CRISPR/Cas9-mediated editing of Bs5 and Bs5L in tomato leads to resistance against Xanthomonas

Bacterial spot, caused by Xanthomonas species, is a devastating disease of tomato (Solanum lycopersicum) and pepper (Capsicum annuum) (Schwartz et al., 2015). The recessively inherited resistance, bacterial spot 5 (bs5), in pepper (hereafter referred to as Cabs5) can confer resistance against different Xanthomonas strains (Jones et al., 2002). The Cabs5 resistance is characterized by the absence of disease symptoms, faint chlorosis at the site of infection, and reduced bacterial growth. Remarkably, commercial pepper varieties containing the bs5 allele show durable resistance, effectively impeding hypervirulent strain emergence in agricultural fields (Vallejos et al., 2010).

The CaBs5 gene, together with its paralog CaBs5-like (CaBs5L), has recently been cloned (Sharma et al., 2023; Szabó et al., 2023). CaBs5 encodes a 92 amino acid long protein possessing a cysteine-rich transmembrane (CYSTM) domain, which is implicated in various biotic and abiotic responses. Typically, the CYSTM domain contains conserved residues composed of four consecutive cysteines, followed by two hydrophobic amino acids. A recent study suggested that Cabs5 mediating the resistance against bacterial spot lacks these two conserved leucine residues within the CYSTM domain (Szabó et al., 2023).

Tomatoes and peppers are close relatives in the Solanaceae family and commonly susceptible to Xanthomonas infection. Based on the current findings in pepper, we hypothesized that modifying the ortholog of CaBs5 in tomato could confer resistance against Xanthomonas. Consequently, putative Bs5 (SlBs5) and Bs5L (SlBs5L) were identified in tomato based on homology to CaBs5. Both SlBs5 and SlBs5L were located on chromosome 9 with the same head-to-head orientation as their pepper homologues on chromosome 3 (Figure 1a). Despite short and highly similar amino acid sequences of SlBs5 and SlBs5L (Figure 1b), the conserved synteny and gene order in pepper and tomato genomes allowed the assignment of orthology for Bs5 and Bs5L.

The mechanism by which the double leucine deletion in Cabs5 leads to resistance against Xanthomonas remains elusive (Figure 1b). Yet, this deletion in the conserved CYSTM domain could potentially impair CaBs5's native functionality (Abell and Mullen, 2011). Following this assumption, we postulated that knocking out SlBs5 would produce similar outcomes to Cabs5. We aimed to disrupt both SlBs5 and SlBs5L to prevent possible functional complementation by SlBs5L, given their greater amino acid sequence similarity compared to CaBs5 and CaBs5L (Figure 1b).

We constructed a binary vector for Cas9 and a single-guide RNA (sgRNA) targeting conserved sequences present in both SlBs5 and SlBs5L (Figure 1c). Tomato variety Fla. 8000 was transformed with Agrobacterium. From the progeny of successful transformants, we selected two homozygous lines, Slbs5-1 and Slbs5-2, containing frameshift mutations in both genes (Figure 1c). These mutant lines were self-pollinated or backcrossed to the wild-type parent variety to segregate the T-DNA containing the Cas9-sgRNA cassette.

The resistance of the two selected mutant lines was qualitatively evaluated against Xanthomonas perforans GE485 with dip inoculation assays (Figure 1d). At 21 days post-inoculation, the wild-type leaves were covered by black spots indicative of Xanthomonas infection, while both Slbs5-1 and Slbs5-2 retained green leaves with fewer visible symptoms. These phenotypes remained consistent in inoculations of X. perforans 4B and Xanthomonas gardneri 153 (Figure S1).

Quantitative evaluation of bacterial growth further supported these findings. At 5 days post-infiltration with a low-density bacterial suspension, Slbs5-1 showed significant decreases in Xanthomonas populations compared to wild-type plants (Figure 1e). Such reductions were consistently observed for Slbs5-2 (Figure S2). However, Slbs5-1 could not significantly hinder Pseudomonas population growth.

We additionally examined the growth penalty associated with Slbs5-1 and Slbs5-2 in controlled conditions (Figure 1f). The height of plants was measured at two different time points, but no significant differences were observed between the wild type and the two mutant lines (Figure 1f; Figure S3). This suggested that the resistance to Xanthomonas species comes at no developmental cost in the vegetative stage in the laboratory setting.

Although Cabs5-mediated immunity is subtle, it has shown practical value in commercial pepper cultivation. To examine the commercial potential of Slbs5, field trials were conducted with both Slbs5-1 and Slbs5-2 lines at the Gulf Coast Research and Education Center in Florida, a major state for tomato production. Along with naturally occurring Xanthomonas populations, a two-isolate cocktail of X. perforans race T4 was inoculated in the field to heighten disease pressure. Plants were grown with recommended fertilizers and pest management programs, excluding the use of any bactericides or activators of systemic acquired resistance.

Despite seasonal variations, Slbs5 mutant lines consistently maintained reduced disease symptoms (Figure 1g). Additionally, no developmental defects, such as stunting, were observed in these mutants (Figure 1h). Quantification of disease severity, based on visible symptoms caused by Xanthomonas infection on plant leaf surfaces, revealed higher percentages of Slbs5-2 leaves with reduced disease symptoms than wild-type leaves in all tested seasons (Figure 1i; Figure S4). Notably, the Slbs5-2 mutants demonstrated effective resistance during three periods of elevated disease pressure, Spring 2018, Fall 2019, and Fall 2023.

The marketable yield of fruits is a critical consideration in tomato cultivation. We quantified total marketable yield across five seasonal trials, except for two seasons impacted by a hurricane (Fall 2022) and extremely dry weather (Spring 2023). Throughout all seasons, there was no statistically significant difference in marketable fruit yields between Slbs5-2 and the wild-type plants (Figure 1j; Figure S5). However, during the three periods of increased disease prevalence in Spring 2018, Fall 2019, and Fall 2023 (Figure 1i), the mutants consistently showed a tendency to produce a greater quantity of marketable tomatoes (Figure 1j). This possibly suggests a correlation between Xanthomonas resistance of the mutant lines and improved fruit yields.

Overall, this study shows that a knockout of SlbBs5 and SlBs5L in tomatoes represents a promising strategy to achieve broad-spectrum resistance to bacterial spot disease. Compared to stronger sources of resistance, the resistance mediated by Slbs5 and Slbs5L may be considered subtle. However, our mutant lines consistently led to a reduced population of Xanthomonas in laboratory and field conditions. This decrease in pathogen populations could lessen the likelihood of hypervirulent strain emergence. Furthermore, when these mutants are combined with other sources of downstream resistance genes, they may serve as a prior layer of defence. This initial protection has the potential to diminish the probability of pathogen effectors directly interacting with and overcoming the resistance genes, possibly extending the efficacy of durable resistance in the agricultural field.

A.O. and B.J.S. conceptualized the project. B.J.S. supervised the project. A.O., D.D. and B.J.S. designed the experiments and helped analyze the data. K.S. and E.S performed bioinformatics analyses. K.S. led statistical analyses and designed the figures. A.S. helped plan the project, designed and tested the guide RNAs and did preliminary genotyping and bacterial disease assays. A.O. did further genotyping, guide RNA testing and conducted disease and phenotype assays of progeny. D.P.T.T., J.V.W., J.B.J, G.M, E.S.O and D.D. performed supplemental bacterial growth assays. E.S. and S.H. conducted field trials. M.J.C. supervised the generation of tomato mutant lines. E.Z. and J.P. conducted tomato transformations. A.O., K.S. and D.P.T.T. analyzed the data and wrote the manuscript.