Comparative Metabolomics Reveals the Adaptive Strategy of White Auricularia cornea to Bamboo Substrate Variation

TL;DR

This study explores how white Auricularia cornea adapts to bamboo substrates, finding that a 58% bamboo mix optimizes growth and provides insights for sustainable mushroom cultivation.

Contribution

The study identifies optimal bamboo substrate ratios and metabolic adaptation mechanisms for white Auricularia cornea cultivation.

Findings

White A. cornea grows best with 58% bamboo substrate, showing a mycelial growth rate of 3.55 mm/d and shortest growth period of 86.2 days.

Metabolomics analysis detected 3779 metabolites, with amino acids and organic acids being the most abundant.

KEGG pathway analysis showed that bamboo substrate ratios influence growth adaptation through nucleotide metabolism and ABC transporters.

Abstract

To address the “fungus-forest conflict” in the edible mushroom industry and the challenge of resource utilization for bamboo substrate waste, this study focused on white Auricularia cornea, and cultivation systems were established with bamboo substrate replacing wood chips at ratios of 0%, 18%, 38%, 58%, and 78%. By integrating liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis with agronomic trait measurements, the study elucidated the metabolic adaptation mechanisms of the substrate. Results indicated that white A. cornea could grow normally in all bamboo substrates, with the 58% bamboo substrate replacement group (D_58) demonstrating the most optimal overall performance. The mycelial growth rate reached 3.55 ± 0.24 mm/d, and the growth period was the shortest (86.2 d), balancing growth efficiency with cost advantages. Metabolomics detected 3779 metabolites, primarily amino acids and their derivatives (42.2%) and organic acids (35.54%). Compared to the control group, each treatment group exhibited 104–528 upregulated and 192–630 downregulated differential metabolites, with 93 shared differential metabolites and numerous unique markers. KEGG pathway enrichment analysis revealed that varying bamboo substrate ratios shaped growth adaptation strategies by regulating core pathways such as nucleotide metabolism and ABC transporters. This study established the feasibility and optimal formulation of bamboo substrate substitution, elucidated the substrate–metabolite–phenotype linkage mechanism, and provided theoretical foundations and practical references for high-quality cultivation of white A. cornea and sustainable development through “substituting bamboo for wood” to reduce carbon emissions.

Figures

Full text

1. Introduction

Auricularia cornea is an important edible mushroom that is widely cultivated and consumed [1]. It features a thick, fleshy texture with a crisp and tender mouthfeel, and is rich in polysaccharides, dietary fiber, essential amino acids, and various trace elements such as iron and calcium [2,3]. White A. cornea (also known as “Yumuer”), as a premium variety of A. cornea [4], has emerged as a hot new cultivar in the edible mushroom industry [5]. Its distinctive advantages—a pure white color and delicate texture—combined with rich functional components such as protein and polysaccharides, have driven its popularity. However, the cultivation of white A. cornea is typically more sensitive to substrate and environmental conditions. Their growth patterns, nutritional metabolism pathways, and ultimate quality formation mechanisms under different cultivation substrates may differ significantly from conventional varieties. Therefore, deciphering their substrate adaptability and secondary metabolic characteristics holds significant strategic importance for expanding high-end edible mushroom categories and enhancing industrial value-added potential.

The growth, development, yield formation, and quality characteristics of edible mushroom are highly dependent on the nutrient supply and physicochemical properties of the cultivation substrate. The composition and ratio of substrate components are among the core factors influencing the economic efficiency of the edible mushroom industry [6]. Traditional edible mushroom cultivation primarily utilizes broadleaf wood chips as the core substrate; however, in recent years, with increasingly strained forest resources, global cultivation practices have rapidly shifted from log-based to substrate substitute-based methods; agricultural and forestry byproducts such as cottonseed hulls, corn cobs, and sugarcane bagasse are now widely employed in the commercial production of both grass and wood rot fungi. Previous studies have demonstrated that, in addition to conventional wood chips, substrates such as cassava straw [7], spent mushroom compost [8], eucalyptus wood chips [9], de-oiled camphor leaves [10], and goji berry wood chips [11] could all support the normal growth of A. cornea. The nutritional richness provided by different substrates varies, indirectly influencing mycelium growth and development, the agronomic traits of fruiting bodies, and the content of nutritional components. Under conditions where 70% of the substrate was replaced with corn cobs, both yield traits and quality traits of Auricularia heimuer showed significant improvement [12]; however, when corn cobs fully replaced the substrate, the free amino acid metabolism pathways in the fruiting bodies exhibited significant enrichment [13]. In an experiment cultivating oyster mushrooms used corn distillers’ grains, researchers found that adding corn distillers’ grains at different ratios altered the composition of both non-volatile and volatile compounds in the mushrooms [14]. Metabolomics analysis of A. cornea cultivated on de-oiled camphor leaf substrate identified 516 differential metabolites, with alkaloids, phenolics, and flavonoids accounting for 26.7% of the total [10]. In another metabolomics study of pink A. cornea growth on goji berry sawdust substrate, 1335 metabolites were identified, primarily comprising lipids and lipoid molecules, organic acids and their derivatives, aromatic compounds, and organic heterocyclic compounds [11]. These studies profoundly demonstrate that altering cultivation substrates directly triggers metabolic reprogramming in edible mushroom, thereby influencing their final quality. However, current metabolomics research on white A. cornea remains relatively scarce, particularly lacking systematic exploration of a substrate substitution’s impact on their metabolic profiles, existing studies remain focused on preliminary optimization of cultivation formulations, failing to deeply reveal the intrinsic relationship between substrate composition and metabolic products.

China is the world’s most abundant source of bamboo resources, ranking first globally in both bamboo forest area and bamboo production, annually, over ten million tons of waste materials—including bamboo substrate and bamboo branches—which are generated from bamboo processing [15,16]. Most of this bamboo substrate is either haphazardly piled or burned, resulting not only in the waste of valuable biomass resources but also imposing severe environmental pressures [17]. Therefore, developing high-value utilization pathways for bamboo substrate is urgently needed. On the other hand, the traditional wood-decaying fungus industry heavily relies on broadleaf wood chips, long-term overconsumption has threatened forest ecosystems, exacerbating the “fungus-forest conflict.” Moreover, excessive logging undermines forest carbon sequestration capacity. Bamboo, however, grows rapidly (maturing in 3–5 years), exhibits carbon fixation efficiency 2–3 times that of timber, and possesses a carbon footprint merely 50–70% of virgin wood chips [18,19]. Therefore, converting bamboo substrates into cultivation materials for edible mushroom is regarded as a viable approach to achieving “substituting grass for wood” and alleviating pressure on forest resources [20]. The chemical composition of bamboo substrate closely resembles that of broadleaf wood chips, being rich in lignocellulosic components such as cellulose, hemicellulose, and lignin. The combined content of cellulose and hemicellulose accounts for approximately 65% [21]. Bamboo is widely available, low-cost, and highly renewable; additionally, it is nutritionally rich and abundant in bioactive compounds like bamboo polysaccharides, offering natural advantages as a substrate for edible mushroom cultivation [22]. A review of existing research indicates that bamboo substrates have been tested for cultivating A. heimuer [23], oyster mushrooms [24,25], shiitake mushrooms [26], common A. cornea [27], Hericium erinaceus [28], and other edible mushroom. There are a few reports exist on the suitability of bamboo substrates for white A. cornea [29], but research on how bamboo substrates replace sawdust to influence white A. cornea metabolites and the mechanisms of lignocellulose conversion remains unexplored, and systematic studies are urgently needed to address this gap.

Based on the aforementioned research background and gaps, this study focuses on white A. cornea. We established cultivation systems substituting wood chips with bamboo substrates at varying ratios. By integrating liquid chromatography–mass spectrometry (LC-MS) non-targeted metabolomics technology, we systematically analyzed the differences in the metabolite profiles of white A. cornea grown on bamboo substrates. The primary objectives of this study can be summarized in three points: First, to validate the acceptable range of bamboo substrate as a single variable affecting fruiting body agronomic traits; second, to map high-resolution metabolic profiles at different substrate levels and screen biomarkers indicative of substrate adaptability; third, to reveal up- or downregulated trends in key metabolic pathways using KEGG pathway enrichment and differential abundance score (DA Score) models, providing molecular endpoints for subsequent precision formulation design. Ultimately, this study aims to clarify the practical feasibility of bamboo substrate substitution, decipher differences in metabolite profiles, and elucidate metabolic reprogramming mechanisms, thereby providing a theoretical basis for substrate optimization in the high-quality and efficient cultivation of white A. cornea.

2. Materials and Methods

2.1. Experimental Design and Mushroom Cultivation

The strain of white A. cornea (deposit number ACW2025-30) used in this study is currently deposited at the College of Horticulture, Jilin Agricultural University. The bamboo substrate employed in the experiments originated from Zunyi City, Guizhou Province, China, and underwent pulverization (Figure S1). The cultivation substrate formulation for white A. cornea cultivation comprised pulverized oak sawdust substrate (purchased from Jilin, China), bamboo substrate, and several auxiliary materials. Specifically, the control group formulation (A_CK) represents the most conventional recipe in mushroom cultivation trials, composed of 78% oak sawdust, 20% wheat bran, 1% lime, and 1% gypsum. Based on the control formulation, four experimental formulations with varying bamboo substrate replacement ratios were designed as follows: B_18 (60% oak sawdust, 18% bamboo substrate, 20% wheat bran, 1% lime, 1% gypsum), C_38 (40% oak sawdust, 38% bamboo substrate, 20% wheat bran, 1% lime, 1% gypsum), D_58 (20% oak sawdust, 58% bamboo substrate, 20% wheat bran, 1% lime, 1% gypsum), and E_78 (78% bamboo substrate, 20% wheat bran, 1% lime, 1% gypsum). The total carbon and total nitrogen contents of the five culture medium formulations are as follows: A_CK: total carbon 458.16 ± 4.66 g/kg, total nitrogen 10.32 ± 0.35 g/kg, C/N 44.41; B_18: total carbon 462.59 ± 5.22 g/kg, total nitrogen 9.84 ± 0.08 g/kg, C/N 47.01; C_38: total carbon 478.14 ± 1.60 g/kg, total nitrogen 9.85 ± 0.22 g/kg, C/N 48.54; D_58 total carbon 463.63 ± 12.81 g/kg, total nitrogen 10.33 ± 0.14 g/kg, C/N 44.87; E_78 total carbon 477.85 ± 7.33 g/kg, total nitrogen 8.41 ± 0.30 g/kg, C/N 56.84 (total carbon and total nitrogen content were determined using elemental analysis, following the standard for determining the composition of agricultural biomass raw materials NY/T 3498—2019 [30]).

Each medium formulation was prepared in 30 replicate bags, each containing 500 g of substrate with a moisture content of 55 ± 2%. After filling, the bags were sterilized at 121 °C for 120 min and inoculated uniformly after cooling. Post-inoculation cultivation proceeded under previously established conditions [31]. Specifically, after mycelium fully colonized the bags, incubation continued for approximately 7 days until physiological maturation concluded. Bags were then perforated using a mechanical perforator for small-hole fruiting [32]. Once mycelium regrew and turned white at the perforation sites within 3–5 days, bags were transferred to the fruiting greenhouse for fruiting trials. Spray water 5–6 times daily and ventilate 2–3 times daily, maintaining alternating dry and moist conditions. Keep greenhouse humidity between 85% and 95% and temperature around 22 °C under normal diffused light. Harvest when the ear is fully expanded with slightly contracted and inward-curled edges, and the underside of the ear shows erect hairs [33].

2.2. Sampling Strategy

After the fruiting bodies matured, five mushroom bags were randomly selected from a total of 30 bags. Three fruiting bodies were harvested from each bag, yielding a total of 15 fruiting bodies. These were then cut into small pieces (1 cm × 1 cm) and mixed together. Subsequently, 5 g of the mixture was transferred to a sterile 50 mL centrifuge tube and immediately immersed in liquid nitrogen for 3–5 min. After thorough pre-cooling, the sample was stored at −80 °C until metabolomics analysis was performed [11].

2.3. Determination of Average Mycelial Growth Rate, Growth Period, Fruiting Body Traits, and Yield

We prepared five sets of culture media formulations according to specified ratios and dispensed 15 g into each glass plate. We compacted the contents, then sterilized them in an autoclave at 121 °C for 120 min. After cooling, we inoculated the activated white A. cornea strain into the plate culture medium, and incubated them at 25 °C for 7 days, then measured and recorded the average mycelium growth rate using the cross-streak method, with 5 replicates per formulation [34]. Agronomic trait assessment followed the Guidelines for Specificity, Uniformity, and Stability Testing of A. cornea (NY/T 3729—2020) [35]. Growth period: Number of days from inoculation to first fruiting body maturity and harvest, with 5 bags per group. After soaking fresh mature ears at room temperature for 4 h, they were drained thoroughly. Ear Length: We measured ear length using a digital vernier caliper. Ear length refers to the distance from the foremost apex of the ear’s leading edge to the base of the ear. Ear Width: We measured ear width using a digital vernier caliper. Ear width refers to the measurement at the widest point perpendicular to the mushroom bag. Ear thickness: We measured the thickness at the center of the ear using a digital vernier caliper, and this value represents the ear thickness. Ear dry weight: Fresh mature fruiting bodies were dried in a 45 °C oven to constant weight, and then we determined the weight of the ear on an electronic analytical balance. Rehydration ratio of fruiting body ears: The ratio of wet weight to dry weight of the ear is the rehydration ratio, commonly expressed numerically. Each group was measured with 15 replicates. Single-bag yield: The total dry weight of fruiting bodies produced per bag of culture medium throughout the entire growth period was measured, with 5 bags per group as replicates [33].

2.4. Extraction of Metabolites

We used vacuum freeze-drying technology, placed the 30 biological samples (five groups of formulations, each with six biological replicates, totaling 30 replicates) in a lyophilizer (Scientz 100F, Scientz Freeze-Drying, Ningbo, China), then grinded (30 Hz, 1.5 min) the samples to powder form using a grinder (MM 400, Retsch, VERDER Company, Haan, Germany). Next, we weighed 30 mg of sample powder using an electronic balance (MS105DM, Mettler Toledo Instruments Co., Ltd., Zurich, Switzerland) and added 1500 uL of −20 °C pre-cooled 70% methanolic aqueous internal standard extract (less than 30 mg added at the rate of 1500 uL extractant per 30 mg sample). They were vortexd once every 30 min for 30 s, for a total of 6 times. After centrifugation (rotation speed 12,000 rpm, 3 min), the supernatant was aspirated, and the sample was filtered through a microporous membrane (0.22 um pore size) and stored in the injection vial for UPLC-MS/MS analysis.

2.5. LC-MS/MS Analysis for Metabolomics

All samples were acquired by the LC-MS (Vanquish, Thermo Scientific, Waltham, MA, USA; Q Exactive HF-X, Thermo Scientific, MA, USA) system followed by machine orders. The analytical conditions were as follows: UPLC: column, Waters ACQUITY UPLC HSS T3 1.8 µm, 2.1 mm × 100 mm; column temperature, 40 °C; flow rate, 0.40 mL/min; injection volume, 4 uL; solvent system, water (0.1% formic acid): acetonitrile (0.1% formic acid); sample measurements were performed with a gradient program that employed the starting conditions of 95% A, 5% B. Within 5 min, a linear gradient to 35% A, 65% Bwas programmed; within 1 min, a linear gradient to 1% A, 99% B was programmed, and kept for 1.5 min. Subsequently, a composition of 95% A, 5.0% B was adjusted within 0.1 min and kept for 2.4 min. All the methods alternated between full scan MS and data-dependent MSn scans using dynamic exclusion. MS analyses were carried out using electrospray ionization in the positive ion mode and negative ion mode using full scan analysis over m/z 84–1250 at 35,000 resolution. Additional MS settings were ion spray voltage 3.5 KV or 3.2 KV in positive or negative modes, respectively. Sheath gas (Arb) 30. Aux gas 5. Ion transfer tube temperature 320 °C. Vaporizer temperature 300 °C. Collision energy 30, 40, 50 V. Signal Intensity Threshold 1 × 10^6^ cps. Top N vs. Top speed 10. Exclusion duration 3 s.

2.6. Data Preprocessing

Raw mass spectrometry data were converted to mzML format using ProteoWizard (v 3.0), followed by peak extraction, alignment, and retention time correction via the XCMS program. Peaks with a missing rate > 50% in any sample group were filtered out, then blank values were imputed using KNN + 1/5 minimum value (1/5 minimum value for blank values > 50%, KNN for blank values < 50%). Peak area correction was performed using the SVR method. Corrected and filtered peaks underwent metabolite identification by searching the lab’s proprietary database, integrating public repositories, utilizing prediction databases, and applying the metDNA method. Finally, compounds with a composite identification score ≥ 0.5 and quality control (QC) samples with coefficient of variation (CV) value < 0.5 were extracted. Positive and negative modes were merged (retaining compounds with the highest qualification level and lowest CV), yielding the all_sample_data.xlsx file.

2.7. Data Processing and Bioinformatics Analysis

Experimental results were expressed as mean ± standard deviation (n ≥ 3). One-way ANOVA and Duncan’s Multiple Range Test were performed using SPSS 26.0 to determine significant differences between samples (p < 0.05). Graphs were generated using OriginPro 2022 software. Instrument stability was assessed using QC samples, with one QC sample inserted every 10 analytical samples. Total ion chromatograms (TICs) from different QC samples were overlaid for analysis, alongside Pearson correlation coefficients and CV to evaluate data reliability. Principal component analysis (PCA) was performed using the built-in prcomp function in R software (v 4.1.2) (www.r-project.org/, accessed on 21 November 2025) [36]. Metabolite content data underwent UV (unit variance scaling) processing, followed by hierarchical clustering analysis and heatmap visualization using the ComplexHeatmap package (v 2.9.4) in R. Orthogonal partial least squares discriminant analysis (OPLS-DA) was applied to the centered raw data [37]. For two-group analysis, differential metabolites (DMs) were determined by VIP (VIP > 1) and absolute Log2FC (|Log2FC| ≥ 1.0). VIP values were extracted from OPLS-DA result, which also contain score plots and permutation plots, and were generated using R package MetaboAnalystR (v 1.0.1). The data was log transformed (Log2) and underwent mean centering before OPLS-DA. In order to avoid overfitting, a permutation test (200 permutations) was performed. Identified metabolites were annotated using KEGG Compound database (http://www.kegg.jp/kegg/compound/, accessed on 21 November 2025), and annotated metabolites were then mapped to KEGG Pathway database (http://www.kegg.jp/kegg/pathway.html, accessed on 21 November 2025) [38].

3. Results

3.1. Changes in Agronomic Traits of White A. cornea Cultivated with Different Proportions of Bamboo Substrate Substitutes

The test strain, white A. cornea, exhibited normal growth on both oak sawdust substrate and four different bamboo substrate replacement formulations (Figure 1). However, significant variations were observed in mycelium growth rate, fruiting body morphology, yield per bag, and growth period. The average mycelial growth rate ranged from 2.75 to 3.55 mm/d. The D_58 group (58% bamboo substrate replacement) exhibited a significantly faster mycelial growth rate (3.55 ± 0.24 mm/d) than other experimental groups, while no significant differences were observed among the remaining four groups (Figure 2a). This growth rate disparity was evident throughout the entire growth period across all five formulations, with the D_58 group exhibiting the shortest growth period (86.2 d), significantly faster than other experimental groups (including the control group, A_CK). This pattern was consistent with the earlier mycelium growth rates on plates. Conversely, the B_18 (18% bamboo substrate replacement) and E_78 (78% full bamboo substrate replacement) groups had the longest growth periods, significantly slower than other experimental groups (Figure 2b).

Among the five formulations, the A_CK group yielded the highest dry fruiting body production per bag (21.86 ± 1.34 g), significantly higher than the other groups. Among the four formulations incorporating bamboo substrate substitution, yield per bag increased with higher substitution levels, peaking at 58% substitution (Figure 2c). This trend likely reflects changes in the carbon-to-nitrogen ratio (C/N) and physicochemical properties of the substrate caused by varying bamboo substrate additions. For morphological differences in fresh fruiting bodies and dry weight of the fruiting body across groups, 15 randomly selected harvested fruiting bodies per group were measured. No significant differences were observed in the fruiting body dry weight among groups. Overall, the fruiting body in the E_78 group were heaviest, while those in the C_38 group (38% bamboo substrate replacement) were lightest (Figure 2d). No significant differences were observed in fresh ear-shaped fruit body length or width. Both dimensions were largest in the E_78 group (46.64 ± 7.89 mm, 57.22 ± 12.20 mm), consistent with the dry weight pattern (Figure 2e,f). The ear thickness was highest in the C_38 group (2.02 ± 0.26 mm) and lowest in the B_18 group (Figure 2g). The rehydration ratio of ear-shaped pieces effectively reflects the quality indicator of abdominal water retention in the fruiting bodies of white A. cornea. The rehydration ratio of the five formulation groups ranged from 6.80 to 7.99, with the highest value observed in the C_38 group (7.99 ± 1.93) and the lowest in the A_CK group (Figure 2h). The control group (A_CK) exhibited relatively poor morphological indicators such as ear length, width, thickness, and rehydration ratio in harvested fruiting bodies. In contrast, the D_58 group produced fruiting bodies with moderately favorable indicators across all aspects, although its yield was lower than the control group, it emerged as the relatively optimal alternative formulation when considering factors like cost. These results indicate that variations in substrate type and proportion significantly influence the metabolic changes in white A. cornea. The metabolic processes of white A. cornea resulted in distinct differences in morphology, growth period, and yield.

3.2. Quality Control for LC-MS/MS Data

QC samples were prepared by pooling sample extracts. During LC-MS/MS analysis, four QC samples were evenly interspersed within each sample analysis run. Overlay analysis of total ion current (TIC) profiles from mass spectrometry detection of different QC samples revealed high curve overlap in metabolite detection (Figure S2A,B), confirming the instrument’s high stability and data output reliability. Pearson correlation analysis of QC samples yielded correlation coefficients exceeding 0.99 in both positive and negative ion modes, indicating minimal inter-sample variation (Figure S2C,D). Further analysis of data dispersion revealed that over 75% of compounds in QC samples exhibited CV values below 0.3, demonstrating exceptional experimental stability (Figure S2E,F). Overall, these observations are satisfactory and fully substantiate the reliability of the experimental results.

3.3. Chemical Classification and Statistical Analysis of Metabolites

This study was conducted in five groups for metabolomics research. After data preprocessing, a total of 3779 metabolites were detected as follows: 2574 in positive ion mode and 1205 in negative ion mode (Table S1). In positive ion mode (Figure 3a,b), metabolites belonged to 20 chemical categories, including amino acids and their derivatives (835), organic acids (300), benzene and its derivatives (227), amines and alcohols (199), alkaloids (163), lipids (162), and heterocyclic compounds (111). In negative ion mode (Figure 3c,d), metabolites were classified into 21 chemical categories, including organic acids (344), amino acids and their derivatives (168), lipids (129), benzene and its derivatives (119), nucleotides and their derivatives (71), glycerophospholipids (62), and phenolic acids (62). Overall, the major chemical classifications were amino acids and their derivatives (42.2%), organic acids (35.54%), benzene and its derivatives (16.74%), lipids (15.12%), amines (9.77%), alkaloids (9.4%), glycerophospholipids (7.66%), phenolic acids (7.48%), and heterocyclic compounds (6.67%) (Figure 3, Table S1). Additionally, OPLS-DA was performed to evaluate metabolite differences in our test material, white A. cornea, grown under different substrate formulations. Results indicated that in comparisons between A_CK and B_18, A_CK and C_38, A_CK and D_58, and A_CK and E_78, the first two principal components explained 43%, 47.4%, 54.76%, and 45.61% of the variance, respectively (Figure S3A,C,E,G). Samples from different groups showed distinct separation, indicating significant differences in metabolite composition. The OPLS-DA model was validated, and its fit and validity were assessed via permutation tests. The Q2 values were exceptionally high (0.934–0.982) and all p-values were below 0.005, indicating the model’s stability and reliability (Figure S3B,D,F,H). Consequently, the VIP values derived from the OPLS-DA model were deemed suitable for subsequent DMs screening.

3.4. Screening, Identification, and Classification of DMs

Metabolomics data are characterized by their “high-dimensionality and massive volume”, necessitating the integration of univariate and multivariate statistical analysis methods. Based on the data’s properties, multi-faceted analysis is required to accurately identify DMs. A total of 3779 metabolites were identified across 30 biological samples. Volcano plots and clustering heatmaps effectively visualized the overall distribution of metabolite differences (Figure 4). These visualizations intuitively displayed the composition of up- and downregulated metabolites between different groups, revealing significant differences in metabolite expression. Compared to the A_CK group, 104 to 528 metabolites were upregulated, while 192 to 630 metabolites were downregulated (Figure 4a,c,e,g). Cluster heatmaps and Venn diagram analyses further revealed distinct changes and expression patterns of significantly different metabolites (SDMs) across groups. Furthermore, certain metabolites exhibited higher relative abundances in the A_CK group, while others were more enriched in samples cultured with varying proportions of bamboo substrate, highlighting the adaptability of white A. cornea to different substrate metabolisms (Figure 4b,d,f,h). Venn diagrams revealed significant differences in SDMs distribution among groups. Analysis showed 93 shared SDMs across groups compared to A_CK, indicating common metabolic responses to substrate changes in white A. cornea (Figure 4i).

Additionally, Figure 4i showed that B_18 exhibits 288 unique differential metabolites compared to the A_CK group, while C_38, D_58, and E_78 display 168, 453, and 307 unique differential metabolites, respectively, when compared to the A_CK group. The results indicated that the metabolic profile of white A. cornea undergoes significant changes due to the addition of bamboo substrate at different ratios, leading to selective accumulation of various bioactive components. Specifically, the volcano plot comparing B_18 with A_CK highlighted 104 upregulated metabolites, including Octadecylamine, Ile-gly, Farnesyl acetate, O,O-Diethyl hydrogen thiophosphate, and 2-(1-methyl-1H-imidazol-4-yl)acetic acid, alongside 489 downregulated metabolites such as Ile-Trp, [10]-Dehydrogingerdione, Muscarine, (S)-2-Propylpiperidine, and cis-10-Heptadecenoic acid. Comparison of C_38 with A_CK revealed 510 upregulated metabolites, including L-Glutaminyl-L-tryptophan, Monocrotaline, Adenosine-5′-triphosphate, Sapropterin, Crotonic acid, and 192 downregulated metabolites including D-Pipecolic acid, 1-Octadecyl Lysophosphatidic Acid, Vanylglycol, 2′-Deoxyguanosine 5′-monophosphate, and cis-3-Hexenyl tiglate. The comparison between D_58 and A_CK groups revealed 528 upregulated metabolites, including Leu-Ile, 4-Methylbenzyl alcohol, 1,2,4-Benzenetriol, Oxoamide, D-(+)-Neopterin, and 626 downregulated metabolites such as 4-Hydroxystyrene, octadecylamine, muscarine, 11-aminoundecanoic acid, and ganoderic acid A. The comparison between E_78 and A_CK revealed 221 upregulated metabolites, such as Lys-Asn, Oxoamide, Benzimidazole, Fraxinol, and 13(S)-Hydroperoxylinolenic acid, along with 630 downregulated metabolites, including 1-(9Z,12Z-octadecadienoyl)-sn-glycero-3-phosphocholine, Ile-Trp, Arachidonoyl Serinol, cis-4-Decen-1-ol, and Steviol.

Meanwhile, distinct substrate compositions shape the unique metabolic development process of white A. cornea. Therefore, a comparative metabolite analysis was conducted between four groups of samples with varying bamboo substrate replacement levels and the CK control group. Results revealed that in B_18 vs. A_CK, 84 unique metabolites were identified, including Adenosine-5′-triphosphate, Pro-Tyr, N-Acetyl-L-arginine, Lys-Val, and Guanosine-5′-monophosphate. In C_38 vs. A_CK, 118 unique metabolites were highlighted, including Aniline, N4-Acetylcytidine, 2H-Pyran-2-one, tetrahydro-6-(2-pentenyl)-, Homoeriodictyol (+/−)-, and Salsolinol. Similarly, 129 unique metabolites were highlighted in D_58 vs. A_CK, including 1,17-Diamino-4,9,13-triazaheptadecane, 2-Benzylsuccinic acid, Fusarin C, Trimethylsilyl (2E)-2-(methoxyimino)propanoate, and N-Methylpyridinium. Finally, 99 unique metabolites were identified in E_78 vs. A_CK, including S-(5-Adenosy)-L-Homocysteine, Uridine-5′-diphosphate, Ethyl maltol, Pinolidoxin, and 14,15-DiHETE (Table S2: Unique metabolite categories for different comparison groups). It could be observed that as the bamboo substrate addition rate increases, the levels of unique metabolites in each group (except E_78) show an upward trend. Additionally, to demonstrate the unique metabolite composition of the test material white A. cornea after adding different proportions of bamboo substrate, a score plot was created for the top 20 metabolites with the highest VIP values (Figure S4).

3.5. Identification and Analysis of Metabolic Biomarkers Based on Bamboo Substrate Treatment

The composition of substrates significantly influences the metabolic levels of white A. cornea. We selected the top 20 significantly up- or downregulated differential biomarkers for bar chart visualization to gain deeper insights into metabolic differences among groups (Figure 5, Table S3). Concurrently, the top five significantly differential biomarkers (VIP > 1) within each comparison group are displayed as violin plots on the right (Figure 6, Table S3). For A_CK vs. B_18, significantly upregulated DMs primarily included N,N-dimethylglycine, 1,3,8-Trihydroxy-4-methyl-2,7-diprenylxanthone, Heneicosanoic acid, Proline betaine, (Z)-2-Decene-4,6,8-triynoic acid, methyl ester. Significantly downregulated metabolites included O-Acetylcypholophine, L-Tyr-L-His, Tetraneurin A, and (6R,7R)-7-[[(5R)-5-azaniumyl-5-carboxylatopentanoyl] amino]-3-(carbamoyloxymethyl)-8-oxo-5-thia-1-azabicyclo[4.2.0]oct-2-ene-2-carboxylate, Tripdiolide, and Diethyl 2-(((2-hydroxyphenyl)amino)methylene)malonate (Figure 5a). In A_CK vs. C_38, significantly upregulated DMs included Ser-Gly-Leu-His-Thr, 3-(3,4-dimethoxyphenyl)-N-[2-(1-hydroxy-3-methoxy-4-oxocyclohexa-2,5-dien-1-yl)ethyl]propanamide, Eudesobovatol A, Pro-Val-Gly, and 1,2-Dioctanoyl-sn-glycero-3-phosphate, while two Ipecoside derivatives and Iprovalicarb were significantly downregulated (Figure 5b). In A_CK vs. D_58, significantly upregulated metabolites included 2-Amino-2-deoxy-D-gluconic acid, Fusarin C, Tumonoic acid E, Phe-Phe-Met, 3,5,7-Trihydroxy-2-(1-hydroxy-3-methoxy-4-oxocyclohexyl)-6-methoxychromen-4-one, S-trans-1-propenyl-L-cysteine. Significantly downregulated metabolites included 2-amino-4-(S-butylsulfonimidoyl)butanoate, 4-Heptanone, (x)-2-Heptanol glucoside, Oxfenicine, nonic acid (Figure 5c). Finally, in A_CK vs. E_78, significantly upregulated metabolites included allyl methyl sulfone, Thioperamide, 4,4′-oxybis-Benzenesulfonamide, 2-Amino-2-deoxy-D-gluconic acid, Ethyl maltol, and 2,2′-Dithiodibenzoic acid, while three classes of metabolites—O-Acetylcypholophine, Tetraneurin A, and Tyr-His—were significantly downregulated. These results indicated that varied bamboo substrate additions significantly influence the metabolic pathways of white A. cornea, thereby identifying distinct biomarkers. Significantly enriched biomarker categories primarily included amino acids and their derivatives, benzene and its derivatives, organic acids, ketones, and amines. These biomarkers were consistently detected across multiple comparison groups and were also discernible in the DMs scatter plot (Figure S5), highlighting their central role in fungal growth and metabolism in response to different substrate substrates.

Additionally, we investigated the relative changes in metabolite content under different bamboo substrate addition levels using K-means clustering analysis (Figure S6). Results indicated that 662 metabolites gradually increased with bamboo substrate addition but abruptly decreased at 78% substrate loading, including Ile-Trp, 4-Methylbenzyl alcohol, 1,2,4-Benzenetriol, 2-Decen-1-ol, and 12-Methyltridecanoic acid (Figure S6A). Conversely, 472 metabolites exhibited a gradual decrease in relative abundance with increasing bamboo substrate addition, including [10]-Dehydrogingerdione, Muscarine, Podofilox, Noroxycodone, and 1-Octadecyl Lysophosphatidic Acid (Figure S6B). Conversely, 249 metabolites exhibited increasing relative abundance with rising bamboo substrate levels, including avocadyne, oxoamide, Ile-gly, betaine aldehyde, and fraxinol (Figure S6D). These findings established a robust foundation for subsequent selection of potential cross-group biomarker metabolites.

3.6. KEGG Functional Annotation and Enrichment Analysis of Differentially Expressed Metabolites

Based on the differential metabolite results, KEGG pathway enrichment analysis was performed, and the annotation results were categorized for display (Figure S7). The bubble chart illustrated the top 20 KEGG metabolic pathways enriched in SDMs compared to the CK group (Figure 7). This enables an in-depth investigation into the functional roles of the differentially metabolized compounds under varying bamboo substrate additions relative to conventional wood chip substrates, providing theoretical insights into the growth and metabolic responses of white A. cornea to bamboo substrate utilization. Results indicated that in A_CK vs. B_18, 80 metabolic pathways were annotated, including valine, leucine and isoleucine degradation, valine, leucine and isoleucine biosynthesis, caprolactam degradation, microbial metabolism in diverse environments, and arachidonic acid metabolism under the metabolism category; phosphotransferase system (PTS), phosphatidylinositol signaling system, and ABC transporters in the category of environmental information processing; and efferocytosis in the cellular processes category (Figure S7A). KEGG pathway enrichment analysis indicated that metabolites were primarily enriched in diterpenoid biosynthesis, sphingolipid metabolism, arachidonic acid metabolism, and cutin, suberine and wax biosynthesis. These metabolic processes synergistically support the core survival strategy of fungi: breaching host defenses and occupying ecological niches through secondary metabolites (diterpenes, AA derivatives); maintaining cellular structural stability and stress resistance via sphingolipids and waxes; ultimately achieving adaptation to environments (e.g., lignocellulosic substrates), host invasion or symbiosis, and species reproduction and persistence. For lignocellulose-degrading fungi, these metabolic processes also interact with cellulose and hemicellulose degradation pathways (e.g., sphingolipid metabolism influences cellulase secretion, while diterpenoids inhibit competing microorganisms), collectively ensuring biomass degradation efficiency (Figure 7a). In A_CK vs. C_38, 103 metabolic pathways were annotated, including metabolic pathways, biosynthesis of secondary metabolites, and microbial metabolism in diverse environments under metabolic processes; vancomycin resistance, pertussis, and pathogenic Escherichia coli infection under human disease; aminoacyl-tRNA biosynthesis under genetic information processing; ABC transporters under environmental information processing; and efferocytosis under cellular processes (Figure S7B). KEGG pathway enrichment analysis showed more significant enrichment in purine metabolism, nucleotide metabolism, and biosynthesis of nucleotide sugars. Purine metabolism provides purine precursors for nucleotide metabolism, while nucleotide metabolism supplies nucleotide skeletons for nucleotide sugar synthesis. Together, these three pathways form the “energy–genetic information–structural function” metabolic network (Figure 7b). In A_CK vs. D_58, 118 metabolic pathways were annotated, encompassing organism systems, metabolism, human diseases, genetic information processing, environmental information processing, and cellular processes (Figure S7C). Enrichment analysis revealed significantly enriched pathways including nucleotide metabolism, purine metabolism, and ABC transporters, which are crucial for nutrient uptake, energy replenishment, and DNA replication (Figure 7c). Finally, in A_CK vs. E_78, a total of 107 metabolic pathways were annotated. The pathways annotated with the highest number of differentially expressed metabolites were primarily metabolic pathways (95 pathways, accounting for 74.22%), biosynthesis of secondary metabolites (52 pathways, 40.62%), and microbial metabolism in diverse environments (28 pathways, 21.88%). Similar annotation patterns were observed in the other three groups, highlighting their importance in the growth and metabolism of white A. cornea (Figure S7D). In enrichment analysis, significantly enriched pathways included biosynthesis of cofactors, biosynthesis of secondary metabolites, and metabolic pathways (Figure 7d).

3.7. Comprehensive Analysis of Differential Metabolite Changes in Metabolic Pathways

The overall changes in differentially expressed metabolites across different metabolic pathways also varied significantly due to differing substrate levels. To visualize the collective impact of these metabolites on key pathways (KPW), a DA score plot was generated, displaying the top 20 pathways ranked by p-value in ascending order (Figure S8). Results indicated that at 18% bamboo substrate substitution (B_18 vs. A_CK), metabolites across all 20 significantly enriched metabolic pathways showed a downregulation trend, particularly in processes such as diterpenoid biosynthesis, sphingolipid metabolism, and arachidonic acid metabolism. However, at 38% bamboo substrate addition (C_38 vs. A_CK), metabolites in the top 20 significantly enriched pathways showed an upregulation trend, including efferocytosis, alpha-Linolenic acid metabolism, and naphthalene degradation. At a bamboo substrate addition of 58% (C_58 vs. A_CK), biological processes like nucleotide metabolism, Chagas disease, and amoebiasis showed an upregulation trend, while metabolic processes such as quorum sensing, phosphatidylinositol signaling system, and methane metabolism exhibited a downregulation trend. Similarly, at a bamboo substrate addition rate of 78% (C_78 vs. A_CK), enriched pathways showed an overall downregulation trend, such as pentose and glucuronate interconversions, carbon metabolism, and pentose phosphate pathway, while the two upregulated pathways were autophagy—yeast and autophagy—other.

4. Discussion

This study aims to systematically elucidate the metabolic response mechanisms of white A. cornea when bamboo substrate replaces oak sawdust substrate at varying substitution levels. Grounded in the dual realities of the acute “fungus-forest conflict” within the edible mushroom industry and the underutilization of bamboo substrate waste resources, this research seeks to provide theoretical support for sustainable “bamboo-substituting-wood” cultivation practices. By establishing five bamboo substrate replacement ratios (0%, 18%, 38%, 58%, and 78%), we pioneered the integration of non-targeted LC-MS/MS metabolomics with conventional agronomic trait measurements in white A. cornea. This approach seeks to address the core scientific question: “How does bamboo substrate reshape the primary and secondary metabolic networks of white A. cornea, thereby determining yield and quality?”. Experimental results indicated that white A. cornea could grow normally in substrates with varying bamboo substrate replacement ratios, with the primary findings exhibiting a “nonlinear dose-metabolism effect”. Significant differences emerged at the agronomic trait level: The 58% bamboo substrate replacement group (D_58) demonstrated optimal performance, with a substrate C/N similar to the A_CK control group. It achieved an average mycelial growth rate of 3.55 ± 0.24 mm/d and the shortest growth period (86.2 d). Considering comprehensive growth efficiency, fruiting body traits, and cost advantages, this group represents the optimal replacement formulation. The 38% bamboo substrate replacement group (C_38) exhibited the highest rehydration ratio (7.99 ± 1.93), making it the preferred option for quality-oriented production. The 78% bamboo substrate replacement group (E_78) demonstrated outstanding ear-shaped morphology metrics and acceptable yield, but exhibited a longer growth period. When the replacement rate increased to 78%, mycelial growth momentum weakened, leading to declining yield and rehydration ratio, and this indicates that the “high bamboo substrate content” formulation encountered a degree of C/N imbalance and metabolic bottlenecks [39]. Metabolomics analysis detected 3779 metabolites, primarily encompassing amino acids and their derivatives (42.2%) and organic acids (35.54%). Compared to the control group, each treatment group exhibited 104–528 differentially upregulated metabolites and 192–630 downregulated metabolites. There were 93 shared DMs and numerous unique metabolites. Different bamboo substrate additions induced distinct dose-dependent metabolic responses, with metabolite levels exhibiting three dynamic patterns along the bamboo substrate gradient: ① continuously increasing (249 metabolites, e.g., Oxoamide, Fraxinol); ② initial increase followed by decrease (662 compounds, e.g., 1,2,4-Benzenetriol, Ile-Trp); and ③ sustained decrease (472 compounds, e.g., Muscarine, 10-Dehydrogingerdione). The number of DMs exhibited a “low-high-low” parabolic trend with bamboo substrate proportion: D_58 vs. A_CK showed 528 upregulated and 626 downregulated compounds—the highest among all comparison groups—perfectly overlapping with the agronomic trait “optimum point,” suggesting metabolic abundance can serve as a predictive biomarker. Notably, the response of white A. cornea to varying bamboo substrate replacement ratios in this study did not follow a simple linear promotion or inhibition pattern, but instead exhibited distinct “optimal range” characteristics. This contrasts with some substrate replacement studies that primarily focus on single replacement ratios or binary comparisons. Previous studies on replacing sawdust with corn cobs or other agricultural wastes for edible mushroom cultivation have mostly focused on “whether substitution is feasible” or “whether high substitution inhibits yield”, with limited systematic discussion on the continuous remodeling of metabolic networks across different substitution gradients [12,23,40,41]. By establishing five replacement gradients, this study first revealed in white A. cornea that substrate replacement exhibits a “functional optimum” jointly determined by metabolic networks. This optimum highly overlaps with agronomic traits and the number of DMs, indicating that the metabolic system’s response to substrate changes possesses intrinsic self-organizing regulatory characteristics rather than passive adaptation. These signature metabolites served as molecular knobs for precise regulation of subsequent “bamboo-wood” formulations. They further confirmed that substrate type and proportion changes could directly influence the morphology, growth period, and yield of white A. cornea by reshaping metabolic processes [11,31].

Compared to previous studies, this research achieved significant breakthroughs in analyzing the substrate adaptability and metabolic mechanisms of white A. cornea on bamboo substrate, realizing a triple leap in “substrate-metabolism” resolution. First, while early reports on bamboo substrate cultivation mostly focused on feasibility descriptions [27,29], we used LC-MS/MS to pinpoint “yield-quality” differences to specific compounds. For instance, the 58% bamboo substrate group showed significant upregulation of 2-Amino-2-deoxy-D-gluconic acid (2-ADGA). This compound plays a pivotal role in multiple biological metabolic pathways, serving as a key node linking primary carbon metabolism to glycoside synthesis [42,43], providing molecular evidence for explaining the “fungus rapid-growth” phenomenon on bamboo substrates. Previous explanations of the “rapid mycelial growth” phenomenon have primarily focused on the physicochemical properties of the substrate (such as porosity, moisture content, or C/N ratio), with limited interpretation at specific metabolic nodes [44]. Through metabolomics, this study further pinpoints this macro-level phenomenon to the significant upregulation of key metabolites like 2-ADGA. This suggests that bamboo substrates might not only promote mycelial growth by improving the physical environment but also accelerate the assimilation efficiency of substrate carbon sources by reshaping glycosamine metabolism and carbon flux allocation. This advancement from “physicochemical explanations” to “metabolic node interpretations” provides a higher-resolution mechanistic perspective for understanding fungal growth differences under varying substrate conditions. Second, studies on substrates like corn cobs and goji berry wood chips indicated that free amino acid metabolism is central to the substrate effect [11,13]. We further refined this to “which amino acids change and how”: the 38% bamboo substrate group showed significant enrichment of Sapropterin and the oligopeptide (Ser-Gly-Leu-His-Thr), indicating activation of the purine-nucleoside-glycosyl cycle, which provides energy and carbon skeletons for rapid mycelial growth [45]. However, when the bamboo substrate content increased to 78%, this pathway was overall downregulated, accompanied by negative DA scores for the pentose phosphate pathway and carbon metabolism, and this explained the biochemical root cause of the “lack of staying power” observed at high substitution levels [45]. Third, previous studies on de-oiled camphor leaf-cultured A. cornea reported increased diterpenes and phenolic acid stress metabolites [10]. We observed similar phenomena in the 18% bamboo substrate group, but the D_58 group exhibited a sudden downregulation of diterpene biosynthesis. This indicates that moderate bamboo substrate content is sufficient to induce stress resistance, while excessive amounts might shut down secondary metabolism due to imbalanced C/N [46]. Functional analysis of key metabolites revealed amino acids and their derivatives as the most abundant differential category. Branched-chain amino acids (Leu-Ile, Ile-Gly) were significantly upregulated in the D_58 group. Serving not only as raw materials for protein synthesis, they also function as energy donors during carbon source imbalance, explaining the rapid mycelial growth observed in this group [47]. Peptides like L-glutaminyl-L-tryptophan enriched in the C_38 group might participate in fruiting body development signaling, consistent with its highest rehydration ratio [48]. Organic acids like Tumonoic acid E and Crotonic acid were upregulated in dominant groups. These acids enhance energy supply efficiency by participating in the tricarboxylic acid cycle while regulating substrate pH to promote cellulase activity, adapting to the high cellulose content characteristic of bamboo substrates [49,50]. Secondary metabolites like flavonoids and phenolics were upregulated in the high bamboo substrate group, potentially enhancing the nutritional value and storage stability of white A. cornea. Conversely, metabolites such as Muscarine and [10]-Dehydrogingerdione were significantly reduced across all bamboo substrate groups, potentially reflecting metabolic regulation triggered by changes in substrate nutrient composition [41,51]. Compared to previous studies that evaluated substrate effects using amino acid metabolism as a holistic indicator [13,52], this research further distinguishes the functional differentiation of various amino acids and peptides across different bamboo substrate gradients: branched-chain amino acids are more involved in growth rate and energy compensation, while specific peptides might be associated with fruiting body development and quality formation. This refined analysis of “same metabolic category, different functional orientations” helps explain why “high yield” and “high quality” phenotypes separate under different substitution ratios. It also provides a basis for subsequent formulation design targeting distinct production goals (yield-oriented or quality-oriented).

Different bamboo substrate addition levels shaped the growth adaptation strategies of white A. cornea by regulating core metabolic pathways. For instance, the 18% bamboo substrate replacement group (B_18) enriched pathways such as diterpene biosynthesis and sphingolipid metabolism. Sphingolipids maintain cell membrane stability, and diterpenes exhibit antimicrobial activity. At low bamboo substrate ratios, fungi may utilize these pathways to counteract substrate nutrient competition [53,54,55]. Furthermore, metabolites from all 20 significantly enriched metabolic pathways in this group showed a downregulated trend, reflecting passive metabolic adaptation under low substitution levels. The 38% bamboo substrate replacement group (C_38) focused on purine metabolism and nucleotide sugar biosynthesis, where purine metabolism provides precursors for nucleotide metabolism, and nucleotide metabolism supplies the backbone for nucleoside glycosylation. Together, these form a metabolic network linking “energy-genetic information-structural function” that underpins fruiting body quality formation [56]. The top 20 significantly enriched metabolic pathways in this group all showed an upregulation trend, indicating active adaptation of the metabolic system to this substrate ratio. The 58% bamboo substrate replacement group (D_58) primarily featured nucleotide metabolism and ABC transporter pathways. ABC transporters facilitate transmembrane transport of lignocellulose degradation products, enhancing mycelial nutrient utilization efficiency [57]. In this group, some pathways were upregulated while others were downregulated, forming a dynamically balanced metabolic regulation pattern consistent with its comprehensive advantages in experimental traits. Finally, the 78% bamboo substrate replacement group (E_78) enriched cofactor synthesis and secondary metabolite biosynthesis pathways. Cofactors participate in enzymatic reactions of degradative enzymes, ensuring cellulase and hemicellulase activity under high bamboo substrate ratios. Overall, enriched pathways showed a downregulated trend, with only autophagy-related pathways upregulated, reflecting stress adaptation of the metabolic system under high replacement levels [58]. This differential regulation of pathways collectively constitutes the gradient adaptation mechanism of white A. cornea to bamboo substrate, demonstrating the flexibility and specificity of fungal metabolic networks.

The potential significance of this study extends across industrial, technological, and scientific dimensions: Industrially, it validated the feasibility of bamboo substrate as a cultivation medium for white A. cornea. From an economic analysis perspective, the cost of bamboo substrate is approximately 400 yuan per ton. With 58% bamboo substrate substitution, the average dry yield per 500 g of wet substrate is 15.88 g of dried mushrooms, in contrast, oak sawdust (A_CK group) costs approximately 1979 yuan per ton, yielding 21.86 g of dry mushrooms per 500 g of substrate. When achieving equivalent mushroom yields, a rough estimate indicates that replacing one ton of sawdust with bamboo substrate in the 58% substitution formula saves at least 855 yuan in substrate costs alone. Therefore, by combining cost advantages, the D_58 group demonstrates a concrete solution for “replacing wood with bamboo”, alleviating pressure on forest resources while utilizing millions of tons of bamboo substrate waste, aligning with circular agriculture development needs. Technologically, it identifies metabolic regulation targets corresponding to different bamboo substrate addition levels, providing molecular evidence for precision optimization of cultivation formulations. Scientifically, it constructs a “substrate–metabolite–phenotype” correlation network, supplementing the substrate adaptation theory of wood-decaying fungi and providing a paradigm for other edible mushroom research. Overall, this study does not merely provide a novel substrate formulation, and through systematic gradient design and metabolomics analysis, it reveals the metabolic regulatory logic by which white A. cornea transitions from passive adaptation to active optimization under “bamboo-substituting-wood” conditions. Compared to previous research models primarily relying on empirical screening, this study emphasizes understanding substrate effects at the metabolic network level, offering a reference analytical framework for subsequent cross-species and cross-substrate substitute research. However, the study still has limitations in several aspects. First, the metabolome represents a snapshot endpoint analysis, failing to resolve dynamic changes across mycelial different developmental stages. Future work should integrate temporal metabolomics with transcriptomics to reveal the full regulatory network spanning “early signaling-mid growth-late fruiting”. Second, bamboo substrate contains unique bamboo flavonoids and silicified cell walls. This study focused solely on intracellular metabolites, neglecting extracellular secreted proteins and polysaccharides. Future work should employ multi-omics approaches (proteomics + metabolomics + ionomics) to elucidate the coupling mechanism between extracellular lignocellulolytic enzymes and intracellular metabolic flux. Third, while DMs correlate with agronomic traits, reverse validation is lacking. Future studies could validate causal roles by exogenously supplementing key metabolites (e.g., 2-ADGA, oxoamide).

5. Conclusions

This study systematically elucidated the metabolic adaptation mechanisms of white A. cornea on bamboo substrates by establishing cultivation systems with varying proportions of bamboo substrate replacing oak sawdust, combined with LC-MS/MS non-targeted metabolomics and agronomic trait measurements. Results indicated that white A. cornea grew normally on all bamboo substrate replacement formulations, with the 58% bamboo substrate replacement group demonstrating optimal overall performance, balancing cost and growth efficiency. Metabolomics detected 3779 metabolites, with differentially expressed metabolites primarily enriched in amino acids and their derivatives, organic acids, and other categories, exhibiting dose-dependent metabolic responses to varying bamboo substrate proportions. This study validated the feasibility of bamboo substrates for the cultivation of white A. cornea, and identified optimal substitution formulations and metabolic regulation targets. This study provides theoretical reference for “substituting wood with bamboo” to alleviate “forest-fungus conflicts” and achieve resource utilization of bamboo substrates.