The tomato (Solanum lycopersicum) cultivar Moneymaker which is susceptible to C. fulvum was used in this work. All tomato plants were grown in pots containing nutrient soil in a plant growth incubator with photoperiod conditions of 16h/8h, control temperatures of 26°C/18°C, and humidity set at 75%. Cladosporium fulvum was kindly provided by the Institute of Plant Protection, Chinese Academy of Agricultural Sciences. Conidia of C. fulvum were harvested from one-week-old PDA plates with distilled water, and adjust the spore suspensions concentration to 1× 10⁶ conidia/ml (de Wit et al., 2012). Tomato plants grown in soil at 3 weeks old were sprayed with a conidial suspension (1× 10⁶ conidia/ml) on both adaxial and abaxial sides of the leaves (Van Esse et al., 2007). After 24 h of inoculation, the third leaves below the growing point were collected from treated and control plants separately, rapidly frozen in liquid nitrogen, and stored at -80°C (Yu et al., 2015). Remaining plants were monitored for disease progression for up to 20 days. Three biological replicates were performed.

Leaf surface sprayed with C. fulvum (1× 10⁶ conidia/ml) or sterile water. At 24 h after inoculation (hai), leaf samples were harvested and total RNA was isolated using the Trizol method (Meng and Feldman, 2010; Gao et al., 2014a). The RNA concentration was quantified by a NanoDrop-2000 nucleic acid spectrophotometer (Thermo Fisher Scientific, Wilmington, DE). After detecting the RNA integrity in 1% agarose gel, mRNA was enriched from a pool of RNA by Oligo-dT magnetic beads. Based on the manufacturer’s instructions, RNA was sheared into small pieces by using RNA fragmentation kit (Illumina, San Diego, CA, USA). cDNA was synthesized by reverse transcription with N6 primers. The cDNA fragments were subjected to end repair and adapter ligation. Following the PCR amplification, the PCR products were used to generate cDNA libraries. Quality control and library construction were entrusted to Huada Gene Technology Company (Wuhan, Hubei, China). The cDNA sequencing was carried out with BGISEQ-500 platform for generating raw reads. The RNA-seq data for C. fulvum inoculated samples (accession no: SRR21437047、SRR21437048、SRR21437049) and control-inoculated samples (accession no: SRR21437044、SRR21437045、SRR21437046) are available at the NCBI gene expression omnibus server (https://www.ncbi.nlm.nih.gov/geo/).

Raw data obtained from BGISEQ-500 platform are subsequently subjected to quality control (QC) to determine whether the sequencing data are suitable for further analysis. The filtering is performed by SOAPnuke software as follows: first, remove the reads containing junction (junction contamination); second, remove the reads with unknown base N content greater than 5%; finally, reads with over half of the component bases with a quality score below 15 were defined as low-quality reads and subsequently removed. After quality control, the filtered clean reads were mapped to the reference genome sequence (GCF_000188115.3_SL2.50, https://www.ncbi.nlm.nih.gov/assembly/GCF_000188115.3/) using HISAT (Hierarchical Indexing for Spliced Alignment of Transcripts). After the alignment, the comparison result is judged by the statistics of the mapping rate and the distribution of reads on the reference sequence to pass the second QC of alignment. The dataset analysis was performed after meeting the requirement.

DEGs with |Log₂ Fold Change| ≥ 1 and Q-value ≤ 0.05 were functionally classified and enriched using the phyper function in R software to calculate P- value, False discovery rate (FDR) correction was then performed to obtain a normalized P-value, also known as Q-value. Gene Ontology (GO, http://geneontology.org/) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.genome.jp/kegg/) pathway enrichment were used to analysis DEGs. KEGG pathways with an adjusted Q-value of ≤0.05 were regarded as significantly enriched (Gao et al., 2014b).

A total of randomly selected 10 genes were used for the measurement of transcript abundance by quantitative real-time PCR to verify the reliability of the transcriptome data. Total RNA was extracted from frozen leaf tissue using Trizol reagent according to the manufacturer’s instructions. The integrity of RNA was analyzed by 1% (w/v) agarose gel electrophoresis (Gao et al., 2017). The reverse transcription was performed using the All-in-One First-Strand Synthesis MasterMix Kit (with dsDNase) (Lablead, Beijing,China), while qRT-PCR analysis was conducted using the TB GreenTM Premix Ex TaqTM kit (TaKaRa, Shanghai, China) with a RealTime PCR System (Bio-Rad, Hercules, CA, USA). The PCR running procedure was as follows: a pre-denaturation at 95°C for 30 seconds, 40 cycles with each cycle employing a denaturation temperature at 95°C for 5 seconds and an annealing/extension temperature at 60°C for 30 seconds, followed by melt curve analysis. For each biological sample, three technical replicates were used and there was a total of three biological replicates. Relative quantification of genes was carried out using the 2^-ΔΔCT method. The primer sequences for the ten genes are listed in Table 1 . SlACTIN (LOC101260631) served as an internal control. The same method was used to detect the relative expression of six genes related to plant hormones signaling pathway with specific primers listed in Table 2 . This experiment was done in three independent biological replicates.

Leaves were taken from 3-week-old plant, and a hole punch was used to make a 6 mm diameter round sample from the same part of every leave. The round leave samples were placed in a container with distilled water for 9 h at room temperature to get rid of the mechanical damage for the leaves. Horseradish peroxidase (Shenggong, Shanghai, China) was dissolved in sterile water to make a working solution with a final concentration of 10 mg/ml. C. fulvum were maintained on potato dextrose agar (PDA) at 22°C prior to use, and diluted the spore suspensions in sterile water to 1×10⁶ conidia/ml (de Wit et al., 2012). Mixed solution with 1μl horseradish peroxidase working solution, 1μl conidial suspension (1×10⁶ conidia/ml) as treatment or 1μl sterile water as the control, and 98μl of ECL (enhanced chemiluminescence) solution (Shenggong, Shanghai, China) was as ROS determination solution. For ROS measurement, the cut leaves were removed from water container and put in ROS determination solution immediately. The machine used to measure ROS production is Glomax microplate luminometer (Promega, Wisconsin, USA). One value was detected every 10 seconds, for a total of 160 values per sample. The trend of the values indicates the degree of ROS accumulation (Robineau et al., 2020). The results are presented as means ± SE of three biological and three technical replicates.

Spore suspensions with C. fulvum (1×10⁶ conidia/ml) were sprayed uniformly on the adaxial and abaxial surfaces of leaves, using sterile water as a control. At 24 h after inoculation, 3-4 leaf samples were taken from three plants for each treatment. The collected samples were immediately frozen in liquid nitrogen and stored at -80°C. The contents of SA, ABA and JA were determined simultaneously by ultra performance liquid chromatography (UPLC) LC-30A coupled with a triple quadrupole mass spectrometer LCMS-8040 (Waters, Massachusetts, USA). The leaf samples prepared in advance were placed on ice and transferred to the rapid nucleic acid extractor. Rapid nucleic acid extraction reagent (900 mg) and 1 ml of ethyl acetate extraction reagent were added and shaken for 45 seconds. The resulting solution was centrifuged at 4°C for 20 min at 16000 g and the supernatant was extracted and evaporated at 30°C under reduced pressure. The dried sample was dissolved in 0.5 ml of 70% methanol solution and centrifuged at 16000 g for 20 min. The supernatant obtained was filtered and put on the machine. The mobile phase consisted of binary gradients of acetonitrile with 0.01% (v/v) formic acid and 0.01% (v/v) aqueous formic acid, with a flowing speed at 0.5 ml/min. A 20 µL portion of each sample was injected into the UPLC-ESI-MS/MS system (Simura et al., 2018; Xin et al., 2020). The results are presented as means ± SE of three biological replicates and every biological repeat includes three technical replicates.

In order to observe the symptoms induced by C. fulvum, a suspension of spores containing the leaf mold pathogen (1× 10⁶ conidia/ml) was sprayed on tomato leaves. Compared to the control group, pathogen-inoculated leaves with yellow spots at 7 days after inoculation. At 20 days after inoculation, yellow spotted area expanded; mold layer can be observed in the abaxial side of leaves and when severe leaves were obviously curled ( Figure 1 ). These symptoms indicated successful infection of C. fulvum and the plant materials we taken at the time point of 24 hours were reliable for experiments.

Detection of high levels of ROS, a signal substance, can reflect the ability of the plant to resist pathogen at some extent (Hernandez et al., 2016). The leave samples were firstly put in distilled water up to 9 hours to avoid impact caused by mechanical damage, then we measured the ROS level of tomato leaves treated with C. fulvum. Results showed that the timing and accumulation of ROS production in tomato leaves in response to C. fulvum were different compared with control treatment. Tomato leaves treated with C. fulvum produced a peak of ROS accumulation and the overall value was higher than the control ( Figure 2 ). This indicates that C. fulvum infection activates an early ROS burst in leaves.

Tomato infected by C. fulvum was used as the treatment group and sterile water treatment was used as the control group for transcriptome sequencing. Three biological repeats were performed for each group. A total of six samples were sequenced using the DNBSEQ platform, and each sample yielded an average of 1.19G of data. We obtained an average of 23.81 M high-quality reads per sample, accounting for > 99% of the raw reads for each sample. The reads with a quality value of 30 accounted for > 95% of the total reads and a quality value of 20 account for > 98% of the total reads, indicating that the sequencing reads are of high quality ( Table 3 ). Hierarchical Indexing for Spliced Alignment of Transcripts (HISAT) was used to align the clean reads to the reference genome sequence (http://daehwankimlab.github.io/hisat2/) after obtaining the clean reads. The average alignment rate to the reference genome of Solanum lycopersicum was 97.54%, indicating the high quality of the data, guaranteeing the reliability of subsequent differentially expressed genes (DEGs) analysis.

Clean reads were compared to reference gene sequences by Bowtie2 software, and then gene expression levels of individual samples were calculated using RNA-Seq by Expectation-Maximization (RSEM) (deweylab.github.io), followed by DEGs screening according to the method described by Michael (Love et al., 2014). The significant DEGs were determined based on the criteria of |Log₂ Fold Change| ≥ 1 and Q-value ≤0.05, the distribution of DEGs based on degree of difference and the significance of the difference were visualized with a volcanic plot ( Figure 3 ). Compared to the control, C. fulvum inoculation plants induced 1909 upregulated DEGs and 2090 downregulated DEGs ( Tables S1 , S2 ).

Gene Ontology (GO) is an international standardized gene function classification system. Based on GO terms, DEGs were classified into three major categories: biological process, cellular component, and molecular function. Within the broad GO category of biological process, a total of 2693 DEGs were involved in 24 GO terms ( Figure 4 ). According to previous reports, terms with “cellular process”, “metabolic process”, “biological regulation”, “regulation of biological process”, “response to stimulus” were associated with plant disease resistance (Zhang et al., 2020). In the classification of cellular component, there were 3280 DEGs involving 16 GO terms ( Figure 4 ), DEGs mainly concentrated in the integral component of membrane. In the molecular function category, more than 85% of the 2346 DEGs were mainly enriched in the “catalytic activity” and “binding” ( Figure 4 ). The enrichment of DEGs in these two terms has been found to associate with the induction of many genes involved in plant hormone signal transduction pathway (Zhang et al., 2020; Singh et al., 2021).

The KEGG (Kyoto Encyclopedia of Genes and Genomes) database is a database that systematically analyzes gene functions, links genomic information and functional information. Through KEGG enrichment analysis, C. fulvum infection induced 830 DEGs involved in 125 KEGG pathways. Fifteen pathways with P- values ≤0.05 were selected based on the number of involved DEGs of corresponding pathways arranged in a descending order ( Table 4 ). The KEGG pathway with the highest number of enriched DEGs was Plant hormone signal transduction pathway (ko04075) involving in 61 DEGs, which includes 40 upregulated DEGs and 21 downregulated DEGs ( Table S3 ). The results above suggest that plant hormone pathways were significantly affected by C. fulvum infection. More importantly, many pathways associated with plant disease resistance were activated, such as Plant-pathogen interaction (ko04626), MAPK signaling pathway – plant (ko04016), and Peroxisome (ko04146) (Nyathi and Baker, 2006; Jiang et al., 2022a). In addition, many metabolic pathways such as Starch and sucrose metabolism (ko00500) and Circadian rhythm – plant (ko04712) were also altered, indicating that pathogenic fungal infection can also affect normal plant growth and development.

In our study, a total of 45 DEGs related with the plant defense responses were identified after C. fulvum inoculation, in which 33 genes were upregulated and 12 genes were downregulated ( Table S4 ). These genes can be considered indicators of plant defense in response to C. fulvum invasion. A total of six serine/threonine-protein kinase receptors or receptor like kinases were identified ( Table 5 ), all of which were upregulated after C. fulvum inoculation. In particular, two genes encoding of LRR receptor-like serine/threonine-protein kinase FLS2 (flagellin sensing 2) and FLS3 (flagellin sensing 3) that binds to flg22 and flgII-28 respectively (Hind et al., 2016; Chi et al., 2021) were significantly induced to 6-fold and 27-fold compared to the control treatment. And gene encoding SERK3B (somatic embryogenesis receptor kinase 3B), which interacts with FLS2 for activating downstream signaling (Ma et al., 2022), was also slightly upregulated. This result indicated that tomato activated the intracellular immune signaling responses after C. fulvum invasion.

In response to various stimuli in plants, Ca²⁺ acts as a second messenger to modulate different target protein activities through CAM/CML receptors, thereby regulating a variety of cellular functions (Zhang et al., 2014). A total of 15 DEGs related with Ca²⁺ signaling were identified upon C. fulvum infection with 9 upregulated and 6 downregulated genes ( Table 5 ). Genes encoding a few of calmodulin (CML) or calmodulin like proteins such as calmodulin (LOC101245298), CML44 (LOC101245539), calcium-binding protein (LOC101264550, CP1), and one potential CDPK protein (LOC101260391, calcium-dependent protein kinase 18-like) were upregulated. Four genes encoding CDPK were downregulated after C. fulvum infection, including CDPK1 (calcium-dependent protein kinase 1), CDPK17-like, CDPK24 and CDPK29, which is consistent with previous report that CDPK1 expression was decreased after 1.5 h treatment with C. fulvum (Chico et al., 2002). This result indicated that differently expression of CDPK genes are involved in the complex signaling network of tomato in response to C. fulvum invasion.

In addition, four genes encoding 3-Ketoacyl-CoA synthase (KCS), which catalyzes a condensation reaction to form 3-ketoacyl-CoA during very long chain fatty acid synthesis, including KCS1, KCS6, 3-ketoacyl-CoA synthase 20-like (LOC101262858) and 3-ketoacyl-CoA synthase 11-like (LOC101268257) were all upregulated in our study ( Table 5 ). These gene expression might contribute to plant cuticular wax and suberin biosynthesis after pathogen invasion (Edqvist et al., 2018). Moreover, two transcriptional factors WRKY26 and WRKY1, which have been reported as key components of resistance in tomato against Alternaria solani (Shinde et al., 2018), were induced in our study. Further, several defense responses related genes such as those encoding pathogenesis-related (PR)1 protein (LOC100191111) (Lavrova et al., 2017), PR1b (Hoegen et al., 2002), and P4 (Pieterse and van Loon, 1999) were significantly induced. Together, our data analysis suggested that plant defense pathways are activated after C. fulvum inoculation in tomato plant.

Next, we chose to analyze DEGs related to plant hormone signal transduction pathway which play key roles in regulating plant defense response against pathogen ( Table S3 ). In the SA signaling pathway, a total of 9 DEGs were identified; 6 genes were upregulated and 3 genes were downregulated. The downregulated genes include encoding the transcription factor TGA1 (TGACG MOTIF-BINDING FACTOR 1), pathogenesis-related leaf protein 4 (LOC101265854) and BOP2 (BLADE-ON-PETIOLE protein), which is associated with leaf and flower development in Arabidopsis (Zhang et al., 2017). The upregulated genes include a SA receptor NPR1 (Nonexpressor of PR1, also known as NIM1) (Li et al., 2019b), NML2 (NPR1/NIM1-like protein), TGA2.2 (Hou et al., 2019) and PR genes (PR-1b1 and P4) ( Table 6 ). This result suggested that SA signaling pathway are activated after C. fulvum invasion given that PR1 is an indicator for activation of the SA signaling pathway in plants.

We further analyzed genes involved in the ABA signaling pathway affected by C. fulvum invasion. The complex consisting of the ABA receptor PYR/PYL/RCAR (Pyrabactin resistance/PYR-like/regulatory components of ABA receptor), PP2C (type 2C protein phosphatase), and SnRK2s (Sucrose Non-fermentation Kinase Subfamily 2) has a key role in ABA signaling (Guo et al., 2011). In our study, genes encoding abscisic acid receptor PYL9 (LOC101246807), PYL3 (LOC101258886), PYR1(LOC101267127), PP2C (protein phosphatase 2C 53)and SRK2C (SNF1-related kinase), analogous to subclass III SnRK2s (Mizoguchi et al., 2010), were all upregulated after C. fulvum invasion. In addition, LOC101249794 encoding protein phosphatase 2C 51-like, SRK21 (LOC101251432) encoding serine/threonine-protein kinase SnRK21 and ABF4 (ABA responsive transcription factor) (Orellana et al., 2010) were downregulated ( Table 6 ). Previous study suggested that once PYR/PYLs binds to ABA, it represses PP2C activity and releases SnRK2 suppression, triggering downstream ABA responses (Yang et al., 2017). Our data analysis indicated that ABA signaling might be activated after C. fulvum invasion.

In the JA synthesis pathway, the genes LoxC (lipoxygenase), AOS2 (allene oxide synthase 2), and AOC (allene oxide cyclase) related to JA synthesis (Liu et al., 2012) were downregulated 2-8 fold ( Table 6 ). And four genes involved in the ethylene signaling pathway were upregulated, including ERF1 (ethylene response transcription factor 1), LOC101246590 (ethylene-responsive transcription factor 1B), EREB (ethylene responsive element binding protein) (Mingchun et al., 2016) and LOC101249950(ETHYLENE INSENSITIVE 3-like 3 protein) ( Table 6 ). These results suggested that C. fulvum inoculation impacted the JA and ET signaling pathway differently in plants.

A total of 24 DEGs related to auxin signaling pathway were identified after C. fulvum invasion, of which 13 were upregulated and 11 were downregulated ( Table 7 ). The SAUR (small auxin-up RNA) genes can respond not only to auxin but also to internal and environmental stress with mounting dynamic spatial-temporal responses (Wang et al., 2020). In particular, 11 SAUR genes were identified to be affected after C. fulvum treatment. Among these 11 DEGs, genes such as SAUR32, SAUR58, an SAUR71 were upregulated and those SAUR71-like (LOC101248065, LOC101257321) genes were downregulated. The other class of DEGs affected by C. fulvum inoculation was IAAs; upregulated genes include IAA4, IAA14, IAA35, IAA7, and downregulated genes include IAA15, IAA17, and IAA19 (Audran-Delalande et al., 2012). One gene ARF5 (auxin response factor 5) was inhibited while the gene ARF1 (auxin response factor 1) was slightly induced (Zouine et al., 2014). Two genes encoding Auxin transporter-like protein LAX2 and LAX5 were downregulated (Pattison and Catala, 2012). The above results indicate that C. fulvum infection does affect the auxin-related pathways in tomato plants in a complicated manner.

Plant phenylpropanoid pathway is an important pathway for the synthesis of plant secondary metabolites, some of which have been proposed as important components of plant defense responses (Naoumkina et al., 2010; Piasecka et al., 2015). Its downstream metabolites mainly include coumarins, flavonoids, terpenoids, anthocyanins, lignin and other phenylpropanoids (Deng and Lu, 2017; Pratyusha and Sarada, 2022). In this study, many tomato DEGs were related with plant phenylpropanoid pathway after C. fulvum invasion. A total of 47 genes was enriched in the phenylpropanoid biosynthesis pathway, with 26 DEGs upregulated and 21 DEGs downregulated ( Table S5 ). Phenylalanine ammonia lyase (PAL) is the first rate-limiting enzyme in the phenylpropane metabolic pathway (Ruili et al., 2016). PAL2 (LOC101249824) was found to be slightly induced after invasion. The other two genes (LOC101244496, LOC101248210) encoding trans-cinnamate 4-monooxygenase and 4-coumarate-CoA ligase, two major players in this pathway, were also upregulated ( Table 8 ). The gene encoding cinnamoyl-CoA reductase (CCR, LOC778359), a key enzyme in the formation of lignin monomers (Fan et al., 2015), was upregulated. Peroxidases (PRXs) are often found in lignifying tissues and they have the capability to oxidize a wide variety of small phenolic compounds, including monolignols (Marjamaa et al., 2009). In our study, there are 9 related genes upregulated such as TAP2 (tomato anionic peroxidase 2, LOC101245316) (Melillo et al., 2014) and CEVI-1 (citrus exocortis viroid) (Vera et al., 1993) encoding an anionic peroxidase in tomato, which could be induced by compatible viral infection (Mayda et al., 2000), while peroxidase 12, peroxidase 18, peroxidase 3, peroxidase 44 like and 45 like genes were downregulated ( Table 8 ). Four genes related with flavonoid biosynthesis, CHS1 (chalcone synthase 1), CHS2 (Heredia et al., 2015), CHI1 (chalcone-flavonoid isomerase 1) (Kang et al., 2014) and DFR (dihydroflavonol 4-reductase) (Bongue-Bartelsman et al., 1994) were significantly downregulated ( Table 8 ). Both ANS (anthocyanidin synthase) and F3H (flavanone 3-dioxygenase) (Li et al., 2019c) involved in anthocyanin synthesis were also downregulated ( Table 8 ). In addition, a few genes encoding beta-glucosidase, catalyzing the hydrolysis of cellobiose and cello-oligosaccharides containing (1 → 4)-beta-glycosidic bonds to glucose, which is crucial in cellulosic ethanol production were downregulated, such as LOC101254239, LOC101263519, LOC101256554, whereas only 1 beta-glucosidase 18-like (LOC101248595) was upregulated ( Table 8 ), suggesting cellulosic ethanol production pathway was inhibited after C. fulvum invasion. Our results indicated that a variety of secondary metabolites related to plant disease resistance was affected by fungal infection.

To validate the reliability of transcriptome data, we randomly picked some upregulated and downregulated genes with one gene related to plant resistance. A total of 10 genes was selected including LOC101249624, LOC109119038, LOC101245298, LOC101266084, LOC101246590, LOC101263535, LOC101265854, LOC104648161, LOC101258353, LOC101267111 ( Figure 5A ). The expression data obtained from qRT-PCR verification had consistent trend with RNA-Seq, with a correlation coefficient of 0.8779 ( Figure 5B ), indicating the RNA-seq data were reliable.

According to our RNA-seq data analysis, some key genes involved in SA and ABA signaling were upregulated after C. fulvum invasion. Thus, we decided to measure the content of SA and ABA. Our results showed that both SA and ABA contents increased after fungal invasion, indicating SA and ABA singling pathways were activated after C. fulvum treatment. We further tested the JA content and showed that JA level decreased after fungal invasion in comparison to the control treatment ( Figure 6A ). This is in consistence with the RNA-seq data in which a few genes related with JA synthesis was inhibited after C. fulvum inoculation.

To further confirm the impact to SA, ABA and JA hormone pathways after C. fulvum invasion, two genes related to each plant hormone signal transduction pathway were selected for qRT-PCR gene expression analysis. Results showed that the relative expression of genes with RNA-seq and qRT-PCR are the similar trend ( Figure 6B ). The qRT-PCR results showed that expression of ABA-related SlSRK2C and SlPYR1) and SA-related genes (SlNPR1 and SlPR1) were upregulated 2-4 folds. In contrast, expression of synthetic genes of JA (SlLoxC and SlAOC) were significantly downregulated by more than 3-fold. Our results further confirmed that these three hormone pathways are affected after C. fulvum invasion.