Plant materials, growth conditions
Seeds of Solanum lycopersicum c.v. Moneymaker was obtained from the C.M. Rick Tomato Genetics Resource Center (TGRC; https://tgrc.ucdavis.edu) at the University of California, Davis. Seeds of XHS072 were obtained from our own stocks. Seeds for all studies were sown in 32-square plug trays filled with a mixture of peat and vermiculite (3:1). Seedlings of parental lines, NILs, recombinant lines and transgenic plants were grown in a growth chamber under the following conditions: 16:8 h light:dark photoperiod, 26/20 °C (day/night) temperature with 300 μmol photons m–2 s–1 irradiance and 60–70% humidity. Thirty-day-old seedlings were transplanted to a greenhouse at the Shunyi experimental station (116°87’ E, 40°17’ N) in Beijing, China. Temperature in the greenhouse was monitored and the conditions were as follows: for the typical (i.e. control) temperature, plants were grown in a greenhouse from March to May, a daily mean temperature of 24 °C with a peak temperature of 27 °C, and for the high temperature (i.e. heat stress), the plants were grown in a greenhouse from May to August, a daily mean temperature of 32 °C with a peak temperature of 40 °C. For a greenhouse at the Nankou experimental station (116°5’ E, 40°13’ N) in Beijing, China. Temperature in the greenhouse was monitored, and the conditions were as follows: for the high temperature (i.e., heat stress), the plants were grown in a greenhouse from August to October, a daily mean temperature of 33.72 °C with a peak temperature of 40.4 °C.
Nicotiana benthamiana seeds were sown in 32-square plug trays filled with a mixture of peat and vermiculite (3:1). Seedlings were grown in a growth chamber under the following conditions: 16:8 h light:dark photoperiod, 26/20 °C (day/night) temperature with 300 μmol photons m–2 s–1 irradiance, 60–70% humidity. Four-week-old seedlings were used for experiments.
Plants sampled for qRT–PCR analysis and auxin dose-response experiments were grown in controlled conditions in growth chambers (Ruihua, AR800) with 12-h light/12-h dark cycle at 26/22°C day/night temperatures, and high-temperature conditions of 12-h light/12-h dark cycle with 35/26 °C day/night temperatures, with 250 μmol photons m–2 s–1 irradiance.
Emasculation of flowers and phenotypic analysis
The artificial emasculation experiment was done as described previously with some modifications13. −2 DPA flowers on mature tomato plants (two to four weeks after the first flowering) were emasculated, and the presence of only pectin and no seeds in the red ripe fruits was considered parthenocarpy. For fruit-bearing ability and parthenocarpy quantification, more than five individual plants of each accession were phenotyped. The fruits of each plant were evaluated without artificial pollination. The percentage of seedless fruit represents the proportion of seedless fruit to the total number of fruits, with 10–40 fruits per plant typically scored. The percentage of fruit set represents the proportion of fruit to the number of flowers. Fruit weight (g) represents the average fruit fresh weight of red ripe fruit. Fruit number represents the total number of fruits per plant and fruit yield (kg) represents the average yield of all fruits fresh weight from more than five plants.
Mapping and identification of TSP4
Non-parthenocarpic Solanum lycopersicum c.v. Moneymaker and the parthenocarpic c.v. XHS072 were used as parental materials. An F2 population was generated by crossing Moneymaker with XHS072 and 320 individual plants were used for bulk-segregant analysis (BSA)23. We selected 31 plants with non-parthenocarpic (seedless fruit rate 0%) and 29 plants with parthenocarpic fruit (seedless fruit rate > 30%) for BSA. Equal amounts of tissue from each plant were pooled for DNA extraction. The genomic DNA was sheared using a Diagenode Bioruptor Plus instrument to obtain ~300 bp fragments. Libraries were prepared using the NEXTflexTM Rapid DNA-Seq Kit for Illumina sequencing (NOVA-5144-08) according to the manufacturer’s protocol. Genomic DNA reads were trimmed by quality using FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/), and paired reads were mapped to the reference tomato genome (SL4.0) using BWA-MEM, with default parameters43. SNP calling was based on alignment results using the Genome Analysis Toolkit GATK 3.1.144.
BSA was performed with modification45. SNPs between two parental genomes were identified for further analysis when the base-quality value was ≥ 20 and the SNP quality value was ≥ 20. On the basis of these criteria and the number of SNPs with a read depth ≥ 5x, an SNP index was calculated for both bulk samples, expressing the proportion of reads harboring SNPs that were identical to those in the parent (Moneymaker). The ∆SNP index was obtained by subtracting the SNP index for the non-parthenocarpic bulk sample from that for the parthenocarpic bulk sample.
Near-isogenic lines (NILs) were generated by selecting heterozygous offspring at the qTSP4 locus for selfing. After six generations, markers to identify plants homozygous for NIL-TSP4MM or NIL-TSP4XHS072 were used to genotype the presence/absence of the introgressed region (860 kb) around the TSP4 locus. On this basis, we generated six alleles with different combinations of TSP4a and TSP4b haplotypes including TSP4aMM4bMM, TSP4aXHS0724bMM, TSP4aMM4bXHS072, TSP4aXHS0724bMM/XHS072, TSP4aMM/XHS0724bXHS072 and TSP4aXHS0724bXHS072. All markers used are listed in Supplementary Data 3.
Plasmid construction and plant transformation
CRISPR–Cas9 mutagenesis was employed to target specific regions within the exons of TSP4a, TSP4b, and SlARF2a, as well as the promoter of TSP4a and the 3’ UTR of TSP4b. The target sites were carefully designed using the CRISPR-P tool (available at http://cbi.hzau.edu.cn/CRISPR/). Vectors were constructed according to the previously described46. Briefly, primers containing the designed sgRNA sequences and BsaI restriction enzyme recognition sites were synthesized. These primers were used to amplify the sgRNAX_U6-26t_SlU6p_sgRNAX fragments from a pCBC_DT1T2_SlU6p vector, after which the PCR fragments were purified and cloned into pTX041 at the BsaI sites47.
For the construction of 35Spro:TSP4a–YFP–HA and 35Spro:TSP4b–YFP–HA, the coding sequences of TSP4a and TSP4b were amplified from Moneymaker cDNA template and fused to the N-terminus of YFP–HA using an In-fusion HD Cloning kit (Vazyme, C113)48. 35Spro:SlARF2aTr–YFP–HA was constructed as described previously with modifications12,15. The SlARF2a coding sequence without the PB1 domain was amplified from the Moneymaker cDNA template and fused to the N-terminus of YFP–HA using an In-fusion HD Cloning kit (Vazyme, C113).
All plasmids were validated by sequencing and then transformed into Agrobacterium tumefaciens AGL149. Wild-type c.v. Moneymaker, XHS072 and NIL-TSP4XHS072 plants were transformed in accordance with A. tumefaciens-mediated cotyledon-explant transformation as described previously47 and selected on 50 mg/L kanamycin. Transgenic plants were validated by PCR and sequencing. All primers used are listed in Supplementary Data 3. tsp4a/4bcr double mutants in the Moneymaker background were obtained by crossing tsp4acr1 and tsp4bcr2 single mutants.
Auxin quantification by UPLC–MS/MS
Auxin was isolated from –2 DPA NIL-TSP4MM and NIL-TSP4XHS072 ovaries grown under control and high-temperature conditions with three biological replicates. Auxin extraction and measurement were as described previously50 with minor modifications. For each sample, 20 mg of tissue was collected into a 2 mL RNA-free centrifuge tube and ground in liquid nitrogen. Empty tubes and tubes plus samples were weighed, and then 1.8 mL chromatography-grade MeOH and 10 μg/mL D-IAA (dissolved in MeOH) were added for use as internal standards. The sample was extracted overnight at −20 °C and then centrifuged for 15 min at 13,400 g at 4 °C. The supernatant was then collected in a new 2 mL tube and evaporated under nitrogen gas. The residue was dissolved in 1 mL chromatography-grade ammonia solution (5%, v/v). Samples were loaded onto an Oasis MAX SPE cartridge, which was sequentially washed with 4 mL chromatography-grade ammonia solution (5%, v/v), 4 mL water then 4 mL MeOH. 4 mL of MeOH containing 10% (v/v) formic acid was used to elute samples, and eluted samples were dried. The dried residue was reconstituted in 200 μL 80% (v/v) MeOH followed by centrifugation for 30 min at 4 °C at 13,000 rpm before UPLC–MS/MS analysis.
Auxin contents were measured using Waters UPLC–MS/MS (ACQUITY UPLC I-Class-Xevo TQ-S Micro, Waters). A Waters CSH C18 column was used as the analytical column (2.1 mm × 100 mm i.d., 1.7 μm). The mobile phase consisted of 0.1% formic acid (FA) and acetonitrile (ACN) as solvent A and 0.1% FA and water as solvent B. The temperatures of the column and autosampler were 35 °C and 4 °C, respectively. Run initially with 20% A, then increase to 73% A in 1 min and increase to 78% A in the next 3 min, decrease to 20% in 1 min, keep 20% A for 2 min (total 7 min) under a flow rate of 0.2 mL/min. The content was calculated according to the internal standard method. Data analysis was performed using MassLynx V4.1 (Waters).
Quantification of ethylene production
Ethylene was isolated from Moneymaker and gene knockout mutants ovaries at –2 DPA, with three biological replicates for each group. Ethylene extraction and measurement were as described previously with minor modifications41. Ethylene production rate was measured by gas chromatography–mass spectrometry (Agilent, 8890-7000E) equipped with a 30 m × 0.25 mm × 0.25 µm capillary column (HP-5ms Ultra Inert). We collected five to eight ovaries of tomatoes, weighed and enclosed them into a 2-mL glass vial, which was incubated at 25 °C for 4 h under darkness. After incubation, 10 µL of gas from the headspace of the 2 mL glass vial was injected into the gas chromatography for analysis. The temperature of injector port was 250 °C. The temperature of the column kept at 40 °C for 1 min, ramped to 100 °C at 20 °C/min, and kept at 100 °C for 1 min. The carrier gas was helium with the pressure of 12.942 psi. The injection mode was in splitless. The content was calculated according to the external standard method. Data analysis was performed using MassHunter Quantitative Analysis 12.0 (Agilent).
Fructose and glucose quantification by UPLC–-MS/MS
Fructose and glucose were isolated from more than five red ripe fruits from NILs and gene-edited lines grown under high-temperature conditions with three biological replicates. Sample pretreatment and measurement were performed as described previously51. Tomato samples (0.1 g) were extracted in 1.9 mL arabinose solution (1 mg/mL) by sonication (500 W, 10 min), followed by centrifugation (13,000 g, 10 min). The supernatant was double-filtered (0.22 μm PES) and mixed 1:1 with acetonitrile before analysis. Fructose and glucose contents were measured by UPLC–MS/MS (ACQUITY UPLC I-Class-Xevo TQ-S Micro, Waters). For saccharide analysis, an ACQUITY UPLC BEH Amide 1.7 µm column was used as the analytical column (2.1 x 100 mm; Waters). Sugar content was calculated according to the internal standard method. Data analysis was performed using MassLynx V4.1 (Waters).
In situ RNA hybridization
In situ RNA hybridization was performed as described with some modifications52,53. All procedures were conducted under stringent RNase-free conditions. The cDNA segments of TSP4a and TSP4b were amplified using primers P18 and P19 (Supplementary Data 3) and subsequently cloned into pEAZY-T3 (TransGen, CT301-01), which contains T7 promoter sequences. In vitro transcription was then carried out using T7 RNA polymerase to generate DIG-labeled antisense and sense probes for in situ hybridization. Ovary tissues were dissected and placed in freshly prepared 4% (w/v) paraformaldehyde in phosphate-buffered saline (PBS, pH 7.2), followed by vacuum infiltration (0.06 MPa) for 30 min. The tissues were then fixed at 4 °C for 16 h. After fixation, the tissues were dehydrated through a graded ethanol series (30–100%, 90 min per step), followed by dehydration in an ethanol-HistoChoice (Sigma, H2779) series. The tissues were then embedded in Paraplast Plus (Leica, P39601095). Following embedding, 10 μm sections were cut, and hybridization with the probes was performed at 55 °C in the dark for 24-48 h. After hybridization, NBT/BCIP (SIGMAFAST™) was used for color development. Sections were rapidly dehydrated using a graded ethanol series (30–100%, 3 seconds per step) and mounted in 50% glycerol for microscopic analysis using an Olympus BX43 microscope.
Phylogenetic analyses
The amino-acid sequences of Arabidopsis and tomato homologs of TSP4a and TSP4b were aligned by MEGAX software. Phylogenetic trees were constructed with the aligned protein sequences using MEGAX software with the neighbor-joining method. Bootstrap values were derived from 1,000 replicates.
Haplotype analyses
SNP calling was based on alignment results using the Genome Analysis Toolkit gatk 3.1.144. SNPs between Moneymaker and XHS072 genomes were identified, wherein the base quality value was ≥ 20 and the SNP quality value was ≥ 20. Haplotype analysis was performed on the 8-kb promoter region of TSP4a and the 2.1-Kb 3’ UTR of TSP4b.
Heat stress and IAA application
Heat treatment was carried out according to a previously published method54 with modifications. To assay TSP4a and TSP4b gene expression in plants grown at 25 °C and 35 °C, seedlings were grown to the 4-leaf stage at 25 °C under 12-h light/12-h darkness and transferred to 25 °C and constant illumination for 5 d. Afterward, seedlings were transferred to 35 °C with constant illumination. Leaf samples were collected at the indicated time points for analysis.
IAA treatment was carried out according to a previously published method55 with minor modifications. Plants were treated with either a mock treatment or 10 μM IAA and 12 d seedlings were incubated on 1/2 MS medium under 25 °C and 12-h light/12-h dark in a growth chamber. Seedlings were soaked in liquid 1/2 MS medium with or without 10 μM IAA. Samples were collected and analyzed at the indicated time points.
RNA extraction and RT–qPCR analysis
Total RNA was extracted using TRNzol Universal reagent (Tiangen, DP424). Reverse transcription was performed with TransScript® II One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen, AT311-02), using 2 μg of total RNA in 20 μL reactions under the following conditions, 42 °C for 60 min, 85 °C for 10 s. Real-time quantitative PCR (RT–qPCR) was performed with 10 μl of 2x Taq Pro Universal SYBR qPCR Master Mix (Vazyme, Q712-02) and 5 μl cDNA on a Bio-Rad CFX-96 Real-Time PCR instrument with the following program: 3 min at 95 °C followed by 40 cycles of 20 s at 95 °C, 30 s at 60 °C, and 20 s at 72 °C. UBI3 (Solyc01g056940) was used as the reference gene for normalize relative expression54. All primers used for RT–qPCR are listed in Supplementary Data 3.
RNA-seq analysis
Total RNA was isolated from –2 DPA NIL-TSP4MM and NIL-TSP4XHS072 ovaries from plants grown under normal and high-temperature conditions in a growth chamber. Three biological replicates were analyzed per line, with each replicate containing at least 15 ovaries from 5 individual plants.
Twelve RNA-seq libraries were constructed and sequenced using Illumina HiSeq2000 at Berry Genomics (http://www.berrygenomics.com/). Filtered clean reads were aligned to the tomato (Solanum lycopersicum) cv. Heinz 1706 genome (ITAG 4.0) using STAR version 2.5.3, and their features were counted by feature Counts v 1.5.356. Principal-component analysis was performed to compare the fragments per kilobase of transcript sequence per million base pairs sequenced (FPKM) values of the expressed gene using the prcomp function in R (R Foundation, Austria). Using the R package DEGseq version 3.0.357, the MA-plot-based method was used to calculate P-values that were adjusted using the Benjamini–Hochberg procedure. Gene-expression fold change between the two genotypes was calculated as FPKM. The thresholds for identification of DEGs were: FPKM > 1 in any tissue, fold change ≥ 4 or ≤ 0.25, and Benjamini–Hochberg adjusted P < 0.001. GO analyses of DEGs were performed using the best match homologous genes in Arabidopsis with DAVID (The Database for Annotation, Visualization, and Integrated Discovery, https://david.ncifcrf. gov).
Subcellular localization
Leaves of 40-d-old 35Spro:TSP4a–YFP–HA and 35Spro:TSP4b–YFP–HA stable-transgenic plants were used for subcellular-localization analysis. The localization of the fluorescent fusion proteins was analyzed by confocal laser-scanning microscopy (Leica DM6CS). Nuclei were detected by 10 μg/ml 4′,6-diamidino-2-phenylindole (DAPI) staining in the dark for 30 s. YFP images were captured with laser excitation at 514 nm and an emission wavelength range from 525–570 nm, and a gain value of around 600. DAPI images were captured with laser excitation at 359 nm and an emission wavelength range from 430–480 nm, and a gain value of around 600.
Auxin dose-response experiments
Auxin dose-response experiments were conducted as described previously10 with minor modifications. Cotyledon explants of 9-d-old seedlings were cultivated in a growth chamber on MS medium containing different NAA concentrations (0 nM, 10 nM, 20 nM, 40 nM and 80 nM) at 25 °C and 35 °C for 10 d. The percentage of rooting plants represents the proportion of the number of rooted cotyledon explants to the total number of cotyledon explants after a 10-d cultivation period.
Hypocotyl fragments from 5-d-old seedlings were isolated at about 1 cm long and immediately moved into MES/sucrose buffer (5 mM MES/KOH and 1% [w/v] sucrose, pH 6.0). After 1 h pre-incubation, hypocotyl segments were randomly distributed to fresh buffer solutions with different NAA concentrations (0.01 μM, 1 μM, 10 μM, 100 μM) with gentle agitation at 25 °C and 35 °C. Elongation of hypocotyl length was measured after 23 h of treatment. The percentage elongation rate represents the proportion of the increase in final length over the initial length after 23 h of incubation.
Prediction of transcription-factor binding sites
The transcription-factor binding sites on the promoters were predicted in the JASPAR database 202458. The promoter and 3’ UTR were searched for the binding sites of transcription factors using the ‘Scan’ function in the database. The relative profile score threshold > 80% was identified as the candidate site.
ChIP–qPCR
Chromatin immunoprecipitation (ChIP) experiment was performed as described with minor modifications54. –2 DPA ovaries-tissue samples (2 g) from NIL-TSP4XHS072/35Spro:TSP4b–YFP–HA and MM/35Spro:SlARF2aTr–YFP–HA transgenic plants were immediately frozen in liquid nitrogen and ground into a fine powder under liquid nitrogen. The powdered tissue was then subjected to cross-linkage in 1% (w/v) formaldehyde (Sigma-Aldrich, 104003) for 10 min at 4 °C, followed by the addition of 0.1 M glycine, which was infiltrated for 5 min at 4 °C. Chromatin was sheared using a Diagenode Bioruptor Plus instrument to obtain ~300-bp DNA fragments. Immunoprecipitation was performed using 5 µg anti-HA antibody (Sigma, H6908). Input as a mock immunoprecipitation (mock-IP) without the antibody was used as a negative control. DNA isolated from the ChIP was used for subsequent qPCR analysis.
For ChIP–qPCR, 10 μL 2x Taq Pro Universal SYBR qPCR Master Mix (Vazyme, Q712-02) and 5 μL DNA were used on a Bio-Rad CFX-96 Real-Time PCR instrument with the following program: 3 min at 95 °C followed by 50 cycles of 20 s at 95 °C, 30 s at 60 °C, and 20 s at 72 °C. The ACTIN (Solyc03g078400) 3’ intergenic region was used as an internal control. Input DNA was used as a negative control. All primers used for qPCR are listed in Supplementary Data 3.
Dual-luciferase reporter assay
For plasmid construction, the TSP4apro:LUC-35Spro:REN, EXP5pro:LUC-35Spro:REN and ACO4pro:LUC-35Spro:REN backbone vectors were obtained from pPZP21159. The 6.8-kb TSP4a, 4-kb EXP5 and ACO4 promoter sequences were amplified using Moneymaker genomic DNA as template and integrated into LUC-35Spro:REN using an In-fusion HD Cloning kit (Vazyme, C113). The 2.1 kb TSP4b 3’ UTR was amplified using Moneymaker genomic DNA as a template and constructed into the 35Sminpro:LUC-35Spro:REN vector to obtain the 35Sminpro:LUC-TSP4butr-35Spro vector. The 35Spro:TSP4a–FLAG, 35Spro:TSP4b–FLAG, 35Spro:SlARF2a–FLAG and 35Spro:SlARF2aTr–FLAG (a truncated variant of ARF2a lacking the PB1 domain) vectors were constructed using the full-length or truncated coding sequence of TSP4a, TSP4b and SlARF2a, respectively. All sequences were amplified from Moneymaker cDNA as a template and cloned into the pCAMBIA2300-p35S:FLAG vector48. All plasmids were validated by Sanger sequencing and transformed into A. tumefaciens EHA10549. A single colony was cultured in LB medium with appropriate antibiotic selection until OD600 nm = 1.0. Agrobacteria were collected by centrifugation (3214 x g, 10 min, 20 °C) and re-suspended in 10 mM MgCl2 and 150 μM acetosyringone until the OD600 nm = 1. Cells containing the effector plasmids, luciferase reporter plasmid, and P19 plasmid were mixed in a ratio of 2:1:3 (v/v/v) and infiltrated into N. benthamiana leaves using a syringe. The leaves were harvested and ground in liquid nitrogen 36 h after infiltration.
The 1 mM luciferin substrate was spread evenly on the underside of the leaves and allowed to dry for 5 min in the dark. Subsequently, luciferase activity signals were detected with the plant in vivo imaging system (LB985 NightSHADE, Germany). Firefly luciferase (LUC) and Renilla (REN) activities were measured using a dual-luciferase reporter assay system (Promega; E1910) on a Promega GLOMAX 20/20 luminometer. REN activities were used as an internal control. Primers used to generate the constructs are listed in Supplementary Data 3.
Yeast two-hybrid
Yeast two-hybrid assays were conducted following the Yeastmaker™ Yeast Transformation System 2 user manual (Clontech, PT1172-1). To explore interactions between TSP4a and SlARF2a, the full-length coding sequence of TSP4a was cloned into bait vector pGBKT7, while the full-length and truncated coding sequences of SlARF2a were cloned into prey vectors pGADT7. The bait and prey plasmids were then cotransformed into the Y2H Gold yeast strain, following the protocol outlined in the Clontech yeast handbook instructions. The resulting transformants were grown on selective plates lacking leucine and tryptophan for 3 d at 30 °C. The interactions were evaluated by growth assays on medium lacking leucine, tryptophan, histidine, and adenine but supplemented with 30 mM 3-Amino-1,2,4-triazole (3-AT), and the growth was monitored for 3 to 5 days.
Bimolecular fluorescence complementation (BiFC) assays
The full-length coding sequences of TSP4a, along with the full-length coding sequence of SlARF2a, or truncated coding sequence of SlARF2aTr and SlARF2aPB1, were amplified and subsequently inserted into the pEarleyGate 201-YN or pEarleyGate 202-YC vectors (using PacI and SpeI restriction enzymes)60, utilizing the In-Fusion HD Cloning Kit (Vazyme, C113). These vectors were subsequently transformed into the A. tumefaciens EHA105, and TSP4a fusion proteins were mixed with intact and truncated SlARF2a fusion proteins in an equal volume of culture and injected into the leaves of 4-week-old N. benthamiana plants. Two days after infiltration, YFP fluorescence was observed using a Leica DM6 CS confocal laser-scanning microscope. Images were captured with a laser excitation wavelength of 514 nm, and collection bandwidth was set between 525–570 nm, with intensity values around 10 and gain value of around 600. All primers used to generate the constructs are listed in Supplementary Data 3.
Statistical analyses
Two-tailed Student’s t-test and two-tailed Mann–Whitney–U test were used to determine statistical significance using GraphPad Prism 8. For comparisons of multiple groups, the data were analyzed by one-way and two-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test using GraphPad Prism 8.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
