Leaf Developmental Stage Influences Disease Resistance in Tomato

ABSTRACT

The processes of morphogenesis and differentiation are crucial for leaf development, with the duration of the morphogenetic window influencing final leaf shape. Leaves at different developmental stages exhibit distinct morphological and physiological characteristics that may influence their ability to resist pathogens, and disease resistance has been linked to developmental stage in many plant species. To understand how leaf development impacts disease resistance, we examined the immunity of leaves at distinct developmental stages, exploring the role of hormonal pathways and the impact of leaf structure and microbial interactions on disease resistance. Our findings reveal that leaves of different developmental stages exhibit distinct disease responses to various pathogens, determined primarily by the ratio between salicylic acid and jasmonic acid. Higher relative jasmonic acid content in later developing leaves was found to result in increased disease resistance to necrotrophs, while higher relative salicylic acid content in earlier developing leaves rendered them more resistant to biotrophs. This phenomenon occurred across plant ages, in several species, and also impacted the plants' response to biocontrol agents, depending on the pathway being primed. We found that structural variations among leaves can also affect disease response, due to differential recognition by the invading pathogen, and possibly also due to alterations in the leaf microbiome. Our results uncover some of the factors influencing developmental immunity in tomato, and highlight the importance of considering plant development when managing disease resistance.

To understand how leaf development impacts disease resistance, we examined the immunity of leaves at distinct developmental stages, exploring the role of hormonal pathways and leaf structure. Our findings reveal that distinct disease responses of different developmental staged leaves are determined primarily by the ratio between salicylic acid and jasmonic acid.

1. Introduction

Plant disease resistance is determined by various factors, including environmental conditions, the type of infected tissue, and the genetic composition of both the host species and the pathogen. The developmental stage of the host plant is crucial as well (Develey‐Rivière and Galiana 2007), but has received less attention in studies of plant–pathogen interactions. This creates a critical gap in our understanding of disease resistance, because susceptibility to many pathogens is influenced by the developmental stage at which the plant is infected.

Plants have developed a two‐tiered innate immune system to recognise and respond to various biotic threats. The plant's initial immune response is activated when pathogen‐associated molecular patterns (PAMPs) are recognised by surface‐localised pattern recognition receptors (PRRs), leading to PAMP‐triggered immunity (PTI). Effector‐triggered immunity (ETI) is triggered by the direct or indirect recognition of pathogen effectors by intracellular nucleotide‐binding leucine‐rich repeat (NLRs) (Jones and Dangl 2006; Yuan et al. 2021). Plant defences can be primed by previous infections, interactions with microorganisms, or treatments, leading to induced resistance to future attacks by pathogens or parasites (Pieterse et al. 2014). Plant‐induced resistance is broadly divided into systemic acquired resistance (SAR) and induced systemic resistance (ISR). SAR typically leads to increased levels of salicylic acid (SA) or SA signalling and pathogenesis‐related (PR) proteins in uninfected tissues following exposure to virulent, avirulent or nonpathogenic microbes, or through chemical treatments (Fu and Dong 2013). ISR, on the other hand, is usually described as being triggered by nonpathogenic soil microbes, providing protection to the plant's shoots against necrotrophic pathogens and insects by regulating the biosynthesis or signalling of jasmonic acid (JA) and ethylene (ET) (Hu et al. 2018; Walters et al. 2013). The biosynthesis of SA and JA or upregulation in their signalling is triggered upon pathogen recognition, and these hormones play essential roles in plant defence. In general, SA‐mediated responses are usually linked to resistance against biotrophic pathogens, while JA‐ and ET‐mediated responses are mostly reported in connection with resistance to necrotrophic pathogens (Di et al. 2017; Glazebrook 2005).

In many flowering plants, disease resistance is associated with host age. As plants develop and age, their immunity profile changes, and resistance to various pathogens is correlated with specific developmental stages (Develey‐Rivière and Galiana 2007; Hu and Yang 2019). Age‐related resistance (ARR) refers to a phenomenon where a plant's susceptibility to disease changes with its age or developmental stage (DeMell et al. 2023; Develey‐Rivière and Galiana 2007; Shah et al. 2015). While plants face dynamic biotic and abiotic conditions throughout their life, they often exhibit a gain or reinforcement of disease resistance as they mature, typically becoming less susceptible to pathogens and herbivores in later stages compared to their juvenile forms (Bruns et al. 2022; DeMell et al. 2023; Develey‐Rivière and Galiana 2007; Hu and Yang 2019). This intrinsic regulation balances the energetic cost of defence with growth, optimising immunity activation at critical times (DeMell et al. 2023; Hu and Yang 2019; Mao et al. 2017). ARR is found across numerous plant species, including economically significant crops like tomato, rice, wheat, potato, sugar beet and the model plant Arabidopsis thaliana (Bruns et al. 2022; Coelho et al. 2009; DeMell et al. 2023; Develey‐Rivière and Galiana 2007; Hu et al. 2023; Hu and Yang 2019; Levy and Lapidot 2008; Liu et al. 2019; Shah et al. 2015; Shibata et al. 2010; Steimetz et al. 2012). While plants are commonly observed to gain or reinforce disease resistance as they mature, becoming less susceptible to a range of pathogens and herbivores (Bruns et al. 2022; Develey‐Rivière and Galiana 2007; Hu and Yang 2019), this is not a universal pattern. Many examples of ARR exist where plants can become more susceptible with age, highlighting the underlying diversity and variability of this response (DeMell et al. 2023; Shah et al. 2015). For instance, mature Arabidopsis plants exhibit increased susceptibility to the necrotrophic fungus Botrytis cinerea and the biotrophic fungus Erysiphe cichoracearum (Rusterucci et al. 2005; Shibata et al. 2010), as well as the necrotrophic bacterial pathogen Erwinia carotovora (Rusterucci et al. 2005). Arabidopsis plants demonstrat a gradual increase in susceptibility to tomato spotted wilt virus (TSWV) with age (DeMell et al. 2023). Furthermore, nonhost resistance in Arabidopsis to Xanthomonas oryzae and Pseudomonas syringae pv. tomato was found to be stronger in juvenile leaves (DeMell et al. 2023). In potato (Solanum tuberosum), resistance to foliar late blight has been observed to decrease as the plant ages, particularly after flowering and during senescence (Millett et al. 2009). Likewise, certain leaves of Brassica oleracea showed increased susceptibility to downy mildew with advancing leaf age (Coelho et al. 2009). These varied responses suggest that ARR is not a simple linear progression of increasing (or decreasing) resistance, but is highly dependent on the specific plant species, pathogen type and even the developmental stage of individual organs (Coelho et al. 2009; DeMell et al. 2023; Develey‐Rivière and Galiana 2007).

Leaf stage‐associated resistance is a form of developmental resistance, linked to both the developmental stage of the plant and that of the specific infected tissue. This differs from age‐related resistance, which, in the past, was not necessarily described as being associated with any specific physiological process or developmental stage (Develey‐Rivière and Galiana 2007; Xu et al. 2018). Studies have shown that leaf stage‐associated resistance occurs in many plant species. For instance, in potato, apical leaves displayed significantly higher resistance to late blight than basal leaves (Visker et al. 2003). In contrast, in grapevine, lower, older leaves show more resistance to powdery mildew (Uncinula necator) compared to the upper, younger leaves (Doster and Schnathorst 1985). In general, previous research has not necessarily related to possible differences between age‐related resistance, which can also involve developmental aspects, and developmental or positional leaf resistance.

Leaves are generally categorised into two basic forms: simple and compound. A simple leaf features a continuous, ‘entire’ lamina, while a compound leaf consists of multiple leaflets, each resembling a simple leaf. Simple leaves differentiate and flatten relatively quickly during development, while compound leaves can be developmentally regarded as intermediate forms between lateral branches and simple leaves (Bar and Ori 2014). Morphogenesis and differentiation are critical phases in the development of leaves. The duration of the morphogenetic window plays a vital role in determining the final shape of the leaf. As a result of a relatively extended morphogenetic window during leaf development, tomato plants ( Solanum lycopersicum ) have compound leaves and great flexibility in leaf development, producing a wide range of leaf sizes and shapes (Shleizer‐Burko et al. 2011; Shwartz et al. 2016). This makes tomato leaf development an excellent model for studying how the balance between morphogenesis and differentiation influences organ shaping (Israeli et al. 2021). Leaf diversity is affected by this balance, as exemplified by, for example, the different lengths of the leaf morphogenetic window in the simple‐leaved A. thaliana and the compound‐leaved tomato (Bar and Ori 2014).

microRNA156 (miR156) and its targets, SQUAMOSA PROMOTER BINDING PROTEIN‐LIKE (SPL) transcription factors, have been reported to play a significant role in age‐related resistance (DeMell et al. 2023; Hu et al. 2023; Hu and Yang 2019; Mao et al. 2017). As plants mature, they undergo a transition from the juvenile to the adult vegetative development stage, known as vegetative phase change (Berardini et al. 2001; Wu and Poethig 2006). The miR156 module significantly regulates this transition (Wei et al. 2023), starting at high expression levels that gradually decrease as the shoot develops (Wu et al. 2009). In tomato plants, elevated miR156 expression triggers early branching and the development of simpler, more juvenile‐like leaves (Eviatar‐Ribak et al. 2013) with shorter developmental programmes; however, in‐depth characterisation of a specific vegetative phase change in tomato is lacking. In young plants, miR156 is highly expressed, which suppresses the accumulation of SPL transcripts and proteins; as plants age, miR156 levels decline, allowing SPLs to accumulate, which has been shown to optimise some defence mechanisms (DeMell et al. 2023; Hu et al. 2023; Hu and Yang 2019; Mao et al. 2017). For instance, resistance against the bacterial pathogen Pseudomonas syringae pv. tomato DC3000 increases during vegetative phase change, correlating with a temporal drop in miR156 levels (Hu et al. 2023).

We previously explored the relationship between leaf developmental status and disease resistance in tomato (Leibman‐Markus, Schneider, et al. 2023; Marash et al. 2023). Because tomato leaves at different developmental stages exhibit distinct hormonal profiles and altered complexity and patterning, we hypothesised that this may underlie observed differences in immune responses. We further hypothesised that the length of the morphogenetic window (representing the period during which the leaf is actively developing and differentiating) may play a crucial role in disease resistance. Here, we investigated the immune response of leaves at different developmental stages and defined the role of hormonal pathways in disease resistance across these stages. In addition, we assessed the possible impact of leaf structure and leaf microbe interactions on disease resistance throughout various developmental stages, finding that the ratio between SA and JA and the leaf structure both influence disease resistance in a leaf‐developmental‐stage‐dependent manner. These insights may allow for developmental‐stage‐targeted disease management strategies.

2. Results

2.1. Impact of Leaf Developmental Stage on Disease Susceptibility

We had previously observed varying disease responses across different leaves on the same plant (Marash et al. 2023). To investigate whether these differences are linked to leaf development, we selected a necrotrophic fungal pathogen, B. cinerea (which causes grey mould), a biotrophic fungal pathogen, Oidium neolycopersici (which causes powdery mildew) and a hemibiotrophic bacterial pathogen, Xanthomonas euvesicatoria (which causes bacterial leaf spot), and examined the disease response of different leaves to these pathogens. We infected leaves 3, 5 and 8 (L3, L5 and L8, respectively) with each of these pathogens. Leaves are numbered from the bottom, with L3 being the first to develop and typically being a simpler leaf, and L8 typically being the last vegetative leaf to develop before transition to flowering and being a more compound leaf. When leaves were infected with the necrotrophic pathogen B. cinerea , we observed that L3 was the most vulnerable to disease, followed by L5. L8 showed the highest resistance (Figure 1A,H). When leaves were infected with the hemibiotrophic pathogen X. euvesicatoria or the biotrophic pathogen O. neolycopersici, we observed the opposite effect: L3 showed the least powdery mildew coverage, followed by L5, while L8 was the most sensitive to O. neolycopersici (Figure 1E,H). Similarly, L3 supported the fewest colony‐forming units (CFUs) of X. euvesicatoria, L5 had a higher count, and L8 had the highest amount of X. euvesicatoria CFUs (Figure 1F). Thus, we found that susceptibility to a necrotrophic pathogen increased with differentiation (and possibly age), while susceptibility to biotrophic pathogens decreased with differentiation (and possibly age).

FIGURE 1.

Effects of leaf stage on immunity and disease resistance. (A) The third (L3), fifth (L5) and eighth (L8) leaves from Solanum lycopersicum ‘M82’5‐week‐old plants were infected with Botrytis cinerea. (B–D) Leaves L3, L5 and L8 were challenged with the immunity elicitor ethylene‐inducing xylanase (EIX; 1 μg/mL). (B) Reactive oxygen species (ROS) production was measured immediately after EIX application every 3 min, using the horseradish peroxidase (HRP)‐luminol method, and expressed as relative luminescent units (RLU). (C) Ethylene production was measured 24 h after EIX application. (D) Conductivity of samples following EIX treatment was measured after 48 h of incubation. (E) Disease coverage of powdery mildew incited by Oidium neolycopersici on leaves L3, L5 and L8 of M82 plants. (F) Bacterial growth, expressed as CFU, in M82 plants infected with Xanthomonas euvesicatoria (Xcv), in leaves L3, L5 and L7. (G) Leaves L3, L5 and L8 were challenged with the immunity elicitor flg22 (1 μM). ROS production was measured immediately after flg22 application every 3 min, using the HRP‐luminol method, and expressed as RLU. (H) Exemplary leaves from A and E. (A–G) Bars depict mean ± SEM, all points shown. Asterisks denote statistical significance among indicated samples in Welch's ANOVA with Dunnett's post hoc test, A: N > 35, B: N > 27, C: N > 9, D: N > 12, E: N > 12, F: N > 42 and G: N > 60. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, non‐significant.

To examine immunity at the cellular level, we exposed leaf samples to two immune elicitors: ethylene‐inducing xylanase (EIX), a fungal elicitor known to induce immune responses primarily through JA/ET‐mediated pathways (Anand et al. 2021; Leibman‐Markus, Schneider, et al. 2023; Pizarro et al. 2020), and flg22, a bacterial‐derived elicitor known to elicit immune responses mainly via the SA pathway (Gupta, Keppanan, et al. 2022; Marash et al. 2022). After EIX exposure, L8 produced the most reactive oxygen species (ROS) per tissue. L5 produced less ROS, while L3 produced the least amount (Figure 1B). ET production and ion leakage in response to EIX followed a similar pattern (Figure 1C,D). In contrast, upon flg22 exposure, L3 had the highest immune response and produced the most ROS, followed by L5 and then L8 (Figure 1G).

In tomato, bottom leaves, which have shorter developmental programmes and differentiate faster, are also chronologically older at any given time than higher positioned leaves, which have longer developmental programmes with more extensive morphogenesis. Therefore, we examined differences in disease susceptibility across a chronological scale, to better assess the role of chronological ‘age’ as part of this developmental analysis. L3, L5 and L8 were infected with B. cinerea on plants of successive ages. We saw that this phenomenon persisted across plant ages, with L3 being the most susceptible to B. cinerea , followed by L5, and L8 showing the highest resistance at any given plant age (Figure 2A). However, both the disease susceptibility of the plants in general and the extent of the differences in susceptibility among the different leaf positions varied with plant age. Younger plants (3–4 weeks old, at the seedling stage) and older plants (above 7–8 weeks old, after flowering and fruit set) showed less variation in disease susceptibility among their leaves. Plants at the height of the vegetative stage, just before flowering (4–6 weeks old) exhibited the greatest differences in the disease susceptibility of L3 as compared with L8 (Figure 2B). Chronological timing of developmental stages is for determinate cultivar M82 plants grown in a greenhouse in 2 L pots.

FIGURE 2.

Botrytis cinerea disease resistance profile in different leaves as a function of plant age. (A) The third (L3), fifth (L5) and eighth (L8) leaves from Solanum lycopersicum ‘M82’ plants were infected with B. cinerea over the course of 7 weeks; plant age indicated in weeks (wo). Plants were grown together, with different plants being used for each experiment. (A) Infection profile of different leaves at indicated plant ages. Bars represent mean ± SEM. Asterisks denote statistical significance among indicated samples in Welch's ANOVA with Dunnett's post hoc test, N > 35, *p < 0.05, ***p < 0.001, ****p < 0.0001, ns, non‐significant. The disease level of L5 on 5‐week‐old plants was set to 100%, and the percentage difference in disease susceptibility are indicated for each comparison. (B) The difference in disease susceptibility among L3 and L8 is plotted over time. Different letters indicate statistically significant differences in the difference between the disease response of L3 and L8 at different plant ages in one‐way ANOVA with Tukey's post hoc test, N > 42, p < 0.011.

Compound leaves such as those of tomato have a prolonged morphogenetic stage, resulting in a wide variety of leaf sizes and shapes (Shleizer‐Burko et al. 2011). Using an additional system to investigate whether differences in disease susceptibility among leaves relate to the length of the leaf developmental programme, we used additional plant species with simple leaves for similar assays. We infected pepper, Nicotiana benthamiana and Arabidopsis plants with B. cinerea and examined the disease response in different leaves. We found the variation in disease response we observed in tomato to be similar in these additional simple‐leaved species (Figure S1). Because different plant species follow different leaf developmental programmes that can vary chronologically, we did not compare the percentage of change in disease susceptibility among different leaves across species.

2.2. The Role of Vegetative Phase Transition in Immunity

The greatest differences in disease response among leaves were observed in 5‐week‐old plants (Figure 2B). The microRNA miR156 has been suggested to play a key role in regulating the transition from juvenile to adult in Arabidopsis (Wu and Poethig 2006) and tomato (Eviatar‐Ribak et al. 2013). Overexpression of miR156 extends the juvenile phase, resulting in simpler, ‘juvenile’ (i.e., earlier to initiate and faster to develop) leaves, which are more similar morphologically to L3 (Figure 3C). To examine whether juvenility or reduction in vegetative phase change can affect disease susceptibility, we infected different leaves of M82 and 35S::miR156 overexpressing plants with B. cinerea and O. neolycopersici. We found that 35S::miR156 overexpressing plants lost the developmental‐stage mediated effects, with all leaves displaying similar susceptibility levels to both pathogens (Figure 3A,B). For B. cinerea , leaves L3, L5 and L8 of miR156 overexpressor plants all displayed a level of sensitivity similar to that of L3 in the wild‐type (WT) M82 plants, suggesting these leaves are indeed more ‘L3‐like’ also in their immune system (Figure 3A). For O. neolycopersici, leaves L3, L5 and L8 of miR156 overexpressor plants all displayed a level of sensitivity similar to that of L5 in the WT M82 plants (Figure 3B), and were more susceptible to O. neolycopersici than L3 of WT M82, but less susceptible than L8.

FIGURE 3.

Differential effects of miR156 overexpression on leaf developmental immunity in tomato. (A) Leaves 3 (L3), 5 (L5) and 8 (L8) from M82 and 35S::miR156‐overexpressing plants, were infected with Botrytis cinerea . (B) Average number of lesions per leaf of Oidium neolycopersici incited powdery mildew disease on L3, L5 and L8 of M82 and 35S::miR156. (C) Exemplary leaves of M82 wild‐type (WT) and 35S::miR156‐overexpressing plants. Bars represent mean ± SEM, all points shown. A: N > 10, B: N > 17. Different letters denote significant differences among samples, in Welch's ANOVA with Dunnett's post hoc test, A, B: p < 0.0001.

2.3. The Involvement of SA and JA in Developmental Immunity

Given the varying disease responses observed across different leaf developmental stages, and in particular the opposite trends observed in the sensitivity to biotrophic and necrotrophic pathogens (Figure 1), we hypothesised that defence hormone levels could underlie this phenomenon. Therefore, next, we examined the levels of hormones involved in resistance to biotrophic and necrotrophic pathogens. We quantified the hormone content of JA and SA in two different developmental stages, L3 and L8. While the quantity of both hormones was greater (per tissue weight) in L8 (Figure 4A), we observed that the ratio of SA to JA was higher in L3 (Figure 4B). In L3, there was 300% more SA than JA, while in L8, there was only 25% more SA than JA (Figure 4A,B). Physiological ageing also affected SA and JA content. When examining L5 across different plant ages, we observed that while the content of both SA and JA increased as the leaf aged, the increase in SA was greater than the increase in JA, leading to an increased SA/JA ratio as the leaf aged (Figure S2). This correlates with the increase in susceptibility to B. cinerea per leaf, as the plant ages (Figure 2A), further supporting the notion that the SA/JA ratio underlies disease susceptibility.

FIGURE 4.

The role of the ratio between jasmonic acid (JA) and salicylic acid (SA) in leaf developmental immunity. (A, B) JA and SA ratios vary across tomato leaves. (A) Quantification of SA and JA in leaf 3 (L3) and leaf 8 (L8) of 6‐week‐old Solanum lycopersicum ‘M82’ plants was done using liquid chomatography‐mass spectrometry (LC–MS). The amount of SA relative to JA in L3 (×3) and L8 (×1.25) is indicated. (B) Ratio of SA/JA in L3 and L8. (C) Relative expression of known SA‐responsive and JA/ET‐responsive genes in L3 and L8. The expression of the SA‐responsive PR1a (Solyc01g106620), NPR1 (Solyc07g040690) and PR1b (Solyc00g174340), and the JA‐responsive PR1b, Pti‐5 (Solyc02g077370), LoxD (Solyc03g122340) and PI2 (Solyc03g020050) was normalised to the geometric mean of the expression of the normalizer genes RPL8 (Solyc10g006580) and CYP (Solyc01g111170). (D–G) The effect of exogenous SA or JA treatment on Botrytis cinerea disease susceptibility in tomato. (D) B. cinerea disease susceptibility of L3 and L8 in 6‐week‐old M82 plants treated with 0.5 mM SA. (E) Increase in lesion area (%) in L3 and L8 in plants treated with SA. (F) B. cinerea disease susceptibility of L3 and L8 in 6‐week‐old M82 plants treated with 0.5 mM JA. (G) Reduction in lesion area (%) in L3 and L8 in plants treated with JA. In (D) and (F), the difference between samples is indicated. Bars represent mean ± SEM; all points indicated in (D) and (F). Asterisks denote statistical significance among indicated samples in one‐way ANOVA with Bonferroni's post hoc test (A, D and F), or two‐tailed t test (B) with Welch's correction (C, E and G). A, B: N = 4; C: N > 6; D, E: N > 45; F, G: N > 40. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, non‐significant.

To further investigate the involvement of the SA and JA pathways in developmental immunity, we examined the expression of genes involved in these hormonal pathways in leaves of different developmental stages (Figure 4C). The SA‐responsive genes PR1a (Gupta, Leibman‐Markus, Pizarro, and Bar 2021; López‐Ráez et al. 2010) and NPR1 (Cao et al. 1997) had a higher relative expression in L3 than L8, while the ET/JA‐responsive genes Pti‐5 (Leibman‐Markus, Gupta, et al. 2023; Thara et al. 1999), LoxD (Mariutto et al. 2011) and PI2 (Ament et al. 2004; Mehari et al. 2015; Pizarro et al. 2020) showed a higher relative expression in L8 than in L3. PR1b has been reported to play a role in both JA and SA signalling pathways (Li et al. 2017; Meller Harel et al. 2014) and showed similar expression levels in both leaves (Figure 4C).

The hormone quantification (Figure 4A,B) and gene expression (Figure 4C) results indicated that the ratio between SA and JA is likely a crucial factor in leaf‐stage‐related disease profiles. Therefore, next, we examined if changing the SA/JA ratio through exogenous treatments could differentially affect disease response in leaves of different developmental stages. Previous results have demonstrated that increasing JA levels reduces B. cinerea susceptibility (Thomma et al. 1998; Zhu and Tian 2012), while treating plants with SA increases B. cinerea susceptibility (Gupta, Leibman‐Markus, et al. 2023). We hypothesised that, given the varying SA/JA ratios in leaves of different stages (Figure 4B), these treatments would affect B. cinerea susceptibility in a differential manner in different leaves. As anticipated, spraying plants with SA led to an increase in B. cinerea susceptibility when compared to mock (water‐sprayed) plants (Figure 4D). The increase in susceptibility was more substantial in L8, with a 60% increase, while L3 showed only a 20% increase in infection (Figure 4D,E). As expected, leaves treated with JA showed reduced susceptibility to B. cinerea in both L3 and L8 (Figure 4F). Here, the reduction in susceptibility was more substantial in L3, with a 19% decrease, compared to a nonsignificant 12% reduction in disease in L8 (Figure 4F,G). Thus, different leaves responded more strongly to the treatment that has a more significant effect on the SA/JA ratio. L3 contains a higher SA/JA ratio and responded more strongly to JA than L8, while L8 contains a lower SA/JA ratio, and thus, responded more strongly to SA than L3.

To further confirm these results, we examined disease susceptibility of leaves of different stages in tomato mutant plants that are disrupted in the SA and JA pathways. We used the JA‐insensitive jai‐1 mutant (background line M82), the reduced JA mutant spr‐2 (background line Castelmart, Cm) and the SA‐deficient transgenic line NahG (background line Moneymaker, MM). For further details on mutant lines see Section 4. Leaves of 6‐week‐old vegetative plants of these mutant lines were infected with B. cinerea . In the SA‐deficient transgenic NahG, L3 had an increase in susceptibility of 69%, while L8 was 45% more susceptible than MM (Figure S3A,B). While NahG plants are expected to display decreased baseline sensitivity to B. cinerea ‐incited disease, their behaviour is somewhat environmentally sensitive, and while they have been previously shown to display a reduction in baseline disease levels with B. cinerea (Gupta et al. 2020; Martínez‐Hidalgo et al. 2015; Mehari et al. 2015), they are also known to display increases in disease susceptibility (Achuo et al. 2004; Audenaert, De Meyer, and Höfte 2002) as well as unchanged levels of disease (Audenaert, Pattery, et al. 2002) when compared to their background line. These differential observations may relate to the presence of unknown abiotic stresses or differential effects of catechol, the primary breakdown product of SA in NahG (Van Wees and Glazebrook 2003). However, consistent with our results (Figure 4), the NahG line still showed a smaller differential in comparison with the background MM line in L8 as compared with L3.

In the JA mutants, L3 of spr‐2 was 32% more susceptible than L3 of Cm, and L8 of spr‐2 had a 50% increase in susceptibility compared to L8 of Cm (Figure S3C,D). The JA‐insensitive jai‐1 plants behaved similarly to spr‐2 plants. L3 of jai‐1 had a 70% increase compared to L3 of M82, while L8 of jai‐1 had 93% more disease than L8 of M82 (Figure S3E,F). Thus, as expected and in agreement with the exogenous treatment results, the differences among the background line and the hormone‐deficient line were greater in L3 in the SA‐deficient line and greater in L8 in the JA‐deficient/insensitive lines (Figure S3).

2.4. The Effect of Developmental Stage on Leaf Response to Biocontrol

The differential immune response in different leaves led us to investigate whether there could be differential responses to priming with biocontrol agents in different leaves. Given the paramount role of the SA/JA ratio in developmental immunity, we suspected that different leaves might have differential responses to priming agents that are known to potentiate either the SA or the JA pathway. To examine this, we used Bion, a chemical SA pathway SAR immunity inducer, and T39, a fungal JA pathway ISR immunity inducer. We have used both biocontrol agents in the past, and confirmed their pathway activation (Leibman‐Markus, Schneider, et al. 2023). Notably, both Bion and T39 are known to be effective against B. cinerea ‐incited disease from past work (Meller Harel et al. 2014). Indeed, both Bion and T39 were able to reduce B. cinerea ‐incited disease levels in both L3 and L8 (Figure 5). Following disease reduction levels with both biocontrol agents over plant age demonstrated that Bion was more active in younger than in older plants, while T39 activity increased as plants aged in L3, and remained highly active in L8 (Figure 5A). Before flowering, at 6–8 weeks of age, both inducers were more active in L8 than in L3 (Figure 5A); after 8 weeks of age, Bion was more active in L3 as compared with L8, while T39 was more active in L8 as compared with L3 (Figure 5A,B). These results demonstrate that biocontrol is mediated by plant hormonal pathways and, as such, can have differential outcomes depending on the developmental stage.

FIGURE 5.

Efficacy of salicylic acid (SA) and jasmonic acid (JA) pathway priming agents in different leaves as a function of plant age. (A) The third (L3), fifth (L5) and eighth (L8) leaves from Solanum lycopersicum ‘M82’ plants were treated with the fungal JA‐pathway biocontrol agent Trichoderma harzianum T39 (10⁷ conidia/mL) or the chemical SA‐pathway biocontrol agent Bion (acibenzolar‐S‐methyl, 0.001%) twice, 4 days and 1 day before being infected with Botrytis cinerea over the course of 6 weeks; plant age indicated in weeks. Plants were grown together, with different plants being used for each experiment. (A) Disease reduction profile in different leaves treated with T39 or Bion, at indicated plant ages. The maximum biocontrol efficacy for each agent in different leaves is indicated with an arrow. (B) The disease reduction percent with Bion and T39 in L3 and L8 in 10‐week‐old plants. Bars represent mean ± SEM. Asterisks denote statistical significance among indicated samples in one‐way ANOVA with Bonferroni's post hoc test, N > 22, *p < 0.05, **p < 0.01.

2.5. Differential Role of Leaf Structure in Developmental Immunity and Disease Resistance

Our results indicate that the ratio between SA and JA is a central factor in leaf‐stage associated immunity. Previously, it was found that B. cinerea can respond to leaf structure (Rombach et al. 2022), and we demonstrated that leaf structure affects disease resistance (Gupta, Elkabetz, Leibman‐Markus, et al. 2021), finding that cytokinin‐rich mutants possess highly compound leaves, correlating with increased immunity, a beneficial bacterial community, and pathogen resistance (Gupta, Elkabetz, Leibman‐Markus, et al. 2021). Therefore, we aimed to determine whether variations in immunity across different leaf developmental stages could also relate to differences in leaf structure. To assess the impact of leaf structure on pathogenic interactions, we created replicas of 6‐week‐old M82 adaxial leaf surfaces of L3 and L8, using the synthetic polymer polydimethylsiloxane (PDMS), according to previously detailed processes (Gupta, Elkabetz, Leibman‐Markus, et al. 2021; Kumari et al. 2020; Rombach et al. 2022). We then examined B. cinerea mycelial growth and spore germination on these replicas. We found that the ratio of B. cinerea growing on the replica as compared to its growth on the unpatterned area surrounding the replica was higher in L3 (Figure 6A,C). On L3 replicas, the fungus grew on the structure as well as it did in the unstructured portion of the agar plate, while on L8, the fungus preferred the unstructured plate and grew less on the leaf structure (Figure 6A,C). Similar results were observed for B. cinerea germination on replicas, with a higher proportion of spores germinating on L3 replicas than on L8 replicas (Figure 6B,D). Therefore, it can be concluded that, in addition to leaf hormonal content, B. cinerea recognises and responds to differences in the structure of leaves at different developmental stages.

FIGURE 6.

The role of leaf structure in disease susceptibility of different leaves. Botrytis cinerea mycelial growth and germination on synthetic leaf replicas of leaflets from different developmental staged leaves. (A, C) Colony area was measured on and off agar leaf replicas, 48 h post‐inoculation with 2000 conidia. In (A), mycelial growth on the patterned replica is shaded in pink, and off the patterned replica area is shaded in blue. Bar = 1 cm. (C) The ratio between the mycelia on the replica structure (pink in A) and off the structure (in the unpatterned surrounding agar, blue in A) is plotted. (B, D) Percentage of conidia germination on polydimethylsiloxane (PDMS) replicas. (B) 2000 conidia in solution (1% glucose/KH₂PO₄, 5 × 10⁵ conidia/mL) were placed on each sterilised positive PDMS replica. Germination was quantified after 4 h. Germinated conidia are shaded in pink. Bar = 20 μM. (C, D) Bars represent mean ± SEM. Asterisks denote statistical significance among indicated samples in two‐tailed t test, A: N > 42, C: N > 35, **p < 0.01, ****p < 0.0001.

2.6. Involvement of the Leaf Microbial Community in Developmental Immunity

Many works have investigated the connection between the plant microbiome and pathogen resistance (Gupta, Elkabetz, Leibman‐Markus, et al. 2021; Gupta and Bar 2020). To examine whether the observed differences in disease levels among leaves of different developmental stages could also relate to microbiome composition, we grew tomato plants in a sterile environment to assess differences between the disease susceptibility of different leaves in the absence of microbes. We found that L3 was more susceptible than L5 when infected with B. cinerea (Figure 7A), as observed in the ambient environment (Figures 1, 2, 3). L8 was not tested as it did not grow sufficiently large in the sterile environment. Many factors in the sterile environment (e.g., humidity, space restriction, etc.) could affect disease susceptibility, and therefore, we did not directly compare the rate of differences in the susceptibility of different leaves among sterile and ambient environments. However, despite our results pointing to the microbial community not having a major role in developmental immunity, we were intrigued by the possibility that different leaves on the same plant might have different microbial communities. Therefore, we proceeded to profile the microbial communities on leaves L3 and L7 from 6‐week‐old M82 plants grown in a greenhouse. The main aim was to investigate changes in microbial colonisation among different leaves on the same plants, which could perhaps be a contributing factor, though minor, to observed differences in disease resistance. Principal component analysis of the sequenced samples showed that L3 and L7 were clustered away from each other (Figure 7B). The L7 samples were closely grouped, indicating greater similarity among them, while the L3 samples were more dispersed, showing more variation among different L3 samples (Figure 7B). Indeed, the Shannon index was higher in L3, supporting a greater diversity within the samples (Figure 7C). The richness was also higher in L3 (Figure 7E) and the distance between samples was highest between L3 and L7 (Figure 7D). Despite originating from the same plants, the relative frequency of microbes at the species level was also different between L3 and L7 (Figure S4A), with a higher abundance of actinobacteria and gamma‐proteobacteria in L7 (Figure S4B,C) and a higher abundance of Bacteroidia, which were almost completely lacking in L7, in L3 (Figure S4D).

FIGURE 7.

The role of the microbial community in developmental immunity. (A) Variation in disease susceptibility among developmental staged leaves of tomato plants grown in sterile conditions. Leaves 3 (L3) and 5 (L5) of Solanum lycopersicum ‘M82’ plants that were grown in sterile conditions were infected with Botrytis cinerea . Microbial community composition varies with leaf development. Leaf microbial communities were prepared and characterised using 16S rRNA sequencing and analysis with Qiime2. (B) Principal component analysis of microbial communities on L3, L5 and L7. (C) Shannon index. (D) Bray–Curtis distance between L3 and L7 communities (beta diversity). (E) Species richness (alpha diversity). (A, C–E) Bars represent mean ± SEM, all points shown. Asterisks denote statistical significance among indicated samples in Welch's t test (A, C and E) or Welch's ANOVA (D). A: N > 16, C: N > 30, D: N > 6 and E: N = 4. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

3. Discussion

This study focuses on how leaf development influences disease resistance. We found that earlier‐developing leaves were more susceptible to the necrotrophic pathogen B. cinerea (Figure 1A) but showed greater resistance to the biotrophic pathogens O. neolycopersici and X. euvesicatoria (Figure 1E,F). Conversely, later‐developing leaves (L8) were more resistant to B. cinerea but demonstrated increased susceptibility to biotrophs. This aligns with previous findings in grapevine, where younger, upper leaves were more susceptible to powdery mildew (Doster and Schnathorst 1985). Our findings suggest that leaf developmental status plays a key role in shaping immunity. We hypothesised that this phenomenon could be due to differential hormonal or structural components.

3.1. On Age‐Related Immunity Versus Developmental or Positional Immunity

Previous research into ARR has uncovered several possible mechanisms for this phenomenon, including physiological changes such as the development of tougher physical barriers and altered chemical content in tissues, as well as shifts in hormonal pathways (SA and JA), and changes in the regulatory activity of microRNAs like miR156 and its targets, SPL transcription factors (DeMell et al. 2023; Hu et al. 2023; Hu and Yang 2019; Mao et al. 2017; Rusterucci et al. 2005; Shibata et al. 2010; Steimetz et al. 2012). Our work confirms the involvement of SA and JA pathways in developmental immunity. Because developmental differences among leaves can also relate to chronological age (elapsed time from initiation), fully separating these factors can be challenging. To address this, particularly as the tomato developmental stage is less defined than in Arabidopsis and is highly dependent on environmental and growth conditions, we followed changes in susceptibility among differently positioned leaves over the course of plant ageing (Figure 2).

3.2. The Role of Morphogenesis in Developmental Immunity

Developmental immunity persisted across different plant ages, with the level of disease resistance variability among leaves following a bell curve (Figure 2). Across all ages tested, later‐developing leaves consistently showed increased resistance to B. cinerea . The greatest variability in resistance among leaves was observed in 5‐ to 6‐week‐old vegetative, preflowering plants, while very young and very old plants showed less difference in susceptibility between leaves (Figure 2). This bell curve suggests that the morphogenetic window—the period of active leaf development and differentiation—may contribute to the observed variability. However, similar disease patterns were observed in simple‐leaved species with shorter leaf morphogenesis (Figure S1), suggesting that the length of the morphogenetic window itself may not be a major determining factor in resistance.

Susceptibility to B. cinerea generally increased with chronological age for all leaves (Figure 2). L3 exhibited increases in susceptibility starting from the seedling stage (3 weeks old), whereas L8 maintained low susceptibility until the plant reached 7 weeks of age. This means earlier‐to‐initiate leaves like L3, which have a shorter developmental programme, increased in susceptibility more quickly than later‐to‐initiate leaves like L8, which have a longer developmental programme. This observation suggests that morphogenetic processes might delay chronological ageing in terms of disease resistance; leaves may only begin to ‘age’ (becoming susceptible to necrotrophs and resistant to biotrophs) once they have differentiated and reached their final form. This differentiation is associated with specific hormonal balances that, in turn, result in alterations in immunity and disease resistance.

3.3. The Role of Vegetative Phase Change in Developmental Immunity

We observed that vegetative phase change may be involved in leaf developmental immunity. Overexpression of miR156A in Arabidopsis leads to increased bacterial growth, while reducing miR156 activity enhances disease resistance, reportedly by enhancing the SA pathway (Hu et al. 2023). Disrupting SA biosynthesis or signalling in Arabidopsis abolishes ARR (Hu et al. 2023). Overexpression of miR156 in tomato results in early branching and simpler, ‘juvenile‐like’ leaves (Eviatar‐Ribak et al. 2013), structurally resembling L3. miR156‐overexpressing plants lost developmental‐stage mediated effects, with all tested leaves displaying ‘L3‐like’ susceptibility to B. cinerea (Figure 3A). While this aligns with findings in Arabidopsis that high miR156 levels suppress resistance, it contrasts with the Arabidopsis findings that miR156 overexpression leads to increased resistance to necrophagous insects by enhancing the JA response (Mao et al. 2017), highlighting potential species‐specific or pathosystem‐specific differences (DeMell et al. 2023). miR156‐overexpressing plants also lost the variability in developmental immune responses to the biotrophic pathogen O. neolycopersici (Figure 3B). This aligns with the Arabidopsis findings that high miR156 levels suppress SA‐mediated immunity (Hu et al. 2023), thus making juvenile stages more susceptible to biotrophic pathogens.

Collectively, these results underscore that SPL proteins and miR156 are central to coordinating age‐dependent defence mechanisms, with their specific roles varying based on the pathogen type and plant species. The loss of variability also strongly suggests altered hormonal profiles and/or pathways in miR156‐overexpressing plants, indicating that the SA/JA ratio may be involved in regulation of vegetative phase change. Hormonal quantification in miR156‐overexpressing plants could shed further light on this in the future.

3.4. SA and JA: The Central Axis of Developmental Immunity

The antagonistic relationship between SA and JA is well established (Gilroy and Breen 2022; Hickman et al. 2017; Pieterse et al. 2012), with SA typically mediating defence against biotrophs and JA mediating defence against necrotrophs (Di et al. 2017). Our results demonstrate that the ratio between these hormones significantly influences disease resistance across developmental stages in tomato. The higher SA/JA ratio in L3 (Figure 4) aligns with its increased resistance to the biotrophic fungus O. neolycopersici (Figure 1E) and the hemibiotrophic bacterium X. euvesicatoria (Figure 1F). One explanation is that SA accumulates more significantly once morphogenesis is complete. Thus, leaves with a shorter morphogenetic window such as L3, accumulate relatively more SA compared to JA. Conversely, the lower SA/JA ratio in L8 (Figure 4) correlates with the stronger resistance to the necrotrophic pathogen B. cinerea (Figures 1A, 2A), aligning with previous reports confirming that JA plays a role in resistance to B. cinerea (Du et al. 2017; El‐Oirdi et al. 2011; Gupta, Leibman‐Markus, Marash, et al. 2021; Gupta, Leibman‐Markus, et al. 2022; Luo et al. 2023). Because JA plays multiple roles in both defence and development (Huang et al. 2017; Wasternack and Hause 2013) and often accumulates more in actively growing and young organs (Kamińska 2021), the greater necrotrophic resistance in L8 (a leaf still undergoing developmental changes) aligns with JA's dual function.

In support of this theory, we found that plants treated with SA showed a significant increase in susceptibility to B. cinerea , as reported previously (El‐Oirdi et al. 2011; Gupta, Leibman‐Markus, et al. 2023), particularly in L8 (Figure 4D,E), while L3 showed a smaller increase. This suggests that SA weakens antinecrotrophic immunity by antagonising the JA pathway, with a greater impact observed in L8 due to its naturally lower SA levels. Concurrently, JA‐treated plants showed reduced disease susceptibility, with a more significant effect in L3 than in L8 (Figure 4F,G). The stronger JA effect in L3 is likely due to the higher native SA/JA ratio in L3; adding JA causes a more pronounced shift in this ratio, whereas L8's naturally low ratio is less impacted. Gene expression analysis (Figure 4C) further confirmed that SA‐mediated immunity is stronger in L3, while JA‐mediated immunity is more prominent in L8. Our results suggest that the central factor influencing leaf stage‐associated resistance is the balance between SA and JA. A model summarising these findings is provided in Figure 8.

FIGURE 8.

Leaf developmental immunity depends on the balance between salicylic acid (SA) and jasmonic acid (JA). In developing leaves still undergoing morphogenesis which are generated during the ‘mature’ vegetative stage, such as L8, the ratio of JA to SA is higher, leading to resistance to necrotrophic pathogens and susceptibility to biotrophic pathogens. In differentiated leaves generated during the ‘juvenile’ vegetative phase, the ratio of JA to SA is lower, leading to resistance to biotrophic pathogens and susceptibility to necrotrophic pathogens. SA in pink, JA in blue. Illustration generated with Biorender.com.

The balance between SA and JA is part of the plant's broader hormonal landscape. Previous work showed that increasing the cytokinin/gibberellin ratio led to reduced and more similar B. cinerea susceptibility between L5 and L8 (Marash et al. 2023). Furthermore, inhibiting TOR‐mediated translational processes reduced disease susceptibility in L3 but had little effect on L8 (where TOR is naturally less active) (Marash et al. 2023). Thus, future investigations should integrate the JA/SA ratio findings with the general hormonal balance and translational activity across developmental stages.

In agreement with our results framing the JA/SA ratio as the central factor in developmental immunity, biocontrol agents that potentiate these pathways showed differential effectiveness based on leaf position and plant age. For instance, Bion (SA pathway SAR inducer) was more effective in younger plants, while T. harzianum T39 (JA pathway ISR inducer) activity increased as plants aged in L3 and remained highly active in L8 (Figure 5A). These findings suggest that future disease management approaches could be optimised by leveraging developmental stage; for example, SA‐potentiating agents might be better suited for younger plants, and JA‐enhancing agents for older plants.

3.5. Leaf Structure and Developmental Immunity

Tomato leaf structure varies significantly across developmental stages, driven by differences in the length of the morphogenetic window (Bar and Ori 2014). L3 has a simpler, flatter structure, while L8 is more complex, rugose and highly patterned. These structural variations appear to influence susceptibility to B. cinerea . Using synthetic replicas of L3 and L8 adaxial leaf surfaces allowed us to eliminate the influence of leaf chemistry and hormones. We found that the ratio of B. cinerea mycelial growth and spore germination was higher on L3 replicas compared to L8 replicas (Figure 6). This suggests that B. cinerea recognised and preferred the L3 structure. The simpler structure of L3 may promote pathogen growth, while the complex patterning of L8 might hinder it. This observation is supported by several studies with other fungi and plant hosts showing that the surface structure influences fungal germination, development and growth (Hoch et al. 1987; Kwon and Hoch 1991; Staples 1983; Wynn 1976). We previously found that the increased resistance of leaves high in cytokinin content relates to their highly compound patterned structure (Gupta, Elkabetz, Leibman‐Markus, et al. 2021). Although we primarily investigated the role of the microbial community in that work, finding that altered leaf structures supported differential communities that could be more (or less) active in promoting disease resistance, we did not rule out a possible direct interaction of specific pathogens with different leaf structures. It is tempting to speculate that fungal pathogens like B. cinerea have evolved to recognise suitable host organs for their feeding based on structural cues. This point will be very interesting to investigate further in the future.

3.6. The Involvement of Microbiome Dynamics in Developmental Immunity

Microbes on the leaf surface are crucial for maintaining phyllosphere balance and preventing pathogen overgrowth (Vorholt 2012). The composition of the phyllosphere microbiome is known to be influenced by host age/developmental status (Ercolani 1991; Gupta, Elkabetz, et al. 2022; Gupta and Bar 2020). We previously demonstrated significant differences in bacterial community richness and diversity across plant developmental stages (Gupta, Elkabetz, Leibman‐Markus, et al. 2021; Gupta, Elkabetz, et al. 2022). Furthermore, plant genetic traits shape leaf chemistry and surface properties, which impact microbial colonisation (Liu et al. 2020), and altering SA and JA pathways can affect microbial community content (Kniskern et al. 2007). When comparing microbial communities of leaves of different stages on the same plants, we observed greater microbial diversity in L3 compared to L7 (Figure 7). One possible explanation is that faster developing ‘older’ leaves may have a more diverse microbial community due to their longer exposure to environmental factors. In contrast, slower developing ‘younger’ leaves undergo prolonged structural changes, potentially restricting the time available for microbial establishment. However, the crucial finding was that susceptibility differences between L3 and L5 were retained (Figure 7A). This suggests that while microbial communities may contribute to disease dynamics, they are a relatively minor factor in determining developmental immunity differences. Therefore, the SA/JA balance remains the central factor influencing developmental leaf immunity.

ARR has been considered in agricultural practices for decades (DeMell et al. 2023; Hu and Yang 2019); however, a comprehensive molecular understanding of ARR is still lacking (Bruns et al. 2022; DeMell et al. 2023; Develey‐Rivière and Galiana 2007; Hu and Yang 2019). Our findings highlight the critical role of the relationship between leaf development and disease resistance. We found differences in disease susceptibility among leaves to be linked to both hormonal pathways and leaf structure. Although the morphogenetic window does not appear to be a major factor influencing disease resistance, our findings do support its involvement to some extent. Significantly, differences in disease resistance were most pronounced during the critical developmental period at the end of the vegetative stage (5–6 weeks old), just before flowering in tomato. It will be interesting to investigate these patterns across ages in other species with shorter morphogenetic windows and explore the connection between vegetative phase change and transition to flowering in influencing the differences in immunity among different leaves.

4. Experimental Procedures

4.1. Plant Materials and Growth Conditions

Tomato ( S. lycopersicum ) cultivar M82 plants were used throughout this study. Other genotypes in this study included the decreased SA transgenic line NahG in the Moneymaker (Mm) background (Brading et al. 2000), the JA biosynthetic mutant spr‐2 in the Castelmart (Cm) background (Li et al. 2003), the JA‐insensitive jai‐1 mutant in the M82 background (Gupta, Leibman‐Markus, et al. 2022; Li et al. 2002) and the 35S::miR156 overexpressor in the M82 (Sp+) background (Eviatar‐Ribak et al. 2013). Additional plants used in this study were pepper (Capsicum annum ‘California Wonder’), Arabidopsis thaliana ‘Columbia’ and Nicotiana benthamiana. Plants were grown in soil (Green 332; Even‐Ari Green, Ashdod, Israel) in a growth chamber set to long‐day conditions (16 h light/8 h dark) at 24°C, or in a greenhouse or net house under natural daylength conditions at ARO, Volcani Institute, Rishon LeZion, Israel.

For this study, leaf numbers 3, 5 and 8 were selected, each representing distinct developmental stages of the tomato leaf. From each leaf, the left hand middle lateral leaflet was consistently used for assays, except where otherwise indicated. Leaves are numbered from the bottom and based on their initiation time, with L3 being the third leaf the plant generates, and L8 being the eighth leaf the plant generates. We focused on leaves numbered 8 and below, as this aligns with the typical development of determinate cultivated tomato varieties that are used in agricultural field settings—like M82, which typically produce eight leaves before the first inflorescence (Dieleman and Heuvelink 1992). This approach allowed vegetative plants to be investigated. In determinate tomato, L3, L5 and L8 have different developmental programmes, with L3 having the shortest—and L8 the longest—morphogenetic window (Shleizer‐Burko et al. 2011). L3 is also earlier to develop, followed by L5 and L8.

4.2. Disease Assays

To examine the difference in disease susceptibility among different leaf developmental stages, we chose several important diseases affecting tomato: grey mould caused by the necrotrophic fungal pathogen B. cinerea (Elad et al. 2016), bacterial leaf spot caused by X. euvesicatoria (An et al. 2019), and powdery mildew caused by the biotrophic fungal pathogen O. neolycopersici (Jacob et al. 2008; Jones et al. 2001).

For B. cinerea infection, isolate B05.10 was cultivated on potato dextrose agar (PDA; Difco) plates at 22°C. Agar plugs (0.4 cm diameter) were used to inoculate detached leaves, which were then placed in a humid chamber at 22°C under long‐day conditions. Necrotic lesions were measured 2–3 days post‐inoculation with ImageJ. Similar infections were conducted on the plant, with similar results (over 4–5 days rather than 2–3 days).

Xanthomonas euvesicatoria was cultured overnight in Luria Bertani (LB) broth with rifampicin (10 μg/mL) at 28°C. A culture diluted to an OD₆₀₀ concentration of 0.0002 in 10 mM MgCl₂, corresponding roughly to 10⁵ CFU/mL (Pizarro et al. 2020), was pressure‐infiltrated into leaves L3, L5 and L8 using a 1‐mL needleless syringe. Eight days post‐inoculation, disease progress was evaluated by crushing infected leaf tissues (two discs of 1 cm diameter per sample) in 1 mL of 10 mM MgCl₂. Bacterial quantification was done by plating 10 μL from 10‐fold serial dilutions and counting the colonies that developed. Controls consisted of 10 mM MgCl₂ without X. euvesicatoria.

Oidium neolycopersici was obtained from young tomato leaves grown in a commercial greenhouse. Conidia were applied to plants by shaking two infected leaves three times above the target leaves. Inoculated plants were then kept in a temperature‐controlled growth chamber at 22°C. The necrotic lesion area or percentage of infected leaf tissue was measured 5–10 days post‐inoculation using ImageJ.

4.3. ROS Production Measurement

ROS measurement was carried out as previously described (Leibman‐Markus et al. 2017; Pizarro et al. 2019). 0.5 cm diameter leaf discs were collected from L3, L5 and L8, and each disc was placed in 250 μL of distilled water within a 96‐well plate (SPL Life Sciences). The plate was gently shaken overnight at room temperature. The next day, the water was removed, and 50 μL of distilled water was added. Before measurement, 100 μL of distilled water containing 1 μM flg22 or 1 μg/mL EIX with luminol 150 μM and horseradish peroxidase (HRP) 15 μg/mL was added to each sample. The emitted light was quantified using a luminometer (Promega GloMax).

4.4. ET Measurement

ET production was measured as previously described (Leibman‐Markus et al. 2017; Marash et al. 2022). Leaf discs, 0.9 cm in diameter, were harvested from plants treated as indicated, and average weight was measured for each plant. Discs were washed in water for 1–2 h. Every six discs were sealed in 10‐mL glass sample bottles with rubber septa, containing 1 mL assay buffer: 10 mM MES pH 6.0, 250 mM sorbitol (with or without the fungal elicitor EIX, 1 μg/mL), and incubated overnight at room temperature. ET content was measured by gas chromatography (Agilent Intuvo 9000).

4.5. Ion Leakage (Conductivity) Measurement

Ion leakage (conductivity) measurement was performed as described (Leibman‐Markus et al. 2017; Pizarro et al. 2020). Leaf discs (0.9 cm diameter) from L3, L5 and L8 were harvested and washed with distilled water for 3 h in a 50‐mL tube. For each sample, five discs were placed in a 10 mL flask with 1 mL of distilled water, with or without 1 μg/mL EIX, for 48 h with agitation. After incubation, 1.5 mL of distilled water was added to each sample and conductivity was measured using a conductivity meter (EUTECH instrument con510).

4.6. RNA Extraction and Reverse Transcription‐Quantitative PCR

For RNA extraction, five 0.9 cm diameter leaf discs were harvested from L3 and L8 of preflowering plants. Isolation of total RNA was performed according to the TRI reagent (Sigma‐Aldrich) procedure, with the application of DNase (EN0521 ThermoFisher) to remove genomic DNA. 1 μg of RNA was used for cDNA synthesis using Maxima reverse transcriptase (ThermoFisher). Quantitative PCR assays were conducted with Power SYBR Green Mix (Life Technologies), using specific primers related to the JA, SA and ET signalling pathway genes (Table S1) (Gupta, Leibman‐Markus, Marash, et al. 2021; Meller Harel et al. 2014; Pizarro et al. 2020), in a StepOnePlus machine (Thermo‐Fisher). Standard curves were achieved by dilutions of one cDNA sample. All primer pairs had efficiencies in the range of 0.97–1.03. Relative expression was quantified by dividing the expression of the relevant gene by the geometric mean of the expression of the normaliser genes RPL8 (Ribosomal protein L2) and CYP (Cyclophillin), using the copy number method for gene expression (D'haene et al. 2010).

4.7. Phytohormone Analysis

Tissue extraction for SA and JA quantification was performed as previously described (Gupta et al. 2020; Shaya et al. 2019). Entire leaflets were harvested from L3, L5 and L8 of 6‐week‐old plants. 200–450 mg of liquid nitrogen‐ground tissue powder was transferred to a 2‐mL tube containing 1 mL of extraction solvent (ES) mixture (79% IPA:20% methanol:1% acetic acid) supplemented with 20 ng of deuterium‐labelled internal standards (IS). The tubes were incubated for 60 min at 4°C with rapid shaking and centrifuged at 14,000 g for 15 min at 4°C. The supernatant was collected and transferred to 2‐mL tubes. The extraction steps were repeated twice, and the combined extracts were evaporated using speed‐vac. Dried samples were dissolved in 200 μL of 50% methanol and filtered with a 0.22‐μm cellulose syringe filter. Five to ten microlitres were injected for each analysis. LC‐MS‐MS analyses were conducted using a UPLC‐Triple Quadrupole MS (Waters Xevo TQMS). Separation was performed on a Waters Acuity UPLC BEH C18 1.7 μm 2.1 × 100 mm column with a VanGuard precolumn (BEH C18 1.7 μm 2.1 × 5 mm). The mobile phase consisted of water (phase A) and acetonitrile (phase B), both containing 0.1% formic acid in the gradient elution mode. The flow rate was 0.3 mL/min, and the column temperature was kept at 35°C. Acquisition of LC–MS data was performed using MassLynx v. 4.1 software (Waters). Quantification was done using isotope‐labelled ISs.

4.8. Hormone Treatments

Hormone treatments were conducted as previously described (Gupta, Leibman‐Markus, et al. 2023; Gupta, Anand, and Bar 2023). Leaves 3, 5 and 8 of 5‐ to 6‐week‐old tomato plants (preflowering) were sprayed to run‐off with 0.5 mM SA or JA. Leaves were sprayed twice: 3 days and 1 day before infection. The leaves were then infected with B. cinerea B5‐10 as described above. Six plants were used per treatment.

4.9. Induced Resistance (Priming Agent) Treatments

For activation of SAR, we used Bion (Marolleau et al. 2017), and for activation of ISR we used T. harzianum T39 (Elad 2000).

Bion, also known as acibenzolar‐S‐methyl (ASM) or BTH (benzothiadiazole), a synthetic analogue of SA, is used commercially to enhance disease resistance (Marolleau et al. 2017; Oostendorp et al. 2001). SA analogues have been found to effectively reduce grey mould ( B. cinerea ) disease in tomato (Meller Harel et al. 2014). Bion has been shown to be processed in vivo by an SA‐processing enzyme (Tripathi et al. 2010). We previously demonstrated that treatment with 0.001% of Bion has no negative effects on plant growth (Leibman‐Markus, Schneider, et al. 2023). Bion was applied to plants at a concentration of 0.001% vol/vol.

Trichoderma spp. are known to antagonise numerous foliar and root pathogens (Gupta and Bar 2020), and many are known to enhance disease resistance by inducing ISR in a JA/ET‐dependent manner (Meller Harel et al. 2014; Zaid et al. 2022). T. harzianum T39 was cultured on PDA plates (Difco) at 25°C in natural light. Conidia were harvested in water and applied to plants at a concentration of 10⁷ conidia/mL.

For disease and immunity assays, Bion or T39 were applied to tomato plants by spraying the fourth and fifth leaves to run‐off. Two treatments were given: 3 days before challenge, and 4 h before challenge.

4.10. Interactions of B. cinerea With the Leaf Surface—Biomimetic Replicas

Biomimetic leaf replicas were prepared as described (Gupta, Elkabetz, Leibman‐Markus, et al. 2021; Rombach et al. 2022). Briefly, a natural leaf was taped to a Petri dish. Liquid polydimethylsiloxane (PDMS) (Sylgard 184; Dow Inc.) at a 10:1 polymer to curing agent ratio was poured over the leaf, and vacuum was applied to ensure thorough coverage of the microstructure. The polymer was allowed to cure overnight at room temperature. Once cured, the leaf was peeled away, leaving a negative replica of the leaf surface. This negative replica was then functionalised to make the polymer hydrophilic and to prevent adhesion between the layers. It was placed in a Petri dish, and a second layer of liquid polymer was poured on top. The replica was again vacuumed for 2 h and left to cure overnight at room temperature. Finally, the negative replica was removed, leaving the synthetic positive replica of the leaf microstructure. Replica imprints on agar plates were prepared by pouring PDA liquid (0.8% agar) onto a Petri dish, and placing a sterilised negative replica onto the liquid agar. Once the agar had fully hardened, the replica was removed, leaving behind the desired imprint on the agar plate.

Botrytis cinerea BcI‐16, an agricultural isolate (Elad et al. 1993) that was previously found to respond to plant structure (Bar and Romanazzi 2023; Rombach et al. 2022) was used for surface interaction assays. For mycelial growth on replicas, conidia of B. cinerea Bcl16 were prepared from 14‐day‐old PDA plates by washing the plate in a solution of 1% glucose/KH₂PO₄ and filtering the solution through two layers of cheesecloth. Conidia concentration was determined using a haemocytometer, and adjusted to 5 × 10⁵ conidia/mL. 2000 conidia were placed on the agar, adjacent to the leaf imprint. The ratio of mycelial growth on and off the leaf imprint was measured after 48 h. 8–10 separate replicas (originating from different plants) were analysed for each leaf type, in two separate experiments.

For conidia germination on replicas, 2000 conidia in solution (1% glucose/KH₂PO₄, 5 × 10⁵ conidia/mL) were placed on each sterilised positive PDMS replica. Germination was quantified after 4 h, in six different replicas per leaf, with three different microscopic fields analysed for each replica, in two separate experiments.

4.11. Sterile Plant Growth and Infection

Seeds of tomato cultivar M82 were sterilised using 100% ethanol for 5 min, followed by 1.5% sodium hypochlorite (NaOCl), and then washed with sterile distilled water three times to remove any excess bleach. The sterilised seeds were germinated in sterile vented Magenta boxes (Duchefa) on half‐strength Murashige and Skoog (MS) medium (Duchefa M0225) and 0.8% plant agar (Duchefa). After 7 weeks, detached L3 and L5 leaves from these plants were inoculated with a 2 μL droplet containing approximately 2000 B. cinerea spores and placed in a humid chamber at 22°C under long‐day conditions. Necrotic lesion dimensions were measured 2–3 days post‐inoculation using ImageJ.

4.12. Microbiome Profiling

For microbial DNA isolation, five leaflet samples per genotype/treatment were collected from the middle lateral leaflets of leaves 3, 5 and 7 of six different plants per sample, using ethanol‐sterilised forceps. Twenty mL of 0.1 M potassium phosphate buffer (pH 8) was added to the tubes. The samples were sonicated in a water bath (47 kHz ±6%) for 2 min and vortexed for 30 s; this step was repeated twice. The pellet of microbes was obtained after centrifugation at 12,000 g for 20 min at 4°C. Sonication was used to minimise undesirable amplification of chloroplast rRNA genes and contamination by endophytic bacteria. The pellet of microbes was resuspended in potassium phosphate buffer. Total DNA from tomato phyllosphere microorganisms was isolated as previously described (Gupta, Elkabetz, Leibman‐Markus, et al. 2021) and used as a template for 16S rRNA PCR amplification. 16S rRNA amplicons were generated with the CS1_515F and CS2_806R primers (Table S1).

Amplicon sequencing was conducted at the RUSH University core sequencing facility, Chicago, United States, using Illumina MiSeq sequencing. We obtained 35–70 K reads per sample, with 91%–93% of reads passing the quality control. QIIME2 (Caporaso et al. 2010) was used for basic bioinformatic analysis: read merging, primer trimming, quality trimming, length trimming, chimera removal, clustering of sequences, annotation of clusters and generation of a biological observation matrix (BIOM; sample‐by‐taxon abundance table). Samples were rarefied at 3800 reads for the analysis. Taxonomy for the operational taxonomic units (OTUs) was assigned using BLAST against the Silva database (silva_132_16S.97) (Glöckner et al. 2017). Alpha and beta diversity, and Shannon index, were performed with the QIIME2 workflow script core_diversity_analysis.py. Permanova analysis showed significant p = 0.03 and q = 0.045 values for differences between the communities present on L3 and L7. The sequence data generated in this study were deposited in the Sequence Read Archive at NCBI under Bioproject PRJNA1250848.

4.13. Statistical Analyses

All data are presented as average ± SEM, optionally with all points shown. Data sets were analysed for normality using the Shapiro–Wilk test. All data sets were found to have normal distribution. Differences between two groups were analysed for statistical significance using a two‐tailed t test, with Welch's correction for samples with unequal variances, where appropriate. Differences among three groups or more were analysed for statistical significance using one‐way ANOVA. Regular ANOVA was used for groups with equal variances, and Welch's ANOVA for groups with unequal variances. When a significant result for a group in an ANOVA was returned, significance in differences between the means of different samples in the group was assessed using a post hoc test. Tukey's or Bonferroni's tests were employed for samples with equal variances, and Dunnett's test was employed for samples with unequal variances. All statistical analyses were conducted using Prism9.

Author Contributions

M.B. was involved in conceptualization. N.L., M.L.‐M., R.G., G.A., I.M., T.D., M.K. and M.B. were involved in methodology. N.L., M.L.‐M., R.G., G.A., I.M., T.D. and M.B. were involved in investigation. N.L., M.L.‐M., R.G. and M.B. were involved in validation. N.L., M.L.‐M., R.G., G.A. and M.B. were involved in formal analysis. M.K. and M.B. were involved in supervision. N.L. and M.B. were involved in writing – original draft. N.L., M.L.‐M., R.G., G.A., I.M., T.D., M.K. and M.B. were involved in writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Additional plant species exhibit similar disease susceptibility variation across developmental stages of leaves. Leaves 3 (L3), 5 (L5) and 8 (L8) were infected with Botrytis cinerea on (A) Pepper plants (Capsicum annum cv. ‘California Wonder’), (B) Nicotiana benthamiana, (C) Old (ORL), middle (MRL) and young (YRL) rosette leaves from Arabidopsis thaliana . Bars represent mean ± SEM, A: N > 36, B: N > 12 and C: N > 42. Asterisks denote statistical significance among indicated samples in one‐way ANOVA with Tukey's post hoc test (A, B), or in Welch's ANOVA with Dunnett's post hoc test (C), *p < 0.05, **p < 0.01, ****p < 0.0001.

Figure S2: The role of the ratio between jasmonic acid (JA) and salicylic acid (SA) in leaf developmental immunity‐ age related results. Quantification of SA and JA in L5 of indicated M82 plant ages was done using LC–MS. The precentage of increase in SA (388%) and JA (218%) during the ageing process is indicated. Bars represent mean ± SEM. Asterisks denote statistical significance among indicated samples in one‐way ANOVA with Bonferroni's post hoc test. N = 4, **p < 0.01, ***p < 0.001.

Figure S3: The role of jasmonic acid (JA) and salicylic acid (SA) in leaf developmental immunity‐ mutant results. (A) Botrytis cinerea incited disease levels in L3 and L8 of 6–7‐week‐old NahG transgenic line (SA deficient) and the background Moneymaker (MM) line. (B) Increase in lesion area (%) in L3 and L8 in NahG plants. (C) B. cinerea incited disease levels in L3 and L8 of 6–7‐week‐old spr‐2 mutant plants (JA deficient) and the background Castelmart (CM) line. (D) Increase in lesion area (%) in L3 and L8 in spr‐2 plants (E) B. cinerea incited disease levels in L3 and L8 of 6–7‐week‐old jai‐1 mutant plants (JA insensitive) and the background M82 line. (F) Increase in lesion area (%) in L3 and L8 in jai‐1 plants. A, C, E: percentage increase in susceptibility in mutant lines is indicated. Bars represent mean ± SEM. Bars represent mean ± SEM; all points indicated in A, C and E. Asterisks denote statistical significance among indicated samples in Welch's ANOVA with Dunnett's post hoc test (A, C, E) or in two‐tailed t test (B, D, F). A, B: N > 25; C, D: N > 30; E, F: N > 40. *p < 0.05, **p < 0.01, ****p < 0.0001.

Figure S4: Differences in the microbial community between different leaf stages. (A) Relative microbial frequency (% of the community) of different microbial taxons in L3 and L7. Proportion of: (B) Actinobacteria, (C) Proteobacteria and (D) Bacteroidia. A: Stacked bars represent part of a whole (%). B–D: Bars represent mean ± SEM, all points shown. Asterisks denote statistical significance among indicated samples in Welch's t test, N = 4, *p < 0.05.

Table S1: Primer pairs used in reverse transcription‐quantitative PCR.

Lindner, N. , Leibman‐Markus M., Gupta R., et al. 2025. “Leaf Developmental Stage Influences Disease Resistance in Tomato.” Molecular Plant Pathology 26, no. 11: e70162. 10.1111/mpp.70162.

Data Availability Statement

The data that support the findings of this study are openly available in Sequence Read Archive at NCBI at https://www.ncbi.nlm.nih.gov/sra/, reference number PRJNA1250848.