Plant A20/AN1 protein serves as the important hub to mediate antiviral immunity

Abstract

Salicylic acid (SA) is a key phytohormone that mediates a broad spectrum of resistance against a diverse range of viruses; however, the downstream pathway of SA governed antiviral immune response remains largely to be explored. Here, we identified an orchid protein containing A20 and AN1 zinc finger domains, designated Pha13. Pha13 is up-regulated upon virus infection, and the transgenic monocot orchid and dicot Arabidopsis overexpressing orchid Pha13 conferred greater resistance to different viruses. In addition, our data showed that Arabidopsis homolog of Pha13, AtSAP5, is also involved in virus resistance. Pha13 and AtSAP5 are early induced by exogenous SA treatment, and participate in the expression of SA-mediated immune responsive genes, including the master regulator gene of plant immunity, NPR1, as well as NPR1-independent virus defense genes. SA also induced the proteasome degradation of Pha13. Functional domain analysis revealed that AN1 domain of Pha13 is involved in expression of orchid NPR1 through its AN1 domain, whereas dual A20/AN1 domains orchestrated the overall virus resistance. Subcellular localization analysis suggested that Pha13 can be found localized in the nucleus. Self-ubiquitination assay revealed that Pha13 confer E3 ligase activity, and the main E3 ligase activity was mapped to the A20 domain. Identification of Pha13 interacting proteins and substrate by yeast two-hybrid screening revealed mainly ubiquitin proteins. Further detailed biochemical analysis revealed that A20 domain of Pha13 binds to various polyubiquitin chains, suggesting that Pha13 may interact with multiple ubiquitinated proteins. Our findings revealed that Pha13 serves as an important regulatory hub in plant antiviral immunity, and uncover a delicate mode of immune regulation through the coordination of A20 and/or AN1 domains, as well as through the modulation of E3 ligase and ubiquitin chain binding activity of Pha13.

Fig 1. Identification of Pha13 involvement in PhaPR1 expression by CymMV-based VIGS system and sequence analysis.

(A) Expression level of Pha13, PhaPR1, and CymMV were analyzed by RT-PCR from leaves of P. aphrodite inoculated with buffer (Mock), or infiltrated with agrobacterium carrying pCymMV (Cy) or pCymMV-Pha13 (Pha13-VIGS). PhaUbiquitin 10 (PhaUBQ10) was used as a loading control, and the relative expression level of corresponding genes are indicated. (B) Domains of Pha13. Open rectangle indicates the entire protein. A20 (black rectangle) and AN1 (diagonally-striped rectangle), nuclear localization signal (open square with vertical lines), and 21-nucleotide position (short black line) used for designing hairpin RNA of Pha13 (hpPha13) are indicated. (C) Sequence alignment of A20/AN1 zinc finger domains of Pha13 with stress-associated proteins from Arabidopsis thaliana (AtSAP5, accession number: AT3G12630), and Oryza sativa (OsSAP3, accession number: LOC_Os01g56040.1; OsSAP5, accession number: LOC_Os02g32840.1) are shown. Conserved amino acid sequences are indicated with a black box. Conserved cysteine (C) and histidine (H) are indicated with a black triangle. (D) Primary sequence organization of A20 and AN1 domains. Xn: the number of amino acid residues between zinc ligands. (B to D) The mutation positions for domain functional analysis are indicated with red triangle.

Fig 2. Pha13 is involved in the expression of PhaPR1 and PhaNPR1, and induced by phytohormone treatment.

(A and B) Expression level of Pha13, PhaNPR1, and PhaPR1 were analyzed by qRT-PCR from leaves of P. aphrodite infiltrated with agrobacterium carrying vector (Vector); hairpin RNA (hpRNA) vector to knock down Pha13 (hpPha13; A); and overexpression vector of Pha13 (Pha13-oe; B). The RNA level of vector was set to 1. (C) Transient silencing of PhaNPR1. Expression level of PhaNPR1 and Pha13 were analyzed by qRT-PCR from leaves of P. aphrodite infiltrated with agrobacterium carrying the vector (Vector) or hairpin RNA (hpRNA) vector to knock down PhaNPR1 (hpPhaNPR1). The RNA level of the vector was set to 1. (A to C) Mean ± SD; n = 3 biological replicates; *, P < 0.05, Student’s t-test compared to vector. (D to F) Time-course expression of Pha13 under different plant hormone treatments in P. aphrodite. Expression level of Pha13 was analyzed by qRT-PCR from leaves treated with SA (D), JA (E), and ET (F) at different hours (h) post-treatment. Inoculation buffer treatment was used as a mock control. Results of qRT-PCR were relative to that of mock at individual time course for relative quantification. The RNA level at 0 hour was set to 1 for comparison between different time courses. PhaPR1 and PhaNPR1 were used as SA marker genes. PhaJAZ1 and PhaACO2 were used as JA and ET marker genes, respectively. Data represent mean ± SD; n = 3 technical replicates; *, P < 0.05, Student’s t-test compared to 0 h. One representative experiment is shown from at least three replicates of similar results. PhaUbiquitin 10 was used as an internal control for normalization.

Fig 3. Pha13 is induced by CymMV and involved in virus accumulation.

(A) The CymMV accumulation level and expression level of Pha13 were analyzed by qRT-PCR in healthy (Mock) and CymMV-infected (Cy-infected) P. aphrodite. The RNA level in the CymMV was set to 1. Data represent mean ± SD; n = 3 biological replicates; *, P < 0.05, Student’s t-test compared to mock. (B and C) Transient silencing or overexpression of Pha13 in CymMV-infected plants. RNA level of Pha13 and CymMV were analyzed by qRT-PCR from leaves of CymMV-infected P. aphrodite infiltrated with agrobacterium carrying vector (Vector); hairpin RNA (hpRNA) vector to knockdown Pha13 (hpPha13; B); or overexpression vector of Pha13 (Pha13-oe; C). The RNA level of vector was set to 1. Data represent mean ± SD; n = 3 biological replicates; *, P < 0.05, Student’s t-test compared to vector. D, The protein level of Pha13 in WT and transgenic P. equestris (Pha13#27, #29 and #30) expressing Pha13, Pha13-oe (35S::Flag-Pha13), were analyzed by anti-Pha13 antibody. The anti-Tubulin antibody (Anti-Tub) was used as a loading control. (E) Three wild-type plants and three plants derived from the same transformed protocorm were used for analysis. RNA level of Pha13 and CymMV were analyzed by qRT-PCR from leaves of WT or transgenic P. equestris (Pha13#27, #29 and #30) inoculated with CymMV. The RNA level of WT was set to 1. Data represent mean ± SD; n = 3 biological replicates; *, P < 0.05, Student’s t-test compared to WT. PhaUbiquitin 10 was used as an internal control for normalization.

Fig 4. Pha13 is involved in the expression of PhaNPR1-dependent and -independent marker genes.

(A) Expression level of PhaRdR1 and PhaGRX were analyzed by qRT-PCR from leaves of P. aphrodite treated with buffer (Mock) or SA, at different times (h) post-treatment. Results of qRT-PCR were relative to that of mock at individual time course for relative quantification. The RNA level at 0 h was set to 1. Data represent mean ± SD; n = 3 technical replicates; *, P < 0.05, Student’s t-test compared to 0 h. One representative experiment is shown from at least three replicates of similar results. PhaUbiquitin 10 was used as an internal control for normalization. (B) Expression level of PhaRdR1, PhaGRX, and CymMV accumulation level were analyzed by semi-quantitative RT-PCR from leaves of P. aphrodite inoculated with buffer (Mock) or infected with CymMV (Cy). PhaUbiquitin 10 was used as a loading control, and relative expression levels of the corresponding genes are indicated. (C) Expression level of PhaNPR1, PhaPR1, PhaRdR1, and PhaGRX were analyzed by qRT-PCR from leaves of P. aphrodite infiltrated with agrobacterium carrying the vector (Vector) or hairpin RNA (hpRNA) vector to knock down PhaNPR1 (hpPhaNPR1). (D and E) Expression level of PhaRdR1, PhaGRX, and CymMV accumulation level were analyzed by qRT-PCR from leaves of CymMV-infected P. aphrodite infiltrated with agrobacterium carrying vector (Vector); hairpin RNA (hpRNA) vector to knockdown PhaRdR1 (hpPhaRdR1, D), or PhaGRX (hpPhaGRX, E). (F and G) Expression level of Pha13, PhaRdR1, and PhaGRX were analyzed by qRT-PCR from leaves of P. aphrodite infiltrated with agrobacterium carrying vector (Vector) or hairpin RNA (hpRNA) vector to knockdown Pha13 (hpPha13; F), or to overexpress Pha13 (Pha13-oe; G). C to G, The RNA level of vector was set to 1. Data represent mean ± SD; n = 3 biological replicates; *, P < 0.05, Student’s t-test compared to vector. PhaUbiquitin 10 was used as an internal control for normalization.

Fig 5. Pha13 A20 and AN1 domains are involved in the expression of PhaNPR1-dependent and -independent immune gene(s) and virus accumulation.

(A and B) Expression level of Pha13, PhaNPR1, PhaRdR1, PhaGRX, and CymMV accumulation level were analyzed by qRT-PCR of healthy P. aphrodite leaves (A), or CymMV pre-infected P. aphrodite (B) and infiltrated with agrobacterium carrying vector (Vector), overexpression clones of Pha13, or the respective A20 and/or AN1 mutant clones. The RNA level of vector was set to 1. Data represent mean ± SD; n = 3 biological replicates; *, P < 0.05, Student’s t-test compared to vector. PhaUbiquitin 10 was used as an internal control for normalization. (C) A model illustrating SA-induced Pha13 transcriptional and post-translational regulation leading to the activation of immune responsive genes. Virus infection caused accumulation of SA and leads to post-translational modification of NPR1, allowing it to enter into the nucleus for the activation of NPR1-dependent immune responsive genes including PR1 and RdR1. On the other hand, increased SA can also regulate Pha13 at both transcriptional and post-translational level and leads to the expression of NPR1-dependent and independent immune responsive genes including NPR1, RdR1 and GRX for virus resistance.

Fig 6. Pha13 confers self-ubiquitination E3 ligase and ubiquitin binding ability.

In vitro ubiquitination analysis was performed on recombinant proteins of Pha13, A20 mutant (Pha13A20m), AN1 mutant (Pha13AN1m), A20/AN1 double mutant (Pha13A20mAN1m) with or without FLAG-tagged ubiquitin (FLAG-Ub), ubiquitin-activating enzyme (E1), or ubiquitin-conjugating enzyme (E2). Ubiquitinated proteins were analyzed with immunoblotting using anti-FLAG antibodies. Coomassie staining was used as a loading control. The ubiquitinated Pha13 (Pha13-2Ub), E1 conjugated with one Ub (E1-Ub), E2 conjugated with one or two Ub (E2-Ub, E2-2Ub), and free Ub are indicated with arrow.

Fig 7. Mapping the ubiquitin-binding region in Pha13 by a yeast two-hybrid assay.

(A) Left panel indicating Pha13 putative functional domains (A20 and AN1) and the sites of truncation. The ubiquitin-binding region is mapped to the A20 domain of Pha13. Various combinations of Pha13 truncations cloned in pGBK vector were co-transformed with pGADPhaUb into yeast AH109 strain. The transformants were spotted on control plates (2DO: Leucine-Trptophan dropped out medium) and selective plates (3DO/X-Gal: Leucine-Trptophan-Histidine dropped out and X-Gal added medium), which were incubated at 30°C for 4 days before photography. Red stars indicate cysteine in A20 domain. Yellow stars indicate cysteine replaced with lysine. (B) In vitro pull-down assay on AtSAP5 (used as positive control) [44], Pha13 and Pha13 A20 mutant (Pha13A20m), with polyubiquitin chains.

Fig 8. SA regulates Pha13 at post-translational level.

(A) Total proteins were extracted from leaves of P. aphrodite without treatment (0 h), treated with H₂O (Mock), or SA at different times (h), and used for in vivo immunoprecipitation (IP) assay using anti-Pha13 antibody. Samples after IP were analyzed by immunoblotting using anti-Pha13 or anti-ubiquitin antibody. (B) P. aphrodite was treated with H₂O (Mock) or SA, and immediately followed by infiltration of DMSO (-) or MG132 (+). Total proteins were extracted from leaves of the treated samples at 3 h post-treatment, and were used for in vivo IP assay using anti-Pha13 antibodies. Samples after IP were analyzed by immunoblotting using anti-Pha13 antibody. (A and B) Extracted total proteins (input) stained by coomassie brilliant blue (CBB) served as a loading control.

Fig 9. Overexpression of Pha13 in engineered transgenic Arabidopsis enhances resistance against viruses and bacteria.

(A to C) Expression level of Pha13 (A) and accumulation level of Tobacco rattle virus (TRV) (B) and Cucumber mosaic virus (CMV) (C) were analyzed by qRT-PCR from leaves of WT (Col-0) or transgenic Arabidopsis. Data represent mean ± SD; n = 4 biological replicates; *, P < 0.05, Student’s t-test compared to vector. Actin was used as an internal control for normalization. (D) Disease symptoms of WT (Col-0) or transgenic Arabidopsis inoculated with CMV at 9 dpi. Images are one representative plant from four replicates; Scale bar, 1 cm. (E) Disease symptoms of WT (Col-0) or transgenic Arabidopsis inoculated with 1 × 10⁷ cfu/ml Pseudomonas syringae pv. tomato DC30000 (PstDC3000) at 3 dpi. Images are three representative leaves from five plants; Scale bar, 1 cm. (F) The growth analysis of PstDC3000 in the infected leaves of WT or transgenic Arabidopsis. Data represent mean ± SD; n = 5 biological replicates. Different letters indicate statistically significant differences analyzed by one-way analysis of variance (ANOVA) Tukey’s test (P < 0.05).

Fig 10. The Arabidopsis homologue of Pha13, AtSAP5, is induced by SA and involved in virus resistance.

(A) The expression of AtSAP5 were analyzed by qRT-PCR in the WT, AtSAP5 overexpression lines (AtSAP5-oe-4 and AtSAP5-oe-11), and RNAi lines (AtSAP5-RNAi-3 and AtSAP5-RNAi-7). Data represent mean ± SD; n = 3 biological replicates *, p<0.05, Student’s t-test compared to WT (B) The accumulation levels of Cucumber mosaic virus (CMV) were analyzed by qRT-PCR from leaves of WT inoculated with buffer (Mock), CMV-inoculated WT, or CMV-inoculated transgenic Arabidopsis. Data represent mean ± SD; n = 5 biological replicates *, p<0.05, Student’s t-test compared to WT (C) Disease symptoms of CMV inoculated WT or transgenic Arabidopsis. The experiments were repeated twice with similar results, and 5 plants were used in each inoculation. Photo of one representative plant from each inoculation taken at 9 days post inoculation are shown here; Scale bar, 1 cm. (D) Time-course expression of AtSAP5 in SA-treated wild-type (WT, Col-0) Arabidopsis. Expression level of AtSAP5 was analyzed by qRT-PCR from leaves at 1 and 6 h post-treatment. Water treatment was used as a mock control. Data represent mean ± SD; n = 3 biological replicates *, p<0.05, Student’s t-test compared to mock. Actin was used as an internal control for normalization.

Fig 11. AtSAP5 is involved in the expression of NPR1 and NPR1-independent genes.

(A-D) The expression level of NPR1 (A), PR1 (B), RdR1 (C), and GRXC9 (D) were analyzed by qRT-PCR in H₂O (Mock) or SA-treated WT, AtSAP5 overexpression (AtSAP5-oe-4 and AtSAP5-oe-11) and RNAi lines (AtSAP5-RNAi-3 and AtSAP5-RNAi-7). Data represent mean ± SD; n = 3 biological replicates; *, p<0.05, Student’s t-test. The dashed lines indicate the sample with significant difference compared to WT. The colored lines indicate the significant difference was analyzed between corresponding samples treated with SA or mock. Actin was used as an internal control for normalization.