Plastids are major regulators of light signaling in Arabidopsis

Abstract

We previously provided evidence that plastid signaling regulates the downstream components of a light signaling network and that this signal integration coordinates chloroplast biogenesis with both the light environment and development by regulating gene expression. We tested these ideas by analyzing light- and plastid-regulated transcriptomes in Arabidopsis (Arabidopsis thaliana). We found that the enrichment of Gene Ontology terms in these transcriptomes is consistent with the integration of light and plastid signaling (1) down-regulating photosynthesis and inducing both repair and stress tolerance in dysfunctional chloroplasts and (2) helping coordinate processes such as growth, the circadian rhythm, and stress responses with the degree of chloroplast function. We then tested whether factors that contribute to this signal integration are also regulated by light and plastid signals by characterizing T-DNA insertion alleles of genes that are regulated by light and plastid signaling and that encode proteins that are annotated as contributing to signaling, transcription, or no known function. We found that a high proportion of these mutant alleles induce chloroplast biogenesis during deetiolation. We quantified the expression of four photosynthesis-related genes in seven of these enhanced deetiolation (end) mutants and found that photosynthesis-related gene expression is attenuated. This attenuation is particularly striking for Photosystem II subunit S expression. We conclude that the integration of light and plastid signaling regulates a number of END genes that help optimize chloroplast function and that at least some END genes affect photosynthesis-related gene expression.

Figure 1.

Lhcb1 and RbcS expression following a fluence-rate shift. A, Lhcb1 expression following a fluence-rate shift. Seedlings were grown for 6 d in 0.5 μmol m⁻² s⁻¹ BR light and then transferred to 60 μmol m⁻² s⁻¹ BR light. Seedlings were collected and RNA was extracted at 0, 0.5, 1, 4, and 24 h following the fluence-rate shift. The levels of Lhcb1 mRNA relative to Lhcb1 mRNA levels at 24 h were determined from four biological replicates and quantified from RNA blots as described by Ruckle et al. (2007). B, RbcS expression following a fluence-rate shift. RNA was extracted and quantified as described in A.

Figure 2.

The light-regulated transcriptomes of lincomycin-treated and untreated seedlings. A, Venn diagram of light- and plastid-regulated genes. Light-regulated genes are defined as those that are expressed 2-fold higher or lower (P ≤ 0.01) at 0.5, 1, 4, or 24 h after the fluence-rate shift than before the shift (0 h). Plastid-regulated genes are those that meet the same fold change and significance criteria used to classify a gene as light regulated when the expression level of a particular gene in lincomycin-treated (+Lin) seedlings is normalized to the expression level in untreated (−Lin) seedlings at the same time point. The numbers of significantly regulated genes are indicated. B, Numbers of genes regulated by light and lincomycin treatment after a BR fluence-rate shift. Numbers of genes that exhibited a significantly different expression level in +Lin or −Lin seedlings at 0.5, 1, 4, and 24 h after a BR fluence-rate shift are indicated. The 3,335 genes that are significantly regulated only by light are indicated with red and light red. The 3,089 genes that are regulated by both light and lincomycin treatment are indicated with purple and light purple. Plastid regulation is presented for the 3,089 genes that are regulated by light and plastid signals in blue and light blue. The plastid regulation for the 680 genes regulated only by the plastid is presented in green and light green. C, Principal component analysis of the lincomycin treatment affecting the light-regulated transcriptome. Trajectory plots show the first principal component (PC1) and the second principal component (PC2), which are two orthogonal factors that describe 61% and 18%, respectively, of the variance caused by the BR fluence-rate shift. D, Principal component analysis of the BR fluence-rate shift affecting the lincomycin-regulated transcriptome. These trajectory plots show PC1 and PC2, which account for 79% and 21%, respectively, of the variance in the data set. E, Agglomerative hierarchical clustering of the 7,104 significantly regulated genes based on their regulation by the BR fluence-rate shift. Nine basic expression patterns were identified (A–I). F, Agglomerative hierarchical clustering of the 7,104 significantly regulated genes based on their regulation by lincomycin treatment. Eight basic expression patterns were identified (J–Q).

Figure 3.

Summary of biological process and cellular component GO terms that are enriched in particular expression patterns. User-defined expression patterns were obtained as described in Supplemental Figure S3, and clusters of expression were obtained as described in Figure 2 and Supplemental Figure S2. Significant enrichment of 19 GO terms defined as biological processes (P) and 16 GO terms defined as cellular components (C) was determined as described in Supplemental Figures S4 and S5. Briefly, Ontologizer 2.0 was used to quantify the significance of GO term enrichment in the user-defined expression patterns, clusters of expression, or the entire data set of 7,104 genes (Supplemental Table S1). For each significantly enriched GO term, the most significantly enriched expression pattern and cluster is presented. Up-regulated (red) and down-regulated (blue) expression is indicated at 0.5, 1, 4, and 24 h following the BR fluence-rate shift in lincomycin-treated (+LIN) and untreated (−LIN) seedlings. Plastid-regulated expression is similarly indicated at 0, 4, and 24 h relative to the fluence-rate shift. Color intensity is proportional to the degree of regulation. Positive correlation describes a similar response to the BR fluence-rate shift regardless of whether seedlings were treated with lincomycin. For genes that exhibit positive correlation, the correlation coefficient between the expression patterns in lincomycin-treated and untreated seedlings is greater than 0.95. The major cluster letters and pattern numbers are defined in Figure 2, E and 2F, and Supplemental Figures S2, A to F, and S3.

Figure 4.

Summary of biological process and biological response to stimulus GO terms that are enriched in particular expression patterns. User-defined expression patterns were obtained as described in Supplemental Figure S3. Clusters of expression were obtained as described in Figure 2 and Supplemental Figure S2. Significant enrichment of four GO terms defined as biological processes (P) and 16 GO terms defined as biological responses to stimulus (R) was determined as described in Supplemental Figures S4 and S5. Briefly, Ontologizer 2.0 was used to quantify the significance of GO term enrichment in the user-defined expression patterns, clusters of expression, or the entire data set of 7,104 genes (Supplemental Table S2). For each significantly enriched GO term, the most significantly enriched expression pattern and cluster is presented. Up-regulated (red) and down-regulated (blue) expression is indicated at 0.5, 1, 4, and 24 h following the BR fluence-rate shift in lincomycin-treated (+LIN) and untreated (−LIN) seedlings. Plastid-regulated expression is similarly indicated at 0, 4, and 24 h relative to the fluence-rate shift. Color intensity is proportional to the degree of regulation. Positive correlation is as described in Figure 3. The major clusters and patterns are defined in Figure 2, E and F, and Supplemental Figures S2, A to F, and S3.

Figure 5.

Light-regulated expression of PsbS and CHS in lincomycin-treated and untreated seedlings. The expression of PsbS and CHS at 0, 4, and 8 h relative to the BR fluence-rate shift was quantified using qRT-PCR. Three biological replicates were analyzed for each time point. For PsbS expression, expression in lincomycin-treated seedlings and untreated seedlings at 4 and 8 h is normalized to expression in lincomycin-treated seedlings at 0 h and untreated seedlings at 0 h, respectively. For CHS expression, expression is normalized to CHS expression in untreated seedlings at 0 h. * Statistically significant difference (P < 0.0001–0.049).

Figure 6.

Distinct light-regulated expression of six genes in lincomycin-treated and untreated seedlings. The expression of CDKB2.2, MCM5, AOS, LOX2, CCA1, and PRR5 at 0, 4, and 24 h relative to the BR fluence-rate shift was quantified using qRT-PCR. Four biological replicates were analyzed for each time point. The expression of a particular gene in untreated seedlings is normalized to the expression of that same gene in untreated seedlings at 0 h. The expression of a particular gene in lincomycin-treated seedlings is normalized to the expression of that same gene in lincomycin-treated seedlings at 0 h. * Statistically significant difference relative to 0 h (P ≤ 0.0001–0.04).

Figure 7.

Chlorophyll phenotypes caused by T-DNA insertion alleles. A, Chlorophyll phenotypes caused by T-DNA insertion alleles of genes that are more highly expressed in lincomycin-treated seedlings than in untreated seedlings following a BR fluence-rate shift. The wild type, gun1-101, and mutants containing T-DNA insertion alleles of genes that are expressed at least 1.5-fold higher in lincomycin-treated relative to untreated seedlings at 1 h following a BR fluence-rate shift were grown for 4 d in the dark and then transferred to continuous, broad-spectrum white light of 125 μmol m⁻² s⁻¹ for 24 h. Chlorophyll was extracted, quantified from four biological replicates for each line, and normalized to the wild type. Mean chlorophyll levels that were at least 2-fold greater than the chlorophyll levels of the wild type are indicated with a red dashed line and red bars. Mean chlorophyll levels that were at least 2-fold less than in the wild type are indicated with a blue dashed line and blue bars. Error bars represent 95% confidence intervals. T-DNA alleles were named using the arbitrary number assigned to each gene (Supplemental Fig. S8) and the last two numbers of the Salk accession code or the last three digits of the SAIL accession code. For example, the T-DNA alleles of gene 1 (At5g24120) are SAIL_1232_H11 and Salk_141383. These alleles are named 1-83 and 1-H11. B, Chlorophyll phenotypes caused by T-DNA insertion alleles of genes that are similarly expressed in lincomycin-treated and untreated seedlings following a BR fluence-rate shift. The deetiolation of mutants and the extraction and quantification of chlorophyll were performed as described in A.

Figure 8.

Deetiolation efficiencies of end mutants in various fluence rates. A, Deetiolation efficiencies of end mutants in three different fluence rates. The end mutants, gun1-101, and the wild type (ecotype Columbia-0 [Col-0]) were grown in the dark for 4 d and then irradiated with broad-spectrum white light at fluence rates of 15, 100, or 300 μmol m⁻² s⁻¹ for 24 h. Chlorophyll was extracted from four biological replicates for each line in each condition. Chlorophyll levels of the wild type (Col-0) are indicated with blue bars. Chlorophyll levels of mutants are indicated with red bars and green bars. Error bars represent 95% confidence intervals. B, Deetiolation rates of end mutants in 1 μmol m⁻² s⁻¹ BR light. The end mutants, gun1-101, and the wild type (Col-0) were grown in the dark for 4 d and then transferred to 1 μmol m⁻² s⁻¹ BR light. Chlorophyll was extracted from four biological replicates for each line at 0, 6, 12, and 24 h from the wild type (Col-0; blue curves) and the end mutants (red and green curves indicate distinct alleles). A Col-0 control was grown on the same plate for each end mutant. Error bars represent 95% confidence intervals.

Figure 9.

Lhcb1.4, RbcS1A, PsbS, and CHS expression in particular end mutants after an increase in fluence rate. A, Expression of Lhcb1.4 in particular end mutants after an increase in fluence rate. The wild type (ecotype Columbia-0 [Col-0]) and the indicated mutants were grown without lincomycin for 6 d in 0.5 μmol m⁻² s⁻¹ BR light and then transferred to 60 μmol m⁻² s⁻¹ BR light. Seedlings were collected immediately before the fluence-rate shift (0 h) and at 4 and 8 h following the fluence-rate shift. We used qRT-PCR to quantify transcript levels. The order of the lines from left to right is as follows: the wild type (Col-0; white bars), gun1-101 (gray bars), hy5 (brown bars), 1-83 (pink bars), 2-31 (red bars), 3-F12 (orange bars), 5-74 (yellow bars), 6-48 (green bars), 19-29 (blue bars), and 25-G09 (purple bars). Three biological replicates were analyzed for the wild type (Col-0) and each mutant in each condition. Error bars indicate sd. * Statistically significant difference between the wild type (Col-0) and a mutant (P = 0.0001–0.049). B, Expression of RbcS1A in particular end mutants after an increase in fluence rate. Analysis of RbcS1A expression was as described in A. * Statistically significant difference between the wild type (Col-0) and a mutant (P = 0.0002–0.03). C, Expression of PsbS in particular end mutants after an increase in fluence rate. Analysis of PsbS expression was as in A. * Statistically significant difference between the wild type (Col-0) and a mutant (P = 0.0001–0.004). D, Expression of CHS in particular end mutants after an increase in fluence rate. Analysis of CHS expression was as described in A. * Statistically significant difference between the wild type (Col-0) and a mutant (P = 0.0001–0.0046).

Figure 10.

Lhcb1.4, RbcS1A, PsbS, and CHS expression in particular end mutants in continuous 60 μmol m⁻² s⁻¹ BR light. A, Expression of Lhcb1.4 in particular end mutants. The wild type (ecotype Columbia-0 [Col-0]) and the indicated mutants were grown without lincomycin for 6 d in continuous 60 μmol m⁻² s⁻¹ BR light. We used qRT-PCR to quantify transcript levels. Three biological replicates were analyzed for the wild type (Col-0) and each mutant in each condition. Error bars indicate sd. * Statistically significant differences between the wild type (Col-0) and a mutant (P = 0.01–0.04). B, Expression of RbcS1A in particular end mutants. Analysis of RbcS1A expression was as described in A. * Statistically significant differences between the wild type (Col-0) and a mutant (P = 0.0002–0.03). C, Expression of PsbS in particular end mutants. Analysis of PsbS expression was as described in A. * Statistically significant differences between the wild type (Col-0) and a mutant (P = 0.02–0.03). D, Expression of CHS in the indicated end mutants. Analysis of CHS expression was as described in A. * Statistically significant differences between the wild type (Col-0) and a mutant (P = 0.04).