Early-Stage Growth Restriction of Hymenoscyphus fraxineus on Tolerant Fraxinus excelsior Is Associated with Constitutive Chemical Defenses

Abstract

Hymenoscyphus fraxineus is the causal agent of ash dieback, a devastating disease of European ash (Fraxinus excelsior). Although the pathogen is believed to have originated in East Asia and has been confirmed in Japan, European ash trees cultivated in the Sapporo Experimental Forest of Hokkaido University remain asymptomatic despite the presence of H. fraxineus. In this study, we investigated the early infection behavior of H. fraxineus and associated host defense responses by comparing asymptomatic F. excelsior with the susceptible control species F. angustifolia. Leaflets were inoculated with ascospores, and fungal development as well as host responses were examined microscopically during early infection stages. In addition, we analyzed the accumulation of selected coumarins, which have been proposed as candidate compounds associated with ash dieback tolerance, and assessed their effects on ascospore germination. We found that fungal growth was consistently restricted on F. excelsior at 7 days post inoculation, particularly at the stage of invasion into adjacent epidermal cells. Fraxetin was detected in F. excelsior leaflets but not in F. angustifolia, and fraxetin treatment significantly reduced ascospore germination in vitro. While typical markers of induced resistance were not clearly detected at the examined time points, these results indicate that constitutive chemical traits, including fraxetin accumulation, may contribute to early-stage suppression of H. fraxineus growth in F. excelsior. Together, our findings provide insight into early host–pathogen interactions associated with ash dieback tolerance and highlight the potential role of constitutive defenses during initial infection.

1. Introduction

Ash dieback is a lethal disease of European common ash (Fraxinus excelsior) that was first recognized in Poland in the early 1990s and has since spread to more than 30 European countries [1,2,3,4]. The causal agent, Hymenoscyphus fraxineus (T. Kowalski) Baral, Queloz & Hosoya, has caused extensive mortality of F. excelsior across Europe and represents a major threat to ash populations [5].

Hymenoscyphus fraxineus is believed to originate from East Asia, where populations exhibit higher genetic diversity than those in Europe [6]. In addition, Asian Fraxinus species show lower susceptibility to H. fraxineus infection, supporting the hypothesis that the pathogen was introduced into Europe from East Asia [7]. In Japan, H. fraxineus is distributed in Hokkaido as well as central and northeastern regions, where F. mandshurica var. japonica occurs [6,8,9]. Notably, ash dieback symptoms have not been reported in these regions.

Previous studies conducted in East Asia have demonstrated that H. fraxineus can persist as an endophyte in living ash leaves, particularly in F. mandshurica var. japonica [10], while exhibiting a saprophytic lifestyle on fallen leaves where apothecia are formed on pseudosclerotial plates [11]. Similar endophytic behavior has also been reported from the Russian Far East, suggesting that asymptomatic host–pathogen interactions are common in the native range of the fungus [12,13]. Early studies on ash dieback further documented that H. fraxineus (formerly Chalara fraxinea T. Kowalski) is capable of colonizing multiple Fraxinus species with variable symptom expression, highlighting that host susceptibility differs markedly among species and environments [14].

In the Sapporo Experimental Forest of Hokkaido University, H. fraxineus is present and apothecia are regularly observed on fallen leaves; however, European common ash trees planted at this site remain healthy. Other ash species growing in the same area, including Manchurian ash and green ash (F. pennsylvanica), also show no dieback symptoms. Our preliminary observations indicate that H. fraxineus exists endophytically within compound leaves of European common ash, suggesting that disease development is actively suppressed in these trees. Despite extensive research on ash dieback epidemiology and host responses at later infection stages, the mechanisms that restrict fungal growth during the earliest phase of infection in asymptomatic or tolerant ash hosts remain poorly understood.

Infection by H. fraxineus is initiated primarily via ascospores, which germinate on leaf surfaces, form appressorium-like structures, and penetrate host tissues through the cuticle before colonizing epidermal and mesophyll cells [15]. Comparative inoculation studies have shown that such early infection processes can proceed similarly across Fraxinus species, whereas subsequent fungal growth and symptom development vary depending on host susceptibility, indicating that post-penetration events are critical determinants of disease outcome [16].

In recent years, increasing attention has been paid to host plant defense responses that restrict the progression of H. fraxineus in tolerant or asymptomatic ash hosts. Accumulating evidence suggests that tolerance to ash dieback is not necessarily associated with strong induction of classical defense responses, such as hypersensitive cell death, but may instead rely on constitutive defense traits that limit fungal growth during early infection stages. In its native range in East Asia, H. fraxineus frequently exhibits an endophytic lifestyle in living ash leaves without causing visible disease symptoms, indicating that early host–pathogen interactions differ fundamentally from those observed in susceptible European ash populations [17]. Histological and biochemical studies have further shown that phenolic compounds can accumulate locally around fungal hyphae in infected ash tissues, even in the absence of extensive necrosis, suggesting that chemical barriers may suppress pathogen spread within host tissues [18]. More recently, comparative metabolomic analyses of F. excelsior genotypes with contrasting susceptibility revealed distinct chemotypes associated with disease tolerance, characterized by the constitutive accumulation of multiple antifungal secondary metabolites, including coumarins, secoiridoids, flavonoids, lignans, and phenylethanoids. Among these, coumarins such as fraxetin and esculetin were strongly associated with reduced susceptibility and exhibited fungistatic activity against H. fraxineus at physiologically relevant concentrations, supporting their proposed role as pre-existing chemical defense barriers rather than stress-induced metabolites [19]. Despite these advances, how such constitutive chemical defenses influence the earliest stages of H. fraxineus infection, particularly penetration, intracellular growth, and invasion into adjacent host cells, remains poorly understood.

In this study, we aimed to elucidate the early-stage defense processes associated with ash dieback suppression in European common ash grown in Japan. Specifically, we compared the early infection behavior of H. fraxineus and host defense responses between asymptomatic F. excelsior and the susceptible control species F. angustifolia which is known as a closely related species to F. excelsior, both species belong to section Fraxius [20]. We hypothesized that early-stage suppression of H. fraxineus in tolerant F. excelsior occurs primarily after host penetration and is associated with constitutive defense traits rather than with strong induction of classical defense responses.

2. Materials and Methods

2.1. Plant Seedlings

Seedlings (10 to 15 cm) of F. excelsior were collected under mature trees at Sapporo Experimental Forest of Hokkaido University, Sapporo, Hokkaido (43°04′ N; 141°20′ E), in July 2019. They might originate from naturally dispersed seeds. Ash dieback-like symptoms were not observed on the seedlings and mature trees of F. excelsior. Wild seedlings were transplanted into plastic pots (approximately 10 cm in depth; approximately 0.5 L in volume, equivalent to a 4-inch pot) containing a mixture (3:1) of Akadama soil and leaf mold. At the time of transplantation, the seedlings were approximately 50 cm in height. The exact age of the seedlings could not be determined because they were collected from natural populations. Three-year-old F. angustifolia ‘Raywood’ seedlings grafted onto F. griffithii rootstock were purchased from Chiyoda Nouen, a commercial nursery farm in Aichi, Japan (35°22′94″ N, 136°77′53″ E; 3 m a.s.l.; http://chiyodanouen.co.jp/ (accessed on 31 January 2026) in 2016. Fraxinus angustifolia seedlings were cultivated in the same soil mixture. Seedlings of both Fraxinus species were grown at the nursery on the campus of University of Tsukuba, Tsukuba, Ibaraki, Japan (36°06′45″ N 140°06′05″ E, 29 m a.s.l.).

2.2. Inoculation Test Using Compound Leaves to Confirm Tolerance of F. excelsior Against H. fraxineus

A single ascospore isolate of H. fraxineus, Su-2003s, which is derived from an apothecium collected on 11 August 2020 at Sugadaira Research Station, Mountain Science Center, University of Tsukuba, Ueda, Nagano, Japan (36°40′44″ N,138°20′50″ E, 1320 m a.s.l.), was used as an inoculum. Fungus was cultured on PDA with 0.1% Ebios (Brewer’s yeast preparation, Asahi Group Foods, Ltd., Tokyo, Japan) in a plate in 9 cm Petri dishes at 17 °C. Seedlings of F. angustifolia and F. excelsior were transferred into a growth cabinet maintained at 24 °C light (photon flux density of 115–160 µmol/m²/s) for 14 h light cycle and 20 °C for 10 h dark cycle before more than one week of inoculation.

For artificial inoculation, 5 compound leaves from one or two seedlings of Fraxinus species were used. A diagonal slit about 1 cm long was made by a scalpel on the petiole with about 2 cm from the base. A small agar block (1 mm × 2 mm × 5 mm) from the edge of the Su-2003s colony was inserted into the slit and tightly sealed with Parafilm. Another five leaves from the same seedlings were selected as the control. The control compound leaves were treated with an agar block without fungus by the same method. After inoculations, seedlings were transferred into same growth cabinet, sprayed sterile distilled water (DW) and covered with a plastic bag to keep moist conditions for 2 days. Plastic bags were removed after 2 days. External symptom development was observed every day for 2 months. Defoliation (fallen leaves or dead leaves which the inoculated leaves turned brown included) were observed; the date was recorded. Means of the date of defoliation and survival rate of the inoculated leaves at the endpoint of observation were calculated. This inoculation assay involves artificial wounding of the petiole and therefore does not represent the natural infection route by ascospores; it was used to confirm relative tolerance and susceptibility under standardized infection pressure. Three independent inoculation tests were demonstrated on 3 May 2021, 17 July 2021 and 17 May 2022.

2.3. Microscopic Observation of the Early Infection Stage Using Whole-Leaf Clearing and Staining Technique

To characterize early infection behavior of H. fraxineus on ash leaflets, microscopic observations were performed as follows. Seedlings of F. angustifolia and F. excelsior were transferred to a growth cabinet from nursery before more than 3 weeks of inoculation. Plants were maintained at 24 °C light (photon flux density of 100~120 µmol/m²/s) for 14 h light cycle and 20 °C for 10 h dark cycle. Fully expanded, but young and soft leaflets were selected for inoculation tests. Three leaflets were arranged on moistened Kimwipes (Kimtech Science brand™, Kimberly-Clark Professional, Irving, TX, USA) with DW in a plastic plate.

Apothecia induced on rachises at laboratory were used as an inoculum. Rachises of F. mandshurica covered with black pseudosclerotial plates of H. fraxineus were sampled from the Sapporo Experimental Forest. Rachises of F. excelsior collected on 21 June 2021 were used for first experiment; rachises of F. mandshurica collected on 21 June and 4 September 2021 were used for second experiment. Rachises were kept at −80 °C until use. To induce apothecia of H. fraxineus, rachises were incubated on moistened sphagnum moss in plastic Petri dishes (140 × 100 × 14.5 mm, Eiken Chemical Co., Ltd., Tokyo, Japan) and continuously exposed to near ultraviolet under black light blue lamp (20W FL20SBL-B, NEC Co., Tokyo, Japan) at 20 to 23 °C. Sprayed with DW once in 2 to 3 days to keep wet condition, 4 to 6 weeks were required for production and maturation of apothecia.

Before inoculation, discharging of ascospores was confirmed on 1.5% agar plates. Apothecia with rachises were fixed inside the lid of the dish over detached leaflets. DW was sprayed to keep wet condition on the surface of the detached leaflets; plastic plates were covered by a plastic bag and incubated at 16 to 17 °C in the moist chamber. After 24 h of inoculations, base of the leaflets was cut and wrapped with a Kimwipe containing 40 ppm gibberellin solution. Plastic plates were kept in a plastic bag and incubated at 16 to 17 °C for 2 days. After 2 days of inoculation, inoculation plates were incubated in the growth cabinet which was used for cultivation of Fraxinus plants. Plates were opened about 5 to 10 mm to stop further epiphytic growth of H. fraxineus to keep dry condition on surface of leaflets. Inoculation tests were demonstrated 2 times. Because of different growth speeds between F. excelsior and F. angustifolia, inoculation was demonstrated on different days in first experiment, for F. excelsior on 11 November 2021 and for F. angustifolia on 18 November 2021. Second experiment was demonstrated on 7 June 2022.

The leaflets were stained using whole-leaf clearing and staining technique following Shipton and Brown (1962) with minor modifications [21], in which trypan blue (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) was used instead of cotton blue or aniline blue. Cleared and stained leaflets were mounted in 50% glycerol on a slide grass and observed under light microscopy (BX51, Olympus, Tokyo, Japan) using ×100 oil lens.

To evaluate the host defense system of F. excelsior during the early infection stage, appressorium formation and invasion hyphae were observed at 5 dpi; invasion into adjacent cells was examined at 7 dpi. Infection stages were quantified as follows: 1. Appressorium formation rate: Ascospores which produce appressoria per 100 ascospores × 100; 2. Invasion rate: Ascospores with invasive hyphae in epidermal cells per 100 ascospores with appressoria × 100; 3. Adjacent cell invasion rate: Intracellular hyphae invading adjacent cells of epidermal cell per 30 invaded ascospores × 100. For each treatment, three detached leaflets were analyzed per experiment. For appressorium formation, at least 100 ascospores were scored per leaflet. Subsequent infection stages were quantified based on the number of infection events reaching each stage, as defined above. Differences between host species were evaluated using the statistical methods described in the Statistical Analysis subsection. Photos were recorded with a Nikon DC5300 camera (Nikon Co., Tokyo, Japan). Experiment was conducted two times.

2.4. Observation of Histochemical Host Defense Responses Observation on Transverse Sections

To evaluate histochemical host defense responses against H. fraxineus, transverse sections were prepared. Inoculations were demonstrated on 7 June 2022; samples of F. angustifolia and F. excelsior were collected at 7 dpi. Samples were boiled in lactic acid:glycerol:absolute ethanol (1:1:3 v/v/v) clearing solution and stored in boiling solution overnight at room temperature. Cleared samples were embedded in FSC22 Frozen Section Compound (Leica Microsystems Co., Wetzlar, Germany) and frozen at −80 °C. Sections at 15 to 20 μm were prepared using a cryomicrotome (Leica CM1520, Leica Microsystems Co., Wetzlar, Germany) and put on a cover glass, rinsed gently with absolute ethanol for a few seconds, and air-dried for 30 s.

Callose depositions and accumulation of lignin and suberin were observed on 20 invaded epidermal cells using each staining method. The presence of callose was verified by the aniline blue staining method. Sections were mounted with 0.02% lacto aniline blue (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) solution and observed under UV exposure (exciter filter 330–385 nm, barrier filter > 420 nm) with fluorescent microscopy (BX50, Olympus, Tokyo, Japan).

To detect lignification, phloroglucinol–HCl reaction was demonstrated following Pomar et al. (2002) with minor modifications [22]. In this study, 2% phloroglucinol ethanol–HCl solution was applied onto the sections and immediately covered with a glass. Sections were subsequently observed under light microscopy within 1 h.

Suberization at the infection sites was confirmed using Safranin-fast green FCF double staining method, following Bond et al. (2008) with minor modifications [23]. Staining procedures were operated on a cover glass in this study. Sections were immersed in 0.05% safranin (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) solution for 30 s and rinsed in gradual concentration of ethanol (50, 75, 90%) for 30 s. Rinsed sections were stained with 0.5% fast green FCF (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) solution for 1 min and rinsed in absolute ethanol for 30 s and mounted in 50% glycerol. Sections were observed under UV exposure (exciter filter 330–385 nm, barrier filter > 420 nm) with fluorescent microscopy. Photos were recorded with a Nikon DC5300 camera. Differences between host species were evaluated using the statistical methods described in the Statistical Analysis subsection. These histochemical analyses were performed on multiple independent experimental days, and similar staining patterns were consistently observed.

2.5. FeSO₄ Staining of Total Phenolic Compounds on Transverse Sections

To confirm the presence of polyphenols on Fraxinus leaflets, FeSO₄ staining method was applied following Ichihara and Yamaji (2009) with minor modification [24]. Samples were post-fixed in 2% osmium tetroxide for 1 h and sections were cut to 15 μm in this study. After 1 h of FeSO₄ staining, samples were observed under light microscopy using ×40 objective lens. Each 3 leaflets replicates of F. angustifolia and F. excelsior were observed. These histochemical analyses were performed on multiple independent experimental days, and similar staining patterns were consistently observed. Representative images from one experimental day are shown.

2.6. Preparation of Methanol Extraction of Leaflet Samples

To examine constitutive and infection-associated accumulation of selected coumarins, metabolite analyses were performed as follows. Leaflet samples were collected before inoculation as control, as well as mock and 7 dpi. Samples were extracted by methanol following Nemesio-Gorriz et al. (2020) [19]. To obtain adequate quantities of coumarin compounds, 100 mg of samples were used. Extraction was conducted once on 24 November 2022.

2.7. Chromatographic Peak Identification

Esculetin, fraxetin, and scopoletin were measured by using the LC–ESI–MS/MS (LCMS-8045; Shimadzu, Kyoto, Japan) under the following conditions: capillary voltage, 4.5 kV; desolvation line, 300 °C; heat block, 500 °C; nebulizer nitrogen gas, 3 L/min; drying gas, 10 L/min. Ion source polarity was set in the negative ion mode. The separation was performed with the LC system equipped with a 150 × 3.0 mm Kinetex^® F5 Column (Phenomenex, Torrance, CA, USA) with a particle and pore size of 2.6 μm and 100 Å, respectively. The initial mobile phase was solvent A: solvent B = 80:20 (solvent A, 0.025% formic acid; solvent B, acetonitrile (LC/MS Grade, Merck KGaA, Darmstadt, Germany). The solvent B concentration was increased to 80% for 10 min and then maintained at that ratio for another 2 min. The column was re-equilibrated for 3 min. The 0.6 mL/min flow rate and the 40 °C column temperature were maintained throughout the analysis. The MRM transitions were m/z 177.10 to 133.1 (for esculetin), m/z 207.10 to 192.05 (for fraxetin) and m/z 191.00 to 176.05 (for scopoletin). The dwell time, Q1 pre-bias, collision energy and Q3 pre-bias were set at 100 ms, 14 V, 17 eV, 21 V for esculetin, 100 ms, 13 V, 15 eV, 19 V for fraxetin, and 100 ms, 22 V, 16 eV, 10 V for scopoletin. The esculetin, fraxetin and scopoletin ion peaks were detected at the retention of 1.4, 1.5 and 2.0 min, respectively. Standards of esculetin (Tokyo Chemical Industry, Tokyo, Japan), fraxetin (Selleck chemicals, Houston, TX, USA) and scopoletin (Tokyo Chemical Industry, Tokyo, Japan) were used.

2.8. Evaluation of the Fungistatic Effect of Fraxetin Against H. fraxineus

To evaluate the fungistatic effect of fraxetin, the germination test of H. fraxineus ascospores was performed. Fraxetin was dissolved in 100% Dimethyl sulfoxide (DMSO) (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) solution and added to 2.5 mL of autoclaved 1.5% water agar (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) in sterilized plastic Petri dish (60 × 15 mm, Kord-Valmark, Brampton, ON, Canada). Final concentration of each chemical in a medium was adjusted to 500 μM fraxetin and 0.1% DMSO. For control, 1.5% water agar plate and mock (0.1% DMSO 1.5% agar) were used. Apothecia were fixed inside the lid of the Petri dish for a few hours to discharge ascospores on the agar plates. The plates with ascospores were kept under same condition as leaflet inoculation. After 2 days incubation, ascospores produce an appressorium or elongate germ tube(s) longer than width of ascospores were counted as germination. A total of 300 ascospores were counted in one apothecium, and the means of germination rate were calculated from 6 apothecia replicates for each treatment. Means for each treatment were compared for significant differences using Tukey–Kramer test at p < 0.05. The germination test was performed three times.

2.9. Statistical Analysis

Statistical analyses were performed using (SPSS version 29). For comparisons between two groups, a two-tailed Student’s t-test was applied. For the compound leaf inoculation assay, defoliation dates and survival rates were analyzed using two-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test. For ascospore germination assays, multiple comparisons were conducted using Tukey–Kramer’s test. A p-value < 0.05 was considered statistically significant.

3. Results

3.1. Inoculation Test Using Compound Leaves to Evaluate Tolerance of Two Fraxinus Species Against H. fraxineus

In all three inoculation experiments, all inoculated F. angustifolia leaflets fell off (

Table 1. Summary of inoculation experiments on compound leaves of two Fraxinus species inoculated with Hymenoscyphus fraxineus.

Species	Inoculum	Experiment 1		Experiment 2		Experiment 3
		Fallen or Dead Leaves 2 MPI (%)	DPI Required for Fallen or Dead Leaves (Average ± SD)	Fallen or Dead Leaves 2 MPI (%)	DPI Required for Fallen or Dead Leaves (Average ± SD)	Fallen or Dead Leaves 2 MPI (%)	DPI Required for Fallen or Dead Leaves (Average ± SD)
Fraxinus excelsior	Hymenoschiphus fraxineus	100 *	18.6 ± 6.4	40 *	43.0 ± 0	20 *	40
	control	0 *	-	0 *	-	0 *	-
F. angustifolia	H. fraxineus	100 *	8.0 ± 2.7	100 *	16.2 ± 8.8	100 *	10.8 ± 2.8
	control	0 *	-	0 *	-	0 *	-

Compound leaves of Fraxinus excelsior and F. angustifolia were inoculated with H. fraxineus, and infection outcomes were assessed at the indicated time points. Five leaves were inoculated per experiment for each species. Statistical significance of categorical data was evaluated using Fisher’s exact test with Bonferroni adjustment (p < 0.05). Values represent means ± SEM from independent experiments. *; indicates significant differences among inoculated and mock treatment leaflets of Fraxinus species.

). The average defoliation dates were 8.0 ± 2.7 dpi, 16.2 ± 8.8 dpi and 10.8 ± 2.8 dpi in each independent experiment. For F. excelsior, all inoculated leaflets fell off only in the first experiment on F. excelsior; the mean of defoliation date was 18.6 ± 6.4 dpi. In the other two experiments, one or two leaflets fell off or died at 40 and 43.0 ± 0 dpi. The survival rates for F. excelsior were 0%, 60%, and 80% in each experiment, respectively. Symptoms that developed on F. excelsior were suppressed and expanded more slowly than those on F. angustifolia. No symptoms were observed on the control leaves of either species.

3.2. Observation of H. fraxineus Behaviors Under Microscopy

In experiments 1 and 2, no consistent results regarding appressorium formation rate and invasion rate were found. In experiment 1, the appressorium formation rate was higher on F. angustifolia (5 dpi: 83.3 ± 12.36%; 7 dpi: 88.7 ± 5.44%) than on F. excelsior (5 dpi: 32.0 ± 5.31%; 7 dpi: 48.7 ± 10.08%) (p < 0.05) (

Table 2. Quantification of early infection behavior of Hymenoscyphus fraxineus on leaflets of Fraxinus excelsior and F. angustifolia using a whole-leaf clearing and staining method.

Species	Experiment 1						Experiment 2
	Appressorium Formation Rate (%) a		Invasion Rate (%) b		Adjacent Cell Invasion Rate (%) c		Appressorium Formation Rate (%) a		Invasion Rate (%) b		Adjacent Cell Invasion Rate (%) c
	5 dpi	7 dpi	5 dpi	7 dpi	5 dpi	7 dpi	5 dpi	7 dpi	5 dpi	7 dpi	5 dpi	7 dpi
Fraxinus excelsior	32.0 ± 5.31 *	48.7 ± 10.0 8 *	30.3 ± 7.59 *	34.0 ± 8.98 *	2.2 ± 1.57	2.2 ± 3.14	76.3 ± 5.31	87.3 ± 4.19	30.7 ± 2.49	53.0 ± 9.80 *	2.2 ± 1.57	11.1 ± 6.85 *
F. angustifolia	83.3 ± 12.36 *	88.7 ± 5.44 *	52.3 ± 6.55 *	55.0 ± 11.31 *	4.4 ± 4.16	30.0 ± 15.15	53.0 ± 27.82	71.7 ± 9.57	28.0 ± 4.55	33.3 ± 1.89 *	10.0 ± 9.81	52.2 ± 8.75 *

Early infection stages were assessed microscopically following ascospore inoculation. (^a) Appressorium formation, expressed as the number of ascospores forming appressoria per 100 ascospores. (^b) Penetration frequency, expressed as the number of ascospores producing invasive hyphae into epidermal cells per 100 ascospores with appressoria. (^c) Restriction of intracellular hyphae, expressed as the number of infection sites in which hyphal growth was restricted within a single epidermal cell per 30 ascospores with intracellular hyphae. For each treatment, three detached leaflets were analyzed per experiment. At least 100 ascospores were scored per leaflet for appressorium formation, and subsequent infection stages were quantified using the denominators described above and in the Section 2. Values represent means ± SEM from independent experiments. *; indicates significant differences between Fraxinus species.

). In contrast, in experiment 2, there was no significant difference in appressorium formation rate between F. excelsior (5 dpi: 76.3 ± 5.31%; 7 dpi: 87.3 ± 4.19%) and F. angustifolia (5 dpi: 53.0 ± 27.82%; 7 dpi: 71.7 ± 9.57%). Similarly, in experiment 1, the invasion rate was significantly lower on F. excelsior (5 dpi: 30.3 ± 7.59%; 7 dpi: 34.0 ± 8.98%) than on F. angustifolia (5 dpi: 52.3 ± 6.55%; 7 dpi: 55.0 ± 11.31%) (p < 0.05). By contrast, in experiment 2, the invasion rate at 7 dpi was significantly higher on F. excelsior (53.0 ± 9.80%) than on F. angustifolia (33.3 ± 1.89%) (p < 0.05), while no significant difference was observed at 5 dpi (F. excelsior: 30.7 ± 2.49%; F. angustifolia:28.0 ± 4.55%). Although appressorium formation rates and initial penetration frequencies showed some variability between independent experiments, invasion into adjacent epidermal cells and subsequent palisade mesophyll invasion were consistently reduced in F. excelsior compared with F. angustifolia.

Invasion to adjacent cells was observed in both Fraxinus species (Figure 1). However, the adjacent cell invasion rate was lower on F. excelsior (experiment 1: 2.2 ± 3.14%, experiment 2: 11.1 ± 6.85%) compared to F. angustifolia (experiment 1: 30.0 ± 15.15%, experiment 2: 52.2 ± 8.75%) on 7 dpi. Experiment 1 showed a decreasing trend and experiment 2 revealed a significant decrease on F. excelsior (p < 0.05). No significant differences in adjacent cell invasion rates were found between F. excelsior (experiment 1: 2.2 ± 1.57%; experiment 2: 2.2 ± 1.57%) and F. angustifolia (experiment 1: 4.4 ± 4.16%; experiment 2: 10.0 ± 9.81%) on 5 dpi.

Suppressed growth of H. fraxineus was observed on the transverse section (Figure 2). Palisade mesophyll cell invasion rate was significantly lower on F. excelsior (experiment 1: 24.0 ± 4.32%; experiment 2: 31.3 ± 16.68%) than F. angustifolia (experiment 1: 84.7 ± 3.40%; experiment 2: 80.7 ± 4.11%) on 7 dpi (p < 0.05,

Table 3. Quantification of early infection behavior of Hymenoscyphus fraxineus on transverse sections of ash leaflets.

Species	Experiment 1	Experiment 2
	Palisade Mesophyll Cell Invasion Rate (%)	Palisade Mesophyll Cell Invasion Rate (%)
Fraxinusexcelsior	24.0 ± 4.32 *	31.3 ± 16.68 *
F. angustifolia	84.7 ± 3.40 *	80.7 ± 4.11 *

Early infection behavior of H. fraxineus was examined on transverse sections of leaflets of Fraxinus excelsior and F. angustifolia at 7 days post inoculation (dpi). Penetration frequency was expressed as the number of ascospores producing invasive hyphae into epidermal cells per 50 ascospores. Three independent biological replicates were analyzed for each species. Statistical differences were evaluated using a t-test (p < 0.05). *; indicates significant differences between Fraxinus species.

). Haustorium-like structures or intercellular hyphae were not observed.

3.3. Host Defense Responses Against H. fraxineus Infection

H₂O₂ accumulation was detected in penetrated epidermal cells of both F. excelsior and F. angustifolia; however, no clear difference in staining patterns was observed between the two species at the examined time points. Callose deposition, lignification, and suberization were not observed in the invaded or surrounding cells of either F. excelsior or F. angustifolia. However, lignification and suberization were detected on the leaf veins of control and inoculated samples of both species. Polyphenols were detected in F. excelsior and F. angustifolia by Fe(II) staining, regardless of inoculation with H. fraxineus (Figure 3).

3.4. Coumarins Detection by LC–ESI–MS/MS and Germination Test

LC–ESI–MS/MS analysis revealed the presence of fraxetin in the leaflets of F. excelsior (

Table 4. Concentration of selected coumarins in leaflets of Fraxinus excelsior and F. angustifolia.

Species	Esculetin (ng/g Fresh Weight)	Fraxetin (ng/g Fresh Weight)	Scopoletin (ng/g Fresh Weight)
Fraxinusexcelsior
Pre-inoculation	n. d.	121.2	n. d.
Non-inoculation	n. d.	87.3	n. d.
Inoculation	n. d.	76.5	n. d.
F. angustifolia
Pre-inoculation	n. d.	n. d.	n. d.
Non-inoculation	n. d.	n. d.	n. d.
Inoculation	n. d.	n. d.	n. d.

The concentrations of esculetin and fraxetin were quantified in methanol-extracted leaflets of F. excelsior and F. angustifolia under the indicated conditions. Values represent mean of independent biological replicates. n.d. indicates that the compound was not detected.

). Neither esculetin nor scopoletin were detected. The amount of fraxetin did not increase in inoculated samples of the tolerant F. excelsior. None of the three coumarins were detected in the samples of F. angustifolia.

To evaluate the effects of fraxetin on ascospore germination, germination tests were demonstrated. The result showed that treatment with 500 μM fraxetin significantly inhibited ascospore germination compared with the DMSO and 1.5% agar controls (Figure 4). Significant differences were confirmed in three independent experiments. This assay quantified ascospore germination and early hyphal growth and does not provide information on fungal viability. Therefore, the observed inhibition reflects the suppression of early developmental processes rather than direct evidence of fungicidal activity.

4. Discussion

In this study, we investigated early-stage interactions between H. fraxineus and two Fraxinus species with contrasting susceptibility, focusing on ascospore-mediated infection. Under artificial inoculation conditions, compound leaves of F. angustifolia rapidly abscised following infection, whereas those of F. excelsior survived longer and exhibited suppressed symptom development (Table 1), indicating a higher level of tolerance in F. excelsior. Microscopic analyses using detached leaflets further demonstrated that fungal growth was significantly restricted on F. excelsior compared with F. angustifolia at 7 days post inoculation (dpi) (Table 2 and Table 3). Notably, although no consistent differences were observed in appressorium formation or initial penetration, invasion into adjacent epidermal cells was markedly restricted in F. excelsior. Together, these results indicate that tolerance in F. excelsior is associated with the suppression of fungal growth after penetration rather than at pre-penetration stages.

Ascospores are a major source of ash dieback disease in natural forests; however, studies examining ascospore-mediated infection processes and early fungal behavior in ash species with different tolerance levels remain limited [25,26,27]. In the present study, the compound leaves inoculation assay involved wounding of the petiole, which may bypass surface-based defense barriers and induce wound-associated responses. Therefore, this assay was not intended to model natural infection processes, but rather to provide a standardized comparison of tolerance between host species (Table 1). For this reason, early infection dynamics were primarily investigated using ascospore inoculation of detached leaflets without wounding (Table 2 and Table 3). A limitation of this study is that the susceptible control species (F. angustifolia) is phylogenetically distinct from F. excelsior, and interspecific differences in leaf anatomy and basal metabolite profiles may influence infection outcomes. Future studies incorporating susceptible F. excelsior genotypes from Europe and additional Fraxinus species will strengthen comparative inference. In addition, because our experiments were conducted using a single Japanese isolate of H. fraxineus (Su-2003s), the observed infection dynamics and early-stage growth suppression may be isolate-dependent. Studies using multiple isolates, including European isolates with different aggressiveness backgrounds, will be required to generalize these findings.

To elucidate the defense mechanisms underlying tolerance in F. excelsior, we focused on the earliest stages of infection. Endophytic behavior of H. fraxineus in F. excelsior growing in the Sapporo Experimental Forest of Hokkaido University has been reported previously [28], leading us to hypothesize that fungal growth is suppressed after host penetration. Consistent with this hypothesis, invasion into adjacent epidermal cells was consistently restricted in F. excelsior, clearly distinguishing it from susceptible F. angustifolia. This restriction became evident after successful penetration of the initially infected epidermal cell, indicating that early-stage suppression of H. fraxineus in F. excelsior is mediated by post-penetration defense processes rather than by pre-penetration resistance mechanisms. Although the underlying mechanisms were not directly examined, constitutive chemical defenses may suppress hyphal extension after penetration, while localized reinforcement of cell walls could further limit cell-to-cell spread. In addition, spatially confined defense responses that do not manifest as classical hypersensitive cell death may contribute to limiting fungal progression. Consistent with this interpretation, we did not observe clear hypersensitive cell death or strong induction of reactive oxygen species during early infection stages. Together, these observations suggest that post-penetration defense processes play a central role in restricting adjacent cell invasion during early host–pathogen interactions.

In addition to constitutive coumarins, alternative hypotheses may also explain the restricted early progression of H. fraxineus on F. excelsior. Physical surface barriers were not examined in this study, although plant cuticles, epicuticular waxes, and trichomes are known to function as biologically active interfaces that influence fungal attachment, penetration, and early infection processes [29,30]. Moreover, fraxetin is unlikely to act alone within the chemical defense layer. Coumarins represent one branch of phenylpropanoid-derived specialized metabolism involved in plant–microbe interactions, and phenylpropanoid- and lignin-associated defense networks may act synergistically to modulate antifungal outcomes [31,32]. Moreover, the leaf-associated microbiome, including endophytic bacteria and fungi, may contribute to the early suppression of H. fraxineus, although this aspect was not examined here. Both bacterial and fungal community composition in ash leaves has been linked to infection intensity and disease outcomes in a host genotype-dependent manner, and several leaf-associated bacteria with putative antagonistic activity against H. fraxineus have been identified [33,34,35]. Experimental inoculation with selected bacterial strains or consortia has also been shown to reduce ash dieback symptoms under controlled conditions [36]. Together, these findings suggest that physical traits, broader constitutive metabolite profiles, and microbiome-mediated interactions may synergize with coumarins to restrict the post-penetration spread of H. fraxineus. Future studies integrating epidermal trait characterization, microbiome profiling, and metabolite analyses will be required to test these complementary mechanisms.

Phenolic compounds were detected in transverse sections of both F. angustifolia and F. excelsior regardless of infection status (Figure 3), indicating that phenolic compounds are constitutively present in ash leaf tissues. However, Fe(II) staining provides only qualitative information and does not resolve the identity, concentration, or spatial distribution of individual polyphenolic compounds. Therefore, the presence of polyphenols in both species does not preclude functionally important differences in polyphenol composition, abundance, or localization that may contribute to tolerance in F. excelsior. Such differences may underlie the observed restriction of fungal progression and warrant further investigation. Accordingly, these assays were not intended for statistical comparison but for qualitative assessment of typical defense marker occurrence.

In this study, we focused on esculetin and fraxetin, two coumarins that have been identified as candidate compounds associated with tolerance to ash dieback disease [19,37,38,39,40]. Previous studies demonstrated fungistatic effects of coumarins and confirmed inhibition of H. fraxineus growth by esculetin and fraxetin [19]. Consistent with these reports, fraxetin was detected constitutively in F. excelsior leaflets but not in F. angustifolia, and its levels did not increase following inoculation (Table 4). These findings suggest that fraxetin functions as a pre-existing chemical barrier rather than as an inducible defense compound. Because metabolite localization at infection sites was not examined in this study, future analyses combining chemical imaging and microscopy will be required to clarify the spatial dynamics of fraxetin during infection.

Germination assays further supported a growth-suppressive role of fraxetin. Although the concentration used in this study was lower than that reported previously [19], significant inhibition of ascospore germination was observed (Figure 4). Because our assays focused on ascospore germination and early hyphal growth, we cannot distinguish whether fraxetin acts in a fungistatic or fungicidal manner against H. fraxineus. Although the precise mode of action was not examined here, coumarins have been proposed to interfere with fungal cellular processes such as membrane integrity, redox homeostasis, or energy metabolism, which could plausibly impair ascospore germination and early hyphal growth [41]. Elucidating the direct molecular targets of fraxetin will require dedicated mechanistic and viability-based analyses in future studies. At present, our data support a fungistatic effect during early infection stages, while definitive assessment of fungicidal activity will require viability-based analyses in future studies. Differences between our results and previous reports [19,42], which did not detect fraxetin in ash leaves, may reflect variation in environmental conditions, populations, or genotypes. Scopoletin, a precursor of fraxetin [43], exhibits fungistatic activity against various fungi [44,45] but was not detected in our samples, suggesting that scopoletin itself is not required for growth suppression but may contribute to fraxetin biosynthesis. While fraxetin inhibited ascospore germination in vitro, its protective role in planta was not directly tested. Exogenous application experiments on intact plants would help clarify the contribution of coumarins to disease suppression, but were beyond the scope of this study and remain an important topic for future investigation.

Other candidate compounds associated with tolerance to H. fraxineus, including iridoid and secoiridoid glycosides and additional phenolic compounds, were not analyzed in this study. Previous studies reported accumulation of polyphenolic compounds around infection sites in F. excelsior and F. mandshurica [17,18], suggesting a contribution of induced resistance. However, our data did not show increased fraxetin accumulation following inoculation, indicating that constitutive and inducible defenses may operate at different stages of infection. Whether fraxetin content correlates with field-level resistance among European ash populations remains unknown but could be relevant for resistance breeding. In addition, the regulatory mechanisms underlying fraxetin biosynthesis remain unclear and will require transcriptomic analyses. Finally, although strong induction of typical defense markers was not detected at early infection stages, inducible responses such as defense priming or systemic acquired resistance may become relevant during later stages of host–pathogen interaction.

In conclusion, our results demonstrate that early-stage growth of H. fraxineus is suppressed on F. excelsior compared with F. angustifolia, and that this suppression is associated with post-penetration defense processes and constitutive chemical traits, including fraxetin. These findings provide new insights into early defense mechanisms against H. fraxineus and contribute to a better understanding of ash dieback tolerance.