Hemin and magnesium-protoporphyrin IX induce global changes in gene expression in Chlamydomonas reinhardtii

Abstract

Retrograde signaling is a pathway of communication from mitochondria and plastids to the nucleus in the context of cell differentiation, development, and stress response. In Chlamydomonas reinhardtii, the tetrapyrroles magnesium-protoporphyrin IX and heme are only synthesized within the chloroplast, and they have been implicated in the retrograde control of nuclear gene expression in this unicellular green alga. Feeding the two tetrapyrroles to Chlamydomonas cultures was previously shown to transiently induce five nuclear genes, three of which encode the heat shock proteins HSP70A, HSP70B, and HSP70E. In contrast, controversial results exist on the possible role of magnesium-protoporphyrin IX in the repression of genes for light-harvesting proteins in higher plants, raising the question of how important this mode of regulation is. Here, we used genome-wide transcriptional profiling to measure the global impact of these tetrapyrroles on gene regulation and the scope of the response. We identified almost 1,000 genes whose expression level changed transiently but significantly. Among them were only a few genes for photosynthetic proteins but several encoding enzymes of the tricarboxylic acid cycle, heme-binding proteins, stress-response proteins, as well as proteins involved in protein folding and degradation. More than 50% of the latter class of genes was also regulated by heat shock. The observed drastic fold changes at the RNA level did not correlate with similar changes in protein concentrations under the tested experimental conditions. Phylogenetic profiling revealed that genes of putative endosymbiontic origin are not overrepresented among the responding genes. This and the transient nature of changes in gene expression suggest a signaling role of both tetrapyrroles as secondary messengers for adaptive responses affecting the entire cell and not only organellar proteins.

Figure 1.

Time course of changes in gene expression upon feeding hemin or MgProto for 0, 60, 120, and 180 min. At time 0, hemin or MgProto was added to cultures incubated in the dark at a final concentration of 9 μm. Samples for RNA isolation were taken after 60, 120, and 180 min. Control cultures were incubated under identical conditions but did not receive tetrapyrroles. The heat map-style color code is based on the absolute expression values from microarrays for the control at 0 min, with normalized intensity values set to blue if −4.46 or less and to red if 4.46 or greater. All probes shown fulfill the criteria of P ≤ 0.05 and display a FC ≥ 4.0 in at least one condition when compared with the respective control.

Figure 2.

Clusters of genes with similar changes in gene expression. A, Hierarchical clustering tree (squared Euclidean distances, Ward’s linkage) based on expression profiles of hemin- or MgProto-treated cells. The color codes are as follows: yellow, RNA levels unchanged; red, RNA levels up-regulated; blue, RNA levels down-regulated. B, Plot of maximal FC in gene expression observed after MgProto feeding versus that seen after hemin feeding. Grouping is based on hierarchical clustering (Manhattan distance, average linkage; performed using the hclust function in R). The pale symbols indicate genes for which only one treatment resulted in a significant (P ≤ 0.05) FC. The circles indicate genes that were also analyzed by northern-blot hybridization (Fig. 3).

Figure 3.

Verification of microarray data by northern-blot hybridization. Twelve genes representing the different regulatory groups 1 to 3 were selected, among them HSP70A as a positive control (Kropat et al., 1997; von Gromoff et al., 2008) for the effects of tetrapyrrole feeding. Two further hybridizations detected the mRNA of CBLP encoding a Gβ-like polypeptide (von Kampen et al., 1994) as a control for equal loading and the mRNA of the light-harvesting protein LHCBM2 as a typical photosynthetic polypeptide.

Figure 4.

Analysis of possible consequences of Proto feeding on gene expression. At 60, 120, and 180 min after Proto feeding, cultures were collected and RNA was prepared. The control cultures were kept under identical conditions except that no Proto was added. For comparison, RNA samples from the MgProto (M) and hemin (He) feeding experiments at 60 and 120 min shown in Figure 3 are also included. Hybridization with the CBLP probe served as a loading control.

Figure 5.

Classification of genes regulated by hemin and MgProto into selected functional groups. The numbers of down-regulated genes (group 2) are shown on the left side (red bars) and those of the up-regulated genes are shown on the right side (blue bars). The classification into functional groups was done manually based on the existing annotation. [See online article for color version of this figure.]

Figure 6.

Venn diagram showing the overlaps of genes that respond to different treatments. The overlap in the number of genes responding to MgProto/hemin feeding, heat shock (our data), anaerobic conditions (−O₂; Mus et al., 2007), or hydrogen-producing conditions (H₂; Nguyen et al., 2008) was determined. The diagram was constructed online (http://bioinfogp.cnb.csic.es/tools/venny/index.html).

Figure 7.

Sequence alignment and logo of PREs. A PRE defined in the promoter of HSP70A has also been discovered in the promoters of HSP70B, HSP70E, HEMA, and CGE1 (von Gromoff et al., 2006). Putative PREs in the promoter regions of genes found to be up-regulated upon the feeding of MgProto and hemin (by northern blotting) were identified using a position specific scoring matrix. The maximum allowed distance was set to 1,000 nucleotides upstream from the start of the first protein-coding exon. Thus, the nucleotide in front of the first translated codon has position −1. The suffix “inv” with the gene name indicates inverse orientation of the PRE in relation to the gene. The sequence logo was created using WebLogo (Crooks et al., 2004).

Figure 8.

Comparison of the phylogenetic origin of Chlamydomonas proteins whose genes respond to tetrapyrrole feeding (inner ring) against the total proteome of Chlamydomonas (outer ring). This graph summarizes the individual best phylogenetic match of Chlamydomonas proteins within the domain bacteria. To be considered, a Chlamydomonas protein had to meet an E-value of 10⁻⁴ or less against at least one protein sequence among all bacterial proteins in the nonredundant set from GenBank. Only the best (minimum E-value) hits were counted. As a result, bacterial homologs for 394 different proteins (corresponding to 397 individual microarray probes) responding to tetrapyrrole feeding (inner ring) and for 3,500 proteins represented by probes on the array (outer ring) were identified. Taxonomy information is based on National Center for Biotechnology Information taxdb downloaded on February 4, 2010.