A feedback loop at the THERMOSENSITIVE PARTHENOCARPY 4 locus controls tomato fruit set under heat stress

Lu, Xiaonan; Wu, Jianxin; Shi, QianQian; Sun, Shuai; Cheng, Yuan; Zhou, Guozhi; Li, Ren; Wang, Huanzhong; van der Knaap, Esther; Cui, Xia

doi:10.1038/s41467-025-59522-7

Download PDF

Article
Open access
Published: 06 May 2025

A feedback loop at the THERMOSENSITIVE PARTHENOCARPY 4 locus controls tomato fruit set under heat stress

Nature Communications volume 16, Article number: 4184 (2025) Cite this article

4635 Accesses
1 Citations
10 Altmetric
Metrics details

Subjects

Abstract

High temperatures compromise crop productivity worldwide, but breeding bottlenecks slow the delivery of climate-resilient crops. By investigating tomato fruit set under high temperatures, we discover a module comprising two linked genes, THERMOSENSITIVE PARTHENOCARPY 4a (TSP4a) and TSP4b, which encode the transcriptional regulators IAA9 and AINTEGUMENTA (ANT), respectively, to control thermosensitive parthenocarpy. TSP4a and TSP4b form a positive feedback loop upon heat stress to repress auxin signaling in ovaries. Natural TSP4a and TSP4b alleles bear regulatory-region polymorphisms and are differentially expressed to overcome the trade-off between fruit set and wider plant development. Gene editing of the TSP4a promoter and TSP4b 3’ UTR in open-chromatin regions results in expression down-regulation, increased parthenocarpy without yield penalties and maintenance of fruit-sugar levels without broad auxin-related pleiotropic defects in greenhouse-grown plants. These mechanistic insights into heat-induced parthenocarpy and auxin signaling in reproductive organs demonstrate breeding utility to safeguard tomato yield under warming scenarios.

Application of 4-CPA or ethanol enhances plant growth and fruit quality of phyA mutant under heat stress

Article Open access 12 September 2025

The temperature sensor TWA1 is required for thermotolerance in Arabidopsis

Article Open access 15 May 2024

Warm temperature triggers JOX and ST2A-mediated jasmonate catabolism to promote plant growth

Article Open access 10 August 2021

Introduction

Climate-change-associated extreme high temperature is a serious threat to global crop production, with every 1 °C increase in temperature, global production of major crops is predicted to decline by 3.1–7.4%¹. Heat stress seriously damages pollen viability and female fertility, thereby heavily penalizing fruit set in tomato^2,3,4, the global market of which grew to $US17.865 billion in 2023⁵. In Florida, United States, which produces the most fresh tomatoes, heat stress causes decreases in production of 52–85%⁶. Therefore, fruit crops with a strong fruit-bearing ability are critical for maintaining consistent yields in variable environmental conditions.

Parthenocarpy is fruit production occurring without fertilization. It offers the possibility of improving fruit set when environmental conditions are adverse for pollen production and fertilization; moreover, the absence of seed is favored by consumers and can improve fruit quality in certain crops (e.g., banana, eggplant, persimmon, grape, and watermelon)⁷. Auxin is the first plant hormone known to induce parthenocarpy⁸. Exogenous application of auxin to unfertilized ovaries is widely used in commercial vegetable production to increase fruit set, but is laborious and often compromises fruit-quality characters⁷. Elevated auxin signaling induces parthenocarpy in fruit crops^7,9. The auxin response protein SlAUX/INDOLE-3-ACETIC-ACID 9 (SlIAA9) is a major repressor of fruit initiation, and its tomato null mutant, entire, shows strong parthenocarpy^10,11. Loss of certain AUXIN RESPONSE FACTOR (ARF) transcription factors also enhances parthenocarpy in tomato^{12,13,14,15,16}. Unfortunately, higher auxin activity invariably results in pleiotropic phenotypes, such as changes to plant and fruit morphology, which prevents widespread deployment in breeding programs^10,15,17,18.

A facultative parthenocarpic tomato line suitable for use in warm climates was developed in 1937¹⁹. Since then, only the HOMEOBOX-LEUCINE ZIPPER PROTEIN 15 A (HB15A) and AGAMOUS-LIKE 6 (AGL6) alleles of historical parthenocarpic mutants pat/pat-1 and pat-K respectively have been identified^20,21,22. However, abnormal ovule development in these mutants results in few seeds that prohibit their exploitation in tomato breeding programs. It is therefore critical to identify other genetic resources that endow parthenocarpic fruit and viable seed, while also overcoming trade-offs between fruit set and plant development. Here, we discovered a locus comprising two genes, THERMOSENSITIVE PARTHENOCARPY 4a (TSP4a) and TSP4b that together form a positive feedback loop to regulate thermosensitive parthenocarpy in tomato. Natural TSP4a and TSP4b alleles function in auxin signaling by fine-tuning each other’s expression levels to overcome the trade-off between fruit set and plant development, specifically under high temperature. Viable seeds produced at control temperatures and strong fruit-bearing ability without loss of desirable fruit sugars during heat stress in the greenhouse confirm their value for breeding climate-resilient crops to cope with changing weather patterns.

Results

TSP4 confers thermosensitive parthenocarpy

Heat stress often causes a dramatic reduction in fruit set in tomato because of pollen sensitivity and impaired fertilization^2,3,4. To identify tomato accessions with sufficient fruit set at high temperature, we evaluated fruit set in 102 cultivated inbred tomato accessions collected from different countries under control-temperature conditions (daily mean temperature: 24 °C, peak temperature: 27 °C) and conditions simulating heat stress (daily mean temperature: 32 °C, peak temperature: 40 °C) near Beijing in 2021. Among them, accession XHS072 had robust fruit set under both conditions. By contrast, fruit set for Moneymaker (MM) was reduced by half under heat-stress conditions from that under control temperatures to about 22% (Fig. 1a, b). Approximately 70% of XHS072 fruits were seedless under heat stress, while seeded fruits were uniformly produced in Moneymaker, implying that most fruits produced by XHS072 were due to pollination-independent fruit set (Fig. 1c). Consistent with this, unpollinated XHS072 ovaries developed into normal-sized seedless fruits only under heat stress compared to unpollinated Moneymaker ovaries, further supporting that parthenocarpic fruits were induced by heat stress in XHS072 (Fig. 1d).

**Fig. 1: The *TSP4* locus regulates thermosensitive parthenocarpy in tomato.**

Toward identifying the locus underlying these phenotypes, we generated an F₂ population by crossing Moneymaker and XHS072 and evaluated parthenocarpy and seedless fruit set under heat stress. Bulked-segregant analysis coupled with whole-genome sequencing (BSA-seq) was performed using two pools with either low or high rates of seedless fruit²³. A statistically significant locus on the long arm of chromosome 4 was identified that we named THERMOSENSITIVE PARTHENOCARPY 4 (TSP4, Fig. 1e). We generated a pair of near-isogenic lines (NILs) to evaluate its contribution to fruit set under heat-stress conditions, named NIL-TSP4^XHS072 and NIL-TSP4^MM that differ by approximately 860 kb at this locus. More seedless fruits were produced by NIL-TSP4^XHS072 plants than the NIL-TSP4^MM only under heat stress (Fig. 1f). This result supports the hypothesis that TSP4 confers thermosensitive parthenocarpy in tomato.

To explore the reason behind the relatively large proportion of unfertilized NIL-TSP4^XHS072 fruits and precocious fruit set prior to anthesis under heat stress, we examined ovary development. Unpollinated ovaries had expanded at –2 days post anthesis (DPA) under heat stress (Supplementary Fig. 1a, b). Such precocious ovary enlargement positioned the stigma out of reach of the stamen, thereby reducing self-pollination and contributing to the development of seedless fruit (Supplementary Fig. 1c). The enlarged NIL-TSP4^XHS072 ovaries were reminiscent of floral auxin application, which is involved in parthenocarpic fruit set^9,24. We therefore quantified indole-3-acetic acid contents in –2 DPA NIL-TSP4^MM and NIL-TSP4^XHS072 by HPLC–MS/MS and saw elevation in ovaries isolated from plants grown under heat stress for both NILs, whereas no obvious differences were found at the control temperature (Supplementary Fig. 1d). These results implied that the elevated auxin contents observed in ovaries upon heat stress are not affected by TSP4.

TSP4a and TSP4b respectively encode SlIAA9 and SlANT transcriptional regulators

To clone the TSP4 gene, we conducted a high-resolution linkage analysis with 4,416 individuals of heterozygous NILs. Unexpectedly, recombinants with the same genotype displayed different parthenocarpic phenotypes. We analyzed the phenotypes of different recombinants and considered that two genes, designated TSP4a and TSP4b, instead of one, were potentially located at this locus. By comparing parthenocarpy between the offspring of NIL-TSP4a^MM/XHS072 TSP4b^XHS072, we delimited the TSP4a locus to a 30-kb region, comprising only one gene Solyc04g076850 encoding SlIAA9 (Fig. 1g and Supplementary Fig. 2a). According to differential parthenocarpy of NIL-TSP4a^XHS072 TSP4b^MM/XHS072 offspring, TSP4b was mapped to a 54-kb region, in which Solyc04g077470, Solyc04g077480 and the 3’ UTR of Solyc04g077490 were located (Fig. 1g). Among them, only Solyc04g077490, encoding a putative ortholog of SlAINTEGUMENTA (SlANT, Supplementary Fig. 2b), was differentially expressed in the –2 DPA ovaries of NILs-TSP4 under high temperature (Fig. 1g, Supplementary Fig. 3), implying that this gene was the most likely the candidate for TSP4b.

Toward confirming whether SlIAA9 and SlANT correspond to TSP4a and TSP4b, we carried out in situ hybridization to determine the expression patterns on transverse sections. Histological analysis of hybridization signals in reproductive tissues showed that TSP4a transcript had significant accumulation throughout the ovary, whereas TSP4b was mainly accumulated in the ovule and placenta. These results implied functionality for these two genes during fruit set (Supplementary Fig. 4).

We next compared Moneymaker and XHS072 genome sequences and found a single-nucleotide polymorphism (SNP) in the SlIAA9 coding region, which causes an amino-acid substitution from isoleucine to serine at position 94 (I94S, Supplementary Fig. 5a). However, this substitution is potentially inconsequential for protein function because it is a conservative substitution. In addition, 4 2 SNPs and 7 indels in the SlIAA9 promoter, and 12 SNPs and 4 indels in the SlANT 3’ UTR were found (Supplementary Fig. 5b). SlIAA9 and SlANT expression levels were statistically significantly reduced in NIL-TSP4^XHS072 compared to NIL-TSP4^MM from –2 DPA to 6 DPA under heat stress, but not at control temperatures (Fig. 1h, Supplementary Fig. 6a), suggesting that SlIAA9 and SlANT are the candidate genes for TSP4a and TSP4b. Similar expression patterns in leaves also confirmed that regulatory changes in the promoter and 3’ UTR respectively account for the variation in their expression levels (Supplementary Fig. 6b).

To verify the functions of TSP4a and TSP4b, we generated knock-out mutants in the non-parthenocarpic parent Moneymaker by CRISPR–Cas9 (Supplementary Fig. 7a, b). The recovered tsp4acr mutants produced 51% seedless fruits under control temperatures (Fig. 1i) similar to other tomato SlIAA9 mutants and knock-down lines^10,25. When the tsp4acr plants were grown under heat-stress conditions, seedless fruit increased to 87%. tsp4bcr plants instead produced 16% seedless fruits under control temperatures and 32% seedless fruits under heat stress. When TSP4b was knocked out in the parthenocarpic XHS072 background, unpollinated ovaries could develop into seedless fruits at control temperatures (Supplementary Fig. 7c, d). The tsp4bcr3 and tsp4bcr4 mutants bore 70% seedless fruits at control temperatures and completely (i.e. 100%) seedless fruits upon heat stress, greater than XHS072 (Supplementary Fig. 7e). Additionally, tsp4a/4bcr double mutants in the Moneymaker background displayed even stronger parthenocarpy ( ~ 100%) under both control and heat-stress conditions (Fig. 1i). These data revealed that XHS072 is a weak allele, in which the functions of TSP4a and TSP4b are not completely lost, and TSP4a and TSP4b play major roles in the regulation of tomato parthenocarpy.

Natural TSP4a and TSP4b alleles improve tomato yield at high temperatures

To examine whether allelic variation in TSP4a and TSP4b affects tomato yield at high temperature, we generated a series of alleles by genetic crossing with different combinations of TSP4a and TSP4b haplotypes. Remarkably, only the combinations of TSP4a^XHS072/4b^MM/XHS072, TSP4a^MM/XHS072/4b^XHS072 and TSP4a^XHS072/4b^XHS072 lines produced seedless fruits under heat stress compared to the TSP4a^MM/4b^MM lines (Fig. 2a). Consistent with this finding, fruit numbers for these haplotypes were increased by 44%, 26% and 81% respectively, under heat stress (Fig. 2b) without appreciably affecting fruit fresh weight (Fig. 2c). Other plant and leaf morphological characteristics were largely comparable compared to the TSP4a^MM/4b^MM lines, although alleles carrying TSP4b^XHS072 flowered earlier (Fig. 2d and Supplementary Fig. 8a). The increase in fruit set and unchanged fruit mass dramatically improved tomato yields, with a 67% and 76% yield increase observed for TSP4a^XHS072/4b^XHS072 under heat stress when grown in Beijing, Nankou in 2023 and Shunyi (both near Beijing) in 2022, respectively (Fig. 2e and Supplementary Fig. 8b). Notably, there was no penalty on fruit sugar contents (an important sensory trait for consumers), with higher fructose and glucose contents seen in red ripe fruits of TSP4a^XHS072 lines (Supplementary Fig. 8c). These results demonstrated that natural alleles of TSP4a and TSP4b increase tomato yield without major negative effects under heat stress.

**Fig. 2: Natural *TSP4a* and *TSP4b* alleles enhance yield under heat stress.**

Different from natural TSP4a and TSP4b alleles, tsp4acr mutants revert from compound to simple leaves (Fig. 2f and Supplementary Fig. 8d), mimicking entire mutants with a shorter stature^10,11. Compared with Moneymaker controls, tsp4acr and tsp4bcr mutants produced smaller fruits and fruit weights were reduced to 35% and 30%, respectively (Fig. 2f, g). Unlike ANT which regulates shoot, flower and ovule development in Arabidopsis^26,27, the tomato tsp4bcr mutants in the Moneymaker and XHS072 backgrounds only exhibit shorter stature (Supplementary Fig. 8d–f). Notably, shorter stature, simple leaves, and smaller and abnormal fruits were observed in the tsp4a/4bcr double mutants (Fig. 2f, g and Supplementary Fig. 8d, g), indicating the genetic interaction between TSP4a and TSP4b in facilitating fruit enlargement and leaf development. Similar to TSP4b^XHS072 alleles, only the tsp4bcr single mutants and tsp4a/4bcr double mutants displayed early flowering, suggesting that only TSP4b contributes to repressing the floral transition (Supplementary Fig. 8h, i). Although the tsp4acr and tsp4a/4bcr mutants produced approximately twice as many fruits as Moneymaker controls under heat stress, non-significant yield increases were observed for tsp4acr mutants and tsp4a/4bcr double mutant under heat stress in the greenhouse (Fig. 2h, i and Supplementary Fig. 8j). These results suggested knock-out mutations in TSP4a and TSP4b led to a reduction in fruit weight and pleiotropic phenotypes that limit yield improvement.

TSP4a and TSP4b comprise a positive feedback regulatory loop to regulate their expression

Certain TSP4a and TSP4b haplotype combinations bear almost no seedless fruits under heat stress (see Results section for Fig. 2a). To further explore the molecular basis of this phenotype, we compared TSP4a and TSP4b expression levels among different haplotype combinations under heat stress. Although TSP4a and TSP4b expression was markedly reduced in –2 DPA ovaries isolated from TSP4a^XHS072/4b^MM and TSP4a^MM/4b^XHS072 plants compared to TSP4a^MM/4b^MM, the lowest expression levels were detected in TSP4a^XHS072/4b^XHS072 ovaries (Fig. 3a), implying that down-regulation of TSP4a and TSP4b is associated with the parthenocarpic phenotype. Similar down-regulation was seen in the gene-edited mutants. These data imply that the expression of TSP4a and TSP4b functions in a positive regulatory loop (Fig. 3b).

**Fig. 3: *TSP4a* and *TSP4b* comprise a positive feedback loop activated in response to heat stress and auxin cues.**

Since high temperature enhances parthenocarpic fruit set for NIL-TSP4^XHS072 plants, we explored whether TSP4a and TSP4b expression is responsive to heat stress. Upon exposure of Moneymaker plants to 35 °C for 8 h, both genes were rapidly induced (Fig. 3c). However, the heat-stress-induced transcription of TSP4a in the tsp4bcr mutants and TSP4b in the tsp4acr mutants was dramatically compromised (Fig. 3d), implying that both TSP4a and TSP4b are required for their induction upon heat stress. TSP4a and TSP4b expression was also induced within 1 h of auxin supplementation (Fig. 3e), similar to other AUX/IAA genes²⁸. This induction was also compromised in the tsp4bcr and tsp4acr mutants, respectively (Fig. 3f), indicating that TSP4a and TSP4b are both necessary for auxin-dependent induction. Taken together, these data support a model whereby TSP4a and TSP4b are required for increasing the expression of each other in response to auxin and heat stress.

SlARF2a–TSP4a and TSP4b directly activate the transcription of TSP4b and TSP4a

To explore how TSP4a and TSP4b regulate each other’s expression, we first analyzed TSP4a promoter and TSP4b 3’ UTR sequences. Based on the conserved motif [gCAC(A/G)N(A/T)TcCC(a/g)ANG(c/t)] bound by the transcription factor ANT²⁹, several binding sites on the TSP4a promoter were identified. Chromatin immunoprecipitation followed by quantitative polymerase chain reaction (ChIP–qPCR) was used to test whether TSP4b directly binds to these sequences in 35S_pro:TSP4b–YFP–HA transgenic plants, which could partially rescue parthenocarpic phenotypes of NIL-TSP4^XHS072 (Supplementary Fig. 9). Enrichment of TSP4b–YFP–HA on the P2 and P3 regions indicates that TSP4b binds to the TSP4a promoter (Fig. 4a). Dual luciferase reporter assays revealed that TSP4b–FLAG activates TSP4a transcription (Fig. 4b). All these results indicate that TSP4b bind to the TSP4a promoter to activate its expression.

Fig. 4: SlARF2a–TSP4a and TSP4b directly regulate the transcriptions of *TSP4b* and *TSP4a.*

SlIAA9 is an AUX/IAA protein that participates in regulating the auxin signaling pathway through its physical interaction with AUXIN RESPONSE FACTORS (ARFs). When SlIAA9 interacts with ARFs, it inhibits the transcriptional regulatory function of ARFs³⁰. We first analyzed sequence variation of the TSP4b 3’ UTR between XHS072 and MM and found that only the binding sites for ARF2 constitute sequence differences. We therefore investigated the physical interaction between TSP4a and SlARF2a or SlARF2b. Yeast two-hybrid assay showed that SlARF2a could interact with TSP4a through its PB1 (Phox and Bem1) domain (Fig. 4c and Supplementary Fig. 10a, b). Bimolecular fluorescence complementation (BiFC) assays revealed this interaction occurs in the nucleus (Fig. 4d and Supplementary Fig. 10c). Overexpression of SlARF2a^Tr–YFP–HA (a truncated variant of ARF2a lacking the PB1 domain) in the MM background demonstrated that SlARF2aOE^Tr-L1 and SlARF2a^TrOE-L2 lines exhibited parthenocarpic phenotype, similar to that observed in tsp4bcr mutants. This result suggests that SlARF2a participates in regulating parthenocarpy (Fig. 4e).

To determine whether SlARF2a could bind to the TSP4b 3’ UTR, we performed ChIP–qPCR using SlARF2a^TrOE-L1 plants. A statistically significant enrichment of SlARF2a^Tr–YFP–HA was observed at P3, P4 and P6 regions, but not at other assayed ARF2a-recognized sites, indicating that SlARF2a could bind to the TSP4b 3’ UTR (Fig. 4f). We also found that TSP4b expression was dramatically reduced in SlARF2a^TrOE plants (Fig. 4g). Moreover, when SlARF2a–FLAG was co-expressed with 35S_pro:LUC:TSP4b_UTR in N. benthamiana leaves, chemiluminescence intensity was obviously reduced compared with the empty vector control, indicating that SlARF2a repress TSP4b transcription. Fluorescence intensity was restored by adding TSP4a-FLAG in the SlARF2a-FLAG/35S_pro:LUC:TSP4b_UTR co-expression system, whereas co-expression with the truncated SlARF2a^Tr-FLAG/35S_pro:LUC:TSP4b_UTR failed to rescue this fluorescence signal (Fig. 4h). These results suggest that TSP4a interacts with SlARF2a, hindering SlARF2a from repressing TSP4b transcription through binding its 3’ UTR.

TSP4a and TSP4b directly regulate downstream auxin-responsive genes

To explore how TSP4a and TSP4b regulate heat stress and auxin responses during fruit initiation, we compared transcriptomes of NIL-TSP4^MM and NIL- TSP4^XHS072 -2 DPA ovaries from plants grown under control and heat-stress conditions (Supplementary Fig. 11a). In NIL-TSP4^MM and NIL-TSP4^XHS072 ovaries, 341 genes were up-regulated and 322 were down-regulated under heat stress (Supplementary Fig. 11b and Supplementary Data 1). The ‘temperature stimulus’ gene-ontology (GO) term was enriched in the up-regulated genes, including several HEAT-SHOCK PROTEIN (HSP) genes (Supplementary Fig. 11c, d). A total of 516 genes and 468 genes were up- or down-regulated respectively in the NIL-TSP4^XHS072 ovaries under heat stress (Supplementary Fig. 11e and Supplementary Data 2), with ‘auxin response’ and ‘floral-organ development’ in up-regulated genes. ‘Jasmonic-acid signaling’ and ‘plant organ identity’ were enriched in the down-regulated genes (Supplementary Fig. 11f). These included the established warm-temperature-induced auxin-signaling genes such as SAUR, AUX/IAA and ARF (Supplementary Fig. 11g). These results indicated that the expression of these temperature-induced auxin-responsive genes depended on both TSP4a and TSP4b.

Among genes in the auxin-response pathway, two early-fruit-development-related genes, EXPANSIN 5 (EXP5, encoding a cell-wall protein involved in cell-wall modifications) and ACC OXIDASE 4 (ACO4, encoding an enzyme involved in ethylene biosynthesis) were specifically activated or repressed in NIL-TSP4^XHS072 under heat stress (Fig. 5a and Supplementary Fig. 11h)¹³. EXP5 expression was higher and ACO4 expression was lower in the tsp4acr and tsp4bcr single mutants, and further up-regulated in the tsp4a/4bcr double mutants under heat stress compared to wild-type plants (Fig. 5a). These results implied that TSP4a and TSP4b together control EXP5 and ACO4 expression.

**Fig. 5: TSP4a and TSP4b co-regulate auxin-responsive genes under heat stress.**

Consistent with their regulatory roles in auxin signaling, TSP4a and TSP4b were localized in the nucleus (Supplementary Fig. 12). However, TSP4a physically interacts with ARF7 to regulate EXP5 and ACO4 expression by ARF7 directly binding to their promoters¹³. To explore whether TSP4b–YFP–HA regulates EXP5 and ACO4 transcription by direct binding, we performed ChIP–qPCR. The results showed that TSP4b–YFP–HA binds to the P3 region of the EXP5 promoter, and the P2 and P4 regions of the ACO4 promoter (Fig. 5b). Transiently co-expression of TSP4b–FLAG with the EXP5_pro:LUC and ACO4_pro:LUC in N. benthamiana leaves revealed that TSP4b-FLAG activates EXP5 and represses ACO4 expression respectively (Supplementary Fig. 13).

Consistent with the model that the TSP4a–TSP4b feedback loop directly controls EXP5 and ACO4 expression, EXP5 induction and ACO4 repression after heat stress and auxin treatment were statistically significantly enhanced in the tsp4acr and tsp4bcr mutants compared to Moneymaker and untreated control plants (Fig. 5c, d). These results indicated that TSP4a and TSP4b participate in auxin signaling by directly regulating downstream genes under heat stress.

TSP4a and TSP4b influence auxin signaling

SlIAA9 is a negative regulator of auxin signaling and this pathway is elevated in entire mutants¹⁰. We compared EXP5 and ACO4 expression in NIL-TSP4^MM and NIL-TSP4^XHS072 plants, together with tsp4acr, tsp4bcr and tsp4a/tsp4bcr mutants and wild-type Moneymaker controls. EXP5 up-regulation and ACO4 down-regulation were greater in tsp4acr mutant ovaries than those of NIL-TSP4^XHS072 when compared to the corresponding controls. Much higher EXP5 expression and extremely low ACO4 expression still was detected in tsp4a/tsp4bcr double-mutant ovaries (see Results section for Fig. 5a).

To test whether different levels of TSP4a and TSP4b expression influence auxin responsiveness, we compared root formation and hypocotyl-segment elongation in plants grown at 25 °C and 35 °C. In tsp4acr mutants, the auxin analog α-naphthalene acetic acid (NAA) promoted root formation at concentrations above 10 nM both at 25 °C and 35 °C, whereas NAA promoted root formation from NIL-TSP4^XHS072 cotyledon explants above concentrations of 10 nM only at 35 °C. At 25 °C, roots were formed from NIL-TSP4^XHS072 at 20 nM, suggesting these plants are more sensitive to auxin under heat stress. Compared with NIL-TSP4^XHS072 cotyledon explants, tsp4acr induced more roots were induced at the same concentration, indicating that tsp4acr mutants are more sensitive than NIL-TSP4^XHS072 plants to auxin (Fig. 5e and Supplementary Fig. 14). Hypocotyl-elongation trends for tsp4acr and NIL-TSP4^XHS072 plants grown at 25 °C and 35 °C in a concentration series support an interpretation that tsp4acr plants are more sensitive to auxin (Fig. 5f), albeit the NAA concentrations used exceed physiological ranges. Taken together, TSP4a-SlARF2a and TSP4b act in a positive regulatory loop to repress auxin signaling during fruit set and plant development under heat stress (Fig. 5g).

Targeted manipulation of TSP4a and TSP4b dosage conditionally improves yield under heat stress in the greenhouse

Down-regulation of TSP4a and TSP4b expression in NIL-TSP4^XHS072 at high temperatures results in parthenocarpy (see Results section for Fig. 3a). This characteristic of the XHS072 haplotype enhances tomato adaptability and improves fruit productivity under heat stress, without obvious negative phenotypes. These features are of substantial interest to commercial tomato production and represent an important appearance trait for consumers of fresh table tomatoes. To estimate the potential breeding value of the TSP4^XHS072 haplotype, we crossed XHS072 with XHS369, which is a tomato accession with the TSP4^MM haplotype with low fruit set at high temperature (Fig. 6a), but has good taste, uniform fruit appearance and bright color. After backcrossing for four generations followed by marker-assisted selection, NIL-XHS369^XHS072 plants displayed a high ratio of seedless fruits without other noticeable phenotypic changes under heat stress in the greenhouse (Fig. 6b and Supplementary Fig. 15a, b). Compared with XHS369, the similar fruit weight enhanced yield by approximately 77% in NIL-XHS369 ^XHS072 under heat-stress conditions (Fig. 6c–e).

**Fig. 6: Customization of *TSP4a* and *TSP4b* expression levels by editing regulatory regions results in conditional parthenocarpy under heat stress.**

Because down-regulation of TSP4a and TSP4b leads to parthenocarpy in XHS072 under heat stress, we wondered whether targeted edits to their regulatory regions, such as the promoter or UTRs, could produce a series of alleles with high parthenocarpic fruit set without undesirable phenotypes by altering expression levels. We generated gene-edited lines in the Moneymaker background with deletions in the TSP4a promoter (‘^pro’) and TSP4b 3’ UTR (‘^utr’) around ATAC-seq peaks (Assay for Transposase-Accessible Chromatin using sequencing), indicative of open chromatin³¹. Of the five tsp4a^pro alleles (Fig. 6f–j) and five tsp4b^utr alleles (Fig. 6k–o) with deletions distributed along these regulatory regions, the tsp4a^pro-4, tsp4a^pro-5 and tsp4b^utr-5 alleles all developed seedless fruits in the greenhouse under heat stress. This was associated with reduced TSP4a and TSP4b expression levels (Fig. 6g, h, l-m). Importantly, the parthenocarpic phenotypes were tightly associated with deletions in the second distal ATAC region in the TSP4a promoter and the second downstream ATAC region in the TSP4b 3’ UTR. Compared to Moneymaker, tsp4a^pro-4, tsp4a^pro-5 and tsp4b^utr-5 alleles had no obvious effects on fruit fresh weight and sugar content of red ripe fruits in the greenhouse (Fig. 6i, n, Supplementary Fig. 15c–d) and leaf morphology (Fig. 6j, o). Consistent with the increase in fruit number, the yield of these plants was also improved under heat stress compared to Moneymaker (Supplementary Fig. 15e–h). Taken together, these results further support the roles of TSP4a and TSP4b in regulating parthenocarpic fruit set. Targeted editing of the TSP4a and TSP4b regulatory regions improves fruit-bearing ability under heat stress.

Discussion

Parthenocarpy is one of the most efficient ways to enhance yields when fruit crops face certain adverse environmental conditions and breeders have endeavored to develop parthenocarpic lines toward ensuring fruit-bearing ability^16,20,32. Although over 10 parthenocarpic loci have been identified in tomato over the past decades^33,34,35,36, only HB15A and AGL6 (underlying pat/pat-1 and pat-K) have been cloned^20,21,22. These two genes are involved in ovule development and also regulate seed production, therefore limiting their utility in tomato breeding in commercial settings. XHS072 is a facultative parthenocarpy accession that undergoes parthenocarpy only under heat stress. Large numbers of seeds can be produced in alleles with TSP4^XHS072 genotypes grown in optimal environments because of successful fertilization. Breeders therefore, could exploit these two genes in contemporary tomato-improvement programs.

In contrast with Arabidopsis ANT, which has a prominent role in the development of ovules, flowers and shoots^26,27, the tsp4bcr mutants did not cause any other obvious defects except for parthenocarpy and smaller fruits. In contrast to the entire parthenocarpic mutant defective in iaa9 with severe pleiotropic phenotypes such as simple leaves and small fruits^10,11,13, alleles with TSP4^XHS072 eliminated the trade-off between parthenocarpy and broader aspects of fruit and plant development. Our data similarly suggest that auxin signaling in the ovaries is a consequence of transcriptional repression of TSP4a and TSP4b under heat stress. However, the different phenotypes for root generation and hypocotyl elongation between the TSP4^XHS072 plants and tsp4acr knock-out lines under distinct auxin treatments further support their differential responses to auxin.

Non-pleiotropic parthenocarpic iaa9 transgenic lines, which express an RNAi construct from the ovary-specific promoter Psol80/60, hint that tissue-specific manipulation of IAA9 abundance is an effective way to improve tomato yield by inducing parthenocarpy³⁷. Here, we obtained non-transgenic gene-edited alleles that induce parthenocarpic fruits only under heat stress by engineering TSP4a and TSP4b cis-regulatory elements. The clear parthenocarpic phenotypes expressed in the greenhouse by the bespoke engineered tsp4a^pro-4 and tsp4b^utr-5 lines support that changes in TSP4a and TSP4b expression levels are important for auxin-signaling responses in fruit initiation and development, and indicate that gene editing in elite germplasm is a promising approach to rapidly improve tomato genetics and safeguard fruit set in the face of climate change.

Our study reveals a regulatory mechanism in which SlARF2a unexpectedly acts as a positive regulator of auxin signaling within the TSP4a-SlARF2a-TSP4b feedback loop to govern tomato parthenocarpy. This finding challenges the established paradigm of SlARF2a as a canonical class B ARF protein, which typically suppresses auxin signaling through transcriptional repression^38,39,40. However, CRISPR/Cas9-generated slarf2a knockout mutants in the Moneymaker background also exhibited pronounced parthenocarpic phenotypes (Supplementary Fig. 16), consistent with previous reports of parthenocarpy induction upon SlARF2a silencing¹². The paradoxical convergence of parthenocarpic phenotypes in both loss-of-function slarf2a mutants and SlARF2a^Tr overexpression transgenic plants is reminiscent of the dual regulatory functions observed in SlARF8A and 8B during tomato parthenocarpy¹⁵. Similar to the PB1-domain-deficient ARF8A/8B overexpression lines¹⁵, our SlARF2a^Tr transgenic plants likely achieve their phenotypic effects through competitive interference with endogenous auxin signaling components. These parallel findings challenge the binary classification of ARF proteins as activators or repressors. Instead, our results reveal an unexpected context-dependent functional plasticity within the ARF proteins.

In addition to auxin, ethylene also plays a role in suppressing tomato fruit set by modifying gibberellin metabolism⁴¹. ACO4 encodes an ACC oxidase, which catalyzes the conversion of 1-aminocyclopropane-1-carboxylic acid (ACC) to ethylene⁴². In the iaa9-3 mutant, low expression levels of ACO4 result in a decrease of ethylene content in fruit⁴¹. Our results showed that the expression level of ACO4 is also dramatically reduced in the tsp4acr and tsp4bcr mutants. Consistent with this, ethylene content is reduced in these mutants compared to that of wild-type Moneymaker (Supplementary Fig. 17). Thus, parthenocarpy is a complex process that involves interactions among different hormone pathways.

In conclusion, we discovered two genes conditionally conferring parthenocarpy in tomatoes under heat stress by optimizing auxin signaling during fruit initiation and development via a tightly controlled feedback loop. The practical utility of this approach was validated in our graded mutant series with varying degrees of TSP4a and TSP4b down-regulation that display durable parthenocarpy in the rigors of the greenhouse and maintain desired fruit sugar contents, demonstrating translational potential to engineer elite varieties favored by industry with high yield to cope with global climate change.

Methods

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-TSP4^MM or NIL-TSP4^XHS072 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 TSP4a^MM4b^MM, TSP4a^XHS0724b^MM, TSP4a^MM4b^XHS072, TSP4a^XHS0724b^MM/XHS072, TSP4a^MM/XHS0724b^XHS072 and TSP4a^XHS0724b^XHS072. 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 35S_pro:TSP4a–YFP–HA and 35S_pro: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. 35S_pro:SlARF2a^Tr–YFP–HA was constructed as described previously with modifications^12,15. 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-TSP4^XHS072 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-TSP4^MM and NIL-TSP4^XHS072 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 modifications^52,53. 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-TSP4^MM and NIL-TSP4^XHS072 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 35S_pro:TSP4a–YFP–HA and 35S_pro: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-TSP4^XHS072/35S_pro:TSP4b–YFP–HA and MM/35S_pro:SlARF2a^Tr–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 TSP4a_pro:LUC-35S_pro:REN, EXP5_pro:LUC-35S_pro:REN and ACO4_pro:LUC-35S_pro: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-35S_pro: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 35S_minpro:LUC-35S_pro:REN vector to obtain the 35S_minpro:LUC-TSP4b_utr-35S_pro vector. The 35S_pro:TSP4a–FLAG, 35S_pro:TSP4b–FLAG, 35S_pro:SlARF2a–FLAG and 35S_pro:SlARF2a^Tr–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 OD_{600 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 OD_{600 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 SlARF2a^Tr and SlARF2a^PB1, 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.

Data availability

The sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive (SRA) database under accession SRR2391865 (Moneymaker DNA-seq), PRJNA1248730 (XHS072 DNA-seq), PRJNA1247138 (BSA-seq) and PRJNA1068180 (RNA-seq). Source data are provided with this paper.

References

Zhao, C. et al. Temperature increase reduces global yields of major crops in four independent estimates. Proc. Natl Acad. Sci. USA 114, 9326–9331 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Picken, A. J. F. A review of pollination and fruit set in the tomato (Lycopersicon esculentum Mill. J. Hortic. Sci. 59, 1–13 (1984).
Peet, M. M., Willits, D. H. & Gardner, R. Response of ovule development and post-pollen production processes in male-sterile tomatoes to chronic, sub-acute high temperature stress. J. Exp. Bot. 48, 101–111 (1997).
Article CAS Google Scholar
Sato, S., Peet, M. M. & Thomas, J. F. Physiological factors limit fruit set of tomato (Lycopersicon esculentum Mill.) under chronic, mild heat stress. Plant Cell Environ. 23, 719–726 (2000).
Article Google Scholar
ReportLinker. Tomatoes Global Market Report 2023. GlobeNewswire News Room. https://www.globenewswire.com/en/newsrelease/2023/06/21/2692386/0/en/Tomatoes-Global-Market-Report-2023.html (2023).
Ayankojo, I. T. & Morgan, K. T. Increasing air temperatures and its effects on growth and productivity of tomato in South Florida. Plants 9, 1245 (2020).
Sharif, R., Su, L., Chen, X. & Qi, X. Hormonal interactions underlying parthenocarpic fruit formation in horticultural crops. Hortic. Res. 9, uhab024 (2022).
Gustafson, F. G. Inducement of fruit development by growth-promoting chemicals. Proc. Natl Acad. Sci. USA 22, 628–36 (1936).
Article ADS CAS PubMed PubMed Central Google Scholar
Pandolfini, T., Molesini, B. & Spena, A. Molecular dissection of the role of auxin in fruit initiation. Trends Plant Sci. 12, 327–329 (2007).
Article CAS PubMed Google Scholar
Wang, H. et al. The tomato Aux/IAA transcription factor IAA9 is involved in fruit development and leaf morphogenesis. Plant Cell 17, 2676–92 (2005).
Article CAS PubMed PubMed Central Google Scholar
Zhang, J. et al. A single-base deletion mutation in SlIAA9 gene causes tomato (Solanum lycopersicum) entire mutant. J. Plant Res. 120, 671–8 (2007).
Article CAS PubMed Google Scholar
Breitel, D. A. et al. AUXIN RESPONSE FACTOR 2 intersects hormonal signals in the regulation of tomato fruit ripening. PLoS Genet 12, e1005903 (2016).
Article PubMed PubMed Central Google Scholar
Hu, J., Israeli, A., Ori, N. & Sun, T. P. The interaction between DELLA and ARF/IAA mediates crosstalk between Gibberellin and Auxin signaling to control fruit initiation in tomato. Plant Cell 30, 1710–1728 (2018).
Article CAS PubMed PubMed Central Google Scholar
Liu, S. et al. Tomato AUXIN RESPONSE FACTOR 5 regulates fruit set and development via the mediation of auxin and gibberellin signaling. Sci. Rep. 8, 2971 (2018).
Article ADS PubMed PubMed Central Google Scholar
Hu, J., Li, X. & Sun, T. P. Four class A AUXIN RESPONSE FACTORs promote tomato fruit growth despite suppressing fruit set. Nat. Plants 9, 706–719 (2023).
Article CAS PubMed PubMed Central Google Scholar
Israeli, A. et al. Modulating auxin response stabilizes tomato fruit set. Plant Physiol. 192, 2336–2355 (2023).
Article CAS PubMed PubMed Central Google Scholar
Molesini, B., Pandolfini, T., Rotino, G. L., Dani, V. & Spena, A. Aucsia gene silencing causes parthenocarpic fruit development in tomato. Plant Physiol. 149, 534–48 (2009).
Article CAS PubMed PubMed Central Google Scholar
Ren, Z. et al. The auxin receptor homologue in Solanum lycopersicum stimulates tomato fruit set and leaf morphogenesis. J. Exp. Bot. 62, 2815–26 (2011).
Article CAS PubMed Google Scholar
Hawthorn, L. R. Seedlessness in tomatoes. Science 85, 199 (1937).
Article ADS CAS PubMed Google Scholar
Clepet, C. et al. The miR166-SlHB15A regulatory module controls ovule development and parthenocarpic fruit set under adverse temperatures in tomato. Mol. Plant 14, 1185–1198 (2021).
Article CAS PubMed Google Scholar
Klap, C. et al. Tomato facultative parthenocarpy results from SlAGAMOUS-LIKE 6 loss of function. Plant Biotechnol. J. 15, 634–647 (2017).
Article CAS PubMed Google Scholar
Takisawa, R. et al. The parthenocarpic gene Pat-k is generated by a natural mutation of SlAGL6 affecting fruit development in tomato (Solanum lycopersicum L.). BMC Plant Biol. 18, 72 (2018).
Article PubMed PubMed Central Google Scholar
Takagi, H. et al. QTL-seq: rapid mapping of quantitative trait loci in rice by whole genome resequencing of DNA from two bulked populations. Plant J. 74, 174–83 (2013).
Article CAS PubMed Google Scholar
de Jong, M., Mariani, C. & Vriezen, W. H. The role of auxin and gibberellin in tomato fruit set. J. Exp. Bot. 60, 1523–32 (2009).
Article PubMed Google Scholar
Ueta, R. et al. Rapid breeding of parthenocarpic tomato plants using CRISPR/Cas9. Sci. Rep. 7, 507 (2017).
Article ADS PubMed PubMed Central Google Scholar
Klucher, K. M., Chow, H., Reiser, L. & Fischer, R. L. The AINTEGUMENTA gene of Arabidopsis required for ovule and female gametophyte development is related to the floral homeotic gene APETALA2. Plant Cell 8, 137–53 (1996).
CAS PubMed PubMed Central Google Scholar
Mizukami, Y. & Fischer, R. L. Plant organ size control: AINTEGUMENTA regulates growth and cell numbers during organogenesis. Proc. Natl Acad. Sci. USA 97, 942–7 (2000).
Article ADS CAS PubMed PubMed Central Google Scholar
Guilfoyle, T. J. Aux/IAA proteins and auxin signal transduction. Trends Plant Sci. 3, 205–207 (1998).
Article Google Scholar
Nole-Wilson, S. & Krizek, B. A. DNA binding properties of the Arabidopsis floral development protein AINTEGUMENTA. Nucleic Acids Res. 28, 4076–82 (2000).
Article CAS PubMed PubMed Central Google Scholar
Mockaitis, K. & Estelle, M. Auxin receptors and plant development: a new signaling paradigm. Annu. Rev. Cell Dev. Biol. 24, 55–80 (2008).
Article CAS PubMed Google Scholar
Hendelman, A. et al. Conserved pleiotropy of an ancient plant homeobox gene uncovered by cis-regulatory dissection. Cell 184, 1724–1739.e16 (2021).
Article CAS PubMed Google Scholar
Gorguet, B., van Heusden, A. W. & Lindhout, P. Parthenocarpic Fruit Development in tomato. Plant Biol. 7, 131–139 (2005).
Article CAS PubMed Google Scholar
Philouze, J. & Milesi, M. Natural parthenocarpy in tomato .4. Study of parthenocarpy with polygenic determinism of line 75/59. Agronomie 9, 63–75 (1989).
Article Google Scholar
Beraldi, D., Picarella, M. E., Soressi, G. P. & Mazzucato, A. Fine mapping of the parthenocarpic fruit (pat) mutation in tomato. Theor. Appl. Genet. 108, 209–16 (2004).
Article CAS PubMed Google Scholar
Gorguet, B. et al. Mapping and characterization of novel parthenocarpy QTLs in tomato. Theor. Appl. Genet. 116, 755–67 (2008).
Article PubMed PubMed Central Google Scholar
Nunome, T. Map-based cloning of tomato parthenocarpic pat-2 gene. Regul. Plant Growth Dev. 51, 37–40 (2016).
CAS Google Scholar
Kim, J. S., Ezura, K., Lee, J., Ariizumi, T. & Ezura, H. Genetic engineering of parthenocarpic tomato plants using transient SlIAA9 knockdown by novel tissue-specific promoters. Sci. Rep. 9, 18871 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Ulmasov, T., Hagen, G. & Guilfoyle, T. J. Activation and repression of transcription by auxin-response factors. Proc. Natl Acad. Sci. USA 96, 5844–9 (1999).
Article ADS CAS PubMed PubMed Central Google Scholar
Okushima, Y., Mitina, I., Quach, H. L. & Theologis, A. AUXIN RESPONSE FACTOR 2 (ARF2): a pleiotropic developmental regulator. Plant J. 43, 29–46 (2005).
Article CAS PubMed Google Scholar
Choi, H. S., Seo, M. & Cho, H. T. Two TPL-binding Motifs of ARF2 are involved in repression of Auxin responses. Front Plant Sci. 9, 372 (2018).
Article PubMed PubMed Central Google Scholar
Shinozaki, Y. et al. Ethylene suppresses tomato (Solanum lycopersicum) fruit set through modification of gibberellin metabolism. Plant J. 83, 237–51 (2015).
Article CAS PubMed Google Scholar
Kende, H. Ethylene biosynthesis. Annu. Rev. Plant Biol. 44, 283–307 (1993).
Article CAS Google Scholar
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–60 (2009).
Article CAS PubMed PubMed Central Google Scholar
Zhang, S. et al. Detection of major loci associated with the variation of 18 important agronomic traits between Solanum pimpinellifolium and cultivated tomatoes. Plant J. 95, 312–323 (2018).
Article CAS PubMed Google Scholar
Lin, T. et al. Genomic analyses provide insights into the history of tomato breeding. Nat. Genet 46, 1220–6 (2014).
Article CAS PubMed Google Scholar
Wang, X. et al. Antagonistic regulation of target genes by the SISTER OF TM3-JOINTLESS2 complex in tomato inflorescence branching. Plant Cell 35, 2062–2078 (2023).
Article PubMed PubMed Central Google Scholar
Deng, L. et al. Efficient generation of pink-fruited tomatoes using CRISPR/Cas9 system. J. Genet. Genom. 45, 51–54 (2018).
Article CAS Google Scholar
Hu, B. et al. LEAF TIP NECROSIS1 plays a pivotal role in the regulation of multiple phosphate starvation responses in rice. Plant Physiol. 156, 1101–15 (2011).
Article CAS PubMed PubMed Central Google Scholar
Hwang, H. H., Yu, M. & Lai, E. M. Agrobacterium-mediated plant transformation: biology and applications. Arabidopsis Book 15, e0186 (2017).
Article PubMed PubMed Central Google Scholar
Fu, J., Chu, J., Sun, X., Wang, J. & Yan, C. Simple, rapid, and simultaneous assay of multiple carboxyl containing phytohormones in wounded tomatoes by UPLC-MS/MS using single SPE purification and isotope dilution. Anal. Sci. 28, 1081–7 (2012).
Article CAS PubMed Google Scholar
Li, R. et al. FIS1 encodes a GA2-oxidase that regulates fruit firmness in tomato. Nat. Commun. 11, 5844 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Liu, X. et al. AGAMOUS terminates floral stem cell maintenance in Arabidopsis by directly repressing WUSCHEL through recruitment of Polycomb Group proteins. Plant Cell 23, 3654–70 (2011).
Article CAS PubMed PubMed Central Google Scholar
Wang, X. T. et al. SISTER OF TM3 activates FRUITFULL1 to regulate inflorescence branching in tomato. Hort. Res. 8, 251 (2021).
Sun, S. et al. Genetic control of thermomorphogenesis in tomato inflorescences. Nat. Commun. 15, 1472 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Audran-Delalande, C. et al. Genome-wide identification, functional analysis and expression profiling of the Aux/IAA gene family in Tomato. Plant Cell Physiol. 53, 659–672 (2012).
Article CAS PubMed Google Scholar
Zhang, S. et al. Spatiotemporal transcriptome provides insights into early fruit development of tomato (Solanum lycopersicum). Sci. Rep. 6, 23173 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Wang, L., Feng, Z., Wang, X., Wang, X. & Zhang, X. DEGseq: an R package for identifying differentially expressed genes from RNA-seq data. Bioinformatics 26, 136–8 (2010).
Article PubMed Google Scholar
Rauluseviciute, I. et al. JASPAR 2024: 20th anniversary of the open-access database of transcription factor binding profiles. Nucleic Acids Res. 52, D174–d182 (2024).
Article CAS PubMed Google Scholar
Hajdukiewicz, P., Svab, Z. & Maliga, P. The small, versatile pPZP family of Agrobacterium binary vectors for plant transformation. Plant Mol. Biol. 25, 989–94 (1994).
Article CAS PubMed Google Scholar
Lu, Q. et al. Arabidopsis homolog of the yeast TREX-2 mRNA export complex: components and anchoring nucleoporin. Plant J. 61, 259–70 (2010).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

This work was supported by the National Key Research and Development Program of China (2022YFF1003002 to X. Cui) and the National Natural Science Foundation of China Grant (32430097 to X. Cui). We are grateful to Prof. Hong Yu (Chinese Academy of Sciences), Prof. Guosheng Xiong (Nanjing Agricultural University), Prof. Chengcai Chu (South China Agricultural University), and A. Prof. Jianchang Gao (Chinese Academy of Agricultural Sciences) for their input on the manuscript.

Author information

These authors contributed equally: Xiaonan Lu, Jianxin Wu.

Authors and Affiliations

State Key Laboratory of Vegetable Biobreeding, Sino–Dutch Joint Laboratory of Horticultural Genomics, Institute of Vegetables and Flowers, Chinese Academy of Agricultural Sciences, Beijing, 100081, China
Xiaonan Lu, Jianxin Wu, QianQian Shi, Shuai Sun, Ren Li & Xia Cui
State Key Laboratory for Managing Biotic and Chemical Threats to the Quality and Safety of Agro-Products, Institute of Vegetables, Zhejiang Academy of Agricultural Sciences, Hangzhou, 310021, China
Yuan Cheng & Guozhi Zhou
Department of Plant Science and Landscape Architecture, University of Connecticut, Storrs, CT, 06269, USA
Huanzhong Wang
Department of Horticulture and Institute of Plant Breeding, Genetics & Genomics University of Georgia, Athens, GA, 30602, USA
Esther van der Knaap
Key Laboratory of Quality and Safety Control for Subtropical Fruit and Vegetable, Ministry of Agriculture and Rural Affairs, College of Horticulture Science, Zhejiang A&F University, Hangzhou, 311300, China
Xia Cui

Authors

Xiaonan Lu
View author publications
Search author on:PubMed Google Scholar
Jianxin Wu
View author publications
Search author on:PubMed Google Scholar
QianQian Shi
View author publications
Search author on:PubMed Google Scholar
Shuai Sun
View author publications
Search author on:PubMed Google Scholar
Yuan Cheng
View author publications
Search author on:PubMed Google Scholar
Guozhi Zhou
View author publications
Search author on:PubMed Google Scholar
Ren Li
View author publications
Search author on:PubMed Google Scholar
Huanzhong Wang
View author publications
Search author on:PubMed Google Scholar
Esther van der Knaap
View author publications
Search author on:PubMed Google Scholar
Xia Cui
View author publications
Search author on:PubMed Google Scholar

Contributions

X.C. conceived the project and designed the research. X.L. carried out most of the experiments; J.W., Q.S., S.S., Y.C., G.Z., and R.L. performed some of the experiments and helped score phenotypes; S.S. helped to analyzed data and revised figures; H.W. and E.v.d.K. revised the manuscript. X.L. and X.C. wrote the manuscript.

Corresponding author

Correspondence to Xia Cui.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Concha Gómez-Mena and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Lu, X., Wu, J., Shi, Q. et al. A feedback loop at the THERMOSENSITIVE PARTHENOCARPY 4 locus controls tomato fruit set under heat stress. Nat Commun 16, 4184 (2025). https://doi.org/10.1038/s41467-025-59522-7

Download citation

Received: 16 July 2024
Accepted: 23 April 2025
Published: 06 May 2025
DOI: https://doi.org/10.1038/s41467-025-59522-7