Bacteria Causing Pith Necrosis and Tissue Discoloration in Tomato

Abstract

Tomato is one of the most important vegetable crops in Costa Rica, where favorable environmental conditions enabled year-round production but also promote bacterial diseases. In recent years, pith necrosis has been frequently observed; nevertheless, the causal agents remain unidentified in the country. This study evaluated bacteria associated with symptomatic plants collected in the Central Valley of Costa Rica. From 32 plants, 61 bacterial isolates were obtained, described morphologically, and characterized through basic biochemical tests. Partial sequencing of the 16S rRNA gene revealed diverse bacterial taxa, predominantly belonging to the genus Pseudomonas. Thirteen isolates were selected for pathogenicity assays, which confirmed variable virulence levels. Multilocus sequence analysis based on concatenated sequences of the 16S rRNA, gyrB, rpoD, and rpoB genes identified Pseudomonas alliivorans LTM 13.1.2, P. flavescens LTM 14.2.2, and P. capsici LTM 78.3.2 as causal agents of pith necrosis. Additionally, P. straminea LTM 78.2.1 and Cedecea sp. LTM 72.2.1 caused tissue discoloration. Whole-genome sequencing of the two most virulent isolates (LTM 13.1.2 and LTM 78.3.2) supported their taxonomic classification and revealed virulence-associated genes and biosynthetic clusters. This study represents the first report of these Pseudomonas species as tomato pathogens in Costa Rica and expands their known distribution and host ranges.

1. Introduction

Tomato (Solanum lycopersicum L.) is one of the most widely produced and consumed vegetables worldwide [1]. In Costa Rica, tomato cultivation primarily occurs in the Central Valley regions, with a production of 56,586 metric tons on 1187 hectares reported in 2023 [1]. The high demand for tomatoes in most countries makes it an intensive production crop in which techniques must be used to maximize yield per unit of land [2]. Also, producing countries must have climatic conditions such as temperatures around 18–32 °C and relative humidity around 55–90% that favor year-round cultivation [3,4] and measures to control the development and spread of fungal and bacterial diseases.

Bacterial infections affect crop yields, causing significant economic losses [5,6] which can reach up to 91% in some cases, such as bacterial wilt caused by Ralstonia solanacearum [7]. As a result, bacteria pose a significant global threat to agricultural production, and their control is crucial for ensuring food security [8,9]. Among the most frequently reported bacterial diseases in Costa Rica, bacterial wilt caused by the Ralstonia species complex is the most important and destructive. In recent years, the damage caused by other bacteria has focused on soft rot caused by Pectobacterium spp., bacterial speck caused by Pseudomonas spp., and bacterial spot caused by Xanthomonas spp. [10,11,12]. In addition to these diseases, others, such as pith necrosis, are either infrequently reported in crops or have not yet been officially recorded in the country. Notably, characteristic symptoms of pith necrosis were repeatedly observed in samples collected in various locations across the Central Valley of Costa Rica between 2021 and 2023.

Tomato pith necrosis was first described in the 1970s, caused by Pseudomonas corrugata [13]. Since then, other causal agents have been identified, including several Pseudomonas species such as P. mediterranea [14], P. cichorii [15], P. viridiflava [16], P. marginalis [17], P. fluorescens and P. putida [18], as well as species from other genera such as Xanthomonas perforans [19] and Pectobacterium carotovorum [20]. Pseudomonas corrugata remains the principal species associated with the disease and is now widely distributed across most tomato-producing regions worldwide [21]. In the American continent, P. corrugata has been reported in Brazil [22], Argentina [23], Uruguay [24], Mexico [25], and the United States of America [26]. In Costa Rica, the presence of P. corrugata is presumed; however, no official reports or targeted studies have addressed the identification of the species responsible for tomato pith necrosis. Disease incidence in tomato plantations can range from 10% to 100%, depending on factors such as the bacterial species, environment, cropping system, and management practices [13,17,26,27]. Additionally, up to 20% losses in fruit weight and size have been reported and related to the presence of pathogens such as P. corrugata and P. mediterranea [28]. In severe infections, the wilting caused by pith necrosis becomes widespread and leads to plant death [24].

These observations underscore the necessity of evaluating the presence of pathogenic bacteria in crops. Therefore, the objectives of this study were to isolate and preliminarily identify bacteria associated with pith necrosis symptoms in tomato plants collected from different locations in the Central Valley of Costa Rica, to assess the pathogenicity of selected isolates, to perform the phylogenetic identification of the pathogenic isolates using multilocus sequence analysis (MLSA), and to analyze the genomic features of the most virulent isolates through whole-genome sequencing.

2. Materials and Methods

2.1. Bacterial Isolation and Culture

From 2021 to 2023, a total of 32 tomato plants exhibiting symptoms of bacterial infection were collected from different regions in Costa Rica, including Paraíso (5), Turrialba (2), and Alvarado (3) in Cartago province; Alajuela (11) and Zarcero (4) in Alajuela province; and Barva (6) and Santa Bárbara (1) in Heredia province, all located in the Central Valley of Costa Rica. Bacterial isolation was performed by surface sterilizing plant samples with 1% sodium hypochlorite for 30 s, followed by 70% ethanol for 1 min, and three rinses with sterile distilled water. Internal stem tissues were aseptically excised from the advancing margin of the infection, cut into small fragments (~5 mm), macerated in 3 mL of sterile 0.85% saline solution, and streaked onto nutrient agar (NA; Oxoid, Basingstoke, Hampshire, UK). Plates were incubated inverted at 28 ± 2 °C for 24–72 h, and individual colonies from each plate were subcultured onto fresh NA to obtain pure isolates.

Each of the isolates obtained (n = 61) was characterized based on colony morphology and stored in nutrient broth (Oxoid, Basingstoke, Hampshire, UK) supplemented with 35% glycerol at −80 °C. Basic biochemical tests were performed on fresh cultures (<24 h) of the isolates, including Gram staining, potassium hydroxide (3% KOH) reaction, oxidase test using Microbact oxidase detection strips (Oxoid, Basingstoke, Hampshire, UK), catalase test with hydrogen peroxide, fluorescens on King’s B medium [29], growth and lactose fermentation on MacConkey agar (Oxoid, Basingstoke, Hampshire, UK), and fermentation of lactose, glucose, and/or sucrose on Triple Sugar Iron (TSI) agar (Oxoid, Basingstoke, Hampshire, UK).

2.2. Identification of Bacterial Isolates

Partial 16S rRNA gene sequencing was performed on the 61 obtained isolates. Genomic DNA was extracted from fresh bacterial colonies (<24 h) using the CTAB protocol [30] with modifications described by Trout [31]. DNA concentration and quality were assessed using a NanoDrop One spectrophotometer (Thermo Scientific, Waltham, MA, USA). PCR reactions were performed in a final volume of 25 µL containing 2.5 µL of 10× Dream Taq buffer (Thermo Scientific, Waltham, MA, USA), 2.5 µL of 2 mM dNTP mix (Thermo Scientific, Waltham, MA, USA), 1.25 µL of 100 µM primers F27 and 1492R [32], 0.25 µL of Dream Taq DNA polymerase 5 U/µL (Thermo Scientific, Waltham, MA, USA), 1 µL of DNA template (60.6 ± 37.5 ng/µL), and ultrapure water to complete the volume. The thermal profile consisted of an initial denaturation at 94 °C for 2 min, followed by 35 cycles of 94 °C for 1 min, 56 °C for 1 min, and 72 °C for 2 min, with a final extension at 72 °C for 7 min. PCR amplification was verified by electrophoresis on a 1% agarose gel at 100 V for 35 min. An aliquot of 5 µL of each PCR product was mixed with 1.5 μL of 6× DNA Loading Dye (Thermo Scientific, Waltham, MA, USA) to stain the amplified fragments, and 1.5 μL of GeneRuler 100 bp Plus DNA Ladder (Thermo Scientific, Waltham, MA, USA) was used as a DNA size reference. Gels were visualized using a UV transilluminator to evaluate the size and quality of the amplified fragments.

PCR products were purified using 1.5 µL of Exonuclease I (20 U/µL; Thermo Scientific, Waltham, MA, USA) and 3 µL of 10× reaction buffer per 15 µL of PCR product, followed by incubation at 37 °C for 15 min and enzyme inactivation at 85 °C for 15 min in a thermocycler. Purified products were sent to Macrogen (Seoul, Republic of Korea) for Sanger sequencing [33] using the same primers as in the PCR amplification. Sequence trimming and consensus assembly were conducted using BioEdit v. 7.0.5.3 [34]. The resulting sequences were compared with type strain entries in the NCBI GenBank database using the BLASTn tool v. 2.16.0 [35].

2.3. Pathogenicity Test

Thirteen isolates were selected for pathogenicity assays based on previous reports of plant-pathogenic or plant-associated bacteria [36,37,38,39,40] to confirm whether the isolates could cause pith necrosis or related stem tissue symptoms in tomato plants. The inoculum was prepared from fresh cultures on tryptic soy agar (TSA); colonies were suspended in phosphate-buffered saline (PBS) to a concentration of 1 × 10⁸ CFU/mL. Five-week-old seedlings of the commercial tomato variety “JR”, which is susceptible to bacterial diseases, and the “Acorazado” variety, developed by the Tomato Breeding Program at the Fabio Baudrit Moreno Agricultural Experiment Station (EEAFBM-UCR) and carrying resistance genes to Ralstonia solanacearum phylotype I and Tomato Yellow Leaf Curl Virus (TYLCV) strain Israel [41], were used as plant material. Seedlings were transplanted into 12.8 × 10 cm pots filled with sterile soil and maintained under greenhouse conditions at 26.3 °C ± 2.9 °C and 58.8% ± 9.1% relative humidity.

Inoculation was performed by directly injecting 100 µL of bacterial suspension into the stem at the height of the first true leaf. Six plants per bacterial isolate and per variety were inoculated, along with six plants per variety as PBS-inoculated controls. In total, 168 plants were evaluated over a 21-day period [21 days after inoculation (dai)]. Plant growth, defined as the difference between final height and initial height (${Height}_{21 dai}-{Height}_{0 dai}$) was measured, and the presence of external and internal symptoms was assessed. Internal symptoms were evaluated by making a longitudinal cut of the stem from the inoculation point towards both ends. The extent of pith necrosis or tissue discoloration (cm) was measured in plants that exhibited the symptoms. The severity scale described by Aysan [42] was used to classify the virulence of the isolates. These were classified as “non-virulent” if no symptoms were observed; “low-virulent” if necrosis extended 0.1–2.0 cm; “virulent” if necrosis extended 2.1–4.0 cm; and “highly virulent” if necrosis extended ≥4.1 cm. To fulfill Koch’s postulates, the bacteria were reisolated from the symptom on nutrient agar (NA), and their identity was confirmed through partial 16S rRNA gene sequencing.

2.4. Classification and Phylogenetics of Pathogenic Isolates

Pseudomonas isolates causing pith necrosis were identified by Multilocus Sequence Analysis (MLSA) using concatenated partial sequences of the 16S rRNA [32], gyrB [43], rpoD [44], and rpoB [45] genes. Genomic DNA was extracted from fresh bacterial colonies (<24 h) using the CTAB protocol [30] with modifications described by Trout [31]. PCR reactions were carried out in a final volume of 25 µL containing 2.5 µL of 10× Dream Taq buffer (Thermo Scientific, Waltham, MA, USA), 2.5 µL of 2 mM dNTP mix (Thermo Scientific, Waltham, MA, USA), 1.25 µL of each 100 µM primer (Table S1), 0.25 µL of Dream Taq DNA polymerase 5 U/µL (Thermo Scientific, Waltham, MA, USA), 1 µL of DNA template, and ultrapure water to complete the volume. The thermal profile for each reaction was programmed using amplification parameters specific to each primer pair (Table S1). PCR product amplification was confirmed by 1% agarose gel electrophoresis and visualized using a UV transilluminator to assess fragment size and quality.

PCR products were purified with Exonuclease I and sequenced at Macrogen (Seoul, Republic of Korea) using the Sanger method [33]. Sequences were edited and assembled in BioEdit v. 7.0.5.3 [34], along with type strain sequences downloaded from the NCBI GenBank database (Table S2). Sequence alignment for each locus was performed using MAFFT v. 7 [46] with default parameters. Aligned and trimmed sequences of the genes 16S rRNA (1266 bp), gyrB (443 bp), rpoD (409 bp), and rpoB (669 bp) were concatenated into a single matrix using Mesquite v. 3.61 [47], totaling 2787 bp. JModelTest v. 2.1.10 [48] was used to determine the best nucleotide substitution model for each locus.

Bayesian phylogenetic analysis was performed in MrBayes v. 3.2.6 [49]; four chains were run under the GTR+I+G model for the 16S rRNA, gyrB, and rpoD genes, and the SYM+I+G model for rpoB, with 1,000,000 Markov chain Monte Carlo (MCMC) generations. Pseudomonas aeruginosa strain DSM 50071^T was used as the outgroup (Table S2). A maximum likelihood phylogenetic analysis was also conducted using RAxML v. 8.2.12 [50] through the raxmlGUI v. 2.0.9 interface [51], with the GTR+I+G model, 1000 replicates, and bootstrap support values. The consensus phylogenetic tree was edited in iTOL v. 6.9.1 [52]. Partial housekeeping gene sequences (gyrB, rpoB, and rpoD) of pathogenic Pseudomonas isolates were deposited in the GenBank database under accession numbers PQ883186–PQ883197.

Non-pathogenic and/or non-Pseudomonas bacteria were identified by phylogenetic analysis based on partial 16S rRNA gene sequences along with type strain sequences from the GenBank database (Table S3), using the same procedure as the multilocus analysis, except for the concatenation step. The GTR+I+G substitution model was used, with a total alignment length of 1346 bp, and Bacillus subtilis strain IAM 12118^T was used as the outgroup (Table S3). Partial 16S rRNA gene sequences of the bacterial isolates from this study were deposited in GenBank within the accession numbers PQ857593–PQ857653.

2.5. Genome Sequencing and Analysis

A single colony of virulent isolates Pseudomonas alliivorans LTM 13.1.2 and Pseudomonas capsici LTM 78.3.2 picked from the NA media plates was used for genomic DNA extraction using the CTAB method [30] with some modifications described by Trout [31]. DNA concentration and quality were assessed using a NanoDrop One spectrophotometer (Thermo Scientific, Waltham, MA, USA). The DNA libraries were prepared using the Illumina DNA Prep Kit (Illumina, San Diego, CA, USA), and the whole genome sequencing was performed using the Illumina NovaSeq Plus X (2 × 151) paired end method at SeqCenter in Pittsburgh, Pennsylvania (USA). The raw reads were trimmed using Trimmomatic v. 0.36 [53] and assembled with Unicycler v. 0.4.8 [54]. The quality of the genomes was assessed with Quast v. 5.2.1 [55]. The taxonomic classification of the genomes was conducted using GTDB-Tk v. 2.0.0 [56], and the functional annotation was conducted using EggNOGmapper v. 2.1.12 [57]. Other genomic features such as secondary metabolites and virulence factors were also annotated using antiSMASH v. 8.0 [58] and MmetaVF toolkit [59], respectively. Genome data can be found in the NCBI BioProject accession number PRJNA1314254. For details on the draft genomes, see [60].

2.6. Statistical Analysis

A linear model was fitted to evaluate the mean plant growth $∆Height$ as a function of bacterial isolate, tomato variety, and their interaction. Model selection based on the Akaike Information Criterion (AIC) supported the inclusion of the interaction term:

$${\mu }_{{∆Height}_{ijk}}={\beta }_0+{\beta }_1{isolate}_j+ {\beta }_2{var}_k+{\beta }_3({isolate}_j\times {var}_k)+{ϵ}_{ijk}$$

(1)

where ${∆Height}_{ijk}$ represents the height increment of plant $i$ exposed to bacterial isolate $j$ and tomato variety $k$; ${\beta }_0$ is the intercept corresponding to the reference isolate and variety; ${\beta }_1$ and ${\beta }_2$ represent the main effects of specific bacterial isolate and tomato variety, respectively; ${\beta }_3$ captures the interaction between bacterial isolate and variety; and ${ϵ}_{ijk}$ is the residual error term, assumed to follow a normal distribution with mean zero and constant variance.

The same linear model structure was used to evaluate the effect of bacterial isolates, tomato variety, and their interaction on internal stem lesions, with lesion length ($Nec$) as the response variable. The model with interaction was selected based on AIC as the best fit for the data:

$${\mu }_{{Nec}_{ijk}}={\beta }_0+{\beta }_1{isolate}_j+ {\beta }_2{var}_k+{\beta }_3({isolate}_j\times {var}_k)+{ϵ}_{ijk}$$

(2)

where ${Nec}_{ijk}$ represents the extent of internal lesion in each plant $i$ for each bacterial isolate $j$ and tomato variety $k$; ${\beta }_0$ is the intercept; ${\beta }_1$ and ${\beta }_2$ represent the main effects of specific bacterial isolate and tomato variety, respectively; ${\beta }_3$ captures the interaction between isolate and variety; and ${ϵ}_{ijk}$ is the residual error term, assumed to follow a normal distribution with mean zero and constant variance.

Fisher’s LSD test for pairwise comparisons was applied to the fitted linear models to determine statistically significant differences among each bacterial isolate–tomato variety combination.

To assess the effect of internal lesion length (pith necrosis) on plant growth (height increment), two multiple linear regression models were fitted: one including an interaction term between necrosis and tomato variety, and another without interaction. Model selection based on AIC favored the model without interaction:

$${\mu }_{{∆Height}_i}={\beta }_0+{\beta }_1{Nec}_i+ {\beta }_2{var}_i+{ϵ}_i$$

(3)

where ${∆Height}_i$ represents the expected increase in height of plant $i$; ${\beta }_0$ is the intercept; ${\beta }_1{Nec}_i$ is the coefficient for the extent of pith necrosis; ${\beta }_2{var}_i$ is the coefficient for the tomato variety; and ${ϵ}_i$ is the residual error term, assumed to follow a normal distribution with mean zero and constant variance.

All analyses were performed in the R environment [61]. Model selection based on AIC was carried out using the ‘performance’ package [62], mean contrast tests were conducted using the ‘agricolae’ package [63], and models were plotted using the ‘ggeffects’ [64], ‘sjPlot’ [65] and ‘ggplot2’ [66] packages.

3. Results

3.1. Description and Identification of Bacterial Isolates

A total of 61 bacterial isolates were obtained from 32 symptomatic tomato plants collected in the Central Valley of Costa Rica. The most frequently observed symptoms were internal stem necrosis (81%), wilting (69%), hollow stem (38%), and chlorosis (25%) (Figure S1). All isolates were characterized based on colony morphology, including shape, margin, elevation, color, texture, and appearance under emitted and reflected light, as well as basic biochemical tests such as Gram staining, catalase, oxidase, fluorescent pigment production on King’s B medium, growth and lactose fermentation on MacConkey agar and fermentation of glucose, lactose and/or sucrose on TSI agar (Table S4). Most isolates were Gram-negative (68.9%), catalase positive (93.4%), oxidase negative (70.5%), did not produce fluorescent pigments on KB medium, and did not ferment glucose, lactose and/or sucrose on TSI agar (73.8%) (Table S4). Preliminary identification based on partial 16S rRNA gene sequencing assigned the isolates to 23 genera, including Pseudomonas (34%), Microbacterium (11%), Stenotrophomonas (5%), Agrobacterium (5%), Serratia, (3%), and Pantoea (3%) (Figure S2).

3.2. Pathogenicity

Thirteen Gram-negative bacterial isolates were evaluated in pathogenicity assays to confirm their ability to cause pith necrosis or related stem tissue symptoms in tomato plants. Five of the thirteen isolates caused internal symptoms in both tomato varieties, three of which caused pith necrosis, and two caused tissue discoloration. Isolates that only caused slight oxidation at the inoculation site without affecting adjacent tissue were considered non-pathogenic. The most virulent isolate was Pseudomonas sp. LTM 13.1.2, with a pith necrosis extension of 11.55 ± 5.25 cm in var. “JR” and 4.57 ± 1.37 cm in var. “Acorazado” and was classified as “highly virulent” according to the severity scale described by Aysan [42] (Table S5, Figure 1). The necrosis caused by this isolate accounted for 27.3% of the final plant height (42.35 ± 2.23 cm) at 21 dai in var. “JR.” This was the only isolate that showed significant differences in necrosis extension between varieties, being 39.57% greater in var. “JR” (p < 0.05) (Figure 1a). The lesions caused by this isolate were dark brown, dry, with disaggregated and collapsed tissue. External symptoms such as discoloration and tissue depression at the inoculation site were also observed (Figure 1b).

The second most virulent isolate was Pseudomonas sp. LTM 78.3.2, also classified as “highly virulent”, with pith necrosis of 4.63 ± 1.03 cm in var. “JR” and 4.12 ± 1.86 cm in var. “Acorazado” (Table S5, Figure 1). The necrotic lesions were very similar to those caused by the isolate LTM 13.1.2, showing more whitish tissue at the edges of the collapsed core. Discoloration and tissue depression were observed on the epidermis at the inoculation site (Figure 1b). Pseudomonas sp. LTM 14.2.2 was classified as “virulent” for causing 3.10 ± 2.49 cm of necrosis in var. “JR” and 2.45 ± 2.37 cm in var. “Acorazado” (Table S5, Figure 1). The lesions were pale brown, and collapsed tissue was observed only in plants of the “JR” variety, similar to the lesions caused by LTM 78.3.2 isolate with whitish tissue at the edges of the collapsed core (Figure 1b). External discoloration at the inoculation site was also noted.

The isolate of Pseudomonas sp. LTM 78.2.1 was classified as “virulent” for causing 2.03 ± 1.83 cm of tissue discoloration in var. “JR”, and Cedecea sp. LTM 72.2.1 as “low-virulent” with 1.70 ± 2.31 cm of tissue discoloration in var. “Acorazado” (Table S5, Figure S3). These isolates showed greater variability than other isolates classified within the “low-virulent” range of the severity scale, which caused only slight oxidation at the inoculation site without spreading through the tissue. A trend of greater necrosis or discoloration extension was recorded in the “JR” variety, except for isolate LTM 72.2.1, which caused greater discoloration in the var. “Acorazado.” In contrast, three isolates of the genus Pseudomonas (LTM 32.2.1, LTM 40.2.3, and LTM 78.1.1), two of Stenotrophomonas sp. (LTM 49.1.1 and LTM 78.2.3), two of Serratia sp. (LTM 71.3.2 and LTM 74.3.1), and one of Pantoea sp. (LTM 78.3.1) were classified as non-pathogenic, causing no visible symptoms in internal stem tissue or only slight oxidation at the inoculation site, which was not considered tissue discoloration as observed with isolates LTM 72.2.1 and LTM 78.2.1. No necrosis or tissue discoloration was observed in control plants inoculated with PBS, confirming that the lesions in affected plants were caused by bacterial infection.

Plant growth (${Height}_{21 dai}-{Height}_{0 dai}$) was variable, ranging from 4.1 to 32.4 cm, with a mean of 19.91 ± 0.36 cm. Plants of the “Acorazado” variety inoculated with Cedecea sp. LTM 72.2.1 showed the lowest average growth (12.38 ± 2.01 cm), followed by those inoculated with Pantoea sp. LTM 78.3.1 (13.62 ± 2.48 cm) of the same variety (Table S5). The highest growth was recorded in the control plants of the “Acorazado” variety (25.33 ± 2.89 cm) and plants of the “JR” variety inoculated with Pseudomonas sp. LTM 78.2.1 (24.47 ± 1.60 cm). A general trend of lower growth was observed in most plants of the “Acorazado” variety compared to those of the “JR” variety when inoculated with the same isolate, with significant differences (p < 0.05) for isolates LTM 72.2.1, LTM 78.2.1, and LTM 78.3.1 (Table S5, Figure 2a). The effect of internal lesion on plant growth was analyzed, and although an average reduction of −0.20 ± 0.12 cm (mean ± standard error) was observed for each cm increase in internal lesion, it was not statistically significant (p > 0.05). However, a significant difference in growth between varieties was found (p < 0.05) (Table S6, Figure 2b).

3.3. Classification and Phylogenetics

Multilocus phylogenetic analysis of pathogenic Pseudomonas species based on concatenated sequences of the 16S rRNA, gyrB, rpoD and rpoB genes through Bayesian inference and maximum likelihood methods showed that isolates LTM 13.1.2 and LTM 78.3.2 clustered within the Pseudomonas syringae group. Isolate LTM 13.1.2 grouped with Pseudomonas alliivorans 20GA0068^T showing a posterior probability (PP) = 1 and bootstrap support (BS) = 100, and was closely related to P. viridiflava, a causal agent of pith necrosis. Isolate LTM 78.3.2 grouped with Pseudomonas capsici Pc19-1^T (PP = 1; BS = 100) and was closely related to P. cichorii, a well-known causal agent of pith necrosis. Isolates LTM 14.2.2 and LTM 78.2.1 clustered within the P. straminea group. Isolate LTM 14.2.2 grouped with Pseudomonas flavescens B62^T (PP = 1; BS = 97), and isolate LTM 78.2.1 with Pseudomonas straminea IAM 1598^T (PP = 1; BS = 98) (Figure 3).

Isolate LTM 72.2.1 was classified as Cedecea sp. with Cedecea neteri JCM 7582^T as the closest type strain (PP = 0.97; BS = 83) through phylogenetic analysis based on partial sequences of the 16S rRNA gene (Figure S4). Non-pathogenic isolates were also classified by 16S rRNA phylogenetic analysis, which identified isolates LTM 32.2.1, LTM 40.2.3, and LTM 78.1.1 as Pseudomonas sp. clustering with the Pseudomonas putida group. Isolates LTM 49.1.1 and LTM 78.2.3 were classified as Stenotrophomonas sp.; isolates LTM 74.3.1 and LTM 71.3.2 as Serratia sp; and isolate LTM 78.3.1 as Pantoea sp. with Pantoea deleyi LMG 24200^T as the closest type strain (PP = 1; BS = 99) (Figure S4).

3.4. Genome Features and Annotations

Genome sequencing of the two most virulent isolates using Illumina technology results in a genome assembly size of 6.15 Mb for isolate LTM 13.1.2 with a coverage of 143×, and 5.87 Mb for isolate LTM 78.3.2 with a coverage of 146×. The number of contigs in each genome assembly were 52 and 54. The N50 of the assemblies were 249,463 bp and 182,980 bp. The total number of predicted genes in each genome were 5527 and 5000. The GC contents of the genomes were 59.1% and 58.37%, respectively (Table S7). Taxonomic classification of the isolates was confirmed with GTDB-Tk tool and database, the Average Nucleotide Identity (ANI) shows that isolate LTM 13.1.2 has an ANI value of 98.86% when compared to P. alliivorans 20GA0068^T, and isolate LTM 78.3.2 has an ANI value of 99.53% when compared to P. capsici Pc19-1^T.

Virulence factors and the gene clusters associated with secondary metabolites were annotated for each genome assembly. Twenty-eight genes related to virulence factors such as Type IV pili (T4P), alginate biosynthesis (mucoid exopolysaccharide), pyoverdine (siderophore) and flagella were detected in the genome of P. alliivorans LTM 13.1.2. For P. capsici LTM 78.3.2 genome assembly, thirty genes associated with virulence factors including T4P, Tap T4P, alginate, pyoverdine, and flagella were detected (Table S8).

Regarding biosynthetic gene clusters, eleven clusters associated with secondary metabolites were detected in P. alliivorans LTM 13.1.2 genome, including arylpolyene, hydrogen cyanide, non-ribosomal peptide synthetase (NRPS), and polyketide synthase (PKS) products, siderophores, N-acetylglutaminylglutamine amide (NAGGN), a redox-cofactor, a terpene, and several NRPS-derived compounds. In P. capsici LTM 78.3.2 genome, fourteen clusters were identified, comprising NRPS, PKS, siderophores, NAGGN, arylpolyene, hydrogen cyanide, a redox-cofactor, and diverse NRPS-derived products (Table S9).

4. Discussion

This study provides the first evidence of five bacterial species affecting tomato plants in Costa Rica. Pseudomonas alliivorans, recently found to be associated with foliar symptoms in onion and cucurbits [38,67], is reported here for the first time as a causal agent of pith necrosis in tomato. Likewise, P. capsici, a species associated with foliar diseases in a wide range of hosts [39,67,68,69], is also reported for the first time causing tomato pith necrosis, as well as its presence in Costa Rica, with the additional relevance that it can induce symptoms at higher temperatures compared to other species like Pseudomonas syringae [67]. Given the symptoms induced by these isolates at early stages of tomato cultivation under experimental conditions, further evaluation of their effect on crop yield is necessary, despite the fact that some studies have reported that such symptomatology caused by related species like P. viridiflava does not always result in significant yield losses [70]. Furthermore, the ability to colonize both foliar and vascular tissues underscores the potential phytosanitary risk these bacteria pose not only to tomato but also to other solanaceous crops in tropical environments. Consequently, it is imperative to implement monitoring strategies to prevent their spread, as well as effective management strategies.

The phylogenetic proximity between P. alliivorans with P. viridiflava [38] and P. capsici with P. cichorii [39], well-known tomato pathogens [15,16,23,24], raises the possibility of historical misidentifications in different regions of the world, as has occurred with several Pseudomonas species [71]. Moreover, recent taxonomic revisions showed that some strains previously identified as Pseudomonas cichorii are in fact P. capsici [69], reinforcing the need for accurate molecular identification and ongoing epidemiological surveillance to correctly map pathogen distributions in tomato-producing regions.

The isolation of P. flavescens from symptomatic tomato plants further illustrates the growing diversity of Pseudomonas taxa capable of colonizing and damaging tomato vascular tissues. As with the previous isolates, this is the first report of P. flavescens causing pith necrosis in tomato. Although this species has been only sporadically reported in association with other hosts [72,73], its association here with necrosis suggests that it may contribute to a broader disease complex in tomato. The bacterial isolates that caused pith necrosis also triggered external symptoms such as tissue discoloration and sunken lesions at the inoculation site, similar to those described for other causal agents of pith necrosis [15,27].

In this study, three Pseudomonas species were identified as causal agents of pith necrosis in tomato plants. However, none of these species had been previously reported causing this symptom. The disease has been mainly associated with Pseudomonas corrugata [27], but a wide range of studies attributes its development to several Pseudomonas species including P. mediterranea [14], P. cichorii [15], P. viridiflava [16], P. marginalis [17], P. fluorescens, and P. putida [18], as well as species from other genera, such as Xanthomonas perforans [19] and Pectobacterium carotovorum [20]. The identification of Pseudomonas species that induce analogous internal symptoms suggests that the etiology of this disease may be more complex than previously recognized. Furthermore, previously reported synergistic interactions among Pseudomonas spp., such as P. straminea with other pathogenic Pseudomonas species and with Xanthomonas perforans [19,74] suggest that pith necrosis may be a result of the combined action of different bacterial taxa or of co-infections involving both pathogenic and opportunistic bacteria, which together exacerbate tissue degradation and symptom expression. This factor may complicate the diagnostic process and reduce the efficacy of single-target control measures.

Although some isolates exhibited low virulence under greenhouse conditions, their detection in symptomatic tissue should not be dismissed as incidental. Opportunistic bacteria detected in this study causing only tissue discoloration, such as Pseudomonas straminea and Cedecea sp., may act as secondary colonizers or potentially emerge as pathogens under specific environmental or host conditions [19,75]. The presence of pathogens from different taxonomic groups, such as Cedecea sp. [40,76], and the lack of reported pathogenicity in vegetables underscore the need for genomic and functional characterization to determine pathogenic potential and epidemiological relevance in vegetable crops.

The variability in plant growth observed across treatments suggests that the response to bacterial inoculation is not uniform and may be influenced by multiple factors, including the intrinsic physiological variability between plants [77]. Although a general trend of reduced growth was observed in the “Acorazado” variety compared to the “JR” variety when inoculated with bacterial isolates, this pattern was not evident in the PBS-inoculated control plants, where the highest growth rate was observed in the “Acorazado” variety, suggesting that the variety itself may not be inherently less vigorous. Notably, inoculation with highly virulent isolates did not result in measurable reductions in plant growth, suggesting that internal damage is not always associated with immediate loss in vegetative development, at least within the evaluation period.

The genome analysis of the two most virulent isolates revealed the presence of several virulence factors and biosynthetic gene clusters that may confer a competitive advantage as plant pathogens. Some of those genomic traits have been reported in well-known plant pathogenic bacteria such as Ralstonia solanacearum, Xylella fastidiosa, Acidovorax citruli, Xanthomonas spp., and Pseudomonas spp., and contribute to their pathogenicity and virulence [78,79,80,81]. Phytotoxin syringomycin detected in Pseudomonas capsici LTM 78.3.2 has been associated with necrosis promotion in the host and enhancement of bacterial growth in a wide variety of plants [82]. This toxin, along with virulence factors, may be involved in the high virulence expressed. Furthermore, genomic traits implicated in protection against plant defense mechanisms [83,84,85], microbial competition [86,87], and environmental stress tolerance [88,89,90,91] may benefit these isolates in microbe–microbe interactions and provide them with efficient means for epiphytic and environmental survival. Despite the damage observed in plants, the present study does not provide evidence that these virulence genes are expressed under in vitro or in vivo conditions.

This study represents the first effort in Costa Rica to identify bacteria associated with tomato pith necrosis. The findings of this study contribute to the existing body of knowledge by including new species of pathogenic bacteria that cause pith necrosis in tomato plants in Costa Rica. This suggests that the range of causal agents may be broader than previously recognized. Furthermore, the presence of bacteria that cause tissue discoloration could weaken the plant and make it more susceptible to attack by more aggressive pathogens. All of the above highlights the importance of considering aspects such as the identity and virulence of pathogenic organisms when diagnosing and implementing plant health management strategies.

5. Conclusions

Five bacterial isolates were identified as causal agents of internal symptoms in tomato: Pseudomonas alliivorans LTM 13.1.2, P. flavescens LTM 14.2.2, and P. capsici LTM 78.3.2 caused pith necrosis, while Cedecea sp. LTM 72.2.1 and P. straminea LTM 78.2.1 caused tissue discoloration. P. alliivorans LTM 13.1.2 was the most virulent isolate, followed by P. capsici LTM 78.3.2. Genome sequences of the most virulent isolates revealed several virulence factors and gene clusters associated with secondary metabolites that may contribute to the pathogenicity of these isolates, while also providing a competitive advantage over other microbes and enhancing environmental survival.