High Throughput Prediction of Citrus Huanglongbing (HLB) Resistance/Susceptibility Protein Interactions Using AlphaFold3

Monday, August 4, 2025

10:30 AM - 10:45 AM HST

Location: Room 313 A

Presenting Author(s)

MJ

Mukesh Jain, PhD

Research Coordinator
Department of Plant Pathology
Gainesville, Florida, United States

Author(s)

ZB

Zhenghong Bao

Bioinformatics Analyst
Department of Plant Pathology
Gainesville, Florida, United States
WB

Wardatou Boukari

Postdoc
Department of Plant Pathology
Gainesville, Florida, United States
TT

Trishna Tiwari

Department of Horticultural Sciences
Gainesville, Florida, United States
JH

Jose C. Huguet-Tapia, PhD

Research Scientist
University of Florida
Gainesville, FL, USA
CM

Carlos D. Messina

Professor
Department of Horticultural Sciences
Gainesville, Florida, United States
RL

Rosemary Loria

Professor
Department of Plant Pathology
Gainesville, Florida, United States

Abstract Text: Candidatus Liberibacter asiaticus (CLas) is an uncultured, psyllid-transmitted, and phloem-limited bacterial pathogen associated with the devastating citrus ‘greening’ disease (a.k.a. Huanglongbing or HLB). Molecular characterization of the citrus/CLas interactions remains challenging because the CLas genome lacks canonical pathogenesis-related secretion systems and effector-encoding genes. SignalP 6.0 identified secreted proteins in 18 CLas strains, which were then clustered into 73 orthogroups using OrthoFinder 2.5, resulting in 88 unique proteins. Database (Phytozome) mining of C. sinensis and C. clementina genomes identified 491 HLB-related and differentially expressed proteins in citrus phloem. AlphaFold3-mediated structural modeling of 43,208 citrus/CLas protein pairs predicted highly confident interactions with ipTM scores for 65 (=0.8-1.0), 983 (=0.6-0.8), and 7,154 (=0.3-0.6) pairs. Functional categorization of the citrus target proteins implicated a role of CLas-secreted proteins in HLB progression via modulation of key biological processes, such as oxidative stress responses, carbohydrate and energy metabolism, and cell wall and membrane biogenesis. These in silico predictions provide a valuable approach for high throughput screening and efficient selection of HLB resistance/susceptibility genes. Empirical validation of high-confidence interactions through differential expression, Yeast-2-Hybrid, and co-immunoprecipitation and pulldown analyses will help establish a prioritized framework for achieving HLB resistance via transgenic and/or gene-editing approaches.