Long bugs to short plants – the Lon protease in protein stability and thermotolerance


(Author for correspondence:
tel +61 8 6488 7245, fax +61 8 6488 4401, email harvey.millar@uwa.edu.au)

Mitochondria are the cellular organelles responsible for respiration, oxidizing organic acids to release carbon dioxide and reducing oxygen to water. In the process, mitochondria synthesize ATP and export it to the cell to drive many energy-utilizing processes in growth and development. To undertake these processes, mitochondria contain many hundreds, and perhaps even thousands, of different proteins. Each protein participates as an enzyme in a complex series of biochemical pathways to complete the task of respiration. In addition, mitochondria in plants are involved in a wide array of other processes like nitrogen metabolism, photorespiration or even making cofactors such as biotin and folate. Mitochondria arose from endosymbiotic bacteria in the ancient eukaryotic cell. Over time, most of the mitochondrial genes were transferred to the nucleus and now the proteins are post-translationally imported into mitochondria from the cytosol. While the transcription, translation, import and processing of these nucleus-encoded mitochondrial proteins has received considerable attention in plants, very little work has been carried out on the array of proteases that are likely to be instrumental in protein stability, turnover and assembly within mitochondria.

‘Arabidopsis Lon1 has been moved from the shadows to the limelight, and the nonredundant role of this class of proteases in plant mitochondrial homeostasis is beginning to be uncovered.’

Proteases in plant mitochondria

Proteases from the Lon, FtsH and Clp families are known to be targeted to mitochondria in plants (Adam et al., 2001). Proteolysis plays an important role in post-translational control by the targeted degradation of short-lived proteins and also helps to maintain protein quality control by removing defective, damaged or even damaging proteins. However, how important each of these proteases is to plant mitochondrial structure and function, and what their targets are, has largely remained a mystery.

In this issue of New Phytologist, Rigas and colleagues (pp. 588–600) report the discovery of a Lon1 mutant through a forward genetic screen for root growth in Arabidopsis. What follows is a very thorough analysis of this mutant, providing both convincing evidence of the importance of a single mitochondrial protease for root growth, coupled to a detailed molecular analysis of the impact of this protease loss on mitochondrial form and function. As a result, Arabidopsis Lon1 has been moved from the shadows to the limelight, and the nonredundant role of this class of proteases in plant mitochondrial homeostasis is beginning to be uncovered.

Lon proteases across the kingdoms

So what is Lon? Lon is an ATP-dependent protease that was first found in bacteria. In fact, it was the first ATP-dependent protease discovered (Chung & Goldberg, 1981). It is a multidomain polypeptide with a variable N-terminal domain, a central ATPase domain and a C-terminal domain that contains the proteolytic activity (Fig. 1). Its name derives from an Escherichia coli K-12 mutant lon, identified in the 1960s and named for the long form of the mutant cells (Donch & Greenberg, 1968). Lon acts as a heat shock protein in bacteria, being transcriptionally induced by heat. Lon proteases have since been found in virtually all living organisms, from Archaea to Eubacteria, to plants and animals.

Figure 1.

The protein domains of the Lon protease. (a) The crystal structures of the three domains of the Escherichia coli Lon protease (Botos et al., 2004a,b; Li et al., 2005). Crystal structures were obtained from the Research Collaboratory for Structural Bioinformatics (RCSB) database http://www.rcsb.org. (b) A schematic representation of the Lon protease showing the variable N-terminal domain, the AAA+ ATP hydrolysis module adjacent to the sensor and substrate-discrimination module (SSD), in the central domain, and the proteolytic P-domain.

The presence of Lon-like proteases in mitochondria was first noted in the 1970s, and an extensive literature on this class of protease in mitochondria from a variety of organisms has traversed the past 30 yr. In the last decade it has become clear that Lon preferentially degrades oxidatively modified proteins in the mitochondrial matrix, notably the oxidized, hydrophobic form of aconitase (Bota & Davies, 2002). As Lon expression and activity are known to decline with age, its inactivation may contribute to the accumulation of the oxidatively modified protein aggregates often observed during aging and in cells from diseased individuals. Beyond its proteolytic role, other functions have been attributed to Lon in mitochondria, including mitochondrial DNA (mtDNA) binding and chaperone activity for the assembly of respiratory chain complexes (Fig. 2; Ngo & Davies, 2007).

Figure 2.

The functional role of Lon and consequences of loss of Lon function. Substrates for proteolysis by Lon protease can be misfolded, oxidatively damaged, overproduced or aggregated proteins in cells. Lon protease has been shown to be involved in the degradation or refolding of such proteins. Organisms lacking or depleted in Lon protease show a variety of phenotypes, as shown. mtDNA, mitochondrial DNA.

Lon proteases in plants

In plant mitochondria, little work on Lon proteases has been reported. The four Arabidopsis isomers of the Lon protease are predicted to reside in either chloroplasts or mitochondria (Adam et al., 2001), and proteomic evidence has identified several in mitochondria (Heazlewood et al., 2007). Transcriptionally they are only minimally affected by stresses such as high light, cold and heat, in contrast to other organelle protease classes that are highly induced by such conditions (Sinvany-Villalobo et al., 2004). The only clear documentation of a molecular function that we are aware of in plants is the report of Lon protease being responsible for degradation of a cytoplasmic male sterility-associated protein, ORF239, in bean mitochondria in vegetative, but not in reproductive, tissues (Sarria et al., 1998).

In this issue of New Phytologist, Rigas et al. report the story of one particular mutant discovered during the screen of an ethane methyl sulfonate (EMS) mutagenized population of Arabidopsis to identify mutants with impaired root-growth phenotypes. This mutant had not only impaired primary root elongation but also retardation of postgerminative growth that persisted throughout its life cycle. Positional cloning identified the locus responsible for the mutation to be a 40-kb region on the upper arm of chromosome V. Using the power and ease of reverse genetics in Arabidopsis, the eight candidate genes in this region were then tested for complementation of the phenotype, and the stably transformed insertion of At5g26860 was able to rescue the phenotype. This gene encoded a protein with a Lon protease-like amino acid sequence. Sequence analysis further revealed that the mutation was caused by the introduction of a premature termination codon in the At5g26860 locus.

To confirm the function of this gene locus, Rigas et al. used a yeast mutant that is missing an ATP-dependent, mitochondrially located protease with 30% identity to the Lon protease from Bacillus brevis (Van Dyck et al., 1994). This mutation leads to disruption of mtDNA and thus to respiratory deficiency in yeast. Complementation of Δpim1 yeast with At-Lon1 allowed growth of the Δpim1 yeast at 30°C on the nonfermentable carbon source, glycerol, that normally prevents Δpim1 growth (van Dyck et al., 1998). However, the authors also showed that At-Lon1 only promoted minimal growth of Δpim1 at 36°C compared with wild-type yeast, suggesting that while there was functional conservation of PIM 1 and At-Lon1, under heat stress At-Lon1 was unable to perform the specific function of PIM1. This difference was further examined by measuring the At-Lon1 transcript abundance when seedlings were grown at elevated temperatures, and the results showed that the Lon-1 transcript is, in fact, mildly downregulated during both acute and prolonged exposure to heat. Interestingly, this result is supported by the observation that maize Lon-1 gene expression also declined in response to thermal stress (Barakat et al., 1998). This is in stark contrast to Lon proteases in many nonplant species where these genes are transcriptionally induced during heat shock (Fig. 2). Interestingly, the authors showed that while Lon is not a heat shock protein in Arabidopsis, loss of Lon in the lon1-1 and lon1-2 lines leads to a temperature-dependent germination phenotype, thus still linking Lon with a heat-induced functional role in plant germination (Fig. 2). Together, these observations suggest that while the plant Lon-1 homologues may perform many similar roles to the yeast PIM1, there are differences in substrate recognition, heat stability and/or transcriptional control that will influence their roles during thermal stress (Fig. 2).

The cellular location of At-Lon1 was confirmed by the insertion of the yellow fluorescent protein (YFP) reporter gene into At-Lon1 cDNA. The YFP fluorescence in root cells was seen to overlay directly in transgenic plant seedling stained with MitoTracker Orange, revealing that the Lon-1-YFP is in planta targeted to mitochondria. To dig deeper into the effect of the absence of At-Lon1 on mitochondrial function, the respiratory capacity of isolated mitochondria was tested using oxygen-consumption assays of respiration and direct analysis of specific mitochondrial enzymes by spectrophotometry. These assays showed that while mitochondria of the mutant maintained outer-membrane integrity, respiratory capacity was reduced when oxidizing succinate and cytochrome c, indicating damage or decrease in complexes II and IV. The largest effect was on the activity of complex IV, the cytochrome c terminal oxidase of the respiratory chain, which was only ∼20% of the wild-type level, providing a strong marker for one molecular change in the mutants. By contrast, the activities of Complex I (CI), the external NADH dehydrogenase and the alternative oxidase, were unaffected by the mutation. The absence of Lon1 also led to significant decreases in the activities of five tricarboxylic acid (TCA) cycle enzymes. A morphological examination of hypocotyl tissue using transmission electron microscopy showed that lon1-1 mitochondria were swollen and had a poorly developed internal membrane structure with few discernable cristae. This suggested either a damaged mitochondrial structure or perhaps even an undeveloped structure. Similar-looking organelles, termed protomitochondria, have been observed in dry seeds but are normally modified during germination in a mitochondrial maturation process (Logan et al., 2001; Howell et al., 2006).

Future perspectives on Lon in plants

The lack of transcriptional up-regulation of Lon by heat stress compared with the temperature-induced germination phenotype of the mutants, suggests that Lon may be more of a constitutive protease in plant mitochondria relative to the other proteases present (Sinvany-Villalobo et al., 2004) but this still needs to be verified. While Lon might have important roles during heat stress, a role as a chaperone or assembly factor might better explain its expression pattern that is correlated to tissues with a high growth rate rather than to environmental factors. The dual role of Lon in yeast as chaperone and protease was clearly shown in mutants lacking a functional C-terminal proteolytic domain (Rep et al., 1996). The chaperone role is probably a result of its central AAA+ ATPase domain, a fold often noted in other proteins with chaperone roles (Neuwald et al., 1999). Sophisticated remodeling of Lon1 and other Lon proteases in plants now awaits researchers interested in untangling both their specialization and their multifaceted roles in plant mitochondrial function.