The Nuclear Localization of the DnaJ-Like Zinc Finger Domain-Containing Protein EDA3 Affects Seed Development in Arabidopsis thaliana

Abstract

The DnaJ-like zinc finger domain-containing proteins are involved in different aspects of plastid function and development. Some of these proteins were recently reported to have dual subcellular localization in the nucleus and plastids. One member of this family, PSA2 (AT2G34860), was found to localize to the thylakoid lumen and regulate the assembly of photosystem I (PSI). However, PSA2 was also annotated as Embryo sac Development Arrest 3 (EDA3) from the observation that its embryo sac development was arrested at the two-nuclear stage. In this study, we characterized the eda3 mutant, and demonstrated that, as compared with the wild-type (WT) plants, the mutant has shorter siliques, fewer siliques per plant, and fewer seeds per silique. Both aborted and undeveloped ovules were observed in siliques of the mutant. By immunoblot analysis, we found that, different from the chloroplast localization in mature leaves, EDA3 localizes in the nucleus in seeds. A nuclear localization signal was identified from the deduced amino acid sequence of EDA3, and also proved to be sufficient for directing its fusion peptide into the nucleus.

Keywords: DnaJ-like; EDA3; PSA2; chloroplast; female gametophyte; nucleus; zinc finger domain.

Figure 1

Expression pattern of EDA3 in different tissues. Transcript abundances of EDA3 in different tissues of Arabidopsis thaliana wild-type plants, together with the whole seedlings, were determined by qRT-PCR and normalized against the levels of ACTIN2 (ACT2) (as a reference). Data represent means ± SEM (n = 3).

Figure 2

Phenotype of the Arabidopsis thaliana eda3 mutant. (A) Heights of the seedlings; (B) Numbers of siliques per seedling; (C) Length of siliques; (D) Numbers of seeds per silique, were measured in the wild-type (WT) and eda3 mutant seedlings. Seedlings for each line were cultivated under normal growth conditions. For (A,B), 6-week-old seedlings were measured. For (C,D), mature (brown) siliques were used for measurements. Data represent means ± SEM (n = 3). Asterisks indicate significant differences between the indicated lines (** p < 0.01, *** p < 0.001, Student’s t-test). (E) SEM observation of mature siliques of WT and eda3. Bar = 100 μm; (F) Microscopic observation of the embryo sacs of the WT and eda3 plants. Vacuole (Vac), synergid (Syn), egg cell (Ec), polar nuclei (Pn), antipodal cells (Ap) are indicated. Bar = 20 μm.

Figure 3

Immunoblot showing the dual localization of EDA3. (A) EDA3 localizes in leaves and seeds in the 12 and 20 kDa forms, corresponding to the truncated and full-length peptides, respectively. Actin was probed as a loading control; (B) EDA3 is a thylakoid protein in leaf cells. Rubisco large subunit (RbcL) was probed as a loading control; (C) EDA3 localizes in the nucleus in seeds. Histone H3 was probed as a loading control. L, leaves; S, seeds; Thy, thylakoid preparation; S-N, seed nuclear preparation.

Figure 4

The nuclear localization signal (NLS) of EDA3 accounts for its nuclear targeting. (A) Schemes of the fusion proteins used in this assay. NLS, nuclear localization signal; cTP, chloroplast transit peptide; EDA3ΔcTP, truncated EDA3 with its cTP removed; EDA3^NLS-PIP2A, PIP2A protein fused beyond the NLS of EDA3. The CaMV 35S promoter was used to drive the transient expression of each protein in tobacco leaves; (B) Transient expression EDA3 and EDA3ΔcTP fused upstream EYFP, and PIP2 and EDA3^NLS-PIP2A fused upstream mCherry, in tobacco leaves by infiltration; (C) Co-localization of the EDA3ΔcTP-EYFP fusion protein with the nuclear localization marker protein VirD2NLS-mCherry. The excitation wavelengths for EYFP, mCherry, and chlorophyll auto-fluorescence were 488, 543, and 543 nm laser, respectively, and their corresponding recoding ranges were 500–530, 580–620, and 680–720 nm, respectively. Empty vectors expressing EYFP and mCherry were used as controls. FP, fluorescent protein. Bar = 10 μm.