Southern root knot nematode (Meloidogyne incognita), collected from Langfang, Hebei, China, and single egg masses were picked for expansion and culture on susceptible tomatoes. Egg masses were manually picked from the root system and second stage juveniles at 25 °C were used for inoculation. Tomato ‘AC’ model variety, provided by the Tomato genetic breeding group, Institute of Horticultural Crops, Xinjiang Academy of Agricultural Sciences. Planted in the Key Laboratory of Genome Research and Genetic Improvement of Xinjiang Characteristic Fruits and Vegetables, the planting conditions were inoculated with 1000 second instar larvae of root-knot nematodes/plant in sterile sandy soil (7: 3) from tomato roots after 3-4 weeks. 5 days after inoculation, tomato roots were collected and four biological replicates were set up for each treatment and stored at -80°C after liquid nitrogen snap freezing for RNA Extraction.

Download the tomato 5.0 genome annotation file from the Solanaceae Genome Database (http://solomics.agis.org.cn/tomato/), and download the S. pennellii genome annotation file from the Solanaceae Database (https://solgenomics.net/). Four AtPAL amino acid sequences (Huang et al., 2010) were downloaded from the TIAR website (http://www.arabidopsis.org) based on known reports. AtPALs were used as genome-wide query sequences in the Phytozome (https://phytozome.jgi.doe.gov) database for BLASTP to obtain PAL members of rice, potato (Solanum tuberosum) and maize (Zea mays) (Goodstein et al., 2012; Fernandez-Pozo et al., 2015). Hidden Markov files (PF00221.21) of PAL structural domains were downloaded from the Pafm database (https://www.ebi.ac.uk/interpro/search/text/), and tomato and S pennellii candidate genes (E-value: 1e-5) were extracted using Hmmsearch (Eddy, 2008; Finn et al., 2011). The protein sequences of the Arabidopsis PAL family genes were used as seed sequences for local construction of BLAST libraries, which were analyzed using Smart (http://smart.embl.de/) and NCBI-CDD (https://www.ncbi.nlm.nih.gov/cdd) online software to validate the tomato and S. pennellii PAL family gene The tomato and S. pennellii PAL family genes were finally identified by deleting genes without typical PAL family domains (Sonnhammer et al., 1997; Schultz et al., 2000). The physicochemical properties of PAL genes were predicted by the online Expasy (https://web.expasy.org/protparam/) tool (Artimo et al., 2012) and by WoLF PSORT (https://www.genscript.com/wolf-psort.html?src=leftbar) online software to predict protein subcellular localization (Horton et al., 2007).

Chromosome lengths and all tomato PAL family loci were obtained based on the genome annotation file GFF3 and then conformed using Map Inspect software (Jiang et al., 2019). The phylogenetic tree was constructed using MEGA7 software to analyze the PAL amino acid sequences of Arabidopsis, rice, maize, potato, tomato, and S. pennellii by multiple sequence alignment and set by Neighbor-joining (NJ) and Maximum likelihood (ML), Bootstrap test 2000 replicates (Kumar et al., 2016). Itools online website (https://itol.embl.de/) was used to display the midpoint root tree. Mapping of intron exons was performed in TBtools software using existing tomato gff files and tomato PAL gene family IDs to map the tomato PAL gene structure. motif was then obtained using meme software to obtain the meme.xml file of the SlPAL family genes, with the search motif value set to 20 and other default parameters (Grundy et al., 1997). The conserved motifs were then predicted for analysis using Visualize MEME/MAST Motif Pattern in TBtools software (Chen et al., 2020). Structural domain analysis was performed using the SlPALs protein sequence in NCBI-cdd to obtain hit data files, which were plotted by TBtools software.

The 2000 bp sequence upstream of all tomato PAL genes was extracted as candidate promoter sequences using tomato genome annotation files, cis-acting element analysis was performed using PlantCare (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) website (Rombauts et al., 1999), and cis-acting element mapping was performed using TBtools software (Chen et al., 2020). For multispecies covariance, MCscanX in TBtools software was used to extract covariate genes of species pairs and finally mapped using Multiple Synteny Plot. The substitution rate of paralogous genes was calculated using KaKs_Calculator 2.0 (Wang et al., 2010). Protein-protein interaction relationships were predicted using the STRING online website (https://string-db.org/) (Szklarczyk et al., 2019).

Plant root RNA was extracted using the Tiangen Plant Polyphenol Polysaccharide Total RNA Extraction Kit (Beijing, China). The extracted total RNA was then subjected to cDNA synthesis using the 5 × All-ln-One RT MasterMix (AccuRT Genomic DNA Removal Kit; G492, ABM, Vancouver, Canada) reverse tanscription kit. qRT-PCR primers for SlPAL family members were designed using DNAMAN6 software. The designed qRT-PCR primers were sent to Biotech Biologicals (Shanghai, China) for synthesis ( Supplementary Table S1 ). Quantitative PCR (qPCR) analysis was subsequently performed on a LightCycler machine using ChamQ Universal SYBR qPCR Master Mix (Q711, Vazyme, Nanjing, China). The Slactin was used as the internal reference gene. Each treatment contains four independent biological replicates, and each replicate contains three technical replicates. Gene expression was calculated using the 2^-ΔΔCt method (Livak and Schmittgen, 2001).

In this study, we used HMM search and Arabidopsis amino acid sequences (AtPALs) for BLAST. The results were finally combined, redundant sequences were removed, and they were named according to their positions on the chromosomes ( Table 1 ). Finally, 14 members of the cultivated tomato PAL family gene were identified and named SlPAL1-SlPAL14, and the 11 members of the wild relatives of S. pennellii PAL family gene were named SpenPAL1-SpenPAL11. Two tomato PAL family amino acids range in length from 218 (SpenPAL4) to 1274 (SlPAL10). Molecular weights were in the range of 23065.33 (SpenPAL4)-139461.81 (SlPAL10). pI ranged from 5.61 (SlPAL3)-9.44 (SpenPAL4). The pI values of 11 PAL genes of cultivated tomatoes were less than 7, and the pI values of 10 genes PAL genes of S. pennellii were also less than 7, indicating that most of these two tomatoes belong to acidic proteins. The pI values ranged from 78.17 (SlPAL6) to 104.10 (SlPAL5), hydrophobic values ranged from 0.055 to -0.432, only SlPAL5 was 0.055, and the rest of the PAL family hydrophilic values of cultivated tomatoes and S. pennellii were less than zero, indicating that these proteins are hydrophobic. In subcellular localization, it can be found that there are 8 SlPAL genes distributed in chloroplasts, one in cytoplasm of 3 distributions, one in mitochondria, plasma membrane and endoplasmic reticulum respectively, 6 SpenPAL genes distributed in chloroplasts, two in cytoplasm and endoplasmic reticulum respectively, and one in plasma membrane.

To fully understand the gene distribution mechanism of the tomato PAL family on chromosomes ( Figure 1 ), 14 SlPAL genes and 11 SpenPAL genes were mapped by TBtools, all distributed on chromosomes 3, 5, 9, and 10, respectively. And the SlPAL and SpenPAL families have 4 and 3 gene clusters, respectively, indicating that tandem replication is likely to be the main amplification mechanism of these two tomato gene families.

To investigate the evolutionary relationships of PAL genes, phylogenetic analysis was performed on Arabidopsis (4), rice (9), maize (13), potato (8), cultivated tomato (14), and S. pennellii (11) altogether PAL genes ( Figure 2 ). A phylogenetic tree diagram was constructed by comparison using MEGA 7.0 software. These PAL genes were clustered into three subclades (GroupI, GroupII, GroupIII). From the figure, we can find that monocotyledons and dicotyledons are more distantly related, and monocotyledons are all clustered on Group II subfamily, and PAL proteins of these subfamilies may evolve in different directions in monocotyledons and dicotyledons. The dicotyledonous plants potato, cultivated tomato, and S. pennellii occupy two subfamilies, Group I and Group III, thus confirming that the PAL genes of these three species are very inwardly related and may have similar biological functions.

To further understand the differences within the tomato PAL family, mapping was performed by TBtools software ( Figure 3 ). The results showed that there were five structural domains in tomato PAL family, Lyase_aromatic, Lyase_I_like superfamily were the typical structural domains. It can be found that the tomato PAL family with Lyase_aromatic structural domains are also relatively uniform in their conserved motifs. Group III subfamily SlPAL10 has two Lyase_aromatic and its conserved motifs are duplicated, SlPAL6, SlPAL7, SpenPAL4 have the structural domains of Lyase_I_like the conserved motifs and structural domains of SlPAL5 and SlPAL8 of Group I subfamily are identical.

The intron (Intron), coding region (CDS), and untranslated region (UTR) of 14 SlPAL and 11 SpenPAL gene families were mapped using TBtools. 11 SpenPAL genes had 1-3 CDSs, 6 of which had 2 CDSs, but none of the 11 SpenPAL genes had a UTR. SlPALs genes have 2-13 CDSs, and SlPAL14 gene has the most CDSs. Ten of the SlPALs genes had UTRs, and five had two UTRs.

Multiple sequence alignment analysis revealed ( Figure 4 ) that the amino acid sequences of both SlPAL and SpenPAL genes contain the active site GTITASGDLV(L)PLSYIAG of PAL, except for SlPAL1, SlPAL5, SlPAL6, SlPAL7, and SpenPAL3. In this site is contains the Ala-Ser-Gly composition of the highly conserved methylene imidazolone (MIO) electrophilic group. In the active sites of SlPAL5 and SpenPAL4, isoleucine (I) at position 3 may be mutated to valine (V), and glycine (G) at position 17 may be mutated to arginine (R) and valine (V). In the active sites of SlPAL11, SlPAL12, SlPAL13, and SpenPAL10, isoleucine (I) at position 3 may be mutated to leucine (L). The alanine (A) at position 16 in the active site of SpenPAL6 may be mutated to valine (V).

Many studies have shown that the PAL gene family is involved in various stress responses. To better understand and elucidate the possible regulatory functions of SlPALs and SpenPALs under different stresses, we identified stress-related and phytohormone-related cis-elements from the 2000 bp promoter region of the start codon ( Figure 5 ).

According to the data obtained from PlantCARE, five stress response control elements were found in the promoters of SlPALs and SpenPALs: anaerobic induction (ARE), drought (MBS), defense and stress response (TC-rich repeats), cold stress (LTR), and hypoxia-specific enhancers (GC-motif). 6 major growth and developmental response control elements were identified: regulation of maize protein metabolism (O2-site), involvement in endosperm expression (GCN 4 _ motif), participation in circadian control (circadian), participation in meristem expression (CAT-box), participation in cell cycle regulation (MSA-like), and participation in palisade mesophyll cell differentiation (HD-Zip 1). All found 20 major photoresponse elements: (AT1-motif, Sp 1, LAMP-element, GT1-motif, G-Box, ATC-motif, ATCT-motif, Box 4, MRE, AE-box, TCC-motif, G-box, TCT-motif, chs-CMA 1 a, ACE, CAG-motif, 3-AF1 binding site, AAAC-motif, GATA-motif, chs-CMA1a). 4 hormone-responsive elements were found: phytoauxin (TGA-element), gibberellin (P-box, TATC-box, GARE-motif), methyljasmonic acid (TGACG-motif, TCA-element, CGTCA-motif), and abscisic acid (ABRE). The heat map made to select the largest number of control elements in stress, growth, light response, and hormones shows that ARE elements in stress, O2-site elements in growth and development, Box 4 and G-box elements in light response, and ABRE elements in hormones are generally more abundant in SlPALs and SpenPALs, while SlPALs and SpenPALs also contain a large number of other regulatory elements. These results suggest that PAL genes are widely involved in a variety of life activities such as plant growth and development and stress response.

To reveal the evolutionary origin and direct homology of the tomato PAL gene family, a covariance analysis of the PAL gene family on the chromosomes of four dicots [tomato (S. lycopersicum, S. pennellii), Arabidopsis, and potato] and two monocots (maize and rice) was performed in this study ( Figure 6 ). The results showed that two genes of SlPALs were direct homologs with AtPALs, among which SlPAL10 had colinearity with AtPAL1 and AtPAL3, and SlPAL14 had colinearity with AtPAL2 and AtPAL4. There are 6 genes of SlPALs that are orthologous with SpenPALs, among which, SlPAL10 has collinearity with SpenPAL6 and SpenPAL11, SlPAL14 has collinearity with SpenPAL6 and SpenPAL11, and the collinearity of other genes corresponds. SlPALs have 1 gene that is direct homolog to OsPALs, but SlPAL10 has colinearity with OsPAL1 and OsPAL5. SlPALs also have 1 gene that is direct homolog to ZmPALs are direct homologs, but also SlPAL10 has colinearity with ZmPAL3 and ZmPAL9. For example, two genes, AtPAL1 and AtPAL3, have colinearity with SlPAL10, suggesting that the two pairs of genes may regulate similar or complementary functions, respectively. It can be found that most PAL genes in Arabidopsis and S. pennellii potato, which are dicotyledons, are direct homologs of the two genes SlPAL10 and SlPAL14. Interestingly, the PAL genes of rice and maize, two monocotyledons, are only direct homologs of SlPAL10, suggesting that the SlPAL10 gene existed before monocotyledon differentiation.

Repeated sequences on the tomato chromosome set were extracted by native BLAST and McscanX software to explore the co-linearity of the PAL gene family within the S. pennellii genome ( Figure 7 ). As shown in the figure, there is one paralogous homologous gene pair each in cultivated tomato and S. pennellii, SlPAL10/SlPAL14 and SpenPAL6/SpenPAL11, and both of these paralogous homologous gene pairs originate from duplication of segments on their respective chromosomes 9 and 10.

The Ka/Ks ratio is usually an important indicator of selection pressure in evolution ( Table 2 ). The ka/ks values of SlPAL10/SlPAL14 and SpenPAL6/SpenPAL11 paralogous homologous gene pairs were calculated by TBtools to be less than 0.3, indicating that the evolution of these two gene pairs was affected by purifying selection pressure, and the gene functions tended to be conserved and the evolutionary rate was slowed down.

The mechanism of action of different tomato PAL proteins was specifically analyzed by constructing the interaction network of SlPALs and SpenPALs proteins ( Figure 8 ). We excluded some proteins with missing annotations and low values, and the results showed that the highest number of proteins interacted with SlPAL8, SlPAL3, SlPAL14 and SlPAL10 in cultivated tomato and they had strong interactions with each other. Among them, SlPAL6 and SlPAL7 had no direct interactions with other PAL proteins. In contrast, the highest number of proteins interacted with SpenPAL4, SpenPAL5, SpenPAL7 and SpenPAL11 in S. pennellii. Among them, there were no direct interactions between SpenPAL6 and other PAL proteins. As can be seen from the figure, different tomatoes exhibited similar protein action mechanisms. Chalcone synthase (CHS) is the first rate-limiting enzyme in the plant flavonoid biosynthetic pathway, and the expression of CHS is closely related to the anthocyanin synthesis and accumulation metabolism. The CHS gene is also involved in plant response to abiotic stresses, with implications for plant flower color formation, growth and development, adversity stress, hormone regulation and transport (Dao et al., 2011). Flavonoid 3’-hydroxylase (F3’H), an intermediate in the synthesis of anthocyanins, often cooperates with chalcone isomerase and chalcone synthase to form different anthocyanins, an important step in flower color formation (Winkel-Shirley and Flavonoid, 2001). PAT proteins are members of lipid droplet proteins family that regulates cellular lipid storage. It may play an important role in the rapid synthesis of lipid droplets, the stabilization of nascent lipid droplets, the regulation of lipid droplet maturation, and the intracellular transport of lipid droplets (Wolins et al., 2005). 4-coumarate: CoA ligase (4CL) regulates plant lignin metabolism and is involved in the synthesis of flavonoids and other secondary metabolites synthesis. Studies have shown that 4CL is important for cell wall formation, water transport, plant stress resistance, and disease resistance (Chen et al., 2019b; Geng et al., 2020; Xiao et al., 2021). There are also some missing annotated proteins in the interaction network map. They have obvious direct or indirect synergistic effects with SlPALs and SpenPALs proteins, but their functions remain unclear.

To identify genes associated with SlPAL expression patterns, we performed a weighted gene co-expression network analysis (WGCNA) using tomato resistance and fruit development-related RNA-seq sequencing data (PRJNA888477). A total of 18224 tomato genes with expression levels (FPKM) ≥10 in at least one sample were calculated for the construction of WGCNA ( Figure 9 ). The results showed that there are two PAL genes present in separate co-expression modules, module A contains 21 co-expressed genes, of which Solyc01T002708.1 and Solyc11T000183.1 have mangiferin biosynthesis function, implying that PAL is involved in mangiferin synthesis together with these genes. Module B contains 32 co-expressed genes, of which Solyc06T000114.1 and Solyc09T002292.1 have successful functions in encoding enzymes and proteins, implying that PAL is involved in enzyme and protein synthesis together with these genes. The genes with the highest connectivity in the 2 modules were selected as the core genes and mapped separately for the interactions network. the functional annotation of the 53 core genes in the 2 key modules is shown in ( Supplementary Table S2 ).

To verify the potential role of SlPALs under root-knot nematode infection, their expression levels were analyzed by qRT-PCR ( Figure 10 ). It could be found that after root-knot nematode infection, the expression of SlPAL5, SlPAL8, SlPAL11, and SlPAL12 genes were higher than the control, and the expression of SlPAL3, SlPAL4, and SlPAL6 genes were lower than the control.