Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease

Lily D Peck; Theo Llewellyn; Bastien Bennetot; Samuel O’Donnell; Reuben W Nowell; Matthew J Ryan; Julie Flood; Ricardo C Rodríguez de la Vega; Jeanne Ropars; Tatiana Giraud; Pietro D Spanu; Timothy G Barraclough

doi:10.1371/journal.pbio.3002480

. 2024 Dec 5;22(12):e3002480. doi: 10.1371/journal.pbio.3002480

Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease

Lily D Peck ^1,^2,^3,^*, Theo Llewellyn ^1,⁴, Bastien Bennetot ⁵, Samuel O’Donnell ⁵, Reuben W Nowell ^6,⁷, Matthew J Ryan ³, Julie Flood ³, Ricardo C Rodríguez de la Vega ⁵, Jeanne Ropars ⁵, Tatiana Giraud ⁵, Pietro D Spanu ⁸, Timothy G Barraclough ^2,⁶

Editor: Sophien Kamoun⁹

PMCID: PMC11620798 PMID: 39637834

Abstract

Outbreaks of fungal diseases have devastated plants and animals throughout history. Over the past century, the repeated emergence of coffee wilt disease caused by the fungal pathogen Fusarium xylarioides severely impacted coffee production across sub-Saharan Africa. To improve the disease management of such pathogens, it is crucial to understand their genetic structure and evolutionary potential. We compared the genomes of 13 historic strains spanning 6 decades and multiple disease outbreaks to investigate population structure and host specialisation. We found that F. xylarioides comprised at least 4 distinct lineages: 1 host-specific to Coffea arabica, 1 to C. canephora var. robusta, and 2 historic lineages isolated from various Coffea species. The presence/absence of large genomic regions across populations, the higher genetic similarities of these regions between species than expected based on genome-wide divergence and their locations in different loci in genomes across populations showed that horizontal transfers of effector genes from members of the F. oxysporum species complex contributed to host specificity. Multiple transfers into F. xylarioides populations matched different parts of the F. oxysporum mobile pathogenicity chromosome and were enriched in effector genes and transposons. Effector genes in this region and other carbohydrate-active enzymes important in the breakdown of plant cell walls were shown by transcriptomics to be highly expressed during infection of C. arabica by the fungal arabica strains. Widespread sharing of specific transposons between F. xylarioides and F. oxysporum, and the correspondence of a putative horizontally transferred regions to a Starship (large mobile element involved in horizontal gene transfers in fungi), reinforce the inference of horizontal transfers and suggest that mobile elements were involved. Our results support the hypothesis that horizontal gene transfers contributed to the repeated emergence of coffee wilt disease.

Outbreaks of coffee wilt disease, caused by the fungus Fusarium xylarioides, have severely impacted coffee production. By sequencing and comparing 13 historical strains spanning 60 years and multiple outbreaks, this study shows that successive horizontal gene transfers from F. oxysporum contributed to host specificity and disease emergence.

Introduction

Fungal diseases have devastated plants and animals throughout history, with the first records dating from biblical plagues [1]. Many cases of famines and harvest failures resulted from outbreaks of fungal diseases with wide-ranging and important human consequences, with between 10% and 23% yields lost to fungal disease, and modern farming practices risk inadvertently helping the emergence of new pathogens [2]. Coffee is a major commodity that is worth over $22 billion to the global economy [3], provides high export earnings (e.g., one third exports in Ethiopia [4]) and supports millions of smallholder farmers [5]. However, as with other crops, there is a constant threat from pests and disease, especially fungal pathogens. Coffee wilt and leaf rust have caused major losses to crop yields and livelihoods in Africa and Central America, respectively [6,7]. Pathogens can evolve rapidly [2], so understanding how new variants emerge is vital to developing sustainable disease control measures.

Coffee wilt disease, caused by the fungal pathogen Fusarium xylarioides, infects coffee plants via roots or wounds to colonise the xylem [8]. Wilting results from the pathogen blocking water uptake, and cell walls are degraded for nutrients [9]. Currently, 2 host-specialised strain types are known to infect the main African coffee crops: arabica (Coffea arabica), grown at high altitudes, and robusta (a variety of C. canephora), grown at low-mid altitudes in tropical Africa [8]. Coffee wilt disease has a well-documented history of emergence and outbreaks since its discovery in 1927 [6]. By the 1950s, coffee crops in west and central Africa were decimated, resulting in the replacement of commercial cultivars with resistant robusta coffee [6]. Yet by the 1970s, coffee wilt disease re-emerged on robusta coffee in central Africa [10], becoming a major problem in east and central Africa by the mid-1990s. Production did not recover in some countries (e.g., Democratic Republic of Congo, [11]). In parallel, coffee wilt disease was first reported on arabica coffee in Ethiopia in the 1950s, becoming widespread in the 1970s, but with low severity of disease [12]. However, one recent study suggests that disease severity in Ethiopia has increased [13]. Key questions for understanding the emergence of coffee wilt disease include: (i) what are the genetic relationships among the strain types on different hosts and at different times; and (ii) what genetic changes underlie their ability to infect different hosts?

There is growing evidence that horizontally transferred regions (HTRs) between distantly related fungi are important for the emergence of new pathogen variants [14]. For example, a 14 kb fragment encoding the ToxA virulence protein and abundant transposons was transferred between 3 wheat fungal pathogens of the order Pleosporales [15]. This fragment was later identified as being part of a Starship [16], a newly described group of large mobile elements found across every major class of filamentous ascomycetes that can conflict or cooperate with their fungal host genome [17]. Starships mediate the movement of large sets of genes within genomes, species, or even between species through horizontal transfer. They were recently linked to genetic and phenotypic variability in a human pathogen [18]. In addition to genomic evidence, there has been experimental evidence of Starships as autonomous transposons [19].

Interspecies transfers can also occur via mobile accessory chromosomes. For example, F. oxysporum, which causes wilt disease in over 100 plant species with host-specific formae speciales [20], contains supernumerary chromosomes which transfer between formae speciales and determine host specificity [21–23]. These chromosomes contain 4 named “supercontigs,” subregions enriched in secreted in xylem effector genes (SIX) and mimp transposons. SIX genes encode small proteins secreted by F. oxysporum into the xylem of a host plant during infection and which promote virulence [24], while mimp transposons are a type of nonautonomous miniature-inverted impala repeat which are always present in the promoter of F. oxysporum SIX genes and consequently associated with virulence [25]. Transposons are hypothesised to promote hyper-variability of sets of effector genes, either directly by transposition or by facilitating recombination to generate new gene assortments [26]. Thus, while core fungal genomes may generally evolve separately within local species gene pools, transfers between more distant relatives can introduce diversity and enable outbreaks of new types of pathogen [27].

A previous study sequenced 6 F. xylarioides genomes comprising arabica, robusta, and strains isolated in 1950s which we describe as “coffea” for their ability to infect multiple Coffea species [28]. Genome-wide variation supported an early split between arabica and other strain types, consistent with pre-agricultural divergence on wild coffee species, and potentially a separate gene pool on C. arabica. In contrast, the robusta strain type from the 1970s onwards appeared to derive from the pre-1970s coffea strain type outbreak. Different strain types featured different putative effector genes, whereby some effectors showed close matches with F. oxysporum, which were lacking in closer F. xylarioides relatives. Strikingly, a 20-kb scaffold showed high sequence similarity to an F. oxysporum pathogenicity chromosome and contained the same genes and mimp transposons, while these were lacking in closer F. xylarioides relatives. This led to the hypothesis that host specialisation and changes in subsequent outbreaks involved divergent strain types receiving HTRs from several F. oxysporum formae speciales (ff. spp.) and thus a greatly expanded gene pool of wilt-associated effector genes.

Here, we test this hypothesis using comparative genomics of 11 Fusarium genomes, including F. oxysporum and F. solani chosen to represent potential HTR donors in crops that grow with or near coffee in Africa (S1 Table). Aligning F. xylarioides genomes with a new long-read reference genome for the arabica strain type, we found that the different outbreaks on diverse coffee types arose in separate populations. The populations differ in the presence and absence of large genomic regions that, when present, show high sequence similarity with multiple different F. oxysporum ff. spp. compared to genome-wide similarity, but are found in non-homologous locations in genomes, contrasting with an otherwise high synteny level; this is consistent with horizontal acquisition from distantly related lineages. Transcriptomics revealed that putative effector genes were highly expressed during coffee plant infection, with several putatively originating from F. oxysporum and absent in closely related species to F. xylarioides. Finally, all F. xylarioides populations share highly similar mimps and other transposable elements associated with F. oxysporum’s mobile pathogenic chromosome, and, notably, a Starship element, which could have mediated horizontal gene transfer (HGT) between these 2 species. Together, these results support the hypothesis that HGT contributed to the emergence of host-specialised populations that caused successive outbreaks of coffee wilt disease.

Results

A high-quality genome assembly for the Fusarium xylarioides arabica host specialist

Assembly of Nanopore long-read data for F. xylarioides IMI 389563 (“arabica563”) generated its first high-quality reference genome with 21 contigs, including 12 contigs over 1 Mb, and 4 contigs over 100 kb. Telomeric repeats were identified on both ends of 3 chromosomes, consistent with these representing complete assembled chromosomes, and at just one end of 7 contigs, probably representing incomplete chromosomes (S5B Fig). The assembly length was 57 Mb (with 55 Mb in the 12 longest contigs) with an N50 greater than 5 Mb (Table 1). The RagTag assemblies ordered and oriented to the reference genome for the remaining F. xylarioides strains allowed the identification of various large arabica-specific regions over 0.7 Mb in contigs 6, 7, 8, 12, and 13, compared to robusta-specific genomes. The regions in contigs 7, 12, and 13 were absent from all robusta- and coffea-specific genomes.

Table 1. Statistics for sequenced Fusarium genomes.

Arabica563 is the reference assembly. Only contigs >1 kb in length were included. Fusarium oxysporum is abbreviated to Fo.

Strain name	Strain ID	Genome size	No. contigs	N50 contig length	Genome completeness	Genome GC	Coding genes
	(IMI)	Mb (>1 kb)	n (> 1kb)	n (> 1kb)	BUSCO (%)	%	n
Arabica563	389563	57.3	21	5,156	100.0	42.2	17,424
Arabica038	507038	51.1	4631	24.5	97.9	43.4	15,705
Robusta268	392268	50.2	4420	27.1	97.6	44.1	15,575
Coffea035	507035	51.1	3880	30.4	97.9	43.9	15,588
Coffea113	507113	48.7	3795	30.1	98.3	44.3	15,115
Coffea676	392676	50.0	4362	27.1	97.9	44.4	15,649
F. solani 280	392280	48.4	1867	45.2	96.9	50.9	14,007
Fo 509	244509	49.7	1923	67.9	98.3	47.4	14,475
Fo cubense 109	141109	48.4	1992	60.1	99.0	47.6	14,993
Fo pisi 221	500221	52.9	4299	28.1	95.5	47.5	16,614
Fo raphani 541	337541	53.8	4285	38.1	98.3	47.7	16,421
Fo vasinfectum 248	292248	49.4	1904	72.5	98.6	47.4	14,846

Open in a new tab

Fusarium xylarioides is a monophyletic species complex with genetically differentiated populations

Long-read sequencing of a new F. xylarioides reference genome revealed the same core chromosomes as F. verticilliodes (Fig 1A), another species of the Fusarium fujikuroi species complex (Fig 1C). Short-read genome sequencing of F. xylarioides strains isolated across the spatial and temporal range of coffee wilt disease (Fig 1B) supported the delimitation of F. xylarioides as a species complex (Figs 1C and S1). Analysing 11 new Illumina genome assemblies of 5 F. xylarioides strains and 6 other Fusarium genomes (S1 Table and Fig 1), together with 30 published Fusarium genomes (S2 Table), found 26,000 orthologous groups of protein-coding genes, of which 3,544 had members in all genomes with 1,685 single-copy genes. Two species trees, a “supermatrix” tree reconstructed from single-copy gene trees (Figs 1C and S1), as well as a “supergene” tree reconstructed from the 3,544 orthologous gene trees (S2 Fig), supported phylogenetic relationships from earlier studies with lower sampling [28,29]. Fusarium xylarioides forms a monophyletic clade within the F. fujikuroi species complex, a result that receives 100% bootstrap branch support in concatenated analysis of single-copy genes (a “supermatrix” tree, 1,000 replicates, Fig 1C) and support from 58% of genes in the “supergene” tree (branch support values, S2 Fig). Average nucleotide identities in coding sequences within F. xylarioides (98%) are high compared to lower identities with other species (F. xylarioides to F. phyllophilum 96%, to the F. fujikuroi complex 91% to 92%, and to F. oxysporum strains 90%, S3 Fig). Multilocus species delimitation based on patterns of congruence/incongruence among gene trees (using the tr2 algorithm) identified significant units that matched named species except for excluding one strain from F. oxysporum (F. oxysporum rapae Tf1208 isolate, Fig 1C). Within F. xylarioides, 4 sub-clades were present: arabica, robusta, and 2 separate clades of coffea genomes (differentiating those isolated from the pre-1970s outbreak, strains 659 and 676, from the others) (Figs 1C and S1). Across all orthogroups present in these sub-clades, 80% genes are shared by all clades (S4 Fig), yet relationships were not highly congruent across gene trees, with branch support values <60% in the “supergene” tree (Fig 1C). This lack of congruence may reflect incomplete lineage sorting due to short internode intervals.

Analyses of genetic variation within F. xylarioides revealed population subdivision consistent with the hypothesis that multiple outbreaks arose from separate gene pools. A NeighbourNet analysis based on 239,944 single-nucleotide polymorphisms (SNPs) from coding and noncoding regions suggests 4 differentiated populations (Fig 2, absolute divergence d_XY >0.001 for all pairwise comparisons, S3 Table): an arabica group, a robusta group, and the same 2 coffea clades as recovered on the species tree (Fig 1C). A small amount of reticulation occurs among strains in the robusta group, indicating current or ancient gene flow. There is no reticulation between the arabica group and either the robusta or coffea groups, suggesting a lack of gene flow and recombination between the arabica clade and the other clades.

Fig 2 — A Neighbour-Net (SplitsTree) analysis based on a single-nucleotide polymorphism distance matrix. The scale bar represents 0.1 substitutions per polymorphic site. The points are shaded according to their host population. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

The populations differ in the presence of multiple large genome regions

We identified numerous large genomic regions (search size 10 kb) that varied in their presence/absence between the F. xylarioides arabica genomes and the other F. xylarioides populations (presence defined as >50% similarity after whole-genome alignment using minimap2, see Methods). In total, 2.35 Mb were differentially present among the F. xylarioides populations, with nearly half (1.7 Mb) specific to the arabica population, comprising 70 mostly-contiguous regions (namely a gap of 20 kb between regions is still considered contiguous) of 10 kb or more on all chromosomes (Fig 3A). The longest contiguous arabica-specific region was a 360 kb region of chromosome 1 (contig 13 in S5 Fig). Together, these regions contained 1,071 predicted genes and 2,623 annotated repeats. We therefore hypothesised that the gain or loss of genome regions could underpin specialisation of the different F. xylarioides host populations.

(A) The genomes of the F. *xylarioides* coffea1, coffea2, and robusta populations compared to the arabica563 reference show high coverage. Y-axis shows coverage (fraction of bases covered) from minimap2 (-cx asm10 --secondary = no --cs) mapping between the arabica reference and each coffea1 (coloured peach, top bar), coffea2 (coloured pink, middle bar), and robusta (coloured turquoise, bottom bar) genomes (data plotted as average of 100 × 100 bp = 10 kb windows). See S1 and S2 Tables for strain details, vertical black bars in each plot denote the original contigs, see S12 Table for contig names. However, some regions are differentially present or absent between populations, annotated with black boxes in the bar below the x-axes (71 in total comprising 2.35 Mb). These indicate the F. *xylarioides* host-specific genome regions, namely: those present only in arabica; in arabica and robusta; in arabica and coffea1; in arabica and coffea2. sc1, supercontig 1; sc2, supercontig 2. (B i) Unrooted ML tree of HTR 1 and matching sequences from different *Fusarium* genomes. A BLAST search against the “nr” database with the F. *xylarioides* HTR 1 sequence as query recovered its sequence in 3 *Fusarium* species: F. *xylarioides* arabica (green tip labels), F. *oxysporum* (orange tip labels), and F. *redolens* (grey tip labels). The ML tree shows high bootstrap support (100% from 1,000 bootstraps) for clustering of the HTR 1 sequences of F. *xylarioides* arabica with that of F. *oxysporum*. The HTR 1 sequence is absent from other F. *xylarioides* populations and F. *fujikuroi* complex species. Bootstrap support for all internal nodes with a support value >70% are shown. Tip labels correspond to NCBI GenBank accession numbers. *FOSC*, F. *oxysporum* species complex. The scale bar indicates branch length (number of substitutions per site). (B ii) Sliding window plot of genome similarity between F. *xylarioides* host-specific populations and F. *oxysporum* as in Fig 3A but zooming in on the HTR 1 region with the addition of F. *oxysporum* f. sp. *lycopersici* GCA_000149955 (orange line) that shares a highly supported clade with F. *xylarioides* (XM 018402122). The location of HTR 1 is indicated with a blue line and is absent from coffea1, coffea2, and robusta populations (each line is at 0 on the y-axis), whereas it is present and similar in F. *oxysporum* f. sp. *lycopersici* GCA_000149955 (2 peaks with >90% similarity). (C i) Unrooted ML tree of HTR 2 from different *Fusarium* genomes. Same as above, but HTR 2 was recovered in F. *xylarioides* arabica, coffea1, and coffea2 populations (green, peach, and pink tip labels, respectively), F. *oxysporum* (orange tip labels), and F. *fujikuroi* species complex genomes (yellow tip labels, F. *verticillioides*, F. *mangiferae*, F. *fujikuroi*, F. *proliferatum*) and distantly related *Fusarium* species (grey tip labels). (C ii) Sliding window plot of genome similarity for HTR 2. Same as above, but with the addition of F. *oxysporum* Fo5176 GCA_030345115 (orange line) because this genome shares a highly supported clade with F. *xylarioides* (CP128298, CP128283, CP128285, CP128282, CP128291, CP128288, CP128296). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

Several population-specific genome regions show close matches to pathogenicity chromosomes of multiple Fusarium oxysporum lineages

We compared population-specific genome regions to all genomes in the NCBI nonredundant (“nr”) database and found that 9 contiguous large genomic regions (>20 kb) were also present in F. oxysporum. Of those regions, 5 were absent from other F. fujikuroi complex genomes and with unexpectedly strong BLAST hits to various F. oxysporum lineages (Evalue<10⁻²⁰, Figs 3B, 3C, S6A–S6C and 4.) An additional region was specific to the robusta and coffea genomes but with a strong BLAST hit to F. oxysporum f. sp. lycopersici (E-value 10⁻¹⁶⁹, S6D and S7 Figs). We hypothesised that these could be putative HTRs shared by F. xylarioides and F. oxysporum, which we named HTR 1 to 6. They range from 20 kb (HTR 4) to 200 kb (HTR 1). Putative HTRs 1, 3, and 4 were found only in the arabica population; HTR 2 was found in arabica, coffea1, and coffea2 populations; HTR 5 was found in all F. xylarioides populations; and HTR 6 in robusta, coffea1, and coffea2 populations. HTR 2 (in arabica and coffeas) and HTR 6 (in robusta and coffeas) showed strong matches to different parts of “supercontig 51” from the mobile pathogenic chromosome in F. oxysporum f. sp. lycopersici. For HTRs 1 to 4, the mean sequence identity to closest matches in other F. xylarioides strains (denoted I Fx) was 5.8%, 5.8%, 4.7%, and 0%, respectively, whereas the mean identity to closest F. oxysporum matches (I_Fo) was 55.2%, 39.3%, 84.4%, and 88.5%. To test if these values were statistical outliers compared to the genome-wide average, we aligned similarly sized windows to both F. xylarioides and F. oxysporum genomes and compared the observed values of the difference (I_Diff = I_Fx - I_Fo) to this background distribution. The proportion of background windows with an I_Diff < the observed value was 0.018%, 6.7%, 0.25%, and 0.54% for HTRs 1 to 4, respectively, indicating that most HTR regions were extreme outliers compared to the background distribution of sequence identity across F. xylarioides and F. oxysporum genomes (S13 Table).

Fig 4 — Long-read assemblies for each genome were used (GCA_000149955.2 for F. *oxysporum* f. sp. *lycopersici* and F. *oxysporum* f. sp. *cubense* [31]). Nucmer whole genome alignments were visualised in RIdeogram. The links indicate matching blocks larger than 200 bp, coloured in grey. Accessory chromosomes, as described by [21,31], are coloured in turquoise. The HTRs are shown on the chromosomes by colours as indicated on the figure key.

Phylogenetic analyses on alignments of each whole putative HTR provided further support for the horizontal transfer hypothesis. In every case, F. xylarioides grouped with F. oxysporum genomes with strong bootstrap support (all >99%). In 5 cases (HTRs 1, 3–6), the putative HTRs were not recovered from any species from the F. fujikori complex (Figs 3B and S6), which should be closer phylogenetic relatives of F. xylarioides (Fig 1C). We report a close relationship between HTR 4 and F. oxysporum 509 (S6D Fig), isolated from a coffee tree and sequenced in this study, as well as F. oxysporum f. sp. pisi, also sequenced here, which clustered with F. xylarioides arabica nodes for HTR 5 (S6C Fig). To test if pairwise divergence between the putative HTRs and the nearest F. oxysporum (D_FxFo) copy were statistical outliers compared to background levels of divergence, we compared putative HTR gene trees with the nearest F. oxysporum copy (D_FxFo), and then compared the observed values of the difference with the metric of divergence in single-copy gene trees. The proportion of HTRs with a D_FxFo ≤ the observed value was: HTR 1, 9.4%; HTR 3, 1.1%; HTR 4, 4.3%; HTR 5, 0.13%; HTR 6, 0.84%, S14 Table. In HTR 2, the same genome region was recovered from one or more F. fujikori complex species, but they clustered with different F. oxysporum copies in the tree (Figs 3C, S5A and S5B). Pairwise divergence between the F. xylarioides and the closest F. oxysporum match (D_FxFo) was unexpectedly lower than the divergence between F. xylarioides and the closest species from the F. fujikori complex species match (D_FxFFC), relative to the expected levels (percentile of D_FxFo < D_FxFFC) across single copy ortholog trees: 0.0%. In all cases, F. oxysporum sequences were highly divergent with long branch lengths in maximum likelihood (ML) trees in Figs 3 and S6, whereas F. xylarioides displayed low variation and short/identical branch lengths in the same ML trees.

The putative HTRs all include BLAST hits to extremities of mobile and pathogenicity chromosomes known to be associated with host specialisation in F. oxysporum formae speciales (Fig 4), but never to a single chromosome. For example, the largest HTR (HTR 1) is on chromosome 1 (contig 1.1 in Fig 1A) in F. xylarioides arabica strains, in a region which is absent from the F. verticillioides chromosome assembly (Figs 1A and S5A). One half of contig 1.1 is found in all strains (e.g., Figs 1A and S5B), albeit with very low sequence identity outside F. xylarioides. The other half, which mostly comprises HTR 1, is unique to the F. xylarioides arabica population and F. oxysporum, with high identity (>92%) in relatively short pieces (F. oxysporum f. sp. vasinfectum CP130309, 22.6 kb; F. oxysporum Fo5176 CP128295, 19.2 kb; F. oxysporum f. sp. lycopersici XM_018402123, 7 kb) compared to the total 150 kb putative HTR region (Fig 3C). Compared to the best-assembled F. oxysporum f. sp. lycopersici and F. oxysporum f. sp. cubense genomes, a synteny alignment reveals that HTR 1 matches in part to all of their mobile chromosomes, namely chromosomes 1, 3, 6, 14, and 15 in F. oxysporum f. sp. lycopersici and chromosomes 1 and 12 (along with other non-accessory regions) in F. oxysporum f. sp. cubense (Fig 4). Within HTR 1, there are 39 genes with a high sequence identity between the F. oxysporum f. sp. lycopersici and F. xylarioides arabica strains (BLAST, >94% nucleotide identity compared with 88% identity in flanking regions and whole genome identity of 90%, S3 and S8 Figs). Two of these genes belong to orthologous groups only found in F. oxysporum and the F. xylarioides arabica strains, with highly similar F. xylarioides arabica copies nested within diverse F. oxysporum copies from various formae speciales (S8B and S8C Fig).

As a second example, HTR 2 comprises a 30-kb region on chromosome 5 (Figs 3C and 4). It contains 3 genes shared by all F. xylarioides arabica and coffea populations (arabica563 gene copies are H9Q71 0016558, H9Q71 0017194, H9Q71 0017193) and which match to F. oxysporum effector proteins: Six7 with 91% identity, Six10 with 92% identity, and Six12 with 94% identity (S7 Fig and S4 Table). Other SIX effector genes (six1, six2, six6, six7, and six11) were identified in members of the F. fujikuroi species complex, with vertical (for six2) and horizontal inheritance from the F. oxysporum species complex suggested [32]. In addition, we previously identified six7 and six10 in arabica and coffea populations [28]. There were no significant BLAST hits in the intergenic regions of HTR 2 outside F. xylarioides. Each SIX gene either contains a mimp in its gene body (six10 and six12) or in its promoter region (six7, mimp 943bp upstream) (S7 Fig). In F. oxysporum f. sp. lycopersici, these 3 genes are found in supercontig 51 on the mobile, pathogenic chromosome, a region enriched in effector genes and mimps, among other transposable elements [25]. Remarkably, HTR 6, in the robusta and coffea populations, is shared with a different part of F. oxysporum f. sp lycopersici’s supercontig 51 to that shared with HTR 2 (Figs 3C, S6D, and S7). In each genome, HTR 6 comprises a single scaffold (approximately 20 kb): All genes in HTR 6 are unique to the F. xylarioides robusta and coffea populations and various F. oxysporum ff. spp., as identified by BLAST (S6D Fig). This region does not contain SIX effectors, but other genes important in infection, including a glycosyltransferase, Cytochrome P450 monooxygenases, a squalene-hopene-cyclase, a methyltransferase-UbiE protein, and a Tri7 homolog.

Putative horizontally transferred regions contain genes that are highly expressed during infection, including effector genes from Fusarium oxysporum

Among the expressed genes and relative to their relevant controls, an average of 2,164 were differentially expressed across the arabica563 samples, 1,713 across the arabica908 samples, and 233 across the coffee plant control samples. In the fungal samples, over half of the differentially expressed genes were up-regulated in vivo infection samples relative to controls, while in the coffee plant samples 80% of the differentially expressed genes (116 in the plants infected with arabica563; 231 with arabica908) were up-regulated in the infected samples. Genes which were differentially expressed in both F. xylarioides arabica strains belong to nearly 700 orthologous groups. The differentially expressed coffee genes are described in S10 Table.

To determine whether genes found on putative HTRs are associated with coffee wilt infection, we analysed up-regulated genes in planta during infection. Diverse carbohydrate-active enzyme (CAZyme) families capable of breaking down cellulose and pectin from plant cell walls, and effectors known in F. oxysporum such as the SIX genes, were previously identified in F. xylarioides arabica and robusta as putative effectors [28], but their expression during infection was unexplored. Analysis of gene expression of 2 strains of the F. xylarioides arabica clade from infected arabica coffee plants, compared with the same strains grown in axenic liquid culture, revealed significant up-regulation in planta of genes encoding CAZyme families important in the breakdown of pectin, xylan, and other plant cell wall polymers (Fig 5A and 5B). In both strains, 7 pectin lyases (within the polysaccharide lyase PL family) were previously predicted as effectors: 3 PL1 genes; 3 PL3 genes including the pelA and pelD effectors characterised by [33]; and 1 PL9 gene (S5 Table). Most of the 23 pectin lyase genes from the subfamilies PL1, PL3, PL4, PL9, and PL26, showed strong up-regulation with 80% differentially expressed and 50% making up the top 20% of expressed genes (see S8 and S9 Tables for details).

Fig 5 — (A) Points represent individual genes plotted by log2 fold-change in expression level on the X-axis and significance (–log10 FDR) on the Y-axis. Significantly up-regulated genes in planta samples were identified using negative binomial generalised linear models and shaded orange and those down-regulated are shaded blue. Up-regulation included the F. *oxysporum SIX*, that is secreted in xylem, effectors (annotated), and pectate lyase and other CAZymes involved in pectin metabolism (shaded dark orange). The CAZymes encompass AA9, CE12, CE5, GH105, GH11, GH12, GH43, GH5, GH51, PL1, PL3, PL4, PL9, and PL26. (B) Arabica coffee plants 99 days post inoculation, infected at left, that is inoculated with *F. xylarioides* arabica563, and control at right, that is mock-inoculated with sterile water. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

Several putative effectors previously identified [28] were found to be significantly up-regulated in F. xylarioides from infected coffee plants, compared with fungi cultured in vitro (S6 and S7 Tables). Six putative effectors known from the mobile pathogenic chromosome of F. oxysporum f. sp. lycopersici were up-regulated in at least 1 arabica strain. These were: H9Q71 0011695 which appears to be a divergent type of six7 effector (referred to as OG0014398 in [28]); 3 novel effector candidates identified by [25] which are all secreted proteins—unidentified FOXG 14254, oxidoreductase orx1, and a catalase-peroxidase found in infected xylem sap FOXG 17460; sge1, the SIX effector gene transcription factor; H9Q71 0017133, a LysM effector (referred to as OG0013477 in [28]); and 2 pectin lyase effectors, pelA and pelD. Only 1 putative effector, H9Q71 0014367, was expressed less in infected plants than control samples in both strains cultured in vitro. Seven remaining putative effectors were expressed but not differentially in both strains; 6 genes were expressed but not differentially in 1 strain (S6 and S7 Tables and S9 Fig). In addition to these putative effectors, reanalysis of the new reference genome with EffectorP 3.0, a predictor of apoplastic and cytoplasmic effectors in fungi [34] detected further pectinolytic enzymes, similar extracellular degradative enzymes, as well as the established fungal effector domains Ecp2 and LysM (see S5 Table for details), which were also up-regulated in infected plants.

There were 496 genes significantly up-regulated in planta in both arabica strains, of these, 19 were absent from F. fujikuroi complex species and mostly did not match any recognised domains (S4 Table). Of these 19 genes, remarkably, 3 were found in HTR 1 and were shared by F. xylarioides arabica and coffea populations only (H9Q71 0016558, H9Q71 0017194, H9Q71 0017193). Each matches to an F. oxysporum effector protein: Six7, Six10, and Six12, respectively (S4 Table). The SIX orthologues were also in the top 20% most highly expressed genes in planta across the arabica strains (Fig 5B). Of the remaining genes absent from F. fujikuroi complex species, several were population specific; some were shared by all F. xylarioides populations; and some were shared between F. xylarioides and other species complexes (S4 Table).

Miniature impala elements and other transposons are shared between Fusarium xylarioides and Fusarium oxysporum

Certain transposable element (TE) families have been implicated in promoting effector gene variability and mobility in F. oxysporum and previously in F. xylarioides [25,28,32], and so we surveyed for the same transposon classes in the new F. xylarioides genomes. A recent study found one or a few mimps in 3 F. fujikuroi- and one F. tricinctum complex species, and linked their presence to horizontal transfer [32]. Fusarium xylarioides genomes contained miniature impala elements (mimps) that are present in F. oxysporum genomes but lacking in more closely related species, as well as other transposons previously identified in mobile F. oxysporum chromosomes. These elements are present in some putative HTRs and close to or inside several highly expressed genes. A total of 450 full-length mimps were detected across the F. xylarioides and F. oxysporum genomes (Fig 6A). Fusarium oxysporum formae speciales genomes contained the most mimps, with F. oxysporum f. sp. lycopersici and F. oxysporum f. sp. raphani in the highest copy numbers (90 and 85, respectively, Fig 6). The mimp families 1, 2, and 4 were the most numerous in all genomes. Within F. xylarioides, robusta and coffea strains contained twice as many mimps as the arabica strains, primarily driven by families 1 and 2 (Fig 6). All F. xylarioides genomes and most of the F. oxysporum genomes contained intact impala transposases (Fig 6A), required for mimp activity [35]. A small number of mimps were found in F. proliferatum (also reported in [32]), but no other mimps or impala transposases were found in the genomes of other species in the F. fujikuroi species complex. Genealogy reconstruction showed that mimps in F. xylarioides are intermingled with those from the F. oxysporum species complex, rather than forming separate clades, consistent with either a frequent and ongoing transfer between the 2 species, or a single transfer of multiple diverse copies: multiple clades with strong bootstrap support contain copies from both F. xylarioides and F. oxysporum (shown for mimp family 1, Figs 6B and S11). The alternative hypothesis of ancestral polymorphism would assume complete loss of all these families in all species from the F. fujikuroi complex.

Fig 6 — (A) Miniature *impala (mimp)* elements across the whole-genome are shown as a count per family. They were identified in 6 representative F. *xylarioides* genomes, all F. *oxysporum* formae speciales genomes sequenced in this study and F. *oxysporum* f. sp. *lycopersici* (GCA_000149955.1), each annotated strain denoted by an asterisk. A BLAST search did not return any hits in other genomes shown in the tree. Colour represents the family to which the mimp sequence belongs. The STAG species tree is shown on the left and the tip label corresponds to the presence (blue) or absence (orange) of an intact *impala* transposase open reading frame. The scale bar indicates branch length (substitutions per site). (B) *Fusarium oxysporum* and F. *xylarioides* miniature *impala* (*mimp*) family 1 sequence genealogy. All annotated *mimp* 1 sequences were aligned, with phylogeny represented by an unrooted tree. Each point corresponds to a *mimp* sequence and the scale bar indicates branch length (0.02 substitutions per site). Tree tip labels are shaded by phylogenetic group: arabica, robusta, coffea1, and coffea2 F. *xylarioides* populations; *FOSC*, F. *oxysporum* species complex. Full strain details in Table 2 and expanded tree in S11 Fig. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

Some mimps were observed in large regions identified as putative HTRs above. For example, in HTR 2 and 6 that match F. oxysporum f. sp. lycopersici’s supercontig 51, F. xylarioides arabica contained at least 3 and robusta at least 2 mimps. Some mimps overlapped with- (six10 and six12) or were close to the six7 effector genes (S7 Fig). The coffea scaffolds that matched the region of supercontig 51 found in F. xylarioides arabica also contained mimps, while those matching the region found in F. xylarioides robusta lacked these. Both robusta and coffea scaffolds contained an unclassified TE (“rnd-6 family-1942”) and newly recognised TEs (Fot6, MGR583-like, Yaret1) found on F. oxysporum f. sp. lycopersici’s mobile pathogenic chromosome (identified by [25]). It is possible, however, that the mimps assembled to different coffea scaffolds, as the robusta scaffolds that match supercontig 51 are only 20 kb each.

In addition to mimps, all HTRs were high in TEs. Across the F. xylarioides arabica563 genome, average TE density (excluding simple and low complexity repeats) was 0.06 TEs/bp, while all HTRs contained more transposons (S11 Table) with an average 0.12 across HTRs. HTRs 2, 5, and 6 had the highest overall TE densities at 0.2, 0.11, and 0.12 TE/bp. HTRs 1–2 and 5–6 all contained diverse TEs which were first identified on the mobile pathogenic chromosome of F. oxysporum f. sp. lycopersici by [25] (S10 Fig) and the unclassified TE family (“rnd-6 family-1942”) which we found with mimps (S7 and S10 Figs and S11 Table). Overall, the forty-four 10-kb windows comprising the HTRs were significantly enriched for mimps (total 6 copies) and all classes of TEs known from F. oxysporum mobile chromosomes (total 38): probability of observing as many or more in randomly selected forty-four 10-kb windows from the genome p < 0.001 (randomisation test, 1,000 trials).

An 80-kb Starship mobile element has inserted between highly expressed genes in the arabica genomes

The role of large mobile elements such as Starships in the mobility of effector genes has recently been discovered [15]. A Starship mobile element was identified in all 3 arabica genomes on chromosome 11 and fully corresponds to HTR 4, thus we re-define HTR 4 as a Starship. This Starship was bounded by flanking target site duplications, the 5′ end of the element contained a “captain” gene encoding a predicted tyrosine recombinase with a DUF3435 domain (Pfam accession PF11917) and comprising 22 genes (“cargo”) (Fig 7A).

Fig 7 — (A) Schematic plot showing known *Starship* tyrosine recombinases (captains), and full-length *Starship* elements coloured red, with predicted gene models shown. GC content is indicated with density plots coloured grey below each genome. Predicted gene models and their direction are indicated, with grey shaded areas indicating synteny. Orthologous gene models are present in each *Starship* (BLAST, E-value <10⁻¹⁵⁰); any gaps or differences are due to annotation errors. Up-regulated genes are annotated and named in the key. (B) Maximum likelihood tree showing phylogenetic relationships of F. *xylarioides* tyrosine recombinases searched against the BLAST nonredundant database. Trees are drawn with 1,000 bootstraps and nodes are annotated with bootstrap support. The scale bar indicates branch length (0.02 substitutions per site). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

Multiple sources of evidence suggest that this Starship is an HTR from F. oxysporum. First, this region was independently determined as an HTR with different methods (as HTR 4) and shows high similarity to F. oxysporum detected using BLASTn (S6B Fig). In addition, this Starship/HTR 4 was found to have high TE density, and specifically TE families which were first identified from the mobile pathogenic chromosome of F. oxysporum by Schmidt and colleagues [25] (S10 Fig). Third, the genealogy of the DUF3435 captain gene reveals the F. xylarioides copies are nested within F. oxysporum copies (Fig 7B). The level of divergence between F. xylarioides and the nearest F. oxysporum copy (D FxFo = 2.7%) falls within the lower 0.1% expected across background genes (S14 Table). Interestingly, this Starship appears to occupy a TE island (with an AT basis and high repeat density 0.06 TE/bp) and has significantly up-regulated effector-like genes directly flanking either side: H9Q71 0016025 was one of the top 20% differentially expressed genes and is a nucleartransport factor 2; H9Q71 0016024 is a glycoside hydrolase 43 CAZyme (GH43) predicted by EffectorP as an apoplastic effector; H9Q71 0001940 is a secreted protein not predicted as an effector by EffectorP. Finally, 11 of 12 TEs (>500 bp) within the Starship are single copy, i.e., there are no other copies (BLAST, >90% grade) outside the Starship in the F. xylarioides arabica genomes, except for 1 TE (TE2649, “rnd-6 family-788”) which is 12,257 bp and is also found in HTR 5 on supercontig 2. These different lines of evidence point to a recent acquisition of the Starship by the common ancestor of F. xylarioides arabica genomes.

Discussion

Fusarium xylarioides is genetically differentiated into multiple host-specific populations isolated from different species of coffee. We found that the arabica population shares several highly similar regions with F. oxysporum, including a large mobile Starship element. These regions are at non-homologous locations in the 2 species and lacking in other closely related species or populations, consistent with their acquisition by horizontal transfer. Transcriptome data following infection revealed high expression of putative effector genes, including some only known to F. oxysporum outside of coffee wilt, as well as pectinolytic enzymes. Finally, we found that all F. xylarioides populations share highly similar mimps and other transposable elements with F. oxysporum that have likely transferred between the species and could have facilitated HGT of wider genome regions.

The F. xylarioides arabica and robusta groups were genetically differentiated from each other and from coffea1 and coffea2, which were themselves 2 differentiated populations. The 2 coffea populations overlapped spatially and temporally (coffea 1 was isolated from Guinea and Central African Republic, coffea2 was isolated from Guinea, the Central African Republic, and Cote d’Ivoire), yet they remained genetically distinct, with the robusta population apparently emerging from within coffea 1. Possible causes of genetic differentiation could be the occurrence of distinct clonal lineages, consistent with the congruence between gene genealogies within populations, or reproductive isolation due to pleiotropy between host adaptation and mate choice [36]. The specific coffee tree hosts from which F. xylarioides coffea population strains were isolated were not recorded, so we do not know if these infected a broad range or specific coffee tree species. Reproductive isolation can be a pleiotropic effect of specialisation if mating occurs within the host [36], as only strains able to infect the same trees can mate. In fact, mating appears to occur in the host in F. xylarioides as perithecia are only observed on coffee wilt diseased trees [37], which could have aided genetic differentiation between geographically overlapping populations. Our findings support earlier work based on DNA markers and crossing experiments [38,39], which suggested that F. xylarioides was a species complex containing distinct arabica and robusta populations. Therefore, we confirm the presence of genetically differentiated populations with different host specificity. Future taxonomic work including mating tests could determine whether these populations should be named as separate species.

Several lines of evidence support the acquisition of large HTRs into F. xylarioides populations over an alternative hypothesis of vertical inheritance followed by loss from other lineages within the F. fujikuroi clade. First, a vertical inheritance and loss scenario would require independent losses in at least 6 lineages across the F. fujikuroi phylogeny. Second, the regions display significantly higher sequence identity with their nearest F. oxysporum match than expected based on background genome levels, whereas vertical inheritance and loss would predict sequence identity to fall within the observed distribution of identities for similarly sized genome regions. Third, HTRs are present in different chromosomal locations in F. xylarioides compared to F. oxysporum, in contrast to the high degree of synteny observed across the genome as a whole. Presence in different chromosomal locations is also not consistent with a scenario of introgression at homologous loci following interspecific sexual hybridisation. In theory, horizontal transfers can occur through transposon-assisted mechanisms or by transfer of an entire chromosome [21]. While we have no direct evidence for mechanisms here, either seem plausible. Filamentous fungi, including the genus Fusarium, species are known to form anastomoses (somatic fusions between mycelia) which may facilitate the transfer of genetic material by whole chromosomes or potentially by other mobile elements [40]. Indeed, hyphal fusions have been shown to occur within the arabica and robusta F. xylarioides populations [41]. The finding of specific shared transposons between F. xylarioides and F. oxysporum, and in particular a Starship element, known in other species to be able to transfer horizontally [15], would be consistent with a role for mobile elements in horizontal transfer.

Each of the HTRs appears to represent a single transfer event into the most recent common ancestor of the F. xylarioides population (or populations for HTR 2 and 5) from F. oxysporum: locations of HTRs are conserved within each population, and there is low sequence variation compared to largescale variation observed among F. oxysporum copies. Transfer of pathogenicity genes via accessory chromosomes was first reported in F. oxysporum f. sp. lycopersici, with horizontal transfer of its entire mobile pathogenic chromosome to 2 previously nonpathogenic strains [21]. Several additional cases have since been reported, each with an association with mimps in transposition [25,32]. In F. xylarioides, a whole mobile chromosome may have been acquired by a common ancestor to the arabica and robusta populations, with later divergence and differential loss of genomic regions and integration of different parts. Loss of regions that are no longer useful to the organism is well reported [42]. Alternatively, only parts of the F. oxysporum chromosome may have been transferred into F. xylarioides, with different parts in arabica and robusta populations.

The association of mimps with the HTRs sheds further light on how transposable elements might interact with host specialisation. Overall, we found 4 putative HTRs with TEs that have been linked to horizontal transfer and the generation of effector variability in F. oxysporum. TEs could generate effector variation either via direct transposition (for example, [43] and as in HTRs 1–3 and 5) or via increasing the potential for recombination events (including ectopic recombination and genome loss). One HTR, the Starship/HTR 4, lacked mimps and the transposons identified from the F. oxysporum mobile pathogenic chromosome, but contained other single-copy TEs. We found few mimps in the F. fujikuroi complex species (see also [32]), whereas arabica and robusta F. xylarioides populations contained different numbers of mimps in total as well as in their HTRs, consistent with different histories of transfer and/or expansion in each population. In F. oxysporum, six10, six12, and six7 make up supercontig 51 and are flanked by mimps [25]. In F. xylarioides, different pieces of supercontig 51 are present in the arabica and coffea genomes, versus the robusta and coffea genomes. The SIX genes in arabica were flanked by mimps, while the robusta region also contained mimps and other class II TEs, as well as other genes involved in pathogenicity. A study of 2 F. oxysporum f. sp. cubense transcriptomes found that both lacked mobile chromosomes of F. oxysporum f. sp. lycopersici, yet still showed differing presence and expression of SIX genes [44]. In F. oxysporum, all effector genes contain a mimp in the promoter region [25], yet in F. xylarioides, this was only the case for the 3 SIX genes. In addition, the fact that the mimp-overlapping SIX genes are only found in the arabica and coffea populations (and not robusta nor any F. fujikuroi complex species) suggests a horizontal transfer origin.

To our knowledge, such patterns of horizontal transfer of pathogenicity between species complexes has not been reported in Fusarium before, while introgression has [45]. We found that the closest match to one of our putative HTRs (HTR 4/Starship) was F. oxysporum 509, sequenced in this study and isolated from an arabica coffee plant in the 1970s in Tanzania, which is spatially and temporally within the robusta coffee wilt disease range. This suggests these patterns of horizontal transfer could be more commonplace than thought. The close phylogenetic relationship between CAZyme gene families from F. oxysporum and the arabica strains provides further evidence of horizontal transfer of gene types that are important in vascular wilt infection. Such horizontal transfer could have occurred in a shared niche, as both F. xylarioides and F. oxysporum are soil-borne pathogens and have been isolated from the roots and wood of coffee wilt-diseased trees in Ethiopia and central Africa [46], as well as from banana roots in an intercropped field in Uganda [47], and on which F. oxysporum f. sp. cubense is widespread [48]. These transfers may have driven the evolution of new host-specific populations and therefore could have facilitated disease emergence and subsequent re-emergence.

Recent comparative genomics studies have elucidated the role of these changes in domesticated fungi and plant pathogens, including a host jump in Phytophthora infestans followed by adaptive specialisation [49], transfer of a 14 kb Starship encoding the ToxA virulence protein and abundant transposons between 3 wheat fungal pathogens [15], Starship activity in the human pathogen Aspergillus fumigatus which promotes strain heterogeneity [18], and the transfer of regions that allow adaptation to cheese environments in distantly related species [50]. Furthermore, transfer of pathogenicity into a previously nonpathogenic strain in F. oxysporum has been demonstrated experimentally [21]. These examples, as well as cross-kingdom transfers (e.g., [51]), demonstrate the occurrence of HTRs across many distantly related branches of the eukaryotic tree of life and their importance for shaping eukaryotic evolution [52]. Previously, the analysis of historic genetic material involved the use of highly degraded museum specimens and molecular markers and target probes in a small number of genes [53]. However, fungal culture collections contain tens of thousands of fungal isolates collected over the past century and preserved in a living state [54]. Connecting collections with new-omics studies provides a powerful way to investigate the role of HTRs in historic fungal outbreaks.

Materials and methods

Strain details

All strains were from the CABI culture collection (Egham, United Kingdom), with IMI 507113 transferred from the Belgian coordinated collection of microorganisms from the Université Catholique de Louvain (Louvain-la-Neuve, Belgium) (under MUCL 47066/ MNHN 709) and IMI 507035 and 507038 transferred from the Agricultural Research Service culture collection (Illinois, United States of America) (with 507035 under NRRL 25804/BBA 62721/CBS 749.49; and 507038 under NRRL 37019/FRC L96/BBA 62458). The 11 additional Fusarium strains were selected for sequencing based on the presence of certain putative effector genes described in [28], which was established by Sanger sequences of PCR data [55].

Long-read reference genome sequencing and assembly

The F. xylarioides strain IMI 389563 (hereafter called “arabica563”) from the arabica host-specific strain group of coffee wilt disease was used for long-read sequencing, and is available from the CABI-IMI culture collection (CABI, Egham, UK). Arabica563 mycelium was grown in GYM broth for 7 days at 25°C and 175 RPM. High-quality high molecular weight genomic DNA was extracted using the chloroform extraction method in [56]. Specifically, 1 to 2 g of mycelium were ground in liquid nitrogen using a lysis buffer containing sodium metabisulfite, 2 M Tris, 500 mM EDTA, 5 M NaCl, and 5% Sarcosyl. The gDNA was extracted from the cell lysate using chloroform, and precipitated using isopropanol and 70% ethanol. The gDNA was eluted in 500 μl TE (Tris EDTA) buffer and incubated overnight at 4°C. RNA contaminants were removed using RNAse A (10 mg/ml) and gDNA was purified using Agencourt AMPure XP beads (Thermo Fisher). The gDNA quantity was determined using a NanoDrop 2000 (Thermo Fisher) and its integrity was determined by electrophoresis with a 0.5% agarose gel. gDNA (1 μg) was diluted in nuclease-free water to a final volume of 47 μl. Half of the DNA was then sheared into 20 kb fragments with a COVARIS g-TUBE (Covaris, Woburn, Massachusetts, USA), centrifuging at 4,200 rpm for 1 min to increase DNA throughput. The sheared DNA was converted into Oxford Nanopore libraries using the SQK-LSK110 library kit and the short fragment buffer (Oxford Nanopore Technologies, United Kingdom). Library prep followed the standard manufacturer’s protocol apart from the final elution, which was done for 30 min at 37°C. The libraries were sequenced using a SpotON R9.4.1 FLO-MIN106 flowcell for 48 h.

Nucleotide bases were called from the raw sequence data using the super high accuracy model (SUP) of the GPU version of Guppy v6.0.1 (config file: dna r9.4.1 450bps sup.cfg). Base-called reads were assembled into contigs by Flye v2.9 with “-nano-hq” parameters [57]. The Flye assembly was error-corrected using both Nanopore and Illumina data with 1 round of medaka v1.4.4 (https://github.com/nanoporetech/medaka) and 1 round of (Hapo-G) polishing [58]. Contig completeness was visualised in Tapestry v1.0.0 [59], including the presence of the terminal telomere repeat sequence TTAGGG, common to eukaryotes. The presence of core Fusarium chromosmes, shared by close and distantly related sister species was assessed by mapping the F. xylarioides arabica563 reference to the F. verticillioides chromosomal assembly. First, the 2 genomes were aligned using nucmer [60], before visualising the alignment in DotPrep.py (https://github.com/marianattestad/dot). Those contigs in F. xylarioides which match to multiple F. verticillioides chromosomes are described as supercontig 1 and supercontig 2.

Illumina sequencing and assembly

An additional 11 Fusarium genomes were selected for Illumina whole-genome sequencing: 5 F. xylarioides, 5 F. oxysporum, and 1 F. solani (Table 1), see Strain details. Of the 5 F. xylarioides strains, 3 were from the pre-1970s outbreak, which was observed to infect multiple Coffea species and which we call “coffea” (coffea035, coffea113, coffea676), and the remaining 2 were the earliest isolates available in the collections for each of the arabica and robusta groups (both isolated from around 1970, arabica038 and robusta268).

Strain morphologies were verified and grown in GYM broth following the same protocol as above. DNA was extracted from 250 to 500 mg of washed mycelium following the NucleoSpin Soil Mini Kit (Macherey Nagel, Germany) standard protocol. For each strain, a single library was prepared with the Illumina DNA PCR-Free Prep and sequenced with the Illumina Novaseq 6000 with 2× 150 bp reads. Low-quality bases (Phred score <20) and adapters (stringency 4) were removed using TrimGalore 0.6.0 by Cutadapt (https://github.com/FelixKrueger/TrimGalore, [61]). Overlapping paired-end reads were stitched together using FLASH 1.2.11 [62] and de novo assembled with MEGAHIT 1.2.9 [63,64]. Assembly metrics including BUSCO scores were computed using QUAST 5.2.0 (Table 1) [65].

Each F. xylarioides MEGAHIT assembly from this study, as well as those assembled previously [28] and the publicly available genomes robusta925 [66] and robustaL0394 [67], was mapped to the reference genome assembly using RagTag v2.1.0 homology-based scaffolding [68]. The parameter “-u” identified unmapped (and therefore absent from arabica563) scaffolds. The F. oxysporum and F. solani MEGAHIT assemblies were mapped to their closest and least fragmented publicly available assembly (Table 2). These RagTag assemblies were used to predict TEs (see later section).

Table 2. Data sources used to annotate Fusarium oxysporum and F. solani genomes sequenced in this study.

The columns describe: species, CABI IMI strain number, RNA-seq data used as evidence in gene prediction, expressed sequence tag (EST) data used as evidence in gene prediction, which publicly available assembly the genome was scaffolded to using RagTag for annotation of repeats. Strain bio-geographic details are in S1 Table.

Species	Strain number (IMI)	RNA-seq data	Expressed sequence tag data	Genome assembly for scaffolding
F. oxysporum f. sp. cubense	141109	ERR10015931	F. oxysporum f.sp. cubense II5 v2.0 [72]	GCA_005930515.1
F. oxysporum f. sp. coffea	244509	SRR10123445	F. oxysporum f.sp. cubense II5 v2.0 [73]	GCA_000149955.1
F. oxysporum f. sp. vasinfectum	292248	SRR12855423	F. oxysporum f.sp. cubense II5 v2.0 [73]	GCA_000260175.2
F. oxysporum f. sp. raphani	337541	SRR14318405	F. oxysporum f.sp. cubense II5 v2.0 [73]	GCA_019157275.1
F. oxysporum f. sp. pisi	500221	SRR12855423	F. oxysporum f. sp. pisi F23 v1.0	GCA_000260075.2
F. solani	392280	SRR19182467	F. solani FFSC 5 v1.0 [73]	GCA_020744495.1

Open in a new tab

The raw sequence data and assembled genomes have been deposited in the International Nucleotide Sequence Database Collaboration (INSDC) database under the BioProject accession number PRJNA1043203.

RNA sequencing to aid annotation of long- and short-read assemblies

The same arabica563 strain used for long-read sequencing was grown under the same culture conditions described above. Mycelia were harvested and total RNA extracted using the standard protocol of the RNeasy Mini kit (Qiagen, Hilden, Germany). The RNA quality assessment (Qubit and Fragment Analyzer), library preparation and Next Generation Sequencing (60 M paired-end reads per sample) were performed by Genewiz/Azenta (Frankfurt, Germany). See RNA extraction section for QC steps. Transcript sequences were de novo assembled using Trinity v2.13.2 with the genome-free method [69].

Gene prediction and annotation

Protein-encoding genes were predicted in all short-read genomes using the Funannotate Eukaryotic Genome Annotation Pipeline v1.8.11 [70]. All 13 F. xylarioides genomes were annotated (Table 1), including re-annotations for the 6 genomes from [28]. Before annotation, repetitive contigs were removed using “funannotate clean” with default parameters and repetitive elements within the assemblies were soft masked using RepeatModeler and RepeatMasker with “funannotate mask” and F. oxysporum parameters.

The trimmed RNA-seq reads and the Trinity assembly were aligned to the arabica563 MEGAHIT assembly using the parameters “funannotate train --jaccard_clip” which runs PASA to model the gene structures. The PASA gene models were parsed to run Augustus, snap, GeneMark, and GlimmerHMM, with the predictions used to run CodingQuarry and the Funannotate Evidence Modeler. The complete set of arabica563 output files were then used to predict protein encoding genes on the remaining MEGAHIT genome assemblies using “funannotate predict -p predict results/fusarium xylarioides 389563.parameters.json.” The arabica563 MEGAHIT assembly was selected over the reference genome to avoid issues with gene prediction on assemblies of different quality when the parameters were applied to the remaining F. xylarioides MEGAHIT assemblies. The gene model predictions were refined (e.g., correcting intron-exon boundaries) and untranscribed regions annotated by re-aligning the RNA-seq data for each genome using “funannotate update” on the output files from “funannotate predict.” The annotations were then mapped to the arabica563 reference genome and long-read RagTag assemblies using Liftoff [71] with default parameters.

Other Fusarium genomes

The same steps were followed for the 5 F. oxysporum genomes and the F. solani genome, using different sources of evidence for the different species. RNA-Seq reads for the closest formae speciales match were downloaded from the sequence read archive (see Table 1 for accession numbers) and aligned to the corresponding assembly using “bbmap.sh”), and expressed sequence tag data for the closest f. sp. match from JGI Mycocosm were used to provide evidence to “funannotate predict.” The expressed sequence tag data were produced by the US Department of Energy Joint Genome Institute (https://www.jgi.doe.gov/) in collaboration with the user community. Evidence sources are listed in Table 2. The same RNA-Seq data was then re-aligned through “funannotate update” as in the section above.

Functional annotation of the protein-coding genes for all genomes was completed with annotations from PFAM, InterPro, UniProtKB, MEROPS, CAZyme, and GO Ontology through “funannotate annotate.” InterProScan v5.56–89.0 was run locally using the parameters “-goterms -iprlookup” and phobius was run through the web browser (https://phobius.sbc.su.se/). All proteins lacking significant hits were annotated as hypothetical proteins.

Transposable elements and repeat annotations

Repeatss and transposable elements were identified from the nucleotide assemblies. A custom repeat library was constructed using all Fusarium RagTag assemblies in RepeatModeler v2.0.3 with the parameter “-LTRStruct” [74]. Added to this library were TEs known to be involved in F. oxysporum transfer of pathogenicity between strains: impalas, a new set of TEs described by [25], as well as manually curated mimps. Sequences for mimp families 1–4 were downloaded from NCBI (accession numbers: AF076624.1, AF076625.1, EU833100.1, and EU833101.1). Sequences for the more divergent mimp families mimp5, mimp6, mn2, mn3, mn4, mn8, and mn14 are from [35]. Manual curation followed the protocol described by [75] and involved: identifying conserved domains with sequence homology to these TEs using BLASTn; only retaining the best BLAST hits (≥ 50% query length and ≥ 80% identity) for each TE family; extracting and aligning nucleotide sequences for TE families including an additional 500 bp up- and down-stream; trimming alignments to the TIRs at their start and end and removing insertions using T-COFFEE [76]. This generated consensus sequences for each mimp family that were added to the RepeatModeler custom library and used to call TEs and repeats from genome scaffolds using RepeatMasker v4.1.2 with the parameters “-lib $LIB -gff -no_is” [77]. The presence of the impala transposase ORF from F. oxysporum f. sp. melonis was found by a BLAST search “tblastn -evalue 1e^-5” using accession AAB33090.2. See Data Availability for custom repeat library and analysis scripts.

The mimp family 1 phylogenetic tree was constructed by aligning all mimp family 1 sequences that were annotated with MAFFT v7.309 [78,79] in Geneious v9.1.8 (https://www.geneious.com). We used Geneious to select the resulting consensus sequence and used BLASTn [80] in Geneious to search for copies in the public genomes used in this study (S2) as well as the NCBI “nr” database. Copies were only selected if they were >80% in length and identity and were a species other than F. xylarioides and F. oxysporum: just 1 additional species was included (F proliferatum). Sequences were re-aligned using MAFFT in Geneious, with maximum likelihood trees constructed in IQ-Tree2 v2.3.5 with 1,000 bootstrap replicates, with the best-fit model automatically selected by ModelFinder and using the ultra-fast bootstrap parameter [81]. Phylogenetic relationships were visualised with ggtree.

Orthologous gene groups, species tree, and genetic structure

Whole-genome similarity between the protein-encoding genes was assessed using NSimScan, part of the QSimScan package [82]. OrthoFinder v2.5.4 was used to determine Fusarium orthologous groups between the genomes sequenced in this study (Table 1) and published genomes detailed in S2 Table. OrthoFinder infers gene trees for all orthogroups and uses gene duplication events via the STRIDE algorithm to root them [83]. Specifically, unrooted gene trees are built for each orthogroup using DendroBLAST [84] and trees are resolved using the OrthoFinder hybrid species-overlap/duplication-loss coalescent model [85]. The “supergene” species tree was built and rooted using STAG in OrthoFinder [86,87] with 3,544 orthogroups in which every strain has at least 1 gene copy. Resolved monophyletic species clades were identified using tr2 software [88]. An alternative “supermatrix” species tree was reconstructed by concatenated alignment of all 1,685 single-gene orthologs using seqkit [89], creation of a partition file and maximum likelihood analysis using IQ-Tree2 v2.3.5, the partition file, the best-fit model automatically selected by ModelFinder and the ultra-fast bootstrap [81].

SNPs were called against the arabica563 reference genome for each set of F. xylarioides reads, with reads for robusta925 downloaded from the Sequence Read Archive (accession PRJNA508603, [66]) and robustaL0394 shared by [67]. SNPs were called using GATK v4.1.2.0 “HaplotypeCaller -ERC GVCF,” which provides 1 genomic variant call format (gVCF) output file per strain. Using Snakemake v5.3.0, gVCFs were combined using GATK “CombineGVCFs,” genotypes were combined with “GATKGenotypeGVCFs,” “SelectVariants -select-type SNP” selected the SNPs which were then filtered with “Variant-Filtration QUAL<30, DP<10, QD<2.0, FS>60.0, MQ<40.0, SOR>3.0, QRankSum<-12.5, ReadPosRankSum<-8.0.” The SNP data set was used in Neighbour-Net analysis using SplitsTree App (version 6.3.37, [90]) to infer population structure. Absolute nucleotide divergence d_xy was calculated using the R package Popgenome [91].

Identification of Fusarium xylarioides population-specific regions and horizontal transfers

Fusarium xylarioides population-specific regions were identified by mapping each F. xylarioides genome to the arabica563 reference using minimap2 with the parameters “-cx asm10 --secondary=no --cs” [92]. The PAF file was converted to bed using “paftools.js splice2bed,” and then coverage was calculated in 100-bp windows using “bedtools coverage.” Finally, coverage was calculated in contiguous 10-kb windows by taking the average of 100 × 100 bp windows using “bedtools map -c 7 -o mean.” Mean mapped coverage was calculated for each population group (arabica, robusta, coffea1, and coffea2) for each 10-kb window. Regions were identified as: arabica-specific if they were present (mean similarity >0.5) in arabica and absent (mean similarity = 0) from all other population groups (48 windows); arabica- and coffea1-specific if the arabica and coffea1 means >0.5 and the coffea2 and robusta means <0.2 (22 windows); arabica- and coffea2-specific if the arabica and coffea2 means >0.5 and the coffea1 and robusta means <0.2 (21 windows); arabica-, coffea1-, and coffea2-specific if these means were >0.5 and robusta was <0.2 (9 windows); and arabica- and robusta-specific if arabica and robusta means were >0.5 and coffea1 and coffea2 were <0.2 (15 windows). See Data Availability for analysis scripts and data files.

Putative horizontal transfers were identified using BLASTn [80] in Geneious v9.1.8 (https://www.geneious.co) to search for these regions in their entirety in the NCBI “nr” database. Only those hits with a grade >50% (grade combines query coverage, E-value and identity values for each hit with weights 0.5, 0.25, and 0.25, respectively; all E-values >1e^-133) were selected, namely the longest, highest identity hits. We describe as putative HTRs where such hits were only present in, or most closely related between, F. xylarioides and F. oxysporum. Selected BLAST hits for each putative HTR were aligned using MAFFT v7.309 [78,79] in Geneious and phylogenetic trees reconstructed for each alignment using IQ-Tree2 v2.3.5 with 1,000 replicates of ultra-fast bootstrap branch support [81,93]. This resulted in 5 putative HTRs identified in the F. xylarioides arabica563 reference genome on chromosomes 1, 5, 8, 11, and supercontig 2. After discovering that HTR 2 matched a chromosomal subregion “supercontig 51,” a region on the F. oxysporum mobile pathogenic chromosome which is enriched in effector genes, a BLAST match of each F. oxysporum chromosomal subregion against the F. xylarioides genomes returned an additional HTR in the robusta genomes (HTR 6). Synteny between the F. xylarioides putative HTRs and accessory regions in 2 long-read F. oxysporum genomes (F. oxysporum f. sp. lycopersici and F. oxysporum f. sp. cubense were visualised by first aligning each to the arabica563 reference using MUMmer’s nucmer [60], then plotting the alignment in Rideogram [94]). The remaining and non-horizontally transferred F. xylarioides population-specific regions were found across the F. fujikuroi, F. oxysporum, F. graminearum, and other Fusarium species complexes. See Data Availability for analysis scripts and data files. The putative HTR bed file and nucleotide sequences are available as S1 and S2 Figs.

Starships were identified using starfish v1.0.0 [95] in conjunction with metaeuk [96], hmmer [97], bedtools [98], blastn [80], mummer4 [60], cnef [99], sourmash [100], mcl [101], and visualised using gggenomes [102], mafft [78], and minimap2 [92]. In total, starfish identified 15 putative elements over 19 genomic regions. By manually inspecting each, we determined that one is a Starship and the remainder dead and/or degraded. We used BLASTp on NCBI (https://blast.ncbi.nlm.nih.gov/) to search for matching hits to the 3 identical F. xylarioides DUF3435 “captain” genes, and aligned using MAFFT in Geneious those hits with a score >1,300 (score is the sum of alignment scores of all segments from the same subject sequence, comprising query coverage and E-value). The phylogenetic tree was drawn using PHYML in Geneious with bootstrap branch support calculated with 100 replicates. The Starship comprises 47 simple and low complexity repeats, and TEs. We took TEs with length >500 bp (n = 12) and used BLAST in Geneious to look for other copies in the F. xylarioides genomes. TEs were counted as present with a >90% grade match to the F. xylarioides arabica563 copies in the Starship. The predicted gene annotations and functions within and flanking the Starship are described in S3 File.

Statistical evaluation of putative horizontal transfers

For each putative HTR, we calculated the average sequence identity (across 10 kb windows) to its closest match in F. oxysporum (defined as I_Fo) and F. xylarioides (excluding F. xylarioides strains in which the HTR was determined to be present, I_Fx). We then used I_Fx - I_Fo as a metric of the identity to F. oxysporum relative to the other F. xylarioides populations: a negative value indicates greater similarity to F. oxysporum. For each HTR, we then calculated an equivalent metric for all similarly sized regions across the genome as a whole, and took the proportion of regions showing a value ≤ that observed for each HTR as the chance of observing the HTR value, given the empirical background distribution. For each HTR and for the captain gene of the Starship in HTR 4, we ran an analogous phylogenetic test. First, we calculated the mean patristic pairwise divergence between F. xylarioides sequences and their closest F. oxysporum match (D_FxFo). Then, we calculated the same metric across all single-copy ortholog gene trees as a sample of expected background levels of divergence. The percentile of the observed value for each HTR (and the captain gene) was then used as an estimate of the chance of observing the same pairwise genetic divergence to F. oxysporum across the genome.

The putative HTRs were associated with mimps and other TEs using a randomisation trial. There are forty-four 10-kb windows which comprise the putative HTRs, which contain mimps (total 6 copies) and all classes of TEs known from F. oxysporum mobile chromosomes (total 38 copies). We randomly selected the same number of windows from the F. xylarioides arabica563 genome and counted how many times they contained more copies of mimps or other TE classes. We repeated this 1,000 times; there were no instances where the randomly selected windows contained more mimps or other TEs than the putative HTRs (p < 0.001).

Infection assays and controls

The F. xylarioides strains IMI 389563 and IMI 375908 (“arabica563” and “arabica908”) from the arabica host-specific coffee wilt disease population from the CABI-IMI culture collection (CABI, Egham, UK) were used. Strains were grown on synthetic low nutrient agar at 25°C, with the spores verified after 5 days using lactophenol cotton blue staining under a compound microscope following [103]. The cultures were harvested with 1 drop of Tween20 and in 10 ml sterile water. Two plates were combined to make each fungal inoculum before conidial concentrations were measured with a haemocytometer and adjusted to 10⁶ spores/ml. A third control inoculum was made in the same way using sterile distilled water added to fresh synthetic low nutrient agar that had not been inoculated. Each inoculum type had 4 technical replicates.

Coffea arabica plants were purchased online from Gardeners Dream (https://www.gardenersdream.co.uk/) and left for 12 weeks to acclimate and grow in a controlled growth room at 25°C with a 12:12 hour light:dark cycle with 120 μmol/m²/s wavelength and 50%:65% day:night humidity.

We compared gene expression of F. xylarioides strains infecting coffee plants and grown in axenic culture, as well as between the inoculated and control-inoculated coffee plants. Plants were infected with 10 μm through stem wounding [39], using a sterile needle at the first green (i.e., non-woody) internode. Each inoculum type (arabica563, arabica908, control inoculum) was used to inoculate 4 plants, making up 16 replicates in total per inoculum type. Plants were grouped by inoculum type and sealed in propagator trays with parafilm to make 4 trays containing 4 plants for each inoculum type. Tray location in the growth chamber was randomised and the plants were grown under shade at 25°C with a 12:12 hour light:dark cycle in a growth room until at least 50% of infected plants showed wilting, yellowing or early leaf senescence as symptoms of infection. The plants were harvested 98 days after inoculation.

Axenic cultures of arabica563 and arabica908 isolates were grown in GYM broth at 25°C with shaking at 175 RPM for 7 days. As with the in planta inocula, the in axenic culture samples included 4 biological replicates (arabica563 1–4 and arabica908 1–4).

RNA extraction, sequencing, and quality control

All aboveground plant material was collected and flash-frozen in liquid nitrogen. The biological replicates were ground separately in liquid nitrogen with PVPP (polyvinyl polypyrrolidone) in a mortar and pestle. Approximately 5 g of each biological replicate was pooled in a 1.5 ml polypropylene tube, precooled in liquid nitrogen, before vortexing and re-freezing.

Total RNA was extracted using the standard protocol of the RNeasy Mini kit (Qiagen, Germany) and following the same methods described above. In total, 20 samples were extracted with 4 technical replicates for each treatment: in planta arabica563-infected coffee plants; in vitro culture arabica563; in planta arabica908-infected coffee plants; in vitro culture arabica908; control (mock-inoculated with water) coffee plants. RNA was quantified using a NanoDrop 2000 (Thermo Fisher), with the RNA quality assessment (Qubit and Fragment Analyzer), library preparation and Next Generation Sequencing (60 M paired-end reads per sample) performed by Genewiz/Azenta (Frankfurt, Germany). Raw sequenced reads were quality- and adapter trimmed using Trimmomatic v0.39 (parameters: ILLUMINACLIP:$TRIMMOMATIC DIR/adapters/TruSeq3PE.fa:2:30:10 SLIDINGWINDOW:4:5 LEADING:5 TRAILING:5 MINLEN:25) [104]. Reads were quality checked with FASTQC v0.11.2 [105] and retained where both pairs were trimmed. To remove ribosomal RNA sequences, reads were mapped to the SILVA rRNA database using BBTools (sourceforge.net/projects/bbmap/—Bushnell B.) “bbmap” with the parameter “outu = filtered R*.fq.gz” to keep only the unmapped reads. Finally, the reads were repaired using BBTools “bbmap” “repair.sh” to re-pair reads that became disordered or had their pair eliminated. These QC steps removed 61 GB of data, see S4 File. Before using the data to quantify gene expression, reads were mapped using “bbmap.sh” to the gene annotations of arabica563 and arabica908 and from C. arabica (GCA_003713225.1, [106]). These mapped read libraries were used for all next steps.

Differential expression analysis

Transcript quantification was performed three times, using the Funannotate gene annotations as the target transcriptomes in a genome-free way using Salmon v1.5.2 [107] on the mapped read libraries: in vitro culture arabica563 and in planta arabica563; in vitro culture arabica908 and in planta arabica908; infected C. arabica plants and control-inoculated C. arabica plants. The relationships within and between biological replicates for each data set were visually examined using the “PtR” script in the Trinity toolkit [69,108]. Each transcript count matrix was tested for differential expression using DESeq2 [109], which uses negative binomial generalised linear models to test for statistical significance. Differential expression was also calculated using edgeR [110], which gave similar results, and limma/voom [111] which handles biases in RNA-seq data differently [112] and was too conservative with the read numbers from the in planta samples for this experiment. The Benjamini–Hochberg method [113] was used to adjust P-values for multiple testing to control the false discovery rate (FDR) and strict significance thresholds were applied, with P-values <0.001 and >4-fold to define a set of differentially expressed genes. The top 20% differentially expressed genes were classified as the “most differentially expressed”: 272 genes in arabica563; and 177 in arabica908.

Functional annotation analysis

Gene Ontology and InterProScan classifications

Arabica coffee gene annotations (accession: GCF 003713225.1) were assigned InterProScan and Gene Ontology classifications using “interproscan.sh -iprlookup–goterms.”

EffectorP

A new set of putative effector proteins were identified using EffectorP 3.0 [34] on secreted proteins which were significantly differentially expressed in both fungal strains (total differentially expressed secreted proteins: 224 in arabica908 and 380 in arabica563). Only secreted proteins significantly up-regulated in both strains were described as effector-like.

Carbohydrate-active enzymes

Carbohydrate Active EnZymes (CAZy) enzymes are grouped by family in the CAZy database [114]. CAZyme-encoding differentially expressed genes were identified in 2 ways, by comparing the proportion of differentially expressed genes to the genome and taking those where >50% of the genes were significantly differentially expressed in both strains, and by identifying enriched gene ontology terms with a cell-wall or pectin-degrading enzymatic function.

Data analysis and presentation

All analyses were performed in R using R Statistical Software (v4.2.1; R Core Team) [115] and in Linux run on the Imperial College Research Computing Service HPC facility (DOI: 10.14469/hpc/2232).

The following R packages were used in analyses and to make figures, all using the latest versions: tidyverse [116], cowplot [117], ggrepel [118], wes anderson [119], pheatmap [120], ggtree [121], treeio [122], aplot (https://cran.r-project.org/package=aplot - Yu G.), RIdeogram [94], circlize [123], ape [124], viridis [125], pals [126], pafr [127], and gggenomes [102].

Supporting information

S1 Fig. The Fusarium xylarioides and F. phyllophylum clade of the supermatrix species tree based on 1,685 concatenated single-copy orthogroups.

Annotated branches indicate bootstrap branch support values. The scale bar indicates branch length (0.001 substitutions per site). Fx, F. xylarioides. Full strain details in S1 and S2 Tables. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s001.tiff^{(49.9MB, tiff)}

S2 Fig. STAG supergene species tree based on 3,544 orthogroups in which every strain has at least 1 gene copy.

Support values show the proportion of gene trees that are congruent with the focal node and blue boxes indicate resolved monophyletic species clades identified by multilocus species delimitation using tr2 (see Methods). Shaded boxes refer to the F. fujikuroi (yellow) and F. oxysporum (orange) species complexes. Scale bar indicates 0.02 substitutions per site. The F. xylarioides strain types arabica, robusta, and coffea are labelled. Fx, F. xylarioides; Fo, F. oxysporum. Full strain details in S1 and S2 Tables. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s002.tiff^{(64.1MB, tiff)}

S3 Fig. Whole-genome similarity between Fusarium oxysporum and Fusarium fujikuroi species complexes.

Cells indicate with colour the nucleotide similarity across all predicted coding sequences between the strains on the x and y axes. Shaded boxes refer to: the Fusarium oxysporum species complex, orange; and the Fusarium fujikuroi species complex, yellow, of which Fusarium xylarioides is a member. Arabica, robusta, coffea1, and coffea2 genomes all belong to Fusarium xylarioides. Excluding F. xylarioides, one genome is shown for each species/formae speciales with strain details in S2 Table. Fo, F. oxysporum. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s003.tiff^{(24.8MB, tiff)}

S4 Fig. Gene sharing across the Fusarium xylarioides host-specific populations.

Orthogroups shared between F. xylarioides arabica, robusta, and the 2 coffea clusters. Drawn 16,006 (excluding 10,567 that were absent from F. xylarioides.

(TIFF)

pbio.3002480.s004.tiff^{(11.1MB, tiff)}

S5 Fig

Whole-genome alignments (A) Fusarium verticillioides chromosomal assembly mapped against the Fusarium xylarioides arabica563 contig assembly and (B) Fusarium xylarioides robusta254 mapped against the Fusarium xylarioides arabica563 contig assembly. (A) Each bar represents one of the 16 arabica563 contigs >100 kb in length that are also present in Fusarium verticillioides, with contig size and position on the x axis. Arabica563 contig numbers are annotated in-line with the bars. Shaded regions indicate the presence of syntenic Fusarium verticillioides chromosomes, the colour of which identifies the specific mapped chromosome. Unshaded regions indicate those found only in arabica563, and arabica563 contig 16 is absent from Fusarium verticillioides. Fusarium verticillioides chromosomes CM010967.1-CM010977.1 correspond to 1–11 in the genome accession GCA_000149555.1. (B) Each bar represents one of the 16 Fusarium xylarioides arabica563 contigs that are also present in Fusarium xylarioides robusta254, with contig size and position on the x axis. Shaded regions indicate the presence of syntenic Fusarium xylarioides robusta254 contigs, the colour of which identifies the specific mapped contig from Fusarium xylarioides robusta254. Unshaded regions indicate contigs found only in Fusarium xylarioides arabica563 that are absent in Fusarium xylarioides robusta254. Telomeric repeats are annotated as blue diamonds on their corresponding scaffolds.

(TIFF)

pbio.3002480.s005.tiff^{(68.8MB, tiff)}

S6 Fig. PHYML phylogenetic trees for HTRs 3, 4, 5, and 6.

All trees are drawn with 100 bootstraps, branch labels reflect bootstrap support values. Tip labels correspond to NCBI GenBank accession numbers or F. xylarioides chromosmes and are shaded by their F. xylarioides population or species complex. Scale bar shows genetic distance. A HTR 3 on chromosome 8 B HTR 4 on chromosome 11 C HTR 5 on supercontig 2 D HTR 6 in robusta, coffea1 and coffea2 genomes. FFC, F. fujikuroi complex; FOSC, F. oxysporum species complex. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s006.tiff^{(92.2MB, tiff)}

S7 Fig. Different parts of Fusarium oxysporum f. sp. lycopersici supercontig 51 is shared with Fusarium xylarioides populations.

A region of supercontig 51 in Fusarium oxysporum f. sp. lycopersici chromosome 14 (NC 030999.1, middle bar) is shared which matches to HTR 2 on arabica563 chromosome 5 (bottom plot, grey rectangle indicates the region which is shared). The shared genes in this 4-kb region are all secreted in xylem (SIX) effector genes, which are shaded according to the legend and labelled. A different part of supercontig 51 is shared with F. xylarioides robusta genomes and matches to a scaffold which is not in the arabica563 reference genome. Repeats of interest are shaded according to the legend and unclassified repeats which occur once are labelled “rnd.” The locations for six12 and six7 are from [25]. Figure drawn to scale. (B) Heatmap showing presence of orthologous groups found in the region found in F. xylarioides robusta from (A), compared with Fusarium phylogeny. Present genes are shaded in dark and all share >96% identity. The phylogenetic species tree was created following S3 Fig. Full strain details in Tables S2 and 1. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s007.tiff^{(79.1MB, tiff)}

S8 Fig. Putative horizontal transfer 1 (HTR 1) between Fusarium oxysporum and Fusarium xylarioides.

(A) HTR 1 on chromosome 1 in Fusarium xylarioides arabica563 (544 kb—739 kb, bottom plot) matches 2 regions in Fusarium oxysporum f. sp. lycopersici chromosome 3 (NC 030988, 4.7 Mb—5.6 Mb, top plot) in 2 pieces, large (200 kb) and small (20 kb). Highly similar genes shared by the 2 species are shaded in purple in top and bottom plots, a subset are labelled by Fusarium oxysporum gene locus tag, and non-matching genes found in the shared region are shaded grey. Genes flanking the shared region in Fusarium oxysporum are shaded yellow, and those which are found in Fusarium xylarioides (BLAST length >80%) have a lower sequence identity and are absent from contig 13. (B) A DendroBLAST rooted gene tree for the arabica563 gene H9Q71 0002169 (OG0015707) orthologous group, found in the shared region. (C) DendroBLAST gene tree for the arabica563 gene H9Q71 0003005 (OG0017996) orthologous group, found in the shared region. (D) The genes shared between the 2 species in the matching region have a higher BLAST sequence identity (mean 94%, dotted line) than those found in the flanking regions either side (mean 88%, dotted line). Figure drawn to scale. The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s008.tiff^{(52.5MB, tiff)}

S9 Fig. Differential expression of genes and putative effector genes in Fusarium xylarioides.

Genes up-regulated in the in planta samples are shaded orange and those down-regulated are shaded blue. Putative effectors identified in [28] are labelled according to their class: yellow, pre-described fungal plant pathogens; purple, small and cysteine-rich proteins; blue, carbohydrate-active enzymes (S6 and S7 Tables). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s009.tiff^{(96.8MB, tiff)}

S10 Fig. Whole-genome view of the loci of transposable elements in the F. xylarioides arabica563 reference genome.

Transposable elements shown are those which were newly identified on the mobile and pathogenic chromosome of F. oxysporum f. sp. lycopersici by [25], as well as the unclassified TE “rnd-6 family-1942.” Chromosomes are labelled with F. xylarioides contig numbers (see S12 Table).

(TIFF)

pbio.3002480.s010.tiff^{(77.4MB, tiff)}

S11 Fig. Miniature impala (mimp) family 1 sequence genealogy.

All mimp family 1 sequences from Fig 6 were aligned using MAFFT in Geneious v9.1.8. Using BLASTn, the consensus sequence was used to find copies in all other genomes used in this study (S2 and in the nr database. Any species with a hit >80% length and identity were selected (additional F. xylarioides and F. oxysporum hits were ignored), all sequences were re-aligned and a maximum likelihood tree with 1,000 bootstrap replicates constructed with IQ-TREE2. Bootstrap branch support is shown for all branches >0.7 (calculated with 1,000 bootstraps). Each point corresponds to a mimp sequence and the scalebar represents 0.1 substitutions per site. Tree tip labels are shaded by phylogenetic group: arabica, robusta, coffea1, and coffea2 F. xylarioides populations; FFC, F. fujikuroi species complex (all FFC tips are F. proliferatum; FOSC, F. oxysporum species complex. F. oxysporum strains are those annotated in this Fig 6). The data underlying this figure can be found in https://doi.org/10.5281/zenodo.13836286.

(TIFF)

pbio.3002480.s011.tiff^{(40.4MB, tiff)}

S1 Table. Bio-geographic details for strains sequenced in this study.

(PDF)

pbio.3002480.s012.pdf^{(63.2KB, pdf)}

S2 Table. Published genomes analysed in this study.

The columns describe: species, short species with strain number, accession number, whether the genome was used in OrthoFinder, whether the genome was used in S3B Fig for whole-genome similarity (Nsimscan).

(PDF)

pbio.3002480.s013.pdf^{(65.2KB, pdf)}

S3 Table. Population genetics statistics related to genetic differentiation (d_xy) in four Fusarium xylarioides populations.

Measured in 100 kb windows.

(PDF)

pbio.3002480.s014.pdf^{(75.3KB, pdf)}

S4 Table. Nineteen up-regulated genes in both Fusarium xylarioides arabica strains are absent from the Fusarium fujikuroi complex (FFC) and differentially present across Fusarium xylarioides, the Fusarium oxysporum species complex (FOSC), the Fusarium graminearum species complex (FGC), and the Fusarium solani species complex (FSC).

Species presence was determined using genealogies from OrthoFinder (see Methods) and BLAST similarity. A caret indicates a genealogy present in Fusarium udum, the other vascular wilt in the Fusarium fujikuroi species complex. Partial shading indicates partial presence across the population. An asterisk represents genealogies which match Fusarium oxysporum effector proteins.

(PDF)

pbio.3002480.s015.pdf^{(75.3KB, pdf)}

S5 Table. Predicted effectors in arabica563 genes by EffectorP [34].

Only those which were up-regulated in both Fusarium xylarioides arabica563 and arabica908 are shown. The numbers in brackets describe the certainty. The orthogroup is shown for those predicted effectors up-regulated in both arabica563 and arabica908, and those which are annotated as carbohydrate-active enzymes are described. H9Q71 0010006 encodes an Ecp2 effector protein and H9Q71 0007533 contains a LysM domain.

(PDF)

pbio.3002480.s016.pdf^{(67KB, pdf)}

S6 Table. Expression data for all putative effectors from [28] across the Fusarium xylarioides arabica563 samples.

Each gene is described as up-regulated in planta (“coffee.up”), in axenic (“culture.up”), or not differentially expressed (“ns”).

(PDF)

pbio.3002480.s017.pdf^{(64.9KB, pdf)}

S7 Table. Expression data for all putative effectors from [28] across the Fusarium xylarioides arabica908 samples.

Each gene is described as up-regulated in planta (“coffee.up”), in axenic (“culture.up”), or not differentially expressed (“ns”).

(PDF)

pbio.3002480.s018.pdf^{(64.6KB, pdf)}

S8 Table. The number of genes expressed for each carbohydrate-active enzyme subfamily in Fusarium xylarioides arabica563.

The in planta up-regulated genes are shaded, where the warmest colours represent the highest up-regulated gene count.

(PDF)

pbio.3002480.s019.pdf^{(61.8KB, pdf)}

S9 Table. The number of genes expressed for each carbohydrate-active enzyme subfamily in Fusarium xylarioides arabica908.

The in planta up-regulated genes are shaded, where the warmest colours represent the highest up-regulated gene count.

(PDF)

pbio.3002480.s020.pdf^{(61.4KB, pdf)}

S10 Table. Gene count and description for significantly enriched InterPro terms in the Coffea arabica in planta samples.

(PDF)

pbio.3002480.s021.pdf^{(88.7KB, pdf)}

S11 Table. Transposable element density in horizontally transferred regions (HTR), Starships and genome-wide in the Fusarium xylarioides arabica563 reference.

Low complexity and simple repeats were excluded from the transposable element (TE) count. “TE of interest” refers to those transposable elements identified on the F. oxysporum mobile pathogenic chromsome in [25] and the unclassified TE “rnd-6 family-1942.” sc2, supercontig 2.

(PDF)

pbio.3002480.s022.pdf^{(47.6KB, pdf)}

S12 Table. The contigs in the Fusarium xylarioides arabica563 reference genome which correspond to each chromosome in the F. verticillioides assembly [21].

The second column represents the contigs as named in the public assembly under accession PRJNA1043203. In the final column, each contig has been renamed to first correspond to its core Fusarium chromosome, and then its order in the whole genome visualisation (Fig 1A).

(PDF)

pbio.3002480.s023.pdf^{(47.6KB, pdf)}

S13 Table. Comparison of divergence patterns for HTR regions to those of background genome regions.

Mean bedtools coverage in 10 kb regions relative to the whole genome. I_Fx = mean similarity with other F. xylarioides populations across 10 kb windows within HTR. I_Fo = mean similarity with closest F. oxysporum match across 10 kb windows within HTR. I_Diff is I_Fx - I_Fo, a metric of how similar is F. xylarioides versus F. oxysporum: positive value = more similar to F. xylarioides; negative = more similar to F. oxysporum. p = proportion of values of I_Diff from all contiguous regions of genome the same size as HTR that are ≤ the observed value of I_Diff for the HTR.

(PDF)

pbio.3002480.s024.pdf^{(58KB, pdf)}

S14 Table. Comparison of divergence patterns for HTR regions to those of background genome regions.

Phylogenetic divergence relative to single copy ortholog gene trees. D_Fx = mean pairwise divergence among F. xylarioides genomes for given HTR (substitutions. per site). D_Fo = mean pairwise divergence among F. oxysporum genomes for a given HTR. D_FxFo = minimum pairwise divergence between F. xylarioides HTR and closest F. oxysporum match. p = proportion of values of D_FxFo across single ortholog gene trees ≤ the observed value of D_FxFo for the HTR. Mean D_FxFo across single ortholog gene trees = 0.067, 95% range 0.024–0.153 substitutions per site.

(PDF)

pbio.3002480.s025.pdf^{(58KB, pdf)}

S1 File. Genomic loci for the 6 putative horizontally transferred regions and the F. xylarioides strain in which they were identified.

This file can be opened in a text editor program.

(BED)

pbio.3002480.s026.bed^{(292B, bed)}

S2 File. Nucleotide sequences for the 6 putative horizontally transferred regions from the F. xylarioides strain in which they were identified.

This file can be opened in a text editor program.

(FA)

pbio.3002480.s027.fa^{(437KB, fa)}

S3 File. Predicted gene annotations and functions for Starship cargo and those annotations which flank the Starship.

(XLSX)

pbio.3002480.s028.xlsx^{(11KB, xlsx)}

S4 File. Sequence read archive (SRA) accessions and data counts for filtered data.

(XLSX)

pbio.3002480.s029.xlsx^{(14.1KB, xlsx)}

Acknowledgments

We thank W. Gordon, J. Meyer, M. Rutherford, T. Kibani, G. Armstong, D. Dring, R. Hillocks, E. Khonga, and R. Mehrotra, as well as other unnamed collectors who originally isolated the fungal pathogens and deposited them in the CABI-IMI collection, and 4 reviewers. With thanks to the Imperial College Stevenson Fund for sponsoring L. D. Peck’s research placement in Génétique et Écologie Évolutives at Université Paris Saclay, and to “i Focus & Write.” With special thanks to A. Snirc for sequencing genomes, J. Vernadet for help with variant-calling, A. van Westerhoven for providing the F. oxysporum f. sp. cubense reference genome, E. Gluck Thaler and A. Vogan for Star-fishing, and S. Gurr and M. Fisher for valuable comments.

Abbreviations

FDR: false discovery rate
gVCF: genomic variant call format
HGT: horizontal gene transfer
HTR: horizontally transferred region
INSDC: International Nucleotide Sequence Database Collaboration
ML: maximum likelihood
SIX: secreted in xylem
SNP: single-nucleotide polymorphism
TE: transposable element

Data Availability

Improved assemblies and raw sequencing data have been deposited in the relevant International Nucleotide Sequence Database Collaboration (INSDC) database with the BioProject ID PRJNA1043203 and the SRA run accession SRP491020 (see S3 File). This study also analysed publicly available data from accession numbers described in S2 Table. The exact versions of genome assemblies and annotations analysed in this manuscript are published in a publicly available Zenodo digital repository (https://doi.org/10.5281/zenodo.13836286), along with source data, custom analysis and plotting scripts, and bioinformatic workflows used in this study.

Funding Statement

This research was supported by the Natural Environment Research Council [Grant number NE/L002515/1 (LDP) and NE/S010866/1 (TGB).] The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Fones HN, Fisher MC, Gurr SJ. Emerging fungal threats to plants and animals challenge agriculture and ecosystem resilience. Microbiol Spectr. 2017;5:787–809. doi: 10.1128/microbiolspec.FUNK-0027-2016 [DOI] [PubMed] [Google Scholar]
2.Fisher MC, Henk DA, Briggs CJ, Brownstein JS, Madoff LC, McCraw SL, et al. Emerging fungal threats to animal, plant and ecosystem health. Nature. 2012;484:186–194. doi: 10.1038/nature10947 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.FAO. FAOSTAT data May 2022. Available from: www.fao.org/faostat. [Google Scholar]
4.USDA. Coffee Annual, Ethiopia tech. rep. (United States Dept of Agriculture, Foreign Agricultural Service, 2023), 1–7.
5.FAO. Statistical Pocket Book of the Food And Agricultural Organization for the United Nations: Coffee 2015 tech. rep. 2015. [Google Scholar]
6.Flood J. in Coffee Wilt Disease (ed Flood J.) chap. 1 (CAB International, Wallingford, UK, 2009). [Google Scholar]
7.Ristaino J B, Anderson PK, Bebber DP, Brauman KA, Cunniffe NJ, Fedoroff NV, et al. The persistent threat of emerging plant disease pandemics to global food security. Proc Natl Acad Sci U S A. 2021;118:e2022239118. doi: 10.1073/pnas.2022239118 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Rutherford M. Current knowledge of coffee wilt disease, a major constraint to coffee production in Africa. Phytopathology. 2006;96:663–666. doi: 10.1094/PHYTO-96-0663 [DOI] [PubMed] [Google Scholar]
9.Yadeta KA, Thomma BPHJ. The xylem as battleground for plant hosts and vascular wilt pathogens. Front Plant Sci. 2013;4. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Phiri N, Baker P. Coffee Wilt in Africa: Final Technical Report of the Regional Coffee Wilt Programme (2000–2007) tech. rep. (CABI, Wallingford, UK: CAB International, 2009), 233.
11.Peck LD, Boa E. Coffee wilt disease: The forgotten threat to coffee. Plant Pathol. 2023;73:506–521. doi: 10.1111/ppa.13833 [DOI] [Google Scholar]
12.Van Der Graaff NA, Pieters R. Resistance levels in Coffea arabica to Gibberella xylarioides and distribution pattern of the disease. Neth J Pl Path. 1978;84:117–120. [Google Scholar]
13.Mulatu A, Shanko D. Incidence and prevalence of coffee wilt disease (Gibberella xylarioides) and its impact on the rural livelihoods in western Guji zone, southern Ethiopia. Am J Biosci. 2019;7:7–15. [Google Scholar]
14.Bredeweg EL, Baker SE. Horizontal gene transfer in fungi. Grand Challenges in Biology and Biotechnology. 2020;317–332. [Google Scholar]
15.McDonald MC, Taranto AP, Hill E, Schwessinger B, Liu Z, Simpfendorfer S, et al. Transposon-mediated horizontal transfer of the host-specific virulence protein ToxA between three fungal wheat pathogens. MBio. 2019;10:01515–01519. doi: 10.1128/mBio.01515-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Gourlie R, McDonald M, Hafez M, Ortega-Polo R, Low KE, Abbott DW, et al. The pangenome of the wheat pathogen Pyrenophora tritici-repentis reveals novel transposons associated with necrotrophic effectors ToxA and ToxB. BMC Biol. 2022;20. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Gluck-Thaler E, Ralston T, Konkel Z, Ocampos CG, Ganeshan VD, Dorrance AE, et al. Giant Starship elements mobilize accessory genes in fungal genomes. Mol Biol Evol. 2022;39. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Gluck-Thaler E, Forsythe A, Puerner C, Stajich JE, Croll D, Cramer RA, et al. Giant transposons promote strain heterogeneity in a major fungal pathogen. Biorxiv. 2024. doi: 10.1101/2024.06.28.601215 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Urquhart AS, Vogan AA, Gardiner DM, Idnurm A. Starships are active eukaryotic transposable elements mobilized by a new family of tyrosine recombinases. Proc Natl Acad Sci U S A. 2023;120:e2214521120. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Michielse CB, Rep M. Pathogen profile update: Fusarium oxysporum. Mol Plant Pathol. 2009;10:311–324. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Ma LJ, Van Der Does HC, Borkovich KA, Coleman JJ, Daboussi MJ, Di Pietro A, et al. Comparative genomics reveals mobile pathogenicity chromosomes in Fusarium. Nature. 2010;464:367–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.van Dam P, de Sain M, ter Horst A, van der Gragt M, Rep M. Use of comparative genomics-based markers for discrimination of host specificity in Fusarium oxysporum. Appl Environ Microbiol. 2018;84:1868–1885. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Li J, Fokkens L, Rep M. A single gene in Fusarium oxysporum limits host range. Mol Plant Pathol. 2020;1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Rep M, van der Does HC, Meijer M, van Wijk R, Houterman PM, Dekker HL, et al. A small, cysteine-rich protein secreted by Fusarium oxysporum during colonization of xylem vessels is required for I-3-mediated resistance in tomato. Mol Microbiol. 2004;53:1373–1383. [DOI] [PubMed] [Google Scholar]
25.Schmidt SM, Houterman PM, Schreiver I, Ma L, Amyotte S, Chellappan B, et al. MITEs in the promoters of effector genes allow prediction of novel virulence genes in Fusarium oxysporum. BMC Genomics. 2013;14:1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Faino L, Seidl MF, Shi-Kunne X, Pauper M, van den Berg GCM, Wittenberg AHJ, et al. Transposons passively and actively contribute to evolution of the two speed genome of a fungal pathogen. Genome Res. 2016;26:1091–1100. doi: 10.1101/gr.204974.116 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Farrer RA, Fisher MC. Describing genomic and epigenomic traits underpinning emerging fungal pathogens. Adv Genet. 2017;100:73–140. ISSN: 0065-2660. doi: 10.1016/bs.adgen.2017.09.009 [DOI] [PubMed] [Google Scholar]
28.Peck LD, Nowell RW, Flood J, Ryan MR, Barraclough TG. Historical genomics reveals the evolutionary mechanisms behind multiple outbreaks of the host specific coffee wilt pathogen Fusarium xylarioides. BMC Genomics. 2021;22:1–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Kvas M, Marasas W, Wingfield B, Wingfield M, Steenkamp E. Diversity and evolution of Fusarium species in the Gibberella fujikuroi complex. Fungal Divers. 2009;34:1–21. [Google Scholar]
30.Oduor G. et al. Surveys to assess the extent and impact of coffee wilt disease in East and Central Africa. Final Technical Report. tech. rep. (CAB International, Nairobi, Kenya, 2003).
31.van Westerhoven AC, Aguilera-Galvez C, Nakasato-Tagami G, Shi-Kunne X, Martinez de la Parte E, Chavarro-Carrero E, et al. Segmental duplications drive the evolution of accessory regions in a major crop pathogen. New Phytol. 2024;242:610–625. doi: 10.1111/nph.19604 [DOI] [PubMed] [Google Scholar]
32.van Dam P, Rep M. The distribution of miniature Impala elements and SIX genes in the Fusarium genus is suggestive of horizontal gene transfer. J Mol Evol. 2017;85:14–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Rogers LM, Kim Y-K, Guo W, Gonzá Lez-Candelas L, Li D, Kolattukudy PE, et al. Requirement for either a host-or pectin-induced pectate lyase for infection of Pisum sativum by Nectria hematococca. Proc Natl Acad Sci U S A. 2000;97. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Sperschneider J, Dodds PN. EffectorP 3.0: Prediction of apoplastic and cytoplasmic effectors in fungi and oomycetes. Mol Plant Microbe Interact. 2022;146–156. [DOI] [PubMed] [Google Scholar]
35.Bergemann M, Lespinet O, M’Barek SB, Daboussi MJ, Dufresne M. Genome wide analysis of the Fusarium oxysporum mimp family of MITEs and mobilization of both native and de novo created mimps. J Mol Evol. 2008;67:631–642. [DOI] [PubMed] [Google Scholar]
36.Giraud T, Gladieux P, Gavrilets S. Linking the emergence of fungal plant diseases with ecological speciation. Trends Ecol Evol. 2010;25:387–395. doi: 10.1016/j.tree.2010.03.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Rutherford M, Bieysse D, Lepoint P, Maraite H. in Coffee Wilt Disease (ed Flood J.) 99–119 (CAB International, Wallingford, UK, 2009). [Google Scholar]
38.Lepoint, P. C. E. Speciation within the African Coffee Wilt Pathogen PhD thesis (Universite Catholique de Louvain., 2006), 208.
39.Girma A. Diversity in Pathogenicity and Genetics of Gibberella xylarioides (Fusarium xylarioides) Populations and Resistance of Coffea spp. in Ethiopia PhD thesis (Rheinischen Friedrich-Wilhelms-Universitat zu Bonn, 2004), 85.
40.Kurian SM, Di Pietro A, Read ND. Live-cell imaging of conidial anastomosis tube fusion during colony initiation in Fusarium oxysporum. PLoS ONE. 2018;13(ed Freitag M.) e0195634. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Lepoint PCE, Munaut FTJ, Maraite HMM. Gibberella xylarioides sensu lato from Coffea canephora: a new mating population in the Gibberella fujikuroi species complex. Appl Environ Microbiol. 2005;71:8466–8471. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Ropars J, Didiot E, Rodríguez de la Vega RC, Bennetot B, Coton M, Poirier E, et al. Domestication of the emblematic white cheese-making fungus Penicillium camemberti and its diversification into two varieties. Curr Biol. 2020;30:4441–4453. [DOI] [PubMed] [Google Scholar]
43.Rouxel T, Grandaubert J, Hane JK, Hoede C, van de Wouw AP, Couloux A, et al. Effector diversification within compartments of the Leptosphaeria maculans genome affected by Repeat-Induced Point mutations. Nat Commun. 2011;2:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Guo L, Han L, Yang L, Zeng H, Fan D, Zhu Y, et al. Genome and transcriptome analysis of the fungal pathogen Fusarium oxysporum f. sp. cubense causing banana vascular wilt disease. PLoS ONE. 2014;9:e95543. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Brankovics B, van Diepeningen AD, de Hoog GS, van der Lee TA, Waalwijk C. Detecting introgression between members of the Fusarium fujikuroi and F. oxysporum species complexes by comparative mitogenomics. Front Microbiol. 2020;11. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Flood J. Epidemiology and variability of Gibberella xylarioides, the coffee wilt pathogen. R8188 (ZA0505). Final Technical Report tech. rep. (CAB International, Wallingford, UK, 2005).
47.Serani S, Taligoola HK, Hakiza GJ. An investigation into Fusarium spp. associated with coffee and banana plants as potential pathogens of robusta coffee. Afr J Ecol. 2007;45:91–95. [Google Scholar]
48.O’Donnell K, Kistler HC, Cigelnik E, Ploetz RC. Multiple evolutionary origins of the fungus causing Panama disease of banana: Concordant evidence from nuclear and mitochondrial gene genealogies. Proc Natl Acad Sci U S A. 1998;95:2044–2049. doi: 10.1073/pnas.95.5.2044 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Raffaele S, Farrer RA, Cano LM, Studholme DJ, MacLean D, Thines M, et al. Genome Evolution Following host jumps in the Irish potato famine pathogen lineage. Science. 2010;330:1540–1543. doi: 10.1126/science.1193070 [DOI] [PubMed] [Google Scholar]
50.Ropars J, Rodríguez De La Vega RC, Ló Pez- M, Dupont L, Branca A, Giraud T. Adaptive horizontal gene transfers between multiple cheese-associated fungi. Curr Biol. 2015;25:2562–2569. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Richards TA, Soanes DM, Jones MDM, Vasieva O, Leonard G, Paszkiewicz K, et al. Horizontal gene transfer facilitated the evolution of plant parasitic mechanisms in the oomycetes. Proc Natl Acad Sci U S A. 2011;108:15258–15263. doi: 10.1073/pnas.1105100108 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Van Etten J, Bhattacharya D. Horizontal gene transfer in eukaryotes: not if, but how much? Trends Genet. 2020;36:915–925. doi: 10.1016/j.tig.2020.08.006 [DOI] [PubMed] [Google Scholar]
53.Brewer GE, Clarkson JJ, Maurin O, Zuntini AR, Barber V, Bellot S, et al. Factors affecting targeted sequencing of 353 nuclear genes from herbarium specimens spanning the diversity of angiosperms. Front Plant Sci. 2019;10:1102. doi: 10.3389/fpls.2019.01102 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Ryan MJ, Peck LD, Smith D, Flood J, Barraclough TG. Culture collections as a source of historic strains for genomic studies in plant pathology. J Plant Pathol. 2022. [Google Scholar]
55.Peck LD. Seventy years of Fusarium wilt—coffee interactions: historical genomics reveals pathogen emergence and divergence PhD thesis (Imperial College London, London, 2023). [Google Scholar]
56.Spanu PD, Abbott JC, Amselem J, Burgis TA, Soanes DM, Stüber K, et al. Genome expansion and gene loss in powdery mildew fungi reveal tradeoffs in extreme parasitism. Science. 2010;330:1543–1546. doi: 10.1126/science.1194573 [DOI] [PubMed] [Google Scholar]
57.Kolmogorov M, Yuan J, Lin Y, Pevzner PA. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019;37:540–546. doi: 10.1038/s41587-019-0072-8 [DOI] [PubMed] [Google Scholar]
58.Aury JM, Istace B. Hapo-G, haplotype-aware polishing of genome assemblies with accurate reads. NAR Genom Bioinform. 2021;3. doi: 10.1093/nargab/lqab034 [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Davey JW, Davis SJ, Mottram JC, Ashton PD. Tapestry: validate and edit small eukaryotic genome assemblies with long reads. bioRxiv. 2020. [Google Scholar]
60.Marçais G, Delcher AL, Phillippy AM, Coston R, Salzberg SL, Zimin A. MUMmer4: A fast and versatile genome alignment system. PLoS Comput Biol. 2018;14:e1005944. doi: 10.1371/journal.pcbi.1005944 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17:10–12. [Google Scholar]
62.Magoč T, Salzberg SL. FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics. 2011;27:2957–2963. doi: 10.1093/bioinformatics/btr507 [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Li D, Liu CM, Luo R, Sadakane K, Lam TW. MEGAHIT: An ultra-fast singlenode solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics. 2015;31:1674–1676. [DOI] [PubMed] [Google Scholar]
64.Li D, Luo R, Liu CM, Leung CM, Ting HF, Sadakane K, et al. MEGAHIT v1.0: A fast and scalable metagenome assembler driven by advanced methodologies and community practices. Methods. 2016;102:3–11. doi: 10.1016/j.ymeth.2016.02.020 [DOI] [PubMed] [Google Scholar]
65.Mikheenko A, Prjibelski A, Saveliev V, Antipov D, Gurevich A. Versatile genome assembly evaluation with QUAST-LG. Bioinformatics. 2018;34:142–150. doi: 10.1093/bioinformatics/bty266 [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Olal S, Bitalo DN, Olango ND, Mulindwa J, Ochwo S, Opiyo SO, et al. novo genome sequence of a Fusarium xylarioides race pathogenic to Robusta coffee (Coffea canephora) in Uganda. Microbiol Resour Announc. 2019;8:e00520–e00519. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Wingfield BD, Fourie A, Simpson MC, Bushula-Njah VS, Aylward J, Barnes I, et al. IMA Genome-F 11. IMA Fungus. 2019;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Alonge M, Soyk S, Ramakrishnan S, Wang X, Goodwin S, Sedlazeck FJ, et al. RaGOO: Fast and accurate reference-guided scaffolding of draft genomes. Genome Biol. 2019;20:224. doi: 10.1186/s13059-019-1829-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol. 2011;29:644–652. doi: 10.1038/nbt.1883 [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Palmer JM, Stajich J. Funannotate v1.8.1: Eukaryotic genome annotation. 2020. [Google Scholar]
71.Shumate A, Salzberg SL. Liftoff: accurate mapping of gene annotations. Bioinformatics. 2021;37:1639–1643. doi: 10.1093/bioinformatics/btaa1016 [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Ma L-J, Zhang Y, Li C, Liu S, Liu C, Mostert D, et al. Accessory genes in tropical race 4 contributed to the recent resurgence of the devastating disease of Fusarium wilt of banana. Res Sq. 2023. [Google Scholar]
73.Mesny F, Miyauchi S, Thiergart T, Pickel B, Atanasova L, Karlsson M, et al. Genetic determinants of endophytism in the Arabidopsis root mycobiome. Nat Commun. 2021;12. doi: 10.1038/s41467-021-27479-y [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Flynn JM, Hubley R, Goubert C, Rosen J, Clark AG, Feschotte C, et al. RepeatModeler2: automated genomic discovery of transposable element families. Proc Natl Acad Sci U S A. 2020;117:9451–9457. [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Goubert C, Craig RJ, Bilat AF, Peona V, Vogan AA, Protasio A, et al. A beginner’s guide to manual curation of transposable elements. Mob DNA. 2022;13:1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Notredame C, Higgins DG, Heringa J. T-Coffee: A novel method for fast and accurate multiple sequence alignment. J Mol Biol. 2000;302:205–217. doi: 10.1006/jmbi.2000.4042 [DOI] [PubMed] [Google Scholar]
77.Smit AFA, Hubley R, Green P. RepeatMasker Open-4.0. 2015. [Google Scholar]
78.Katoh K, Standley DM. MAFFT Multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30:772–780. doi: 10.1093/molbev/mst010 [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Katoh K, Misawa K, Kuma K, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002;30:3059–3066. doi: 10.1093/nar/gkf436 [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, et al. BLAST+: architecture and applications. BMC Bioinform. 2009;10:1–9. doi: 10.1186/1471-2105-10-421 [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Hoang DT, Chernomor O, von Haeseler A, Minh BQ, Vinh LS. UFBoot2: Improving the Ultrafast Bootstrap Approximation. Mol Biol Evol. 2018;35:518–522. doi: 10.1093/molbev/msx281 [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Novichkov V, Kaznadzey A, Alexandrova N, Kaznadzey D. NSimScan: DNA comparison tool with increased speed, sensitivity and accuracy. Bioinformatics (Oxford, England). 2016;32:2380–2381. doi: 10.1093/bioinformatics/btw126 [DOI] [PubMed] [Google Scholar]
83.Emms DM, Kelly S. OrthoFinder: Phylogenetic orthology inference for comparative genomics. Genome Biol. 2019;20:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
84.Kelly S, Maini PK. DendroBLAST: Approximate phylogenetic trees in the absence of multiple sequence alignments. PLoS ONE. 2013;8:e58537. doi: 10.1371/journal.pone.0058537 [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Emms DM, Kelly S. OrthoFinder: solving fundamental biases in whole genome comparisons dramatically improves orthogroup inference accuracy. Genome Biol. 2015;16:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
86.Emms DM, Kelly S. STAG: species tree inference from all genes. BioRxiv. 2018;1–29. [Google Scholar]
87.Emms D, Kelly S. STRIDE: species tree root inference from gene duplication events. Mol Biol Evol. 2017;34:3267–3278. doi: 10.1093/molbev/msx259 [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Fujisawa T, Aswad A, Barraclough TG. A rapid and scalable method for multilocus species delimitation using Bayesian model comparison and rooted triplets. Syst Biol. 2016;65:759–771. doi: 10.1093/sysbio/syw028 [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Shen W, Le S, Li Y, Hu F. SeqKit: A cross-platform and ultrafast toolkit for FASTA/Q file manipulation. PLoS ONE. 2016;11:e0163962. doi: 10.1371/journal.pone.0163962 [DOI] [PMC free article] [PubMed] [Google Scholar]
90.Huson DH, Bryant D. Application of Phylogenetic Networks in Evolutionary Studies. Mol Biol Evol. 2006;23(2):254–267. doi: 10.1093/molbev/msj030 [DOI] [PubMed] [Google Scholar]
91.Pfeifer B, Wittelsburger U, Ramos-Onsins SE, Lercher MJ. PopGenome: An efficient swiss army knife for population genomic analyses in R. Mol Biol Evol. 2014;31:1929–1936. doi: 10.1093/molbev/msu136 [DOI] [PMC free article] [PubMed] [Google Scholar]
92.Li H. Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018;34:3094–3100. doi: 10.1093/bioinformatics/bty191 [DOI] [PMC free article] [PubMed] [Google Scholar]
93.Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, et al. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol. 2020;37:1530–1534. https://academic.oup.com/mbe/article/37/5/1530/ 5721363. doi: 10.1093/molbev/msaa015 [DOI] [PMC free article] [PubMed] [Google Scholar]
94.Hao Z, Lv D, Ge Y, Shi J, Weijers D, Yu G, et al. RIdeogram: Drawing SVG graphics to visualize and map genome-wide data on the idiograms. PeerJ Comput Sci. 2020;6:1–11. ISSN: 23765992. doi: 10.7717/peerj-cs.251 [DOI] [PMC free article] [PubMed] [Google Scholar]
95.Gluck-Thaler E, Vogan AA. Systematic identification of cargo-carrying genetic elements reveals new dimensions of eukaryotic diversity. Biorxiv. 2023. doi: 10.1101/2023.10.24.563810 [DOI] [PMC free article] [PubMed] [Google Scholar]
96.Levy Karin E, Mirdita M, Soding J. MetaEuk—sensitive, high-throughput gene¨ discovery, and annotation for large-scale eukaryotic metagenomics. Microbiome. 2020;8:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
97.Eddy SR. Accelerated profile HMM searches. PLoS Comput Biol. 2011;7:e1002195. doi: 10.1371/journal.pcbi.1002195 [DOI] [PMC free article] [PubMed] [Google Scholar]
98.Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26:841–842. doi: 10.1093/bioinformatics/btq033 [DOI] [PMC free article] [PubMed] [Google Scholar]
99.Ayad LAK, Pissis SP, Polychronopoulos D. CNEFinder: finding conserved noncoding elements in genomes. Bioinformatics. 2018;34:i743–i747. doi: 10.1093/bioinformatics/bty601 [DOI] [PMC free article] [PubMed] [Google Scholar]
100.Pierce NT, Irber L, Reiter T, Brooks P, Brown CT. Large-scale sequence comparisons with sourmash. F1000Res. 2019;8. [DOI] [PMC free article] [PubMed] [Google Scholar]
101.Enright AJ, Van Dongen S, Ouzounis CA. An efficient algorithm for large-scale detection of protein families. Nucleic Acids Res. 2002;30:1575–1584. doi: 10.1093/nar/30.7.1575 [DOI] [PMC free article] [PubMed] [Google Scholar]
102.Hackl T, Ankenbrand MJ, van Adrichem B. gggenomes: A Grammar of Graphics for Comparative Genomics. 2024. [Google Scholar]
103.Geiser DM, Lewis Ivey ML, Hakiza G, Juba JH, Miller SA. Gibberella xylarioides (anamorph: Fusarium xylarioides), a causative agent of coffee wilt disease in Africa, is a previously unrecognized member of the G. fujikuroi species complex. Mycologia. 2005;97:191–201. [DOI] [PubMed] [Google Scholar]
104.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30:2114. doi: 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
105.Andrews S, Krueger F, Seconds-Pichon A, Biggins F, Wingett S. FastQC. A Quality Control tool for High Throughput Sequence Data. Babraham Bioinformatics. 2015. [Google Scholar]
106.Zimin A, Yepes M, Maldonado CE, Navarro L, Kovaka S, Pertea M, et al. Coffea arabica variety Caturra and C. eugenioides Genome Assembly and Annotation in Plant and Animal Genome Conference XXVII. 2019. [Google Scholar]
107.Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017;14:417–419. doi: 10.1038/nmeth.4197 [DOI] [PMC free article] [PubMed] [Google Scholar]
108.Haas BJ, Papanicolaou A, Yassour M, Grabherr M, Blood PD, Bowden J, et al. De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nat Protoc. 2013:1494–1512. [DOI] [PMC free article] [PubMed] [Google Scholar]
109.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:1–21. doi: 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
110.Chen Y, Yin H, Gao M, Zhu H, Zhang Q, Wang Y. Comparative transcriptomics atlases reveals different gene expression pattern related to Fusarium wilt disease resistance and susceptibility in two Vernicia species. Front Plant Sci. 2016;7. [DOI] [PMC free article] [PubMed] [Google Scholar]
111.Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43:e47–e47. doi: 10.1093/nar/gkv007 [DOI] [PMC free article] [PubMed] [Google Scholar]
112.Soneson C, Robinson MD. Bias, robustness and scalability in single-cell differential expression analysis. Nat Methods. 2018;15:255–261. doi: 10.1038/nmeth.4612 [DOI] [PubMed] [Google Scholar]
113.Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc B Methodol. 1995;57:289–300. [Google Scholar]
114.Cantarel BI, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B. The Carbohydrate-Active enZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res. 2009;37. doi: 10.1093/nar/gkn663 [DOI] [PMC free article] [PubMed] [Google Scholar]
115.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria, 2021.
116.Wickham H, Averick M, Bryan J, Chang WDL, Mcgowan A, François R, et al. Welcome to the Tidyverse. J Open Source Softw. 2019;4:1686. [Google Scholar]
117.Wilke C. Cowplot: Streamlined Plot Theme and Plot Annotations for ‘ggplot2’. 2022. [Google Scholar]
118.Slowikowski K. ggrepel: Automatically Position Non-Overlapping Text Labels with ‘ggplot2’. Comprehensive R Archive Network (CRAN). 2021. [Google Scholar]
119.Ram K, Wickham H. wesanderson: A Wes Anderson Palette Generator Comprehensive R Archive Network (CRAN). 2018. [Google Scholar]
120.Kolde R. pheatmap: Pretty Heatmaps Comprehensive R Archive Network (CRAN). 2019. [Google Scholar]
121.Yu G, Smith DK, Zhu H, Guan Y, Lam TTY. ggtree: an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods Ecol Evol. 2017;8:28–36. [Google Scholar]
122.Wang LG, Lam TTY, Xu S, Dai Z, Zhou L, Feng T, et al. Treeio: An R package for phylogenetic tree input and output with richly annotated and associated data. Mol Biol Evol. 2020;37:599–603. doi: 10.1093/molbev/msz240 [DOI] [PMC free article] [PubMed] [Google Scholar]
123.Gu Z, Gu L, Eils R, Schlesner M, Brors B. Circlize implements and enhances circular visualization in R. Bioinformatics. 2014;30:2811–2812. doi: 10.1093/bioinformatics/btu393 [DOI] [PubMed] [Google Scholar]
124.Paradis E, Schliep K. Ape 5.0: An environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics. 2019;35:526–528. doi: 10.1093/bioinformatics/bty633 [DOI] [PubMed] [Google Scholar]
125.Garnier S, Ross N, Rudis R, Camargo PA, Sciaini M, Scherer C. viridis(Lite)—Colorblind-Friendly Color Maps for R. Comprehensive R Archive Network (CRAN). 2023. [Google Scholar]
126.Wright K. pals: Color Palettes, Colormaps, and Tools to Evaluate Them. Comprehensive R Archive Network (CRAN). 2021. [Google Scholar]
127.Winter D. pafr: Read, manipulate and visualize ‘Pairwise mApping Format’ Data. Comprehensive R Archive Network (CRAN). 2022. [Google Scholar]

PLoS Biol. doi: 10.1371/journal.pbio.3002480.r001

Decision Letter 0

Melissa Vazquez Hernandez

19 Dec 2023

Dear Dr Peck,

Thank you for submitting your manuscript entitled "Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease" for consideration as a Research Article by PLOS Biology.

Your manuscript has now been evaluated by the PLOS Biology editorial staff as well as by an academic editor with relevant expertise and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. After your manuscript has passed the checks it will be sent out for review. To provide the metadata for your submission, please Login to Editorial Manager (https://www.editorialmanager.com/pbiology) within two working days, i.e. by Dec 21 2023 11:59PM.

If your manuscript has been previously peer-reviewed at another journal, PLOS Biology is willing to work with those reviews in order to avoid re-starting the process. Submission of the previous reviews is entirely optional and our ability to use them effectively will depend on the willingness of the previous journal to confirm the content of the reports and share the reviewer identities. Please note that we reserve the right to invite additional reviewers if we consider that additional/independent reviewers are needed, although we aim to avoid this as far as possible. In our experience, working with previous reviews does save time.

If you would like us to consider previous reviewer reports, please edit your cover letter to let us know and include the name of the journal where the work was previously considered and the manuscript ID it was given. In addition, please upload a response to the reviews as a 'Prior Peer Review' file type, which should include the reports in full and a point-by-point reply detailing how you have or plan to address the reviewers' concerns.

During the process of completing your manuscript submission, you will be invited to opt-in to posting your pre-review manuscript as a bioRxiv preprint. Visit http://journals.plos.org/plosbiology/s/preprints for full details. If you consent to posting your current manuscript as a preprint, please upload a single Preprint PDF.

Feel free to email us at plosbiology@plos.org if you have any queries relating to your submission.

Kind regards,

Melissa

Melissa Vazquez Hernandez, Ph.D.

Associate Editor

PLOS Biology

mvazquezhernandez@plos.org

PLoS Biol. doi: 10.1371/journal.pbio.3002480.r002

Decision Letter 1

Melissa Vazquez Hernandez

5 Feb 2024

Dear Dr Peck,

Thank you for your patience while your manuscript "Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease" was peer-reviewed at PLOS Biology. I apologized for the long waiting time. Your manuscript has been evaluated by the PLOS Biology editors, an Academic Editor with relevant expertise, and by several independent reviewers.

As you will see in the reviewer reports, while there is recognition of the potential significance of your findings, a significant number of crucial concerns have been raised. Following discussions with the Academic Editor, it is evident that substantial revisions would be necessary to meet the criteria for publication in PLOS Biology. Given our and the reviewers' interest in your study, we are open to considering a comprehensive revision; however, this would require thorough addressing of all reviewers' comments. A successful revision must address issues such as strengthening the HTR hypothesis, correcting the inappropriate use of FST, incorporating available genomic data, including bootstrap support in the phylogenetic analysis, providing support for MIMPS, and enhancing statistical support. Additionally, the revision should clarify the lineage sorting hypothesis and thoroughly discuss it.

IMPORTANT: Please note that the Academic Editor has kindly provided some additional guidance as to what needs to be addressed before we can consider the study further (see the foot of this email). All other concerns raised by the reviewers should also be addressed.

Given the extent of revision that would be needed, we cannot make a decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript would need to be seen by the reviewers again, but please note that we would not engage them unless their main concerns have been addressed.

We appreciate that these requests represent a great deal of extra work, and we are willing to relax our standard revision time to allow you 6 months to revise your study. Please email us (plosbiology@plos.org) if you have any questions or concerns, or envision needing a (short) extension.

At this stage, your manuscript remains formally under active consideration at our journal; please notify us by email if you do not intend to submit a revision so that we may withdraw it.

**IMPORTANT - SUBMITTING YOUR REVISION**

Your revisions should address the specific points made by each reviewer. Please submit the following files along with your revised manuscript:

1. A 'Response to Reviewers' file - this should detail your responses to the editorial requests, present a point-by-point response to all of the reviewers' comments, and indicate the changes made to the manuscript.

*NOTE: In your point by point response to the reviewers, please provide the full context of each review. Do not selectively quote paragraphs or sentences to reply to. The entire set of reviewer comments should be present in full and each specific point should be responded to individually, point by point.

You should also cite any additional relevant literature that has been published since the original submission and mention any additional citations in your response.

2. In addition to a clean copy of the manuscript, please also upload a 'track-changes' version of your manuscript that specifies the edits made. This should be uploaded as a "Revised Article with Changes Highlighted " file type.

*Resubmission Checklist*

When you are ready to resubmit your revised manuscript, please refer to this resubmission checklist: https://plos.io/Biology_Checklist

To submit a revised version of your manuscript, please go to https://www.editorialmanager.com/pbiology/ and log in as an Author. Click the link labelled 'Submissions Needing Revision' where you will find your submission record.

Please make sure to read the following important policies and guidelines while preparing your revision:

*Published Peer Review*

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out. Please see here for more details:

https://blogs.plos.org/plos/2019/05/plos-journals-now-open-for-published-peer-review/

*PLOS Data Policy*

Please note that as a condition of publication PLOS' data policy (http://journals.plos.org/plosbiology/s/data-availability) requires that you make available all data used to draw the conclusions arrived at in your manuscript. If you have not already done so, you must include any data used in your manuscript either in appropriate repositories, within the body of the manuscript, or as supporting information (N.B. this includes any numerical values that were used to generate graphs, histograms etc.). For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5

*Blot and Gel Data Policy*

We require the original, uncropped and minimally adjusted images supporting all blot and gel results reported in an article's figures or Supporting Information files. We will require these files before a manuscript can be accepted so please prepare them now, if you have not already uploaded them. Please carefully read our guidelines for how to prepare and upload this data: https://journals.plos.org/plosbiology/s/figures#loc-blot-and-gel-reporting-requirements

*Protocols deposition*

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for your submission to our journal. We hope that our editorial process has been constructive thus far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Melissa

Melissa Vazquez Hernandez, Ph.D.

Associate Editor

PLOS Biology

mvazquezhernandez@plos.org

-----------------------------------------

COMMENTS FROM THE ACADEMIC EDITOR:

The critical issues that need addressing, in my opinion, are:

The available genome datasets that weren't used, especially the more comprehensive chromosome-level assemblies and the species that were excluded.

The problem of lineage sorting. It's unacceptable for this to be mentioned only once and in an ambiguous manner (page 7), leaving room for this alternative hypothesis without further detail. This needs to be directly tackled (as even reviewer 2 suggested) and discussed more thoroughly in the paper. Personally, I'm skeptical of horizontal gene transfer (HGT/HTR) claims when the boundaries of the transferred segment are undefined.

Reviewer 1 expressed several important criticisms, notably regarding the inappropriate use of FST.

Reviewer 2 criticized the bootstrap support for the phylogenetic trees and the classification into four clades, which is a fundamental issue. Reviewer 3 also highlighted the inadequate bootstrap support, which is really concerning as this is so fundamental to these analyses.

-----------------------------------------

REVIEWER'S COMMENTS:

— — — — —

Reviewer #1:

In this manuscript the authors investigate an hypothesis that was proposed previously that there are some regions of the genomes of F. xylarioides were potentially as the result of horizontal gene transfer (HTR) from another Fusarium species, F. oxysporum. To do this additional genomes have been sequenced, both F. xylarioides and F. oxysporum. In addition the authors interrogated whether genes in the putative HTR regions were highly expressed during infection of plants by two of the isolates studied.

Overall I found that the manuscript seems to consist of two stories, one is about the differences and similarities between F. xylarioides isolates, which is in itself interesting and useful with regards to managing these pathogens. The second is the hypothesis that some of genome segments in F. xylarioides seem to have originated from another source and share significant similarity with equivalent sequences in F. oxysporum

The authors make a convincing argument with regards to HTR but I miss a number of aspects. The first is that there is a long read genome for F. xylarioides and while they suggest that a full chromosome assembly was not possible I think that it would be useful to perhaps assign some of the contigs to chromosomes. Most species in the Fusarium fujikuroi species complex have 12 "core" chromosomes and then some have accessory chromosomes. Given that there are chromosome assemblies of some related species this assignation should be possible?

Different strain types featured different putative effector genes, whereby some effectors showed close matches with F. oxysporum, but were absent from closer relatives of F. xylarioides. But the authors do not seem to have considered many other species that are in the Fusarium fujikuroi species complex - and there are many more genomes available for comparison than those that are used in this study. This would be an easy comparison to make and potentially strengthens the argument for HTR.

It is also not clear why the authors chose to sequence the genomes of F. oxysporum isolates in this study (I understand the rationale in terms of which forma specialis were chosen) given that there are many genomes already available including (with one exception) the forma specialis which were chosen for this study. These other genomes could also be considered in this analysis for the presence of the HTR regions - if only to show that these are not isolate specific in F. oxysporum?

An aspect of this study that is discussed is that there are in fact potentially more than one species within the F. xylarioides complex. I think that the authors show good evidence that there are in fact at least two species and to consider these just different populations is confusing and not strictly accurate. I think that this aspect does make understanding this study more complicated. The fact that the authors "rename" the isolates arabica, robusta and coffea actually speaks volumes to the fact that these are probably three different species.

I do not agree that one can reasonably do an FST calculation on such a small number of isolates.

In supplementary figure 1 the asterix is not explained?

The names of the genomes and isolates is very complicated and I found myself getting very confused. For example in figures 3 and 6 the isolatename I039 appears but is not in any of the tables, I had to go back to the previous publication to understand this and then noted that this isolate name is in fact in the supplementary section named as I-394. Also there is the F. verticilliodes isolate that is alternatively 14953 and also 953. The naming's of the isolates is different in supplementary figure 1 and 2 (for some reason in the supplementary table 2 F. xylarioides 379925 and L0394 are written in italics) I suggest that a single table providing all the names of all the isolates for which genomes have been used in the study is necessary. In addition the same name needs to be used for a particular isolate throughout the entire manuscript (the supplementary table 2 goes some way to providing this information but then does not include the isolates sequenced in this study). This way when one compares figures it is possible to easily understand what is going on.

One of the most perplexing aspects of this manuscript is that in some cases analyses are done using all the genomes and in others the authors have focused only on the genomes that were sequenced in this study. This is rather unbalanced and does not seem to be needed?

The last sentence of the abstract states "Our results support the hypothesis that horizontal gene transfers contributed to the repeated emergence of this fungal disease." This sentence is problematic because I think it refers to F. xylarioides but the last species referred to in the previous sentence is in fact F. oxysporum. Another aspect is that while fungi can be causal agents of a plant disease, one cannot strictly refer to them as causing a fungal disease. Surely a fungal disease is a disease of a fungus?

— — — — —

Reviewer #2: This tour de force by Peck et al seeks to elucidate how horizontal transfers between Fusarium species has contributed to the outbreaks of devastating coffee wilt disease. I must commend the authors on such a thorough study, and I must say I enjoy reading it. Therefore, the comments I have seek to improve the overall study and tie up a couple of loose ends.

My main comments on this is that the foundational analyses and findings need a little more support. The follow on findings will then be bolstered by this. Primarily, the authors state in the Results that 'there are four well resolved groups', yet my issue with this is that there isn't basic bootstrap support on the phylogenetic analysis to support this. Then, when talking about coffea strains falling into two distinct clades, again, bootstrap support would help here. For these two clades it would be worth noting what makes them distinct, i.e. are there differences in gene content (this may have been done and I didn't see it, in which case, make it a bit clearer).

Perhaps my biggest comments are:

- does temporal analysis using BEAST/BEAST2 confirm your findings? This may be beyond the scope of this study, but would be nice.

- I think there needs to be a bit more consideration of recombination and/or gene flow, which would be resolved by including an ADMIXTURE and/or fineSTRUCTURE analysis. Indeed, the latter would help resolve ancestral donation and inferring the 'donor strains'.

Minor comments:

Data availability: for me, I get the notification that there's been no data deposited or its awaiting submission/validation.

Introduction: a couple of sentences describing the yield losses (amounts/percentage) attributed to fungi would help non-Fusarium/fungal specialist readers. Also, perhaps expanding the recent disease severity in Ethiopia. Is this due to climate change? Also in the introduction, it should be noted that Starships haven't been proven to be involved in HGT (i.e. no one has 'seen' them move) - see Bucknell et al 2023 Mol Microbiology. At the moment, it's all sequence evidence, which is fine, but we are not completely 100% sure of how they are moving.

Methods: 'other Fusarium genomes' - more details needed in the text on the sequence read archive or refer to Table 1. Is the TE/repeat annotation library available for community use? Importantly, how were orthologous gene group trees rooted (needs mention in Methods)? How does your Starship identification align with the 'official' starfish method for discovering these?

Figures: a description of what the scale is describing is needed in the legend for all phylogenies.

I couldn't find Table S1!

— — — — —

Reviewer #3: The manuscript by Peck et al. analyzes the outbreaks caused by the coffee wilt disease agent Fusarium xylarioides. The authors take advantage of historical samples for a total of 13 strains to investigate the genetic structure and host associations. The authors indicate that virulence was acquired at least in part through horizontal transfer of effector genes. Roles in virulence were supported by transcriptomic analyses during infection.

The manuscript reports additional genomes for Fx over existing historical genome datasets and provides new dimensions of data analyses. However, much of the major claims of horizontal gene transfer and virulence components are not supported by robust analyses allowing for hypothesis testing. Hence, most of the manuscript remains largely descriptive and suggestive. The main claim stated in the title and elsewhere that horizontal transfer contributed to outbreaks, even successive outbreaks, lacks support on multiple levels.

Evidence for HTRs: the authors describe "putative HTRs" by identifying similar genomic regions between F. oxysporum and F. xylarioides. I am struggling to understand how the evidence for this was evaluated. Neither the main results nor Figure 4 provide direct support. Trees should include bootstrap values.

As Fo and Fx are closely related, how was the alternative explanation of incomplete lineage sorting evaluated?

The results should also more clearly state how cut-offs for similarity were defined using genome-wide average divergence, i.e. how much more similar are putative HTRs compared to the genome-wide average divergence?

Was similarity restricted to coding sequences or did similarity expand across the entire HTR including non-coding sequences? The issue of similarity analyses extends also to the Six gene analyses where it remains unclear whether the reported similarities are true outliers compared to expectations. Six genes are expected to diverge more rapidly. What divergence or similarity is hence considered "unusual", falling outside of the expectation for vertical inheritance? Also for HTR 4, the authors report 88% flanking sequence identity, 90% genome-wide identity and >94% defined as "high" identity. Are these true outliers / what percentiles?

Do Starship regions suspected to be HTR share higher similarity than other regions? Can you provide more insights into the "overlap" of a Starship and HTR3? What is the suspected mechanism? That the Starship was the mechanism of HT?

Overall, without proper statistical support, it is difficult to discussing evidence for HTR in this system.

Upregulation of CAZymes and predicted effectors: This section suggests that there is an overlap between gene upregulation and candidate genes involved in virulence. I lack though a clear conclusion from the section. Was the aim to show that effector candidates tend to be upregulated in planta? Were effector candidates more likely to be upregulated? Furthermore, how is the statement "It appears that a subset of CAZyme families have been acquired..." related to the HTR section above? What HTR region is this? How strong is the evidence for HT there? Is the pattern of upregulation conserved among the homologous segments in the putative donor and recipient strain genomes?

Analyses of miniature impala elements and HTR: Here I also lack a clear statistical support whether e.g. mimps are more likely to be putatively horizontally transferred. The authors simply refer to the fact that some mimps were found in HTRs.

Is "shared" TEs between Fx and Fo to be understood as being horizontally transferred or shared by recent, common descent? What statistical analyses support either of these interpretations?

Additional comments:

- Figure 3B: The authors attribute the small amount of reticulation in the network to gene flow. However, this could be explained by a lack of resolution of the network and/or historic recombination. Later this possibility is mentioned Clarifying the reticulation pattern would be helpful.

- FST analyses among genetic groups seem inappropriate as it remains unclear whether each group indeed consists of a population of interbreeding individuals (a requirement for calculating FST). A group size of ~2-4 haploid individuals is also inappropriate to estimate allele frequencies (necessary for FST calculations).

- Discussing the existence of "populations" with such small group sizes of genetic clusters seems tenuous. Without further sampling efforts, terms like "lineage" or "genetic cluster" seem more appropriate.

- The Circos plot identifies identities spanning 0-1. What is the interpretation of similarities of e.g. 0.2? Isn't 20% similarity at the level of random expectation for sequence similarity? Similarities would typically only be reported for syntenic or aligned regions. So 70-80% might be the lower limit for this. Clarifying this in the legend would be helpful.

— — — — —

Reviewer #4: This manuscript "Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease" investigate the coffee wilt disease caused by the fungal pathogen Fusarium xylarioides focusing on the sub-Saharan Africa area. Authors sequenced and compared 13 genomes, including 11 Fx, F6 o and Fs, generated one Fx genome using long-read to be used as the Fx reference genome, and conducted comparative genomics and transcriptomics analysis. This is an important research topic.

Authors have done tremendous work, trying to answer at least 3 rather complex biological questions: 1) populations structure of Fx isolates 2) HGT, and 3) mechanisms of HGT. Unfortunately, I am not able to derive the same conclusions authors presented here based on the data presented. I believe more careful analysis that testing specific hypothesis proposed in the manuscript is necessary.

Specific comments.

First, authors concluded that F. xylarioides comprises at least four distinct lineages: one host-specific to Coffea arabica, one to C. canephora var. robusta, and two historic lineages isolated

from various Coffea species. The data presented in the manuscript doesn't support this.

* Figure 2a supports the monophyletic nature of the Fx isolates. 2b shows no differences among F. x genomes. Current phylogenetic analyses, including Figure 3 didn't provide strong evidence supporting the distinct affiliation with different coffee hosts.

* In addition, Figure 2 should use the single-copy genes instead of genes have members in all genome. Bootstrap values are needed on the phylogeny.

* Also a detailed table capture all Fx strains included in the analysis, not only sequenced strains, may help to see the relationships among these strains from both geological and historical perspectives.

Second: authors concluded that host-specificity appears to be acquired through horizontal transfer of effector genes from members of the F. oxysporum species complex by comparing a F. oxysporum tomato pathogen to the Fx coffee pathogen and identified regions on contigs 3, 6, 10, 11 and 13.

* Why not the F. oxysporum coffee pathogen?

* As presented in Figure 1 and stated in the manuscript, "HTRs were differentially present across the F. xylarioides populations". If these regions are important for Fx pathogenicity to coffee, why they are so different among different Fx isolates?

* Do these present/absent polymorphisms have any relationship to the Fx populations structure?

Third, authors presented the affiliation of Mimps between F. oxysporum genome and Fx. And concluded that facilitate HGT. Turning affiliation into facilitation is too big a leap.

Overall, authors added unsupported speculation in the discussion section. For instance, authors claim that "the F. xylarioides formae speciales proposed nearly two decades ago can now be recognised as genetically differentiated populations with different host-specificity". In the result section, authors stated that there are 4 distinct populations. Then

which strains will be classified as F. xylarioides f. sp. abyssiniae and F. xylarioides f. sp. canephorae, respectively?

PLoS Biol. 2024 Dec 5;22(12):e3002480. doi: 10.1371/journal.pbio.3002480.r003

Author response to Decision Letter 1

5 Aug 2024

Attachment

Submitted filename: Reviewer-comments-final.docx

pbio.3002480.s030.docx^{(67.7KB, docx)}

PLoS Biol. doi: 10.1371/journal.pbio.3002480.r004

Decision Letter 2

Melissa Vazquez Hernandez

11 Sep 2024

Dear Dr Peck,

Thank you for your patience while we considered your revised manuscript "Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease" for publication as a Research Article at PLOS Biology. This revised version of your manuscript has been evaluated by the PLOS Biology editors, the Academic Editor and the original reviewers, being Brenda Wingfield reviewer #1.

Based on the reviews and on our Academic Editor's assessment of your revision, we are likely to accept this manuscript for publication, provided you satisfactorily address the remaining points raised by the reviewers, however it is not necessary to perform the additional analyses suggested by Reviewer #3. Please also make sure to address the following data and other policy-related requests.

a) We would strongly recommend you to publish the reports. This would be an option once the study is accepted.

b) We do not have a word limit and we believe having all material and methods in the main text allows them to be more accessible for readers. Please place the Supplementary Material and Methods in the main text.

c) You may be aware of the PLOS Data Policy, which requires that all data be made available without restriction: http://journals.plos.org/plosbiology/s/data-availability. For more information, please also see this editorial: http://dx.doi.org/10.1371/journal.pbio.1001797

Please supply the numerical values either in the a supplementary file or as a permanent DOI’d deposition for the following figures:

Figure 3BC, 5, 6B, S3, S6ABC, S7B, S8D, S9

NOTE: the numerical data provided should include all replicates AND the way in which the plotted mean and errors were derived (it should not present only the mean/average values).

d) Please cite the location of the data clearly in all relevant main and supplementary Figure legends, e.g. “The data underlying this Figure can be found in S1 Data” or “The data underlying this Figure can be found in https://doi.org/10.5281/zenodo.XXXXX”

e) Please provide the tree files for Figures Fig 1C, 2, 3BC, 6A, 7B, S1, S2, S6ABCD, S7B, S8BC, S11

f) Please ensure that your Data Statement in the submission system accurately describes where your data can be found and is in final format, as it will be published as written there.

g) Per journal policy, if you have generated any custom code during the course of this investigation, please make it available without restrictions upon publication. Please ensure that the code is sufficiently well documented and reusable, and that your Data Statement in the Editorial Manager submission system accurately describes where your code can be found.

Please note that we cannot accept sole deposition of code in GitHub, as this could be changed after publication. However, you can archive this version of your publicly available GitHub code to Zenodo. Once you do this, it will generate a DOI number, which you will need to provide in the Data Accessibility Statement (you are welcome to also provide the GitHub access information). See the process for doing this here: https://docs.github.com/en/repositories/archiving-a-github-repository/referencing-and-citing-content

As you address these items, please take this last chance to review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the cover letter that accompanies your revised manuscript.

We expect to receive your revised manuscript within two weeks.

To submit your revision, please go to https://www.editorialmanager.com/pbiology/ and log in as an Author. Click the link labelled 'Submissions Needing Revision' to find your submission record. Your revised submission must include the following:

- a cover letter that should detail your responses to any editorial requests, if applicable, and whether changes have been made to the reference list

- a Response to Reviewers file that provides a detailed response to the reviewers' comments (if applicable, if not applicable please do not delete your existing 'Response to Reviewers' file.)

- a track-changes file indicating any changes that you have made to the manuscript.

NOTE: If Supporting Information files are included with your article, note that these are not copyedited and will be published as they are submitted. Please ensure that these files are legible and of high quality (at least 300 dpi) in an easily accessible file format. For this reason, please be aware that any references listed in an SI file will not be indexed. For more information, see our Supporting Information guidelines:

https://journals.plos.org/plosbiology/s/supporting-information

*Published Peer Review History*

Please note that you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out. Please see here for more details:

https://blogs.plos.org/plos/2019/05/plos-journals-now-open-for-published-peer-review/

*Press*

Should you, your institution's press office or the journal office choose to press release your paper, please ensure you have opted out of Early Article Posting on the submission form. We ask that you notify us as soon as possible if you or your institution is planning to press release the article.

*Protocols deposition*

Please do not hesitate to contact me should you have any questions.

Sincerely,

Melissa

Melissa Vazquez Hernandez, Ph.D.

Associate Editor

mvazquezhernandez@plos.org

PLOS Biology

------------------------------------------------------------------------

REVIEWERS' COMMENTS:

------------------------------------------------------------------------

Reviewer #1:

please correct in the supplementary table 2 - F. xylarioides 379925 and L0394 are written in italics

------------------------------------------------------------------------

Reviewer #2:

I have no comments on the revision - all my comments were addressed in the previous revision round. I recommend accept

------------------------------------------------------------------------

Reviewer #3:

The authors have conducted extensive new analyses and re-analyses in an attempt to support many of the claims made in the previous version.

Notable improvements include analyzing a more complete set of the already existing genomes and evaluating alternative hypotheses to the claim of horizontal gene transfer.

The new version now also lacks several analyses that were not replaced with more appropriate investigations (e.g. population differentiation analyses, etc.).

The evidence for horizontal gene transfer is now clearly and more transparently laid out. The authors primarily can show that the regions labeled as HT are outliers in the genome for their phylogenetic signal (or similarity patterns), which is consistent with HT. Rare incomplete sorting would also be an outlier obviously. I take the finding of a Starship in one of the HT regions as an encouraging sign though. This would indeed provide a mechanism for how DNA of common descent can appear in distinct lineages outside of vertical transmission.

Finally, I remain confused about what "identities" spanning from 0-1 represent. The tool that was used analyzes mapped read coverage but not what is typically called "sequence identity" or homology. If sequence coverage is shown by the 0-1 range, then the term "identity" is wrong. This would simply reflect that there were alignable reads without any statement about how perfectly this read aligned (i.e- how identical the two genomes are at this locus).

PLoS Biol. 2024 Dec 5;22(12):e3002480. doi: 10.1371/journal.pbio.3002480.r005

Author response to Decision Letter 2

25 Sep 2024

Attachment

Submitted filename: Reviewer-comments-final.docx

pbio.3002480.s031.docx^{(67.7KB, docx)}

PLoS Biol. doi: 10.1371/journal.pbio.3002480.r006

Decision Letter 3

Melissa Vazquez Hernandez

30 Sep 2024

Dear Dr Peck,

Thank you for the submission of your revised Research Article "Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease" for publication in PLOS Biology. On behalf of my colleagues and the Academic Editor, Sophien Kamoun, I am pleased to say that we can in principle accept your manuscript for publication, provided you address any remaining formatting and reporting issues. These will be detailed in an email you should receive within 2-3 business days from our colleagues in the journal operations team; no action is required from you until then. Please note that we will not be able to formally accept your manuscript and schedule it for publication until you have completed any requested changes.

Please take a minute to log into Editorial Manager at http://www.editorialmanager.com/pbiology/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with biologypress@plos.org. If you have previously opted in to the early version process, we ask that you notify us immediately of any press plans so that we may opt out on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

Thank you again for choosing PLOS Biology for publication and supporting Open Access publishing. We look forward to publishing your study.

Sincerely,

Melissa

Melissa Vazquez Hernandez, Ph.D., Ph.D.

Associate Editor

PLOS Biology

mvazquezhernandez@plos.org

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Fig. The Fusarium xylarioides and F. phyllophylum clade of the supermatrix species tree based on 1,685 concatenated single-copy orthogroups.

(TIFF)

pbio.3002480.s001.tiff^{(49.9MB, tiff)}

S2 Fig. STAG supergene species tree based on 3,544 orthogroups in which every strain has at least 1 gene copy.

(TIFF)

pbio.3002480.s002.tiff^{(64.1MB, tiff)}

S3 Fig. Whole-genome similarity between Fusarium oxysporum and Fusarium fujikuroi species complexes.

(TIFF)

pbio.3002480.s003.tiff^{(24.8MB, tiff)}

S4 Fig. Gene sharing across the Fusarium xylarioides host-specific populations.

Orthogroups shared between F. xylarioides arabica, robusta, and the 2 coffea clusters. Drawn 16,006 (excluding 10,567 that were absent from F. xylarioides.

(TIFF)

pbio.3002480.s004.tiff^{(11.1MB, tiff)}

S5 Fig

(TIFF)

pbio.3002480.s005.tiff^{(68.8MB, tiff)}

S6 Fig. PHYML phylogenetic trees for HTRs 3, 4, 5, and 6.

(TIFF)

pbio.3002480.s006.tiff^{(92.2MB, tiff)}

S7 Fig. Different parts of Fusarium oxysporum f. sp. lycopersici supercontig 51 is shared with Fusarium xylarioides populations.

(TIFF)

pbio.3002480.s007.tiff^{(79.1MB, tiff)}

S8 Fig. Putative horizontal transfer 1 (HTR 1) between Fusarium oxysporum and Fusarium xylarioides.

(TIFF)

pbio.3002480.s008.tiff^{(52.5MB, tiff)}

S9 Fig. Differential expression of genes and putative effector genes in Fusarium xylarioides.

(TIFF)

pbio.3002480.s009.tiff^{(96.8MB, tiff)}

S10 Fig. Whole-genome view of the loci of transposable elements in the F. xylarioides arabica563 reference genome.

(TIFF)

pbio.3002480.s010.tiff^{(77.4MB, tiff)}

S11 Fig. Miniature impala (mimp) family 1 sequence genealogy.

(TIFF)

pbio.3002480.s011.tiff^{(40.4MB, tiff)}

S1 Table. Bio-geographic details for strains sequenced in this study.

(PDF)

pbio.3002480.s012.pdf^{(63.2KB, pdf)}

S2 Table. Published genomes analysed in this study.

(PDF)

pbio.3002480.s013.pdf^{(65.2KB, pdf)}

S3 Table. Population genetics statistics related to genetic differentiation (d_xy) in four Fusarium xylarioides populations.

Measured in 100 kb windows.

(PDF)

pbio.3002480.s014.pdf^{(75.3KB, pdf)}

(PDF)

pbio.3002480.s015.pdf^{(75.3KB, pdf)}

S5 Table. Predicted effectors in arabica563 genes by EffectorP [34].

(PDF)

pbio.3002480.s016.pdf^{(67KB, pdf)}

S6 Table. Expression data for all putative effectors from [28] across the Fusarium xylarioides arabica563 samples.

Each gene is described as up-regulated in planta (“coffee.up”), in axenic (“culture.up”), or not differentially expressed (“ns”).

(PDF)

pbio.3002480.s017.pdf^{(64.9KB, pdf)}

S7 Table. Expression data for all putative effectors from [28] across the Fusarium xylarioides arabica908 samples.

Each gene is described as up-regulated in planta (“coffee.up”), in axenic (“culture.up”), or not differentially expressed (“ns”).

(PDF)

pbio.3002480.s018.pdf^{(64.6KB, pdf)}

S8 Table. The number of genes expressed for each carbohydrate-active enzyme subfamily in Fusarium xylarioides arabica563.

The in planta up-regulated genes are shaded, where the warmest colours represent the highest up-regulated gene count.

(PDF)

pbio.3002480.s019.pdf^{(61.8KB, pdf)}

S9 Table. The number of genes expressed for each carbohydrate-active enzyme subfamily in Fusarium xylarioides arabica908.

The in planta up-regulated genes are shaded, where the warmest colours represent the highest up-regulated gene count.

(PDF)

pbio.3002480.s020.pdf^{(61.4KB, pdf)}

S10 Table. Gene count and description for significantly enriched InterPro terms in the Coffea arabica in planta samples.

(PDF)

pbio.3002480.s021.pdf^{(88.7KB, pdf)}

S11 Table. Transposable element density in horizontally transferred regions (HTR), Starships and genome-wide in the Fusarium xylarioides arabica563 reference.

(PDF)

pbio.3002480.s022.pdf^{(47.6KB, pdf)}

S12 Table. The contigs in the Fusarium xylarioides arabica563 reference genome which correspond to each chromosome in the F. verticillioides assembly [21].

(PDF)

pbio.3002480.s023.pdf^{(47.6KB, pdf)}

S13 Table. Comparison of divergence patterns for HTR regions to those of background genome regions.

(PDF)

pbio.3002480.s024.pdf^{(58KB, pdf)}

S14 Table. Comparison of divergence patterns for HTR regions to those of background genome regions.

(PDF)

pbio.3002480.s025.pdf^{(58KB, pdf)}

S1 File. Genomic loci for the 6 putative horizontally transferred regions and the F. xylarioides strain in which they were identified.

This file can be opened in a text editor program.

(BED)

pbio.3002480.s026.bed^{(292B, bed)}

S2 File. Nucleotide sequences for the 6 putative horizontally transferred regions from the F. xylarioides strain in which they were identified.

This file can be opened in a text editor program.

(FA)

pbio.3002480.s027.fa^{(437KB, fa)}

S3 File. Predicted gene annotations and functions for Starship cargo and those annotations which flank the Starship.

(XLSX)

pbio.3002480.s028.xlsx^{(11KB, xlsx)}

S4 File. Sequence read archive (SRA) accessions and data counts for filtered data.

(XLSX)

pbio.3002480.s029.xlsx^{(14.1KB, xlsx)}

Attachment

Submitted filename: Reviewer-comments-final.docx

pbio.3002480.s030.docx^{(67.7KB, docx)}

Attachment

Submitted filename: Reviewer-comments-final.docx

pbio.3002480.s031.docx^{(67.7KB, docx)}

Data Availability Statement

[pbio.3002480.ref001] 1.Fones HN, Fisher MC, Gurr SJ. Emerging fungal threats to plants and animals challenge agriculture and ecosystem resilience. Microbiol Spectr. 2017;5:787–809. doi: 10.1128/microbiolspec.FUNK-0027-2016 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref002] 2.Fisher MC, Henk DA, Briggs CJ, Brownstein JS, Madoff LC, McCraw SL, et al. Emerging fungal threats to animal, plant and ecosystem health. Nature. 2012;484:186–194. doi: 10.1038/nature10947 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref003] 3.FAO. FAOSTAT data May 2022. Available from: www.fao.org/faostat. [Google Scholar]

[pbio.3002480.ref004] 4.USDA. Coffee Annual, Ethiopia tech. rep. (United States Dept of Agriculture, Foreign Agricultural Service, 2023), 1–7.

[pbio.3002480.ref005] 5.FAO. Statistical Pocket Book of the Food And Agricultural Organization for the United Nations: Coffee 2015 tech. rep. 2015. [Google Scholar]

[pbio.3002480.ref006] 6.Flood J. in Coffee Wilt Disease (ed Flood J.) chap. 1 (CAB International, Wallingford, UK, 2009). [Google Scholar]

[pbio.3002480.ref007] 7.Ristaino J B, Anderson PK, Bebber DP, Brauman KA, Cunniffe NJ, Fedoroff NV, et al. The persistent threat of emerging plant disease pandemics to global food security. Proc Natl Acad Sci U S A. 2021;118:e2022239118. doi: 10.1073/pnas.2022239118 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref008] 8.Rutherford M. Current knowledge of coffee wilt disease, a major constraint to coffee production in Africa. Phytopathology. 2006;96:663–666. doi: 10.1094/PHYTO-96-0663 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref009] 9.Yadeta KA, Thomma BPHJ. The xylem as battleground for plant hosts and vascular wilt pathogens. Front Plant Sci. 2013;4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref010] 10.Phiri N, Baker P. Coffee Wilt in Africa: Final Technical Report of the Regional Coffee Wilt Programme (2000–2007) tech. rep. (CABI, Wallingford, UK: CAB International, 2009), 233.

[pbio.3002480.ref011] 11.Peck LD, Boa E. Coffee wilt disease: The forgotten threat to coffee. Plant Pathol. 2023;73:506–521. doi: 10.1111/ppa.13833 [DOI] [Google Scholar]

[pbio.3002480.ref012] 12.Van Der Graaff NA, Pieters R. Resistance levels in Coffea arabica to Gibberella xylarioides and distribution pattern of the disease. Neth J Pl Path. 1978;84:117–120. [Google Scholar]

[pbio.3002480.ref013] 13.Mulatu A, Shanko D. Incidence and prevalence of coffee wilt disease (Gibberella xylarioides) and its impact on the rural livelihoods in western Guji zone, southern Ethiopia. Am J Biosci. 2019;7:7–15. [Google Scholar]

[pbio.3002480.ref014] 14.Bredeweg EL, Baker SE. Horizontal gene transfer in fungi. Grand Challenges in Biology and Biotechnology. 2020;317–332. [Google Scholar]

[pbio.3002480.ref015] 15.McDonald MC, Taranto AP, Hill E, Schwessinger B, Liu Z, Simpfendorfer S, et al. Transposon-mediated horizontal transfer of the host-specific virulence protein ToxA between three fungal wheat pathogens. MBio. 2019;10:01515–01519. doi: 10.1128/mBio.01515-19 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref016] 16.Gourlie R, McDonald M, Hafez M, Ortega-Polo R, Low KE, Abbott DW, et al. The pangenome of the wheat pathogen Pyrenophora tritici-repentis reveals novel transposons associated with necrotrophic effectors ToxA and ToxB. BMC Biol. 2022;20. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref017] 17.Gluck-Thaler E, Ralston T, Konkel Z, Ocampos CG, Ganeshan VD, Dorrance AE, et al. Giant Starship elements mobilize accessory genes in fungal genomes. Mol Biol Evol. 2022;39. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref018] 18.Gluck-Thaler E, Forsythe A, Puerner C, Stajich JE, Croll D, Cramer RA, et al. Giant transposons promote strain heterogeneity in a major fungal pathogen. Biorxiv. 2024. doi: 10.1101/2024.06.28.601215 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref019] 19.Urquhart AS, Vogan AA, Gardiner DM, Idnurm A. Starships are active eukaryotic transposable elements mobilized by a new family of tyrosine recombinases. Proc Natl Acad Sci U S A. 2023;120:e2214521120. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref020] 20.Michielse CB, Rep M. Pathogen profile update: Fusarium oxysporum. Mol Plant Pathol. 2009;10:311–324. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref021] 21.Ma LJ, Van Der Does HC, Borkovich KA, Coleman JJ, Daboussi MJ, Di Pietro A, et al. Comparative genomics reveals mobile pathogenicity chromosomes in Fusarium. Nature. 2010;464:367–373. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref022] 22.van Dam P, de Sain M, ter Horst A, van der Gragt M, Rep M. Use of comparative genomics-based markers for discrimination of host specificity in Fusarium oxysporum. Appl Environ Microbiol. 2018;84:1868–1885. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref023] 23.Li J, Fokkens L, Rep M. A single gene in Fusarium oxysporum limits host range. Mol Plant Pathol. 2020;1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref024] 24.Rep M, van der Does HC, Meijer M, van Wijk R, Houterman PM, Dekker HL, et al. A small, cysteine-rich protein secreted by Fusarium oxysporum during colonization of xylem vessels is required for I-3-mediated resistance in tomato. Mol Microbiol. 2004;53:1373–1383. [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref025] 25.Schmidt SM, Houterman PM, Schreiver I, Ma L, Amyotte S, Chellappan B, et al. MITEs in the promoters of effector genes allow prediction of novel virulence genes in Fusarium oxysporum. BMC Genomics. 2013;14:1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref026] 26.Faino L, Seidl MF, Shi-Kunne X, Pauper M, van den Berg GCM, Wittenberg AHJ, et al. Transposons passively and actively contribute to evolution of the two speed genome of a fungal pathogen. Genome Res. 2016;26:1091–1100. doi: 10.1101/gr.204974.116 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref027] 27.Farrer RA, Fisher MC. Describing genomic and epigenomic traits underpinning emerging fungal pathogens. Adv Genet. 2017;100:73–140. ISSN: 0065-2660. doi: 10.1016/bs.adgen.2017.09.009 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref028] 28.Peck LD, Nowell RW, Flood J, Ryan MR, Barraclough TG. Historical genomics reveals the evolutionary mechanisms behind multiple outbreaks of the host specific coffee wilt pathogen Fusarium xylarioides. BMC Genomics. 2021;22:1–24. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref029] 29.Kvas M, Marasas W, Wingfield B, Wingfield M, Steenkamp E. Diversity and evolution of Fusarium species in the Gibberella fujikuroi complex. Fungal Divers. 2009;34:1–21. [Google Scholar]

[pbio.3002480.ref030] 30.Oduor G. et al. Surveys to assess the extent and impact of coffee wilt disease in East and Central Africa. Final Technical Report. tech. rep. (CAB International, Nairobi, Kenya, 2003).

[pbio.3002480.ref031] 31.van Westerhoven AC, Aguilera-Galvez C, Nakasato-Tagami G, Shi-Kunne X, Martinez de la Parte E, Chavarro-Carrero E, et al. Segmental duplications drive the evolution of accessory regions in a major crop pathogen. New Phytol. 2024;242:610–625. doi: 10.1111/nph.19604 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref032] 32.van Dam P, Rep M. The distribution of miniature Impala elements and SIX genes in the Fusarium genus is suggestive of horizontal gene transfer. J Mol Evol. 2017;85:14–25. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref033] 33.Rogers LM, Kim Y-K, Guo W, Gonzá Lez-Candelas L, Li D, Kolattukudy PE, et al. Requirement for either a host-or pectin-induced pectate lyase for infection of Pisum sativum by Nectria hematococca. Proc Natl Acad Sci U S A. 2000;97. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref034] 34.Sperschneider J, Dodds PN. EffectorP 3.0: Prediction of apoplastic and cytoplasmic effectors in fungi and oomycetes. Mol Plant Microbe Interact. 2022;146–156. [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref035] 35.Bergemann M, Lespinet O, M’Barek SB, Daboussi MJ, Dufresne M. Genome wide analysis of the Fusarium oxysporum mimp family of MITEs and mobilization of both native and de novo created mimps. J Mol Evol. 2008;67:631–642. [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref036] 36.Giraud T, Gladieux P, Gavrilets S. Linking the emergence of fungal plant diseases with ecological speciation. Trends Ecol Evol. 2010;25:387–395. doi: 10.1016/j.tree.2010.03.006 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref037] 37.Rutherford M, Bieysse D, Lepoint P, Maraite H. in Coffee Wilt Disease (ed Flood J.) 99–119 (CAB International, Wallingford, UK, 2009). [Google Scholar]

[pbio.3002480.ref038] 38.Lepoint, P. C. E. Speciation within the African Coffee Wilt Pathogen PhD thesis (Universite Catholique de Louvain., 2006), 208.

[pbio.3002480.ref039] 39.Girma A. Diversity in Pathogenicity and Genetics of Gibberella xylarioides (Fusarium xylarioides) Populations and Resistance of Coffea spp. in Ethiopia PhD thesis (Rheinischen Friedrich-Wilhelms-Universitat zu Bonn, 2004), 85.

[pbio.3002480.ref040] 40.Kurian SM, Di Pietro A, Read ND. Live-cell imaging of conidial anastomosis tube fusion during colony initiation in Fusarium oxysporum. PLoS ONE. 2018;13(ed Freitag M.) e0195634. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref041] 41.Lepoint PCE, Munaut FTJ, Maraite HMM. Gibberella xylarioides sensu lato from Coffea canephora: a new mating population in the Gibberella fujikuroi species complex. Appl Environ Microbiol. 2005;71:8466–8471. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref042] 42.Ropars J, Didiot E, Rodríguez de la Vega RC, Bennetot B, Coton M, Poirier E, et al. Domestication of the emblematic white cheese-making fungus Penicillium camemberti and its diversification into two varieties. Curr Biol. 2020;30:4441–4453. [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref043] 43.Rouxel T, Grandaubert J, Hane JK, Hoede C, van de Wouw AP, Couloux A, et al. Effector diversification within compartments of the Leptosphaeria maculans genome affected by Repeat-Induced Point mutations. Nat Commun. 2011;2:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref044] 44.Guo L, Han L, Yang L, Zeng H, Fan D, Zhu Y, et al. Genome and transcriptome analysis of the fungal pathogen Fusarium oxysporum f. sp. cubense causing banana vascular wilt disease. PLoS ONE. 2014;9:e95543. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref045] 45.Brankovics B, van Diepeningen AD, de Hoog GS, van der Lee TA, Waalwijk C. Detecting introgression between members of the Fusarium fujikuroi and F. oxysporum species complexes by comparative mitogenomics. Front Microbiol. 2020;11. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref046] 46.Flood J. Epidemiology and variability of Gibberella xylarioides, the coffee wilt pathogen. R8188 (ZA0505). Final Technical Report tech. rep. (CAB International, Wallingford, UK, 2005).

[pbio.3002480.ref047] 47.Serani S, Taligoola HK, Hakiza GJ. An investigation into Fusarium spp. associated with coffee and banana plants as potential pathogens of robusta coffee. Afr J Ecol. 2007;45:91–95. [Google Scholar]

[pbio.3002480.ref048] 48.O’Donnell K, Kistler HC, Cigelnik E, Ploetz RC. Multiple evolutionary origins of the fungus causing Panama disease of banana: Concordant evidence from nuclear and mitochondrial gene genealogies. Proc Natl Acad Sci U S A. 1998;95:2044–2049. doi: 10.1073/pnas.95.5.2044 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref049] 49.Raffaele S, Farrer RA, Cano LM, Studholme DJ, MacLean D, Thines M, et al. Genome Evolution Following host jumps in the Irish potato famine pathogen lineage. Science. 2010;330:1540–1543. doi: 10.1126/science.1193070 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref050] 50.Ropars J, Rodríguez De La Vega RC, Ló Pez- M, Dupont L, Branca A, Giraud T. Adaptive horizontal gene transfers between multiple cheese-associated fungi. Curr Biol. 2015;25:2562–2569. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref051] 51.Richards TA, Soanes DM, Jones MDM, Vasieva O, Leonard G, Paszkiewicz K, et al. Horizontal gene transfer facilitated the evolution of plant parasitic mechanisms in the oomycetes. Proc Natl Acad Sci U S A. 2011;108:15258–15263. doi: 10.1073/pnas.1105100108 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref052] 52.Van Etten J, Bhattacharya D. Horizontal gene transfer in eukaryotes: not if, but how much? Trends Genet. 2020;36:915–925. doi: 10.1016/j.tig.2020.08.006 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref053] 53.Brewer GE, Clarkson JJ, Maurin O, Zuntini AR, Barber V, Bellot S, et al. Factors affecting targeted sequencing of 353 nuclear genes from herbarium specimens spanning the diversity of angiosperms. Front Plant Sci. 2019;10:1102. doi: 10.3389/fpls.2019.01102 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref054] 54.Ryan MJ, Peck LD, Smith D, Flood J, Barraclough TG. Culture collections as a source of historic strains for genomic studies in plant pathology. J Plant Pathol. 2022. [Google Scholar]

[pbio.3002480.ref055] 55.Peck LD. Seventy years of Fusarium wilt—coffee interactions: historical genomics reveals pathogen emergence and divergence PhD thesis (Imperial College London, London, 2023). [Google Scholar]

[pbio.3002480.ref056] 56.Spanu PD, Abbott JC, Amselem J, Burgis TA, Soanes DM, Stüber K, et al. Genome expansion and gene loss in powdery mildew fungi reveal tradeoffs in extreme parasitism. Science. 2010;330:1543–1546. doi: 10.1126/science.1194573 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref057] 57.Kolmogorov M, Yuan J, Lin Y, Pevzner PA. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019;37:540–546. doi: 10.1038/s41587-019-0072-8 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref058] 58.Aury JM, Istace B. Hapo-G, haplotype-aware polishing of genome assemblies with accurate reads. NAR Genom Bioinform. 2021;3. doi: 10.1093/nargab/lqab034 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref059] 59.Davey JW, Davis SJ, Mottram JC, Ashton PD. Tapestry: validate and edit small eukaryotic genome assemblies with long reads. bioRxiv. 2020. [Google Scholar]

[pbio.3002480.ref060] 60.Marçais G, Delcher AL, Phillippy AM, Coston R, Salzberg SL, Zimin A. MUMmer4: A fast and versatile genome alignment system. PLoS Comput Biol. 2018;14:e1005944. doi: 10.1371/journal.pcbi.1005944 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref061] 61.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17:10–12. [Google Scholar]

[pbio.3002480.ref062] 62.Magoč T, Salzberg SL. FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics. 2011;27:2957–2963. doi: 10.1093/bioinformatics/btr507 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref063] 63.Li D, Liu CM, Luo R, Sadakane K, Lam TW. MEGAHIT: An ultra-fast singlenode solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics. 2015;31:1674–1676. [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref064] 64.Li D, Luo R, Liu CM, Leung CM, Ting HF, Sadakane K, et al. MEGAHIT v1.0: A fast and scalable metagenome assembler driven by advanced methodologies and community practices. Methods. 2016;102:3–11. doi: 10.1016/j.ymeth.2016.02.020 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref065] 65.Mikheenko A, Prjibelski A, Saveliev V, Antipov D, Gurevich A. Versatile genome assembly evaluation with QUAST-LG. Bioinformatics. 2018;34:142–150. doi: 10.1093/bioinformatics/bty266 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref066] 66.Olal S, Bitalo DN, Olango ND, Mulindwa J, Ochwo S, Opiyo SO, et al. novo genome sequence of a Fusarium xylarioides race pathogenic to Robusta coffee (Coffea canephora) in Uganda. Microbiol Resour Announc. 2019;8:e00520–e00519. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref067] 67.Wingfield BD, Fourie A, Simpson MC, Bushula-Njah VS, Aylward J, Barnes I, et al. IMA Genome-F 11. IMA Fungus. 2019;10. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref068] 68.Alonge M, Soyk S, Ramakrishnan S, Wang X, Goodwin S, Sedlazeck FJ, et al. RaGOO: Fast and accurate reference-guided scaffolding of draft genomes. Genome Biol. 2019;20:224. doi: 10.1186/s13059-019-1829-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref069] 69.Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol. 2011;29:644–652. doi: 10.1038/nbt.1883 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref070] 70.Palmer JM, Stajich J. Funannotate v1.8.1: Eukaryotic genome annotation. 2020. [Google Scholar]

[pbio.3002480.ref071] 71.Shumate A, Salzberg SL. Liftoff: accurate mapping of gene annotations. Bioinformatics. 2021;37:1639–1643. doi: 10.1093/bioinformatics/btaa1016 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref072] 72.Ma L-J, Zhang Y, Li C, Liu S, Liu C, Mostert D, et al. Accessory genes in tropical race 4 contributed to the recent resurgence of the devastating disease of Fusarium wilt of banana. Res Sq. 2023. [Google Scholar]

[pbio.3002480.ref073] 73.Mesny F, Miyauchi S, Thiergart T, Pickel B, Atanasova L, Karlsson M, et al. Genetic determinants of endophytism in the Arabidopsis root mycobiome. Nat Commun. 2021;12. doi: 10.1038/s41467-021-27479-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref074] 74.Flynn JM, Hubley R, Goubert C, Rosen J, Clark AG, Feschotte C, et al. RepeatModeler2: automated genomic discovery of transposable element families. Proc Natl Acad Sci U S A. 2020;117:9451–9457. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref075] 75.Goubert C, Craig RJ, Bilat AF, Peona V, Vogan AA, Protasio A, et al. A beginner’s guide to manual curation of transposable elements. Mob DNA. 2022;13:1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref076] 76.Notredame C, Higgins DG, Heringa J. T-Coffee: A novel method for fast and accurate multiple sequence alignment. J Mol Biol. 2000;302:205–217. doi: 10.1006/jmbi.2000.4042 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref077] 77.Smit AFA, Hubley R, Green P. RepeatMasker Open-4.0. 2015. [Google Scholar]

[pbio.3002480.ref078] 78.Katoh K, Standley DM. MAFFT Multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30:772–780. doi: 10.1093/molbev/mst010 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref079] 79.Katoh K, Misawa K, Kuma K, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002;30:3059–3066. doi: 10.1093/nar/gkf436 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref080] 80.Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, et al. BLAST+: architecture and applications. BMC Bioinform. 2009;10:1–9. doi: 10.1186/1471-2105-10-421 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref081] 81.Hoang DT, Chernomor O, von Haeseler A, Minh BQ, Vinh LS. UFBoot2: Improving the Ultrafast Bootstrap Approximation. Mol Biol Evol. 2018;35:518–522. doi: 10.1093/molbev/msx281 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref082] 82.Novichkov V, Kaznadzey A, Alexandrova N, Kaznadzey D. NSimScan: DNA comparison tool with increased speed, sensitivity and accuracy. Bioinformatics (Oxford, England). 2016;32:2380–2381. doi: 10.1093/bioinformatics/btw126 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref083] 83.Emms DM, Kelly S. OrthoFinder: Phylogenetic orthology inference for comparative genomics. Genome Biol. 2019;20:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref084] 84.Kelly S, Maini PK. DendroBLAST: Approximate phylogenetic trees in the absence of multiple sequence alignments. PLoS ONE. 2013;8:e58537. doi: 10.1371/journal.pone.0058537 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref085] 85.Emms DM, Kelly S. OrthoFinder: solving fundamental biases in whole genome comparisons dramatically improves orthogroup inference accuracy. Genome Biol. 2015;16:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref086] 86.Emms DM, Kelly S. STAG: species tree inference from all genes. BioRxiv. 2018;1–29. [Google Scholar]

[pbio.3002480.ref087] 87.Emms D, Kelly S. STRIDE: species tree root inference from gene duplication events. Mol Biol Evol. 2017;34:3267–3278. doi: 10.1093/molbev/msx259 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref088] 88.Fujisawa T, Aswad A, Barraclough TG. A rapid and scalable method for multilocus species delimitation using Bayesian model comparison and rooted triplets. Syst Biol. 2016;65:759–771. doi: 10.1093/sysbio/syw028 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref089] 89.Shen W, Le S, Li Y, Hu F. SeqKit: A cross-platform and ultrafast toolkit for FASTA/Q file manipulation. PLoS ONE. 2016;11:e0163962. doi: 10.1371/journal.pone.0163962 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref090] 90.Huson DH, Bryant D. Application of Phylogenetic Networks in Evolutionary Studies. Mol Biol Evol. 2006;23(2):254–267. doi: 10.1093/molbev/msj030 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref091] 91.Pfeifer B, Wittelsburger U, Ramos-Onsins SE, Lercher MJ. PopGenome: An efficient swiss army knife for population genomic analyses in R. Mol Biol Evol. 2014;31:1929–1936. doi: 10.1093/molbev/msu136 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref092] 92.Li H. Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018;34:3094–3100. doi: 10.1093/bioinformatics/bty191 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref093] 93.Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, et al. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol. 2020;37:1530–1534. https://academic.oup.com/mbe/article/37/5/1530/ 5721363. doi: 10.1093/molbev/msaa015 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref094] 94.Hao Z, Lv D, Ge Y, Shi J, Weijers D, Yu G, et al. RIdeogram: Drawing SVG graphics to visualize and map genome-wide data on the idiograms. PeerJ Comput Sci. 2020;6:1–11. ISSN: 23765992. doi: 10.7717/peerj-cs.251 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref095] 95.Gluck-Thaler E, Vogan AA. Systematic identification of cargo-carrying genetic elements reveals new dimensions of eukaryotic diversity. Biorxiv. 2023. doi: 10.1101/2023.10.24.563810 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref096] 96.Levy Karin E, Mirdita M, Soding J. MetaEuk—sensitive, high-throughput gene¨ discovery, and annotation for large-scale eukaryotic metagenomics. Microbiome. 2020;8:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref097] 97.Eddy SR. Accelerated profile HMM searches. PLoS Comput Biol. 2011;7:e1002195. doi: 10.1371/journal.pcbi.1002195 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref098] 98.Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26:841–842. doi: 10.1093/bioinformatics/btq033 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref099] 99.Ayad LAK, Pissis SP, Polychronopoulos D. CNEFinder: finding conserved noncoding elements in genomes. Bioinformatics. 2018;34:i743–i747. doi: 10.1093/bioinformatics/bty601 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref100] 100.Pierce NT, Irber L, Reiter T, Brooks P, Brown CT. Large-scale sequence comparisons with sourmash. F1000Res. 2019;8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref101] 101.Enright AJ, Van Dongen S, Ouzounis CA. An efficient algorithm for large-scale detection of protein families. Nucleic Acids Res. 2002;30:1575–1584. doi: 10.1093/nar/30.7.1575 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref102] 102.Hackl T, Ankenbrand MJ, van Adrichem B. gggenomes: A Grammar of Graphics for Comparative Genomics. 2024. [Google Scholar]

[pbio.3002480.ref103] 103.Geiser DM, Lewis Ivey ML, Hakiza G, Juba JH, Miller SA. Gibberella xylarioides (anamorph: Fusarium xylarioides), a causative agent of coffee wilt disease in Africa, is a previously unrecognized member of the G. fujikuroi species complex. Mycologia. 2005;97:191–201. [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref104] 104.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30:2114. doi: 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref105] 105.Andrews S, Krueger F, Seconds-Pichon A, Biggins F, Wingett S. FastQC. A Quality Control tool for High Throughput Sequence Data. Babraham Bioinformatics. 2015. [Google Scholar]

[pbio.3002480.ref106] 106.Zimin A, Yepes M, Maldonado CE, Navarro L, Kovaka S, Pertea M, et al. Coffea arabica variety Caturra and C. eugenioides Genome Assembly and Annotation in Plant and Animal Genome Conference XXVII. 2019. [Google Scholar]

[pbio.3002480.ref107] 107.Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017;14:417–419. doi: 10.1038/nmeth.4197 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref108] 108.Haas BJ, Papanicolaou A, Yassour M, Grabherr M, Blood PD, Bowden J, et al. De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nat Protoc. 2013:1494–1512. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref109] 109.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:1–21. doi: 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref110] 110.Chen Y, Yin H, Gao M, Zhu H, Zhang Q, Wang Y. Comparative transcriptomics atlases reveals different gene expression pattern related to Fusarium wilt disease resistance and susceptibility in two Vernicia species. Front Plant Sci. 2016;7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref111] 111.Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43:e47–e47. doi: 10.1093/nar/gkv007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref112] 112.Soneson C, Robinson MD. Bias, robustness and scalability in single-cell differential expression analysis. Nat Methods. 2018;15:255–261. doi: 10.1038/nmeth.4612 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref113] 113.Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc B Methodol. 1995;57:289–300. [Google Scholar]

[pbio.3002480.ref114] 114.Cantarel BI, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B. The Carbohydrate-Active enZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res. 2009;37. doi: 10.1093/nar/gkn663 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref115] 115.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria, 2021.

[pbio.3002480.ref116] 116.Wickham H, Averick M, Bryan J, Chang WDL, Mcgowan A, François R, et al. Welcome to the Tidyverse. J Open Source Softw. 2019;4:1686. [Google Scholar]

[pbio.3002480.ref117] 117.Wilke C. Cowplot: Streamlined Plot Theme and Plot Annotations for ‘ggplot2’. 2022. [Google Scholar]

[pbio.3002480.ref118] 118.Slowikowski K. ggrepel: Automatically Position Non-Overlapping Text Labels with ‘ggplot2’. Comprehensive R Archive Network (CRAN). 2021. [Google Scholar]

[pbio.3002480.ref119] 119.Ram K, Wickham H. wesanderson: A Wes Anderson Palette Generator Comprehensive R Archive Network (CRAN). 2018. [Google Scholar]

[pbio.3002480.ref120] 120.Kolde R. pheatmap: Pretty Heatmaps Comprehensive R Archive Network (CRAN). 2019. [Google Scholar]

[pbio.3002480.ref121] 121.Yu G, Smith DK, Zhu H, Guan Y, Lam TTY. ggtree: an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods Ecol Evol. 2017;8:28–36. [Google Scholar]

[pbio.3002480.ref122] 122.Wang LG, Lam TTY, Xu S, Dai Z, Zhou L, Feng T, et al. Treeio: An R package for phylogenetic tree input and output with richly annotated and associated data. Mol Biol Evol. 2020;37:599–603. doi: 10.1093/molbev/msz240 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002480.ref123] 123.Gu Z, Gu L, Eils R, Schlesner M, Brors B. Circlize implements and enhances circular visualization in R. Bioinformatics. 2014;30:2811–2812. doi: 10.1093/bioinformatics/btu393 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref124] 124.Paradis E, Schliep K. Ape 5.0: An environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics. 2019;35:526–528. doi: 10.1093/bioinformatics/bty633 [DOI] [PubMed] [Google Scholar]

[pbio.3002480.ref125] 125.Garnier S, Ross N, Rudis R, Camargo PA, Sciaini M, Scherer C. viridis(Lite)—Colorblind-Friendly Color Maps for R. Comprehensive R Archive Network (CRAN). 2023. [Google Scholar]

[pbio.3002480.ref126] 126.Wright K. pals: Color Palettes, Colormaps, and Tools to Evaluate Them. Comprehensive R Archive Network (CRAN). 2021. [Google Scholar]

[pbio.3002480.ref127] 127.Winter D. pafr: Read, manipulate and visualize ‘Pairwise mApping Format’ Data. Comprehensive R Archive Network (CRAN). 2022. [Google Scholar]

PERMALINK

Horizontal transfers between fungal Fusarium species contributed to successive outbreaks of coffee wilt disease

Lily D Peck

Theo Llewellyn

Bastien Bennetot

Samuel O’Donnell

Reuben W Nowell

Matthew J Ryan

Julie Flood

Ricardo C Rodríguez de la Vega

Jeanne Ropars

Tatiana Giraud

Pietro D Spanu

Timothy G Barraclough

Roles

Abstract

Introduction

Results

A high-quality genome assembly for the Fusarium xylarioides arabica host specialist

Table 1. Statistics for sequenced Fusarium genomes.

Fusarium xylarioides is a monophyletic species complex with genetically differentiated populations

Fig 1. The Fusarium xylarioides species complex.

Fig 2. Genetic clusters in Fusarium xylarioides.

The populations differ in the presence of multiple large genome regions

Fig 3. Host-specific genome regions between members of the Fusarium xylarioides species complex comprise putative horizontally transferred regions.

Several population-specific genome regions show close matches to pathogenicity chromosomes of multiple Fusarium oxysporum lineages

Fig 4. Synteny between the Fusarium xylarioides arabica reference genome and 2 Fusarium oxysporum genomes for the putative HTRs.

Putative horizontally transferred regions contain genes that are highly expressed during infection, including effector genes from Fusarium oxysporum

Fig 5. Up-regulated genes in Fusarium xylarioides coffee wilt infection.

Miniature impala elements and other transposons are shared between Fusarium xylarioides and Fusarium oxysporum

Fig 6. Miniature impala composition of Fusarium genomes.

An 80-kb Starship mobile element has inserted between highly expressed genes in the arabica genomes

Fig 7. A Starship mobile element in the Fusarium xylarioides arabica genomes.

Discussion

Materials and methods

Strain details

Long-read reference genome sequencing and assembly

Illumina sequencing and assembly

Table 2. Data sources used to annotate Fusarium oxysporum and F. solani genomes sequenced in this study.

RNA sequencing to aid annotation of long- and short-read assemblies

Gene prediction and annotation

Other Fusarium genomes

Transposable elements and repeat annotations

Orthologous gene groups, species tree, and genetic structure

Identification of Fusarium xylarioides population-specific regions and horizontal transfers

Statistical evaluation of putative horizontal transfers

Infection assays and controls

RNA extraction, sequencing, and quality control

Differential expression analysis

Functional annotation analysis

Gene Ontology and InterProScan classifications

EffectorP

Carbohydrate-active enzymes

Data analysis and presentation

Supporting information

Acknowledgments

Abbreviations

Data Availability

Funding Statement

References

Decision Letter 0

Melissa Vazquez Hernandez

Roles

Decision Letter 1

Melissa Vazquez Hernandez

Roles

Author response to Decision Letter 1

Decision Letter 2

Melissa Vazquez Hernandez

Roles

Author response to Decision Letter 2

Decision Letter 3

Melissa Vazquez Hernandez

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES