Tomato Ubiquitinome Reveals How Two Bacterial Haplotypes Drive Different Disease Outcomes

A devastating bacterial pathogen that travels inside the bodies of tiny psyllid insects has long puzzled plant scientists: two closely related strains can infect the same tomato plant, yet one leaves it stunted while the other kills it outright. A new study published in the journal Crop Health has now mapped, in unprecedented molecular detail, how these two strains reshape the chemical tagging system that plants use to regulate their own proteins. The findings point to the ubiquitin–proteasome system, a cellular recycling and signaling machinery, as a central battleground in the struggle between tomato and the bacterium ‘Candidatus Liberibacter solanacearum’, known as Lso.

Lso is transmitted by psyllids and causes serious diseases in solanaceous crops, including the notorious zebra chip disorder of potato. Among its known genetic variants, haplotypes A and B both infect tomato and other members of the nightshade family in the Americas and New Zealand, but their consequences diverge sharply. Four weeks after infection, tomato plants carrying either haplotype begin to show leaf deformation and chlorosis, and LsoA-infected plants go on to develop small leaves, stunted growth, and shoots from axillary buds. LsoB-infected plants, by contrast, deteriorate much faster: necrosis appears and, by around eight weeks, many plants are dead. Understanding why two such similar bacteria produce such different diseases has been a long-standing goal of the research group led by Julien G. Levy and Cecilia Tamborindeguy at Texas A&M University.

The team had a specific molecular suspect in mind. Earlier work from the same laboratory identified HPE1, a protein secreted by Lso that interacts with tomato RAD23 proteins. RAD23 proteins act as shuttle receptors in the ubiquitin–proteasome system: they carry proteins that have been tagged with ubiquitin, a small 76-amino-acid marker, to the 26S proteasome, the cell’s protein-destruction machine. When HPE1 was expressed in plant leaves, ubiquitinated proteins accumulated, and a similar pile-up was seen in LsoB-infected tomato leaves. Because the ubiquitin–proteasome system is a well-known target of microbial effectors, phytoplasmas being a classic example with their SAP54 phyllogen that hijacks RAD23 to destroy plant transcription factors, the researchers reasoned that cataloging the full set of ubiquitinated proteins, the ubiquitinome, in infected tomato leaves could reveal how Lso rewires plant physiology.

Timing was critical to the experimental design. Lso distributes itself unevenly within plants, and the titers of the two haplotypes diverge as disease progresses. The researchers therefore sampled the top leaves of tomato plants exactly four weeks after psyllid-mediated infection, a point at which LsoA and LsoB titers are similar in those leaves and before the most severe symptoms of LsoB infection emerge. Control plants were infested with Lso-free psyllids to account for the effects of insect feeding itself. Using ubiquitin remnant profiling, a mass spectrometry technique that enriches for peptides carrying the GlyGly footprint left by ubiquitin on lysine residues, the team identified 48,605 peptide precursors, of which 4,070 were ubiquitinated, corresponding to 941 distinct ubiquitinated proteins.

The results showed that each haplotype leaves a distinct fingerprint on the tomato ubiquitinome. Principal component analysis cleanly separated the three groups of plants, and LsoB infection produced a broader perturbation than LsoA: 710 proteins were differentially represented in the LsoB comparison with uninfected controls, compared with 357 for LsoA, echoing earlier transcriptomic work showing that LsoB triggers more extensive gene expression changes. Importantly, 281 proteins changed in both infections, representing a core response to Lso that likely underlies shared symptoms, while only 144 proteins distinguished the two haplotypes directly. This suggests that early in disease the two bacteria push the plant along largely similar molecular trajectories, and that the dramatic difference in outcomes may hinge on a smaller set of decisive changes.

Functional enrichment analyses revealed what those trajectories involve. In LsoA-infected plants, differentially ubiquitinated proteins were enriched for glucose metabolism and proteasome-mediated, ubiquitin-dependent protein catabolism, and pathway analysis highlighted carbon fixation through the Calvin cycle and general carbon metabolism. LsoB produced a much wider sweep of enriched terms spanning carbohydrate and energy metabolism, transport, translation, proteasome activity, and oxidative stress. Notably, most Calvin cycle proteins were under-represented in the ubiquitinome of infected plants even before chlorosis became visible, indicating that the molecular events leading to the yellowing symptom begin well before it can be seen. Because Lso is confined to the phloem and disrupts sugar transport, these metabolic shifts likely reflect both the pathogen’s energetic demands and a disturbed source-sink balance in the plant.

The team then tested whether changes in the ubiquitinome translated into actual protein degradation. They expressed seven candidate tomato proteins in Nicotiana benthamiana, a model host in which LsoA and LsoB produce disease outcomes similar to those in tomato, and used the proteasome inhibitor MG132 to probe degradation. Among the four tomato RAD23 proteins, RAD23a and RAD23e were detectable only when the proteasome was blocked, regardless of infection status, confirming their constitutive turnover. RAD23d, however, behaved differently depending on the pathogen: in uninfected and LsoB-infected leaves it accumulated even without inhibitor, but in LsoA-infected leaves it was detectable only with MG132, meaning LsoA infection drove its degradation. Since RAD23d is one of the RAD23 proteins that interacts with the HPE1 effector, this haplotype-specific effect hints that LsoA and LsoB manipulate the RAD23-mediated degradation pathway through different mechanisms, possibly involving effectors that have not yet been characterized.

Two additional proteins told a striking story about LsoB virulence. PLANT UBX DOMAIN-CONTAINING PROTEIN 4, a cofactor of the p97 ATPase involved in extracting proteins from the endoplasmic reticulum during ER-associated degradation, and HOP-INTERACTING PROTEIN THI111, a DDI1-like shuttle that delivers ubiquitinated cargo to the proteasome, were both over-represented in the LsoB ubiquitinome yet undetectable in LsoB-infected plants unless MG132 was applied. Their targeted degradation in LsoB-infected plants could help explain the accumulation of ubiquitinated proteins previously observed in these infections. Conversely, NEDD8-CONJUGATING ENZYME UBC12, an E2 enzyme central to ubiquitin-like protein transfer, was under-represented in LsoB samples and was detected in infected plants even without inhibitor, consistent with reduced ubiquitination pathway activity. RT-qPCR confirmed that Rad23 transcript levels were unchanged by infection, indicating that these alterations occur at the protein level rather than through gene regulation.

Among the proteins that distinguished the haplotypes, superoxide dismutase stood out: it was over-represented in LsoB infection but under-represented in LsoA, aligning with evidence that LsoB-infected plants accumulate more reactive oxygen species and that oxidative stress correlates with symptom severity in Liberibacter diseases. Enrichment of water-channel and water-stress-related terms in LsoB also fits the known link between Liberibacter infection and plant water deficit. Patellin-3, a membrane-trafficking protein that can influence viral cell-to-cell movement, was over-represented in LsoB and under-represented in LsoA, making it an intriguing candidate for regulating how the phloem-restricted bacterium distributes itself within the plant. The authors caution that ubiquitination does not always lead to degradation and that some detected changes may reflect altered protein abundance, newly synthesized proteins failing quality control, or ER stress rather than pure regulatory tagging. Even so, the study substantially expands the inventory of ubiquitinated tomato proteins and their modification sites, and it establishes the ubiquitin–proteasome system as a key arena in which Lso haplotypes diverge. By pinpointing the proteins and pathways altered before visible symptoms diverge, the work offers breeders and plant pathologists concrete molecular targets for developing tomato varieties that tolerate or resist these psyllid-borne bacteria, and it underscores how a pathogen’s ability to co-opt a host’s protein recycling machinery can determine whether infection means stunted growth or death.

Subject of Research: Haplotype-specific changes in the tomato ubiquitinome during infection by the psyllid-transmitted bacterium 'Candidatus Liberibacter solanacearum'

Article Title: Tomato ubiquitinome in response to ‘Candidatus Liberibacter solanacearum’ haplotypes A and B

Article References: Levy, J. G., Liu, J., Oh, J., Mendoza Herrera, A., & Tamborindeguy, C. (2026). Tomato ubiquitinome in response to ‘Candidatus Liberibacter solanacearum’ haplotypes A and B. Crop Health, 4(1), Article 15. https://doi.org/10.1007/s44297-026-00075-6

Image Credits: AI Generated

DOI: 10.1007/s44297-026-00075-6

Keywords: tomato, Candidatus Liberibacter solanacearum, ubiquitinome, ubiquitin-proteasome system, plant pathology, psyllid, Lso haplotypes, HPE1 effector, RAD23 proteins, proteomics, Calvin cycle, oxidative stress