A Near-T2T Genome Assembly of Elsinoe fawcettii Provides Insights into Host Adaptation Driven by Cis-Regulatory Evolution

Abstract

Elsinoe fawcettii is a devastating citrus pathogen worldwide, yet high-quality genomic resources are lacking, limiting insights into its adaptive mechanisms. Seventeen strains collected from 13 host species across 5 Chinese provinces were confirmed as E. fawcettii by multi-loci (ITS, rpb2, tef1-α) phylogenetic and morphological analyses. A near-telomere-to-telomere (near-T2T) genome for representative strain FJ-Y-3 was constructed using integrated PacBio and Hi-C sequencing. The 24.40 Mb assembly was organized into 11 chromosomes with exceptional completeness (BUSCO: 97.1%) and continuity (scaffold N50: 2.18 Mb). Pan-genome analysis revealed a closed structure, with core genes representing 77.19% of the total, suggesting evolutionary adaptation through fine-regulation of conserved elements rather than extensive gene content variation. Accessory genes were significantly enriched in terpenoid/polyketide metabolism, cell surface remodeling, and xenobiotic degradation, underscoring metabolic plasticity. Whole-genome resequencing showed single-nucleotide polymorphisms as the dominant variant, with ~60% residing in regulatory regions, implicating cis-regulation as a key adaptive mechanism. This work provides a high-quality genome and multi-omics framework for E. fawcettii, establishing a crucial molecular foundation for understanding pathogen adaptation and developing sustainable disease management strategies.

1. Introduction

Fungi of the genus Elsinoe represent significant pathogens affecting a wide range of economically important crops worldwide [1]. Elsinoe fawcettii, the causal agent of citrus scab, infects fruits and leaves, resulting in malformation, premature fruit drop, and quality deterioration. E. fawcettii causes significant economic losses to the global citrus industry [2]. The pathogen’s broad host range across numerous Citrus species enables diverse transmission routes, greatly complicating control efforts [3].

Current studies on citrus scab disease primarily focus on morphological characteristics and molecular identification based on a limited set of genetic markers (e.g., ITS, rbp2, tef1-α) [4]. Although studies have provided initial insights into the intraspecific diversity of E. fawcettii, a comprehensive understanding of its genetic variation and host adaptation is still lacking. The absence of a high-quality reference genome for E. fawcettii stands in contrast to the resources available for species like E. ampelina, E. arachidis, E. annonae, E. batata, and presents a major obstacle to advanced functional genomics studies on this pathogen [5,6,7,8,9].

Plant pathogenic fungi achieve host adaptation through regulatory mechanisms including reprogramming, cis-regulatory variation, DNA methylation, histone modification, and chromatin remodeling [10,11]. For instance, cis-regulatory variation enables pathogens to adjust gene expression patterns, facilitating infection of new hosts and expansion from narrow to broad host ranges [10].

This study employs an integrated approach, combining PacBio HiFi long-read sequencing with Hi-C chromatin conformation capture, to generate the first near-telomere-to-telomere (Near-T2T), chromosome-level reference genome for the pomelo (Citrus maxima) scab pathogen, E. fawcettii FJ-Y-3. Furthermore, we conduct comparative genomic analyses on a collection of E. fawcettii strains derived from diverse hosts (e.g., Citrus, Camellia, Ficus, etc.) with the following objectives: (1) to establish a high-quality genomic framework for this pathogen; (2) to characterize its pan-genome structure and functional evolutionary patterns; (3) to identify genetic variations SNPs associated with host adaptation. These integrated analyses will elucidate the genetic basis of pathogenicity, ultimately identifying molecular targets for precision disease control strategies.

2. Materials and Methods

2.1. Strains Collection and Identification

Seventeen Elsinoe fawcettii strains were obtained from diverse host plants (including Citrus, Camellia, Ficus, etc.) in the Chinese provinces of Fujian, Jiangxi, Shanxi, Yunnan and Zhejiang through tissue isolation and purification. The isolates were cultured on potato dextrose agar (PDA) at 25 °C in the dark for 21 days to assess colony morphology. Conidiomata and conidia were examined using stereomicroscopy and light microscopy to document key morphological features, with 50 conidia randomly selected for dimensional measurements. Genomic DNA was subsequently extracted from the harvested mycelia using a modified cetyltrimethylammonium bromide (CTAB) protocol. The quality and concentration of the DNA were assessed by NanoDrop 2000C (Thermo Fisher Scientific, Waltham, MA, USA) spectrophotometry (OD260/280 = 1.8–2.0) and confirmed by electrophoresis on a 1% agarose gel, with qualified samples stored for subsequent analyses.

All strains were subjected to PCR amplification of 3 gene regions: the internal transcribed spacer (ITS), RNA polymerase II second largest subunit (rpb2), and translation elongation factor 1-alpha (tef1-α). The primers used were as follows: ITS1-F (CTTGGTCATTTAGAGGAAGTAA)/ITS4 (TCCTCCGCTTATTGATATGC) for ITS, RPB2-5F2 (GGGGWGAYCAGAAGAAGGC)/fRPB2-7Cr (CCCATRGCTTGYTTRCCCAT) for rpb2, and elongation-1-F (AGCCCCTCCGTCTTCCTCTCCAG)/elongation-1-R (CGGTACGGCGGTCAATCTTCTCG) for tef1-α [12,13,14]. The amplification protocol consisted of an initial denaturation at 94 °C for 5 min; 35 cycles of 94 °C for 30 s, annealing at 54 °C (ITS) for 30 s, 56 °C (rpb2) for 60 s, or 60 °C (tef1-α) for 60 s, respectively, and extension at 72 °C for 45 s; followed by a final extension at 72 °C for 10 min. Molecular identification was performed through a multi-loci phylogenetic analysis based on the concatenated sequences (Supplement Table S1) [4]. Maximum likelihood (ML) trees were constructed with IQ-TREE 2.2.0 as implemented in PhyloSuite v1.2.3, with branch support evaluated using 1000 bootstrap replicates [15,16]. Bayesian inference (BI) analyses were conducted in parallel using MrBayes v3.2.7 to assess topological consistency [17].

2.2. Library Construction and Genome Sequencing

Library preparation was conducted as follows: For PacBio sequencing, DNA was randomly sheared using Covaris M220 (Covaris, Woburn, MA, USA) ultrasonication, and large fragments were selected using magnetic beads. Following end repair and hairpin adapter ligation, the fragments were purified with exonuclease, yielding libraries with insert sizes of approximately 15 kb and 20 kb. For Illumina sequencing, DNA was sheared to ~400 bp using Covaris, followed by end repair/5′ phosphorylation, A-tailing, adapter ligation, and final PCR enrichment. For Hi-C sequencing, crosslinked DNA was digested with MboI to create sticky ends. After end repair incorporating biotin labeling, spatially proximate fragments were ligated and circularized in situ using T4 DNA ligase. Following de-crosslinking, the DNA was sheared again, and biotin-labeled fragments were captured with streptavidin beads for library construction.

A differential sequencing strategy was employed for various strains. Strain FJ-NM was sequenced using PacBio Continuous Long Read (CLR, PacBio, Menlo Park, CA, USA); strain FJ-Y-3 was sequenced using both PacBio CLR and Circular Consensus Sequencing (CCS, PacBio, Menlo Park, CA, USA), supplemented with Hi-C sequencing; the remaining strains from diverse hosts (e.g., Citrus, Camellia, Ficus, etc.) were subjected to 150 bp paired-end sequencing on Illumina HiSeq platforms(Illumina, San Diego, CA, USA).

2.3. Genome Assembly

Adapter sequences were removed from raw Illumina short reads, and low-quality reads (Q-score < 30) were filtered out using Fastp v0.23 [18]. De novo assembly of Illumina data was performed with SPAdes v4.0.0 (parameters: -k 21,33,55,77,99 --careful --cov-cutoff auto) [19]. Then, a chromosome-level assembly of FJ-Y-3 was generated by integrating PacBio HiFi long-reads with Hi-C data using the following assembly pipelines: PacBio CLR data were processed using Hifiasm v0.20.0 [20]; PacBio HiFi reads were assembled with Canu v2.3 (parameters: genomeSize = 30 m) [21]; Hi-C data were utilized for chromosome-scale scaffolding with ALLHIC v0.9.8 (parameters: --minREs 50 --maxlinkdensity 3 --NonInformativeRabio 2) [22]. Genome completeness was assessed with BUSCO v6.0.0 against the ascomycota_odb12 database [23].

2.4. Repetitive Sequences Annotation

Repetitive sequences were identified combining de novo prediction with homology-based methods. For de novo prediction, a custom database was first constructed with BuildDatabase, followed by RepeatModeler v1.0.11 analysis to generate a species-specific repeat library [24]. This de novo library was then merged with a curated repeat library for the related class Dothideomycetes, which was retrieved from the Dfam database, to create a final, integrated reference library. Finally, RepeatMasker v4.1.2 was applied to perform a homology-based repeat search throughout the whole genome [25].

2.5. Gene Prediction and Annotation

For strains FJ-Y-3 and FJ-NM, protein-coding genes were annotated using BRAKER2, which integrates both ab initio gene predictions generated by AUGUSTUS (v.3.5.0) and homology evidence from E. fawcettii (Genbank: GCA_012977835.1), E. murrayae (Genbank: GCA_002895985.1), E. batatas (Genbank: GCA_017309325.2), E. australis (Genbank: GCA_003013795.1), E. ampelina (Genbank: GCA_010093995.1), E. arachidis (GWH: GWHBFXO00000000.1) and E. annonae (GWH: GWHBKHL00000000.1), as well as a transcriptome assembly generated from self-generate RNA-seq datasets [26]. The annotation of protein-coding genes for other strains of E. fawcettii was performed using a combination of homology alignment and an ab initio approach. The longest transcripts were generated via TBtools v2.0 and tRNAs were identified using tRNAscan-SE v2.0.9 [27].

Multiple gene families were also identified, including carbohydrate-active enzyme (CAZymes) by dbCAN v5.0, secreted proteins by SignalP v6.0, and secondary metabolite biosynthetic gene clusters (SMBGCs) by antiSMASH v7.0 (parameters: --taxon fungi) [28,29,30]. Additionally, eggNOG-Mapper v2.0 was used to perform gene functional annotations, such as those related to the GO and KEGG pathways, across all strains [31].

2.6. Pan-Genome Analysis

Orthologous gene families (orthogroups) were identified across all analyzed strains using Orthofinder v2.5.4 [32]. Based on their distribution frequency among the 21 strains (17 obtained in this study and 4 from the NCBI public database, SRX7956662, SRX7950446, SRX7945230, SRX5307824), gene clusters were categorized as core genes (present in all strains), accessory genes (present in two or more but not all strains), or strain-specific genes (unique to individual strains). To characterize the pan-genome dynamics, saturation curves were constructed by randomly sub-sampling of n_strains (ranging from 1 to 21) for 1000 bootstrap iterations, and the mean sizes of the core genome (shared gene clusters) and the total pangenome (non-redundant gene clusters) at each n were plotted to determine its openness-distinguished by a significant linear expansion (open) or an asymptotic plateau (closed) as new strains are added. Finally, functional enrichment analyses of accessory genes were performed by integrating eggNOG-Mapper annotations for all strains, constructing a custom organism database (orgdb) in R v4.4.2 (https://cran.r-project.org/bin/windows/base/old/4.4.2/, accessed on 2 February 2026), and executing KEGG pathway enrichment by TBtools v2.0.

2.7. Detection of SNP Variations

Following quality control, raw Illumina reads were aligned with the reference genome using BWA v0.7.18 [33]. The modified alignments were sorted, and PCR duplicates were marked. Variant calling for single-nucleotide polymorphisms (SNPs) was performed with GATK v4.0.0 [34]. For the cohort of samples, a joint genotyping approach was applied by first generating gVCF files for each sample, followed by merging and joint genotyping. High-quality variants were obtained by applying stringent filters. Finally, functional annotation of these filtered variants was conducted using a custom SnpEff v5.3.0a (parameters: -ud 2000) [35].

3. Results

3.1. Isolation, Identification and Selection of Elsinoe Isolates

A total of 15 diseased plant tissue samples (leaves and fruits) from 13 distinct host species were collected from five provinces in China (Fujian, Jiangxi, Yunnan, Shaanxi, Zhejiang) (Figure 1). Seventeen strains were obtained through isolation and purification. Preliminary identification based on colony morphology and ITS sequence alignment confirmed that 17 of these isolates belonged to the genus Elsinoe (Supplement Table S2).

3.2. Morphological Characterization and Multi-Loci Phylogenetic Analysis of Elsinoe fawcettii

Concatenated phylogenetic analyses based on the ITS, rpb2, and tef1-α gene sequences were performed on the 17 obtained strains. Both maximum likelihood (ML) and Bayesian inference (BI) trees revealed that all strains clustered into a single, well-supported monophyletic clade with the reference E. fawcettii strain CBS 139.25, exhibiting 100% bootstrap support and a Bayesian posterior probability (BPP) of 0.99 (Figure 2). Morphological characteristics of the representative strain FJ-Y-3 were examined in detail. After 21 days of incubation on PDA medium at 25 °C, the colony diameter reached 1.8 cm. The colony appeared fleshy and wrinkled, with a reddish-brown reverse. Conidiogenous cells were hyaline, aseptate, solitary and phialidic. Conidia were unicellular, subglobose to ellipsoid, measuring 5.42–8.25 μm × 2.06–3.82 μm (n = 50) (Figure 3). Based on the concordant results from the multi-loci phylogenetic analyses and morphological observations, all 17 strains were conclusively identified as E. fawcettii. All voucher strains have been deposited in the China General Microbiological Culture Collection Center (CGMCC).

3.3. Near-T2T Genome Assembly and Annotation of Elsinoe fawcettii

In this study, whole-genome sequencing of E. fawcettii strain FJ-Y-3, a representative strain of the pomelo (Citrus maxima) scab pathogen, was performed using an integrated approach combining PacBio Continuous Long Read (CLR) (>120 × coverage), PacBio High Fidelity (HiFi) (>120 × coverage), and Hi-C (>120 × coverage) technologies. The PacBio CLR and HiFi platforms generated approximately 12.38 Gb and 8.05 Gb of raw data (Supplement Table S3), respectively. Initial assembly with Canu yielded 57 contigs, with 1.99 Mb of contigs N50. These contigs were subsequently scaffolded into chromosomes using the ALLHIC, resulting in the first near-telomere-to-telomere (near-T2T) level genome assembly for E. fawcettii. The final nuclear genome size is 24.40 Mb, anchored onto 11 chromosomes (

Table 1. Genomic features of E. fawcettii.

Features	Elsinoe fawcettii
	FJ-Y-3; CGMCC3.24394
Scaffold Assembly size (bp)	24,401,734
Scaffold number	11
Scaffold N50 (bp)	2,181,953 (L = 5)
Scaffold N90 (bp)	1,706,596 (L = 10)
Maximum Scaffold length (bp)	3,384,137
GC content (%)	52.52
Gap length (bp)	400
BUSCO (%)	97.1
Repeat sequences (bp/%)	576,723/2.36
Retroelements(number/%)	132/2.12
SINEs (number/%)	0
LINEs (number/%)	26/1.36
LTR elements (number/%)	106/0.76
DNA transposons (number/%)	33/0.10
Rolling-circles (number/%)	0
Unclassified interspersed repeats (number/%)	79/0.15
Simple repeats (number/%)	0
Small RNA (number/%)	0
Satellites (number/%)	0
Low complexity (number/%)	0
tRNA	43
Number of genes	9818
Pfam annotated genes	7575
GO annotated genes	3815
KEGG annotated genes	4153
KOG annotated genes	8869
CAZy annotated genes	342
Candidate effectors	1061
SMBGCs	25

and Figure 4). Telomere analysis revealed that the characteristic repeat unit (TTAGGG) n was detected at both ends of all chromosomes. The assembly achieved a scaffold N50 of 2.18 Mb and an N90 of 1.70 Mb. Assessment with BUSCO indicated a high completeness of 97.1%, demonstrating the exceptional quality and contiguity of the genome assembly.

Only 2.36% of the assembled FJ-Y-3 genome was annotated as repetitive sequences and a total of 43 tRNAs were identified. The whole genome contains 9818 protein-coding genes, which spans 19.67% (4.81 Mb) of the total genome length. The BUSCO completeness values were 97.1% for gene models, which demonstrated the high completeness of our gene annotations. Furthermore, A total of 1061 genes encoding signal peptides, 342 CAZyme genes, and 25 BGCs were predicted, which are potentially involved in the pathogenicity and environmental adaptation of the fungus (Table 1 and Figure 4).

3.4. The Pangenome Structure of Elsinoe fawcettii

Based on the whole-genome sequences of twenty-one E. fawcettii strains from diverse hosts and geographical origins, a pan-genome was constructed using OrthoFinder2, in which a total of 11,333 gene families were identified. These gene families were classified into core genes (present in all strains), accounting for 77.19% (8749 orthogroups), accessory genes (present in two to twenty strains), constituting 19.17% (2172 orthogroups), and private genes (present in only one strain), representing 3.64% (412 orthogroups) (Figure 5). Strains isolated from Rutaceae hosts exhibited distinct core and accessory gene compositions compared to strains from other hosts. Furthermore, the fitting of the pan-genome saturation curve (1000 permutations) using Heaps’ law yielded a pan-genome index of b = 0.0295 (approaching zero), with a good model fit (R² = 0.95) (Figure 6). This indicates that E. fawcettii possesses a typical closed pan-genome structure. The discovery rate of new gene families decreased rapidly and reached saturation as the number of strains increased, suggesting a relatively fixed species gene pool where genetic diversity is primarily confined to a limited set of accessory gene regions. This characteristic implies that the evolutionary strategy of this pathogen relies more on fine regulation and subtle variations within the core genome, rather than large-scale gene gain or loss, providing a novel framework for understanding its host adaptation and pathogenic differentiation.

3.5. KEGG Enrichment Analyses of Accessory Genes

KEGG pathway enrichment analyses of accessory genes in the E. fawcettii pangenome revealed significant metabolic plasticity. Among the top 20 enriched pathways (of 219 identified) (Supplement Table S4; Figure 7), those related to terpenoid/polyketide metabolism, cell surface polysaccharide remodeling, and xenobiotic degradation were prominently represented. Key terpenoid and polyketide pathways, including terpenoid backbone biosynthesis (344 genes) and terpenoid-quinone synthesis (224 genes) were exceptionally enriched, implying their functional significance in expanding the chemical arsenal for defense and ecological interactions. Concurrently, pathways regulating cell surface plasticity, such as O-glycan biosynthesis and glycosaminoglycan degradation, were also significantly enriched. Notably, the prokaryotic defense system pathway exhibited the most striking enrichment (125 genes, p = 2.54 × 10⁻¹⁶). In summary, these findings indicate that the accessory genome contributes to environmental adaptation and metabolic flexibility through the expansion of specialized metabolite biosynthesis, cell surface reconfiguration, and xenobiotic responsiveness.

3.6. Genetic Variations Across the Whole Genomes of 21 Elsinoe fawcettii Strains

The strains were sequenced to an average of 390× depth with an Illumina 150 bp paired-end reads protocol (Supplement Table S3). A total of 825,416 high-quality SNPs with a minor allele frequency greater than 0.01 were identified following the SNP calling procedure. The distribution of SNP density was significantly clustered and non-random both within and across chromosomes. Specific high-density regions (e.g., central Chr01/Chr09, proximal-central Chr05/Chr10, terminal Chr07) exhibited SNP densities substantially exceeding the genomic background, implying elevated recombination or local selection, whereas Chr08 was characterized by broad low-SNP-density segments (Figure 8). Interestingly, the distribution of SNPs was highly biased, with approximately 60% clustered in putative gene regulatory regions. Specifically, 32.59% and 29.30% of SNPs were located downstream and upstream of genes, respectively; these genomic locations are typically enriched for cis-regulatory elements such as promoters and enhancers. In contrast, a minority of SNPs (18.05%) were located within exonic regions, which directly alter protein sequences, and only a minimal proportion were found in intronic or splice-site regions (Figure 9). This distinct pattern strongly supports the predominance of cis-regulatory evolution as a key evolutionary mechanism in this fungus.

4. Discussion

4.1. Multi-Loci Phylogenetic Analysis Resolves Taxonomic Identity and Uncovers Limited Intraspecific Diversity in Elsinoe fawcettii Strains

Multi-loci sequence analyses (ITS, rpb2, tef1-α) unequivocally assigned all examined strains to Elsinoe fawcettii, confirming taxonomic consistency and establishing a reliable species framework for comparative genomic studies [36,37]. The phylogenetic tree exhibited short branch lengths with tight clustering patterns, indicating limited genetic diversity among the sampled E. fawcettii strains. This pattern may reflect either a relatively young evolutionary stage with insufficient time for genetic divergence or the concentration of genetic variation in specific genomic regions not captured by the conserved genes (rpb2, tef1-α) used for phylogeny reconstruction [37].While useful for preliminary classification, multi-loci markers may underestimate the actual genetic variation, as single-nucleotide polymorphisms (SNPs) analysis reveals a more extensive diversity at the genome-wide level.

4.2. A Near-T2T Genome for Elsinoe fawcettii as a Foundation for In-Depth Functional Genomic Studies

The newly constructed near-telomere-to-telomere (Near-T2T) genome for E. fawcettii provides a critical foundation for elucidating the genomic basis of its pathogenicity and evolutionary mechanisms [38,39]. In comparison to other sequenced Elsinoe species (e.g., E. ampelina, E. arachidis, and E. batata), Elsinoe fawcettii possesses a smaller genome size and a reduced repetitive sequence content, yet it maintains a comparable number of predicted protein-coding gene [5,6,8]. This high-quality genome facilitates the precise annotation of gene models and regulatory elements. It thereby provides a reliable structural framework and a benchmark for species identification, which will support subsequent comparative and functional genomics studies [39]. The power of T2T genomes for localizing pathogenicity genes is exemplified by two key cases. In Pyricularia oryzae (Synonym: Magnaporthe oryzae), T2T assembly closed about 20% of previous gaps, achieving complete chromosomes and facilitating the precise annotation of 493 effector genes to create a comprehensive pathogenicity atlas [40]. Similarly, research on Diplocarpon coronariae showed that a recent chromosome (Chr15) emerged from LTR retrotransposon bursts, highlighting how repetitive sequences drive major chromosomal changes and rapid adaptation [41]. These examples underscore the value of the new E. fawcettii near T2T genome for probing chromosomal evolution, repeat-driven plasticity, and pathogenicity mechanisms, thus providing a molecular basis for understanding its virulence evolution.

4.3. Closed Pan-Genome Architecture of Elsinoe fawcettii Reveals Accessory Genome-Driven Host Adaptation

Closed pan-genomes rely on regulatory plasticity of a stable gene set to adapt, whereas open pan-genomes evolve through frequent genes gain [42,43]. The pan-genome of E. fawcettii is characterized by a closed architecture, with core genes comprising 77.2% of all gene families. Notably, strains from Rutaceae hosts exhibit significant divergence in accessory genome composition, highlighting the importance of this genomic compartment in host adaptation. This observation parallels findings in the wheat scab pathogen Fusarium graminearum, where agricultural practices have driven virulence differentiation [44]. These cases collectively indicate that host-imposed selection is a major driver of pathogen evolution. Methodologically, conventional GWAS relying on a single reference genome may fail to detect key loci associated with structural variants. This limitation was overcome in the pan-genome-wide association studies of the wheat glume blotch pathogen Parastagonospora nodorum, which identified fungicide resistance genes that standard approaches missed [45]. Therefore, future research on E. fawcettii should integrate multi-reference genomes and pan-genomic analyses to systematically uncover the cryptic genetic variation underlying its pathogenicity. In summary, these findings demonstrate that E. fawcettii achieves broad-host-range adaptation through functional diversification of a limited accessory genome within a closed genomic framework. This resolves the apparent paradox of high adaptability coupled with low genetic variation and provides a molecular basis for developing broad-spectrum control strategies targeting conserved core pathogenicity pathways.

4.4. Elsinoe fawcettii Orchestrates Broad-Host Adaptation via an Integrated Attack–Defense–Stability Strategy Encoded in Its Accessory Genome

Functional enrichment analysis of the accessory genome in E. fawcettii reveals a comprehensive adaptive strategy centered on the synthesis and regulation of virulence genes, masking of pathogen-associated molecular patterns (PAMPs), and genomic stabilization. This integrated approach, leveraging functional diversification within a constrained genome, facilitates efficient host adaptation under a closed pan-genome framework. Significant enrichment of accessory genes in terpenoid and polyketide metabolic pathways underscores the critical role of secondary metabolites in pathogenesis, wherein polyketide toxins such as elsinochrome directly destroy host membrane integrity, while terpenoid derivatives potentially disrupt immune recognition via allelochemical signaling [46]. Concurrently, the pronounced enrichment of pathways governing cell surface polysaccharide remodeling, particularly O-mannan biosynthesis, enables dynamic masking of pathogen-associated molecular patterns (PAMPs) on the fungal cell wall [47]. Strikingly, the accessory genome exhibits enrichment of prokaryotic-derived defense systems, suggesting the horizontal acquisition of exogenous mechanisms that maintain genomic stability by restricting transposable element proliferation, preventing genomic redundancy, and preserving core pathogenicity functions [48]. Overall, E. fawcettii employs an integrated adaptive network through specialization of its accessory genome, focused on the synthesis and regulation of virulence genes, PAMPs, and maintenance of genomic stability, thereby ensuring adaptation within a constrained evolutionary landscape.

4.5. Closed Pan-Genome of Elsinoe fawcettii Reveals Broad Host Adaptation Through Cis-Regulatory Evolution over Gene Gain or Loss

Whole-genome analysis of E. fawcettii reveals a markedly non-random distribution of genetic variation, dominated by SNPs and exceptional enrichment in cis-regulatory regions flanking genes. This pattern, in which low coding diversity (reflected in a conserved core genome and limited coding-region SNPs) coexists with high regulatory diversity, suggests that E. fawcettii achieves host adaptation primarily through cis-regulatory evolution rather than through changes to protein-coding sequences. This pattern indicates an evolutionary strategy that favors cis-regulatory evolution, which enables rapid host adaptation through sequence variations in regulatory elements that finely regulate gene expression. This strategy avoids the potentially higher fitness costs associated with extensive alterations to protein-coding sequences. This strategy appears to be broadly conserved across phytopathogenic fungi. For instance, F. graminearum exhibits a biphasic genome structure, with virulence-related genes enriched in a rapidly evolving, recombination-active subgenome that is specifically activated during infection [49]. This example highlights the crucial role of regulatory and epigenetic mechanisms in adaptive evolution. From the perspective of genome structure, differential evolutionary pressures across chromosomes suggest functional compartmentalization. Notably, chromosome 1, 5, 7, 9, 10 emerges as a variation hotspot, analogous to the fast-evolving subgenome of F. graminearum, and acts as an evolutionary incubator for effector gene clusters. In contrast, chromosome 8 remains relatively conserved and is likely to harbor essential housekeeping genes. This core-accessory coexistence represents a fundamental mechanism for balancing environmental adaptability with genomic stability in pathogenic fungi [50]. Thus, the regulatory sequence-driven evolution observed in E. fawcettii aligns with adaptive mechanisms observed in diverse pathogens. This paradigm, which prioritizes precise expression control over protein structural changes and is coupled with intragenomic functional partitioning, likely constitutes a key evolutionary framework. This framework enables filamentous pathogenic fungi to rapidly adapt to hosts while preserving the integrity of their core genome. These findings provide novel perspectives for understanding pathogenicity evolution and developing sustainable disease management strategies.

5. Conclusions

An integrated phylogenetic analysis of multi-loci (ITS, rpb2, tef1-α) and morphological assessment definitively identified 17 strains from 13 host species across 5 Chinese provinces as Elsinoe fawcettii. A high-quality, near-telomere-to-telomere (near-T2T) genome assembly of E. fawcettii strain FJ-Y-3 was produced. Pan-genomic analysis revealed a characteristic closed structure, wherein core genes comprised 77.19% of all gene families. The saturation of the gene pool with sequential strain addition indicated an evolutionary strategy reliant on fine regulation of the core genome rather than extensive gene gain or loss. Functional enrichment analysis demonstrated that accessory genes were significantly overrepresented in terpenoid/polyketide metabolism and cell surface polysaccharide remodeling pathways, which provides a molecular basis for metabolic plasticity and immune evasion. Genome-wide variation analysis further established that approximately 60% of SNPs were densely clustered within gene regulatory regions, strongly supporting cis-regulatory evolution as the core mechanism for rapid host adaptation of E. fawcettii. In summary, this study systematically elucidates a novel adaptive paradigm of E. fawcettii, wherein environmental adaptation is achieved through diversification of cis-regulatory elements within a constrained genomic framework.