Variation in the response of tomato (Solanum lycopersicum ) breeding lines to the effects of benzo (1,2,3) thiadiazole-7-carbothioic acid S-methyl ester (BTH) on systemic acquired resistance and seed germination

Genetic variation may play a major role in how plants respond to activators of systemic acquired resistance. To examine this, the defence activator benzo(1,2,3)thiadiazole‐7‐carbothioic acid S‐methyl ester (BTH) was applied to seed of different breeding lines of tomato (Solanum lycopersicum) with diverse pedigrees, and the levels of induced resistance against Pseudomonas syringae pv. tomato, changes in defence gene expression and detrimental effects on seed germination and seedling emergence were measured. Two breeding lines, 7007 and 7024, were selected as non‐responsive and responsive to BTH. The SAR‐associated genes, SlPR1a and SlPR3b, were induced earlier or more strongly over the control prior to inoculation for line 7024 but not for line 7007. This was not observed for the ISR‐related genes, SlPin2 and SlPR2b. BTH inhibition of seed germination and seedling emergence was more delayed in line 7024 than 7007. However, applying BTH as a seed or soil drip reduced the delay. Thus, greater levels of BTH response have both positive (i.e., induced resistance and expression of SAR‐related gene expression) and negative (i.e., inhibition of seed germination and seedling emergence) effects and can differ significantly between genotypes. Thus, recommendations for use of induced resistance activators should include plant genotype recommendations and consider possible negative impacts of greater responsiveness.

bacterial speck; RT‐PCR; seed germination; systemic acquired resistance

Systemic acquired resistance (SAR) can be activated by the commercial product, benzo(l,2,3)thiadiazole‐7‐carbothioic acid S‐methyl ester (BTH) produced by Syngenta AG, which is an analogue of salicylic acid (SA) that induces the same set of defence genes as an induction by SA (Ryals et al., [35] ). BTH inactivates the enzymes catalase, ascorbate peroxidase and a mitochondrial oxidase resulting in an accumulation of reactive oxygen species (ROSs) coupled with increased phenolic compounds and SA accumulation (Schreiber & Desveaux, [37] ).

Resistance induced by SAR varies among genotypes within plant species. For example, BTH activated SAR against Sclerotinia homoeocarpa in Agrostis stolonifera ‘Crenshaw’, ‘Penncross’ and ‘Providence’ relative to the water‐treated controls; however, ‘L‐93’ did not show increased resistance (Lee, Fry, & Tisserat, [18] ). Another example is SAR activated by BTH in Glycine max against Sclerotinia sclerotiorum, which was most increased relative to the controls in the highly susceptible ‘Elgin 87’ and ‘Williams 82’, but was not activated as highly in the moderately resistant ‘Corsoy 79’ and ‘NKS19‐90’ (Dann, Diers, Byrum, & Hammerschmidt, [6] ). Variation in the resistance response to SAR activators also occurs in tomato (Solanum lycopersicum). The level of induced resistance against Fusarium oxysporum f.sp. lycopersici by validamycin, a SAR activator, varied from over 80% to less than 30% for 20 different tomato cultivars (Ishikawa, Shirouzu, Nakashita, Teraoka, & Arie, [15] ).

Variation of SAR among plant genotypes could be due to differences in the induction of defence gene expression. The defence genes activated by BTH are SA‐mediated and have been found to have homologs in many dicots (Kombrink & Schmelzer, [17] ). For tomato, BTH treatment increased expression of SlPR1a (Herman, Restrepo, & Smart, [12] ), Phytophthora‐inhibited protease 1 (PIP1), cysteine protease (RCR3) and pathogenesis‐related protein P2 (PR4) (Shabab et al., [38] ). Following application of the defence activator burdock fructooligosaccharide (BFO) to tomato fruit, expression of many genes was up‐regulated including two chitinases, (SlPR3a and SlPR3b) and an acidic beta‐1,3‐glucanase (SlPR2a) (Wang, Feng, & Chen, [42] ). Herman et al. ([12] ) compared defence gene expression in healthy non‐infected plants of three tomato cultivars treated with BTH, and the induction of acidic SlPR1a by BTH was highest in ‘Rutgers’, followed by ‘Supersonic’ and then ‘Rio Grande’. However, the lack of pathogen challenge in the study meant that it is unknown if those gene expression differences correlated to the level of induced disease resistance.

Seed application of BTH can be an effective way to induce SAR, although thus far only foliar BTH application has been reported against the bacterial speck pathogen, P. syringae pv. tomato (Pst), in tomato (Graves & Alexander, [9] ; Louws, Wilson, & Campbell, [21] ; Thaler, Fidantsef, & Duffey, [41] ). For example, BTH seed application was used to induce resistance against Peronospora parasitica and Rhizoctonia solani in Brassica spp. (Jensen, Latunde‐Dada, Hudson, & Lucas, [16] ), Sphaerotheca fuliginea in cucumber (Cucumis sativus) (Ramasamy, Bokshi, Phan‐Thien, & McConchie, [33] ) and S. sclerotiorum and Didymella bryoniae in melon (Cucumis melo L.) (Buzi, Chilosi, Sillo, & Magro, [4] ). However, one concern with seed application of BTH is that it can negatively affect seed germination. Ramasamy et al. ([33] ) reported reduced seed germination soaking with 0.0005 mm BTH for 12 hr or 0.0001 to 0.0005 mm BTH soaking for 24 hr in cucumber. Melon seed germination was delayed after exposure to 0.0002 mm BTH for 12 hr (Buzi et al., [4] ), and tomato seed germination was affected with soaking in 0.6 mm or higher BTH for 30 min (Pitblado, [30] ). Thus far, no studies have examined whether this effect of BTH is affected by plant genotype or whether such an effect is related to the level of SAR.

The goal of this study was to examine a collection of tomato breeding lines with diverse pedigrees for their variation in the effectiveness of seed applied BTH to protect against Pst. As cultivated tomato varieties contain <5% of the genetic variation of their wild relatives (Moyle & Graham, [24] ), breeding lines were used that had been generated by backcrosses that originated from interspecific crosses between cultivated and wild tomato species (Poysa, [32] ). One line that was highly responsive and one line that was weakly responsive to BTH for induced disease resistance were then examined for changes in defence gene expression that are been associated with SAR or another form of induced resistance, induced systemic resistance (ISR), which is associated with jasmonic acid (JA) rather than SA in plants. The same lines were also observed for the effects of BTH treatment on seed germination and seedling emergence. For comparison, BTH was applied by drip application to seed or soil to mimic the Phyto‐Drip® seed treatment application system, which has been used to apply other types of seed treatments to crops, including tomato (http://www.phyto-drip.com/).

Seeds of tomato breeding lines, TD99‐0057 (line 7001), TI99‐0145‐0‐2‐5DBA (line 7004), TI00‐0106 (line 7007), TD00‐0035 (line 7008), P351 (line 7009), P139 (line 7011), Q002 (line 7020) and Q038 (line 7024) were obtained from the plant breeding programme of Steve Loewen (Ridgetown Campus, University of Guelph, Ridgetown, ON, Canada). The Pst isolate 06T2 was provided by Dr. Diane Cupples (retired, Agriculture and Agri‐Food Canada, London, Ontario, Canada), which was isolated in 2006 from a tomato ‘H9478’ leaf.

Seeds were disinfested with 15 ml 50% bleach shaking at 60 rpm for 10 min. The bleach was removed, and the seeds washed three times in distilled water for five minutes each shaking at 60 rpm. Approximately, 45 seeds were either submersed in 15 ml 0.3 mm BTH (active ingredient of Actigard 50WP) (Syngenta, Guelph, Ontario, Canada) or distilled water, and then shaken at 60 rpm for 20 hr before planting.

Plants were grown in Sunshine aggregate plus mix LA4 (SunGro Horticulture Canada Ltd., Vancouver, British Columbia, Canada) with 8 hr dark and 16 hr light at 40 μmol m−2 s−1. Pst was grown at 22°C in tryptic soy broth at 125 rpm for 24 hr and adjusted to 1 × 107 CFU/ml with sterile distilled water and 0.025% Silwet L‐77 (Momentive, Columbus, OH, USA). At 14 dps (days postseeding), the upper and lower sides of the terminal leaflets of the first two fully mature leaves were sprayed to run‐off with the bacterial suspension. Mature leaves were defined as being approximately 80% or greater the size of fully mature leaves. Inoculated plants were placed in a 7” Mondi propagation dome (Sunlight Supply, Vancouver, WA, USA) to maintain high humidity, and the plants were misted with water at 2 dpi (days postinoculation). At 7 dpi, lesion number and leaf area were determined. Leaf surface area was measured using a Leaf Area Meter model 3100 (Li‐Cor Environmental, Lincoln, NE, USA). Lesions/cm2 was calculated.

The first two fully mature terminal leaflets of tomato were harvested at 12, 14, 15, 17, 19 and 21 days post‐treatment (dpt) and stored at −80°C. Frozen leaves were homogenized in a chilled mortar in 1 ml ONE STEP‐RNA Reagent following the manufacturer's instructions (BioBasic, Markham, ON, Canada). RNA was stored in DEPC‐treated dH2O at −80°C.

The tomato ubiquitin 3 gene (SlUBI3) was chosen as the constitutively expressed housekeeping gene (Mascia, Santovito, Gallitelli, & Cillo, [23] ). It has been used as a housekeeping gene for RT‐PCR examining biotic interactions, such as root‐knot nematode (Bhattarai, Li, Liu, Dinesh‐Kumar, & Kaloshian, [2] ), Ralstonia solanacearum (Ishihara, Mitsuhara, Takahashi, & Nakaho, [14] ) and Agrobacterium tumefaciens (Rotenberg, Thompson, German, & Willis, [34] ).

Primer sequences with the predicted amplification product sizes are listed in Table [NaN] . GeneRunner (Hastings Software, Hastings, NY, USA) was used to examine the primer characteristics. All primers were synthesized at the University of Guelph Laboratory Services Division (Guelph, ON, Canada).

Primers used for measuring gene expression of constitutive, SAR and ISR‐related genes

qScript cDNA Superscript (Quanta Biosciences, Gaithersburg, MD, USA) master mix with total RNA was used to synthesize single‐stranded cDNA according to the manufacturer's instructions. Relative RT‐PCR, which involves the co‐amplification of a gene of interest with the constitutively expressed control gene, SlUBI3, was used to assess relative transcript levels.

RT‐PCRs were conducted in 15 μl reaction volumes with 0.5 μl cDNA, 0.75 units Tsg polymerase (Biobasic, Toronto, ON, Canada), 10× Tsg polymerase buffer, 10 mm dNTPs, 4 mm Mg2+ and 0.5 μl of 25 mm primer for the gene of interest and the constitutive control (Dean, Goodwin, & Hsiang, [7] ). The concentration of cDNA was adjusted to give similar levels of amplification of SlUBI3 in each reaction. Amplification conditions consisted of 1 cycle at 94°C for 3 min followed by 30 cycles of 94°C for 30 s, a primer‐specific temperature for 1 min and 72°C for 1 min, followed by a final extension cycle of 10 min at 72°C. The primer‐specific annealing temperatures were 54°C for SlPR1a, SlPR3b and SlPin2 and 58°C for SlPR2b. All reactions were carried out in an Eppendorf AG 22331 Master Cycler (Eppendorf, Hamburg, Germany).

RT‐PCR products were separated on 1% TAE agarose gels and stained with ethidium bromide. Gel images were taken using a CCTV camera fitted with a 23A orange filter and visualized with an UV transilluminator. Images were saved as TIFF files and quantified using NIH Image (Scion Corporation, Frederick, MD, USA). Relative expression for the gene of interest was calculated by taking a ratio of the number of pixels for that band intensity over the number of pixels of SlUBI3 in the same lane (Dean et al., [7] ). Relative expression units (REU) are the means of three treatment replications with the PCRs repeated twice per treatment replication.

To confirm that the number of cycles used provided accurate quantification of relative expression values, relative RT‐PCR was also performed with 25 cycles (five fewer than that used for quantification). For all genes examined in this study (SlPR1a, SlPR3b, SlPin2 and SlPR2b), the expression patterns with 25 cycles were highly similar to that obtained with 30 cycles showing that 30 cycles were not too many for quantification.

Ten seeds were used per treatment replicate, and the experiment was repeated four times for line 7024 and five times for line 7007 using a randomized complete block design with five replications per treatment. For seed soak treatment, seeds were disinfested and soaked in 0.3 mm BTH or sterile distilled water as described previously except 20 ml of solution was used. After shaking for 20 hr, seeds were rinsed briefly in tap water, placed on a moistened paper towel inside a petri dish and incubated at 24°C in the dark. For seed drip treatment, disinfested seeds were transferred to moistened paper towel inside petri plates and then 0.3 ml of 0.3 mm BTH or sterile distilled water was applied on top of each seed. The plates were incubated as per the seed soak treatments. Seeds were checked for germination by the appearance of the radicle at 1, 2, 3, 4, 5 and 8 dpt.

Seedling emergence was evaluated by monitoring hypocotyl emergence from seeds planted in Sunshine aggregate plus mix LA4 (SunGro Horticulture Canada Ltd.) and covered by 3‐4 mm of vermiculite. The soak and drip treatments were applied as previously described, except the drip solutions were applied to the soil surface after planting. Ten seeds were used per treatment replicate with one seed per pot cell (38 × 38 × 55 mm). The experiment was repeated three times with five replications per treatment in each trial and arranged in a randomized complete block design. Plants were grown with 16 hr light and 8 hr dark at 24°C, and emergence monitored by monitoring hypocotyl emergence at 6, 7, 8, 9 and 10 dpt.

Disease severity was analysed using the student's t test (p ≤ .05) with SAS v9.1 (SAS Institute, Cary, NC, USA). Significant differences in REUs of gene expression and areas under the gene expression curve were determined with SAS v9.4 using proc mixed (p ≤ .05) with treatment as fixed effect and replication as random effect. Normality was tested with the Shapiro–Wilk statistic, and outliers were identified using Lund's test of standardized residuals. Means were separated using Tukey's HSD when ANOVA indicated a significant treatment effect. The area under the gene expression curve was as follows: area under the curve = ∑ [((Yi + Yi‐1) (Xi − Xi‐1))/2]. Regression analysis for REUs vs. time was performed using proc reg to determine model significance and proc glm to obtain parameter estimates with SAS v9.4. When initial REUs were zero, only the zero value before the first time point of detectable expression was included in the regression.

For seed germination and seedling emergence, normality and outliers were identified as described above. Data for line 7007 germination soak treatments at 5 dpi were arcsine‐transformed, and data for all germination drip treatments were logit transformed to meet assumptions of normality. Back‐transformed means are presented. ANOVA was completed as described above, with treatment and experiment as fixed effect and replication within experiment as random effect. Data from different experiments were pooled because ANOVA indicated no significant treatment x experiment interaction (p ≤ .05). For seed germination, bars show confidence intervals (95%) because standard error of back‐transformed means could not be computed.

The disease severity from Pst in the first two fully mature terminal leaflets was determined for eight tomato breeding lines where the seeds were soaked with BTH or water prior to planting (Table [NaN] ). The disease severity of the water control revealed that the eight lines varied in their inherent susceptibility from line 7020 being moderately resistant with 1.30 lesions/cm2 to line 7024 being highly susceptible with 6.76 lesions/cm2. While the disease severity was reduced by BTH in all lines except line 7020, only line 7024 showed a significant difference between BTH and the water control.

Effect of seed applied BTH on severity of Pst on eight breeding lines of tomato

1 Values are means from three biological replications with three plants per replication (two leaves per plant).

• 2 Student's t test was used to separate means within rows. Means followed by a letter in common are not significantly different (p = .05).
• 3 Treated with distilled water.

From the above results, lines 7007 and 7024 were chosen as weakly and highly responsive, respectively, to seed applied BTH for an examination of tomato gene expression and seed germination. Retesting their response to BTH using more plants per replication showed that the difference in disease severity between the water control and BTH was significantly different for only line 7024 (Table [NaN] ).

Effect of seed applied BTH on severity of Pst on two breeding lines of tomato

• 4 Values shown are pooled from three biological replications with six plants per replication (two leaves per plant).
• 5 Student's t test was used to separate means within rows. Means followed by a letter in common are not significantly different (p = .05).
• 6 Treated with distilled water.

Before inoculation (12 and 14 dpt), expression of SlPR1a in water‐treated lines 7007 and 7024 was undetectable (Figure [NaN] ). At 12 dpt, BTH‐treated leaves of line 7007 had undetectable expression, but SlPR1a expression was detected in BTH‐treated line 7024. At 14 dpt, however, expression was detected with BTH for both lines 7007 and 7024. These results show that SlPR1a expression was induced at least 2 days earlier by BTH in line 7024 than 7007.

After inoculation (15–21 dpt), both water‐treated lines 7007 and 7024 showed continuously increasing SlPR1a expression over time with the best‐fitting curve for line 7007 (y = −4.47 + 0.32x, R2 = 0.91, p = .0007) being relatively similar to that of line 7024 (y = −3.76 + 0.27x, R2 = 0.87, p = .0035). BTH treatment of both lines 7007 and 7024 also showed increased SlPR1a expression after inoculation. However, the best‐fitting curve for expression in BTH‐treated line 7007 (y = 37.91 − 8.00x + 0.54x2 − 0.01x3, R2 = 0.94, p < .0001) showed a peak at 19 dpt followed by a decline at 21 dpt. In contrast, the best‐fitting curve for expression in BTH‐treated line 7024 (y = −2.38 + 0.19x, R2 = 0.75, p = .0198) showed a continuous increase in expression after inoculation, gradually merging with that of the water control. This indicates that elevated SlPR1a expression with BTH treatment after inoculation persisted longer with line 7024 than 7007.

The area under the curve for the gene expression over the period of the experiment was calculated to give an indication of the total level of gene expression. For line 7007, the area was 8.42 and 9.98 for the water and BTH‐treated plants, respectively, which were not significantly different (p = .1434). However, for line 7024, it was 7.27 and 10.91 for the water and BTH‐treated plants, respectively, which was significantly different (p = .0314), indicating that line 7024 was responding to BTH over the course of the experiment, unlike line 7007.

Before inoculation, expression of SlPR3b was detected in water‐treated lines 7007 and 7024 (Figure [NaN] ). Expression in BTH‐treated line 7007 was equivalent to the water control before inoculation. In contrast, expression in BTH‐treated line 7024 was significantly higher than the water control at 12 dpt, indicating that BTH induced expression of SlPR3b over the control before inoculation for line 7024, but not for line 7007.

After inoculation, both water‐treated lines 7007 and 7024 had a continuous increase in expression of SlPR3b, but the best‐fitting curve for line 7007 (y = −1.46 + 0.16x, R2 = 0.72, p = .0305) showed an initially slower increase in expression than that for line 7024 (y = −10.20 + 0.1.17x − 0.03x2, R2 = 0.93, p < .0001). For BTH‐treated line 7007, the best‐fitting curve (y = 37.87 − 7.80x + 0.53x2 − 0.01x3, R2 = 0.83, p = .0038) revealed a faster increase than the water control, and expression of SlPR3b was significantly higher than the water control on 17 dpt (Figure [NaN] ). For BTH‐treated line 7024, the best‐fitting curve (y = −10.51 + 1.37x − 0.04x2, R2 = 0.69, p = .0499) showed that expression increased more slowly than the water control and was significantly lower than the water control at 21 dpt (Figure [NaN] ). These results indicate that the increase in SlPR3b expression over the control in line 7024 only occurred prior to inoculation, but there was a stronger response in line 7007 after inoculation relative to the water controls.

The area under the curve for the gene expression of line 7007 was 11.58 for the water control, which was not significantly different from 13.25 for the BTH treatment (p = .1942). The area for line 7024 was 14.20 for the water control, which was also not significantly different from 14.72 for the BTH treatment (p = .7496). This indicates that despite the significant induction of SlPR3b expression by BTH over the control in line 7024 vs. line 7007 before inoculation at 12 dpt, expression did not respond more strongly after that compared to the controls for both lines 7007 and 7024.

Before inoculation, there was no SlPin2 expression detected in water‐treated and BTH‐treated plants for both lines (Figure [NaN] ). This showed that BTH did not induce expression of this gene.

Following inoculation, water‐treated line 7007 first showed increased SlPin2 expression at 17 dpt, and the best‐fitting curve (y = −42.42 + 4.54x − 0.11x2, R2 = 0.94, p = .0015) revealed a continuous increase until peaking at 19–21 dpt. In contrast, expression for water‐treated line 7024 increased earlier starting at 15 dpt, and the best‐fitting curve (y = −26.02 + 0.84x − 0.07x2, R2 = 0.84, p = .0071) also revealed a peak at 19–21 dpt. For BTH‐treated line 7007, the best‐fitting curve (y = −37.23 + 4.21x − 0.11x2, R2 = 0.87, p = .0034) showed that expression increased and peaked earlier than with the water treatment, but that was not significantly different. For BTH‐treated line 7024, the best‐fitting curve (y = −24.56 + 2.79x − 0.07x2, R2 = 0.88, p = .0022) showed a progressively slower increase than the water control, and so by 21 dpt, expression was significantly higher in water than BTH‐treated line 7024.

For line 7007, the area under the curve for SlPin2 expression was 11.34 and 14.22 for the water and BTH‐treated plants, respectively, which were not significantly different (p = .0619). For line 7024, the area was 13.49 and 10.63 for water and BTH‐treated plants, respectively, which were not significantly different (p = .2545). This shows that there was no statistically significant priming or induction of SlPin2 expression in either line.

Before inoculation, there was no SlPR2b expression detected for water‐treated and BTH‐treated plants for both lines (Figure [NaN] ). This showed no induction of SlPR2b by BTH, like as observed for SlPin2.

After inoculation, the best‐fitting curves for both water‐treated line 7007 (y = −1783.09 + 404.80x − 34.25x2 +1.28x3 − 0.02x4, R2 = 0.93, p = .0014) and line 7024 (y = −1284.19 + 290.44x − 24.45x2 + 0.91x3 − 0.01x4, R2 = 0.90, p = .0056) showed that SlPR2b expression increased rapidly at 15 dpt and then fluctuated without any clear upward or downward trends. For BTH‐treated line 7007, the best‐fitting curve for SlPR2b expression (y = −239.19 + 40.64x − 2.27x2 + 0.04x3, R2 = 0.80, p = .0182) also revealed a rapid increase at 15 dpt, and then expression fluctuated without any clear trends like the water‐treated line 7007. For BTH‐treated line 7024, the best‐fitting curve (y = −15.58 + 1.79x − 0.05x2, R2 = 0.79, p = .0209) showed a slightly slower induction of SlPR2b expression than the water control, but there were no significant differences with the water‐treated line 7024.

The area under the curve for SlPR2b expression with line 7007 was 11.76 and 12.22 for water‐ and BTH‐treated plants, respectively, which was not significantly different (p = .6526). The area under the curve with line 7024 was 8.28 and 9.20 for water‐ and BTH‐treated plants, respectively, which was also not significantly different (p = .5989). This indicates that SlPR2b expression was not affected by BTH treatment in either line, even though it was induced by infection like the other genes examined in this study.

Seed germination following BTH seed soak treatment was lower than the water control beginning at 1.67 dpt for both line 7007 (Figure [NaN] ) and line 7024 (Figure [NaN] ), indicating that BTH inhibited germination in both lines. With BTH, seed germination for line 7007 surpassed 85% at 3 days after the control, whereas that did not occur until 5 days after the control for line 7024. However, seed germination in line 7024 increased more rapidly near the end of the experiment, and thus by 7.67 dpt, seed germination was lower in BTH‐treated seed (93%) than the water control seed (98%) for line 7007, but not for line 7024. While both lines were negatively affected by BTH seed soak treatment, the highly responsive line 7024 showed more of a delay in seed germination by BTH than the weakly responsive line 7007.

Seedling emergence surpassed 75% at 2 days later with BTH than the control for line 7007 (Figure [NaN] ) vs. 3 days later for line 7024 (Figure [NaN] ). In addition, line 7007 emergence was only significantly affected by BTH at 6 dpt (Figure [NaN] ) compared to all time points tested for line 7024 (Figure [NaN] ). This indicates seed soak BTH treatment also reduced seedling emergence more for line 7024 than line 7007.

BTH seed drip treatment delayed seed germination surpassing 85% by 2.00 days than the control for line 7007 but 4.84 days for line 7024 (Figures [NaN] and [NaN] ). At the end of the experiment, seed germination with the BTH seed drip treatment was not significantly lower than the water control for line 7024 but not line 7007 (Figures [NaN] and [NaN] ). Thus, seed germination was also delayed by applying BTH as a seed drip treatment; with once again, the highly responsive line 7024 being more negatively affected than the weakly responsive line 7007.

BTH soil drip treatment did not affect seedling emergence in line 7007 at any time points (Figure [NaN] ), but it did result in lower emergence than the control on 6 and 10 dpt for line 7024 (Figure [NaN] ). Seedling emergence surpassed 75% on the same day as the water control for lines 7007 but was delayed 1 day for line 7024 (Figures [NaN] and [NaN] ). By the end of the experiment, seedling emergence was reduced by 5% compared to the water control in line 7024 but not significantly for line 7007. Although BTH soil drip treatment had the least negative effects compared to other application methods, it still negatively affected line 7024, while there was no negative effect on line 7007.

Seed applied BTH has been reported in tomato to induce SAR against root galls (Meloidogyne spp.) (Mutar & Fattah, [25] ), but this is the first report of using it for control of Pst. However, its effectiveness is highly dependent upon the genotype of tomato. Breeding line 7024 was the only line of the eight to show a significant reduction in Pst infection. Bacterial speck levels in line 7024 were approximately 30% of the control with BTH seed application, which is comparable to that reported for BTH foliar application with disease severity approximately 43% of the control in ‘Sunchief VF’ (Herman, Davidson, & Smart, [11] ), 32% in the field for an unspecified cultivar (Thaler et al., [41] ) and 24% in the field for ‘Sunpride’ (Graves & Alexander, [9] ).

Induction of SAR by BTH is common in commercial tomato cultivars when applied foliarly (Louws et al., [21] ), but only one breeding line showed SAR with BTH in this study. One reason may be that inoculation with the pathogen occurred 14 days after seed treatment with BTH in this study, and the seedlings needed to grow sufficiently so that the leaves were large enough to obtain accurate disease severity measurements. Extended protection has not been observed in tomatoes with BTH. Louws et al. ([21] ) applied BTH on a 7 or 14 day spray schedule to achieve control of Pst and other bacterial pathogens of tomatoes in the field with different tomato varieties at multiple locations in the USA, but 7 and 14 day intervals were not compared directly. For other bacterial pathogens in tomato, 14 day intervals are less effective than shorter intervals (Huang et al., [13] ; Pontes et al., [31] ). Graves and Alexander ([9] ) used a 7–10 day spray schedule against Pst and other bacterial pathogens of tomato in the field with several tomato varieties in Virginia. In the glasshouse, Herman et al. ([11] ) inoculated Pst 5 days after foliar treatment with BTH. Thus, 14 days is longer than typically used for BTH‐induced resistance against Pst, although this study shows that it is possible at least for certain tomato genotypes.

All the lines tested in this study have cultivated tomato pedigrees, but the BTH‐responsive breeding line 7024 was unique among the eight lines in having S. lycopersicoides in its pedigree. Among the BTH non‐responsive breeding lines, lines 7001 and 7008 comprise only S. lycopersicum in the pedigrees, while lines 7007 and 7020 also have a S. habrochaites parent, line 7011 has a S. pennelli parent, and lines 7004 and 7009 have S. habrochaites, S. peruvianum and S. chilense parents or S. arcanum and S. peruvianum parents, respectively (S. Loewen, personal communication). Solanum lycopersicoides is a rare wild tomato‐like nightshade found in only a few drainage areas on the western slopes of the Andean cordillera near the Peru–Chile border (Albrecht, Escobar, & Chetelat, [1] ). It is considered to belong to an immediate outgroup of Solanum section Lycopersicoides, which contains cultivated tomato (Peralta, Spooner, & Knapp, [27] ). Thus, it is the most distant species from cultivated tomato used to create the breeding lines in this study. Only limited work has been done with S. lycopersicoides in tomato breeding programmes because of the severe reproductive barriers with S. lycopersicum (Canady, Meglic, & Chetelat, [5] ; Gleddie, Keller, & Poysa, [8] ). Further work is needed to determine whether S. lycopersicoides is a good source of BTH responsiveness and the mode of inheritance of this trait for its introduction into other cultivated tomato genotypes.

PR1a genes are one of the best characterized SA‐induced genes associated with BTH application in various plants (Loake & Grant, [19] ). For example, Herman et al. ([11] ) found that BTH application to the tomato ‘Sunchief VF’ caused an up‐regulation of SlPR1a expression prior to inoculation with Pst. However, Mandal et al. ([22] ) and Sanz‐Alférez, Mateos, Alvarado, and Sánchez ([36] ) found no up‐regulation following BTH application without pathogen infection in tomato ‘K326’ and ‘Marmande’, respectively. In this study, increased expression of SlPR1a was detectable for line 7024, but not line 7007 before Pst inoculation. After inoculation, SlPR1a expression increased for both the BTH‐ and water‐treated plants of lines 7007 and 7024 with expression being higher in the BTH‐treated plants for several days after inoculation. However, the difference between BTH‐ and water‐treated plants over the entire experiment was only significant for the highly responsive line 7024. Herman et al. ([12] ) also reported tomato genotype differences in the level of induction of SlPR1a by BTH between tomato ‘Supersonic’, ‘Rutgers’ and ‘Rio Grande’.

PR3 genes encode for class I chitinases, which play a role in disease resistance by acting against fungal and bacterial cell walls (Hamel, Boivin, Tremblay, & Bellemare, [10] ). SlPR3b was up‐regulated in tomato fruit following either the SAR‐inducers, BFO or chitosan oligosaccharide, responding similar to that of SlPR1a expression (Wang et al., [42] ). In this study, SlPR3b expression with BTH treatment was significantly greater than the water control in line 7024 but not line 7007 prior to inoculation. Similar to SlPR1a expression, this demonstrated that the more BTH‐responsive genotype also showed greater early BTH induction of SlPR3b than the less BTH‐responsive genotype. After inoculation, however, line 7007 initially showed a greater increase in SlPR3b expression over the control compared to line 7024, and thus the postinoculation response of SlPR3b expression did not correlate with BTH‐activated resistance responsiveness.

Proteinase inhibitor II (Pin2) is a serine proteinase inhibitor with trypsin and chymotrypsin inhibitory activities (Pieterse & van Loon, [29] ). SlPin2 has been used in a number of ISR studies in tomato. SlPin2 expression was up‐regulated after but not before Pst inoculation when ISR was activated by the PGPRs Bacillus subtilis GB03 and B. amyloliquefaciens IN937a (Herman et al., [11] ), whereas it was up‐regulated both before and after Pst inoculation when ISR was activated by the rhizobacterium, Bacillus cereus AR156 (Niu et al., [26] ). In this study, there was never detectable SlPin2 expression prior to inoculation. It was induced after inoculation, but there were no significant differences between BTH‐treated and water‐treated plants after inoculation, except for a significantly lower expression in line 7024 at the end of the experiment. There was no indication that SlPin2 expression was induced by BTH treatment, which was expected for a gene related to ISR rather than SAR, Herman et al. ([11] ) and Niu et al. ([26] ) also found that SlPin2 expression in BTH‐treated plants were not increased compared to non‐treated controls after inoculation with Pst.

PR2 protein is a β ‐1,3‐glucanase that is able to degrade glucans in fungal cell walls and therefore inhibit pathogen growth directly (van Loon, [20] ). SlPR2b has been used as a marker for ISR in tomato. Root application of the arbuscular mycorrhizal fungus, Glomus mosseae, to cause ISR against Alternaria solani highly induced SlR2b expression in treated plants after infection relative to the control (Song et al., [40] ). Similarly, root application of the rhizobacterium, P. fluorescens WCS417r, to cause ISR against Bemisia tabaci significantly induced SlPR2b expression relative to the control (Shavit, Ofek‐Lalzar, Burdman, & Morin, [39] ). In this study, BTH had no effect on SlPR2b expression prior to inoculation, similar to SlPin2 expression. Expression in the current study increased after inoculation, but the changes were almost identical for both water‐ and BTH‐treated plants of both lines. Thus, the expression of this ISR‐related gene did not correlate with the BTH responsiveness, as might be expected for a gene related to ISR rather than SAR. The patterns of expression of SlPin2 and SlPR2b indicate that the differences observed between lines 7007 and 7024 in SlPR1a and SlPR3b expression are not due to universal changes in defence gene expression with BTH, but are due to SAR‐related gene induction.

While a genotype being more responsive to seed applied BTH for inducing SAR and SAR‐related gene expression would be considered positive, it is possible that the same genotype would also be more responsive to seed applied BTH for traits considered to be negative. BTH seed soak treatment has been reported to negatively affect seed germination in plants, including tomato (Pitblado, [30] ), cucumber (Ramasamy et al., [33] ) and melons (Buzi et al., [4] ). Both lines 7007 and 7024 showed that soaking the seed in BTH resulted in a delay in tomato seed germination and seedling emergence. However, the BTH‐responsive genotype for SAR, line 7024, was more negatively affected than line 7007, indicating that a stronger response to BTH relates to both induced disease resistance and delayed seed germination. BTH causes an accumulation of phenolic compounds and SA (Schreiber & Desveaux, [37] ), and SA potentiates the generation of ROSs that have been reported to result in tissue damage during seed germination under abiotic stresses (Borsani, Valpuesta, & Botella, [3] ). Thus, the negative effect of BTH seed soak treatment reported by this study and others (Buzi et al., [4] ; Pitblado, [30] ; Ramasamy et al., [33] ) could be due to ROSs generation from BTH. Future work is needed to determine whether the traits linked to a greater response to BTH to cause SAR in line 7024 are the same as the traits linked to a greater response to BTH to cause inhibition of seed germination and seedling emergence.

To reduce the phytotoxicity of pesticides applied to seed, like neonicotinoids, the Phyto‐Drip® method was developed applying pesticides to seed as a single drop per seed when sowing, and is currently being used commercially in the Netherlands and Belgium (Petersen, van der Maarel, Rosman, & Bletscher, [28] ). In this study, seed and soil drip application methods were used to apply BTH, which are similar to the Phyto‐Drip® method. The seed drip method reduced the negative germination effects of BTH, and the soil drip application further reduced the negative germination effects of BTH on germination. However, future work is needed to see whether those methods can also effectively induce SAR.

This study shows that the effects of BTH to induce SAR, increase expression of certain defence genes and delay seed germination are dependent upon plant genotype. Testing plant genotypes with greater levels of genetic diversity, such as those that include introgression from wild species, will help to determine the potential range of response to BTH within a plant species. It is possible that conflicting reports about the success of BTH could be due to a limited range of genotypes for a particular crop chosen to be tested with the defence activator. However, negative effects of SAR activators on plant growth and development also need to be considered when choosing a plant genotype to test with BTH as that trait may also be directly linked to the stronger induced defence response to BTH.

Funding for this study was partially provided by the Ontario Ministry of Agriculture, Food and Rural Affairs—University of Guelph Partnership Program, the Ontario Tomato Research Institute and Syngenta Canada.

The authors have no conflict of interests.

Graph: Relative RT‐PCR of SlPR1a in breeding lines, (a) 7007 and (b) 7024 after seed treatment with 0.3 m m BTH ( ) or water ( ). Bars represent standard error. ‘*’ indicates significant difference at a time point, p  ≤  .05, Tukey's HSD

Graph: Relative RT‐PCR of SlPR3b in breeding lines, (a) 7007 and (b) 7024 after seed treatment with 0.3 m m BTH ( ) or water ( ). Bars represent standard error. ‘*’ indicates significant difference at a time point, p  ≤  .05, Tukey's HSD

Graph: Relative RT‐PCR of SlPin2 in breeding lines, (a) 7007 and (b) 7024 after seed treatment with 0.3 m m BTH ( ) or water ( ). Bars represent standard error. ‘*’ indicates significant difference at a time point, p  ≤  .05, Tukey's HSD

Graph: Relative RT‐PCR of SlPR2b in breeding lines, (a) 7007 and (b) 7024 after seed treatment with 0.3 m m BTH ( ) or water ( ). Bars represent standard error. ‘*’ indicates significant difference at a time point, p  ≤  .05, Tukey's HSD

Graph: Seed germination of breeding line 7007 following BTH seed soak ( ), water seed soak ( ), BTH soil drip ( ) or water soil drip ( ). Bars represent confidence intervals (95%). A ‘*’ indicates significant difference for seed soak, and a ‘¡’ indicates a significant difference for seed drip at a time point, p  ≤  .05, Tukey's HSD

Graph: Seed germination of breeding line 7024 following BTH seed soak ( ), water seed soak ( ), BTH soil drip ( ) or water soil drip ( ). Bars represent confidence intervals (95%). A ‘*’ indicates significant difference for seed soak, and a ‘¡’ indicates a significant difference for seed drip at a time point, p  ≤  .05, Tukey's HSD

Graph: Seedling emergence of breeding line 7007 following BTH seed soak ( ), water seed soak ( ), BTH soil drip ( ) or water soil drip ( ). Bars represent standard error of the mean. A ‘*’ indicates significant difference for seed soak, and a ‘¡’ indicates a significant difference for soil drip at a time point, p  ≤  .05, Tukey's HSD

Graph: Seedling emergence of breeding line 7024 following BTH seed soak ( ), water seed soak ( ), BTH soil drip ( ) or water soil drip ( ). Bars represent standard error of the mean. A ‘*’ indicates significant difference for seed soak, and a ‘¡’ indicates significant difference for soil drip at a time point, p  ≤  .05, Tukey's HSD