 biologically independent samples). The median values are
with Jensen-Shannon distance). Genes listed in each cluster were subjected to GO described within each boxplot. Asterisks indicate significantly different fresh
analysis (The full gene list is available in Supplementary Data 3). d, e Expression of weight ratio (Ct3/Mock) compared with Col-0 (*p < 0.1, **p < 0.05, two tailed t-test
ABA-responsive A. thaliana genes (AtRD29A and AtMAPKKK18) at 10 dpi with Ct3 or (Col-0 vs. abi1-1C: p = 0.047, Col-0 vs. aba1: p = 0.08, Col-0 vs. aba2: p = 0.021, Col-0
Ct4 under low Pi (±SD, p = 0.025352 (d), p = 0.000267 (e), two tailed t-test) (n = 3 vs. pyr1 pyl1 2 4 5: p = 0.033)).
conservation of their cluster across these Ct genomes. Their encoded Our comparative genomic analysis indicated that putative ABA
amino acid sequences were nearly identical (particularly in Ct4; Sup- and BOT biosynthesis genes are conserved as a single cluster in ben-
plementary Fig. 4o: see below). These genes were silenced at 24 dpi eficial and pathogenic Ct genomes but are separate in the B. cinerea
when plant growth inhibited by Ct3 was alleviated (Supplementary genome (Fig. 4b). Furthermore, these genes, in particular of the BOT
Fig. 2g; Supplementary Table 4). Furthermore, pathogenic C. incanum cluster genes, show very high degrees of sequence conservation
displayed an upregulation of CiABA3 and CiBOT5 during root infection between Colletotrichum spp. (class Sordariomycetes) and B. cinerea
(Supplementary Fig. 4a). Therefore, co-activation of putative ABA and (class Leotiomycetes). Such sequence conservation was unexpected,
BOT biosynthesis genes was specifically associated with pathogenic given the considerable phylogenetic distance between the two fungal
lifestyles in the root-infecting Colletotrichum species. lineages, which diverged approximately 261.6 million years ago21.

a Normal Pi
217 63 24
Low Pi
b
ABA and BOT biosynthesis clusters
Botrytis cinerea B05.10
ABA4
Chromosome 8
ABA2 ABA1 ABA3
XP_001553966.2 XP_001553967.1 XP_024550393.1

XP_024550389.1
Chromosome 12
BOT4 BOT1 BOT2 BOT3 BOT5 BOT6
XP_024552380.1 BOT7
Putative ABA-BOT biosynthesis cluster

Colletotrichum tofieldiae Ct3
ABA2 BOT6 BOT7 ABA3 BOT2
contig5
BOT3 BOT5 BOT4 ABA1 BOT1
GKT60935.1 GKT60934.1 GKT60925.1 GKT60924.1
ABA4
contig2
GKT56071.1 GKT56074.1 GKT56075.1

GKT56072.1
c d
Putative ABA cluster

250
ABA+BOT Colletotrichum strains ABA BOT1-5 BOT6,7
200 ABA BOT1-5

ABA1-3 Diaporthe helianthi BOT6,7
ABA1-4
150 Botrytis cinerea B05.10 ABA BOT1-5 BOT6,7
FPKM
Monosporascus sp. CRB-8-3 ABA BOT1-5 BOT6,7

100
BOT cluster
HGT of BOT
50
Colletotrichum strains ABA BOT1-5 BOT6,7
ABA+BOT Botrytis cinerea B05.10 ABA BOT1-5 BOT6,7

0
Pi + - +- + - + - BOT
Ct3 Ct4 Ct3 Ct4 Diaporthe helianthi ABA BOT1-5 BOT6,7
CtABA3 CtBOT5
Monosporascus sp. CRB-8-3 ABA BOT1-5 BOT6,7
Fig. 4 | Expression and distribution of the putative ABA and BOT biosynthesis- respectively. Yellow arrows represent other genes. c CtABA3 and CtBOT5 mRNA
related genes in Ct are associated with fungal pathogenesis in Ct. a Venn dia- levels in Ct3 and Ct4 at 10 dpi, shown with the FPKM values being derived from RNA
gram showing Ct3 genes differentially upregulated in early colonization (10 dpi) sequencing analysis. d Phylogenetic relationship and genomic structures of puta-
compared with late colonization (24 dpi) phases ( | log2FC | >1, FDR < 0.05, Sup- tive ABA and BOT gene clusters. All four genera carrying putative ABA and BOT
plementary Data 4). Sixty-three Ct3 genes were significantly upregulated during gene clusters in their genomes were used for comparison. Organisms harboring
root colonization at 10 dpi in both Pi conditions. putative ABA and BOT bio- adjacent putative ABA and BOT clusters (ABA + BOT (ABA-BOT) clusters) in their
synthesis genes were included in the Venn diagram. b Genomic structures of Ct3 genomes are labeled with gray backgrounds. Phylogenetic analyses of these genes
putative ABA-BOT and B. cinerea ABA and BOT biosynthesis gene clusters. The blue in the clusters are shown in Supplementary Fig. 4b–l.
and red arrows represent the genes related to ABA and BOT biosynthesis,
To assess the evolutionary origin(s) of putative ABA and BOT but (4) that the putative ABA and BOT clusters have different evolu-
biosynthesis genes, we generated molecular phylogenetic trees for tionary histories, indicated by their divergence in molecular phyloge-
these genes from the Colletotrichum strains (Ct3, Ct4, C. liriopes, C. netic trees (Fig. 4d and Supplementary Fig. 4b–m). The substantial
spaethianum34, belonging to the spaethianum clade), and from the differences between species and gene trees suggest horizontal gene
GenBank non-redundant protein sequences (NR) database (Fig. 4d, transfers (HGTs) distributing the putative ABA and BOT biosynthesis
Supplementary Data 6 and 7). Phylogenetic analyses showed (1) that gene clusters among these plant-associated fungi. Our comparative
putative ABA and BOT cluster genes were distributed across distantly genomic analysis revealed that Diaporthe helianthi, a plant pathogen,
related plant-associated fungi in a manner different from their phylo- also possesses a single putative ABA-BOT biosynthesis gene cluster of
genic relationships (Supplementary Fig. 4b–m), (2) that putative ABA high synteny when compared with Ct (Supplementary Fig. 4n). On an
and BOT cluster genes each formed monophyletic groups (Supple- assumption that the putative ABA-BOT cluster originated only once
mentary Fig. 4b–m), (3) that genes in the putative ABA or BOT cluster according to the principle of parsimony, the results suggest a HGT of
each have similar evolutionary histories (Supplementary Fig. 4b–m), the BOT cluster to Botrytis, (Fig. 4d and Supplementary Fig. 4m, n).

Given that the putative ABA and BOT clusters have likely arisen mul- responses (Fig. 6b). Interestingly, AtMAPKKK18 induction was also
tiple times in different phytopathogenic fungi, it is conceivable that absent in the roots inoculated with Ct3Δbot fungi (Fig. 6b), suggesting
the acquisition of putative ABA and BOT biosynthesis gene cluster(s) a role for Ct3BOT genes in activating plant ABA responses. Further-
contributes to pathogenesis evolution in plant-associated fungi. more, root fungal biomass was dramatically reduced for Ct3Δaba and
Gene composition in putative ABA and BOT biosynthesis clus- Ct3Δbot fungi compared with WT fungi (Fig. 6c and Supplementary
ters was slightly different across species. The putative ABA bio- Fig. 7d-e). indicating that these fungal genes contribute to root infec-
synthesis cluster consists of ABA1-ABA3 in, e.g., Ct, C. incanum, and D. tion of Ct3. Since Ct3 but not Ct4 increases fungal growth in the pre-
helianthi or ABA1-ABA4 in, e.g., B. cinerea (Fig. 4d, Supplementary sence of sucrose (Supplementary Fig. 2h), we tested whether Ct3 root
Fig. 4b–e, and 4m, n). CtABA4 was separately located in Ct genomes infection influences sucrose accumulation in the host. Ct3 colonization
(Fig. 4b), and unlike Ct3ABA1-Ct3ABA3, its expression was not resulted in the accumulation of sucrose in roots, which was found to be
detected in Ct3 during root colonization (Supplementary Table 4), dependent on the Ct3ABA2 and Ct3BOT5 genes (Supplementary
suggesting that Ct3ABA4 is dispensable for Ct3 root colonization or Fig. 7f). These observations suggest that activation of the putative
pathogenesis in Ct3. The BOT gene cluster typically consists of BOT1- ABA-BOT cluster promotes sugar accumulation, which may turn, be
BOT7, where BOT6 and BOT7 correspond to Zn_clus (Zn2Cys6 tran- exploited by Ct3 for extensive fungal growth in roots.
scription factor) and Adh_short (retinol dehydrogenase 8) genes, Plant growth inhibition was also reduced when inoculated with
respectively, named in B. cinerea33. BOT6 and BOT7 show similar Ct3Δaba2 and Ct3Δaba3 fungi compared with WT fungi (Fig. 6d and
phylogenetic distributions with the other five BOT genes within Supplementary Fig. 7g, h). Ct3 virulence was, however, restored when
Colletotrichum and Botrytis strains, indicating functional coupling of ABA was exogenously applied (Supplementary Fig. 7i). Ct3ABA3-
these genes in these fungal lineages (Supplementary Fig. 4f–n). We mediated plant growth inhibition was reduced in the host mutants
confirmed that Ct3BOT6 and Ct3BOT7 were strongly expressed dur- defective in ABA signaling or biosynthesis (Fig. 6d). The results suggest
ing Ct3 pathogenesis (Supplementary Table 4). that biosynthesis of ABA by Ct3 requires the plant ABA core pathway to
promote fungal pathogenesis. Notably, exogenous ABA application
Ct3 produces intermediate metabolites of BOT biosynthesis also suppressed both plant growth and Ct4-mediated PGP under low Pi
during culture growth conditions (Supplementary Fig. 7j). The results are consistent with
Our phylogenetic analyses revealed a previously unsuspected diversity fungal-derived ABA playing a role in shifting from a beneficial to a
in the repertories of biosynthetic genes for putative ABA and BOT pathogenic lifestyle in Ct. Remarkably, Ct3Δbot5 fungi did not inhibit
among plant-associated fungi. We examined whether the biosynthetic but rather promoted plant growth under low Pi, reminiscent of Ct4
genes in Ct3 mediate the production of these metabolites as predicted, (Fig. 6e and Supplementary Fig. 7k). The mutation effects were abol-
like in B. cinerea. We generated fungal knockout mutants for Ct3ABA2, ished when Ct3BOT5 was introduced back into the Ct3Δbot5 (Supple-
Ct3ABA3, and Ct3BOT5 via homologous recombination. We then pro- mentary Fig. 7l). The results suggest that BOT biosynthesis is required
filed metabolites related to BOT biosynthesis in the WT, Ct3Δaba2, for Ct3 virulence, and its disruption even renders Ct3 beneficial for
Ct3Δaba3, and Ct3Δbot5 strains when cultivated in three distinct the host.
nutrient media (Mathurs, MPY, rice-based media). This detected two
previously reported intermediate metabolites in BOT biosynthesis, 4β- Ct3 shifts between pathogenic and beneficial lifestyles in a
acetoxyprobotryan-9β-ol (9) and 4β-acetoxyprobotryane-9β,15α-diol manner dependent on the temperature and host AtPHR1/AtPHL1
(10)35,36, previously reported in B. cinerea in Ct3 (Fig. 5; Supplementary We noticed that Ct3 inoculation promoted plant growth under low Pi
Fig. 5), specifically when cultivated in rice-based media (Fig. 5). at 26 °C (Fig. 7a). This was accompanied by a decrease in fungal root
Importantly, the accumulation of these metabolites was abolished in colonization at 26 °C compared to 22 °C (Fig. 7b). Plant growth pro-
Ct3Δbot5, disrupted at a putative catalyzing step upstream of their motion and fungal colonization at high temperatures were unaf-
production, whereas it was unaffected in Ct3Δaba (Fig. 5). The results fected for Ct3Δaba3 or Ct3Δbot5 fungi (Fig. 7a), consistent with a
indicate that the Ct3BOT contributes to the biosynthesis of BOT (6). In great decrease in Ct3ABA3 and Ct3BOT5 expression during root
contrast, ABA (S1) or related precursor metabolites (S3, S5) were not colonization at 26 °C (Fig. 7c). Limitation of Ct3 root colonization at
detected under the conditions tested (Supplementary Fig. 6), implying high temperatures was thus associated with low expression of the
a separation between ABA and BOT gene functioning, despite their co- putative fungal ABA-BOT cluster genes. However, in phr1phl1 plants,
clustering in the fungal genome and transcriptional co-regulation Ct3 inhibited plant growth even at 26 °C in a manner dependent on
during root colonization. Alternatively, this may imply that the gene the putative ABA-BOT cluster (Fig. 7a), suggesting that the suppres-
cluster is involved in the production of a metabolite or metabolites sion of fungal pathogenesis via the putative ABA-BOT at high tem-
other than ABA. peratures requires the host AtPHR1/AtPHL1. Consistently, Ct3Δaba3
and Ct3Δbot5 fungi both promoted plant growth even in phr1 phl1
Genetic disruption of putative fungal ABA and BOT biosynthesis plants at 26 °C (Fig. 7a). This also suggests that AtPHR1/AtPHL1 are
genes leads to switching from pathogenic to beneficial lifestyles not required for PGP per se at high temperatures when the putative
in Ct3 fungal ABA-BOT cluster is disrupted. Non-inoculated WT and phr1
We next tested whether putative fungal ABA and BOT biosynthesis phl1 plants were indistinguishable in shoot growth at 26 °C even
genes are required for Ct3 pathogenesis. Ct3Δaba2, Ct3Δaba3, under low Pi, at least in our settings (Fig. 7a). These results indicate a
Ct3Δbot1, Ct3Δbot3, and Ct3Δbot5 mutant fungi showed WT-like critical role for the putative ABA–BOT cluster and its negative reg-
growth and spore formation on nutrient-rich media and WT-like ulation by host AtPHR1/AtPHL1, in the pathogen-mutualist transition
hyphal growth on glass slides (Supplementary Fig. 7a–c). Ct3Δaba2, of Ct3.
Ct3Δaba3, and Ct3Δbot5 fungi showed WT-like growth on normal or
low Pi media under our conditions (Fig. 6a), suggesting that these Host nitrogen and iron uptake genes are targeted by the puta-
genes are dispensable for culture growth in Ct3. tive fungal ABA-BOT cluster
We next examined a possible role for these fungal genes in the CtABA1, CtABA2, CtBOT1, CtBOT3, and CtBOT4, all annotated to encode
induction of host ABA responses after root inoculation. Compared to P450, constitute a large monophyletic group (Fig. 8a and Supple-
WT fungi, ABA-dependent AtMAPKKK18 induction was reduced in the mentary Fig. 8a). The simultaneous expression of these genes in con-
roots inoculated with Ct3Δaba2 and Ct3Δaba3 fungi, suggesting that junction with other Ct3 genes, suggests their interdependent
the fungal ABA biosynthesis genes are required for the host ABA regulation. Initially, we examined whether the expression of the 92 Ct3

OH OH OH
BcBOT2 BcBOT4 BcBOT5
OPP
H H H
HO AcO
FPP
7 8 4 -acetoxyprobotryan-9 -ol
(9 )
OH
BcBOT3 CHO CHO
OH OH
BcBOT1 BcBOT7
H H
AcO AcO
4 -acetoxyprobotryane-9 ,15 -diol botrydial ( 6 )
( 10 )
(m/z = 164) (m/z = 180)
9 10
(A) Standard
(B) Ct3 wild type

(Mathur s medium)
(C) Ct3 wild type

(MPY medium)
9 10
(D) Ct3 wild type
(rice medium)
9 10
(E) Ct3Δaba2
(rice medium)
(F) Ct3Δaba3 9 10
(rice medium)
(G) Ct3Δbot5
(rice medium)
10
(H) Ct4 wild type 9
(rice medium)
12.0 15.0 18.0 (min) 12.0 15.0 18.0 (min)
Fig. 5 | Ct3 produces intermediate metabolites of BOT biosynthesis. GC–MS profiles for 9 and 10 were obtained in the indicated Ct3 genotypes under the indicated
medium conditions.
genes, which were co-expressed with putative ABA and BOT genes manner dependent on the putative Ct3 ABA-BOT cluster. We examined
during pathogenesis, was altered in the Ct3Δaba2, Ct3Δaba3, and root transcriptome at 10 dpi with WT, Ct3Δaba2, Ct3Δaba3, and
Ct3Δbot5 mutant strains in comparison to the wild type (WT) (Sup- Ct3Δbot5 of Ct3, and with WT Ct4 under low Pi, where the
plementary Data 8). Our findings indicate that the expression levels of Ct3ABA/Ct3BOT expression status greatly influences Ct3 lifestyles
these Ct3 genes remained largely unaffected in the mutant strains (Supplementary Data 10).
(Supplementary Data 9), implying a separate regulation of putative Plant responses were similar among Ct3Δaba, Ct3Δbot, and Ct4
ABA and BOT genes from the other genes. (Supplementary Fig. 8b), consistent with their PGP effects under low
The requirements of putative ABA and BOT genes for activating Pi (Fig. 6e). Plant DEGs following inoculation with WT Ct3, compared
the plant core ABA pathway during Ct3 infection implied the existence with Ct3Δaba2, Ct3Δaba3, and Ct3Δbot5 mutants and WT Ct4,
of a common host target(s) for the two fungal biosynthesis pathways. included 288 Ct3-induced genes and 375 Ct3-repressed genes
To gain insight into the host pathways affected by the putative fungal (Fig. 8b–c and Supplementary Data 11 and 12). Ct3-induced genes
ABA-BOT cluster, we next assembled A. thaliana genes specifically were overrepresented with the genes responsive to oxidative stress,
induced or repressed by pathogenic Ct3, but not beneficial Ct4, in a chitin, ABA, and cellular response to phosphate starvation. DNA

Fig. 6 | Disruption of the putative fungal ABA and BOT biosynthesis genes p < 0.05, two-tailed t-test) in (b, c). d, e Shoot fresh weight at 24 dpi with the
results in a pathogen-to-mutualist lifestyle transition in Ct3 under low Pi. indicated fungi under normal Pi (d) and low Pi (e). The median values are described
a Fungal colony diameter under normal Pi and low Pi conditions. Bars represent within each boxplot. Numerals indicate theΔaba3 /Ct3 ratio, and asterisks indicate a
±SD (n = 25 biologically independent samples). b AtMAPKKK18 expression at 10 dpi significant difference compared to Col-0 (p < 0.05, two-tailed t-test) in (d) (n = 17 or
with the indicated Ct genotypes under low Pi (±SD, n = 3 biologically independent 18 (Col-0_Mock, aba2_Mock, and aba2_Ct3) biologically independent samples).
samples). c Fungal biomass indicated by Colletotrichum ACTIN relative abundance Different letters indicate significantly different statistical groups in (e) (ANOVA,
in A. thaliana roots at 10 dpi under low Pi (±SD, n = 3 biologically independent Tukey-HSD test, p < 0.05 (Mock: n = 66, Ct3: n = 68, Ct3Δaba2: n = 61, Ct3Δaba3:
samples). The photos represent the roots at 16 dpi with Ct3, Δaba3, and Δbot5. Ct n = 56, Ct3Δbot5: n = 60, Ct4: n = 60 biologically independent samples)). Repre-
hyphae were stained by WGA-Lectin (Bars = 100 µm). Asterisks indicate significantly sentative plant photos are shown (Bar = 1 cm).
different means compared with Ct3 (n = 3 biologically independent samples,
motif analysis revealed DNA sequences bound by NAC transcription To test the biological significance of this suppression, we exam-
factors related to drought tolerance37 being enriched within 1000 bp ined whether the genes repressed by Ct3ABA and Ct3BOT genes and by
upstream of their transcription initiation sites (Supplementary host ABA pathways under low Pi contribute to plant growth (Fig. 8e).
Fig. 8c). Notably, the RNA-seq analyses also indicated that PSR gene Ct3-repressible genes included AtNRT1.1 and AtNRT2.1, two major
activation at 10 dpi with Ct3 was dependent on the putative fungal transporters for nitrate uptake38, and AtFIT, FER-like iron deficiency-
ABA-BOT cluster (Fig. 8d). We further showed that expression of induced transcription factor inducing Fe uptake genes39. Ct3 repres-
AtAT4, an AtPHR1-regulon during PSR, was significantly lowered in sion of AtNRT1.1 and AtNRT2.1 expression was alleviated in aba2 and
abi1-1C and aba2 plants when colonized with Ct3 (Supplementary abi1-1C plants, respectively, pointing to its dependence on ABA (Sup-
Fig. 8d), indicating a role for the host ABA pathway in Ct3 induction plementary Fig. 8f, g). Disruption of AtNRT1.1, AtNRT2.1, or AtFIT
of host PSR genes. Interestingly, a large subset of GOs over- resulted in strong growth deficits or chlorosis under normal Pi in the
represented in the 375 Ct3-repressed plant genes were related to absence of Ct3, whereas their disruption did not show significant
hosting nutrition, such as inorganic anion transport, cation home- effects under low Pi (Fig. 8f and Supplementary Fig. 8h, i). Although
ostasis, and inorganic ion homeostasis (Supplementary Data 12 and this has hampered assessing Ct3 infection phenotypes in these mutant
Fig. 8c, e). Consistently, Ct3 root colonization changed host shoot plants, the results suggest that these nutrition-related genes are rate-
nutrition status depending on putative ABA-BOT (Supplementary limiting in plant growth under Pi sufficiency and that they define a host
Fig. 8e). The results suggest that Ct3 root infection suppresses host target for the fungal ABA-BOT cluster during Ct3-mediated plant
nutrient uptake and impacts host mineral homeostasis. growth inhibition.

Fig. 7 | Dynamic pathogenic-to-mutualistic lifestyle transitioning in Ct3 test). b Fungal biomass in roots at 24 dpi with the indicated fungi at 22 °C and 26 °C,
dependent on temperature and AtPHR1/AtPHL1. a Shoot fresh weight at 24 dpi indicated by RT-qPCR analysis (Ct ACTIN/Plant ACTIN) (±SD, Ct3: n = 4, Ct3Δaba3:
with the indicated fungi at 26 °C under low Pi. The median values are described n = 4, Ct3Δbot5: n = 4, Ct3_22: n = 3 biologically independent samples, p = 0.034187,
within each boxplot. Asterisks indicate significantly different mean (Mock_Col-0: two-tailed t-test). c, d Ct3ABA3 or Ct3BOT5 mRNA levels relative to CtACTIN in roots
n = 32, Ct3_Col-0: n = 28, Ct3Δaba3_Col-0: n = 30, Ct3Δbot5_Col-0: n = 32, Mock_phr1 at 22 °C and 26 °C (±SD, Ct3: n = 4, Ct3Δaba3: n = 4, Ct3Δbot5: n = 4, Ct3_22: n = 3
phl1: n = 20, Ct3_phr1 phl1: n = 19, Ct3Δaba3_phr1 phl1: n = 22, Ct3Δbot5_phr1 phl1: biologically independent samples, p = 0.000106 (c), p = 0.002197 (d), two tailed
n = 18 biologically independent samples, Mock vs. Ct3 (Col-0): p < 0.05, two tailed t- t-test).
Discussion How does the Ct3 putative ABA-BOT cluster inhibit plant growth?
In this study, we obtain evidence in a Ct strain (Ct3) that a dynamic Under our conditions, Ct3 and Ct4 are largely indistinguishable in root
transition from pathogenic to mutualistic lifestyles during the root colonization levels (Fig. 1e–h, Fig. 6c), making it unlikely that Ct3
colonization is achieved by altering the expression of a single fungal pathogenesis is caused by fungal overgrowth in the roots. It has been
secondary metabolism cluster (putative ABA-BOT) (Fig. 9). It has been described that bacterial and fungal pathogens mobilize the host ABA
well documented that the presence or absence of one genetic com- pathway to suppress plant immunity, in particular, salicylic acid (SA)-
ponent(s) in the genome, e.g., an island, plasmid, or (mini) chromo- based defenses28. However, our transcriptomic analyses on Ct3 wild-
some, is associated with the distinction between pathogenic and type and Ct3Δaba/Ct3Δbot fungi did not detect ABA/BOT-dependent
nonpathogenic (or potentially beneficial) lifestyles in different bacteria alterations in plant induction of defense-related genes. Consistently,
and fungi40–42. In contrast, our studies reveal that the beneficial and beneficial Ct61 promotes plant growth in the plants simultaneously
pathogenic Ct strains share a nearly identical putative ABA-BOT cluster disrupted with defense-related hormones, SA, jasmonate, ethylene,
in their genomes and that its expression status plays a critical role in and defense regulator AtPAD4, as well as in the WT plants, without
the diversification of fungal lifestyles. fungal overgrowth3. These data suggest that Ct3-mediated virulence is
Our studies further indicate that altering Ct3 putative ABA-BOT not expressed through ABA-SA antagonism. Rather, our transcriptome
cluster expression enables its lifestyle transition on the same host. The and genetic data suggest that Ct3 utilizes the putative ABA-BOT cluster
divergence between adapted and non-adapted (nonpathogenic) fungal to suppress the host genes required for the acquisition of different key
pathogens was often attributed to transcriptional induction of nutrients, such as nitrogen, iron, and zinc, which are rate-limiting in
virulence-related genes in the former43,44. Our findings extend this view plant growth, particularly when phosphate is sufficient. These results
that the loss of the putative ABA-BOT cluster not only results in the suggest that Ct3 virulence via putative ABA-BOT occurs at least in part
suppression of Ct3 pathogenesis but the expression of PGP function through perturbation of host nutrition. Indeed, Ct3 putative ABA-BOT
under low Pi, reminiscent of other beneficial Ct strains. This is also the contributes to sucrose accumulation in the host roots (Supplementary
case at high temperatures, where not only Ct3 virulence suppression Fig. 7f), which is likely effectively exploited by the fungus for increasing
but PGP is also achieved, likely through the host AtPHR1/AtPHL1- hyphal growth.
dependent suppression of putative ABA-BOT cluster expression. The Host plants require AtPHR1/AtPHL1 for suppression of fungal
results highlight an important role played by a secondary metabolism overgrowth in beneficial interactions with Ct61, a prerequisite for PGP3.
gene cluster in the plastic lifestyle transition of plant-infecting fungi This work further shows the critical role of AtPHR1/AtPHL1 in restrict-
and the convergence of environmental and host modulations on its ing Ct3 pathogenesis. Consistently, PSR-related AtPHR1/AtPHL1 reg-
transcriptional regulation. Notably, the investigated Ct strains, albeit ulons are induced during both pathogenic and beneficial interactions,
classified into the Ct species through the widely accepted method for albeit at different timings. In Ct3, putative ABA-BOT genes serve to
Colletotrichum species, harbor substantial genomic sequence varia- accelerate the early induction of host PSR-related genes (at 10 dpi)
tions. Although there are no sequence differences in the putative ABA- even under Pi-sufficient conditions, likely through the host ABA core
BOT region, there are subtle variations outside the putative ABA-BOT pathways. This may reflect a positive role for ABA in PSR under low
region between beneficial and pathogenic Ct strains, which might Pi45,46. In contrast, beneficial Ct strains induce host PSR-related genes,
contribute to the divergence in the expression profiles of putative ABA- specifically under low Pi, at a later phase (24 dpi), coincident with the
BOT genes and in fungal infection modes. Further studies are war- appearance of PGP effects. Given the absence of CtABA/CtBOT gene
ranted to examine this hypothesis and the underlying mechanisms. expression in beneficial Ct strains, their PSR activation seems

a b c
Bc A
Color Key
Up-regulated A. thaliana genes −2 −1 0

Row Z−Score
1 2
TZ5
Ct AT5G52830
3
AT5G53110
AT1G79040
during Ct3 colonization

AT2G28160
AT2G47560
AT3G16530
AT1G49320
AT3G21260
AT2G39518
G
AT5G07080
AT1G70890
AT5G22920
AT2G28780
AT5G56870
8
AT5G02090
AT4G25170
EN
AT2G39200
AT2G44370
AT3G18560
AT5G59090
AT1G60140
669
AT1G61740
AT1G23870
AT1G21680
AT4G23700
AT3G23190
AT4G39780
AT5G06800
AT1G12780
E
AT5G50760
AT5G62720
AT1G70880
AT4G08290
Δa
AT4G31875
AT3G48360
AT5G44585
AT4G12545
AT2G13360
00
AT1G80440
AT4G37390
AT4G15990
Δaba2
AT4G07820
AT4G12550
288 A. thaliana genes

.1
AT4G30140
AT1G80050
ba
AT3G07310
AT1G02610
AT3G52480
00
AT2G15880
AT3G29780
AT1G80650
AT2G27660
AT2G42060
AT1G72200
AT5G65925
3
AT2G28960
AT1G22500
AT3G20380
53
AT3G59940
AT1G71030
AT5G65690
AT5G51390
AT5G59080
AT1G63580
AT5G47240
AT2G47400
AT1G75030
AT5G22555
AT2G37180
69 Enriched GO Term
AT5G24230
AT1G52050
AT4G00910
AT5G06570
AT1G13080
AT1G52060
AT3G28345
AT5G47560
AT2G39380
Δb
AT3G26290
AT3G12977
AT2G02710
AT3G51910
AT5G24490
AT5G47450
AT2G44130
AT3G61390
AT3G14060
AT5G14120
AT3G51350
AT5G28770
AT5G49450
AT1G76590
AT1G02640
ot
AT1G78290
AT1G49860
AT2G22122
AT3G16560
AT4G20820
AT2G37820
AT4G15740
AT4G25835
AT3G49790
5
AT3G55150
AT4G15340
AT5G24410
AT4G04640
AT3G14415
・Response to oxidative stress

AT1G31885
AT1G05700
AT2G40230
AT1G51790
AT4G27700
AT4G27710
AT1G07700
CYP7A1
AT4G24180
AT2G05160
AT1G73885
AT1G22430
AT4G09770
AT5G35940
AT1G77145
AT4G15330
AT2G17820
AT1G23205
AT2G01430
AT2G26820
AT3G03500
AT4G06534
AT4G13620
AT2G31083
AT4G28270
AT4G12330
・Cellular response to
AT3G23800
BOT1
AT5G52790
AT4G01430
AT4G30450
AT1G11080
AT3G01260
Hs N
AT5G07680
AT1G51860
AT1G60740
AT1G78090
AT5G65980
AT5G10210
AT1G67110
P 000
AT1G19900
AT2G13810
AT2G29060
AT4G25080
AT4G27400
AT4G18940
AT4G40070
AT1G70860
AT4G15400
771.2
AT3G01175
AT1G51270
AT1G07175
Bc
phosphate starvation (PSR)
AT3G18773
AT1G08090
AT3G19430
AT3G21352
AT5G59680
AT1G62280
AT4G39675
AT1G11530
AT1G65486
AT3G52840
3
AT1G63180
AT5G63160
AT3G60270
AT3G03670
AT1G53635
58
AT5G45080
AT2G34000
Ct
AT2G28270
AT1G63560
AT2G36690
AT4G29740
AT2G16060
AT4G33070
AT5G43330
AT2G30550
・Response to chitin
AT5G63850
AT4G33030
AT5G51440
AT3G50930
AT3G13080
AT5G22500
AT4G37370
AT2G32020
AT4G04610
AT1G54575
AT5G36130
AT1G18870
AT5G24660
AT3G28510
AT5G39050
AT1G75280
AT4G00670
AT5G21105
AT3G53620
AT3G03270
AT1G05060
AT4G17260
AT1G12200
AT1G22065
AT2G46630
・Response to abscisic acid

AT2G18150
AT3G02040
AT5G20150
AT5G01220
AT3G47420
AT5G10180
AT3G17790
AT1G73220
AT4G33550
AT2G38940
AT2G21640
AT5G09570
AT5G13210
AT3G08860
BOT3
AT2G38823
AT5G24640
AT2G04050
AT4G12735
AT5G39120
AT5G39110
AT2G48090
AT5G53830
100
AT4G17670
AT2G47010
AT1G72330
AT5G49630
Bc
AT1G68740
AT5G57560
AT3G26830
AT5G18470
AT1G46480
AT5G53880
AT4G39740
AT1G08310
AT4G17030
AT1G43910
AT1G04770
AT3G43110
AT3G28580
AT1G65481
AT4G13395
AT5G18560
AT5G09800
Ct3
AT2G37870
AT3G03530
AT5G06720
AT3G49580
AT3G49570
AT2G16890
AT2G29460
AT1G73880
AT3G60540
AT4G14020
AT3G16330
AT5G36140
AT3G45970
10
AT3G18080
AT2G44260
AT2G27550
AT5G01790
AT2G39510
AT4G25810
AT5G06730
AT1G60470
AT2G35480
0
AT1G32870
AT3G13433
AT5G44400
100
AT2G47520
AT4G00165
AT3G54530
AT5G52940
AT3G58150
84
AT5G39150
AT5G39180
AT1G08165
AT1G22150
AT5G17340
AT5G04120
Down-regulated A. thaliana genes

AT5G20790
AT5G15960
AT1G73010
AT3G09922
AT4G33560
AT1G68570
AT4G05100
AT1G69870
AT1G18100
AT2G41180
AT3G49690
AT3G56770
AT4G37290
AT3G52340
AT2G36650
Ct3: Colletotrichum tofieldiae Ct3

AT1G33260
AT1G63820
AT3G27150
56 AT4G34950
AT2G02010
AT3G55090
AT2G31420
during Ct3 colonization

AT2G39710
AT4G39210
AT4G16680
375 A. thaliana genes

AT1G01480
AT1G29860
AT2G28830
AT5G39130
AT3G45960
AT3G62730
Bc: Botrytis cinerea B05.10

AT4G38420
AT1G19200
AT2G42690
AT3G43190
AT1G01380
AT3G05630
AT5G24655
AT1G70440
ATMG00640
AT1G62420
AT1G52410
Δa
AT4G35190
AT5G50260
AT5G19880
AT4G37800
Δaba2
AT1G55525
AT5G47220
Hs: Homo sapiens

AT5G44578
AT5G39100
AT3G01600
AT5G37490
99
AT5G06510
ba
AT5G38700
Enriched GO Term
AT2G38250
AT1G43800
ATMG01130
AT1G50400
AT3G56275
AT1G75290
AT1G35140
AT3G22840
AT3G15650
AT5G40000
3
AT3G27250
AT3G05400
AT5G64510
AT5G45970
AT5G22530
AT1G17830
AT1G28370
AT1G36180
AT5G59220
AT1G17380
AT3G56620
AT3G05858
AT5G50800
AT5G01900
10
AT5G05390
AT2G43720
AT2G25625
AT4G19970
AT1G12320
AT1G22990
AT3G56040
AT3G48610
AT1G58170
0
AT3G23250
AT4G39235
AT3G56710
Δb
AT1G72070
AT4G26270
AT4G31800
100
AT3G50220
AT4G14680
・Inorganic anion transport

AT3G24310
AT1G77120
AT4G26200
AT1G62710
AT3G21560
0
AT1G69880
AT5G52310
AT5G22860
10
AT1G04330
AT1G67600
ot
AT3G29670
AT3G06020
AT5G38000
AT5G42380
AT4G01380
AT5G17450
AT3G21720
AT1G32350
AT2G04070
5
AT1G65500
AT5G39190
Ct3
AT2G39980
AT1G17020
AT5G52640
AT5G19260
AT3G08970
AT1G05100
AT1G61930
・Cation homeostasis
AT2G35700
AT5G44580
AT3G46110
AT2G34850
AT3G04220
AT3G26840
AT4G23496
Bc
AT3G58270
AT5G59760
AT3G25655
AT1G07400
AT5G03210
AT1G32960
AT5G11930
Bc
AT2G41231
AT4G01060
AT3G17110
AT5G66815
AT5G45630
AT4G33467
AT4G12290
AT5G57123
AT1G54160
AT4G10120
AT1G74650
BOT4
AT2G47950
AT1G20990
・Inorganic ion homeostasis

AT1G55230
AT4G23000
AT5G53820
AT1G52560
AT1G50750
AT5G23060
ABA1
AT3G51220
AT1G09950
AT5G28520
AT2G32860
AT3G23290
AT5G20410
AT4G36850
AT2G30540
AT5G24770
AT1G16110
AT4G15417
AT3G50820
Ct3
AT1G15890
AT4G40065
AT5G25140
AT1G08980
AT2G16380
Bc
AT4G37060
AT3G23510
Ct
AT5G14070
AT2G31081
・Secondary metabolic process

AT2G26975
AT1G06120
AT5G38100
AT1G73270
AT2G26410
AT4G09420
AT1G61480
AT1G55580
AT5G42010
AT4G16950
AT1G16150
AT2G20180
3
AT5G45840
AT5G10140
AT4G22640
AT1G53950
0.5
AT4G33790
AT2G44000
AT2G13910
AT2G38460
AT5G57180
AT2G17060
AT3G12240
AT2G21610
AT1G05650
AT1G60160
ABA2
AT3G61300
AT3G04900
AT2G05180
AT2G43600
AT4G31710
AT2G21650
AT3G14810
AT3G47050
AT4G33020
AT1G64210
AT2G17050
AT1G49200
AT2G46660
AT5G22560
AT1G44050
AT4G34930
AT4G12490
AT3G56980
AT3G01080
AT4G14780
AT3G14185
AT3G02850
AT2G14510
AT5G46650
AT5G16410
AT5G51850
AT3G02430
AT4G16870
AT3G46400
AT3G45440
AT3G15840
AT5G07650
AT2G25680
AT5G45090
AT3G45390
AT3G43890
AT1G55410
AT2G15350
AT3G13840
AT1G44020
AT4G35440
AT2G19500
AT1G06160
AT1G13710
AT1G50050
AT4G00780
AT5G54585
AT2G29320
AT4G04630
AT5G49780
AT5G54300
AT1G58340
AT3G51340
AT3G14260
AT2G25090
AT3G63110
AT4G25050
AT3G47040
AT4G38390
AT4G26555
AT3G26220
AT1G10340
AT5G49770
AT1G07560
AT5G58360
AT2G38600
AT3G04110
AT2G24720
AT3G60700
AT4G12500
AT2G44340
AT2G28990
AT5G19970
AT2G21020
AT2G32660
AT1G30260
AT5G05400
AT1G21360
AT4G39770
AT3G09240
AT1G65570
AT1G45015
AT3G18450
AT2G19060
AT3G20360
AT4G16000
AT5G24920
AT4G08300
AT4G33720
AT5G56100
AT5G55250
AT3G46190
d
AT5G16980
AT4G32950
AT1G80830
e
AT5G43180
AT1G30720
AT5G43520
AT4G22212
AT3G50740
AT2G32270
AT5G23980
AT5G60530
AT4G24015
AT1G62510
AT2G34420
AT5G54370
AT4G34138
AT1G20840
AT2G46750
AT2G25900
AT1G30820
AT5G61590
AT2G15890
AT5G43780
AT1G78000
AT5G23990
AT1G21100
AT1G52070
AT1G52200
AT5G48430
AT5G04950
AT4G38470
AT1G80180
AT3G51330
AT3G06390
AT3G10020
AT4G05070
AT5G20250
AT3G26740
AT3G09220
AT4G17340
AT2G22330
AT1G49310
AT2G44380
AT4G29905
AT2G33790
AT2G36460
AT1G05680
AT5G19550
AT4G14060
PSR genes
AT2G03760
AT5G54100
AT4G19810
AT5G62480
AT3G22370
AT3G08590
AT4G15150
AT2G41730
AT3G52060
AT5G43350
AT5G65207
AT2G21045
AT5G63600
AT5G50200
AT1G15380
AT5G44020
AT3G61430
AT5G43580
AT3G25190
AT1G17180
AT2G43535
AT2G05510
AT2G36830
AT1G32450
AT3G53420
AT3G15450
AT1G26250
AT1G65845
AT1G07610
AT3G54040
AT1G66200
AT1G28330
AT2G33830
2
Δaock
Δbba3
C5
M t3
t4
Δa ba
ot
C
NRT1.1
Nitrogen
SULTR1;2
Iron
f
FRO5 Zinc a c b b c c c c
FRO4 Phosphate 0.04
Sulfate
ZIP3
Glutamate
NITR2;1 Shoot Fresh Weight (g)
0.03
FIT
PHT1;1
NRT1.5
0.02
VTL5
ZF14
GLR1.1
0.01
NRT2.1
ZIP9
Δa t3
Δa ba2
Mba3
Δbock
C5
t4
0.00
ot
C
nr ol-0
nr nrt1 _1
nr 2
C 2
t1 0
nr nrt 1_1
nr 2
.2
.1 _
.
nr ol-
.1 1_
t2
t2
.1
t2 .1
C
.
t2 1.
t1
+ Pi - Pi
Δa Ct3
M a2
ba k
3
ΔbCt4
5
Δa oc
ot
b
Fig. 8 | Ct3ABA-BOT gene cluster suppresses nitrogen uptake genes while enlisted, respectively. Red and blue lines next to the gene IDs represent their
activating PSR genes in the host during Ct3 colonization. a Maximum Like- upregulation or downregulation, respectively. d Hierarchical clustering of A.
lihood tree of ABA and BOT genes in Cytochrome P450 family using IQ-TREE ver- thaliana PSR-related 193 genes. Overrepresented (red to yellow) and under-
sion 1.6.11. Ct3: C. tofieldiae Ct3; Bc: B. cinerea B05.10; Hs: Homo sapiens. Homologs represented (yellow to blue) modules are depicted as log10 (fpkm + 1).
to CYP7A1 were used as an outgroup. The phylogenetic relationship of the ABA and e Hierarchical clustering of A. thaliana genes related to nutrient uptake and sup-
BOT genes and other 409 P450 genes in C. tofieldiae, B. cinerea, and H. sapiens is pressed by Ct3 through the ABA-BOT cluster (FDR < 0.05). Overrepresented (red to
indicated in Supplementary Fig. 8a. Ultrabootstrap probability is shown on the yellow) and underrepresented (yellow to blue) modules are depicted as log10
branches. The scale bar represents substitutions per site. b Venn diagram repre- (fpkm + 1). f Shoot fresh weight of the indicated Arabidopsis genotypes under
senting significantly up (Upper)- or down (Down)-regulated Arabidopsis genes at 10 normal (+) or low (−) Pi at 24 days (Col-0: n = 18, nrt1.1_1: n = 15, nrt1.1_2: n = 17, nrt2.1
dpi with Ct3 compared with the other Ct genotypes (Ct4, Ct3Δaba2, Ct3Δaba3, and nrt2.2: n = 18, Col-0_low Pi: n = 18, nrt1.1_1_low Pi: n = 17, nrt1.1_2_low Pi: n = 16, nrt2.1
Ct3Δbot5) ( | log2FC | >1, FDR < 0.05) under low Pi. c Transcript profiling of 288 nrt2.2_low Pi: n = 15 biologically independent samples). The median values are
commonly up-regulated and 375 commonly down-regulated Arabidopsis genes in described within each boxplot. The fresh weight of each plant was measured after
the roots following inoculation with all the examined Ct genotypes. Over- 24 days of incubation. Different letters indicate significantly different statistical
represented (red to yellow) and underrepresented (yellow to blue) modules are groups (ANOVA, Tukey-HSD test, p < 0.05).
depicted as log10 (fpkm + 1). The major enriched GOs of 288 or 375 genes IDs are

Beneficial Pathogenic
22oC 22oC 26oC AtPHR1/AtPHL1
Putative ABA-BOT cluster Putative ABA-BOT cluster Putative ABA-BOT cluster
silenced activated silenced
Ct4
Ct61 Ct3 Ct3
ABA / BOT
botrydial
AtPYL
AtABA
AtABI1
ABA core pathway
Growth promotion Growth promotion

AtNRTs
Sugar PSR
AtFit
Growth inhibition
Fig. 9 | A model depicting the pathogenic-mutualistic lifestyle transition in Ct3. host ABA core pathway, thereby leading to the suppression of nutrient uptake-
Unlike beneficial Ct strains (Ct4 and Ct61), Ct3 induces putative ABA-BOT gene related genes and the activation of PSR-related genes in A. thaliana. Perturbation of
expression during an initial phase of root colonization. This putatively leads to the host nutrition pathways is then associated with sucrose accumulation in the
fungal production of ABA, its precursors, and/or related metabolites from the ABA roots, which is likely to facilitate Ct3 hyphal growth. At 26 °C, Ct3 colonizes the
biosynthetic genes as summarized previously16 and botrydial production from the roots with the putative ABA-BOT genes silenced through the host AtPHR1/AtPHL1
BOT biosynthetic genes as shown in Fig. 5. These metabolites, in turn, involve the and promotes plant growth under low Pi.
independent of putative fungal ABA or BOT biosynthesis genes. Consistent with our molecular phylogenetic analysis that the
Indeed, like Ct3, beneficial Ct strains efficiently colonize A. thaliana beneficial strains are derived from the pathogenic ancestors (e.g., C.
roots under low Pi without expressing CtABA/CtBOT genes (Figs. 1e, f incanum or C. liriopes), the present evidence indicates that the acqui-
and 6c). Therefore, separate fungal mechanisms, in terms of ABA sition and expression of putative ABA-BOT gene cluster (or putative
dependence, are likely to confer host PSR activation between patho- ABA and BOT gene clusters) contribute to fungal pathogenesis, likely
genic Ct3 and beneficial Ct interactions. Conversely, AtPHR1/AtPHL1 through the activation of host ABA biosynthesis and signaling during
restricts fungal growth under low Pi through the suppression of infection. It seems that putative ABA and BOT gene clusters also con-
separate fungal infection strategies between Ct3 and beneficial Ct tribute to fungal pathogenesis in Botrytis, as shown for Ct since host
strains, diverging in putative ABA-BOT dependence. Our evidence ABA signaling is required for the infection and pathogenesis of the
attributes AtPHR1/AtPHL1-mediated alleviation of Ct3 pathogenesis to necrotrophic fungus18,19. Although definite proof remains to be
the suppression of fungal putative ABA-BOT expression, under low Pi obtained, putative ABA-BOT gene clusters are likely to confer virulence
and at high temperatures. Plant transcriptome data not detecting in phylogenetically diverse fungi beyond Ct species. The additional
differential defense gene regulation imply that AtPHR1/AtPHL1-medi- requirements for host ABA biosynthesis in fungal pathogenesis imply
ated control of fungal growth is also distinct from their suppression of that these fungal genes alone are not sufficient for effective ABA bio-
pattern-triggered immunity10,11. synthesis and virulence. Future studies will be required to elucidate the
Interestingly, the loss of AtPHR1/AtPHL1 allows Ct3Δaba3 and precise mechanisms by which fungal and plant ABA pathways work in
Ct3Δbot5 fungi to promote plant growth in phr1 phl1 at elevated concert to promote fungal virulence.
temperatures, which is not seen in the WT plants (Fig. 7a). This indi-
cates that AtPHR1/AtPHL1 are not required for the PGP per se conferred Methods
by Ct3 at high temperatures when the putative ABA-BOT cluster is Plant material and growth conditions
disrupted. In addition to stimulating host ABA signaling, putative ABA- Arabidopsis thaliana Col-0, the abi1-1C27, aba149, aba2-1250, pyr1-1 pyl1
BOT may suppress the PGP function through promoting host Pi pyl2 pyl4 pyl551, phf152, phr1 phl19, nrt1.1 (chl1-553), nrt2.1nrt2.254, and fit1-
acquisition, which is conserved in Ct species, including Ct3 (suggested 239 mutants (Col-0 background) and nrt1.1_2 (Ler background) were
by Fig. 7a). This also seems to apply to pathogenic C. incanum, in which used in this study. We also used 15 A. thaliana WT accessions as
inherent putative ABA-BOT expression is associated with a massive described in Supplementary Fig. 2. Seeds were surface sterilized with
decrease in phosphorus transfer to the host compared to beneficial 70% ethanol for the 30 s, followed by 6% sodium hypochlorite with
Ct613. Notably, however, in Ct3, this PGP function is likely suppressed 0.01% Triton X. After being washed three times in sterilized water,
by AtPHR1/AtPHL1, suggested by the absence of PGP in WT plants when seeds were placed in cold treatment at 4 °C for 24 h before sowing.
inoculated with Ct3Δaba3/Ct3Δbot5 fungi (Fig. 7a). It is conceivable
that ABA-dependent activation of AtPHR1/AtPHL1 regulons activated at Fungal inoculation assay
10 dpi with Ct3 contributes to the suppression of PGP, given the pre- Sterilized A. thaliana seeds were sowed on half-strength MS agarose
viously described negative role for AtPHR1 in nitrate transporter medium containing defined Pi concentrations. Ct spores (3 μL of
expression, including AtNRT1.1, required for plant growth47,48. The 1 × 104/ml) were inoculated 3 cm below the sowed seeds. Plates were
mechanisms underlying the multifaceted functions of AtPHR1/AtPHL1 placed vertically in a plant growth chamber under a 10:12 h light: dark
in the complex host–fungus interactions require further studies. cycle at ~22 °C ( ± 1 °C) (80 μmol/m2s) around 50% humidity unless

otherwise described. For 26 °C assays, plants were incubated in a plant flanking the target gene. Around 1.5-kb fragments of 5′ (5 F) and 3′
growth chamber under a 10:12 h light: dark cycle at ~26 °C (80 μmol/ sequences (3 F) flanking the gene ORF were amplified from Ct3
m2s) during light and ~22 °C during dark. Effects on plant growth by genomic DNA by PrimeSTAR HS DNA polymerase (TaKaRa). The pur-
fungal inoculation were evaluated by measuring shoot fresh weight, ified PCR products 5 F and 3 F were mixed with SalI-treated
root length, or chlorophyll a and b from ~15 plants per experiment. Ct pBIG4MRHrev for In-Fusion reactions. This generated plasmid was
hyphae were mixed with soils for soil inoculation, and B. rapa seeds then introduced into Agrobacterium C58C1 strains by electroporation
were incubated for 24 days on the soils before measuring shoot fresh using the default setting for the Eppendorf Eporator. The transformed
weight (SFW). Agrobacterium strains (OD600 = 0.4) were mixed with Ct3 spores
(1 × 107) in a 1:1 ratio on a paper filter attached to incubation media
Plant growth conditions containing 200 μM acetosyringone (BLD Pharm). Selection of hygro-
Half-strength MS medium used in this study is based on the previous mycin (150 µM)-resistant strains was conducted by transferring the
work55, with minor modifications. Agar granulated (Difco) was initi- paper filter to PDA media with Hyg, cefotaxime, and spectinomycin (all
ally used as an agar containing limited phosphate in this study. at 50 μg/mL) and incubated for two days before the paper was elimi-
However, as the new batches of the agars caused unstable growth in nated from the media. The resistant strains from the media were tested
mock-treated plants under low Pi, INA Agar BA-10 (INA food) with by PCR using primers targeting the outside sequence and a Hyg
further limited phosphate was used as an alternative. pH was adjus- sequence to determine whether the Hyg resistance cassette success-
ted to 5.1. 55 ml of the mixture was then poured into square Petri fully replaced the interesting genes. The PCR primers used are listed in
plates (15 × 15 cm, Greiner). For Ct inoculation in soils, Ct spores Supplementary Table 5. The Ct3 and Ct4 lines constitutively expres-
(50 mL of 1 × 104/ml) or distilled water (mock treatment) were incu- sing GFP were generated as previously described3.
bated on mixtures of autoclaved 180 g sawdust, 60 g rice bran, 60 g
bran, and 180 ml water for 2 weeks, where mixtures with or without Fluorescence microscopy
Ct hyphae were mixed with soils containing Kanuma soil, Exo sand, Inoculated A. thaliana roots were observed through confocal micro-
and black soil (15:42.5:42.5) with 3-5% weight. Before measuring SFW, scopy. To trace the colonization process of Ct3-GFP or Ct4-GFP in the
B. rapa var. perviridis seeds (ʻMisakiʼ, Sakata seeds) were then sowed roots, we utilized a confocal laser scanning microscope (Nikon C2 plus)
in the soil and incubated for 24 days. equipped with filters optimized for visualization of GFP or mCherry to
distinguish fungal hyphae and plant plasma membranes. Fungal lectin
Fungal growth assay in vitro staining was performed using a confocal laser scanning microscope
Agar plugs of ten-day-old fungal cultures grown on Mathur’s media Olympus FV1000 with filter settings suitable for fluorescein (FITC), i.e.,
were transferred to new Mathur’s plates. Alternatively, 3 µL of fungal blue excitation light. WGA-FITC conjugate (Sigma) was used to specify
spore suspension, comprising approximately 1 × 104/ml, was put onto the fungal cell walls. Fungal-inoculated roots were incubated overnight
standard Pi media (625 µM KH2PO4) and low Pi media (50 µM KH2PO4) in a 1:3 mixture of chloroform and ethanol. Then, the roots were
without sucrose. Colony formation was determined three days after transferred to chloral hydrate (2 g/mL in water) and incubated for 2 h.
inoculation by measuring the colony radius from center to edge. After PBS washed roots, the roots were incubated in wheat germ
Fungal spore counts were conducted as described previously56. For agglutinin (WGA) conjugated to a FITC solution (5 µg/ml; Sigma) for 1 h
fungal growth on glass slides, fungal spores were placed on glass slides at room temperature.
and incubated for 3 days.
Genome sequencing and assembly
Leaf chlorophyll measurements DNA extraction, genome sequencing, and assembly were conducted as
The method of chlorophyll quantification was adapted from the pre- described34. De novo assembly of Ct4 genomes was conducted by
vious report57. Samples were prepared from ~100 mg of leaf tissue FALCON-integrate (v. 1.8.18). Illumina HiSeq for 100 bp paired-end
pooled from ~4 plants per sample and weighed. Chlorophyll was short reads against Ct4 genomes was also conducted by Macrogen
extracted by adding 800 µL chilled cold 100% acetone, and samples (TrueSeq DNA PCR-Free kit), resulting in 59 million filtered reads. The
were shaken for 10 min until plant tissue was transparent. After the filtered sequence reads were used for error collection by mapping
samples were diluted four times with 80% acetone, the absorbance of Hiseq reads to the PacBio assembly using Pilon (v. 1.21).
tissue-free chlorophyll extract was measured at 646 nm, 663 nm, and
750 nm with an Eppendorf Biospectrometer following the formula. Syntenies of Ct3, Ct4, and Ct61 genomes
A: Chlorophyll a (µg/ml) = 12.25 × (OD663.6 − OD750) − 2.85 × Syntenies of Ct3, Ct4, and Ct61 genomes were detected as locally
(OD646.6 − OD750) collinear blocks (LCB) partly by performing whole-genome alignment
B: Chlorophyll b (µg/ml) = 20.31 × (OD646.6 − OD750) − 4.91 × in the Mauve program v1.1.158. The location of LCBs and synteny cor-
(OD663.6 − OD750) relations were visualized as circos diagram by package circlize
v.0.4.359 in R.
Quantitative real-time PCR
cDNA was synthesized from 300 to 500 ng total RNA using the Pri- Phylogenetic tree analyses for fungal classification
meScript RT Master Mix (Takara) in a volume of 10 µL. We then The maximum-likelihood phylogenetic tree was reconstructed using 6
amplified 3 µL of cDNA (10 ng/µL) in Power SYBR (Thermo) with 1.6 µM fungal marker nucleotide sequences of 34 strains in 12 species of the
primers using the AriaMx real-time PCR system (Agilent) in a volume of genus Colletotrichum. The 6 sequences used were the nuclear 5.8 S
12 µL. Primers used in this study are listed in Supplementary Table 5. ribosomal RNA gene with the two flanking internal transcribed spacers
(ITS), a 200-bp intron of the glyceraldehyde-3-phosphate dehy-
Fungal transformation via Agrobacterium-mediated drogenase gene (GAPDH), partial sequences of the actin (ACT), chitin
transformation synthase 1 (CHS-1), beta-tubulin (TUB2), histone 3 (HIS3) genes, as used
For targeted gene replacement, we used the previously described previously to identify the species and species complexes of the genus
method3. Briefly, to generate replacement mutants lacking Ct3 puta- Colletotrichum23,25. Sequence information for all the strains was
tive ABA biosynthesis genes (ABA2 and ABA3) or BOT biosynthesis retrieved using BLAST. The sequences of C. incanum MAFF 238704 and
genes (BOT1, BOT3, and BOT5), we constructed a plasmid in which the C. liriopes MAFF 242679 in GenBank were also used. Nucleotide
Hyg resistance gene cassette was inserted between DNA sequences sequences were aligned using MAFFT60 and concatenated manually.

The maximum-likelihood phylogenetic tree was reconstructed using column (0.25 mm × 30 m, 0.25 μm film thickness; SUPELCO) in the
IQ-TREE (v.2.1.2) with the options -m MFP -T AUTO -B 100061. following conditions (Method A); 100 °C for 3 min, 100-230 °C (rate:
The maximum-likelihood phylogenetic tree was also recon- 14 °C/min), 230 °C for 5 min at a flow rate of 1.2 mL/min (helium carrier
structed using single-copy orthologs of all genomes of the genus gas). The followings are summary of other setting conditions; Column
Colletotrichum in GenBank, those obtained in this study, and that of B. Oven Temp.: 100 °C, Injection Temp.: 230 °C, Pressure: 89.4 kPa, Total
cinerea B5.10 as an outgroup. The predicted proteins of the 71 fungal Flow: 28.1 mL/min, Column Flow: 1.2 mL/min, Linear Velocity: 40.7 cm/
strains were clustered into single-copy orthologous groups using sec, Purge Flow: 3.0 mL/min, Split Ratio: 20.0, Ion Source Temp.:
Orthofinder v.2.5.562 and Mirlo (https://github.com/mthon/mirlo). 200 °C, Interface Temp.: 230 °C, Solvent Cut Time: 2.5 min, Micro Scan
Amino-acid sequences of the 1,509 single-copy orthologous groups Width: 0 u, and Mass range: 50-500. GCMSsolution software (version
were aligned using MAFFT with default settings60. The maximum- 4.45) was used for data acquisition and data analysis.
likelihood tree was reconstructed using IQ-TREE (v.2.0.3) with the
options -m MFP -B 100061. For ABA intermediates
After filtration of the crude extracts, the filtrates were directly analyzed
Construction of phylogenetic trees of ABA and BOT biosynth- by a UPLC-MS equipped with ACQUITY UPLC BEH C18 (2.1 × 50 mm) in
esis genes the following conditions (Method B); 0-1 min = 10% B, 1-3 min = 10-95%
For this analysis, we re-annotated putative ABA and BOT biosynthesis B, 3-5 min = 95% B (A: H2O + 0.1% of formic acid, B: CH3CN + 0.1% of
Ct (Ct3, Ct4, and Ct61) genes based on both genome assembly and formic acid) at a flow rate of 0.7 mL/min; column temp. 40 °C. The
RNA-seq expression data for each gene. Sequences similar to ABA and followings are a summary of other setting conditions; λ Range: 210 nm-
BOT genes of Colletotrichum strains, B. cinerea, and some other strains 400 nm, Resolution: 1.2 nm, Sampling Rate: 20 points/s, Filter Time
were collected with BLASTP search using Biopython (v.1.7.6; E-value: Constant: 0.2 s, Exposure Time: auto. MassLynx V4.1 was used for data
1e-4 or 1e-5) against newly analyzed genomes in this study, Homo acquisition and data analysis.
sapiens genome assembly GRCh38.p13, and GenBank NR database on
February/March 2020. Those containing large indels or that were Profiling fungal metabolites during solid medium incubation
highly diverged were omitted from the data set of sequences Mycelia of Ct was inoculated into a solid medium containing polished
(Supplementary Table 3). Sequences were aligned with MAFFT using rice (1 g) in a 5 mL test tube. The culture was incubated at 25 °C for
the E-INS-i strategy (v.7.273)60. Phylogenetic analysis was conducted 5 days. After extraction with Ethyl acetate, the extract was con-
with IQ-TREE (v.1.6.11; -m MFP -nt AUTO -b 100)61. Microsynteny of ABA centrated in vacuo to afford crude extracts.
and BOT genes was analyzed with GenomeMatcher (v.3.00)63. The
fungal species tree in Fig. 4m is followed by64. For botrydial intermediates
After filtration of the crude extracts, the filtrates were directly analyzed
RNA extraction and RNA-sequencing analyses by a GC-MS in Method A (see above).
Total RNA was extracted from the whole root compartments using
the NucleoSpin RNA Plant kit (Macherey-Nagel). RNA samples (1 µg For ABA intermediates
each) were then sent to BGI for quality assurance, library preparation, The crude extracts were evaporated, and the residues were then dis-
and subsequent sequencing using their default process. The gener- solved in MeOH. After filtration, the filtrates were directly analyzed by
ated libraries were sequenced by HiSeq-illumina, resulting in a UPLC-MS in Method B (see above).
approximately 40 million reads per sample, paired-end 150 bp
(Figs. 3 and 4) or BGIseq (Fig. 8), resulting in about 40 million reads Isolation of the biosynthetic intermediates
per sample, paired-end 100 bp. Tophat2 with default settings, except The ABA biosynthetic intermediates were isolated from the transfor-
that average fragment size was specified65, was used for sequence mants constructed in our previous studies17,70. The BOT biosynthetic
mapping based on the A. thaliana genome (TAIR10) or fungal gen- intermediates, 9 and 10, were isolated from the A. oryzae transfor-
omes. The generated BAM files from the Tophat2 platform were used mants possessing either BcBOT2/4/5 or BcBOT1/2/3/4/5, which were
to analyze DEGs by cuffdiff with default settings66. The following constructed by incorporating those biosynthetic genes into A. oryzae
statistical analysis and data visualization were conducted in R Bio- applying an established hot spot-knock-in method. Their 1H- and
conductor packages, such as Cummerbund, gplots, and ggplot266,67. 13
C-NMR spectra were in good agreement with the reported data35,36.
As judged by the Cummerbund platform, DEGs were identified using
an FDR threshold of < 0.05. Heatmaps representing gene expression ABA measurement in plants
profiles were generated with the R package heatmap.2 in gplots. GO Extraction, purification, and quantification of ABA were carried out as
analysis for A. thaliana genes was conducted by AgriGO v. 2 described in the previous report71 using around 5-mg dry weights of
(FDR < 0.0568). Motif analysis was conducted by CentriMo69 with the around 10-day-seedling roots per replicate (equivalent to around 260
default setting using 1000 bp upstream of the target A. thaliana (Mock or Ct4)-320 (Ct3) roots).
genes extracted via TAIR Bulk Data Retrieval.
Mineral measurements
Profiling fungal metabolites during liquid culture growth Plant samples were dried for 3 days in a 70 °C oven. Dried samples were
The mycelia of Ct were inoculated into 1 mL of both MPY and Mathur’s set as 10-15 mg in one biological sample. Each sample was digested
medium in a 5 mL test tube. The cultures were incubated at 25 °C with a with nitric acid and hydrogen peroxide. The dried pellets, after
constant rotational speed of 200 rpm for a period of 5 days. Following digestion, were dissolved in 0.08 M HNO3. Elemental concentrations in
extraction with 2 mL of acetone, the extracts were evaporated. Sub- samples were measured by inductively coupled plasma mass spec-
sequently, 1M-HCl was added to the mixture, and the resulting mixture trometry (ICP-MS) Agilent 7800 (Agilent Technologies Co., Ltd., Japan)
was extracted with ethyl acetate. The crude extracts were then dis- according to the manufacturer’s instructions. To determine the con-
solved in methanol. centration of P, ICP-MS NexION 350 S (PerkinElmer, Waltham, MA,
USA) was used. Clustering analysis was done by using mineral con-
For botrydial intermediates centrations of the samples. NbClust package (version 3.0) of R was
After filtration of the crude extracts, the filtrates were directly analyzed used for k-means clustering (Hartigan-Wong algorithm). An optimal
by a GC-MS equipped with a Beta DEXTM 120 fused silica capillary number of clusters was determined by silhouette score.

Sucrose measurement in plants 15. Marumo, S. et al. Microbial-production of abscisic-acid by Botrytis

Sucrose extraction from the freeze-dried sample and its derivatiza- cinerea. Agr. Biol. Chem. Tokyo 46, 1967–1968 (1982).
tion for GC/MS analysis were essentially carried out as previously 16. Endo A, Okamoto M & Koshiba T. ABA biosynthetic and catabolic
described71. GC-EIMS used in this study was a 7890B GC-MS system pathways. In (ed Zhang, DP) Abscisic Acid: Metabolism, Transport
(Agilent Technologies) equipped with a DB-5 MS + Dura Guard and Signaling. (Springer, Dordrecht, 2014).
(30 m × 0.25 mm i.d., film thickness of 0.5 μm and 10 m Dura Guard, 17. Takino, J. et al. Unveiling biosynthesis of the phytohormone abscisic
Agilent Technologies). The injection temperature was 250 °C, and acid in fungi: unprecedented mechanism of core scaffold formation
the helium gas flow rate through the column was 0.9 mL/min. The catalyzed by an unusual sesquiterpene synthase. J. Am. Chem. Soc.
column temperature was held at 60 °C for 1 min and then was raised 140, 12392–12395 (2018).
by 10 °C/min to 325 °C and was held there for 10 min isothermally. 18. Audenaert, K., De Meyer, G. B. & Hofte, M. M. Abscisic acid deter-
The retention time for sucrose-8TMS was 16.4 min under the mines basal susceptibility of tomato to Botrytis cinerea and sup-
condition. presses salicylic acid-dependent signaling mechanisms. Plant
Physiol. 128, 491–501 (2002).
Reporting summary 19. Liu, S. A., Kracher, B., Ziegler, J., Birkenbihl, R. P. & Somssich, I. E.
Further information on research design is available in the Nature Negative regulation of ABA signaling by WRKY33 is critical for
Portfolio Reporting Summary linked to this article. Arabidopsis immunity towards Botrytis cinerea 2100. Elife 4,
e07295 (2015).
Data availability 20. Brakhage, A. A. & Schroeckh, V. Fungal secondary metabolites—
Accession codes: Newly obtained fungal whole-genome data have strategies to activate silent gene clusters. Fungal Genet. Biol. 48,
been deposited in DDBJ/NCBI (Ct61: BQXX01, Ct3: BQXV01, Ct4: 15–22 (2011).
BQXW01, KHC23: BPPY01, C. liriopes: BPPX01). Additionally, raw 21. Hacquard, S. et al. Survival trade-offs in plant roots during coloni-
sequencing data of RNA-seq transcriptomic data have been deposited zation by closely related beneficial and pathogenic fungi. Nat.
in DDBJ (DRA012868, DRA012867). Microscopic pictures have been Commun. 7, 11362 (2016).
deposited in BioImage Archived (S-BIAD827). Source data are provided 22. Darma, R., Lutz, A., Elliott, C. E. & Idnurm, A. Identification of a gene
in this paper. cluster for the synthesis of the plant hormone abscisic acid in the
plant pathogen Leptosphaeria maculans. Fungal Genet Biol. 130,
References 62–71 (2019).
1. Drew, G. C., Stevens, E. J. & King, K. C. Microbial evolution and 23. Damm, U., Woudenberg, J. H. C., Cannon, P. F. & Crous, P. W.
transitions along the parasite-mutualist continuum. Nat. Rev. Colletotrichum species with curved conidia from herbaceous hosts.
Microbiol. 19, 623–638 (2021). Fungal Divers 39, 45–87 (2009).
2. Cheng, Y. T., Zhang, L. & He, S. Y. Plant-microbe interactions facing 24. Sato, T., Moriwaki, J. & Kaneko, S. Anthracnose fungi with curved
environmental challenge. Cell Host Microbe 26, 183–192 (2019). conidia, Colletotrichum spp. belonging to Ribosomal Groups 9-13,
3. Hiruma, K. et al. Root endophyte Colletotrichum tofieldiae confers and their host ranges in Japan. Jarq Jpn Agr. Res. Q. 49,
plant fitness benefits that are phosphate status dependent. Cell 351–362 (2015).
165, 464–474 (2016). 25. Jayawardena, R. S., Bhunjun, C. S., Hyde, K. D., Gentekaki, E. &
4. Morcillo, R. J. L. et al. Rhizobacterium-derived diacetyl modulates Itthayakorn, P. Colletotrichum: lifestyles, biology, morpho-species,
plant immunity in a phosphate-dependent manner. Embo J. 39, species complexes and accepted species. Mycosphere 12,
e102602 (2020). 519–669 (2021).
5. Fesel, P. H. & Zuccaro, A. Dissecting endophytic lifestyle along the 26. Liu, F. et al. Updating species diversity of Colletotrichum, with a
parasitism/mutualism continuum in Arabidopsis. Curr. Opin. phylogenomic overview. Stud. Mycol. 101, 1–56 (2022).
Microbiol. 32, 103–112 (2016). 27. Umezawa, T. et al. Type 2C protein phosphatases directly regulate
6. Kia, S. H. et al. Influence of phylogenetic conservatism and trait abscisic acid-activated protein kinases in Arabidopsis. Proc. Natl.
convergence on the interactions between fungal root endophytes Acad. Sci. USA 106, 17588–17593 (2009).
and plants. ISME J. 11, 777–790 (2017). 28. Zabala, M. D., Bennett, M. H., Truman, W. H. & Grant, M. R. Antag-
7. Puga, M. I. et al. Novel signals in the regulation of Pi starvation onism between salicylic and abscisic acid reflects early host-
responses in plants: facts and promises. Curr. Opin. Plant Biol. 39, pathogen conflict and moulds plant defence responses. Plant J. 59,
40–49 (2017). 375–386 (2009).
8. Rubio, V. et al. A conserved MYB transcription factor involved in 29. Yamaguchishinozaki, K. & Shinozaki, K. Characterization of the
phosphate starvation signaling both in vascular plants and in uni- expression of a desiccation-responsive Rd29 Gene of Arabidopsis
cellular algae. Gene Dev. 15, 2122–2133 (2001). thaliana and analysis of its promoter in transgenic plants. Mol. Gen.
9. Bustos, R. et al. A central regulatory system largely controls tran- Genet 236, 331–340 (1993).
scriptional activation and repression responses to phosphate star- 30. Okamoto, M. et al. Activation of dimeric ABA receptors elicits guard
vation in Arabidopsis. Plos Genet. 6, e1001102 (2010). cell closure, ABA-regulated gene expression, and drought toler-
10. Castrillo, G. et al. Root microbiota drive direct integration of phos- ance. Proc. Natl. Acad. Sci. USA 110, 12132–12137 (2013).
phate stress and immunity. Nature 543, 513–518 (2017). 31. Siewers, V., Kokkelink, L., Smedsgaard, J. & Tudzynski, P. Identifi-
11. Tang, J. et al. Plant immunity suppression via PHR1-RALF-FERONIA cation of an abscisic acid gene cluster in the grey mold Botrytis
shapes the root microbiome to alleviate phosphate starvation. cinerea. Appl Environ. Microbiol. 72, 4619–4626 (2006).
EMBO J. 41, e109102 (2022). 32. Siewers, V. et al. Functional analysis of the cytochrome P450
12. Nutzmann, H. W., Scazzocchio, C. & Osbourn, A. Metabolic gene monooxygenase gene bcbot1 of Botrytis cinerea indicates that
clusters in eukaryotes. Annu. Rev. Genet. 52, 159–183 (2018). botrydial is a strain-specific virulence factor. Mol. Plant Microbe 18,
13. Hoffmeister, D. & Keller, N. P. Natural products of filamentous fungi: 602–612 (2005).
enzymes, genes, and their regulation. Nat. Prod. Rep. 24, 33. Porquier, A. et al. The botrydial biosynthetic gene cluster of Botrytis
393–416 (2007). cinerea displays a bipartite genomic structure and is positively
14. Keller, N. P. Fungal secondary metabolism: regulation, function and regulated by the putative Zn(II)(2)Cys(6) transcription factor
drug discovery. Nat. Rev. Microbiol. 17, 167–180 (2019). BcBot6. Fungal Genet. Biol. 96, 33–46 (2016).

34. Utami, Y. & Hiruma, K. Genome resource of Colletotrichum spae- 54. Li, W. et al. Dissection of the AtNRT2.1:AtNRT2.2 inducible high-
thianum, the causal agent of leaf anthracnose in Polygonatum fal- affinity nitrate transporter gene cluster. Plant Physiol. 143,
catum. PhytoFrontiers 2, 152–155 (2022). 425–433 (2007).
35. Duran-Patron, R., Colmenares, A. J., Hernandez-Galan, R. & Collado, 55. Gruber, B. D., Giehl, R. F., Friedel, S. & von Wiren, N. Plasticity of the
I. G. Some key metabolic intermediates in the biosynthesis of Arabidopsis root system under nutrient deficiencies. Plant Physiol.
botrydial and related compounds. Tetrahedron 57, 1929–1933 163, 161–179 (2013).
(2001). 56. Hiruma, K. & Saijo, Y. Methods for long-term stable storage of
36. Zhang, W. et al. Asymmetric total synthesis of the highly strained 4 Colletotrichum species. Methods Mol. Biol. 1398, 309–312 (2016).
beta-acetoxyprobotryane-9 beta,15 alpha-diol. J. Am. Chem. Soc. 57. Porra, R. J., Thompson, W. A. & Kriedemann, P. E. Determination of
142, 19868–19873 (2020). accurate extinction coefficients and simultaneous-equations for
37. Sakuraba, Y., Kim, Y. S., Han, S. H., Lee, B. D. & Paek, N. C. The assaying chlorophyll-a and chlorophyll-B extracted with 4 different
Arabidopsis transcription factor NAC016 promotes drought stress solvents—verification of the concentration of chlorophyll standards
responses by repressing AREB1 transcription through a trifurcate by atomic-absorption spectroscopy. Biochim. Biophys. Acta 975,
feed-forward regulatory loop involving NAP. Plant Cell 27, 384–394 (1989).
1771–1787 (2015). 58. Darling, A. C., Mau, B., Blattner, F. R. & Perna, N. T. Mauve: multiple
38. Okamoto, M., Vidmar, J. J. & Glass, A. D. Regulation of NRT1 and alignment of conserved genomic sequence with rearrangements.
NRT2 gene families of Arabidopsis thaliana: responses to nitrate Genome Res. 14, 1394–1403 (2004).
provision. Plant Cell Physiol. 44, 304–317 (2003). 59. Gu, Z., Gu, L., Eils, R., Schlesner, M. & Brors, B. circlize Implements
39. Colangelo, E. P. & Guerinot, M. L. The essential basic helix-loop- and enhances circular visualization in R. Bioinformatics 30,
helix protein FIT1 is required for the iron deficiency response. Plant 2811–2812 (2014).
Cell 16, 3400–3412 (2004). 60. Katoh, K. & Standley, D. M. MAFFT multiple sequence alignment
40. Ma, L. J. et al. Comparative genomics reveals mobile pathogenicity software version 7: improvements in performance and usability.
chromosomes in Fusarium. Nature 464, 367–373 (2010). Mol. Biol. Evol. 30, 772–780 (2013).
41. Savory, E. A. et al. Evolutionary transitions between beneficial and 61. Nguyen, L. T., Schmidt, H. A., von Haeseler, A. & Minh, B. Q. IQ-TREE:
phytopathogenic Rhodococcus challenge disease management. a fast and effective stochastic algorithm for estimating maximum-
Elife 6, e30925 (2017). likelihood phylogenies. Mol. Biol. Evol. 32, 268–274 (2015).
42. Melnyk, R. A., Hossain, S. S. & Haney, C. H. Convergent gain and loss 62. Emms, D. M. & Kelly, S. OrthoFinder: phylogenetic orthology
of genomic islands drive lifestyle changes in plant-associated inference for comparative genomics. Genome Biol. 20, 238 (2019).
Pseudomonas. Isme J. 13, 1575–1588 (2019). 63. Ohtsubo, Y., Ikeda-Ohtsubo, W., Nagata, Y. & Tsuda, M. Genome-
43. Baetsen-Young, A., Chen, H., Shiu, S. H. & Day, B. Contrasting Matcher: a graphical user interface for DNA sequence comparison.
transcriptional responses to Fusarium virguliforme colonization in BMC Bioinforma. 9, 376 (2008).
symptomatic and asymptomatic hosts. Plant Cell 33, 64. Li, Y. et al. A genome-scale phylogeny of the kingdom Fungi. Curr.
224–247 (2021). Biol. 31, 1653–1665 e1655 (2021).
44. Kusch, S. et al. Transcriptional response to host chemical cues 65. Kim, D. et al. TopHat2: accurate alignment of transcriptomes in the
underpins the expansion of host range in a fungal plant pathogen presence of insertions, deletions and gene fusions. Genome Biol.
lineage. ISME J. 16, 138–148 (2022). 14, R36 (2013).
45. Aleksza, D., Horvath, G. V., Sandor, G. & Szabados, L. Proline 66. Trapnell, C. et al. Differential gene and transcript expression ana-
accumulation is regulated by transcription factors associated with lysis of RNA-seq experiments with TopHat and Cufflinks. Nat. Pro-
phosphate starvation. Plant Physiol. 175, 555–567 (2017). toc. 7, 562–578 (2012).
46. Zhang, Y. et al. Abscisic acid facilitates phosphate acquisition 67. Ginestet, C. ggplot2: Elegant Graphics for Data Analysis. J. R. Stat.
through the transcription factor ABA INSENSITIVE5 in Arabidopsis. Soc. Stat. 174, 245–245 (2011).
Plant J. 111, 269–281 (2022). 68. Tian, T. et al. agriGO v2.0: a GO analysis toolkit for the agricultural
47. Medici, A. et al. AtNIGT1/HRS1 integrates nitrate and phosphate community, 2017 update. Nucleic Acids Res. 45, W122–W129 (2017).
signals at the Arabidopsis root tip. Nat. Commun. 6, 6274 69. Bailey, T. L. & Machanick, P. Inferring direct DNA binding from ChIP-
(2015). seq. Nucleic Acids Res. 40, e128 (2012).
48. Maeda, Y. et al. A NIGT1-centred transcriptional cascade regulates 70. Takino, J. et al. Elucidation of biosynthetic pathway of a plant hor-
nitrate signalling and incorporates phosphorus starvation signals in mone abscisic acid in phytopathogenic fungi. Biosci. Biotechnol.
Arabidopsis. Nat. Commun. 9, 1376 (2018). Biochem. 83, 1642–1649 (2019).
49. Morris, E. R., Chevalier, D. & Walker, J. C. DAWDLE, a forkhead- 71. Saika, H. et al. Ethylene promotes submergence-induced expres-
associated domain gene, regulates multiple aspects of plant sion of OsABA8ox1, a gene that encodes ABA 8 ‘-hydroxylase in rice.
development. Plant Physiol. 141, 932–941 (2006). Plant Cell Physiol. 48, 287–298 (2007).
50. Gonzalez-Guzman, M. et al. The short-chain alcohol dehydrogenase
ABA2 catalyzes the conversion of xanthoxin to abscisic aldehyde. Acknowledgements
Plant Cell 14, 1833–1846 (2002). We thank Mie Matsubara, Akemi Uchiyama, Shunsuke Imai, Takafumi
51. Mega, R., Tsujimoto, H. & Okamoto, M. Genetic manipulation of Shimizu, Sachiko Minezaki, and Hiromi Haba for their technical assis-
abscisic acid receptors enables modulation of water use efficiency. tance. We thank Paul Schulze-Lefert, Soledad Sacristán, Ryo Tabata,
Plant Signal Behav. 14, e1642039 (2019). Chika Tateda, Momoko Takagi, Yuri Tajima, and Ren Ujimatsu for the
52. Gonzalez, E., Solano, R., Rubio, V., Leyva, A. & Paz-Ares, J. PHOS- fruitful discussion. We thank Yasuyuki Kubo, Maarten Koornneef, Javier
PHATE TRANSPORTER TRAFFIC FACILITATOR1 is a plant-specific Paz-Ares, Shigetaka Yasuda, Haruhiko Inoue, and Richard O’Connell for
SEC12-related protein that enables the endoplasmic reticulum exit published materials. This work was supported in part by the JSPS
of a high-affinity phosphate transporter in Arabidopsis. Plant Cell 17, KAKENHI Grant (16H06279, 18K14466, 18H04822, 19H05688,
3500–3512 (2005). 20H02986, 21H05150, and 22H02204 (A.M.)), the JST grant
53. Tsay, Y. F., Schroeder, J. I., Feldmann, K. A. & Crawford, N. M. The (JPMJPR16Q7, JPMJCR19S2, JPMJSC1702, and JPMJFR200A) and The
herbicide sensitivity gene CHL1 of Arabidopsis encodes a nitrate- Uehara Memorial Foundation (A.M.). Computations were partially per-
inducible nitrate transporter. Cell 72, 705–713 (1993). formed on the NIG supercomputer at ROIS National Institute of Genetics.

Author contributions Reprints and permissions information is available at

K.H. initiated the project. K.H. and Y.S. directed the research. K.H., J.T., http://www.nature.com/reprints
T.H., A.S., M.O., N.K., M.N., N.K., Y.O., R.S., and K.T. conducted the
experiments. Y.U. conducted comparative genomic analyses except for Publisher’s note Springer Nature remains neutral with regard to jur-
molecular phylogenetic tree analyses. S.A. conducted analyses of isdictional claims in published maps and institutional affiliations.
molecular phylogenetic trees, supervised by W.I. K.H. conducted RNA-
seq analyses. J.T. conducted a metabolites analysis supervised by H.O. Open Access This article is licensed under a Creative Commons
and A.M. T.S. provided unique material. K.H., S.A., T.H., M.O., Y.U., N.K., Attribution 4.0 International License, which permits use, sharing,
M.N., Y.O., R.S., K.T., A.M., and W.I. analyzed the data. K.H., S.A., W.I., and adaptation, distribution and reproduction in any medium or format, as
Y.S. wrote the paper with feedback from all authors. long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license, and indicate if
Competing interests changes were made. The images or other third party material in this
The authors declare no competing interests. article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not
Additional information included in the article’s Creative Commons license and your intended
Supplementary information The online version contains use is not permitted by statutory regulation or exceeds the permitted
supplementary material available at use, you will need to obtain permission directly from the copyright
https://doi.org/10.1038/s41467-023-40867-w. holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
Correspondence and requests for materials should be addressed to Kei
Hiruma. © The Author(s) 2023
Peer review information Nature Communications thanks Riccardo Bar-

oncelli and the other anonymous reviewer(s) for their contribution to the
peer review of this work. A peer review file is available.

A Fungal Sesquiterpene Biosynthesis Gene Cluster Critical For Mutualist-Pathogen Transition in Colletotrichum Tofieldiae

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Fungal Sesquiterpene Biosynthesis Gene Cluster Critical For Mutualist-Pathogen Transition in Colletotrichum Tofieldiae

Uploaded by

Copyright:

Available Formats

Article https://doi.org/10.

A fungal sesquiterpene biosynthesis gene

Plant-associated fungi show diverse lifestyles from pathogenic to mutualistic

Nature Communications | (2023)14:5288 1

Nature Communications | (2023)14:5288 2

Chlorophyll a + b (μg/g SFW)

Mock Ct61 Ct4 Ct3

g Intra-cellular hyphae Inter-cellular hyphae h Intra-cellular hyphae Inter-cellular hyphae

Nature Communications | (2023)14:5288 3

Nature Communications | (2023)14:5288 4

Nature Communications | (2023)14:5288 5

Normalized expression value

= > > Ct3_m_0 Ct61_m_2 Ct61_P_0 a p_Ct4e_2 3

Nature Communications | (2023)14:5288 6

XP_001553966.2 XP_001553967.1 XP_024550393.1

Putative ABA-BOT biosynthesis cluster

GKT56071.1 GKT56074.1 GKT56075.1

Putative ABA cluster

200 ABA BOT1-5

Monosporascus sp. CRB-8-3 ABA BOT1-5 BOT6,7

ABA+BOT Botrytis cinerea B05.10 ABA BOT1-5 BOT6,7

Nature Communications | (2023)14:5288 7

Nature Communications | (2023)14:5288 8

(m/z = 164) (m/z = 180)

(B) Ct3 wild type

(C) Ct3 wild type

12.0 15.0 18.0 (min) 12.0 15.0 18.0 (min)

Nature Communications | (2023)14:5288 9

Nature Communications | (2023)14:5288 10

Nature Communications | (2023)14:5288 11

Up-regulated A. thaliana genes −2 −1 0

during Ct3 colonization

288 A. thaliana genes

・Response to oxidative stress

・Response to abscisic acid

Down-regulated A. thaliana genes

Ct3: Colletotrichum tofieldiae Ct3

during Ct3 colonization

375 A. thaliana genes

Bc: Botrytis cinerea B05.10

Hs: Homo sapiens

・Inorganic anion transport

・Inorganic ion homeostasis

・Secondary metabolic process

Nature Communications | (2023)14:5288 12

Growth promotion Growth promotion

Nature Communications | (2023)14:5288 13

Nature Communications | (2023)14:5288 14

Nature Communications | (2023)14:5288 15

Sucrose measurement in plants 15. Marumo, S. et al. Microbial-production of abscisic-acid by Botrytis

Nature Communications | (2023)14:5288 16

Nature Communications | (2023)14:5288 17

Author contributions Reprints and permissions information is available at

Peer review information Nature Communications thanks Riccardo Bar-

Nature Communications | (2023)14:5288 18

You might also like