Summary
The variability in leaf form in nature is immense. Leaf patterning occurs by differential growth that occurs during a limited window of morphogenetic activity at the leaf marginal meristem. While many regulators have been implicated in the designation of the morphogenetic window and in leaf patterning, how these effectors interact to generate a particular form is still not well understood.
We addressed the interaction among different effectors of tomato compound leaf development, using genetic and molecular analyses.
Mutations in the tomato auxin response factor SlARF5/SlMP, which promotes leaflet formation, suppressed the increased leaf complexity of mutants with extended morphogenetic window. Impaired activity of the NAC/CUC transcription factor GOBLET (GOB), which specifies leaflet boundaries, also reduced leaf complexity in these backgrounds. Analysis of genetic interactions showed that the patterning factors SlMP, GOB and the MYB transcription factor LYRATE (LYR) act in parallel to promote leaflet formation.
This work places an array of developmental regulators in a morphogenetic context. It reveals how organ-level differentiation rate and local growth are coordinated to sculpture an organ. These concepts and findings are applicable to other plant species and developmental processes that are regulated by patterning and differentiation.
Introduction
Leaf shape ranges from simple leaves, with a single leaf blade, to compound leaves, in which the leaf is composed of separate blade units termed leaflets (Bar & Ori, 2015; Du et al., 2018; Efroni et al., 2010). Leaf-primordia margins maintain a transient window of morphogenetic activity (Alvarez et al., 2016). The elaboration of compound leaves requires a prolonged morphogenetic window (Hagemann & Gleissberg, 1996), and tuning this window enables a species-dependent variability in leaflet number (Blein et al., 2010; Chen et al., 2010; Hagemann & Gleissberg, 1996; Ori et al., 2007; Zhou et al., 2014). In tomato, the TCP transcription factor LANCEOLATE (LA) and the MYB transcription factor CLAUSA (CLAU) promote differentiation, thus restricting the morphogenetic window at the leaf margin (Bar et al., 2015, 2016; Dengler, 1984; Jasinski et al., 2007; Kang & Sinha, 2010; Maltnan & Jenkins, 1962; Ori et al., 2007). In contrast, the MYB transcription factor TRIFOLATE (TF) and the tomato KNOTTED1-LIKE HOMEOBOX (KNOXI/TKN2) proteins delay leaf differentiation and preserve the meristematic identity of the leaf margin (Bharathan et al., 2002; Blein et al., 2013; Chen et al., 2010; Hake et al., 2004; Hay & Tsiantis, 2009; He et al., 2020; Janssen et al., 1998; Kimura et al., 2008; Naz et al., 2013; Peng et al., 2011; Shani et al., 2009; Zhou et al., 2014).
Both simple and compound leaf shape is patterned at the meristematic leaf margin by localized differential growth, producing serrations, lobes, or leaflets (Barkoulas et al., 2008; Bar & Ori, 2015; Bilsborough et al., 2011; Efroni et al., 2010; Nikovics et al., 2006; Kawamura et al., 2010). The formation of distinct leaflets involves the definition of regions of leaflet initiation and blade growth, alongside intercalary and boundary domains in which growth is inhibited (Fig1a and (Ben-Gera et al., 2012, 2016; Bilsborough et al., 2011; Blein et al., 2008; Fleming, 2006; Koenig et al., 2009; Vlad et al., 2014)). The plant hormone auxin plays a central role in this patterning mechanism, together with transcription factors such as the CUP-SHAPED COTYLEDONS (CUC) family, which interact with auxin in the domain specification process (Ben-Gera et al., 2012; Bilsborough et al., 2011; Blein et al., 2008; Kierzkowski et al., 2019; Xiong et al., 2019). In tomato, the Auxin Response Factor (ARF) SlMP/SlARF5 was recently shown acts downstream to auxin to promote leaflet initiation and growth together with additional class A ARF proteins. The activity of these ARFs is antagonized in the intercalary domain by the AUX/IAA protein ENTIRE (E)/IAA9 (Israeli et al., 2019). Leaflet initiation and growth is also promoted by the MYB transcription factor LYRATE (LYR) (David-Schwartz et al., 2009). While we learned a lot in recent years about the individual functions of these factors in leaf development, it is still not clear how leaf patterning regulators such as auxin and SlMP, LYR and CUC, interact with regulators of the transient morphogenetic window of the leaf margin, such as KNOXI, TF, CLAU and LA. Furthermore, it is not clear how the regulators of the growth, intercalary and boundary domains interact to coordinate leaf patterning.
Here, we examined the genetic and molecular interactions between the described effectors, to investigate how patterning and differentiation interact to achieve leaf shape diversity. We show that an extended morphogenetic window and specific patterning events are both required for stable leaflet production. We further show that the different regulators of the growth and boundary domains act via parallel pathways to pattern leaflets. Therefore, a coordinated network has developed to enable flexible leaf elaboration.
Materials and methods
Plant material and growth conditions
Tomato seeds (Solanum lycopersicum cv M82) were germinated and grown for three to four weeks in a growth room or a growth chamber in a 16/8 light/dark regime at 25C0 under fluorescent light. Seedlings were then transferred to a greenhouse or to an open field with natural day length and 250C/200C day/night temperature. The La-2, la-6, clau, Gob4d, gob-3, lyr, slmp-1 and slmp-2 alleles are from a tomato EMS mutagenesis populations, in the M82 background, and have been described before (Bassel et al., 2008; Bar et al., 2015; Berger et al., 2009; Jasinski et al., 2008; Menda et al., 2004; Ori et al., 2007). The Pts (LA2532) and bip (LA0663) mutants were obtained from the TGRC and backcrossed to M82. The transgenic lines 35S:Kn1, BLS≫TKN2, FIL≫miR164, FIL≫miR319 and the DR5:VENUS were generated in the M82 background and described before (Berger et al., 2009; Hareven et al., 1996; Ori et al., 2007; Shani et al., 2009, 2010). EMS Mutant screens were carried out on M82, La-2/+ gob/+ and la-6 seeds as described before (Menda et al., 2004; Ori et al., 2007).
Trans-activation system
We used the Promoter:LhG4 (p) and Operator (OP) system as described before (Moore et al., 1998; Shani et al., 2009). Briefly, in this system, driver lines expressing the synthetic transcription factor LhG4 under the control of a specific promoter are crossed to responder lines containing a gene of interest under the control of the E.coli operator, which is recognized by the LhG4 transcription factor but not by any endogenous plant transcription factor. A cross between a driver and a responder line produces an F1 plant in which the selected gene is expressed under the control of the selected promoter (p≫GENE).
RNA extraction and qRT-PCR analysis
RNA was extracted using the Plant/Fungi Total RNA Purification Kit (Norgen Biotek, Thorold, ON, Canada) according to the manufacturer’s instructions, including DNase treatment. cDNA synthesis was performed using the Verso cDNA Kit (Thermo Scientific, Waltham, MA, USA) or SuperScript II reverse transcriptase (18064014; Invitrogen, Waltham, MA, USA) using 1 mg of RNA. qRT-PCR analysis was carried out using a Corbett Rotor-Gene 6000 real-time PCR machine, with SYBR Premix for all other genes. Levels of mRNA were calculated relative to the EXPRESSED (EXP) or TUBULIN (TUB) genes as internal controls as follows: in each biological repeat, the expression levels of the assayed gene and EXP/TUB were separately calculated relative to a standard curve obtained by a dilution series of a reference sample. The gene expression level in each biological repeat was calculated by dividing the gene expression value by that of EXP/TUB. Average expression values were then calculated and presented as ‘relative gene expression’. Each biological repeat included between 3-15 leaf primordia, depending on the developmental stage. Primers used for the qRT-PCR analysis are detailed in Table S2.
Phenotyping, Imaging and Scanning Electron Microscopy (SEM)
Images analyzing the early developmental stages of the whole leaf primordium were captured using an Olympus SZX7 stereo microscope (http://www.olympus.com/) equipped with a Nikon DXM1200 camera and ACTA software, or a Nikon SMZ1270 stereo micro- scope equipped with a Nikon DS-Ri2 camera and NIS-ELEMENTS software. The expression pattern of the DR5::VENUS reporter was detected by a Stereomicroscope, as described before (Shani et al., 2010; Bar et al., 2016). For scanning electron microscopy (SEM), tissues were fixed in 30% Ethanol and vacuumed for 10 min, followed by dehydration in an increasing ethanol series up to 100% ethanol. Fixed tissues were critical-point dried, mounted on a copper plate and coated with gold. Samples were viewed using a JEOL 5410 LV microscope (Tokyo, Japan).
Leaf quantification
Phenotyping and quantification of leaf form, petiole length and shoot architecture were performed on field- or greenhouse-grown plants. Collected Representative mature intact leaves or mature plants, were photographed using a Nikon D5200 camera and the photographs used for quantification of leaf-shape phenotype. Leaflet order was defined, and leaflet number quantified as described before (Bar et al., 2015; Shani et al., 2010; Yanai et al., 2011). Briefly, primary leaflets are separated by a rachis, and some of them develop secondary and tertiary leaflets. Intercalary leaflets are lateral leaflet that develop from the rachis later than the primary leaflets and between them. Each genotype was represented by at least 3 biological replicates, consisting of leaves from different plants. Mean values were statistically analyzed using the Student’s t-test. Student’s t-test (two-tailed) was used for comparison of means, which were deemed significantly different at pv - 0.05. Images were manipulated uniformly using adobe Photoshop.
Accession Numbers
Sequence data used in this study can be found in the Sol Genomic Network under the following accession numbers: SlMP - solyc04g081240; ENTIRE/SlIAA9 - Solyc04g076850; LYR - Solyc05g009380; GOBLET - Solyc07g062840; LANCEOLATE - Solyc07g062680; CLAUSA - Solyc04g008480; TKN2 - Solyc02g081120; TRIFOLIATE - Solyc05g007870; PETROSELINUM - Solyc06g072480.
Results
SlMP and LYRATE promotes growth in parallel pathways
We have previously shown that the tomato ARF transcription factor SlMP/SlARF5 promotes organ initiation and growth (Israeli et al., 2019). Previous studies have shown that the MYB transcription factor LYRATE (LYR) also promotes leaflet initiation and blade growth (Bar et al., 2015; David-Schwartz et al., 2009). lyr mutants have less primary and intercalary leaflets, similar to slmp mutants (Fig 1), while plants that overexpress LYR have ectopic blade growth in the intercalary domain, similar to entire (e) mutants, in which auxin response and SlMP/SlARF5 activity are enhanced in the intercalary domain (David-Schwartz et al., 2009). We examined how SlMP/SlARF5 and LYR interact to promote the growth domain, and whether they act in the same genetic pathway, by investigating their genetic and molecular interactions. Strikingly, lyr and slmp enhanced each other, with the double mutants showing further reduction in primary leaflet number in comparison to the single mutants (Fig 1b-e). lyr slmp leaves had a range of phenotypes, with the most severe leaves being flattened, nearly bladeless, and lacking leaflets (Fig 1b-e and S1). The most severe phenotypes occurred in early leaves and leaves of axillary branches. The substantial enhancement of the single mutants suggests that SlMP and LYR promote leaflet initiation and growth via parallel pathways (Fig 1o). It was previously shown that lyr partially suppresses the ectopic blade phenotype of e in the VF36 background (David-Schwartz et al., 2009). We generated a e lyr double mutant in the M82 background and compared it to the e slmp double mutant. This comparison highlighted the differences between the two double mutants: While e and slmp mutually suppressed each other, restoring a wild-type leaf form (Israeli et al., 2019), lyr only partially suppressed the ectopic blade growth of e, mainly in the basal region of the leaf (Fig 1f-h). We further investigated the effect of these two mutants on auxin response, by comparing the effect of the single and double mutants on the expression of the auxin response marker DR5. While in wild-type leaf primordia DR5 is expressed specifically in the growth domain, marking leaflet initiation sites (Fig 1i and (Shani et al., 2010)), in the slmp mutant DR5 expression is expanded into the intercalary domain (Fig 1j and (Israeli et al., 2019)). In contrast, in the lyr mutant DR5 expression was comparable to that of the wild type, and in lyr slmp double mutants, DR5 expression was similar to single slmp mutants, despite the reduced blade growth domain (Fig 1k-l). In agreement with the genetic and DR5 analyses, we did not detect an effect of the lyr and slmp mutants on the expression of SlMP and LYR, respectively (Fig 1m, n). Therefore, we concluded that SlMP and LYR promote leaflet initiation and growth via separate pathways and the lack of both severely impairs marginal growth (Fig 1o).
SlMP and GOBLET jointly promote leaflet formation
The tomato CUC gene GOBLET (GOB) plays an important role in boundary specification at the leaf margin, and altered GOB/CUC2 expression or activity substantially alters leaf patterning (Fig 2, 5 and (Ben-Gera et al., 2012; Berger et al., 2009)). Plants overexpressing microRNA164 (MIR164), which targets GOB/CUC2, or loss-of-function gob-3 mutants, have less leaflets and smooth leaf margins (Fig 2g, 5c red and purple arrowheads and (Berger et al., 2009; Brand et al., 2007)). The gain-of-function mutant Gob4d/+, which has a point mutation in the MIR164 target site leading to loss of MIR164 recognition and thus deregulation, has deeply lobed leaflets (Fig 2e, 5i blue arrowheads and (Berger et al., 2009; Brand et al., 2007;)). In Arabidopsis, CUC2, auxin and auxin transport regulate serrations in a coordinated feedback loop: CUC2 promotes serrations by regulating auxin transport, and auxin in turn represses CUC2 (Bilsborough et al., 2011; Galbiati et al., 2013). To understand how these tools are utilized and coordinated in the context of compound-leaf development and leaflet formation, we examined the molecular and genetic interaction between SlMP and GOB/CUC2. We first analyzed the effect of SlMP and GOB activities on the expression of each other. SlMP expression was similar between wild type, Gob4d, FIL≫GOBm, and FIL≫miR164 (Fig 2a). Similarly, GOB expression was not affected by the slmp mutant (Fig 2b). Gob4d and slmp both have elongated petioles and less leaflets (Fig 2c-e and i, j). In agreement with the expression analysis, slmp Gob4d/+ double mutants had partially additive phenotypes: slmp Gob4d/+ double mutants had fewer primary leaflets and longer petioles than did each of the single mutants (Fig 2c-f and i, j). slmp FIL≫miR164 plants also showed additive phenotypes (Fig 2g, h). These results suggest that SlMP-mediated leaflet initiation and GOB/CUC2-mediated boundary specification are parallel pathways acting together to determine the number and location of leaflets (Fig 2k). Therefore, GOB/CUC2 affects leaflet patterning not only through auxin. The effect of GOB/CUC2 appears to differ between lobe patterning and leaflet patterning. The effect on lobes is conserved with Arabidopsis, while the effect on leaflets is different, suggesting that leaflet and lobe patterning utilizes similar tools in a different manner.
SlMP promotes leaflet formation within a transient developmental window
The results above suggest that several regulators act in parallel within the same domain (MP and LYR), and that different patterning factors (MP and GOB) act in parallel to promote the growth and intercalary domain, respectively. We then asked how these patterning processes are integrated with the morphogenetic activity of the leaf marginal meristem. First, we studied in more detail SlMP expression during leaf development (Fig 3). Successive leaves vary in their complexity and maturation rate, such that at the P5 stage, the first leaf (L1) has matured and ceased generating leaflets, while later leaves still make leaflets (Bilsborough et al., 2011; Shleizer-Burko et al., 2011). SlMP expression showed a gradual increase in P5 primordia of successive leaves and peaked earlier in leaf development in the first leaf, in agreement with a role in a specific developmental window of leaflet formation (Fig 3a, b). In addition, we found higher expression of SlMP in the margins of P7-P8 compared with inner tissues (Fig 3c, d). We therefore assessed the effect of slmp on the number of leaflets in successive leaves. All slmp leaves showed a reduction in the number of leaflets compared to the wild type, and the effect increased with the increase in leaf complexity (Fig 3e, f, g). This suggests that SlMP acts in the leaf margin at a specific spatial and temporal context to promote leaflet initiation and growth.
We genetically examined this idea by studying the relationship between SlMP and regulators of the morphogenetic window. LA/TCP4 restricts the morphogenetic potential of tomato leaves by promoting leaf differentiation (Nancy G. Dengler, 1984; Ori et al., 2007). The dominant, gain-of-function mutant La-2 shows accelerated leaf maturation and differentiation, resulting in small leaves and reduced leaflet formation (Fig 4a, b and (Ori et al., 2007)). Conversely, the MYB transcription factor TRIFOLIATE (TF) promotes morphogenesis and delays differentiation, and leaf development terminates preciously in tf loss-of-function mutants leading to small leaves with only one terminal and two lateral leaflets (Fig S2i and (Naz et al., 2013)). Leaves of La-2/+ slmp double mutants were similar to those of La-2/+, but were slightly larger with longer petioles, which are characteristics of single slmp mutants (Fig 4a-d, S2l). Similarly, tf slmp double mutants were almost identical to tf single mutant, but slightly larger (Fig S2i, j, l). This finding is in agreement with the expression dynamics of SlMP (Fig3) and suggests that La-2/+ and tf leaves mature fast and cease morphogenesis before the window of SlMP activity. In agreement, SlMP expression was reduced in La-2/+ primordia compared with the wild type (Fig 4i).
To further characterize the spatial and temporal context of SlMP activity, we generated double mutants between slmp and several mutants with increased leaf complexity. The loss-of-function mutants la-6 and clau have elaborated and dissected compound leaves due to extended morphogenetic activity (Fig 4e, g and (Avivi et al., 2000; Bar et al., 2016)). la-6 slmp and clau slmp double mutants showed reduced complexity relative to the single la-6 or clau mutants (Fig 4e-h and Fig S2m, missing leaflets are marked with red arrowheads). slmp also substantially suppressed the increased leaflet formation of mutants and transgenic lines with elevated expression or activity of KNOXI genes, including Me/+, BLS≫TKN2, bippinata (bip) and Petroselinum (Pts) (Fig S2a-h, m and (Kimura et al., 2008; Parnis et al., 1997; Shani et al., 2009)). In agreement, we found elevated expression level of SlMP in clau leaf primordia (Fig 4i). Interestingly, TKN2 expression decreased in slmp leaf primordia, suggesting a more complex interaction between these factors (Fig S2k). Therefore, mutants that shorten the morphogenetic window are epistatic to slmp, while increased leaf complexity of mutants that prolong the morphogenetic window depends on intact SlMP. Cumulatively, these results indicate that marginal activity largely depends on SlMP-mediated leaflet formation and that similar developmental programs mediate initiation and growth from the leaf margins.
GOBLET specifies leaflet boundaries within a transient developmental window
To genetically examine the respective contribution of boundary specification and maturation rate to leaf shape, we introgressed the different gob alleles into genotypes with altered morphogenetic windows. MIR164 overexpression suppressed the la-6 increased leaflet number and lobing (Fig 5a-e, red and blue arrowheads). Strikingly, expression of MIR164 substantially suppressed the phenotype of the super compound leaves and undifferentiated margins caused by overexpression of MIR319, which targets LA/TCP4 and three additional class II TCPs (Fig S3a-c). Similarly, MIR164 overexpression or gob-3 suppressed the increased dissection caused by overexpression of the maize KNOXI gene kn1 or by Me/+, respectively (Fig 5f, g and S3d-i). This suggests that proper specification of the boundary domain by GOB/CUC2 is essential for the enhanced complexity that results from a prolonged morphogenetic window (Fig 5 and S3). In contrast, La-2/+ was nearly epistatic to alteration of GOB/CUC2 activity in La-2/+ FIL≫MIR164 and La-2/+ Gob4d/+ leaves, although the increased and decreased lobing caused by Gob4d/+ and FIL≫MIR164, respectively, were still apparent in La-2/+ Gob4d/+ and in La-2/+ FIL≫ MIR164 (Fig 5h-k). These results suggest that GOB/CUC2 acts within the morphogenetic window that is defined partially by LA/TCP4 to specify leaflet boundaries. However, the effect of GOB/CUC2 on lobing is less dependent on the morphogenetic window than the number of leaflets, or LA/TCP4 has a more prominent role in leaf maturation than in leaflet maturation. Together, these genetic interactions suggest that marginal activity depends on both a prolonged morphogenetic window at the leaf margin, and proper specification of growth and boundary domains within this window.
Prolonged morphogenetic window is essential for leaf complexity
The genetic interactions presented above suggest that a prolonged morphogenetic window is essential for the manifestation of leaflet patterning events. To examine this idea in a broader view, we performed genetic screens in two genetic background with opposite effects on the morphogenetic window: The La-2 mutant, with rapid maturation and a short morphogenetic window and the la-6 mutant with prolonged maturation and extended morphogenetic window. We hypothesized that the La-2 background will be relatively insensitive to the identification of new patterning regulators, while la-6 will serve as a sensitive background for the potential identification of new modifiers. We previously mutagenized La-2/+ gob/+ progeny with ethyl methane sulfonate (EMS), a population that we used for the identification of the la-6 loss-of-function allele (Menda et al., 2004; Ori et al., 2007). We have screened ∼1800 M2 families from this population, however, very few mutants that affected the La-2 leaf phenotype were identified (Fig 6 and S4), in agreement with La-2/+ being a relatively insensitive background for the identification of leaf shape mutants. Three La-2/+ enhancers (h1413, h586 and h1241 the latter two being alleles of the same gene) were identified (Fig 6, S4 and table 1). In contrast to the screen in the La-2/+ background, from the ∼1800 M2 families that were screened in the la-6 background, we identified 37 suppressors and 14 enhancers of the la-6 leaf phenotype (Fig 6, S5 and table 1). Therefore, the La-2/+ genetic background appears epistatic to mutations in many potential patterning regulators (similarly to slmp and gob-3) and the la-6 genetic background enabled the identification of potential new patterning regulators. In general, this forward genetics, unbiased approach strongly supports the importance of both extended morphogenesis and marginal patterning for leaf-shape diversity.
Discussion
Morphological diversity is achieved by the combination of tuning global differentiation in time and space, and patterning by local growth (Kierzkowski et al., 2019; Tsukaya, 2019). Here, using the tomato compound leaf as a model, we examined how these two processes cooperate to pattern the leaf and generate leaf-shape diversity. We found that coordinated specification of growth and boundary domains takes place within a transient morphogenetic window, jointly enabling the elaboration of leaflets and lobes (Fig 7). Within each domain, several pathways act in parallel to ensure robust shape patterning.
Multiple regulators of morphogenesis and patterning
Leaf development is a continuous process composed of several partially overlapping stages (Dengler & Tsukaya, 2001; Hagemann & Gleissberg, 1996; Poethig, 1997;). Shortly after leaf initiation from the periphery of the SAM, the primordium margins undergo primary morphogenesis, during which the main patterning events, including the initiation of marginal structures, take place. The initiation of separate blade units depends on the specification of three differential growth domains: the growth, intercalary and boundary domains (Berger et al., 2009; Bilsborough et al., 2011; Hagemann & Gleissberg, 1996; Israeli et al., 2019; Poethig, 1997). Genetic evidences suggest that specification and maintenance of these domains is regulated by the activity of several, partially overlapping regulators. For example, three main boundary specification regulators were identified in tomato: GOB/CUC2, LATERAL SUPRESSOR (LS) and POTATO LEAF (C). While GOB and LS were shown to act in the same genetic pathway (Rossmann et al., 2015), GOB/CUC2 and C likely act via different pathways to specify the boundary domain, as gob c double mutants enhance the single mutants and have a completely entire margin with no formation of leaflets (Busch et al., 2011). We have previously shown that several A-ARFs have partially overlapping function in promoting growth from the leaf margin (Israeli et al., 2019). Here, we show that SlMP and LYR act in parallel pathways to promote the growth domain. While SlMP is expressed throughout the leaf margins (Israeli et al., 2019), LYR is more specifically expressed at the sites of leaflet initiation (David-Schwartz et al., 2009). In addition, LYR expression is not affected in the slmp mutant (Israeli et al., 2019). Therefore, LYR is probably not regulated by SlMP but acts downstream of a different auxin mediator and/or by another, yet unknown factor. Therefore, several independent factors regulate the specification of each domain, and their activities only partially overlap, such that in the absence of one of them, patterning is compromised, but the basic structure is retained. This partial overlap thus contributes to the balancing between robustness and diversity (Abley et al., 2016; Israeli et al., 2019).
Conservation and divergence in leaf patterning
Several lines of evidence suggest that NAM/CUC and class I KNOX (KNOXI) proteins positively affect the expression and activity of each other. In Arabidopsis, tomato and Cardamine KNOXI proteins act downstream of CUC in the establishment and maintenance of the SAM. KNOXI expression is activated by CUC and is absent in cuc mutants, which lack a SAM (Aida et al., 1999; Brand et al., 2007; Blein et al., 2008; Hay et al., 2006; Rast-somssich et al., 2015). Here, we find that compromised GOB/CUC2 activity substantially suppresses the increase in leaf complexity caused by KNOXI overexpression. Similarly, in Cardamine leaf development, elevated KNOXI activity leads to increased leaf complexity (Hay & Tsiantis, 2006), and reduced CUC activity suppresses this effect of KNOXI (Blein et al., 2008). This suggests that the relationship between CUC and KNOXI is conserved in several species and developmental processes (Alvarez et al., 2016; Floyd & Bowman, 2010). Interestingly, Arabidopsis CUC2 and CUC3 were dispensable for marginal elaboration, unlike tomato (Alvarez et al., 2016). Therefore, the regulation of marginal elaboration is distinct between tomato and Arabidopsis. However, other factors appear to mediate marginal elaboration in both Arabidopsis and tomato. Mutations in genes encoding growth-promoting factors such as WUSCHEL-related homeobox (WOX1) and (PRESSED FLOWER) PRS suppressed the indeterminate margin caused by miR-TCP-NGA overexpression (Alvarez et al., 2016). These factors were shown to act downstream of auxin and SlMP (Guan et al., 2017). In tomato, slmp mutants are shown here to suppress the enhanced complexity caused by prolonged marginal activity in mutants such as la-6/tcp4, clau, Me, bip and Pts. Therefore, growth promoting factors have conserved roles in mediating marginal activity in both tomato and Arabidopsis. Mutations in differentiation promoting factors such as LA/TCP4 and CLAU and plants with increased activity of morphogenetic promoting factors such as KNOXI differ in their phenotypes, as well as in their interactions with slmp and gob. slmp and gob reduced leaf complexity of la-6 and clau to a similar extent as they did in the wild type. Conversely, slmp and gob phenotypes were partially enhanced by Me/+ and Kn1 overexpression, suggesting that these factors interact in additional manners.
Context-dependent interaction
A gradient of TCP expression correlates with leaf maturation, and leaves with reduced TCP activity produce larger and more compound leaves, while leaves with increased TCP activity are smaller and simpler (Alvarez et al., 2016; Efroni et al., 2008; Koyama et al., 2010, 2017; Nikolov et al., 2019; Ori et al., 2007; Palatnik et al., 2003). While TCP3 negatively regulates CUC expression (Koyama et al., 2007), loss of CUC activity can only partially explain the reduced leaf complexity of the dominant and smaller La-2 gain-of-function mutant, because gob-3 leaves are comparable in size to those of the wild type (Berger et al., 2009; Brand et al., 2007). The relationship between TCP and CUC is also connected to the age-dependent change in leaf shape, termed heteroblasty (Kerstetter & Poethig, 1998). Both simple and compound leaves display heteroblasty (Naz et al., 2013; Rubio-Somoza et al., 2014; Shleizer-Burko et al., 2011;). In tomato and other Solanum species, LA/TCP4 expression increases earlier in early leaves relative to later leaves, correlating with a gradual increase in leaf complexity in later leaves (Shleizer-Burko et al., 2011). In Arabidopsis and Cardamine, higher TCP expression in early leaves delimits the activity of CUC via MIR164 regulated and non-regulated pathways (Rubio-Somoza et al., 2014). With plant maturation, the inhibitory effect of TCP on CUC activity is reduced, releasing CUC to increase leaf complexity. It will be interesting to examine whether a similar age-dependent interaction exists between LA/TCP4 and GOB/CUC2 in tomato. The negative interaction between LA/TCP4 and GOB/CUC2 may underlie the retained effect of GOB/CUC2 and MIR319 overexpression on leaf lobing in backgrounds with a short morphogenetic window. The age-related changes in SlMP expression and phenotype shown here suggest that the window of SlMP activity changes in successive leaves, likely responding to the changing dynamics of TCP activity.
In addition, GOB/CUC2 and auxin have distinct effects on leaflet and lobe patterning, and their interaction with the morphogenetic window also differs between these contexts. While the effect on lobing appears similar to Arabidopsis, this is not the case for leaflet patterning. Therefore, common tools are combined differently to pattern leaflets and lobes. This is consistent with the view that leaflets are more similar to simple leaves than the entire compound leaf (Bharathan et al., 2002; Efroni et al., 2010; Poethig & Sussex, 1985; Runions et al., 2017).
Tweaking agricultural traits
Phenotypic diversity has been of great interest in research and agriculture for many years (Eshed & Lippman, 2019; Theophrastus, 1916). Domestication and breeding themes have selected genetic variants with beneficial traits such as a determinate growth habit, early flowering, fruit size, non-shattering seed dispersal and non-dormant seeds (Abbo et al., 2014). Many of these traits are related to maturation and differentiation rate, and are controlled by plant hormones, such as Florigen and Gibberellin (Boden et al., 2015; Cong et al., 2008; Eshed & Lippman, 2019; Lemmon et al., 2018; Müller et al., 2015; Pourkheirandish et al., 2015; Soyk et al., 2019; Studer et al., 2011; Zhu et al., 2013). The current results highlight the potential of tweaking growth patterning within the context of a maturation program to achieve shape diversity. This may inform future studies in other agriculturally important developmental processes that are regulated by patterning and maturation, such as inflorescence structure, flower and fruit development (Eshed & Lippman, 2019; Park et al., 2012, 2014; Rodríguez-Leal et al., 2017;). In addition, future breeding programs may use this approach to increase yield in a range of crops by combining subtle changes in relevant developmental traits.
Funding
This research was supported by grants from the Israel Science foundation (2407/18 and 248/19) and the U.S. –Israel Binational Science Foundation (2015093). Alon Israeli is grateful to the Azrieli Foundation for the award of an Azrieli Fellowship.
Disclosures
The authors declare that they have no conflict of interest
Author contribution
A.I, N.O., and O.B.H. design of the research; A.I., O.B.H., Y.B., I.S., H.B.G., S.H.S., M.B., I.E. and N.O. performed the experiments, collected, analyzed and interpreted the data; A.I. and N.O. wrote the manuscript with input from all authors.
Acknowledgements
We thank Aya Refael Cohen for initial analysis of the La-2 genetic screen and members of the Ori group for continuous discussion and support.