Biological roles of the Podospora anserina mitochondrial Lon protease and the importance of its N-domain

Abstract

Mitochondria have their own ATP-dependent proteases that maintain the functional state of the organelle. All multicellular eukaryotes, including filamentous fungi, possess the same set of mitochondrial proteases, unlike in unicellular yeasts, where ClpXP, one of the two matricial proteases, is absent. Despite the presence of ClpXP in the filamentous fungus Podospora anserina, deletion of the gene encoding the other matricial protease, PaLon1, leads to lethality at high and low temperatures, indicating that PaLON1 plays a main role in protein quality control. Under normal physiological conditions, the PaLon1 deletion is viable but decreases life span. PaLon1 deletion also leads to defects in two steps during development, ascospore germination and sexual reproduction, which suggests that PaLON1 ensures important regulatory functions during fungal development. Mitochondrial Lon proteases are composed of a central ATPase domain flanked by a large non-catalytic N-domain and a C-terminal protease domain. We found that three mutations in the N-domain of PaLON1 affected fungal life cycle, PaLON1 protein expression and mitochondrial proteolytic activity, which reveals the functional importance of the N-domain of the mitochondrial Lon protease. All PaLon1 mutations affected the C-terminal part of the N-domain. Considering that the C-terminal part is predicted to have an α helical arrangement in which the number, length and position of the helices are conserved with the solved structure of its bacterial homologs, we propose that this all-helical structure participates in Lon substrate interaction.

Figure 1. The mitochondrial PaLON1 protein.

(A) Schematic representation of the PaLON1 protein outlining the three domains present in both prokaryotes and eukaryotes. The N-domain, which is the most divergent domain between Lon proteins, is followed by the highly conserved ATPase and protease domains. Within the N-domain, the most conserved region is within the C-terminal part (hatched). The line referring to residues 382 to 619 indicates the part of the protein presented in (B). Diamond (S423L), point (L430P), and inverted triangles (Δ514–567) mark changes induced by PaLon1-31, PaLon1-1 and PaLon1-f, respectively. (B). Primary sequence and secondary structure of the C-terminal part of the N-domain of B. subtilis, E. coli, and P. anserina Lon proteases. Sequences were aligned using the Clustal W program. Conserved amino acids are boxed in black (identical) and gray (similar). For the P. anserina sequence (PODAN), changes induced by PaLon1 mutations are represented by the same symbols as in (A). The GenBank accession numbers for B. subtilis (BACSU) and E. coli (ESCCO) proteins are CAA99540.1 and AAC36871.1, respectively. The Walker A motif of the central ATPase domain is boxed and begins at position 607, 356, and 354 in P. anserina, E. coli and B. subtilis proteins, respectively. The predicted consensus secondary structure of the PaLON1 region was determined on the NPS@ Web server using a combination of available methods. For the same region, the secondary structure information available for E. coli and B. subtilis proteins ends at residue 245 or contains a gap of 36 amino acids (dotted line), respectively . For the B. subtilis protein, structure information was not available after the last α helix just before the Walker A motif. Secondary structures are indicated above each sequence as follows: lines, α helices; c letter, random coil (no secondary structure); and question mark (?), ambiguous state.

Figure 2. MtDNA profile of the

ΔLon1 strain. MtDNA profiles were determined by HaeIII digestion of total DNA extracted from mycelium for the indicated dying strains. HaeIII digestion of mtDNA from a young wild-type mycelium was used as a control. The arrows indicate the senDNAα multimeric subgenomic molecules present in dying strains (right and left panels). The identity of the senDNAα molecules was assessed by Southern blot analysis using an intron α specific probe (right panel). In addition to senDNAα (2.5 kb), hybridization revealed two HaeIII fragments on the intact mitochondrial genome (1.9 and 0.8 kb). Note that these two bands were only detected in young wild-type (Control) and dying ΔLon1 strains.

Figure 3. Phenotypic characteristics of PaLon1 mutants.

(A) Mycelium phenotype of PaLon1-1 germinating ascospores on germination medium at 27°C after 2 days of growth. A cross between PaLon1-1 and the wild-type strain gave rise to a progeny of PaLon1-1 germinating ascospores with a less dense mycelium (letters: b, c, e, f, h, i, k) than that of the wild type (letters: a, d, g, j, l). (B) Growth phenotype exhibited by the PaLon1 mutants. Strains were grown on M2 standard medium for 2 days (27°C and 36°C), 3 days (18°C), or 7 days (11°C). The genotype of each strain is shown in the table, except for the rmp1-1 (mat−) and rmp1-2 (mat+) alleles that are represented by a gray and white tone, respectively. (C) DASPMI staining of mitochondria. Mitochondria of growing strains (2 days at 27°C on M2) were stained with DASPMI, a vital mitochondrion-specific dye. For each indicated strain, filaments were gently mixed with a drop of DASPMI (25 mg/ml) directly on microscope slides and observed immediately with a fluorescence microscope (450–490/500–550 nm). All panels are at the same magnification and the scale bar corresponds to 5 μm. (D) Hydrogen peroxide (H₂O₂) sensitivity. Nine subcultures of each strain were inoculated in M2 or M2 supplemented with 0.005% (1.47 mM) or 0.01% (2.94 mM) H₂O₂. Growth (cm) was determined by measuring the radius of each thallus after 2 days at 27°C in the dark. Error bars indicate standard deviation. The statistically significant increase in the sensitivity of the PaLon1-f and ΔLon1 strains to 2.94 mM H₂O₂ is marked by an asterisk. In each case, the p-value (0.001) is below 0.05, as determined by the Mann-Whitney test.

Figure 4. Detection of PaLON1 mutant proteins.

Mitochondrial extracts (70 μg) purified from the indicated strains were resolved on an SDS-polyacrylamide gel and subjected to immunoblotting. The PaLON1 protein and the β-subunit of mitochondrial ATPase were detected by a P. anserina anti-PaLON antibody and an S. cerevisiae anti-Atp2 antibody, respectively. Mitochondrial extraction and western blotting were repeated at least twice for each strain. The left and right panels correspond to two independent membranes.