Gene regulatory networks elucidating huanglongbing disease mechanisms

Abstract

Next-generation sequencing was exploited to gain deeper insight into the response to infection by Candidatus liberibacter asiaticus (CaLas), especially the immune disregulation and metabolic dysfunction caused by source-sink disruption. Previous fruit transcriptome data were compared with additional RNA-Seq data in three tissues: immature fruit, and young and mature leaves. Four categories of orchard trees were studied: symptomatic, asymptomatic, apparently healthy, and healthy. Principal component analysis found distinct expression patterns between immature and mature fruits and leaf samples for all four categories of trees. A predicted protein - protein interaction network identified HLB-regulated genes for sugar transporters playing key roles in the overall plant responses. Gene set and pathway enrichment analyses highlight the role of sucrose and starch metabolism in disease symptom development in all tissues. HLB-regulated genes (glucose-phosphate-transporter, invertase, starch-related genes) would likely determine the source-sink relationship disruption. In infected leaves, transcriptomic changes were observed for light reactions genes (downregulation), sucrose metabolism (upregulation), and starch biosynthesis (upregulation). In parallel, symptomatic fruits over-expressed genes involved in photosynthesis, sucrose and raffinose metabolism, and downregulated starch biosynthesis. We visualized gene networks between tissues inducing a source-sink shift. CaLas alters the hormone crosstalk, resulting in weak and ineffective tissue-specific plant immune responses necessary for bacterial clearance. Accordingly, expression of WRKYs (including WRKY70) was higher in fruits than in leaves. Systemic acquired responses were inadequately activated in young leaves, generally considered the sites where most new infections occur.

Figure 1. Gene Set Enrichment Analysis based on sparse principal component analysis (sPCA).

GO terms representing differentially regulated functional classes of genes in at least two of the four tissues (p<0.05). Color values correspond to p-values, as indicated by color bars. Total genes in each gene set (more than ten genes) are given in parentheses. Each column represents one of four tissues: IF, immature fruits; MF, mature fruits; YL, young leaves; ML, mature leaves. Complete GSEA results, including sets with <10 genes, and enriched in one tissue only, are given in Dataset S2.

Figure 2. HLB-regulation of photosynthesis and carbohydrate metabolism.

Overview of changes induced by HLB in the expression of genes affecting photosynthesis and small carbohydrate metabolism (AH vs. SY samples; see Table 1 for a key to abbreviations for sample types).

Figure 3. HLB-induced modulation of hormone-mediated immune responses.

Four AH vs. SY tissues were analyzed. Regulatory interactions between pathways are shown. See Table 1 for a key to abbreviations for sample types.

Figure 4. Predicted protein-protein interaction networks of Citrus responses to HLB disease.

Four citrus tissues were analyzed using the dataset of HLB-regulated genes in SY versus AH samples based on an Arabidopsis knowledgebase. HLB-regulated proteins are represented by the larger nodes. (Fold ratio >1 or <−1). See Table 1 for a key to abbreviations for sample types.

Figure 5. Transcriptional regulation of fruit and leaf metabolism and sugar transport during HLB disease.

qRT-PCR data compared AH and SY tissues. White triangles within filled squares indicate up- and down-regulated genes. Arrows pointing up and down likewise indicate increased and decreased metabolite concentration. In each bar graph, AH appears on the right, and SY on the left, where vertical axis units are relative ratios of the test gene and housekeeping gene abundance.