SELF-PRUNING Acts Synergistically with DIAGEOTROPICA to Guide Auxin Responses and Proper Growth Form

The antiflorigenic signal SELF-PRUNING controls tomato growth habit by affecting auxin transport, signaling, and metabolism.

Abstract

The SELF PRUNING (SP) gene is a key regulator of growth habit in tomato (Solanum lycopersicum). It is an ortholog of TERMINAL FLOWER1, a phosphatidylethanolamine-binding protein with antiflorigenic activity in Arabidopsis (Arabidopsis thaliana). A spontaneous loss-of-function mutation (sp) has been bred into several industrial tomato cultivars, as it produces a suite of pleiotropic effects that are favorable for mechanical harvesting, including determinate growth habit, short plant stature, and simultaneous fruit ripening. However, the physiological basis for these phenotypic differences has not been thoroughly explained. Here, we show that the sp mutation alters polar auxin transport as well as auxin responses, such as gravitropic curvature and elongation of excised hypocotyl segments. We also demonstrate that free auxin levels and auxin-regulated gene expression patterns are altered in sp mutants. Furthermore, diageotropica, a mutation in a gene encoding a cyclophilin A protein, appears to confer epistatic effects with sp. Our results indicate that SP affects the tomato growth habit at least in part by influencing auxin transport and responsiveness. These findings suggest potential novel targets that could be manipulated for controlling plant growth habit and improving productivity.

Shoot architecture is a key agricultural trait determined mainly by side branching, internode elongation, and shoot determinacy (Wang and Li, 2008). Each of these parameters is an active research area where considerable theoretical and applied knowledge has been gained over the last decade. Shoot determinacy is a domestication trait found in several crop species, such as soybean (Glycine max), common bean (Phaseolus vulgaris), and tomato (Solanum lycopersicum; Pnueli et al., 1998; Tian et al., 2010; Repinski et al., 2012). Shoot determinacy is notably important for tomato, which is a perennial species cultivated as an annual crop. Wild tomatoes display indeterminate growth, resulting from a sequential addition of modules (sympodial units) formed by three leaves and an inflorescence. Sympodial growth starts in tomato when the vegetative apical meristem is converted into an inflorescence meristem after a series of eight to 12 internodes with leaves (Samach and Lotan, 2007). Vegetative growth, however, continues vigorously at the top-most axillary meristem, displacing the inflorescence to the side and producing a new sympodial unit with three leaves and an inflorescence. This process is iterated indefinitely by a concatenation of stacked sympodial units. However, a spontaneous recessive mutant with a compact, bushy growth habit and a reduced number of leaves in successive sympodial units was discovered in 1914 (Yeager, 1927; MacArthur, 1934). It was later shown that the mutation is a single-nucleotide substitution in the SELF-PRUNING (SP) gene (Pnueli et al., 1998), which shares sequence similarity with phosphatidylethanolamine-binding proteins (PEBPs), a group of mammalian polypeptides involved in cell signaling (Hengst et al., 2001; Kroslak et al., 2001). Breeding the SP mutation into industrial tomato cultivars was instrumental in the advent of mechanical harvest (Rick, 1978; Stevens and Rick, 1986). The loss-of-function sp mutant exhibits a determinate growth habit, as opposed to the indeterminate growth habit of wild-type tomatoes. The determinate growth habit occurs via a progressively reduced number of leaves per sympodium until termination in two consecutive inflorescences that top the vertical growth of the plant (Samach and Lotan, 2007). Hence, this phenotype leads to simultaneous fruit ripening, allowing mechanical harvest in field-grown processing tomatoes (Stevens and Rick, 1986).

SP belongs to the CETS gene family, which comprises CENTRORADIALIS (CEN) and TERMINAL FLOWER1 (TFL1) of Antirrhinum spp. and Arabidopsis (Arabidopsis thaliana), respectively (Wickland and Hanzawa, 2015). SINGLE FLOWER TRUSS (SFT)/SP3D, a homolog of FLOWERING LOCUS T and HEADING DATE3A in Arabidopsis and rice (Oryza sativa), respectively, is another CETS gene involved in controlling tomato growth habit (Alvarez et al., 1992; Kojima et al., 2002). Unlike sp mutants, which do not affect flowering time, tomato sft loss-of-function mutants are late flowering and show a disruption in sympodial growth pattern. These mutants produce a single and highly vegetative inflorescence as well as alternating solitary flowers and leaves (Molinero-Rosales et al., 2004). The final phenotypic outcome produced by SP and SFT depends on their local ratio, with the former maintaining meristems in an indeterminate state and the latter promoting the transition to flowering (Park et al., 2014). Heterozygous sft mutants in a homozygous sp mutant background display yield heterosis in tomato (Krieger et al., 2010). Hence, the SP/SFT genetic module has been proposed as a target to increase crop yield via changes in plant architecture (McGarry and Ayre, 2012; Zsögön et al., 2017). It also has been suggested that SP function could be linked to auxin (Pnueli et al., 2001), a hormone with strong effects on plant morphogenesis (Berleth and Sachs, 2001).

Auxin is a key controller of plant development; however, its role in the regulation of plant growth habit is still unclear. An aspect that sets auxin apart from other plant hormones is the relatively well-understood nature of its transport through the plant body (Friml, 2003; Petrášek and Friml, 2009). Polar auxin transport (PAT), which occurs basipetally from the apical meristem, is crucial for the distribution of auxin within plant tissues (Rubery and Sheldrake, 1974; Sheldrake, 1974). PAT works as an organizer of apical-basal polarity in the plant body (Friml et al., 2006), thus controlling a multiplicity of developmental processes (Reinhardt et al., 2003; Blilou et al., 2005; Scarpella et al., 2006).

It was shown recently that DIAGEOTROPICA (DGT) affects PAT in tomato (Ivanchenko et al., 2015). DGT is a cyclophilin A protein with peptidyl-prolyl trans/cis-isomerase (PPIase) enzymatic activity (Takahashi et al., 1989; Oh et al., 2006). Cyclophilins catalyze not only rate-limiting steps in the protein-folding pathway but also can participate in the folding process as molecular chaperones (Kumari et al., 2013). DGT function is highly conserved across plant taxa (Lavy et al., 2012). In tomato, some of the most significant phenotypic defects caused by the lack of functional DGT protein are horizontal shoot growth, thin stems, altered secondary vascular differentiation, and roots lacking lateral branches (Zobel, 1973; Muday et al., 1995; Coenen et al., 2003). Here, we investigated whether SP affects auxin responses by itself and in combination with DGT. We produced four combinations of functional and loss-of-function mutant alleles of SP and DGT (i.e. SP DGT, SP dgt, sp DGT, and sp dgt) in a single tomato genetic background (cv Micro-Tom [MT]) and assessed a series of physiological responses to auxin. We found that free auxin levels, PAT, and gravitropic curvature of the shoot apex are all altered by SP. Our results further show that SP and DGT reciprocally affect AUXIN/INDOLE-3-ACETIC ACID INDUCIBLE (Aux/IAA) and AUXIN RESPONSE FACTOR (ARF) transcript abundances at the sympodial meristem, the key niche of SP function in growth habit.

RESULTS

Comparison of the four combinations of homozygous wild-type and mutant lines for SP and DGT (i.e. SP DGT, SP dgt, sp DGT, and sp dgt) showed that growth habit was affected solely by SP and not by DGT (Fig. 1). Regardless of their DGT or dgt allele, SP plants showed indeterminate growth whereas sp mutants were always determinate (Fig. 1). Time to flowering, however, was affected by both genes in a combinatorial fashion: dgt plants flowered late, independently of the SP allele (Fig. 1). The sp DGT genotype showed consistently precocious flowering, and this was confirmed in an independent experiment by analysis of the rate of shoot apical meristem maturation (Supplemental Fig. S1). The number of leaves to the first inflorescence also was affected by the combination of alleles (Fig. 1), albeit not reflecting the time to flowering. The dgt mutant produced more leaves before flowering, but this effect was abolished in the double mutant sp dgt. Regardless of their SP allele, dgt mutants exhibited markedly reduced transcript abundance of the flowering inducer SFT compared with DGT plants (Fig. 1), which fits with the delayed flowering in these mutants in both SP and sp backgrounds (Fig. 1).

Figure 1.

Additive phenotype of the sp and dgt mutations in tomato cv MT. A, Representative plants of SP DGT, SP dgt, sp DGT (cv MT), and sp dgt at 90 d after germination (dag). Note the simultaneous fruit ripening in sp compared with SP, a well-known effect of the sp mutation. The dgt mutation delays fruit ripening (at least in part due to its late flowering, as indicated in B) in either genetic background. Bar = 5 cm. B, Chronological time to flowering in sp and dgt mutants. The percentage of plants (n = 15) with at least one open flower is shown. The cv MT (sp DGT) plants flower earlier than the wild-type plants (SP DGT), whereas dgt mutants are late flowering. C, Developmental time to flowering in sp and dgt mutants. The number of leaves produced before the first inflorescence was reduced in sp DGT (cv MT) and increased in genotypes carrying the functional allele of SP. Letters indicate statistically significant differences (Dunn’s multiple comparisons test, P < 0.05). D, sp and dgt alter the expression of the flowering inducer SFT. The dgt mutation leads to lower SFT expression and, thus, delays flowering. A minor influence from SP reducing SFT levels also is noticeable. Asterisks indicate statistically significant differences from the wild-type SP DGT (Student’s t test, P < 0.05). E, Effects of sp and dgt on side branching. Pie charts depict the distribution of side branches in each genotype at 60 dag (n = 15 plants). Gray denotes absence of an axillary bud, yellow denotes a visible bud (greater than 1 cm), and dark green denotes a full branch (with one or multiple leaves). Letters indicate statistically significant differences (Dunn’s multiple comparisons test, P < 0.05).

The tomato cv MT harbors a mutation in DWARF (D), a gene coding for a key enzyme in the brassinosteroid biosynthesis pathway (Bishop et al., 1999). Since brassinosteroids are known to influence the flowering induction network (Domagalska et al., 2010; Li et al., 2010), we ascertained whether D could be influencing the effects of SP on flowering time. Using a near-isogenic MT line harboring the functional D allele (Carvalho et al., 2011), we constructed four allelic combinations of SP and D (i.e. SP D, SP d, sp D, and sp d; Supplemental Fig. S2) and assessed their flowering time. The results show an effect of D on flowering time (Supplemental Fig. S2) but not on the number of leaves produced to the first inflorescence, which was again reduced exclusively by the presence of the sp mutant allele (Supplemental Fig. S2). Axillary branching was affected mainly by the SP gene, which led to reduced bud outgrowth in plants carrying the wild-type allele; the dgt mutation, however, exacerbated this repressive effect (Fig. 1). sp mutants, on the other hand, branched more profusely when combined with dgt than when combined with DGT (Fig. 1). Thus, dgt can enhance apical dominance or increase branching, depending on the presence or absence of a functional SP allele, respectively. The number of leaves on the primary shoot was increased by SP, regardless of the DGT allele (Table I). Plant height was additively controlled by both genes, whereas no difference between genotypes was found in the length of the fourth internode or the leaf insertion angle (Table I). Stem diameter was increased by functional DGT, irrespective of the SP allele (Table I). The number of inflorescences was determined synergistically by both genes, whereby the pairing of functional SP and DGT led to an increased number compared with all other allele combinations (Table I).

Table I. Parameters that define growth habit.

Parameters are as follows: number of leaves on the primary shoot (number of leaves up to the first inflorescence); height of the primary shoot (cm); internode length (cm); leaf insertion angle (°); diameter of the stem (mm); number of flowers per inflorescence and number of inflorescences. Measurements were performed at 60 dag. Data are means ± se (n = 10 plants). Different letters indicate statistically significant differences (Tukey’s test, P < 0.05) among genotypes.

Parameters	SP DGT	SP dgt	sp DGT	sp dgt
No. of leaves on the primary shoot	9.00 ± 0.36 a	9.09 ± 0.28 a	7.50 ± 0.48 b	7.67 ± 0.58 b
Height of the main shoot	17.45 ± 0.79 a	12.74 ± 0.61 b	13.45 ± 0.49 b	10.70 ± 1.07 b
Length of the fourth internode	1.08 ± 0.12 a	1.11 ± 0.07 a	1.28 ± 0.13 a	1.22 ± 0.22 a
Leaf insertion angle	74.81 ± 3.41 a	74.87 ± 3.26 a	73.01 ± 4.38 a	63.54 ± 5.22 a
Stem diameter	5.46 ± 0.26 a	4.36 ± 0.12 a,b	5.36 ± 0.21 a	4.45 ± 0.1 b
Flowers per inflorescence	7.00 ± 0.71 b	7.80 ± 0.45 a	7.00 ± 0.00 b	7.00 ± 0.5 b
No. of inflorescences	12.4 ± 1.14 a	9.20 ± 1.30 b	7.40 ± 0.89 b	8.80 ± 1.48 b

Next, the endogenous levels of free indolyl-3-acetic acid (IAA), which is the most abundant auxin in plants (Bartel and Fink, 1995), was determined in three sections of tomato seedlings: leaves plus cotyledons, hypocotyls, and roots (Fig. 2). In leaves plus cotyledons, IAA concentration was more than 2 times higher in the sp dgt double mutant than in the other three genotypes (Fig. 2). In the hypocotyl tissues, SP DGT seedlings had the lowest free IAA content, the sp and dgt single mutants had intermediate levels, and the double mutant (sp dgt) exhibited the highest IAA levels (Fig. 2). Although root IAA levels were clearly higher than in the other hypocotyl regions analyzed, no statistically significant differences in root IAA content were observed between the four genotypes (Fig. 2).

Figure 2.

Auxin levels in tomato seedlings are affected synergistically by the sp and dgt mutations. A, Representative 7-d-old seedling showing the dissection points for auxin quantitation (bar = 1 cm). B to D, Free IAA levels in leaves + cotyledons (B), hypocotyls (C), and roots (D). Data are means ± se (n = 10). Different letters indicate statistically significant differences (Tukey’s test, P < 0.05) among genotypes. FW, Fresh weight.

To understand the variation in endogenous free IAA levels within the seedling tissues and among the four genotypes, we next quantified PAT in hypocotyl segments. PAT was highest in SP DGT, intermediate in sp and dgt single mutants, and lowest in the double mutant (Fig. 3). This indicates that both sp and dgt alleles reduce PAT and that their effects are additive. As PAT and auxin concentration are known to influence vascular patterning (Scarpella, 2017), we also analyzed xylem anatomy in cross sections of stems in adult plants (Fig. 3). Quantification of xylem vessel density and mean vessel size revealed an antagonistic relationship between SP and DGT. Whereas SP tends to reduce vessel density and increase their size, DGT increases vessel density with concomitantly lower vessel sizes (Fig. 3). These results, however, obscure a more complex pattern, which is revealed when analyzing the vessel size distributions. The functional DGT allele increased the incidence of larger vessels (greater than 800 µm² cross-sectional area), particularly in sp mutant plants (Fig. 3). Another physiological response affected by PAT is negative gravitropism of the shoot (Morita, 2010). The kinetics of gravitropic curvature in seedling shoots was affected by both SP and DGT (Fig. 4). Loss of SP function decreased the gravitropic response in both DGT and dgt backgrounds. Hypocotyl elongation in response to exogenous auxin and in vitro rhizogenesis from cotyledon explants are assays to determine auxin sensitivity (Cary et al., 2001). The dgt mutation considerably reduces hypocotyl responsivity to auxin in all concentrations, as described previously (Kelly and Bradford, 1986; Rice and Lomax, 2000). The functional SP allele increased hypocotyl elongation in the DGT background and also exerted a significant compensatory effect on the elongation response in the dgt mutant (Fig. 4). In the in vitro root regeneration assay, as expected, root formation was reduced in dgt mutants (Coenen and Lomax, 1998) but also in sp compared with SP in the presence of a functional DGT allele (Supplemental Fig. S3).

Figure 3.

A, The sp mutation exacerbates defective PAT in hypocotyls caused by dgt. Basipetal [³H]IAA transport is shown in 10-mm hypocotyl sections of the wild type (SP DGT), SP dgt, sp DGT (cv MT; also the negative control treated with 1-N-naphthylphthalamic acid [NPA]), and double mutant sp dgt roots. Data are means ± se (n = 10). Letters indicate statistically significant differences between treatments (Tukey's test, P < 0.05). B to E, Vascular patterning in sp and dgt stems. Cross sections of the fifth internode taken at 45 dag are shown. Bars = 100 µm. F and G, Vessel density (F) and mean vessel size (G) in sp and dgt stems. Letters indicate significant differences (P < 0.05, ANOVA and Tukey’s test). H, Vessel size distribution in the xylem of sp and dgt mutants. The x axis shows the upper values of cross-sectional area for each vessel size category. The bars within each category represent a single individual plant (n = 4 per genotype).

Figure 4.

Impact of the sp mutation on auxin responses in planta. A, Kinetics of the gravitropic response in the shoot. Shoot angle is shown after placing plants horizontally at time point 0 (n = 5). B, Elongation of excised hypocotyls in response to naphthaleneacetic acid (NAA). Six-millimeter hypocotyl sections were incubated in the indicated NAA concentration for 24 h before measurement (n = 15). C and D, Time course of in vitro root elongation of seedlings in control and 10 µm NAA-containing Murashige and Skoog (MS) medium (n = 25). In all graphs, error bars indicate se and asterisks indicate statistically significant differences between SP and sp plants harboring the same DGT allele (*, P ≤ 0.05 and **, P ≤ 0.01, Student’s t test).

Histochemical analysis of DR5 promoter activity revealed no discernible staining difference in both SP and sp seedlings incubated in water, although roots of the sp mutant showed a shorter trace of GUS precipitate in the vascular cylinder (Fig. 5). Exogenous IAA, however, strongly induced GUS expression in SP compared with sp plants, which was evident both in seedlings and in root tips and confirmed by fluorimetric GUS quantitation (Fig. 5). Fainter GUS staining was observed for both auxin-treated and untreated roots in the dgt mutant (Supplemental Fig. S4). As PIN-FORMED (PIN) auxin efflux transporters are key players determining auxin distribution in plants, we quantified the relative expression of the PIN1, PIN2, and PIN3 genes in roots with or without prior auxin incubation. Auxin treatment induced PIN1 and PIN3 expression in all genotypes, except in the sp dgt double mutant (Fig. 5). PIN2 expression was reduced by auxin incubation in SP DGT, SP dgt, and sp dgt but not in sp DGT (i.e. cv MT), where a low basal level of expression was observed for all three genes.

Figure 5.

Effects of SP on the auxin signaling and transport machinery in planta. A, Expression of the GUS reporter driven by the auxin-inducible DR5 promoter. Representative wild-type (SP) and mutant (sp) seedlings (bars = 2 cm) and their root tips (bars = 250 µm) are shown in the absence or presence of exogenous auxin (20 µm IAA, 3 h) at 15 dag. B to D, Fluorimetric quantification of GUS precipitate. Seedlings were sampled at 15 dag, after treatment with exogenous auxin (20 µm IAA, 3 h) or mock solution. Values are means ± se (n = 4). Letters indicate significant differences between genotypes within the same treatment (P < 0.05, ANOVA and Tukey’s test). E to G, Relative gene expression of PIN transporters in roots. Letters indicate significant differences between genotypes within the same treatment (P < 0.05, ANOVA and Tukey’s test).

Finally, we determined whether auxin affects SP at the transcriptional level, as suggested by the presence of auxin-response elements (TGTCTC and the degenerate version, TGTCNC; Ulmasov et al., 1995) in the 3′ and 5′ flanking regions of the SP gene in tomato and related Solanaceae species (Fig. 6). Analyzing SP mRNA levels in seedlings of SP DGT and sp DGT plants sprayed with IAA or a mock solution revealed that SP expression was induced by IAA treatment in both genotypes. Importantly, SP transcript levels were significantly higher (Student’s t test, P < 0.05) in dgt mutant plants than in wild-type DGT, both in IAA-treated and control seedlings (Fig. 6). We further assessed the effect of SP and DGT on the mRNA levels of key players in the auxin signaling cascades, including some members of the ARF and Aux/IAA gene families, which were chosen by their high auxin inducibility (Audran-Delalande et al., 2012). SP and DGT had combinatorial effects on the expression levels of IAA1, IAA2, IAA9, ARF8, and ARF10, whereas functional DGT decreased the expression of IAA3 (Fig. 6).

Figure 6.

SP and auxin signaling gene expression is altered by the dgt mutation. A, Genomic structures of the SP gene in solanaceous species: tomato, its wild relatives Solanum pimpinellifolium and Solanum pennellii, and potato (Solanum tuberosum). The coding sequence is indicated in yellow (exons, thick bars; introns, thin bars). Red blocks indicate the presence of a conserved or degenerate auxin-response element (AuxRE), TGTCNC. B and C, Relative transcript accumulation of SP (B) and auxin signaling genes (C) in sympodial meristems. Tissues were sampled from 10-d-old plants 24 h after 10 µm IAA or mock spray. Asterisks indicate significant differences with respect to the wild-type SP DGT (P < 0.05, Student’s t test).

DISCUSSION

Impact of SP Alleles, and Their Interaction with Auxin, in the Control of Shoot Architecture

Although it has been demonstrated previously that the sp mutation does not alter tomato flowering time or the number of leaves before termination of the shoot (Pnueli et al., 1998; Shalit et al., 2009), SP orthologs vary in this respect depending on the species. Flowering time is not affected in cen mutants in Antirrhinum spp. (Bradley et al., 1996), whereas Arabidopsis tfl1 mutants flower earlier and TFL1 overexpression delays flowering by preventing the meristem transition from vegetative to floral (Ratcliffe et al., 1998). In soybean, where large intraspecific variation exists in time to flowering, association mapping recently linked this important agronomic trait to the Dt1 locus, a CEN/TFL1/SP ortholog (Zhang et al., 2015). Comparison of determinate and indeterminate near-isogenic soybean lines consistently showed earlier flowering in the former across different locations and planting seasons (Ouattara and Weaver, 1994). Our data show that loss of SP function (sp allele) leads to slightly but consistently earlier flowering in tomato, measured either in dag or the reduction of the number of nodes before the first inflorescence. Using a near-isogenic line harboring the wild-type D allele, which codes for a brassinosteroid biosynthesis gene (Martí et al., 2006), we demonstrate that this effect is not related to reduced brassinosteroid levels in cv MT. This is in agreement with the observation that the phenotypes of the sp and dgt individual mutants in the cv MT background closely resemble those published for the same mutations in other tomato cultivars (Carvalho et al., 2011). Therefore, it is unlikely that the combination of both mutations (sp and dgt) would be affected epistatically by the d allele (Campos et al., 2010). Interestingly, the dgt mutation delays the number of days to flowering in both SP and sp backgrounds (Balbi and Lomax, 2003), apparently by reducing the expression of the SFT gene, which encodes the florigen (Evans, 1971; Shalit et al., 2009). It does not, however, significantly affect the number of leaves produced before termination, which is a proven effect of SFT and its orthologs in tomato and other species (Kojima et al., 2002; Lifschitz and Eshed, 2006; Navarro et al., 2015). Hence, the loss-of-function sft mutant produced 130% more leaves on the primary shoot than the control cv MT (Vicente et al., 2015). Conversely, transgenic tomato plants overexpressing SFT flower after producing only three or four leaves (Molinero-Rosales et al., 2004; Shalit et al., 2009).

Axillary branching was increased in sp mutants in both DGT and dgt allele backgrounds. The expression of SP is higher in axillary meristems, suggesting a possible role for SP in the control of apical dominance (Thouet et al., 2008). Our results reinforce this notion, as sp mutants are more profusely branched than wild-type plants. This also agrees with the effects of the SP ortholog Dt1 in soybean, where comparison of determinate and indeterminate isogenic lines revealed an increased propensity to side branching in the former (Gai et al., 1984). The dgt mutant responds to auxin treatment of decapitated shoots, which inhibits bud outgrowth to the same extent as in wild-type plants (Cline, 1994). Apical dominance has been reported to be reduced in intact dgt plants (Coenen et al., 2003), but caution should be exercised when interpreting these results, as published work on dgt has been conducted in tomato cultivars differing in their SP alleles (Supplemental Table S1). Our results indicate a strong and complex interaction between SP and DGT in the control of apical dominance: the dgt mutation increased it in the wild-type SP background but also increased axillary bud outgrowth in the sp background, enhancing its branching phenotype.

Control of Endogenous Auxin Levels and PAT by SP and DGT

Endogenous IAA concentration and distribution within tissues determine a wide range of plant developmental processes, including apical dominance, stem growth, vascular patterning, root development, and others (Petrášek and Friml, 2009; Ljung, 2013). IAA synthesis is maximal in younger, developing parts of the plant such as leaflets and root apices (Ljung et al., 2001). IAA levels in dark-grown seedlings (Fujino et al., 1988) and roots (Muday et al., 1995) of sp dgt and SP DGT plants are indistinguishable. However, free IAA levels in aerial parts of 7-d-old light-grown seedlings of sp dgt plants were twice as high as in SP DGT plants (Fig. 2), suggesting a light-dependent, synergistic influence of the sp and dgt alleles on auxin synthesis, degradation, or transport.

The above results could reflect changes in IAA biosynthesis, degradation, or transport. The reduction in PAT produced by the dgt mutation was described previously (Ivanchenko et al., 2015), but the synergistic effect of the sp mutation described here was unexpected. The differences in IAA concentration in the aerial part of the seedlings could be due to altered PAT caused by both the sp and dgt mutations. PAT from the shoot organs to the root tips induces the formation of the entire plant vascular system (Aloni, 2013; Marcos and Berleth, 2014), as evidenced by the repression of protoxylem formation upon treatment with the auxin transport inhibitor NPA (Bishopp et al., 2011) and the polar localization of PIN1 in preprocambial cells (Scarpella et al., 2006). A lack of large secondary xylem vessels was conspicuous in the dgt mutant, as described previously (Zobel, 1974). In plants harboring the functional DGT allele, the sp mutation led to larger (greater than 800 µm² cross-sectional area) vessels compared with wild-type SP plants. In tree species, there is evidence that the relationship between xylem vessel density and size involves differential regulation of the duration of tracheid expansion along the longitudinal (Anfodillo et al., 2012; Sorce et al., 2013) and radial (Tuominen et al., 1997) axes. Tree stature has a strong influence on vessel width due to an allometric scaling effect (Morris et al., 2018). It remains to be seen if this also is the case in herbs and if the effect of the SP gene on xylem width is caused indirectly by its control of plant height or directly by its influence on PAT. Increased PAT in the polycotyledon tomato mutant, for instance, leads to an altered vascular pattern in the hypocotyl (Al-Hammadi et al., 2003; Kharshiing et al., 2010).

SP Affects Excised Hypocotyl Elongation, Gravitropic Responses, and Root Regeneration and Elongation

The elongation of excised hypocotyl segments in response to different concentrations of exogenous auxin is a classical assay for auxin sensitivity (Gendreau et al., 1997; Collett et al., 2000). The hypocotyl elongation response of dgt has been described in the background of tomato cv VFN8, a mutant for sp (Supplemental Table S1). In both intact or excised hypocotyl segments, a reduced response to exogenous auxin was observed for the sp dgt double mutant (Kelly and Bradford, 1986; Rice and Lomax, 2000). We confirmed these results but show that a functional SP allele leads to increased elongation in either the DGT or dgt background. Collectively, these results indicate that some compensatory effect can be ascribed to SP in this response. Hypocotyl elongation in Arabidopsis relies on auxin-induced changes in the activity of plasma membrane H⁺-ATPases, which leads to increased H⁺ extrusion and cell expansion, through expansin-mediated cell wall loosening, according to the acid growth theory (Takahashi et al., 2012). Hypocotyl elongation upon exogenous auxin application points to a positive effect of SP on the activity of plasma membrane H⁺-ATPases. Interestingly, the activity of both plasma membrane H⁺-ATPases and PIN efflux transporters, which also are influenced by SP at the transcriptional level (Fig. 4), is regulated by changes in their phosphorylation state (Takahashi et al., 2012; Zourelidou et al., 2014; Weller et al., 2017). This fits with earlier suggestions that SP, which encodes a PEBP, exerts at least some of its effects on membrane proteins through interaction with kinases (Pnueli et al., 2001).

The Cholodny-Went hypothesis is a classical model suggesting that differential auxin distribution is the cause of directional plant bending with respect to an exogenous stimulus such as light or gravity (Went, 1974). DGT is required for a correct gravitropic response of roots and shoots, but the explanation at the molecular level is still lacking (Muday et al., 1995; Rice and Lomax, 2000). Functional SP enhances shoot gravitropism in horizontally positioned plants of either the DGT or dgt background. Functional SP produces taller plants, so it is tempting to speculate that they should have a stronger gravitropic response to facilitate the bending of a larger stem. In Arabidopsis, the IAA efflux transporter PIN3 mediates the lateral redistribution of auxin and, therefore, is involved in hypocotyl and root tropisms (Friml et al., 2002). It seems reasonable to suggest a link between SP and PIN3 in the face of our PIN gene expression profiles (Fig. 5). Remarkably, both types of efflux transporters, PIN1 and PIN3, have been shown to relocate at the subcellular level via the same mechanism: vesicle trafficking along the actin cytoskeleton between the plasma membrane and endosomes (Geldner et al., 2001; Friml et al., 2002). Dissecting the intertwined mechanisms involved in this possible coregulation will be required to fully understand the extent to which and exactly how SP affects auxin distribution.

High concentrations of exogenous auxin inhibit root elongation. As expected, the dgt mutation reduced auxin-induced inhibition (Coenen and Lomax, 1998) in root elongation; however, a functional SP allele led to lower inhibition than in the double sp dgt mutant. This result could be ascribed to a new balance in auxin transport and signaling produced by the combination of SP and DGT. Root growth increases with the strength of auxin signaling up to a certain optimum and then begins to decline, probably following a parabolic trajectory (Sibout et al., 2006). In vitro root regeneration, on the other hand, is stimulated by low concentrations of auxin, and dgt is relatively insensitive to this exogenous treatment (Coenen and Lomax, 1998). Interestingly, the sp mutation also reduces rhizogenesis (Supplemental Fig. S3), which reinforces the notion of SP positively influencing PAT, as the PAT inhibitor 2,3,5-triiodobenzoic acid has been shown to reduce in vitro root formation in tomato (Tyburski and Tretyn, 2004).

Interactions between SP and the Auxin Signaling Machinery

Auxin signaling output can be estimated by following the pattern of DR5 promoter activation (Ulmasov et al., 1995; Liao et al., 2015). For example, GUS staining of DR5::GUS revealed that auxin flux at the root tips proceeds acropetally up to the root cap, where it is redistributed via lateral efflux transporters toward a peripheral basipetal transport route (Benková et al., 2003; Paciorek et al., 2005). Exogenous auxin application leads to greater GUS signal in seedlings with a functional SP allele, probably owing to alterations in the auxin signaling machinery produced by SP, such as the expression of Aux/IAA and ARF family genes (Fig. 6).

Auxin signaling is strongly dependent on auxin levels and the responsiveness of target cells. At low IAA levels, a suite of repressor proteins, including Aux/IAA and TOPLESS, repress ARFs, a group of transcription factors that regulate the expression of auxin-responsive genes (Causier et al., 2012; Bargmann and Estelle, 2014; Chandler, 2016). At high IAA levels, auxin acts as a molecular glue to stabilize the TIR1/AFB receptor binding and tagging of Aux/IAAs for 26S proteasome degradation (Hayashi, 2012). This, in turn, frees ARFs bound to AuxRE in the genome (TGTCTC or its degenerate, but also functional, form, TGTCNC) to activate or repress gene expression (Ulmasov et al., 1995). Our in silico analyses demonstrated the presence of conserved AuxRE elements both 5′ (upstream) and 3′ (downstream) of the SP coding sequence in the genome of tomato and closely related species. The 3′ region of TFL1 in Arabidopsis contains multiple auxin cis-regulatory elements key for the control of spatiotemporal expression of the gene (Serrano-Mislata et al., 2016). It remains to be determined whether such cis-regulatory elements also are functional in tomato and if they are involved in the response of SP expression to auxin.

Tomato has 25 Aux/IAA and 22 ARF genes (Audran-Delalande et al., 2012; Zouine et al., 2014), indicating that auxin signaling is very complex. DGT can alter the expression of genes related to auxin signaling (Mito and Bennett, 1995), including Aux/IAA genes (Nebenführ et al., 2000). PPIases catalyze the cis/trans-isomerization of peptide bonds preceding Pro residues of target peptides, including Aux/IAAs (Jing et al., 2015). Only Aux/IAA peptides of the right conformation can bind to the TIR1 receptor and be tagged for degradation, and PPIases, such as DGT, are believed to play a key role in auxin perception (Su et al., 2015). It is likely that some transcriptional feedback exists when the right conformers are not produced, as suggested by the increased transcript levels of IAA3 and the reduced levels of IAA9 in dgt mutants. Furthermore, our in silico analysis shows that the sp mutation occurs in a highly conserved cis-Pro residue in a DPDxPxn10H consensus region in the PEBP domain (Supplemental Fig. S5), which is a potential target for PPIases. Whether this putative molecular interaction between SP and DGT could account for the phenotypic outcomes shown here remains to be determined.

CONCLUSION

Auxin gradients are critical for organogenesis in the shoot apex; however, the influence of this hormone on shoot determinacy, which is a key determinant of growth habit, remains unclear. Our data link auxin and the antiflorigenic protein SP, the main switch between indeterminate and determinate growth habits in tomato. Although it is not clear whether auxin itself can affect growth habit, a physiological interaction between this hormone and members of the CETS family was clearly demonstrated here. Hence, SP alleles affected various auxin-related responses (e.g. apical dominance, PIN1-mediated PAT, vascular differentiation, H⁺ extrusion, and gravitropism responses), different SP orthologs presented AuxREs, and the auxin mutant dgt down-regulated SFT and up-regulated SP expression. Increasing evidence shows that the SP/SFT genetic module is a hub in crop productivity, affecting heterosis for yield (Krieger et al., 2010) and improving plant architecture and the vegetative-to-reproductive balance (McGarry and Ayre, 2012; Vicente et al., 2015; Zsögön et al., 2017). Our results suggest that at least part of the effect of the SP/SFT module on yield is mediated by auxin. This knowledge may inspire novel and more precise manipulation of this hormone for applications in agriculture.

MATERIALS AND METHODS

Plant Material

Seeds of tomato (Solanum lycopersicum ‘MT’) were donated by Dr. Avram Levy (Weizmann Institute of Science) in 1998 and subsequently maintained (through self-pollination) as a true-to-type cultivar. The cv MT seeds carrying the synthetic auxin-responsive (DR5) promoter fused to the reporter gene uid (encoding a GUS) were obtained from Dr. José Luiz García-Martínez (Universidad Politécnica de Valencia). The dgt mutation was introgressed into cv MT from its original background in cv VFN8 (LA1529), donated by Dr. Roger Chetelat (Tomato Genetics Resource Center, University of California, Davis). The functional allele of SP was introgressed from cv Moneymaker (LA2706).

The introgression of mutations into cv MT was described previously (Carvalho et al., 2011). A comparison between indeterminate (SP/SP) and determinate (sp/sp) plants in the cv MT background has been published previously (Vicente et al., 2015). Both sp and dgt mutations were confirmed by cleaved-amplified polymorphic sequence (CAPS) marker analyses and sequencing. All experiments were conducted on BC₆F₃ plants or subsequent generations (Sestari et al., 2014). In vitro seedling cultivation was conducted under controlled conditions (16-h/8-h day/night, approximately 45 µmol m⁻² s⁻¹ photosynthetically active radiation, and 25°C ± 1°C) in flasks with 30 mL of half-strength Murashige-Skoog medium gellified with 0.5% (w/v) agar, pH 5.8. Seeds were surface sterilized by agitation in 30% (v/v) commercial bleach (2.7% [w/v] sodium hypochlorite) for 15 min followed by three rinses with sterile distilled water.

Growth Conditions

Plants were grown in a greenhouse in Viçosa (642 m above sea level, 20°45′S, 42°51′W), Minas Gerais, Brazil, under semicontrolled conditions: mean temperature of 28°C, 11.5-h/13-h (winter/summer) photoperiod, 250 to 350 μmol m⁻² s⁻¹ PAR, and irrigation to field capacity twice per day. Seeds were germinated in 350-mL pots with a 1:1 (v/v) mixture of commercial potting mix (Basaplan; Base Agro) and expanded vermiculite supplemented with 1 g L⁻¹ 10:10:10 NPK and 4 g L⁻¹ dolomite limestone (MgCO₃ + CaCO₃). Upon appearance of the first true leaf, seedlings of each genotype were transplanted to pots containing the soil mix described above, except for the NPK supplementation, which was increased to 8 g L⁻¹.

Auxin Quantification

Endogenous IAA levels were determined by gas chromatography-tandem mass spectrometry-selecting ion monitoring (Shimadzu model GCMS-QP2010 SE). Samples (50–100 mg fresh weight) were extracted and methylated as described (Rigui et al., 2015). About 0.25 μg of the labeled standard [¹³C₆]IAA (Cambridge Isotopes) was added to each sample as an internal standard. The chromatograph was equipped with a fused-silica capillary column (30 m i.d., 0.25 mm, 0.5-μm-thick internal film) DB-5 MS stationary phase using helium as the carrier gas at a flow rate of 4.5 mL min^–1 in the following program: 2 min at 100°C, followed by a ramp of 10°C min^–1 to 140°C, 25°C min^–1 to 160°C, 35°C min^–1 to 250°C, 20°C min^–1 to 270°C, and 30°C min^–1 to 300°C. The injector temperature was 250°C, and the following mass spectrometer operating parameters were used: ionization voltage, 70 eV (electron impact ionization); ion source temperature, 230°C; and interface temperature, 260°C. Ions with mass-to-charge ratios of 130 and 189 (corresponding to endogenous IAA) and 136 and 195 (corresponding to [¹³C₆]IAA) were monitored, and endogenous IAA concentrations were calculated based on extracted chromatograms at mass-to-charge ratios of 130 and 136.

PAT Analysis

PAT was assayed in hypocotyl segments of 2-week-old seedlings according to the protocol originally described by Al-Hammadi et al. (2003), with some modifications. Briefly, 10-mm hypocotyl sections were excised and incubated in 5 mm phosphate buffer (pH 5.8) containing 1 µm IAA for 2 h at 25°C ± 2°C on a rotary shaker (200 rpm). These segments were placed between receiver blocks (1% [w/v] agar in water) and donor blocks (1% [w/v] agar in 5 mm phosphate buffer [pH 5.8] containing 1 µm IAA and 100 nm [³H]IAA) oriented with their apical ends toward the donor blocks. After 4 h of incubation inside a humid chamber at 25°C ± 2°C, the receiver blocks were removed and stored in a 3-mL scintillation cocktail (Ultima Gold; PerkinElmer). Receiver blocks plus scintillation cocktail were shaken overnight at 100 rpm and 28°C ± 2°C before analysis in a scintillation counter. As a negative control, some hypocotyl segments were sandwiched for 30 min between NPA-containing blocks (1% [w/v] agar in water containing 20 µm NPA) prior to the auxin transport assays. ³H dpm was converted to fmol of auxin transported as described by Lewis and Muday (2009).

Auxin Sensitivity Assays

Root regeneration from cotyledon explants was conducted as described previously (Cary et al., 2001). Briefly, cotyledon explants were obtained from 8-d-old seedlings germinated in vitro in one-half-strength MS medium. The explants were then incubated on petri dishes containing MS medium with or without supplementation with 0.4 µm NAA. After 8 d, the number of explants with visible roots (determined using a magnifying glass) was counted.

For hypocotyl elongation assays, hypocotyls were excised from 2-week-old seedlings and cut into 5-mm sections. Between 15 and 20 segments were preincubated for each treatment and floated on buffer (10 mm KCl, 1 mm MES-KOH [pH 6], and 1% [w/v] Suc) for 2 h at 25°C in the dark to deplete endogenous auxin. Segments were then incubated on buffer (10 mm KCl, 1 mm MES-KOH [pH 6], 1%[w/v] Suc, and 0.4 μM NAA) for 24 h on a shaker at 25°C under white light. Segments were photographed to determine their length using ImageJ (NIH). The experiment was repeated three times with similar results.

For the gravitropism assays, plants were germinated in 350-mL pots and transferred to 50-mL Falcon tubes for 2 dag. The gravitropic response was assessed at 10 dag by placing five plants of each genotype horizontally and photographing them in 30-min intervals. The angle of shoot bending at each time point was determined using AutoCad 2016 (Autodesk). Sterilized seeds were germinated in petri dishes onto two layers of filter paper moistened with distilled water and incubated for 4 d at 25°C in the dark. Ten germinated seeds with radicles of 5 to 10 mm were transferred to vertically oriented square petri dishes (120 mm × 120 mm) aligned on each plate with the radicles pointing down. The plates contained MS medium supplemented with vitamins, pH 5.7, 3% (w/v) Suc, 0.8% (w/v) agar, and 10 μm NAA for the auxin treatment. Plates were incubated in a growth chamber in the dark.

In vitro root elongation in response to exogenous auxin was assessed as follows. Seeds were surface sterilized and imbibed for 2 d at 4°C in the dark on agar plates containing one-half-strength MS growth medium (Murashige and Skoog, 1962) and then transferred to a growth chamber under control conditions (12-h photoperiod, 150 μmol m⁻² s⁻¹ white light, 22°C/20°C throughout the day/night cycle, and 60% relative humidity). After 4 d, 10 seedlings per plate were transferred to one-half-strength MS medium with or without 10 µm NAA (Sigma-Aldrich) and covered completely with aluminum foil for 8 d. Root elongation was assessed every 2 d under dim light conditions.

Histochemical Assays

Transgenic DR5::GUS plants were incubated overnight at 37°C in GUS staining solution [100 mm NaH₂PO₄, 10 mm EDTA, 0.5 mm K₄Fe(CN)₆, 0.05% v/v Triton X-100, and 1 mm 5-bromo-4-chloro-3-indolyl-β-d-GlcA]. Following GUS staining, samples were washed in a graded ethanol series to remove chlorophyll. Samples were then photographed using a Leica S8AP0 microscope set to 80× magnification coupled to a Leica DFC295 camera. Quantitative GUS activity was assayed according to Jefferson et al. (1987) with some modifications. Briefly, samples were ground in liquid nitrogen and subsequently homogenized in MUG extraction buffer composed of 50 mm HEPES-KOH (pH 7), 5 mm DTT, and 0.5% (w/v) polyvinylpyrrolidone. After centrifugation, 200-μL aliquots of the supernatant were mixed with 200 μL of GUS assay buffer composed of 50 mm HEPES-KOH (pH 7), 5 mm DTT, 10 mm EDTA, and 2 mm 4-methylumbelliferyl-β-d-glucuronide (MUG) and incubated at 37°C for 30 min. Subsequently, aliquots of 100 μL were taken from each tube, and the reactions were stopped and fluorescence was analyzed using a spectrofluorometer (LS55; PerkinElmer) with 365-nm excitation and 460-nm emission wavelengths (5-nm bandwidth).

Gene Expression Analyses

Total RNA was extracted from approximately 30 mg fresh weight of sympodial meristems of 10-d-old plants following the protocol of the manufacturer (Promega SV total RNA isolation system). For auxin treatments, plants were sprayed previously with 10 µm IAA or mock sprayed 24 h prior to RNA extraction. Four biological replicates were used for subsequent cDNA synthesis, where each replicate consisted of a pool of three plants. Each replicate was used for the analyses, since sympodial meristems are small and did not provide enough biological material for RNA extraction. Two technical replicates were then performed on each of the four samples. RNA integrity was analyzed on a 1% w/v agarose gel, and RNA concentration was estimated before and after treatment with DNase I (Amplification Grade DNase I; Invitrogen). Total RNA was transcribed into cDNA using the enzyme reverse transcriptase and the Universal RiboClone cDNA Synthesis System (Promega), following the manufacturer’s protocols.

For gene expression analyses, Power SYBR Green PCR Master Mix was used on MicroAmp Optical 96-well reaction plates (both from Applied Biosystems) and MicroAmp Optical adhesive film (Applied Biosystems). The number of reactions from the cycle threshold (CT) as well as the efficiency of the reaction were estimated using the Real-Time PCR Miner tool (Zhao and Fernald, 2005).

Relative expression was normalized using ACTIN and UBIQUITIN; ACTIN was used to calculate ΔΔCT assuming 100% efficiency of amplification of genes (2^−ΔΔCT). Primer sequences used are shown in Supplemental Table S2. Melting curves were checked for unspecific amplifications and primer dimerization.

In Silico Sequence Analyses

SP gene alignments was performed using the ClustalW alignment option of the Geneious R9 (Biomatters) software package.

Statistical Analysis

ANOVA and Tukey’s honestly significant difference test were performed using Assistat 7.6 beta (http://assistat.com). Percentage data were converted to inverse function (1/X) before analysis.

Accession Numbers

Gene accession numbers are as follows: JN379431 (IAA1), JN379432 (IAA2), JN379433 (IAA3), JN379437 (IAA9), EF667342 (ARF8), JF911788 (ARF10), U84140 (SP), AY186735 (SFT), HQ127074 (PIN1), HQ127077 (PIN2), HQ127079 (PIN3), NM_001247559 (DGT), NM_001308447 (ACTIN), and X58253 (UBIQUITIN).

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. The rate of meristem maturation is affected by the combination of SP and DGT alleles.

• Supplemental Figure S2. The effects of SP on flowering are not dependent on brassinosteroid biosynthesis.

• Supplemental Figure S3. Impact of the sp mutation on in vitro root regeneration.

• Supplemental Figure S4. Expression pattern of the GUS reporter driven by the auxin-inducible DR5 promoter in the sp dgt double mutant.

• Supplemental Figure S5. The tomato sp mutation occurs in a highly conserved Pro residue.

• Supplemental Table S1. List of publications featuring the tomato dgt mutant.

• Supplemental Table S2. Primers used for quantitative PCR analyses in this work.