FRUITFULL2 controls tomato fertility through style length and pollen quality

Research Paper FRUITFULL2 controls tomato fertility through style length and pollen quality Xiaowei Wang a,b , Monica Lanzoni Rossi c , Adriana Pinheiro Martinelli c , Gerco C. Angenent a,b,) , and Ruud A. de Maagd b,) a Laboratory of Molecular Biology, Wageningen University &Research, Wageningen, 6708 PB, the Netherlands b Business Unit Bioscience, Wageningen University &Research, Wageningen, 6708 PB, the Netherlands c Center for Nuclear Energy in Agriculture, University of Sao Paulo, Piracicaba, 13416-903, Brazil Received 16 August 2024; Accepted 2 January 2025; Available online 19 February 2025 ABSTRACT Tomato reproductive success and yield are particularly vulnerable to the negative effect of heat stress leading to stigma exsertion (protrusion) and lower pollen viability, both interfering with fertilization. Thus, understanding the regulation of these two traits in tomato is crucial for the yield and quality of the crop. Here, we found that knocking out the tomato MADS-domain transcription factor FRUITFULL2 (FUL2) function leads to a higher incidence of parthenocarpy in tomato. This phenotype was primarily due to impeded selfpollination as a consequence of the higher frequency of stigma exsertion and lower fertilization rates due to reduced pollen quality. Stigma exsertion in ful2 mutants, in contrast to heat stress-induced exsertion, was caused by style elongation, particularly in the younger flowers of a truss. Interestingly, Quantitative Trait Loci for style elongation, stigma exsertion, and pollen viability map close to the position of FUL2 on chromosome 3, making it a candidate gene underlying these QTLs. At the molecular level, ful2 mutant styles have higher expression of Style2.1 and SE3.1, which are known as positive regulators of style length. In addition, after reducing the impact of style exsertion and low pollen quality by manual pollination with wild-type pollen, ful2 mutants exhibited reduced fruit size independent of seed number. This study reveals the contributions of flower number, style length, and pollen quality, as affected by FUL2,totomato fertility and fruit size. Keywords: Tomato; Fertilization; Stigma exsertion; Pollen viability; FRUITFULL2 1. Introduction Fruits develop from ovaries after pollination and fertilization, coinciding with seed maturation (Quinet et al., 2019). However, fertilization and fruit development are occasionally disconnected when the ovary develops into a fruit without fertilization and seed formation, known as parthenocarpy (Fos et al., 2000;Medina et al., 2013). As a self-pollinating species, the cultivated tomato (Solanum lycopersicum var. lycopersicum) bears flowers with a unique floral architecture, where the style and stigma are enclosed in the cone of interlocking anthers, maximizing the probability of successful reproduction (Glover et al., 2004;Castellanos et al., 2006;Zhu and Tian, 2015). Slight alterations in any of these features can lead to no or abnormal fruits. Multiple environmental stresses can )Corresponding authors. E-mail addresses: gerco[email protected];[email protected] Peer review under responsibility of Chinese Society of Horticultural Science (CSHS) and Institute of Vegetables and Flowers (IVF), Chinese Academy of Agricultural Sciences (CAAS). September 2025. Horticultural Plant Journal, 11 (5): 1896e1904 Horticultural Plant Journal Available online at www.sciencedirect.com The journal's homepage: http://www.keaipublishing.com/en/journals/horticultural-plant-journal https://doi.org/10.1016/j.hpj.2025.01.002 2468-0141/Copyright ©2025 Chinese Society for Horticultural Science (CSHS) and Institute of Vegetables and Flowers (IVF), Chinese Academy of Agricultural Sciences (CAAS). Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). markedly transform the position of the stigma relative to the anthers, resulting in, relative to the tip of the anther cone, socalled flush or exserted stigmas (Pan et al., 2019). Early genetic studies on the variation of stigma position identified a single major QTL on chromosome 2, designated stigma exsertion 2.1 or se2.1. The QTL was subsequently shown to be a complex locus composed of at least five closely linked genes, three controlling stamen length, one conditioning anther dehiscence, and one controlling style length. The latter gene (designated Style 2.1) accounts for the most significant change in stigma exsertion (Chen et al., 2007;Riccini et al., 2021;Wu et al., 2024). The allele in cultivated tomato contains two 50-upstream deletions of 450 and 750 bp and several smaller indels that likely are the cause of its downregulation and recessed stigmas compared to wild tomato species (Chen et al., 2007). A second mutation, which likely occurred in the tomato improvement stage following domestication, underlies the SE3.1 QTL and controls the transition from flush to inserted stigmas. The affected gene encodes a zinc finger transcription factor and contains a premature termination codon in cultivated tomatoes that improves self-fertilization and productivity (Shang et al., 2021). Precisely coordinated growth of stamens and pistils apparently determines the fertility of plants (Pan et al., 2019). Flowers with exserted stigmas largely lose their capacity for self-fertilization, but the physiological and genetic mechanisms underlying stigma position control are not entirely known. The tomato FRUITFULL1 (FUL1/TM4/TDR4)andFRUITFULL2 (FUL2/MBP7, hereafter called FUL2) genes encode transcription factors of the MADS-box (MCM1, AGAMOUS, DEFICIENS, and SRF) family (Schwarz-Sommer et al., 1990) that jointly regulate fruit ripening and are orthologs of the single FRUITFULL gene in Arabidopsis(Bemeretal.,2012).InArabidopsis,FUL isexpressedindistinct tissues during plant development, and the roles of FUL during fruit and flower development are well-studied (Gu et al., 1998;Bemer et al., 2017;van Mourik et al., 2023). FUL functions as a positive regulator of flowering and is essential for meristem identity (Ferr andiz et al., 2000;Bemer et al., 2012). It is also involved in determining silique valve identity and ful mutant siliques fail to expand to the wild-type (WT) length (Gu et al., 1998). Unlike its FUL ortholog regulating fruit dehiscence in Arabidopsis, the cucumber CsFUL1 represses fruit elongation (Zhao et al., 2019). Additionally, in tomato, producing non-dehiscent fleshy fruits, ful2 knockout mutants show increased inflorescence branching and more locules but smaller fruits with surface cracks (Wang et al., 2019;Jiang et al., 2022). Thus, orthologs of FUL exhibit diverse regulatory functions depending on thefruit type and the evolutionary context (Litt and Irish 2003;Cevik et al., 2010;Bemer et al., 2017;Zhao et al., 2019). Here, we studied tomato ful2 mutants to determine the contribution of FUL2 to other traits such as pollination and fertilization as well as to fruit size. Surprisingly, we found that FUL2 has a multifaceted role in fertility and fruit development. 2. Materials and methods 2.1. Plant materials and growing conditions The ful2 CRISPR/Cas9-mutant lines were previously generated in cv. Moneyberg (Wang et al., 2019; Jiang et al., 2021). Mutant and wild-type plants were grown in the greenhouse facilities of Unifarm (Wageningen University &Research) and under controlled temperatures and natural light, supplemented with artificial sodium lights. 2.2. Flower phenotyping Three plants per genotype were used for all phenotyping. Tomato flowers were labeled at anthesis and vibrated regularly to enhance pollination. The number of flowersaswellasthe number in each pistil length category (inserted, flush, or exserted stigma) per truss were counted. To assess the effect of pistil length and pollen quality, we conducted manual pollination on flowers with either flush or exserted stigmas using WT pollen. 2.3. Pistil imaging Flowers were collected at different stages, and sepals, petals, anthers were removed to expose the pistils. For measuring style, stamen and pistil length, whole pistils were imaged using a stereomicroscope (Stemi 508, Zeiss) with a coupled camera (AxioCam IC, Zeiss, Germany). For measuring style cell length by Scanning Electron Microscopy (SEM), styles were fixed [2 % Glutaraldehyde in 0.1 mol$L 1 Phosphate Buffer, 4 C] for 24e36 h. Next, samples were dehydrated in an ethanol gradient of 40 %, 50 %, 60 %, 70 %, 80 %, 90 % (at least) 30 min each, and finally in 100 % Ethanol, three times, for 30 min. Fixed samples were then dried in a Leica CPD300 Critical Point Dryer mounted on a copper plate, and gold coated under a Leica EM ACE600 High Vacuum Sputter Coater Samples were viewed in a Jeol 5410 LV Scanning Electron Microscope (Jeol, Tokyo, Japan). 2.4. Fruit phenotyping To assess the influence of floral architecture on final fruit size, we established two treatments: 1) Limited fruit number treatment, leaving 4 fruits per truss (with three trusses for each plant), and pruning other flowers and young fruits. 2) Normal nonpruned fruit number treatment, plants (without removing flowers or fruits) to examine parthenocarpy. For both treatments, the height, width, and weight of ripe fruit were measured 14 days after the breaker stage. 2.5. Pollen phenotyping Three flowers per truss were pooled for one biological replicate, and three replicates per genotype were used for phenotyping. To count the pollen number, we took a double concave slide, collected pollen by dissection of the anthers from a 0 DPA flower (with three replicates), and evenly sprinkled on the surface of the liquid culture medium (120 g$L 1 sucrose, 120 mg$L 1 boric acid, 4mg$L 1 erythromycin). Samples were imaged using a Nikon Optiphot microscope (Eclipse 80i) with a coupled camera (AxioCam IC, Zeiss Germany). After 4 h incubation at 25 C, check pollen germination and take pictures again. For pollen viability staining, we used the same method to collect pollen grains and then transferred them to 2 % 2,3,5-Triphenyl-2H-tetrazolium chloride (TTC) solution in the dark at 37 C for 15 min. All samples were imaged under the same microscope. Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. 1897 2.6. Gene expression analysis Five styles were collected randomly from WT and ful2-cr plants and were immediately frozen in liquid nitrogen (n¼3). Total RNA was extracted using the CTAB/LiCl method (Porebski et al., 1997), followed by DNase treatment with Ambion Turbo DNase (AM1907) and then used for cDNA synthesis using the iScript cDNA synthesis kit (Bio-Rad). Real-Time qPCR was performed with the iQ SYBR Green Supermix (Bio-Rad). Actin (Solyc03g078400) was used as a reference gene (all primer sequences are listed in Table S1). 2.7. Statistical analysis Differences between the two groups were assessed using a two-sided Student'st-test. Asterisks indicate significant differences ( ) P<0.05, )) P<0.01). For multiple comparisons among several treatments and groups, two-way ANOVA with Tukey's post hoc test was performed. Different letters indicate statistically significant differences (P<0.05). For the correlation analysis of two or more categorical variables, the Pearson Chisquared test was used ( ) P<0.05, )) P<0.01). 3. Results 3.1. ful2 mutations result in smaller fruits independent of the number of developing fruits In this study, we used two different mutant alleles of FUL2 (ful2-cr1 and ful2-cr2), resulting in functionally deficient proteins Fig. 1 Flower and fruit phenotypes of wild-type and ful2 mutant fruits (a) Fruits of WT and ful2-cr mutants at different stages. Bar ¼2 cm. (b) Flower number and (c) Fruit (seeded) number per truss of WT and ful2-cr mutants. (d) Percentage of normally seeded fruits and seedless fruits of WT and ful2-cr (n¼9). (e) Schematic diagram of comparison between plants with untreated and pruned trusses. (f)Ripe whole and transversely sliced fruits at BRþ14. Bar ¼2cm.(g)Fruit width, (h) height, and (j) weight of fruits from two WT and from two of each ful2-cr plants at BRþ14 after pruning and untreated (n>15). Error bars indicate mean ±SE. Asterisks indicate significant differences between WT and ful2-cr ( ) P<0.05, )) P<0.01; Student'st-tests). 1898 Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. due to a premature stop codon after a frameshift mutation (ful2cr1, truncated protein) and two closely spaced frameshift mutations in the first exon encoding the MADS domain (ful2-cr2) (Wang et al., 2019). Although ful2-cr2 encodes a full-length protein in which only two amino acids were substituted, those were in a region that is conserved throughout the Angiosperms and thus was very likely deleterious with SIFT (Sorting Intolerant From Tolerant) scores of 0.00 (Fig. S1). As previously reported (Wang et al., 2019), FUL2 was found to regulate final ripe fruit size (Fig. 1, a) and inflorescence branching, with the knockout mutants having smaller fruits and more branched inflorescences. Indeed, both ful2 mutants produced more flowers per truss relative to the WT due to more bifurcation of the inflorescence (Fig. 1, b). Furthermore, although having more flowers in each truss did not lead to more normal seeded fruits (Fig. 1, c), instead more than 35 % of ful2-cr flowers formed seedless fruits (Fig. 1, d). Normally developing seeded mutant fruits were smaller than those of the WT (Fig. 1, a). To avoid a potential effect of extra flowers and fruits on overall fruit size in each truss, we pruned trusses to a fixed developing fruit number (4 per truss in all plants) and compared these with fruits from plants with unpruned trusses (Fig. 1, e). Pruning was applied at an early stage, as soon as four fruits per truss developed normally (containing seeds as determined afterward). We then measured fruit size parameters (height, width, and fresh weight) in ripe fruits (Breakerþ14 days, Fig. 1, f), to determine the effect of pruning. As shown in Fig. 1, g-h, WT fruit size parameters increased slightly when the fruit number was limited to four, but not so for ful2-cr fruits, which remained at approximately 75 % of the dimensions of the wild-type fruits. Concurrent with the reduction in fruit width and height, the normal (seeded) fruit weight of ful2-cr fruits was only approximately half that of WT fruits and was not affected by pruning flowers and fruits (Fig. 1, i). These results indicated that the smaller ful2-cr fruit size was not caused by a higher number of (seedless) fruits on the same truss. 3.2. FUL2 controls stigma exsertion depending on the position of the flower Next, we investigated the mechanisms leading to a higher seedless fruit frequency in ful2-cr mutants. In tomato, most seedless phenotypes are likely due to failed or blocked fertilization. We checked for related or potential causal phenotypes in the ful2-cr flowers and found that many exhibited stigmas protruding from the anther cone (Fig. 2, a), while stamen length remained the same as in WT (Fig. S2). To better characterize this phenotype Fig. 2 Mutation of FUL2 leads to longer styles (a) Stigma exsertion phenotype of WT and ful2-cr mutants at 0 DPA. Bar ¼2 mm. (b) Definition of three different positions used in this study. (c) Distribution of stigma positions (%) of WT and ful2-cr mutants at 0 DPA (n>20). The P-value indicates differences between WT and ful2-cr (Student'st-tests). (d) Positions of flowers. (e) Distributions of stigma positions in flowers at different positions per truss (n¼12). ful2-cr trusses occasionally had more than 10 flowers per branch. The P-value indicates differences between WT and ful2-cr (Pearson Chi-squared test). Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. 1899 of stigma exsertion (also called “protrusion”), we distinguished and counted inserted, flush, and exserted stigmas (Fig. 2, b and c). We found that flowers of the ful2-cr plants showed significantly more exserted or flush stigmas and fewer inserted stigmas compared to WT (Fig. 2, c). To investigate the distribution of exserted stigmas along the inflorescence, we categorized all the flowers of the second to fifth trusses according to the order of opening (Fig. 2, d). Because ful2-cr trusses have more branches and flowers (Fig. 1, b), we adapted the numbering for ful2-cr (Fig. 2, d). For WT, which generally showed no inflorescence branching, we numbered flowers ''100 to ''1000 in the order of opening. For ful2cr, which typically had bifurcated inflorescences, flowers in the equivalent position on the two branches opened simultaneously and were given identical numbers (Fig. 2, d). We found that flush stigmas also occur in the younger WT flowers, with flowers at positions ''1e600 showing around 15 % flush stigmas, and 50 % or more upwards from position ''800, but almost no flowers with exserted stigmas. In ful2-cr, we observed the same trend as in the WT flowers, with additionally the increasing occurrence of flowers with exserted stigmas towards the tip of the inflorescence (Fig. 2, e). However, from position ''100 onwards, ful2-cr1 already had more than half of the flowers with flush or exserted stigmas, while ful2-cr2 had a slightly milder phenotype, possibly resulting from the nature of the mutant protein, yet significantly different from WT (Fig. 2, e). Thus, exserted stigmas occur almost exclusively in mutant flowers with a trend towards higher frequencies in the younger flowers in the inflorescence (Fig. 2, c-e). 3.3. ful2 mutations inhibit pollination through longer styles and poorer pollen quality We found that ful2-cr plants had more flowers with exserted stigmas. To characterize the cause of the protrusion of stigmas, we removed the stamens, which showed no significant difference Fig. 3 Effects of the ful2 mutation on style size parameters and pollen production or vigor (a) Pistil morphology of flowers at the distal end of the truss (Bar ¼2mm).(b)Style length at 2 DPA, 0 DPA, and 2 DPA (n>15). (c) Scanning electron microscopic analysis of styles at 0 DPA (n¼3). Bar ¼100 mm. (d) Style width and (e) Style cell length of WT and ful2-cr at 0 DPA (n¼3). (f-g) Phenotype of pollen number and pollen tube germination (after 4 h) of WT and ful2-cr per flower (n¼3). Bar ¼5mm. Error bars indicate mean ±SE. Asterisks indicate significant differences between WT and ful2-cr ( ) P<0.05, )) P<0.01; Student'st-tests). 1900 Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. between WT and mutants (Fig. 3,a,Fig. S2), and measured the length of the styles (Fig. 3, b). The stylar lengths of ful2-cr flowers were, on average, longer than those of the WT from 2 DPA to 2 DPA (Fig. 3, a and b), which was consistent with the mutants'pistil length (Fig. S3). ful2-cr mutants even had slightly larger ovaries (Fig. S4). A higher frequency of flowers with flush or exserted stigmas, negatively affecting self-pollination, likely leads to more seedless fruits. To further investigate the underlying cause of style elongation, we investigated style cell characteristics by scanning electron microscopy in ful2-cr and WT flowers at 0 DPA, at which stage elongated pistils could already be detected (Fig. 3,c). Compared with WT flowers with inserted or flush stigmas, the styles of ful2-cr flowers were thinner (Fig. 3,candd,Fig. S5). Styles with flush or exserted stigmas were both thinner than the inserted ones. Thus, style length was inversely correlated with style width, both in WT and mutant flowers (Fig. 3, c and d), although WT and ful2-cr showed a significant difference in all types of styles (Fig. 3,b, c, d). Consistent with the longer style, the style cells in ful2-cr styles were approximately 20 % longer than those in WT styles (Fig. 3, cee). These results show that stigma protrusion is correlated with longer and thinner styles. In tomato, pollen germinates on the stigma and then forms a pollen tube, which grows through the style, followed by double fertilization. We examined pollen production by carefully opening anthers and counting the released pollen grains and found that pollen production of ful2-cr mutants was lower than that of WT (Fig. 3, f and g, Fig. S6). Furthermore, we assessed pollen vigor in a germination assay and viability staining, and found that Fig. 4 Parthenocarpy, seed number, fruit size, and style gene expression (a) Percentage of seeded and parthenocarpic fruits after natural and manual pollination, respectively. "N" means natural fertilization (with vibration), and "X00 means manual pollination. (b) Seed number and fruit weight of fruits. "I, F, E00 means inserted, flush, and exserted stigma, respectively. (c) The contribution of different stigma positions to parthenocarpy with natural and manual pollination, respectively. (d) Height and width of fruits. (e) Seed number and (f) fruit weight of WT and ful2-cr fruits from inserted stigma or after manual pollination. (g) Relative gene expression of Style2.1 and (h) SE3.1 in WT and ful2-cr styles at 0 DPA (n¼3). Error bars indicate mean ±SE. Statistically significant differences were determined using a two-way ANOVA with Tukey's post hoc test. Different letters indicate statistically significant differences (P<0.05). Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. 1901 nearly 60 % of ful2-cr mutants'pollen failed to grow a pollen tube after 4 h in germination medium, whereas less than 35 % of WT pollen failed (Fig. 3, f and g), which were consistent with staining observation (Fig. S6). Taken together, these results indicate that in addition to regulating fertility through style length, FUL2 has a role in pollen production and vigor. 3.4. Pollination defects additionally contribute to parthenocarpy in ful2 mutants The presence of more seedless fruits in the ful2-cr mutants (Fig. 1, d) was due to longer styles leading to stigma exsertion (Fig. 3, b) and lower pollen vigor (Fig. 3, g). Therefore, we designated ful2-cr seedless fruits as parthenocarpic. We attempted to rescue normal fruit development by manual pollination with fresh WT pollen in ful2-cr flowers and compared that with naturally self-pollinated flowers (i.e., by vibration). We pollinated ful2-cr flowers that had flush or exserted stigmas. After manual pollination with WT pollen, the percentage of parthenocarpic fruits in ful2-cr mutants fell to a level not significantly different from WT (Fig. 4, a). However, the seed numbers were still less than those in WT fruits (Fig. 4, b). We further determined from which type of flowers the parthenocarpic fruits developed and found that not only flowers with flush or exserted stigmas resulted in seedless fruits, but also those with inserted stigma, supporting the conclusion that ful2-cr mutants'poor pollen production and vigor also contributed to parthenocarpy (Figs. 3, ge4, c). Manual pollination with WT pollen could restore seed development (Fig. 4, b) and significantly reduce the percentage of parthenocarpic fruits. It fully restored the formation of seedcontaining fruits from flowers with exserted stigmas (Fig. 4, c). These results show that pollen quality and style length indeed affected fruit sets and seed development in ful2-cr. 3.5. ful2 mutations inhibit fruit growth even after compensating for pollination and fertilization defects It is worth noting that not only the seed number but also the fruit weight of ful2-cr mutants was lower than that of WT (Fig. 4, b). This suggests that style elongation affects fruit weight by influencing the number of seeds in the final fruit (Fig. 2, ce4, b). We manually pollinated the mutant flowers with WT pollen to verify whether successful pollination and fertilization could restore fruit size. Intriguingly, after manual pollination, the seed number increased obviously but the fruit weight was still lower than WT (Fig. 4, b). Fruit width and height showed a similar trend with fruit weight, which increased mildly after manual pollination but still did not reach the level of WT (Fig. 4, d). Yet, analysis of the relation between seed number and fruit weight revealed that when comparing fruits with an approximately equal seed number, there was still a difference in fruit weight between mutantand WT fruits (Fig. 4, e and f, Fig. S7). A part of the ful2-cr fruits developing from flowers with inserted stigmas or after manual pollination contained the same number of seeds as WT (Fig. 4, e), yet their fruit weight was still lower than WT (Fig. 4, f). In conclusion, compensating for the defects in pollination and seed set only partially restored the fruit sizeof ful2-crmutants.Thus,next tothe multiple roles of FUL2 in flowering, flower development, and ripening, it is also involved in fruit growth control in tomato. 3.6. FUL2 functions in a well-known tomato regulatory pathway determining style length To further study FUL2's position relative to the known transcriptional regulators of style length and stigma exsertion, we compared the expression of Style2.1 and SE3.1 in styles from WT and ful2-cr mutants. As expected both genes were significantly upregulated in ful2-cr styles at 0 DPA (Fig. 4, g-h), identifying a role of FUL2 in conjunction with these two genes. 4. Discussion 4.1. Factors causing stigma exsertion Changes in the position of pollen-bearing anthers relative to the surface of the stigma are often associated with the evolution of autogamy. Flowers with stigmas that are exserted beyond their anthers are more likely to receive pollen from neighboring plants, whereas flowers with stigmas that are recessed relative to their anthers are more likely to self-pollinate. This is reflected in the domestication of cultivated tomato (S. lycopersicum). Most wild tomato species, such as Solanum pennellii, Solanum peruvianum,and Solanum chilense are heterogamous, with highly exserted stigmas, and are also mainly self-incompatible (Rick and Lamm, 1995;Chen et al., 2007;Wu et al., 2024). The transition from exserted to inserted stigmas, in two steps, corresponding to the presence of modern Style2.1 (Chen et al., 2007)andSE3.1 (Shang et al., 2021), and the loss of self-incompatibility have occurred in parallel during tomato domestication and improvement. In our study, cv. Moneyberg (Fig. 1,aef) is a modern cultivar derived from cv. Moneymaker (van Rengs et al., 2022). We confirmed the presence of the modern alleles of both Style2.1 and SE3.1 in this cultivar. Stigma exsertion in cultivated tomatoes can be caused by heat stress, but in that case, it is correlated with shortened stamens rather than longer pistils (Liu et al., 2015;Pan et al., 2019). Stamens from ful2 mutants with exserted stigmas were not shorter but had the same length as WT or were even longer (Figs. 2,ae3, a, Fig. S2). FUL2 has considerable expression in the pistil, matching a role in pistil development (Jiang et al., 2021). Styles of ful2-cr flowers were longer, and their elongation was due to longer stylar cells, as is the case in Near Isogenic Lines (NILs) with the long style allele of Style2.1, suggesting that the two genes target the same developmental process (Fig. 3,c)(Chen et al., 2007). Through SE3.1, style length and style cell length are also positively regulated by three HD-ZIP transcription factors that control the production of trichomes at anther margins that interlock to form the closed anther cone (Wu et al., 2024). Here, we have shown that FUL2 represses Style2.1 and SE3.1 gene expression in the style. While the higher expression of SE3.1 likely has no effect on style length because, in this background, the gene contains a premature termination codon, the higher expression of Style2.1 leads to elongated styles. Thus, FUL2 is in the same regulatory pathway as these two genes and regulates their activity. As the measurements in our study were done under equal environmental conditions for WT and mutants, the phenotype of ful2-cr was not determined by differences in ambient temperature. Nonetheless, while stigma exsertion under heat stress is caused by shortened stamens, combining genotype (ful2) and environment (high temperature) may show a syner1902 Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. gistic effect and more stigma exsertion. A strong interaction between genotype and environment for stigma protrusion was found in field experiments (Farinon et al., 2022). Interestingly, overlapping QTLs for stigma protrusion (SP) and style length (SL), phenotypes that were highly correlated, were identified in an F2 population from a cross of two tomato accessions (Xu et al., 2017). The most associated SolCap marker solcap_snp_sl_7942 linked to both traits is located at position 56.3 Mb on chromosome 3 (SL4.0), which is close to the position of FUL2 (59.1 Mb), as well as that of SE3.1 (54.9 Mb). A meta-QTL analysis located a QTL for the clustered traits of stigma protrusion, style length, and pollen viability at 59.8e60.5 Mb on chromosome 3, closer to FUL2 (Ayenan et al., 2019). Thus, although more precise mapping of the QTL may be required, this puts forward FUL2 as a candidate for a gene underlying these QTLs. 4.2. Many factors affect pollen and pollination During plant reproduction, specific changes occur in the male reproductive organs (anthers and filaments), including anther differentiation, functional pollen production, and anther dehiscence, ultimately leading to the release of mature pollen. Failure of any of these processes can lead to male sterility, restricted reproduction, and reduced crop yields (Yang et al., 2023). In tomato, most male-sterile mutants exhibit sporogenous male sterility. Accordingly, they have morphologically normal flowers that produce little or no viable pollen (German and McCormick, 1997). When checking the source of parthenocarpy in our mutants, we found that flowerswithaninserted stigma could also develop into seedless fruits (Fig. 4,c),supporting the notion that poor fertilization was involved in parthenocarpy development. Moreover, manual pollination of ful2 flowers having flush or exserted stigmas with WT pollen gave more normal seeded fruits (Fig. 4, a). Pollen that is made incapable of fertilization by X-ray treatment can still induce fruit set when used for pollination (Tran et al., 2023). This leads us to hypothesize that residual pollination of exserted or flush stigmas with ful2-cr pollen, having lower vigor and not capable of seed initiation through fertilization, may induce parthenocarpic fruit set through a similar mechanism. Finally, in this study, ful2 mutations in tomato led to smaller fruits due to pollination and fertilization defects, but even when those were compensated by manual pollination there was a residual effect on seeded fruit growth (Fig. 4, e and f), which requires further study. 5. Conclusion It is well known that tomato plants are sensitive to heat stress, and extended periods of warmer temperatures can lead to shorter anthers, causing stigma exsertion. However, in our study, we dentified a novel role to FUL2 and found that ful2-cr leads to stigma exsertion not through shorter anthers but rather longer styles, which in this comparison was not caused by differences in temperature. Thus, we have added, besides Style2.1 and SE3.1, another component in this pathway regulating style length and stigma exsertion. We speculate that the combination of these two factors could result in even poorer fertility. Given the increasing impact of global warming, it is prudent to investigate how such genetic effects interact with environmental changes. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This work was supported by a fellowship from the China Scholarship Council (CSC), award nr. 202106300012 to X.W., and from EU-RISE Evofruland: grant agreement no. 101007738 to X.W. and R.d.M. We also thank Prof. Elliot Kitajima (LME/ ESALQ/USP) for supporting access to the scanning electron microscope. Supplementary materials Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.1016/j.hpj.2025.01.002. REFERENCES Ayenan, M.A.T., Danquah, A., Hanson, P., Ampomah-Dwamena, C., Sodedji, F.A.K., Asante, I.K., Danquah, E.Y., 2019. Accelerating breeding for heat tolerance in tomato (Solanum lycopersicum L.): an integrated approach. Agronomy, 9: 720. Bemer, M., Van Mourik, H., Mui~ no, J.M., Ferr andiz, C., Kaufmann, K., Angenent, G.C., 2017. FRUITFULL controls SAUR10 expression and regulates Arabidopsis growth and architecture. J Exp Bot, 68: 3391e3403. Bemer, M., Karlova, R., Ballester, A.R., Tikunov, Y.M., Bovy, A.G., Wolters-Arts, M., de Barros Rossetto, P., Angenent, G.C., de Maagd, R.A., 2012. The tomato FRUITFULL homologs TDR4/FUL1 and MBP7/FUL2 regulate ethylene-independent aspects of fruit ripening. Plant Cell, 24: 4437e4451. Castellanos, M.C., Wilson, P., Keller, S.J., Wolfe, A.D., Thomson, J.D., 2006. Anther evolution: pollen presentation strategies when pollinators differ. Am Nat, 167: 288e296. Cevik, V., Ryder, C.D., Popovich, A., Manning, K., King, G.J., Seymour, G.B., 2010. A FRUITFULL-like gene is associated with genetic variation for fruit flesh firmness in apple (Malus domestica Borkh.). Tree Genet Genomes, 6: 271e279. Chen, K.Y., Cong, B., Wing, R., Vrebalov, J., Tanksley, S.D., 2007. Changes in regulation of a transcription factor lead to autogamy in cultivated tomatoes. Science, 318: 643e645. Farinon, B., Picarella, M.E., Mazzucato, A., 2022. Dynamics of fertilityrelated traits in tomato landraces under mild and severe heat stress. Plants, 11: 881. Ferr andiz, C., Gu, Q., Martienssen, R., Yanofsky, M., 2000. FRUITFULL promotes flower development. Development, 127: 725e734. Fos, M., Nuez, F., García-Martínez, J.L., 2000. The gene pat-2, which induces natural parthenocarpy, alters the gibberellin content in unpollinated tomato ovaries. Plant Physiol, 122: 471e480. German, S.W., McCormick, S., 1997. Male sterility in tomato. CRC Crit Rev Plant Sci, 16: 31e53. Glover, B.J., Bunnewell, S., Martin, C., 2004. Convergent evolution within the genus Solanum: the specialised anther cone develops through alternative pathways. Gene, 331: 1e7. Gu, Q., Ferr andiz, C., Yanofsky, M.F., Martienssen, R., 1998. The FRUITFULL MADS-box gene mediates cell differentiation Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904. 1903 during Arabidopsis fruit development. Development, 125: 1509e1517. Jiang, X., Lubini, G., Hernandes-Lopes, J., Rijnsburger, K., Veltkamp, V., de Maagd, R.A., Angenent, G.C., Bemer, M., 2022. FRUITFULL-like genes regulate flowering time and inflorescence architecture in tomato. Plant Cell, 34: 1527e1530. Litt, A., Irish, V.F., 2003. Duplication and diversification in the APETALA1/FRUITFULL floral homeotic gene lineage: implications for the evolution of floral development. Genetics, 165: 821e833. Liu, J., Feng, L., Li, J., He, Z., 2015. Genetic and epigenetic control of plant heat responses. Front Plant Sci, 6: 133426. Medina, M., Roque, E., Pineda, B., Ca~ nas, L., RodriguezConcepci on, M., Beltr an, J.P., G omez-Mena, C., 2013. Early anther ablation triggers parthenocarpic fruit development in tomato. Plant Biotechnol J, 11: 770e779. Pan, C., Yang, D., Zhao, X., Jian, C., Yan, Y., Lamin-Samu, A., Wang, Q., Xu, X., Fei, Z., Lu, G., 2019. Tomato stigma exsertion induced by high temperature is associated with the jasmonate signaling pathway. Plant Cell Environ, 42: 1205e1221. Porebski, S., Bailey, L.G., Baum, B.R., 1997. Modification of a CTAB DNA extraction protocol for plants containing high polysaccharide and polyphenol components. Plant Mol Biol Rep, 15: 8e15. Quinet, M., Angosto, T., Yuste-Lisbona, F.J., Blanchard-Gros, R., Bigot, S., Martinez, J.P., Lutts, S., 2019. Tomato fruit development and metabolism. Front Plant Sci, 10: 475784. van Mourik, H., Chen, P., Smaczniak, C., Boeren, S., Kaufmann, K., Bemer, M., Angenent, G.C., Muino, J.M., 2023. Dual specificity and target gene selection by the MADS-domain protein FRUITFULL. Nat Plants, 9: 473e485. van Rengs, W.M.J., Schmidt, M.H.W., Effgen, S., Le, D.B., Wang, Y., Zaidan, M.W.A.M., Huettel, B., Schouten, H.J., Usadel, B., Underwood, C.J., 2022. A chromosome scale tomato genome built from complementary PacBio and Nanopore sequences alone reveals extensive linkage drag during breeding. Plant J, 110: 572e588. Riccini, A., Picarella, M.E., De Angelis, F., Mazzucato, A., 2021. Bulk RNA-Seq analysis to dissect the regulation of stigma position in tomato. Plant Mol Biol, 105: 263e285. Rick, C., Lamm, R., 1995. Biosystematic studies on the status of Lycopersicon chilense. Am J Bot, 42: 663e675. Schwarz-Sommer,Z.,Huijser,P.,Nacken,W.,Saedler,H., Sommer, H., 1990. Genetic control of flower development by homeotic genes in Antirrhinum majus. Science, 250: 931e936. Shang, L., Song, J., Yu, H., Wang, X., Yu, C., Wang, Y., Li, F., Lu, Y., Wang, T., Ouyang, B., Zhang, J., Larkin, R.M., Ye, Z., Zhang, Y., 2021. A mutation in a C2H2-type zinc finger transcription factor contributed to the transition toward self-pollination in cultivated tomato. Plant Cell, 33: 3293e3308. Tran, L.T., Sugimoto, K., Kasozi, M., Mitalo, O.W., Ezura, H., 2023. Pollination, pollen tube growth, and fertilization independently contribute to fruit set and development in tomato. Front Plant Sci, 14: 1205816. Wang, R., Tavano, E.C. da R., Lammers, M., Martinelli, A.P., Angenent, G.C., de Maagd, R.A., 2019. Re-evaluation of transcription factor function in tomato fruit development and ripening with CRISPR/Cas9-mutagenesis. Sci Rep, 9: 1696. Wu, M., Bian, X., Huang, B., Du, Y., Hu, S., Wang, Y., Shen, J., Wu, S., 2024. HD-Zip proteins modify floral structures for self-pollination in tomato. Science, 384: 124e130. Xu, J., Driedonks, N., Rutten, M.J.M., Vriezen, W.H., de Boer, G.J., Rieu, I., 2017. Mapping quantitative trait loci for heat tolerance of reproductive traits in tomato (Solanum lycopersicum). Mol Breed, 37: 1e9. Yang,D.,Wang,Z.,Huang,X.,Xu,C.,2023.Molecularregulation of tomato male reproductive development. aBIOTECH, 4: 72e82. Zhao, J., Jiang, L., Che, G., Pan, Y., Li, Y., Hou, Y., Zhao, W., Zhong, Y., Ding, L., Yan, S., Sun, C., Liu, R., Yan, L., Wu, T., Li, X., Weng, Y., Zhanga, X., 2019. A functional allele of CsFUL1 regulates fruit length through repressing CsSUP and inhibiting auxin transport in cucumber. Plant Cell, 31: 1289e1307. Zhu, Y., Tian, H.Q., 2015. The distribution features of polysaccharide and lipid drops in the development of tomato anther. Acta Horic Sin, 42: 672e678. 1904 Xiaowei Wang et al. 2025. Horticultural Plant Journal, 11 (5): 1896e1904.