The non-JAZ TIFY protein TIFY8 from Arabidopsis thaliana is a transcriptional repressor

Abstract

Jasmonate (JA) signalling is mediated by the JASMONATE-ZIM DOMAIN (JAZ) repressor proteins, which are degraded upon JA perception to release downstream responses. The ZIM protein domain is characteristic of the larger TIFY protein family. It is currently unknown if the atypical member TIFY8 is involved in JA signalling. Here we show that the TIFY8 ZIM domain is functional and mediated interaction with PEAPOD proteins and NINJA. TIFY8 interacted with TOPLESS through NINJA and accordingly acted as a transcriptional repressor. TIFY8 expression was inversely correlated with JAZ expression during development and after infection with Pseudomonas syringae. Nevertheless, transgenic lines with altered TIFY8 expression did not show changes in JA sensitivity. Despite the functional ZIM domain, no interaction with JAZ proteins could be found. In contrast, TIFY8 was found in protein complexes involved in regulation of dephosphorylation, deubiquitination and O-linked N-acetylglucosamine modification suggesting an important role in nuclear signal transduction.

Figure 1. The TIFY protein family in Arabidopsis.

Phylogenetic tree of the Arabidopsis TIFY family members based on the ZIM domain (Z) protein sequence. AT4G27110 and AT3G20580 were chosen as the outgroup. AT4G27110 contains a TIFY motif but is not conserved in the domain outside this motif. Consequently, it is not considered to be a real TIFY protein. The second protein, AT3G20580, is its closest homologue within the parsed region. The numbers above the branches are bootstrap values from 100 replicates and assess the robustness of the tree. Additional protein domains are shown. C: CONSTANS, CO-like, and TOC1 (CCT) domain; G: C2C2-GATA Zn-finger; P: PEAPOD domain; J: Jas domain; J* Jas-like domain; E: EAR domain. Figure adapted from .

Figure 2. The ZIM domain of TIFY8 is functional.

A, Y2H analysis of TIFY8 interaction with class II TIFY proteins. NINJA was included as a positive control for JAZ interaction. B, C, Analysis of TIFY8 truncations to map the interaction domain with NINJA (B) and PPD2 (C). Co-transformation of the PJ69-4A yeast strain with TIFY8 or NINJA and all TIFY family members in Gateway-compatible pGADT7 and pGBKT7 vectors, respectively. Transformed yeasts were spotted on control medium lacking Leu and Trp (-2) or selective medium additionally lacking His (-3). AD: activation domain; BD: DNA-binding domain. Controls for autoactivation are provided by transformation with the corresponding empty vector. D, Immunoblot analysis of 7 day-old Arabidopsis seedlings overexpressing the TIFY8- or JAZ1-GS fusions after 1 h treatment with either 50 µM JA or ethanol (mock). Immunoblot using the Peroxidase Anti-Peroxidase (PAP) (top) and anti-cdc2 (bottom) antibodies.

Figure 3. TIFY8 is a nuclear transcriptional repressor.

A–C, TIFY8 localizes to the nucleus. Confocal root tip imaging of 4-day-old Arabidopsis seedlings overexpressing the TIFY8-GFP fusion protein (A), free GFP (B) or SV40-NLS fused GFP (C), respectively. Propidium iodide staining was performed prior to imaging to enhance the visualization of the cells. D, NINJA, but not TIFY8, interacts directly with TPL in Y2H assays. Co-transformation of the PJ69-4A yeast strain with TIFY8 or NINJA and the N-terminal fragment of TPL (TPL-N) in pGADT7 or pGBKT7 vectors, respectively. Transformed yeast were spotted on control medium lacking Leu and Trp (-2) or selective medium additionally lacking His (-3). E, TIFY8 recruits TPL through interaction with NINJA in Y3H assays. Co-transformation of the PJ69-4A yeast strain with TIFY8 and TPL-N in Gateway-compatible pGADT7 and pGBKT7 vectors, respectively, together with NLS-3xFLAG-6xHis tagged NINJA in the pMG426 vector. Transformed yeast were spotted on control medium lacking Leu, Trp and Ura (-3) or selective medium additionally lacking His (-4). A negative control is provided by substitution of NINJA by the empty pMG426 vector. F, TIFY8 acts as a transcriptional repressor in transient expression assays. Transactivation activity in tobacco protoplasts transfected with a pUAS–fLUC reporter construct, effector constructs fused to GAL4DBD, and a 35S:rLUC normalization construct. Error bars represent ±SE of eight biological replicates. Asterisks represent significant differences (***, p<0.001, one-way ANOVA, Tukey HSD's Post Hoc test).

Figure 4. TIFY8 is not induced by JA and is repressed by Pst DC3000 infection.

A–B, RT-PCR analysis of JAZ10 (A) TIFY8 (B) expression after JA treatment. Arabidopsis wild-type seeds were germinated on MS media and, after 8 days, transferred to liquid MS media supplied with different concentrations of JA or COR or equivalent amounts of DMSO (mock treatment). Transcript levels were studied 24 h after treatment. Error bars represent ±SE of four biological replicates. C–D, JAZ10 and TIFY8 expression after infection with Pst DC3000. Transcript levels were studied in 5-week-old Arabidopsis rosette leaves prior to or 24 h after inoculation with Pst DC3000. Error bars represent ±SE of three biological replicates. UBC (AT5G25760) was used as internal control and expression values were normalized to those of the wild-type after mock treatment. (NS, p>0.05; ***, p<0.001, t-test).

Figure 5. Overview or TIFY8 and JAZ1 promoter gene expression.

GUS stains of Arabidopsis plants expressing either TIFY8 or JAZ1 promoter fusions to GUS and GFP. A, B. TIFY8 promoter expression pattern in 5 and 7 day-old seedlings, respectively. C, D. TIFY8 (C) and JAZ1 (D) promoter expression in 5 day-old cotyledons. E, F. TIFY8 (E) and JAZ1 (F) promoter expression in the shoot apical meristem and emerging leaves of 5 day-old seedlings. G, H. TIFY8 (G) and JAZ1 (H) promoter expression in the root tip of 5 day-old Arabidopsis seedlings. I. TIFY8 promoter expression in lateral root of a 10 day-old seedling. J, K. TIFY8 (J) and JAZ1 (K) promoter expression in 14 day-old seedlings. L–O. TIFY8 and JAZ1 promoter expression in flowers (L, M) and siliques (N, O).

Figure 6. Characterization of transgenic lines with altered TIFY8 expression.

A, RT-PCR analysis of TIFY8 expression levels in two independent TIFY8-OE lines and compared to wild type. Transcript levels were studied in two-week-old Arabidopsis TIFY8-OE and wildtype seedlings. UBC (AT5G25760) was used as internal control and expression values were normalized to those of the wildtype. Error bars represent ±SE of three technical replicates. B–C, Analysis of JA-responsiveness of the TIFY8-OE lines compared to wild-type. Root growth inhibition was scored on 11 days after stratification (DAS) (B) while anthocyanins were extracted for the same samples used for root growth but harvested 14 DAS (C). Four technical repeats per line and treatment, consisting on up to eight seedlings per repeat (20≤n≤32), were analysed. Bars represent average ± SE. Differences between the transgenic lines assayed and wild-type in control conditions are shown (*: p<0.05; t-test). Statistically significant differences for the interaction between genotype and treatment were not found (NS, p>0.05, one-way ANOVA). D, Schematic diagram of the TIFY8 (At4g32570) locus. Black bars, black lines and grey bars represent exons, introns and the untranslated regions, respectively. The T-DNA in the tify8-1T line (GK_738B03) is inserted immediately after the start codon of TIFY8, and the T-DNA contains the 35S promoter sequence next to the right border (RB). Arrows and numbers indicate different primer combinations covering different regions of TIFY8. Primer sequences can be found in Table S2. E, RT-PCR analysis of TIFY8 transcripts in the tify8-1T and tify8-2T lines. Transcript levels were studied in 1-week-old seedlings. Numbers represent the primer combination used, described in (D). UBC (AT5G25760) was used as internal control and expression values (Y-axis) were normalized to those of the wildtype. Error bars represent ±SE of three biological replicates. F, Analysis of JA-responsive root growth inhibition of the tify8-1T and tify8-2T lines compared to wild type performed as in (B). G, JA-responsive anthocyanin accumulation in tify8-1T performed as in (C).