For the analysis of transcriptional changes undergoing arbuscular mycorrhization and SlDLK2 overexpression, raw RNA-seq data obtained from previous experiments were used (NCBI BioProjects PRJNA509606 and PRJNA523214, respectively) (Ho-Plágaro et al., 2021). Data belonged to three root pool samples for each condition: mycorrhized roots inoculated with the AM fungus Rhizophagus irregularis and non-inoculated roots (Experiment 1), and SlDLK2 overexpressing (SlDLK2 OE) hairy roots and control roots transformed with the empty vector (Experiment 2). RNA-seq sequence processing was performed as detailed in Ho-Plágaro et al. (2021). Gene Ontology enrichment analysis was performed using the PANTHER database (Mi et al., 2021). To identify possible altered hormone-related pathways in response to SlDLK2 OE on tomato roots, overrepresentation analyses on the significantly SlDLK2 OE-induced and repressed genes (fold change >2 or < −2, respectively, and p-value<0.05) were performed using MetGenMap (Joung et al., 2009). Gene expression heatmaps were generated by z-normalization of log2 count values of selected genes using Heatmapper (Babicki et al., 2016).

For the analysis of transcriptional changes undergoing SlDLK2 overexpression in mycorrhizal plants, an experiment with mycorrhizal composite plants overexpressing SlDLK2 was set up.

The pUBIcGFP-DR: SlDLK2 vector obtained in Ho-Plágaro et al. (2021) was used for A. rhizogenes transformation, and hairy root composite plants were obtained as described in Ho-Plágaro et al. (2018) and Ho-Plágaro et al. (2021). Composite tomato plantlets were inoculated with Rhizophagus irregularis and grown as explained in the “Plant growth and AM inoculation” methodological section.

50 days after AM inoculation, root samples from three control plants transformed with the empty vector (31.67 ± 2.62% mycorrhizal colonization) and three SlDLK2 OE composite plants (19.67 ± 3.72% mycorrhizal colonization) were collected. Total RNA was extracted using the Rneasy Plant Mini Kit (Qiagen). The quality and quantity of total RNA samples were assessed using a NanoDrop 1,000 spectrophotometer (Thermo Scientific), and samples were normalized at the same concentration (6 μg, 300 ng μl⁻¹). Later, samples were sent to Sistemas Genómicos SL (Paterna, Valencia, Spain) for cDNA library preparation and sequencing using an Illumina HiSeq1000 machine.

For RNA-seq sequence processing, the TOPHAT v.2.1.0 algorithm (Trapnell et al., 2009) was used to align reads from the RNA-seq experiment to the Tomato Genome Reference Sequence SL3.0 provided by the Sol Genomics consortium at¹, using the ITAG 3.10 annotation. Then, low-quality reads were removed from the map through Picard Tools², and high-quality reads were selected for assembly and identification through Bayesian inference using the CUFFLINKS v.2.2.1 algorithm proposed by Trapnell et al. (2010). Gene quantification process was performed by the HTSEQ-COUNT 0.6.1p1 tool (Anders et al., 2015). Isoform quantification and differential expression was carried out through the DESEQ2method (Anders et al., 2015). The RNA-seq data have been deposited in the NCBI Short Read Archive (SRA) with accession no. PRJNA509606.

Tomato stable transformation was carried out in the laboratory of Tissue culture and plant breeding at the Institute for Plant Molecular and Cellular Biology (IBMCP, Valencia, Spain). The genetic construction in the pK7FWG2 plasmid was introduced into Agrobacterium tumefaciens LBA4404, and Agrobacterium-mediated transformation of Moneymaker tomato cultivar cotyledons was performed as previously described (Ellul et al., 2003). The empty vector pK7FWG2 was used for the obtention of the control tomato lines.

To induce rooting, the elongated shoots obtained after subculture of the buds were grown in Murashige and Skoog medium (Murashige and Skoog, 1962) supplemented with 0.1 mg l⁻¹ indole-3-acetic acid (IAA) and 50 mg l⁻¹ kanamycin. T0 plants were grown in soil under standardized greenhouse conditions. After successive self-pollination events, T2 progenies were screened for kanamycin resistance conferred by the NEOMYCIN PHOSPHOTRANSFERASE II (NPTII) marker gene, and azygous (null resistance to kanamycin) and homozygous (100% kanamycin resistance in the progeny plants) lines were identified according to results from this test. Kanamycin test was carried out by sowing seeds on Murashige and Skoog (MS) agar medium supplemented with sucrose (10 g l⁻¹) and kanamycin (100 mg l⁻¹). PCR analysis on DNA extracted from leaves was performed to corroborate the presence of the kanamycin-resistant nptII gene insert. Two azygous control lines transformed with the empty vector (WT-1 and WT-2) and three homozygous independent T2 lines overexpressing SlDLK2 (OE-1; OE-2 and OE-3) were selected for further studies, and the DLK2 expression level was analyzed by quantitative real-time polymerase chain reaction (RT-qPCR) using specific primers (Supplementary Table 1). Azygous plants are considered ideal controls because they have been submitted to the entire process of transformation for generating transgenic plants but they have lost the transgene through segregation.

Plant hormone analysis was performed at the Plant Hormones Quantification platform (IBMCP, Valencia, Spain) by Ultra-Performance Liquid Chromatography–Mass Spectrometry (UPLC-MS), using a Thermo Scientific™ Q Exactive™ Hybrid Quadrupole-Orbitrap Mass Spectrometer. In brief, 50 mg of lyophilized leaves or roots material was ground in liquid nitrogen, homogenized in 80% methanol −1% acetic acid containing internal standards, and subjected to gentle agitation for 1 h at 4°C. The resulting extract was maintained at 20°C overnight and then centrifuged, and the supernatant was dried in a vacuum evaporator. The dry residue was suspended in 1% acetic acid, filtered through an Oasis HLB column (Waters Corp., Milford, MA, USA), and subjected to chromatographic separation (Hernandez et al., 2021).

As strigolactone (SL) production in the roots is promoted under Pi-deficiency conditions (López-Ráez et al., 2008), a phosphate-starvation experiment was set up for strigolactone analysis in tomato root exudates. Plants were grown in a 1:1 mixture of washed vermiculite and sand in 500 mL pots. Initially, plants were watered with 20 mL of complete Long Ashton nutrient solution (Hewitt, 1966) three times a week for 2 weeks. The substrate was then washed with 1 liter of tap water before starting the phosphorus (P) treatments. For the following 2 weeks, plants were watered daily with 25 mL of either standard phosphorus Long Ashton solution for the control treatment (5.2 mM Pi, +P) or phosphorus-free solution for the P-starvation treatment (0 mM Pi, –P) as Pi-deficient culture conditions promote exudation and detection of SLs such as orobanchol, solanacol, and didehydro-orobanchol(s) (López-Ráez et al., 2008; Rial et al., 2019). Subsequently, the substrate was washed with 500 mL of the respective nutrient solution (+P or –P) to remove accumulated compounds. Plants were kept in a growth chamber for 48 h and irrigated to field capacity with the corresponding nutrient solution after 24 h. After this period, fresh root exudates were collected by washing the substrate with 1 liter of tap water, and roots were weighed and stored at −80°C for future analysis. Exudates were vacuum-filtered through glass filters, concentrated, and purified using Telos C18 SPE columns (Telos, Kinesis, UK) and an SPE vacuum manifold (Supelco). SPE columns were first pre-equilibrated with 5 mL of 100% acetone and washed with 5 mL of distilled H₂O, and then, a liter of each exudate solution was loaded onto the pre-equilibrated columns. Each column was washed with 5 mL of 40% acetone, and the exudates were eluted with 5 mL of 60% acetone and collected in 10 mL amber tubes. Purified root exudates were stored at −80°C until use.

For strigolactone (SL) analysis, a 15 μL aliquot of 25 nM GR24 (internal standard) was added to 150 μL of purified root exudate. The mixture was evaporated to dryness, redissolved in 15 μL of acetonitrile (ACN), and analyzed using Nexera X2 UHPLC coupled with MS-8050. Chromatographic separation was performed on an ACQUITY BEH C18 column with specific gradient elution parameters. Mass spectra were obtained using electrospray ionization in positive mode, and SLs were identified by comparing retention times and MRM transitions with authentic standards. Data processing was performed using LabSolutions 5.72 software.

Solanum lycopersicum seeds were surface-sterilized with a 2.35% sodium hypochlorite solution for 5 min, shaken at room temperature for 1 day in the dark, and germinated on sterilized moistened filter paper at 25°C in the dark for 4 days. Germinated seeds were then placed on vermiculite for 1 week to allow hypocotyl elongation. Each seedling was transferred to a 500 mL pot containing a sterilized mixture of expanded clay, vermiculite, and coconut fiber (2:2:1). For arbuscular mycorrhizal (AM) treatments, plants were inoculated with 200 spores of Rhizophagus irregularis. Plants were grown in a growth chamber with a 16-h light/8-h dark cycle at 24°C/20°C and 50% relative humidity. 1 week after planting, and weekly thereafter, pots received 20 mL of modified Long Ashton nutrient solution with 325 μM phosphorus to avoid mycorrhizal inhibition. Non-mycorrhizal plants received the same nutrient solution. Plants were harvested at 57 and 77 days post-inoculation (dpi), with roots washed and rinsed for different measurements. The non-vital trypan blue histochemical staining procedure and the assessment of the intensity of root cortex colonization by AM fungus were performed as described by Ho-Plágaro et al. (2020).

For the quantitative reverse transcription-PCR (RT-qPCR) experiments, representative root samples were collected, frozen in liquid nitrogen, and stored at −80°C until RNA extraction. Approximately 0.2 g of root samples were used to isolate total RNA with the RNeasy Plant Mini Kit (Qiagen), followed by DNase treatment. One microgram of DNase-treated RNA was reverse-transcribed into cDNA using the iScript™ cDNA synthesis kit (BioRad). For the qPCR, a 20 μL reaction mixture was prepared containing 1 μL of diluted cDNA (1:10), 10 μL of 2× SYBR Green Supermix (BioRad), and 200 nM of each primer in a 96-well plate. A negative control without reverse transcription was included to check for DNA contamination. The PCR program included a 3-min incubation at 95°C, followed by 35 cycles of 30 s at 95°C, 30 s at 58–63°C, and 30 s at 72°C, with a melting curve analysis performed after the final cycle. Experiments were conducted on three biological replicates, with each biological replicate having three technical replicates. The threshold cycle (Ct) values were normalized to the geometric mean of Ct values from the housekeeping genes SlEF-1a (accession no. X14449) and SlActin2 (NM_001321306.1). Relative transcription levels were calculated using the 2^−ΔΔCt method (Livak and Schmittgen, 2001), and the RT-qPCR data were shown as relative expression compared to a reference treatment (to which a value of 1 was assigned). The genes analyzed and corresponding primers are listed in Supplementary Table 1.

Comparisons among all means were performed using a one-way analysis of variance (ANOVA) followed by Holm–Sidak’s multiple comparison test. The GraphPad Prism v 8.0.2 (GraphPad Software, Boston, MA, USA) was used to determine statistical significance. Differences at a p-value of < 0.05 were considered significant.

For Gene Ontology (GO) analyses, significantly induced or repressed genes (fold change >2 or < −2, respectively; p-value <0.05) upon SlDLK2 overexpression (BioProject PRJNA523214) and mycorrhization (BioProject PRJNA509606) were submitted to the Panther tool. As shown in Table 1 and Supplementary Table 2, an overall overrepresentation of GO terms commonly associated with genes repressed by SlDLK2 OE and induced by mycorrhization was observed, reflecting the negative regulatory role of SlDLK2 on mycorrhization previously described by Ho-Plágaro et al. (2021). Among the SlDLK2 OE-repressed and AM-induced gene sets, we found that many overrepresented GO terms are well-known to be activated during mycorrhization, such as the “cell wall modification,” “response to wounding,” “response to external stimulus,” “defense response,” and “transmembrane transport” GO terms from the biological processes category; the “transmembrane transporter activity,” “hydrolase activity,” “transcription regulatory activity,” and “antioxidant activity” GO terms from the molecular function category; and the “cell wall,” “cell periphery,” and “vacuole” GO terms from the cellular component category.

Among the different GO terms commonly regulated by SlDLK2 OE and mycorrhization, it caught our attention different GO terms associated with hormone regulation, response and signaling, such as “response to cytokinin,” “response to auxin,” “regulation of jasmonic acid mediated signaling pathway,” “abscisic-acid activated signaling pathway,” “isoprenoid catabolic process,” and “isoprenoid binding,” with fold enrichment values ranging from 2.5 to 9.8 (Tables 1, 2; Supplementary Table 2).

To identify hormonal pathways altered by SlDLK2 OE, significantly induced or repressed genes (fold change >2 or < −2, respectively; p-value <0.05) upon SlDLK2 overexpression (BioProject PRJNA523214) were submitted to the Plant MetGenMap tool. As shown in Table 3, the abscisic acid (ABA) and indoleacetic acid (IAA) biosynthesis pathways and the gibberellin inactivation pathway were significantly repressed, while the phaseic acid biosynthesis pathway, a process that is related to ABA degradation, was significantly induced. Regarding genes from the brassinosteroid biosynthesis pathway, the expression of both the SlDLK2 OE-induced and SlDLK2-repressed gene sets significantly changed (Table 3).

In a more detailed analysis, we pinpointed tomato genes that are putatively involved in different hormone pathways and we confirmed a clear repression of marker genes related to ABA and IAA biosynthesis, including the AAO3, ABA3, NCED1, NCED2, TAR2a, AMI1, and AAO1 genes. Moreover, we found that ACO1, ACO6, and ACS1 genes related to ethylene production, and OPR3, AOC, AOS, and LOX genes putatively involved in jasmonic acid biosynthesis were also repressed upon SlDLK2 overexpression (left heatmap of Figure 1A). The repression of most of these hormone-related genes upon SlDLK2 overexpression also occurred in mycorrhizal conditions (middle graph of Figure 1A), as shows the RNA-seq data from our experiment with SlDLK2 OE composite plants and control plants transformed with the empty vector, both inoculated with the AM fungus Rhizophagus irregularis (Supplementary Tables 3, 4). More interestingly, a transcriptomic analysis comparing non-inoculated and mycorrhizal tomato roots (BioProject PRJNA509606) showed that the expressions of all these hormone biosynthesis genes that are repressed by SlDLK2 are AM-induced genes (right heatmap of Figure 1A). In summary, our results show that SlDLK2 overexpression triggers a negative regulation of a number of hormone biosynthesis genes, with a particular relevance of genes related to the biosynthesis of carotenoid and isoprenoid-derived plant hormones that are important molecules during mycorrhization such as apocarotenoids, gibberellins, ABA, α-ionols, or SLs, as illustrated in Figure 1B.

Moreover, a deeper analysis considering a wider number of putative tomato homologs of isoprenoid-related genes previously identified by Ezquerro et al. (2023) showed a repression of many genes related to the biosynthesis of other isoprenoid related molecules such as tocopherols and plastoquinone (Supplementary Table 5). Interestingly, the formation of monoterpenes, diterpenes, and tetraterpenes (carotenoids) and the prenyl moieties of chlorophyll, plastoquinone, and tocopherol requires the plastidial isopentenyl diphosphate (IPP) precursor (Pu et al., 2021), and indeed several genes from the methylerythritol phosphate (MEP) pathway responsible for the production of the plastidial IPP precursor are also repressed in SlDLK2 OE roots. By contrast, the expression of genes from the mevalonate (MVA) pathway for the biosynthesis of cytosolic IPP pools which is related to the biosynthesis of other different compounds (sesquiterpene, triterpene, and polyterpene end products) was rather positively affected upon SlDLK2 overexpression (Supplementary Table 5).

Stable tomato transgenic lines expressing p35S::SlDLK2 were obtained using the pK7FWG2::SlDLK2 binary vector and Agrobacterium tumefaciens LBA4404 for transformation. Two transgenic-negative controls (WT-1 and WT-2) and three homozygous T3 overexpressing lines (OE-1, OE-2, and OE-3) were selected on kanamycin-containing medium, and a mycorrhizal experiment with the AM fungus Rhizophagus irregularis was set up. We tested SlDLK2 expression levels on the roots from these lines, in non-inoculated and mycorrhizal plants at 57 and 77 days post-inoculation (dpi). A successful overexpression of the SlDLK2 gene was obtained in the SlDLK2 OE plants in both conditions and at both harvesting points (Figure 2A).

First, we analyzed mycorrhizal colonization and transcriptional activity of isoprenoid biosynthesis-related genes. In agreement with the results obtained in our previous experiments with tomato composite plants overexpressing SlDLK2 (Ho-Plágaro et al., 2021), a decreased mycorrhizal colonization was observed in the SlDLK2 OE lines with respect to the control ones at 57 and 77 dpi (Figure 2B). Second, the expression of the phytoene synthase 3 (SlPSY3) and zaxinone synthase (SlZAS4) genes putatively related with isoprenoid biosynthesis during mycorrhization was induced by mycorrhization and repressed by SlDLK2 OE in the absence of the AM fungi (Figures 2C,D), as expected based on the results of our previous RNA-seq analyses (Figure 1A). Moreover, we observed that SlPSY3 and SlZAS4 expressions were also reduced in the SlDLK2 OE lines in mycorrhizal conditions at 57 and 77 dpi (Figures 2C,D), indicating that the negative effect of SlDLK2 OE on isoprenoid biosynthesis genes is also occurring in AM plants.

To confirm whether strigolactone (SL) contents were reduced upon SlDLK2 overexpression, we performed an experiment where SlDLK2 OE and control plants were subjected to P-starvation conditions to promote SL production. As expected, SLs were not detected in the control treatment with P nutrition. By contrast, under P-starvation conditions, solanacol, which is one of the most abundant SLs in tomato (Kohlen et al., 2013), was detected, showing decreased levels in the SlDLK2 OE roots (Figure 2E). These results confirm that SlDLK2 overexpression causes a reduction of SL contents in roots.

To determine whether transcriptional changes in hormonal-related genes triggered by SlDLK2 overexpression are accompanied by differences in hormonal contents, we selected non-mycorrhizal SlDLK2 OE tomato lines grown for 57 days from the experiment explained above (Figures 2A–D). Mycorrhizal SlDLK2 OE plants were not selected for hormonal content analysis to avoid possible effects due to the lower mycorrhization levels in these plants and not to a direct effect of SlDLK2 overexpression. In the SlDLK2 overexpressing roots, jasmonic acid content was strongly reduced (nearly a 10-fold decrease), and the ABA contents were significantly reduced to half with respect to the control tomato plants (Figures 3A,B). In addition, indoleacetic acid contents were slightly decreased in roots of SlDLK2 OE plants, although this reduction was not significant for all tomato lines (Figure 3C). Not significant alterations were detected in the root content of gibberellins (GA1 and GA4), salicylic acid, and cytokinins (Figures 3D–G). As expected, the increase on the content of these hormones in SlDLK2 OE roots was accompanied by a decreased expression of marker genes involved in the biosynthesis of JA (AOS3), ABA (NCED1), and IAA (TAR2a) (Itoh et al., 2002; Thompson et al., 2004; Mashiguchi et al., 2011), and an increased expression of the GH3.4 gene related to auxin inactivation (Chen et al., 2022) (Figures 3H–K).

Hormone content analysis was also performed on the leaves of the SlDLK2 OE plants. In leaves, DHZ-type cytokinins showed a slight trend toward induction in the SlDLK2 OE leaves, and indoleacetic acid was the only hormone found to be significantly affected by SlDLK2 overexpression, showing increased levels in SlDLK2 OE leaves with respect to the control leaves (Figure 4). Curiously, this result was opposite to the reduced levels of the indoleacetic acid observed in SlDLK2 OE roots.