S-nitrosylation triggers ABI5 degradation to promote seed germination and seedling growth

Abstract

Plant survival depends on seed germination and progression through post-germinative developmental checkpoints. These processes are controlled by the stress phytohormone abscisic acid (ABA). ABA regulates the basic leucine zipper transcriptional factor ABI5, a central hub of growth repression, while the reactive nitrogen molecule nitric oxide (NO) counteracts ABA during seed germination. However, the molecular mechanisms by which seeds sense more favourable conditions and start germinating have remained elusive. Here we show that ABI5 promotes growth via NO, and that ABI5 accumulation is altered in genetic backgrounds with impaired NO homeostasis. S-nitrosylation of ABI5 at cysteine-153 facilitates its degradation through CULLIN4-based and KEEP ON GOING E3 ligases, and promotes seed germination. Conversely, mutation of ABI5 at cysteine-153 deregulates protein stability and inhibition of seed germination by NO depletion. These findings suggest an inverse molecular link between NO and ABA hormone signalling through distinct posttranslational modifications of ABI5 during early seedling development.

Figure 1. NO depletion phenotypes of ABI5 loss-of-function mutants during seed germination.

(a) Insensitivity of abi5 mutants to NO scavenging by cPTIO during seed germination. Photographs of 2-day-old germinated seeds after imbibition of wild type (Col-0) and the ABA-insensitive abi5-1 and abi5-7 mutants, in the absence of (Control) or the presence of 100 μM cPTIO. Scale bar, 1 mm. (b) Germination of wild-type (Col-0), abi5-1 and abi5-7 seeds in media containing 0 and 100 μM cPTIO after 2 days. Error bars represent±s.e. (n=3). Asterisk indicates significant differences compared with Col-0 (Control) (t-test, P<0.05). (c) Co-localization of ABI5 expression, protein localization and NO production. pABI5:ABI5-GUS seeds were stratified for 3 days at 4 °C and grown for 1 to 2 days at 21 °C on MS agar plates and then subjected to DAF-2DA incubation or GUS staining after treatment with NO scavenger (cPTIO) and donor (SNAP). Arrows indicate high NO accumulation (left), and ABI5 expression and protein localization (middle). Scale bars, 100 μm. (d) qRT–PCR analysis of ABI5 relative transcript abundance in Col-0 seeds untreated (Control) and after treatments with ABA, cPTIO, SNAP and GSNO after 24, 48 and 72 h and in the abi5-1 background. Error bars represent±s.e. (n=3).

Figure 2. Increased NO levels reduce ABI5 protein accumulation in a proteasome pathway-dependent manner.

(a) qRT–PCR analysis of ABI5 relative transcript abundance in Col-0 seeds untreated (C) and after treatments with GSNO, SNAP, cPTIO, cycloheximide (CHX), MG132 proteasome inhibitor and the combinations indicated after 3 h. Error bars represent±s.e. (n=3). (b) SNAP and GSNO treatments promote ABI5 degradation in dormant seeds (DS). Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 and abi5-1, treated with or without (C) NO scavenger (cPTIO) and donors (GSNO and SNAP), and the MG132 proteasome inhibitor. Actin protein levels are shown as a loading control. (c) Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 dormant seeds treated with or without (C) NO donor (GSNO), the proteasome inhibitor cocktail (PIC, composed of MG115, MG132 and epoxomicin), cycloheximide (CHX) and the combinations indicated. Actin protein levels are shown as a loading control. (d) GSNO treatment promotes ABI5 degradation in 2 days ABA (5 μM)-treated after-ripened seeds. Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0, treated with or without (C) NO scavenger (cPTIO) and donor (GSNO). Actin protein levels are shown as a loading control. (e) ABI5 protein levels in wild type (WT; Col-0), abi5-1 and NO-deficient (atnoa1-1, nia1;nia2, atnoa1-2;nia1;nia2) mutant backgrounds. Stratified seeds were sown on control MS (Control, left) and 0.1 μM ABA after 48 h (right). Immunoblot analysis of ABI5 protein levels in seed extracts of WT and NO-deficient mutants. Actin protein levels are shown as a loading control. (f) ABI5 protein levels in WT, 35S:AHb1 (H3, H7) and 35S:antiAHb1 (L1, L3) lines. Stratified seeds were sown on MS (Control) and 0.1 μM ABA for 96 h. Immunoblot analysis of ABI5 protein levels in seed extracts of AHb1-overexpressing and -silencing lines. Actin protein levels are shown as a loading control.

Figure 3. S-nitrosylation of ABI5 in vivo and in vitro.

(a) Mass spectrometric analyses identify C153 as the S-nitrosylation site. MS/MS spectra of C153 from the tryptic fragment QGSLTLPAPLCR (peptide MS/MS spectra shown with Cys modified by biotin-HPDP). (b) The LC–MS spectra of the corresponding peaks (*562 m/z (3+) and 842,49 m/z (+2)) of this peptide fragment is shown in the inset. (c) The C153S mutation blocks S-nitrosylation of ABI5. In vitro S-nitrosylation of wild-type ABI5 and mutant ABI5C153S recombinant proteins by the NO donors GSNO (200 μM) and SNAP (200 μM). This modification is reversed by treatment with DTT (20 mM). No signal was observed with glutathione (200 μM) treatment showing specificity of the biotin-switch assay. ABI5 protein loading was detected by anti-His antibody. (d,e) S-nitrosylation of ABI5 induced by GSNO in after-ripened seed extracts. Samples were initially immunopurified with anti-biotin before immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 (d), 35S:ABI5 (e) and abi5-1 untreated (C) or treated with the indicated compounds. No signal was observed in the absence of biotin (−Biotin) or after DTT (20 mM) treatment. Actin protein levels are shown as a loading control. (f) In-vivo S-nitrosylation of ABI5 in abi5-1;35S:cMyc-ABI5 and abi5-1;35S:cMyc-ABI5C153S after-ripened seed extracts 24 h after proteasome inhibitor MG132 (100 μM) incubation. Immunoblot analysis of in vivo ABI5 protein levels after immunopurification of S-nitrosylated proteins. No signal was observed in the absence of sodium ascorbate (−Asc) or after cPTIO (1 mM) treatment. Input protein levels were also determined using anti-ABI5 anti-serum.

Figure 4. The ABI5 C153S mutant shows decreased proteasomal degradation by CUL4 and KEG, and confers NO insensitivity during seed germination.

(a) Immunoblot analysis of ABI5 protein levels in 8-day-old seedling extracts of similar germination stages abi5-1;35S:cMyc-ABI5 and abi5-1;35S:cMyc-ABI5C153S, in the presence of cycloheximide (1 mM) and cycloheximide (1 mM) plus GSNO (500 μM) from 0 to 9 h. Actin protein levels are shown as a loading control. (b) ABI5 protein levels in wild type (Col-0), cul4cs, keg4 and abi5-1 mutant backgrounds. Stratified seeds were incubated with 5 μM ABA for 48 h and treated with GSNO (1 mM) for 6 h after ABA removal. Immunoblot analysis of ABI5 protein levels in seed extracts of wild type and mutants. Actin protein levels are shown as a loading control. (c) Co-immunoprecipitation assays between CUL4 and transgenic ABI5/ABI5C153S proteins in the presence of GSNO. Input protein levels were also determined using anti-FLAG and anti-MYC antisera, respectively. (d) Co-immunoprecipitation assays between KEG and ABI5/ABI5C153S proteins in the presence of GSNO. Input protein levels were also determined using anti-HA and anti-MYC antisera, respectively. (e) NO-insensitive inhibition of seed germination to NO scavenging in 35S:ABI5C153S lines as compared with 35S:ABI5 plants. Total seed germination of wild type (Col-0), abi5-1, abi5-7 and two (1, 2) 35S:ABI5- and 35S:ABI5C153S-independent lines grown for 2 days on MS agar plates untreated (Control) or supplemented with 50 and 100 μM of the NO-scavenger cPTIO. Values represent the mean ±s.e. (n=3). Asterisks indicate significant differences compared with 0 μM cPTIO (t-test, *P<0.05, **P<0.01). (f) ABI5 levels in 35S:ABI5 and 35S:ABI5C153S transgenic lines used for the germination assay. Immunoblot analysis of ABI5 protein levels in seed extracts. Actin protein levels are shown as a loading control. (g) NaCl- and mannitol-hypersensitive inhibition of post-germinative growth in two 35S:ABI5 and 35S:ABI5C153S lines as compared with wild-type plants. Seedling growth of wild type (Col-0), abi5-1, 35S:ABI5 and 35S:ABI5C153S lines grown for 9 days on MS agar plates untreated (Control) or supplemented with 100 mM of NaCl and 250 mM of mannitol. Values represent the mean±s.e. (n=3). Letters indicate significant differences compared with wild-type (Col-0) (a), 35S:ABI5-1 (b), 35S:ABI5-2 (c), abi5-1;35S:ABI5-1 (e), abi5-1;35S:ABI5-2 (d), (t-test, P<0.05).

Figure 5. ABI5 accumulation and degradation in prt6-1 mutant background during seed germination and post germination.

(a,b) ABA treatment promotes ABI5 accumulation in germinating seeds. Immunoblot analysis of ABI5 protein levels in seed extracts of Col-0 and prt6-1 before (a) and after 48 h treatment with (+) or without (−) 0.25 μM ABA (b). Actin protein levels are shown as a loading control. (c) Post-germinative ABI5 accumulation in seedlings. Immunoblot analysis of ABI5 protein levels in 10-day-old extracts of Col-0, prt6-1 and abi5-1, treated with or without (−) 0.25, 0.5 and 1 μM ABA. Actin protein levels are shown as a loading control and Rubisco large subunit (RbcL) detection is indicated. (d) ABI5 protein levels in wild-type (Col-0), prt6-1 and abi5-1 mutant backgrounds. Stratified seeds were incubated with 5 μM ABA for 48 h (T0) and treated after ABA removal with H₂O, ABA, SNAP (1 mM) and GSNO (1 mM) for 12 h. Immunoblot analysis of ABI5 protein levels in seed extracts of wild type and mutants. Actin protein levels are shown as a loading control.

Figure 6. Model showing a role for NO in the regulation of ABI5 during seed germination.

Transcriptional control of ABI5 expression via the group VII ERFs and posttranslational S-nitrosylation of ABI5 protein are included. Dormant and dry seeds accumulate high levels of the ABA-induced ABI5 growth reppressor. On seed imbibition, a burst of NO is early produced to degrade group VII ERFs via the N-end rule pathway of targeted proteolysis (PRT6) and induces ABI5 S-nitrosylation promoting the interaction with CUL4-based and KEG E3 ligases. Consequently, ABI5 is rapidly degraded by the proteasome during seed germination.