Trichoderma genome and multiomics insight into promoting yield and reducing grain cadmium in barley and wheat

ABSTRACT

Soil cadmium (Cd) contamination poses significant risks to human health and environmental sustainability. Despite advances in bioremediation, effective bioagents with clear mechanistic insights for Cd detoxification are lacking. We first deciphered the whole‐genome sequence of a novel Cd‐tolerant Trichoderma nigricans T32781 and its in vivo heavy metal tolerance. In five independent pot and field trials, we revealed the T32781‐induced alleviation mechanisms of plant–microbe–soil interactions in wheat and barley in response to Cd toxicity using a combination of agronomic, physiological, microbiome and metabolome approaches. We discovered that T32781 inoculation in soil significantly increased grain yield and decreased grain Cd concentration in barley and wheat exposed to different soil Cd levels. T32781 predominantly colonized soils, mitigating Cd toxicity by reducing soil Cd availability and promoting beneficial soil microbial communities and metabolites. These beneficial effects were further validated in the field, where the exogenous application of key metabolites induced by T32781 inoculation in soils and plants significantly increased grain yield and reduced grain Cd concentration in barley. This work highlights the potential of T32781 to enhance plant‒microbe–soil interactions and support sustainable and safe crop production in Cd‐contaminated soils, addressing the increasing global demand for cereal production for food and feed.

The fungal strain Trichoderma nigricans T32781 improved yield and reduced the grain cadmium concentrations of wheat and barley in cadmium‐polluted soils through microbial and metabolic interactions, offering a potential strategy for enhancing crop productivity and food safety in cadmium‐contaminated fields.

INTRODUCTION

Food security and food safety are critical challenges in the context of global population growth and environmental pollution (King et al., 2017; Dillard, 2019). Cadmium (Cd) is one of the most hazardous and widespread heavy metal contaminants and is produced mainly from anthropogenic activities (including smelting, mining and application of fertilizers), threatening agricultural productivity and food safety (Bah et al., 2011; Shi et al., 2019). Cadmium contamination jeopardizes food safety, as crops grown in such soils usually accumulate Cd in edible parts, which can readily transfer to humans via the food chain, causing serious health risks (e.g., renal and cardiovascular dysfunctions and Itai‐Itai disease), even at very low concentrations (Tellez‐Plaza et al., 2007; Chaney, 2015; Du et al., 2020). Given that crops are primary sources of dietary Cd intake for humans, reducing Cd uptake from soil and its translocation to the edible parts of crops is essential for safeguarding human health (Clemens and Ma, 2016; Fu et al., 2019).

Various physicochemical soil remediation strategies, such as excavation and in situ bioremediation, have been extensively explored to reduce Cd bioavailability in soil (Shen et al., 2022; Mubeen et al., 2023). Additionally, exogenous applications of nanoparticles, signaling molecules, and biochar have shown potential in alleviating heavy metal stress in crops. However, these techniques are often limited by high cost, low efficiency, and the risk of secondary pollution (Zhou et al., 2020; Jiang et al., 2022; Liu et al., 2023b; Hou et al., 2024). Alternatively, biological remediation, particularly the use of plant growth‐promoting microbes (PGPM) to mitigate the levels of toxic metals and metalloids, including Cd, has emerged as one of the more promising strategies because of its eco‐friendly characteristics, cost‐effectiveness, and broad public acceptance (Deng et al., 2021; Alves et al., 2022). Nevertheless, the effectiveness of PGPM‐mediated Cd remediation is influenced by the complex interactions among microbial characteristics, environmental conditions, and specific plant traits, emphasizing the critical need for a deeper understanding of plant–microbe–soil interactions (Alori et al., 2022). Therefore, it is imperative to identify and utilize novel PGPM that can efficiently minimize plant Cd absorption with broad adaptability across varying environmental conditions.

Wheat (Triticum aestivum L.) is one of the most important crops and a staple food for the global population (Senapati et al., 2022); barley (Hordeum vulgare L.), the fourth most widely produced cereal crop, plays a vital role in livestock feed, food, and beverages (Lei et al., 2020). However, neither crop is able to thrive in Cd‐contaminated soils (Rizwan et al., 2016; Wang et al., 2023a), posing serious issues for food safety and security in many countries. Barley, with its smaller diploid genome compared to the hexaploid genome of wheat, serves as an ideal model cereal crop for elucidating molecular and physiological pathways involved in Cd stress responses (Mascher et al., 2017). For instance, our previous study revealed that 47.2% of 600 barley genotypes exceed the threshold of grain Cd concentration even when they are grown in non‐Cd‐polluted fields in China (Chen et al., 2007a). Given these findings, it is critical to explore the mechanisms underlying Cd stress tolerance and to develop effective strategies, such as the application of PGPM, to minimize Cd transfer from soil to grains of key cereal crops such as wheat and barley.

Trichoderma nigricans T32781 (T32781 thereafter) effectively alleviated Cd toxicity in tobacco (Nicotiana tabacum L.), notably without requiring root colonization, making it a promising candidate for broader agricultural applications (Zhang et al., 2023). In this study, we explored the potential of the novel fungal strain T. nigricans T32781, identified through PGPM screening, to mitigate Cd stress in cereal crops. We hypothesized that inoculation with T32781 would reduce Cd availability and alter the soil microbiome and metabolome of plants and soil under various environmental conditions, thereby increasing yield and reducing Cd accumulation in wheat and barley grains. Therefore, we deciphered the whole‐genome sequence of T32781 and examined its impacts on the yield and quality of wheat and barley grown in Cd‐polluted soils with varying levels of Cd contamination (Figure 1). We further conducted a comprehensive set of experiments using agronomic, physiological, microbiome, and metabolomic techniques to understand the plant–microbe–soil interactions that contribute to improving barley Cd tolerance. We revealed a novel cross‐talk‐based solution through plant–soil–microbiome interactions to increase crop yield and food sustainability in response to the pervasive global challenge of soil heavy metal contamination.

Figure 1.

Overview of the experimental flow charts for this study

(A) Tolerance assays for heavy metals (HMs) and scanning electron microscopy with energy‐dispersive X‐ray spectroscopy (SEM‒EDS) analysis were performed to analyze the growth of Trichoderma nigricans T32781 (T) under Cd exposure across different growth media. (B) In Experiments 1–5, T32781 inoculum was prepared after fermentation and used in pot and field trials to evaluate its effects on wheat and barley growth and Cd concentrations. Microbiome and metabolome analyses were conducted to explore plant–microbe–soil interactions. In Experiment 6, key metabolites that responded to T32781 inoculation, including 4‐aminobutyric acid (GABA), betaine, L‐citrulline, and N‐acetyl‐methionine (NALM), were added at two different concentrations to evaluate their ability to mitigate the impact of Cd exposure on barley growth in a field trial.

RESULTS

T32781 genome assembly and analysis

Assembly quality was assessed using BUSCO (Benchmarking Universal Single‐Copy Orthologs), which resulted in 25 contigs, with a maximum contig length of 4,976,569 bp. The N50 contig length was 3,799,607 bp, and the GC content was 48.7%, with 98.8% complete genome assembly quality. A detailed circular genome diagram is shown in Figure 2A. A total of 10,280 genes were predicted, with the messenger RNAs (mRNAs) being the most populous type of RNA, at 10,046, followed by 234 transfer RNAs (tRNAs) and 44 ribosomal RNAs (rRNAs), with an average length of 1,644 bp. With respect to transposable elements, 42 short interspersed nuclear elements (SINEs), 270 long interspersed nuclear elements (LINEs), 424 long terminal repeat (LTR) elements, and 217 DNA elements accounted for 0.01%, 0.05%, 0.15%, and 0.04% of the total genome, respectively (Figure 2B). There were 5,816 genes annotated in the Clusters of Orthologous Genes (COG) database, which were enriched mainly in carbohydrate transport and metabolism (595 genes), intracellular trafficking, secretion, and vesicular transport (465 genes), and amino acid transport and metabolism (462 genes) (Figure 2C; Table S1). A Maximum Likelihood (ML) phylogenetic tree revealed that T32781 belongs to the Atroviride clade (Figure S1). To explore the potential mechanisms of Cd tolerance in T32781, we surveyed candidate genes based on functional annotation (Table S2). The data suggested that the tolerance was mediated by regulation of glutathione (GSH) metabolism (GSH1 and GSH2) for reactive oxygen species (ROS) scavenging, together with vacuolar sequestration of Cd²⁺ via both YCF1‐mediated Cd‐GSH complex transport and NHX1/ZRC1‐mediated free Cd²⁺ transport pathway (Figure 2D).

Figure 2.

Circular genome diagram, mechanisms of Cd tolerance, Cd adsorption and scanning electron microscopy with SEM‒EDS analysis of T32781

(A) Circular genome diagram of T32781. A snapshot view is shown from the outermost to the innermost rings of the locations of predicted messenger RNAs (mRNAs), DNA coding sequences (CDSs), and transfer RNAs (tRNAs) within the genome. GC: guanine + cytosine content of the genome. GC skew where green and red indicate above‐ and below‐average GC skew, respectively. (B) Genome assembly overview and BUSCO (Benchmarking Universal Single‐Copy Orthologs) assessment. (C) Clusters of Orthologous Genes (COG) functional annotation. (D) Potential mechanisms of Cd tolerance in T32781. (E–G) SEM images of T32781 after 7 d of inoculation in 1/10 Luria–Bertani (LB) liquid medium: (E) no CdCl₂, (F) 0.5 mmol/L CdCl₂ with a phosphate‐buffered saline (PBS) wash, and (G) 0.5 mmol/L CdCl₂ without a PBS wash. (H, I) Enlarged spectrum of the red box followed by EDS analysis. (J) Intra‐ and extracellular Cd concentrations of T32781 after 4 and 7 d in 1/10 LB liquid medium with 0.5 mmol/L CdCl₂. Independent‐samples t‐tests were used to compare the differences in Cd concentration between 4‐ and 7‐d inoculations. The data are the means ± SEs (n = 3). **indicates significance at P < 0.01.

Antibiotic resistance profiling revealed a single vancomycin resistance gene (42.11% identity to glycopeptide resistance homologs) that operates through target alteration (Table S3). Hemolysin screening yielded 56 UniProt entries and eight Pfam accessions, with four candidate ABC transporter‐containing proteins lacking toxic domains (Tables S4–S6). A putative RTX (repeats‐in‐toxin) toxin translocation protein (32% identity to known cytolysins) was identified but similarly lacked characteristic toxic motifs (Tables S7, S8), collectively suggesting a non‐toxic role in T32781.

Heavy metal‐tolerant T32781 increases wheat and barley yields in Cd‐contaminated soil

T32781 survived exposure to most heavy metals at 0.1 mmol/L in both Luria–Bertani (LB) and potato dextrose (PD) liquid media, where fungal cells settled at the bottom. On solid medium, T32781 exhibited enhanced growth under Cd exposure compared with other heavy metals, tolerating concentrations up to 5 mmol/L Cd in LB medium (Figure S2; Tables S9, S10). Scanning electron microscopy (SEM) revealed that compared with the untreated control, Cd toxicity caused significant structural damage to T32781, leading to fractured and shrunken fungal mycelia in liquid medium (Figure 2E–G). Energy‐dispersive X‐ray spectroscopy (EDS) analysis revealed that the precipitate that formed on the cell surface consisted of Cd (Figure 2H, I), which was easily removed by washing with phosphate‐buffered saline (PBS) (Figure 2F). The results of Cd concentration assessments in T32781 revealed that the intracellular Cd concentration significantly decreased from 9.8 to 3.1 mg/kg over time, whereas Cd concentration on the cell surfaces significantly increased from 0.5 to 0.9 mg/kg between the 4‐ and 7‐d inoculation periods (Figure 2J). Therefore, the novel Cd tolerance property of T32781 promoted further investigation of the mechanisms underlying its role in improving food safety of wheat and barley grown in Cd‐contaminated soils.

We first evaluated growth and grain yield of barley cultivated in low‐Cd‐polluted soil with T32781 inoculum in pots (supplied with 5 mg/kg Cd) and a polluted paddy field (total Cd concentration: 1.79 mg/kg; available Cd concentration: 0.84 mg/kg) in Experiments 1 and 2 during 2021–2022 (Figure S3; Table S11). A pot trial revealed that Cd stress reduced barley growth and grain yield, significantly decreasing the 1,000‐grain weight and increasing Cd accumulation in grains. Notably, compared with Cd, the T32781 inoculum in soil significantly alleviated Cd toxicity in barley by increasing grain yield by 20.9%. Moreover, T32781 significantly decreased Cd concentration in grain and the soil‐available Cd by 37.6% and 15.7%, respectively. In the field trial, T32781 significantly increased panicles per plant by 19.6% but dramatically decreased Cd concentration in grain‐ and soil‐available Cd by 21.7% and 8.2%, respectively.

Given the great potential of T32781 to alleviate Cd stress in barley, we extended its application to wheat and further evaluated its effects on barley under severe Cd‐contaminated soil with T32781 inoculum in pots (supplied with 10 mg/kg Cd) and in the field (total Cd concentration: 2.07 mg/kg; available Cd concentration: 1.00 mg/kg) during 2022–2023 (Figures 3, S4; Tables S12–S17). At the heading stage, compared with plants under Cd stress alone, those under T32781 inoculation presented a 10.1% increase in plant height, a 16.7% reduction in shoot Cd concentration (from 24.42 to 20.34 mg/kg), a 27.2% decrease in TF (translocation factor calculated as the shoot/root Cd concentration, from 0.11 to 0.08), and a 16.6% reduction in soil‐available Cd (from 6.45 to 5.38 mg/kg). With respect to the barley plants in the pot trials in Experiment 4, T32781 inoculation alleviated Cd toxicity by increasing shoot fresh weight and dry weight by 122.4% and 112.2%, respectively. In contrast to the Cd treatment, T32781 significantly reduced the Cd concentrations in shoots, roots, and soil by 18.1% (from 19.17 to 15.70 mg/kg), 13.3% (from 218.40 to 189.44 mg/kg) and 32.5% (from 7.99 to 5.39 mg/kg), respectively.

Figure 3.

Agronomic traits and Cd concentrations in wheat and barley in response to T32781 inoculation under Cd exposure in Experiments 3, 4, and 5

(A, B) Agronomic traits and Cd concentrations in wheat at the heading and ripening stages in Experiment 3. (C–F) Agronomic traits and Cd concentrations in barley at the heading and ripening stages in the pot trial of Experiment 4 (C, D) and in field trials of Experiment 5 (E, F). The translocation factor (TF) was calculated as shoot/root Cd concentration. Independent‐samples t‐tests were used to compare the differences between treatments with and without T32781 inoculation in Experiment 3 (W‐T vs. W‐Cd), Experiment 4 (B‐T vs. B‐Cd) and Experiment 5 (T vs. Mock). Response to T (%) = (Trait _+T−Trait _−T)/Trait _−T × 100%. The data are presented as the means ± SEs (n = 5 for the pot trial; n = 30 for the field trial). *P < 0.05 and **P < 0.01.

Similar alleviating effects were observed in barley plants in the field trial with T32781 inoculation in Experiment 5, where T32781 inoculation improved plant growth and decreased Cd concentrations in plants and soils (Figure 3A, C, E). At the ripening stage, T32781 significantly alleviated Cd stress in wheat, increasing grain weight per panicle by 14.1% and reducing Cd availability and TF by 41.3% (from 9.49 to 5.57 mg/kg) and 18.2% (from 0.33 to 0.27 mg/kg), respectively (Figure S5). This resulted in a 15.4% reduction in shoot Cd concentration (from 24.51 to 20.74 mg/kg) and a 16.0% decrease in grain Cd concentration (from 7.81 to 6.56 mg/kg) (Figure 3B). In barley, T32781 increased grain weight per plant and effective spikes by 74.1% and 85.0%, respectively in the pot trial, and increased panicles and grain yield per square meter by 33.0% and 35.9%, respectively, in the field trial. Additionally, T32781 significantly decreased shoot and grain Cd concentrations and Cd availability by 24.2% (from 20.61 to 15.62 mg/kg), 21.3% (from 5.25 to 4.13 mg/kg) and 45.1% (from 5.21 to 2.86 mg/kg) in the pot trial and 16.7% (from 3.95 to 3.29 mg/kg), 10.8% (from 0.65 to 0.58 mg/kg) and 8.6% (from 1.40 to 1.28 mg/kg) in the field trial (Figure 3D, F). Taken together, the results of the pot and field trials strongly demonstrated that T32781 inoculation effectively alleviated Cd stress to increase grain yield with low Cd contamination of wheat and barley under both controlled and open field conditions.

Enrichment of T32781 in Cd‐contaminated soil

To investigate the impact of T32781 inoculation on the microbial community structure in the rhizosphere soil of wheat and barley, we employed 16S rRNA and Internal Transcribed Spacer (ITS) sequencing during the heading stage in pot and field trials (Experiments 3–5; Tables S18–S30). T32781 inoculation significantly altered beta diversity of fungal communities in both wheat and barley in the pot trial but had no effect on alpha diversity. However, in the field trial, T32781 significantly affected both alpha and beta diversity levels in barley (Figures S6–S8). In terms of fungal composition, T32781 inoculation resulted in a marked dominance of the genus Trichoderma in rhizosphere soil, constituting 95.0% of fungal composition in wheat, 93.6% in barley in the pot trial, and 42.3% in barley in the field trial (Figure 4A–C). Furthermore, we analyzed the microbial compositions among the different treatments (Figures S9–S11). Our results revealed that seven fungal genera were significantly altered in wheat, including three in which abundances decreased and four in which abundances increased (Rhizophlyctidaceae_gen_Incertae_sedis, Trichoderma, Meyerozyma, and Neurospora); increases in genera were also found under Cd stress alone (Figures 4D, S9E). In barley, five fungal genera significantly decreased and seven fungal genera significantly increased (such as Meyerozyma and Trichoderma) in the pot trial, and four fungal genera significantly decreased and 16 fungal genera increased in the field trial (such as Trichoderma) (Figure 4E, F). With respect to bacterial community structure, we found that T32781 inoculation altered neither alpha nor the beta diversity. Nonetheless, bacterial genera in wheat and barley were found to be affected by T32781 inoculation in the pot and field trials. Among them, we identified 11 key bacterial genera that increased following T32781 inoculation in both wheat and barley in pot trials, including Exiguobacterium, Delftia, and Pseudomonas (Figure S12; Table S31).

Figure 4.

Alterations in the soil microbial community in the rhizosphere by T32781 inoculation in wheat and barley plants in Experiments 3, 4, and 5

(A–C) Relative abundances (%) of the top 10 most abundant fungal genera in the rhizosphere of wheat and barley in Experiment 3 (W‐CK, W‐Cd, W‐T), Experiment 4 (B‐CK, B‐Cd, B‐T) and Experiment 5 (Mock, T). (D–F) Volcano plots displaying decreased and increased fungal responses to T32781 inoculation at the genus level. The differential fungi with |Log₂(Fold Change)| > 1 and false discovery rate (FDR) values < 0.05 are marked with different colors, and nonsignificant differences remain uncolored. (G–I) Microbial co‐occurrence networks in the rhizosphere of wheat and barley with T32781 inoculation. The relative abundance of more than 1% of the amplicon sequence variants (ASVs) was used to determine co‐occurrence. Nodes with different bacterial and fungal phyla are colored for clearer visualization. Positive correlations are marked by red edges, and negative correlations are marked by blue edges. (J, K) Semi‐quantitative analysis of tef1 in the rhizosphere of wheat and barley and T32781 DNA.

With respect to alterations in bacterial and fungal composition, we analyzed the microbial co‐occurrence network in wheat and barley in the pot and field trials, respectively. Our results revealed that T32781 fostered a more interconnected and clustered network under Cd stress in wheat and barley in the pot trials, whereas T32781 induced a more dispersed and less interconnected network structure in barley in the field trial, likely because of its competitive effects on native soil microbes (Figure 4G–I).

Furthermore, gel electrophoresis results revealed a markedly brighter band for T32781 in the pot and field trials of wheat and barley (W‐T, B‐T, and T) with T32781 inoculation, indicating a dramatic increase in its abundance compared with that under conditions without inoculation (Figure 4J, K). Because the introduction of T32781 resulted in a subsequent alteration in the microbial composition, we subsequently used barley as a model crop to determine the profiles of soil and plant metabolites with untargeted metabolomics.

T32781 drives the alteration of soil and plant metabolomic profiles

A total of 1,030 metabolites were identified in the soil samples, predominantly classified as lipid and lipid‐like molecules, organic acids and derivatives (Figure 5A), according to the ClassyFire chemical ontology at Class I level. These metabolites were found to be associated with amino acid metabolism and nucleotide metabolism based on Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway classification (Figure 5B). Interestingly, the soil metabolite profiles were notably altered by T32781 inoculation in both pot and field trials (Figures S13, S14). In the pot trial, we identified 124 differentially abundant metabolites in soil, 20 of which were significantly down‐regulated and 104 of which were up‐regulated, such as, N‐acetyl‐methionine (NALM) and 4‐aminobutyric acid (GABA) in response to T32781 inoculation. KEGG analysis revealed that these differentially abundant metabolites were predominantly enriched in pathways such as tyrosine metabolism and fatty acid biosynthesis. In the field trial, we identified 54 down‐regulated and 81 up‐regulated differentially abundant metabolites in soils (e.g., gibberellin A4 and GABA) after T32781 inoculation, and KEGG analysis indicated significant enrichment in β‐alanine metabolism, ubiquinone and other terpenoid‐quinone biosynthesis, and biotin metabolism (Tables S32, S33). Moreover, 27 shared differentially abundant metabolites among field and pot trials were identified and predominantly classified as organic acids and derivatives. Among them, 21 soil metabolites, including GABA, NALM, and betaine, increased following T32781 inoculation in both field and pot trials (Figure 5C; Table S34). These metabolites are known to play significant roles in plant abiotic stress tolerance (Chen and Murata, 2002; Chen et al., 2007b; Huber et al., 2019; Ren et al., 2021), thus highlighting the beneficial effects of T32781 inoculation on barley plants.

Figure 5.

Metabolic profiles of soil and leaves in barley plants in response to T32781 inoculation at the heading stage in Experiments 4 and 5

(A, B) The total number of identified metabolites in soil classified under Class I (ClassyFire, class level) with KEGG pathway annotation. (C) Fold changes (FCs) of overlapping differentially abundant metabolites between field and pot trials. (D, E) The total number of identified metabolites in leaves in field and pot trials classified under Class I and mapped to KEGG pathways. (F–K) Principal component analysis, volcano plots, and KEGG enrichment analysis of differentially abundant metabolites in leaves driven by T32781 inoculation in pot (F–H) and field (I–K) trials, respectively. “NA” denotes metabolites without a specific classification at the Class I category. Differentially abundant metabolites were identified and marked with different colors based on P‐value < 0.05, FC > 1.2 or < 0.83 and variable importance in projection (VIP) > 1.

Moreover, we identified 1,156 metabolites in barley leaves at the heading stage in the field and pot trials. Based on ClassyFire, we determined that most metabolites were in the categories of lipid and lipid‐like molecules (21.6%), phenylpropanoids and polyketides (12.0%), and organic acids and derivatives (10.9%). Unlike those in the soil, the plant metabolites were predominantly involved in the metabolic pathways of amino acids and carbohydrates (Figure 5D, E). In the pot trial, T32781 inoculation substantially changed leaf metabolites, with 23 down‐regulated and 45 up‐regulated differentially abundant metabolites. KEGG enrichment analysis demonstrated that these differentially abundant metabolites are associated mainly with the metabolism of glyoxylate and dicarboxylate, as well as the biosynthesis and metabolism of arginine and proline (Figure 5F–H; Table S35). In the field trial, 28 differentially abundant metabolites in leaves (11 down‐regulated and 17 up‐regulated) were found to be involved mainly in the biosynthesis of amino acids (Figure 5I–K; Table S36). Furthermore, four shared differentially abundant metabolites were identified, with two metabolites increased (e.g., L‐citrulline) in both pot and field trials (Table S37).

Metabolites that were both enriched and shared between soil and leaves were identified and regarded as key metabolites, among which NALM, GABA, betaine, and citrulline were selected to investigate their roles in barley growth in a contaminated paddy field (Figure S15). Agronomic measurements revealed that the application of NALM and citrulline significantly increased plant height, panicle number, 1,000‐grain weight, and grain yield. Exogenous application of these key metabolites significantly reduced the grain Cd concentration, with NALM showing the strongest alleviation of Cd toxicity in barley. Thus, our results further confirmed that the T32781‐induced enrichment of soil and leaf metabolites is vital for improving growth and yield and decreasing grain Cd contamination under Cd exposure in barley.

The co‐occurrence network between the microbiome and metabolites reveals complex plant–microbe–soil interactions in Cd‐polluted soils

We then constructed a co‐occurrence network to analyze the associations between differential bacterial and fungal genera and soil and leaf metabolites. Significant (r > 0.7 or r < −0.7, P‐value < 0.05) correlations were identified, comprising 207 nodes and 700 edges (B‐T vs. B‐Cd) in the pot trial and 166 nodes and 381 edges (T vs. Mock) in the field trial (Figure 6A, B). Furthermore, we found that most soil metabolites were classified as lipids and lipid‐like molecules, whereas leaf metabolites were classified as organic acids and derivatives (Figure S16). The topological properties included an increased clustering coefficient, higher network density, and greater average number of neighbors in the pot trial, indicating a more interconnected and locally clustered network (Table S38). These changes may be associated with the relatively high abundance of Trichoderma observed in the pot experiments. A subnetwork specifically focused on Trichoderma‐related microbes and metabolites showed distinct co‐occurrence patterns between the field and pot experiments. In the pot trial, Trichoderma showed strong positively correlation with PGPM, such as Parasegetibacter and Aurantisolimonas (Figure 6C). In the field trial, soil metabolites such as organic acids (N‐acetyl‐DL‐glutamic acid) and nucleoside analogs (S‐(5′‐adenosyl)‐L‐homocysteine (SAH)) were strongly correlated with Trichoderma, whereas L‐serine in leaves was negatively correlated (Figure 6D).

Figure 6.

Network analysis of leaf and soil metabolites and microbiota at the genus level in Experiments 4 and 5

(A, B) T32781‐induced differentially abundant metabolites and microbes in both field (T vs. Mock) and pot (B‐T vs. B‐Cd) trials were selected for network construction based on the Pearson's correlation coefficient (P‐value < 0.05, |r | > 0.7). Node shapes are as follows: fungi, ellipses; bacteria, diamonds; leaf metabolites, rectangles; and soil metabolites, hexagons. Different colors indicate metabolites with specific classifications at the ClassyFire Class I. (C, D) Subnetwork specifically highlights Trichoderma‐related metabolites and microbes with positive (red) and negative (blue) relationships.

In summary, our results present a significant discovery of T32781 for cleaner production of barley and wheat (Figure 7), which is cost‐effective and environmentally friendly for the bioremediation of industrial Cd‐polluted soil with fewer physical and chemical interventions.

Figure 7.

Schematic diagram summarizing the potential underlying mechanisms

In the rhizosphere of wheat and barley, T32781 predominantly colonizes the soil, reducing soil‐available Cd and promoting the recruitment of other plant growth‐promoting bacteria (PGPB), for example, Exiguobacterium, Pseudomonas, and Delftia. In wheat plants, T32781 inoculation improved safe yield production, decreased Cd translocation and increased the abundance of the core fungal community upon Cd exposure. In barley plants, T32781 inoculation decreased Cd absorption with reduced shoot and grain Cd concentrations; facilitated plant‒microbe–soil interactions through the recruitment of plant growth‐promoting microbes (PGPM), for example, Pseudomonas, Flavobacterium and Aurantisolimonas; and facilitated the accumulation of beneficial metabolites, for example, betaine, 4‐aminobutyric acid (GABA), and N‐acetyl‐methionine (NALM), in rhizosphere soil and organic acids in leaves, collectively improving barley growth and grain yield.

DISCUSSION

T32781 is a novel PGP fungus in Cd‐polluted soil

Microbial applications for improving food production and quality are environmentally friendly approaches for addressing global population growth and food shortages. However, these PGPM do not always promote plant growth under field conditions because of soil nutrients, heavy metal concentrations, water availability, light intensity, temperature, and the complexity of microbe‒soil–plant interactions (Malgioglio et al., 2022). Here, we conducted whole‐genome sequencing of T32781 and identified numerous genes involved in producing key metabolites responsive to heavy metal stress, including those involved in intracellular transport and amino acid and carbohydrate metabolism (Figure 2B). In addition, we confirmed the classification and non‐toxic role of T32781 (Tables S5–S7). We subsequently elucidated the potential tolerance mechanisms of T32781 under Cd exposure (Figure 2C). The entry of Cd²⁺ induces oxidative stress, leading to the production of ROS, which subsequently activate the transcriptional coactivator yes‐associated protein 1 (YAP1) and the transcription factor skinhead 7 (SKN7) in the oxidative stress response system (Lee et al., 1999). These activations trigger the expression of glutathione synthetase genes (GSH1 and GSH2), resulting in the synthesis of the crucial antioxidant GSH, which directly neutralizes ROS (Ng et al., 2008). GSH also binds to intracellular Cd²⁺ to form the Cd–GS₂ complex. Driven by the proton gradient generated by vacuolar membrane ATPases (VAMs), the Cd–GS₂ complex is transported into the vacuole via the yeast cadmium factor protein (YCF1); Cd²⁺ is also transported into the vacuole through the sodium‐hydrogen exchanger (NHX1) and zinc resistance protein (ZRC1) (Li et al., 1997; Delaunay et al., 2000; Hirschi, 2001; MacDiarmid et al., 2002; Mendoza‐Cózatl et al., 2005; Grosjean et al., 2018).

Furthermore, we assessed the heavy metal resistance of T32781 using different media. The strain showed significant tolerance to Cd stress at a very high concentration (5 mmol/L) in 1/10 LB medium, with tolerance to various heavy metals at low concentrations (0.1 mmol/L) (Figure S2). Similarly, several Trichoderma spp. isolated from contaminated wastewater and polluted soil have demonstrated significant tolerance and absorption capacities for heavy metals such as Ni, Cu, and Cr (Ng et al., 2013; De Padua and Dela Cruz, 2021; Kumar and Dwivedi, 2021). In the present study, we observed a significant increase in the intracellular Cd concentration after 4 d of growth compared with 7 d in liquid medium (Figure 2J). These findings suggested that during the early growth stage, T32781 likely absorbed and sequestered Cd intracellularly as part of its detoxification strategy, thereby preventing Cd from interfering with essential cellular functions. In contrast, during the later growth stage, T32781 might activate additional Cd tolerance mechanisms to reduce intracellular Cd accumulation and alleviate the toxic effects of prolonged exposure. Furthermore, SEM‒EDS analysis revealed that Cd precipitated on the cell surface, where it could be removed by PBS (Figure 2E–I). These findings suggest that the Cd tolerance of T32781 is achieved by increased Cd efflux and the chelation of Cd, which might be attributed to genes involved in intracellular transport and metabolism. This dual mechanism highlights the unique role of T32781 in Cd immobilization in Cd‐polluted soil to improve Cd tolerance in both wheat and barley.

T32781 inoculation alleviates Cd toxicity by decreasing Cd concentrations and altering microbe‒soil interactions in wheat and barley

Soil Cd contamination is a severe environmental challenge that threatens food production and human health (Chaney, 2015). Microbial inoculants, known for their natural and sustainable potential as bioagents, have been extensively studied for their ability to promote plant growth and alleviate metal toxicity (Zhu et al., 2023). Among these, Trichoderma spp. are prevalent in soils and can promote plant growth under stressful conditions through various mechanisms, including the modulation of signaling pathways, activation of defense, and enhancement of nutrient absorption in plants (Hermosa et al., 2012; Li et al., 2015; Nieto‐Jacobo et al., 2017). Our results demonstrated the substantial Cd alleviation effects of T32781 on wheat and barley under different conditions in Cd‐polluted soil across five independent experiments (Figures 3, S3). We also observed differences in microbial composition, Cd concentration and growth conditions between the pot and field trials (Figure 4). These results indicated that T32781 inoculation exerted a prolonged positive effect on barley and wheat by significantly increasing biomass accumulation and grain yield under Cd exposure under both rainout shelter and field conditions.

Cadmium contamination in the grain of cereal crops increases the risk of Cd toxicity to humans when large quantities of contaminated food are consumed, emphasizing the need for effective strategies to ensure food safety. Recent research has focused on Cd immobilization in soil as a key strategy to minimize plant uptake and maintain healthy soils. For example, the Cd‐resistant bacterium Pseudomonas sp. B7 was reported to increase plant health in Cd‐polluted soil through Cd immobilization (Wu et al., 2022). Our results revealed the potential mechanisms underlying the reduction in Cd availability: T32781 not only adsorbs soil Cd²⁺ but also immobilizes Cd²⁺ by forming precipitates, which are then removed from the rhizosphere (Figures 2, 3). Hence, T32781 decreased Cd availability in soils, thereby reducing Cd absorption and translocation in plants and contributing to lower Cd concentrations. Our results indicate that T32781 inoculation alleviates Cd toxicity and decreases grain Cd concentrations mainly by reducing Cd absorption and translocation in plants through the immobilization of Cd in the rhizosphere.

Soil microbes play a pivotal role in the alleviation of plant growth under heavy metal stress (Tiwari and Lata, 2018). Although several PGPM have been reported to increase seedling growth and reduce heavy metal accumulation under controlled conditions, in‐depth molecular and agronomical evidence supporting their efficacy in improving crop yield and practical implementation is lacking (Pramanik et al., 2021). In this study, the analysis of fungal composition revealed that Trichoderma became the dominant fungal genus following T32781 inoculation. This inoculation enabled the successful establishment of T32781 in soils under different conditions, allowing this non‐native microorganism to persist and exert a sustained influence on wheat and barley under Cd contamination (Figure 4). After T32781 colonized the soils, 11 bacterial genera were enriched in both the wheat and barley rhizosphere soils (Figure S12). Among those shared genera, for instance, Exiguobacterium was reported to be tolerant to Cd‐induced oxidative stress and can immobilize soil Cd to alleviate heavy metal stress in plants (Kaur et al., 2024; Heisi et al., 2025); Pseudomonas was found to reduce Cd accumulation and enhance plant growth (Nabizade et al., 2023). In addition, we observed differential alterations in the microbial communities of wheat and barley after T32781 inoculation. In wheat, we found that T32781 increased the abundances of core fungal genera that were enriched under Cd stress alone, including Rhizophlyctidaceae_gen_Incertae_sedis, Trichoderma, Meyerozyma, and Neurospora. These results suggest that these genera play critical roles in wheat under Cd stress and that T32781 alleviates Cd toxicity in wheat by increasing the abundances of the core fungal genera. In barley, we found that T32781 inoculation altered the bacterial composition with the recruitment of beneficial PGP microbial genera, such as Flavobacterium and Acidovorax (Figure S10), both of which are known to play roles in bioremediation (Díaz‐Torres et al., 2024). Subnetwork analysis further revealed strong positive correlations between Trichoderma and microbes such as Parasegetibacter and Aurantisolimonas (Figure 6). Notably, these microbes were reported to be significantly enriched under stress conditions, suggesting potential roles in alleviation (Yuan et al., 2022; Yu et al., 2025). Therefore, our results indicated that T32781 actively modulates microbe‒soil interactions, resulting in significant changes in the soil microbial community composition, ultimately enhancing plant growth and Cd tolerance in both wheat and barley.

T32781 alters microbe‒plant‒soil interactions in barley through metabolic adjustment

Extensive studies have shown that plants adapt their metabolic pathways to mitigate the toxicity of heavy metals. For instance, Cd exposure led to a significant accumulation of amino acids in Compositae plants (Zhu et al., 2018) and increased the sugar concentration and shikimic acid metabolism in maize (Wang et al., 2021). These studies highlight the complex metabolic changes that occur in response to Cd stress. We demonstrated that T32781 inoculation induced significant plant–microbe interactions in barley, resulting in the accumulation of metabolites: primarily lipids and other molecules in the pot trial, and organic acids and derivatives in the field trial (Tables S35, S36). Four differentially abundant metabolites were shared between the two trials, suggesting that T32781 promotes barley growth and alleviates Cd toxicity through metabolic adjustment (Table S37). In summary, these findings strongly demonstrated that T32781 has potential as a bioagent to increase the resilience of crops to soil Cd contamination.

Soil metabolites, including root exudates and small molecules synthesized by microorganisms, are crucial in plant‒microbe interactions (Hu et al., 2018; Wong‐Bajracharya et al., 2020). These metabolites are known to influence heavy metal availability, soil redox potential, and metal complexation, thus affecting their translocation to plants (Violante et al., 2010; Li et al., 2013). Here, we investigated plant‒microbe–soil interactions by identifying key metabolites in leaves and soil that respond to T32781 in pot and field trials and constructing co‐occurrence networks (Figures 5, 6). Previous studies have reported that these key metabolites, including betaine, GABA, and NALM, play crucial roles in alleviating the effects of abiotic stresses on plant growth. Betaine has been shown to alleviate salt, drought, and chromium stresses (Hoque et al., 2007; Farooq et al., 2016; Gupta et al., 2021), whereas GABA is involved in the responses to salinity, drought, and Cd stress (Gilliham and Tyerman, 2016; Seifikalhor et al., 2020; Hasan et al., 2021). Additionally, NALM is also known to contribute to ROS scavenging, which could potentially alleviate Cd‐induced oxidative stress in plants (Kouno et al., 2014). We confirmed the roles of these key metabolites, along with citrulline, in enhancing barley growth and reducing the Cd concentration in barley grain in a field trial (Figure S15). Furthermore, co‐occurrence network analysis revealed strong correlations between plant metabolites, soil metabolites and microbes (Figure 6). The identified metabolites were primarily lipid molecules and organic acids, which play key roles in plant‒microbe interactions (Rajkumar et al., 2012; Macabuhay et al., 2022). Taken together, these results confirm that T32781 inoculation actively altered plant–microbe–soil interactions and enriched beneficial metabolites in both plants and soil, thereby alleviating Cd toxicity in barley.

In summary, inoculation with T32781 significantly increased yield and decreased grain Cd concentration in wheat and barley under Cd stress, suggesting that it is a promising strategy for mitigating Cd accumulation in crops and limiting human exposure via the food chain. The superior adaptation of T32781 positively alters plant‒microbe–soil interactions, promoting Cd adsorption and immobilization and thereby reducing soil Cd availability. Overall, we revealed a promising beneficial fungus, T32781, with great potential for sustainable and safe food production, particularly in regions with heavy metal‐contaminated soils.

MATERIALS AND METHODS

Whole‐genome sequencing, assembly and annotation

T32781 was initially isolated from peach rhizosphere and preserved at both 4°C and −80°C in a refrigerator in our laboratory, College of Agriculture and Biotechnology, Zhejiang University, China. Meanwhile, the holotype and ex‐type culture were deposited in the Chinese General Microbiological Culture Collection Center (https://www.cgmcc.net/english/; CGMCC40314) (Zhao et al., 2023). T32781 was cultured in 1/10 LB medium at 28°C with shaking at 200 rpm. The freshly harvested fungal biomass was frozen in liquid nitrogen, and total genomic DNA was extracted using a modified cetyltrimethylammonium bromide method (Watanabe et al., 2010). The complete genome of T32781 was sequenced using a hybrid whole‐genome shotgun strategy with the Illumina NovaSeq PE150 platform and the Nanopore PromethIon long‐read sequencing platform. Genome assembly was performed using Flye (version 2.9.1‐b1780), and polished with Pilon software (Walker et al., 2014). Assembly quality was controlled using BUSCO to evaluate genome completeness (Simão et al., 2015). Genome component prediction for T32781 included the identification of coding genes, repetitive sequences and noncoding RNAs (Liu et al., 2024). Functional annotation was subsequently performed using the COG database (Tatusov et al., 2003).

The pathogenic potential of strain T32781 was systematically investigated with the Pathogen–Host Interactions (PHI) database and the Virulence Factor Database (VFDB) to identify virulence genes and classify their effects (Baldwin et al., 2006; Chen et al., 2016). Antibiotic resistance genes were detected using the resistance gene identifier (RGI) tool and compared to reference clusters ( Alcock et al., 2020), and putative toxins (e.g., hemolysins/cytolysins) were screened through domain‐based homology searches (Stutz et al., 2022).

Experimental design and measurements for wheat and barley

We conducted a total of six independent experiments (Figure 1) from 2021 to 2023, including both pot and field trials, to assess the alleviating effects of T32781 on wheat (Triticum aestivum L., YM20, W) and barley (Hordeum vulgare L., Hua30, B). Pot trials in a controlled rainout shelter (Zhejiang Campus, Hangzhou) compared (i) CK (control), (ii) Cd (Cd stress), and (iii) T (T32781 inoculation + Cd). Field trials in Cd‐polluted paddy fields (Huanshan Village, Hangzhou) tested (i) Mock (polluted soil) and (ii) T (T32781‐inoculated soil). We also explored insights into plant–soil–microbiome interactions and the effects of exogenous application of identified key metabolites using barley plants (Experiment 6). Detailed methods are provided in the Supplementary Information.

Scanning electron microscopy

Scanning electron microscopy with energy‐dispersive X‐ray spectroscopy (SEM‒EDS; GeminiSEM 300, Oberkochen, Germany) was employed to analyze Cd adsorption. T32781 was cultured in 1/10 LB liquid medium for 7 d, either without (control) or with Cd treatment (0.5 mmol/L CdCl₂). After centrifugation at 1,500 × g, the biomass from the Cd‐treated cultures was divided into two portions: One was washed three times with sterilized PBS (pH = 7), and the other was washed three times with sterile deionized water. All samples, including untreated controls, were then freeze‐dried at low temperature for 24 h. The resulting powders were mounted to sample stages using a conductive adhesive. After gold sputtering, the fungal morphology before and after Cd adsorption was observed using SEM (Hitachi SU‐8010), and elemental analysis was conducted with EDS (Liu et al., 2023a).

Cadmium concentrations in T32781 and barley

Fungal cells were cultured for 7 d in 1/10 LB with 0.5 mmol/L CdCl₂ and then harvested and washed twice with sterilized deionized water. Cell samples were further washed with 20 mmol/L ethylenediaminetetraacetic acid (EDTA)–2Na for 10 min to remove Cd bound to the cell surfaces. The supernatant was collected to measure extracellular Cd absorption, and the cell pellets were analyzed for their intracellular Cd concentration. The extra‐ and intracellular Cd concentrations were analyzed by inductively coupled plasma‐mass spectrometry (ICP‐MS), and each treatment was performed three times (Han et al., 2018). To analyze Cd concentration, plant shoots (comprising leaves and stems) and roots were collected at the heading stage, and shoots, roots, and grains were harvested at the ripening stage. Plant tissues were digested with HNO₃, and Cd concentration was analyzed using ICP‐MS (Chen et al., 2007b). Simultaneously, the available Cd in soil was extracted using diethylenetriaminepentaacetic acid (DPTA) according to the China Environmental Protection Standards (HJ 804‐2016). The Cd translocation factor (TF) was calculated as follows:

$${\text{TF}}_{\text{shoot}/\text{root}}=\frac{C_{\text{shoot}}}{C_{\text{root}}}$$

where C _shoot and C _root denote the Cd concentrations (mg/kg) in the shoots and roots, respectively.

Fungal inoculum

The T32781 inoculum was prepared using a solid fermentation method. Initially, T32781 was cultured in test tubes containing 1/10 LB medium at 28°C for 7 d. A sterilized fermentation substrate was prepared using a mixture of rice husks and bran (1:1, w:w) supplemented with 2% sucrose, 1% KH₂PO₄, 0.05% CaCl₂, and 0.05% MgSO₄. A 10‐mL aliquot of T32781 spore suspension was introduced into the fermentation substrate and incubated at 28°C (12 h/light, 12 h/dark) for 7 d. The final T32781 inoculum was extracted by washing the solid fermentation substrate with water (1:20, w:v), resulting in a final concentration of ∼3–8 × 10⁶ spores/L.

Semi‐quantitative polymerase chain reaction for T32781 in soil

To detect the colonization of T32781 in soil, a semi‐quantitative polymerase chain reaction (PCR) method was employed (Zhao et al., 2023). DNA was extracted from soil samples using a Magnetic Soil and Stool DNA Kit (TianGen, Beijing, China). The tef1 gene of T32781, which encodes translation elongation factor 1‐alpha and is highly conserved in all eukaryotes, including fungi, was targeted for amplification using the following primers: forward (5′‐ATCGTAAGTTTGCAATCCTCG‐3′) and reverse (5′‐GTTGGCACCCCACTAAAAGC‐3′). PCR products were resolved on a 1.5% agarose gel and visualized using nucleic acid staining under UV light to highlight the DNA bands.

Analysis of soil microbial community structure

To amplify the bacterial sequences, the hypervariable V4–V5 region of the 16S rRNA was selected with the primer pair 515F (GTGCCAGCMGCCGCGGTAA) and 907R (CCGTCAATTCCTTTGAGTTT). Similarly, the fungal ITS sequences were amplified using the primer pair ITS1F (CTTGGTCATTTAGAGGAAGTAA) and ITS2R (GCTGCGTTCTTCATCGATGC). A standard bioinformatics analysis pipeline was used to analyze the microbial community structure (Liu et al., 2023c). Alpha diversity indices, including Chao1 and Shannon, were calculated from the amplicon sequence variant (ASV) to assess within‐group diversity, with differences between groups evaluated using Kruskal‒Wallis test followed by Dunn's test. Beta diversity was examined through principal coordinate analysis (PCoA) by calculating the Bray‒Curtis distance matrix to evaluate dissimilarities among microbial communities. To detect differences in ASV abundance among samples, we implemented a negative binomial generalized linear model using the edgeR package (Zhang et al., 2022). We specifically compared the effects of Cd stress (B‐Cd vs. B‐CK) and T32781 inoculation under Cd exposure (B‐T vs. B‐Cd in the pot trial; T vs. Mock in the field trial). The analysis began by constructing a DGEList object from the ASV table, followed by filtering out ASVs with zero counts in both pot and field trials to ensure robust statistical results. The calcNormFactors function was then used to calculate normalization factors for the count data. Tagwise, common, and trended dispersions were estimated using the estimateDisp function. To assess differential ASV abundance, we applied the glmFit function to fit a generalized linear model to the normalized counts. Finally, differential genera were identified based on an absolute Log₂‐fold change ( | Log₂FC | ) > 1 and a false discovery rate (FDR)‐adjusted P‐value < 0.05. Microbial co‐occurrence networks were constructed to describe the intricate interactions between bacteria and fungi under different treatments (Feng et al., 2024). ASVs with a relative abundance greater than 1% were included, and significant correlations were determined based on Pearson correlation coefficients ( | r | > 0.8, P‐value < 0.01).

Metabolomic profiles of leaves and soil

Untargeted metabolomics was conducted using liquid chromatography‐mass spectrometry (LC‐MS) to investigate the metabolic responses to T32781 inoculation in Cd‐contaminated soil (An et al., 2021). First, the soil samples were passed through a 2 mm sieve and then homogenized in liquid nitrogen. Similarly, leaf samples were pulverized with liquid nitrogen and resuspended in prechilled 80% methanol. The samples were subsequently chilled on ice and then centrifuged at 15,000 × g at 4°C for 20 min. Then, the concentration of the supernatant was determined by LC‐MS. The raw data files obtained from the analyses were processed with Compound Discoverer 3.3 (CD3.3; Thermo Fisher), which facilitated peak alignment, peak selection, and quantification of the metabolites.

Metabolite annotation was conducted using the KEGG database (Kanehisa et al., 2008) and assigned with ClassyFire ontology (Djoumbou Feunang et al., 2016). Principal component analysis (PCA) was used to elucidate the changes in metabolomic profile resulting from T32781 inoculation, which was further complemented by pathway enrichment analyses. Differentially abundant metabolites were identified based on a set of criteria: A variable importance in projection (VIP) score >1, P‐value < 0.05, and FC > 1.2 or < 0.83.

Co‐occurrence network of metabolites and the microbiome

To elucidate the plant‒microbe–soil interactions in response to T32781 inoculation under Cd exposure, we constructed a co‐occurrence network using the relative abundance of metabolites and microbes (B‐T vs. B‐Cd in pots and T vs. Mock in the field) in barley. Only T32781‐induced differential microbial genera and metabolites of leaves or soil were considered, and significant correlations were determined based on Pearson correlation coefficients, with r > 0.7 indicating strong positive correlations and r < −0.7 indicating strong negative correlations, with a significance threshold of P‐value < 0.05 (Wu et al., 2023). During construction, only plant‒microbe‒soil pairs with significant correlations were included; interactions within each individual category (leaf metabolites, soil metabolites, bacterial, or fungal communities) were discarded (Baker et al., 2024). The resulting correlation matrix was used to construct networks, which were visualized using Cytoscape (v. 3.10.1), and their topological characteristics were systematically analyzed to provide insights into the influence of T32781 inoculation on microbial and metabolic dynamics.

Statistical analyses

To evaluate the alleviation effects of T32781 inoculation on barley under Cd exposure, we calculated the response value for each trait and Cd concentration according to the following equation:

$$\text{Response }(\%)=({\text{Trait}}_{+\mathrm{T}}-{\text{Trait}}_{-\mathrm{T}})/{\text{Trait}}_{-\mathrm{T}}\times 100\%$$

where Response (%) is the percentage improvement, Trait _+T is the trait value with Cd stress and T32781 inoculation (treatment “W‐T” and “B‐T” in the pot trial and “T” in the field trial), and Trait_−T is the trait value with Cd stress only (“W‐Cd” and “B‐Cd” treatments in the pot trial and “Mock” in the field experiment). Independent‐sample t‐tests were used to compare differences between the pot and field trials in SPSS 20.0 (IBM, Armonk, NY, USA). Statistical analyses were conducted in Excel 2019 (Microsoft, Redmond, WA, USA).

CONFLICTS OF INTEREST

The authors declare no competing interests.

AUTHOR CONTRIBUTIONS

F.W. conceived and designed the study and coordination, supervision, resources, funding acquisition, writing–review and editing the manuscript. S.Z. performed the experiments, formal analysis, writing–original draft, review and editing. C‐W.Q., J.S., S‐H. S., Z‐F.G. investigation–performed the experiments and plant physiochemical measurements. W.H.M. genome assembly and analysis. C.Z. resources–strains. M.A.A. reviewed and edited the manuscript. Z‐H.C. writing–review and editing the manuscript. All authors read and approved the final version of the paper.

Supporting information

Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.70063/suppinfo

Figure S1. Phylogenic tree generated by maximum likelihood analysis according to the concatenated sequences of rpb2 and tef1 loci of the Trichoderma genus

Figure S2. Heavy metal tolerance of Trichoderma nigricans T32781

Figure S3. Response values of agronomic traits and Cd concentrations in barley grains at the ripening stage in pot and field trials in Experiments 1 and 2

Figure S4. Phenotypes of barley and wheat after inoculation of T32781 in Experiments 3–5

Figure S5. Cd concentrations in wheat and barley plants at the ripening stage

Figure S6. Microbial communities in different treatments in wheat rhizosphere soil in Experiment 3

Figure S7. Microbial communities in different treatments in barley rhizosphere soil in Experiment 4

Figure S8. Microbial communities in different treatments in barley rhizosphere soil in Experiment 5

Figure S9. Impact of different treatments on microbial communities in wheat rhizosphere soil in Experiment 3

Figure S10. Impact of different treatments on microbial communities in barley rhizosphere soil in Experiment 4

Figure S11. Impact of different treatments on microbial communities in barley rhizosphere soil in Experiment 5

Figure S12. Upset plots of differential bacterial and fungal genera in Experiments 3–5

Figure S13. Overview of soil metabolome data between B‐T and B‐Cd in the pot trial

Figure S14. Overview of soil metabolome data between T and Mock in the field trials

Figure S15. Impact of T32781‐induced metabolites in soil and leaves on barley growth under open field conditions in Experiment 6

Figure S16. Classification of differential soil and leaf metabolites in co‐occurrence networks at the Kyoto Encyclopedia of Genes and Genomes (KEGG) Class I

Table S1. Functional annotation of predicted genes in T32781 based on the COG database

Table S2. Genes involved in Cd tolerance

Table S3. Conserved protein families of antibiotic resistance candidate proteins

Table S4. BLAST results for hemolysin‐related proteins in the UniProt database

Table S5. BLAST results for hemolysin‐related proteins in the Pfam database

Table S6. Conserved families of hemolysin candidate proteins

Table S7. BLAST results for cytolysin‐related proteins in the UniProt database

Table S8. Conserved protein families of cytolysin candidate proteins

Table S9. Measurement of the size of T32781 colonies grown on 1/10 LB or PDA solid medium supplemented with different concentrations of heavy metals after 15 d

Table S10. Analysis of T32781 biomass when grown in 1/10 LB or PD liquid media with different heavy metal concentrations after 4 d

Table S11. Agronomic traits and grain Cd concentrations of barley and soil‐available Cd at the ripening stage in both the pot and field trials in Experiments 1 and 2

Table S12. Agronomic traits and Cd concentrations in wheat plants at the heading stage in Experiment 3

Table S13. Agronomic traits and Cd concentrations in wheat plants at the ripening stage in Experiment 3

Table S14. Agronomic traits and Cd concentrations in barley plants at the heading stage in the pot trial of Experiment 4

Table S15. Agronomic traits and Cd concentrations in barley plants at the ripening stage in the pot trial of Experiment 4

Table S16. Agronomic traits and Cd concentrations in barley plants at the heading stage in the field trial of Experiment 5

Table S17. Agronomic traits and Cd concentrations in barley plants at the ripening stage in the field trial of Experiment 5.

Table S18. Differential bacterial genera in wheat under Cd stress in Experiment 3

Table S19. Differential fungal genera in wheat under Cd stress in Experiment 3

Table S20. Differential bacterial genera in wheat under Cd stress and T32781 inoculation in Experiment 3

Table S21. Differential fungal genera in wheat under Cd stress and T32781 inoculation in Experiment 3.

Table S22. Topological properties of the co‐occurrence networks at the genus level in wheat of Experiment 3

Table S23. Differential bacterial genera in barley under Cd stress in the pot trial of Experiment 4

Table S24. Differential fungal genera in barley under Cd stress in the pot trial of Experiment 4

Table S25. Differential bacterial genera in barley under Cd stress and T32781 inoculation in the pot trial of Experiment 4

Table S26. Differential fungal genera in barley under Cd stress and T32781 inoculation in the pot trial of Experiment 4

Table S27. Topological properties of the co‐occurrence networks at the genus level in barley of Experiment 4

Table S28. Differential bacterial genera in barley under T32781 inoculation in the field trial of Experiment 5

Table S29. Differential fungal genera in barley under T32781 inoculation in the field trial of Experiment 5

Table S30. Topological properties of the co‐occurrence networks at the genus level in barley of Experiment 5

Table S31. Overlaps of increased bacterial genera in both wheat and barley under Cd stress and T32781 inoculation of Experiments 3 and 4

Table S32. Differential soil metabolites in response to T32781 inoculation under Cd stress in the pot trial of Experiment 4

Table S33. Differential soil metabolites in response to T32781 inoculation in the field trial of Experiment 5

Table S34. Overlaps of differential soil metabolites in response to T32781 inoculation between Experiments 4 and 5

Table S35. Differential leaf metabolites in response to T32781 inoculation under Cd stress in the pot trial of Experiment 4

Table S36. Differential leaf metabolites in response to T32781 inoculation in the field trial of Experiment 5

Table S37. Overlaps of differential leaf metabolites in response to T32781 inoculation between Experiments 4 and 5

Table S38. Topological properties of the co‐occurrence networks (total or Trichoderma‐related) related to differentially abundant metabolites and microbes in response to T32781 inoculation in Experiments 4 and 5

Biographies

Zhang, S. , Qiu, C.‐W. , Shi, S.‐H. , Sreesaeng, J. , Gao, Z.‐F. , Zhang, C. , Mao, W. , Ali, M. A. , Chen, Z.‐H. , and Wu, F . (2026). Trichoderma genome and multiomics insight into promoting yield and reducing grain cadmium in barley and wheat. J. Integr. Plant Biol. 68: 39–56.

Data availability statement

The authors confirm the data supporting the findings of this study are available within the article. Genome sequence data of T32781 have been deposited in the National Microbiology Data Center (NMDC) with the accession number NMDC60214225 (https://nmdc.cn/resource/genomics/genome/detail/NMDC60214225). Further raw data is available from the corresponding author Prof. Feibo Wu, upon acceptance and reasonable request.