The cytoskeleton and the peroxisomal-targeted snowy cotyledon3 protein are required for chloroplast development in Arabidopsis

Abstract

Here, we describe the snowy cotyledon3 (sco3-1) mutation, which impairs chloroplast and etioplast development in Arabidopsis thaliana seedlings. SCO3 is a member of a largely uncharacterized protein family unique to the plant kingdom. The sco3-1 mutation alters chloroplast morphology and development, reduces chlorophyll accumulation, impairs thylakoid formation and photosynthesis in seedlings, and results in photoinhibition under extreme CO(2) concentrations in mature leaves. There are no readily apparent changes to chloroplast biology, such as transcription or assembly that explain the disruption to chloroplast biogenesis. Indeed, SCO3 is actually targeted to another organelle, specifically to the periphery of peroxisomes. However, impaired chloroplast development cannot be attributed to perturbed peroxisomal metabolic processes involving germination, fatty acid β-oxidation or photorespiration, though there are so far undescribed changes in low and high CO(2) sensitivity in seedlings and young true leaves. Many of the chloroplasts are bilobed, and some have persistent membranous extensions that encircle other cellular components. Significantly, there are changes to the cytoskeleton in sco3-1, and microtubule inhibitors have similar effects on chloroplast biogenesis as sco3-1 does. The localization of SCO3 to the periphery of the peroxisomes was shown to be dependent on a functional microtubule cytoskeleton. Therefore, the microtubule and peroxisome-associated SCO3 protein is required for chloroplast development, and sco3-1, along with microtubule inhibitors, demonstrates an unexpected role for the cytoskeleton and peroxisomes in chloroplast biogenesis.

Figure 1.

Phenotype of sco3-1. (A) Phenotype of 14-d-old sco3-1 seedlings compared with Col. (B) and (C) TEMs of etioplasts (B) and chloroplasts (C) in cotyledons of Col and sco3-1. (D) Cross sections of chloroplast membranous extensions in sco3-1 (indicated by an arrow). Note the mitochondrial (m) and peroxisomal (p) structures within these plastids and the engulfment of a mitochondrion in the top image. S, stromule; cw, cell wall. Bars = 1 μm.

Figure 2.

Transcript Analyses of Plastid-Encoded and Nuclear-Encoded Genes. (A) RT-PCR analysis of transcripts of plastid genes dependent on nuclear-encoded (NEP) or plastid-encoded (PEP) polymerase as well as of nuclear-encoded genes regulated by retrograde signaling in 4-d-old seedlings (4d) and from cotyledons (cot) and true leaves (TL) of 14-d-old seedlings of sco3-1 and Col. “0” is the negative control with no cDNA added to the reaction. (B) Quantitative RT-PCR analysis of the plastid-encoded genes of 4-d-old seedlings of sco3-1 compared with Col. Analysis was performed with three independent replicates. (C) RT-PCR analysis of nuclear-encoded genes regulated by retrograde signaling in 4-d-old seedlings of sco3, sco3 phyB, and phyB compared with Col. (D) Comparison of transcript levels of LHCB1.5 and LHCB2.3 compared with Col (WT). *P < 0.05; **P < 0.005; compared with wild type/compared with sco3-1. Error bars indicate the sd. Analysis was performed in triplicates. (E) Predicted subcellular localization of down- and upregulated genes in sco3-1 compared with Col from the microarray analysis. Numbers in brackets indicate the number of photosynthesis-related genes.

Figure 3.

Characterization of the SCO3 Gene. (A) Structure of the SCO3 gene. The mutation in sco3-1 is marked with a star. The positions of the T-DNA insertion lines from SALK are indicated. (B) Embryo-lethal phenotype of the sco3 T-DNA insertion mutant (shown for SALK_120239). Embryo development stopped at a very early developmental stage (see arrow for an early aborted embryo). (C) Complementation analysis of the sco3 mutant with the SCO3 cDNA using the GFP:SCO3 construct. (D) RT-PCR analysis of SCO3 transcripts in the sco3-1 mutant and in Col in 4-d-old seedlings grown on MS media without (−NF) or with norflurazon (+NF) using 18S rRNA as loading control. (E) Domain structure of the SCO3 protein. The star marks the position of the mutation in sco3-1. pred. PTS, predicted peroxisome targeting sequence. (F) Transcript abundance of SCO3 and SCO3-sv2 in different plant organs and developmental stages. Note that 25 cycles were used for the RT-PCR for the 18SrRNA control and 50 cycles for SCO3 in (D) and (F). y, young; sen., senecescent; 0, negative control without DNA added into the reaction; gDNA, genomic DNA control for specificity for cDNA.

Figure 4.

Localization of the SCO3 Protein. (A) and (B) Localization studies were performed using N-terminal GFP fusions with the complete SCO3 (A) or mutated SCO3 (Msco3-1). RFP:SRL is used to visualize peroxisomes using peroxisome-targeted RFP (B). Insets in (A) and (B) magnify the GFP signal on peroxisomes. (C) Localization of GFP:SCO3 to the periphery of peroxisomes in a heterologous system, onion (see Supplemental Movie 1 online). (D) Complementation analyses, as determined by chlorophyll content in cotyledons using different wild-type (WT) SCO3 cDNA constructs with an N-terminal GFP (green box) and with or without the PTS (orange box) and N-terminal GFP fused to SCO3 cDNA containing the sco3-1 mutation (star).

Figure 5.

The SCO3 Protein Is Not Localized to the Chloroplast. (A) In vitro import analysis of radioactively labeled SCO3 and mutated SCO3 (SCO3mut) into isolated Arabidopsis chloroplasts. pSSU, the precursor of the small subunit of Rubisco was used as a positive control and the mitochondrial-targeted pAOX (alternative oxidase) as a negative control for chloroplast import of proteins. Copper chloride is inhibiting chloroplast protein import, whereas thermolysin degrades all not-imported proteins in the solution. (B) Localization studies were performed using the first 100 amino acids of SCO3 or mutated SCO3 (m100AA, with the sco3-1 mutation) fused to C-terminal GFP. SSU:RFP was coexpressed to visualize chloroplasts. (C) Complementation analyses showing the complementing GFP:SCO3 and using additional SCO3 constructs with a C-terminal GFP (green box, in which the PTS is not accessible) to unmask the putative chloroplast transit peptide and with the fusion of the chloroplast targeting sequence (Chl cTP) of rcbS (SSUcTP; dark-green box), respectively.

Figure 6.

Analysis of Peroxisomal Function. (A) Analysis of β-oxidation processes using IBA, which is β-oxidized in peroxisomes to auxin. Hypocotyl length of etiolated seedlings at different concentrations of IBA was measured and is expressed as the percentage of the hypocotyl length of seedlings of the same genotype grown without IBA application. The error bars define the sd of three independent replicates. WT, wild type. (B) Effects on photorespiration and photoinhibition. Seedlings of Col (black box) and sco3-1 (white box) were grown for 7 d under high CO₂ (hCO₂, 0.6%) or at ambient CO₂ concentration (air) and, after measuring the activity of PSII (F_v/F_m), were transferred for 24 h to altered CO₂ concentrations (0 CO₂, air, or hCO₂). F_v/F_m was measured again. (C) Phenotype of plants grown for 5 weeks either at ambient CO₂ (air) or under high CO₂ (hCO₂).

Figure 7.

Chloroplasts and the Cytoskeleton. (A) Analysis of chloroplast development in cotyledons under the confocal microscope. Chloroplasts are visible due to the autofluorescence of chlorophyll. In sco3-1, many plastids resemble bilobed chloroplasts (indicated by an arrow), which is magnified in the bottom right insets. In Col, no such structure could be observed. Top right insets show cpTP:GFP to visualize plastids without chlorophyll autofluorescence (chlorophyll autofluoroescence is visualized by the red color). Bars = 50 μm. (B) Effect of cytoskeleton inhibitors APM (microtubule inhibitor) and CD (actin inhibitor) on chloroplast biogenesis in seedlings grown for 3 d on MS media before being transferred to inhibitor-containing media for an additional 2 d. Chloroplasts were visualized under the same magnification using chlorophyll autofluorescence with confocal laser scanning microscopy. The inset in APM-treated Col seedlings shows an example of bilobed chloroplasts in Col. Arrows indicate the bilobed plastids. (C) Visualization of the actin cytoskeleton using Talin:GFP in transient transformation of seedlings of sco3-1 and Col. Two examples for each line are shown. Arrows indicate the fine structure missing in sco3-1 but present in Col, respectively. cp, chloroplast; n, nucleus.

Figure 8.

Model for the sco3-1 Phenotype. Despite the pale cotyledon phenotype in sco3-1, the SCO3 protein (shown as a GFP fusion) is not localized to the chloroplast (cf. the GFP signal to the red signal of chloroplast-localized SSU:RFP in the left image) but to the periphery of peroxisomes (cf. GFP to the peroxisomal RFP:SRL in the middle image). This peripheral localization on peroxisomes of SCO3 is disrupted when the microtubule cytoskeleton assembly is inhibited by APM (+APM, right image). Application of this inhibitor also results in reduced chlorophyll content as well as impaired chloroplast biogenesis and bilobed chloroplasts in the wild type that resemble those in untreated sco3-1 (refer to Figure 7B). This demonstrates a requirement for the cytoskeleton and SCO3 in chloroplast and peroxisome biogenesis and morphology.