Small RNA profiling reveals phosphorus deficiency as a contributing factor in symptom expression for citrus huanglongbing disease

Abstract

Huanglongbing (HLB) is a devastating citrus disease that is associated with bacteria of the genus 'Candidatus Liberibacter' (Ca. L.). Powerful diagnostic tools and management strategies are desired to control HLB. Host small RNAs (sRNA) play a vital role in regulating host responses to pathogen infection and are used as early diagnostic markers for many human diseases, including cancers. To determine whether citrus sRNAs regulate host responses to HLB, sRNAs were profiled from Citrus sinensis 10 and 14 weeks post grafting with Ca. L. asiaticus (Las)-positive or healthy tissue. Ten new microRNAs (miRNAs), 76 conserved miRNAs, and many small interfering RNAs (siRNAs) were discovered. Several miRNAs and siRNAs were highly induced by Las infection, and can be potentially developed into early diagnosis markers of HLB. miR399, which is induced by phosphorus starvation in other plant species, was induced specifically by infection of Las but not Spiroplasma citri that causes citrus stubborn-a disease with symptoms similar to HLB. We found a 35% reduction of phosphorus in Las-positive citrus trees compared to healthy trees. Applying phosphorus oxyanion solutions to HLB-positive sweet orange trees reduced HLB symptom severity and significantly improved fruit production during a 3-year field trial in south-west Florida. Our molecular, physiological, and field data suggest that phosphorus deficiency is linked to HLB disease symptomology.

Figure 1.

Some csi-miRNAs and csi-siRNAs Were Differentially Expressed in Las-Positive and Las-Negative Plants. Expression levels at 10 wpi (A) and 14 wpi (B) are presented as fold changes (log2) in Las-treated samples over corresponding healthy samples. The csi-miRNA IDs are listed at the bottom while the fold change is at the left. ‘+1’ indicates a twofold induction; ‘–1’ indicates a twofold reduction; HLB, Las-infected; wpi, weeks post inoculation. Red bar highlights miR399, which is involved in C. sinensis P homeostasis. (C, D) Relative expression levels of some csi-siRNAs that changed more than fourfold at 10 and 14 wpi, respectively.

Figure 2.

csi-miR399 and csi-miR159 Were Induced in Las-Positive Plants. (A) Expression levels of some csi-miRNAs were examined by Northern hybridization. Total RNA was extracted from healthy and Las-infected plants at 10 and 14 wpi, as indicated at the top. 100 μg of total RNA (15 μg for miR159) was loaded into each well. Each blot was hybridized with a probe reverse complementary to the target small RNA, as indicated to the left. U6 was used as an internal control for quantification, shown below as relative abundance (RA). The RA of healthy samples was assigned to 1. (B) Expression level of csi-miR399 was examined using real-time PCR. RNAs were extracted from Las-infected ‘Valencia’ sweet orange, ‘Duncan’ grapefruit, ‘Sun Chu Sha’ mandarin, and ‘Hamlin’ sweet orange plants. Relative transcript levels were measured by real-time RT–PCR. Csi-actin was used as an internal control. Standard deviations were calculated from three technical replicates. Similar results were obtained from three biological replicates.

Figure 3.

Infection of Spiroplasma citri or Pseudomonas syringae Had No Obvious Effect on the Expression of csi-miR399 in Citrus and Arabidopsis, Respectively. (A) Total RNA was extracted from healthy and S. citri-infected Citrus sinensis, and subjected to Northern blot analysis as in Figure 2. Stubborn, S. citri-infected. (B) In Arabidopsis, miR399 expression was not affected by Pst (avrRpt2) infection. R2, avrRpt2 treated; hpi, hours post inoculation.

Figure 4.

Las-Positive Plants Had Reduced P Content Compared to Negative Control Plants and miR399-Mediated Regulation of P Homeostasis Is Conserved in Citrus. (A) Foliar levels of P (%), K (%), and Zn (ppm) were measured by ICP–OES spectrometry (IRIS 1000 HR Duo, ThermoElemental). Carbon (C; %) was analyzed using a NC 2100 analyzer (CE Elantech, Inc.). 15–20 leaves were collected from each of five sweet orange trees for analysis. Leaves from five non-infected trees were collected for comparison. Levels of each element in non-infected samples were assigned a value of 1. Experiments were repeated twice and similar results were obtained. (B) Expression of csi-PHO2 was down-regulated and PTs were up-regulated in Las-positive plants as compared with the Las-negative control. Expression levels of csi-PHO2 (gi|38053156), csi-PT2 (gi|56530333), and csi-PHT2;1 (gi|219250130) relative to citrus actin (gi|45420693) were determined by quantitative real-time PCR (mean ± SE). Expression level of untreated samples is assigned to 1. Experiments were repeated three times and similar results were obtained.

Figure 5.

Applying P Solution Alleviated Disease Symptom of HLB and Increased the Fruit Yield of the Diseased Trees. Phosphorus oxyanion solution was applied to Las-infected Las-positive sweet orange trees three times each year for 3 years. Application was synchronized with the initiation of new vegetative flushes in spring (March), summer (June), and fall (September). The treatment was a 3–18–20 liquid fertilizer of P oxyanion solution with 56% mono- and dipotassium salts of P acid (K-Phite, Plant Food Systems, Inc.), plus 3.8 kg of spray-grade potassium nitrate (KNO₃) and 18.9 L of 435 citrus spray oil. Control trees received KNO₃ and citrus spray oil only. Photos of trees with and without phosphorus oxyanion solution treatment are shown in (A), and photos of random-picked leaves from control and phosphorus oxyanion solution-treated trees are shown in (B). Treatments and the year the photos were taken are indicated on the right and bottom, respectively. Pictures were taken around the same time of the year and at the same location with similar camera settings.