• Maintenance of protein quality control and turnover is essential for cellular homeostasis. In plant organelles this biological process is predominantly performed by ATP-dependent proteases.
• Here, a genetic screen was performed that led to the identification of Arabidopsis thaliana Lon1 protease mutants that exhibit a post-embryonic growth retardation phenotype.
• Translational fusion to yellow fluorescent protein revealed AtLon1 subcellular localization in plant mitochondria, and the AtLon1 gene could complement the respiratory-deficient phenotype of the yeast PIM1 gene homolog. AtLon1 is highly expressed in rapidly growing plant organs of embryonic origin, including cotyledons and primary roots, and in inflorescences, which have increased mitochondria numbers per cell to fulfill their high energy requirements. In lon1 mutants, the expression of both mitochondrial and nuclear genes encoding respiratory proteins was normal. However, mitochondria isolated from lon1 mutants had a lower capacity for respiration of succinate and cytochrome c via complexes II and IV, respectively. Furthermore, the activity of key enzymes of the tricarboxylic acid (TCA) cycle was significantly reduced. Additionally, mitochondria in lon1 mutants had an aberrant morphology.
• These results shed light on the developmental mechanisms of selective proteolysis in plant mitochondria and suggest a critical role for AtLon1 protease in organelle biogenesis and seedling establishment.
Most intracellular events rely on protein cycling and homeostasis controlled by regulated biosynthesis of new polypeptides and precise degradation of pre-existing protein molecules. The ubiquitin/26S proteasome is responsible for the main mechanism of cellular protein turnover (Hershko & Ciechanover, 1998). This proteolytic mechanism modulates plant development by affecting embryogenesis, photomorphogenesis, floral development, circadian rhythms, senescence, and hormone signaling and response (Smalle & Vierstra, 2004). However, the protein targets of the ubiquitin/26S proteasome are restricted in the cytoplasm and nucleus. Plant organelles possess a distinct quality control system consisting of a range of ATP-dependent proteases belonging to the Clp, FtsH and Lon families (Adam et al., 2001; Ståhl et al., 2002; Sinvany-Villalobo et al., 2004; Janska, 2005; Sakamoto, 2006). Like the 26S proteasome, these proteases belong to the AAA+ (ATPases associated with diverse cellular activities) protein superfamily (Neuwald et al., 1999; Iyer et al., 2004). While significant progress has been made in characterizing the Clp and FtsH proteases, rather less is known about the Lon proteases in plants.
In yeast and mammals, Lon proteases are mitochondrial and control the selective turnover of nonassembled or misfolded proteins that accumulate in the matrix (Suzuki et al., 1994; van Dyck et al., 1994). Mutants in the mitochondrial Lon protease are associated with lack of mitochondrial genome integrity and consequent respiratory defects in yeast (Suzuki et al., 1994; van Dyck et al., 1994), and accumulation of oxidized protein aggregates in mammals (Lee et al., 1999; Bota & Davies, 2002). It has also been proposed that Lon protease serves as a chaperone in the assembly of protein complexes, a function that is independent of its proteolytic activity (Rep et al., 1996).
While it is likely that Lon proteases also fulfill similar roles in plants, the situation is complicated by the presence of a small gene family encoding several Lon isoforms that are predicted to be targeted to different subcellular organelles, including mitochondria, chloroplasts and peroxisomes (Sarria et al., 1998; Adam et al., 2001; Janska, 2005). Direct empirical evidence concerning the subcellular localization, biochemical activity and physiological role of Lon proteases in plants is rather limited. There are indications that plant Lon proteases, like their bacterial and eukaryotic homologs, combine proteolytic and chaperone-like activities (Janska, 2005; Sakamoto, 2006). There is also some evidence that a Lon-like protease may be induced in mitochondria during oxidative stress (Sweetlove et al., 2002; Lister et al., 2004). However, to date the only proven function of Lon proteases in plants is in the control of cytoplasmic male sterility (CMS) (Sarria et al., 1998).
Although the evidence is somewhat patchy, it is generally accepted that Lon proteases function in control of protein turnover and/or protein complex assembly in plant organelles. However, the physiological and developmental roles of Lon proteases have yet to be considered. In this study, we have isolated mutant alleles of the nuclear encoded Lon1 protease in a genetic screen for Arabidopsis thaliana seedlings with impaired primary root growth. We present a physiological and biochemical characterization of the mutants that suggests that Lon1 is required to maintain mitochondrial function, which is critical during post-germinative growth leading to seedling establishment.
Materials and Methods
Plant material and growth conditions
The lon1-1 mutant allele was isolated from a genetic screen of an M2 ethylmethanesulfonate (EMS)-mutagenized Arabidopsis thaliana Columbia (Col-0) background seed population. Before the genetic mapping, lon1-1 seeds were backcrossed four times to their Col-0 wild-type genetic background. The lon1-2 mutant is represented by the Salk T-DNA collection unicode SALK_012797 (http://signal.salk.edu) and was obtained from the Nottingham Arabidopsis Stock Centre (NASC, Nottingham, UK). Arabidopsis thaliana seeds were surface-sterilized and sown in 120 × 120 mm square Petri dishes containing 0.5× Murashige and Skoog (MS) medium (Duchefa, Haarlem, the Netherlands), pH 5.7, supplemented with 1% sucrose and solidified with 0.4% phytagel (Sigma, Gillingham, UK). After 48 h of stratification at 4°C, seedlings were positioned to grow vertically for 2–6 d after germination at 22°C in a Fitotron growth chamber (Weiss Gallenkamp Ltd, Loughborough, UK). Seed germination efficiencies (as determined by radicle emergence) were scored 6 d after plates had been transferred to growth chambers maintained at either 22°C or 32°C under a 16 : 8 h day:night cycle. Seeds used for all germination experiments were age-matched and stored under similar conditions for at least 1 month before the experiments were performed.
Measurements of lon1 and Col-0 primary root elongation were performed by scanning 5-d-old seedlings grown vertically. Morphometric analyses of mature plant and organ sizes were performed by taking digital photographs with a Sony DSC-F707 camera. Images obtained were further analyzed using the imagej software package (http://rsb.info.nih.gov/ij/) and statistically processed.
Positional cloning and gene isolation
The lon1-1 mutant was crossed with the A. thaliana polymorphic ecotype Landsberg erecta. Genomic DNA was isolated from 524 F2 individuals exhibiting the lon1-1 phenotype. Positional cloning was performed using combinations of single sequence length polymorphism (SSLP) and cleaved amplified polymorphic sequence (CAPS) markers (Supporting Information Table S1). The molecular markers were designed based on data for A. thaliana DNA polymorphisms available from the Monsanto Company (St. Louis, MO, USA) and The Arabidopsis Information Resource (TAIR) (http://www.arabidopsis.org; Jander et al., 2002). Genes with Arabidopsis Genome Initiative (AGI) annotation numbers At5g26840–At5g26910 were amplified using template DNA from BAC clone F2P16 with Phusion™ High-Fidelity DNA Polymerase (Finnzymes Oy, Espoo, Finland). Each open reading frame (ORF) was cloned into the SmaI site of the pGPTV-HPT multiple cloning site binary vector. The Agrobacterium tumefaciens strain C58C1 RifR containing the nononcogenic Ti plasmid pGV3101 was transformed with the pGPTV-HPT constructs by electroporation (Gene Pulser II; Bio-Rad, Hercules, CA, USA ). The constructs were introduced into lon1-1 and lon1-2 mutant alleles by the vacuum infiltration method as described previously (Haralampidis et al., 2002). Transgenic plants were selected on 30 mg l−1 hygromycin and grown to maturity for analysis of their ability to complement the lon1 mutant phenotype. The At5g26860 locus was amplified using as template genomic DNA from lon1-1 seedlings with a set of primers resulting in overlapping polymerase chain reaction (PCR) products of average size 600 bp. Sequencing analysis of these products led to the identification of the EMS mutation.
Yellow fluorescent protein (YFP) imaging
Citrine-YFP was PCR-amplified with 5′-AAACGCGT GGAGGTGGAGGTGGAGCTGTG-3′ forward and 5′-AAACGCGTGGCCCCAGCGGCCGCAGCAGC-3′ reverse primers to introduce a MluI restriction site (underlined sequence) at its flanks. The PCR product was cloned into the unique MluI restriction site of the U19013 cDNA clone. The resulting AtLon1-ΥFP transgene was cloned into the XbaI site of the pGPTV-HPT multiple cloning site binary vector, which already contained the constitutively expressed CaMV35S promoter. The resulting P35S:AtLon1-ΥFP construct was introduced into A. thaliana Col-0 wild-type plants by A. tumefaciens-mediated stable transformation.
Transgenic seedlings of the T2 and T3 generations were grown vertically and then stained with the mitochondrial-specific dye MitoTracker® Orange CMTMRos (Invitrogen, Paisley, UK). For staining, 4-d-old whole seedlings were submerged in 0.5× MS liquid medium supplemented with 10 nM MitoTracker® for 30 min and washed several times with 0.5× MS liquid medium. The roots of stained plants were sliced with razor blades and mounted between a slide and cover-slip in tap water. Root-hair cells were examined using an Olympus BX-50 (Olympus, Tokyo, Japan) light microscope equipped with epifluorescence. YFP fluorescence was visualized using the fluorescein filter #41017, Endow GFP Bandpass Emission Filter (Chroma Technology Corp., Rockingham, VT, USA), while the dye fluorescence was visualized with the U-MSWG rhodamine filter set (Olympus). Images were taken with the Olympus DP71 camera, using Cell^A (Olympus Soft Imaging Solutions, Münster, Germany). Final merging of images was performed using Adobe Photoshop CS2 (version 9.01) software.
RNA isolation and analysis
Total RNA was isolated from seedlings using the phenol-sodium dodecyl sulfate (SDS) extraction method as described previously (Haralampidis et al., 2002). RNA concentrations were determined spectrophotometrically and verified by ethidium bromide staining on agarose gels. DNA contaminations were eliminated by treating total RNA with RQ1 RNase free DNase (Promega, Madison, WI, USA). Reverse transcription (RT) was performed on 3 µg of total DNA-free RNA using Superscript II Reverse Transcriptase (Invitrogen) according to the manufacturer's instructions. First-strand cDNA synthesis and RT-PCR analysis for each transcript were performed using the pair of gene-specific primers shown in Table S2. The linear PCR amplification of each transcript was confirmed in preliminary experiments by comparing the relative amounts of PCR products under low and high RT-PCR cycles of amplification. Two biological and at least three technical replicates were performed in order to confirm that the amplification of each transcript was in the logarithmic phase.The products were analyzed by agarose gel electrophoresis and viewed with ethidium bromide staining. For comparative analysis and normalization, glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was chosen as an endogenous control.
The At5g26860 full-length cDNA clone U19013 (GenBank accession number AY117355) was PCR-amplified using the 5′-AATCTAGA CCACCATGTTAAAGCTCTTCACTTC-3′ and 5′-AATCTAGAGACAAAAGGTCAACGTGGCAATACA CA-3′ set of primers as forward and reverse, respectively. The underlined sequence denotes the XbaI restriction site introduced artificially to facilitate cloning into the shuttle vector pVT100U. This vector carries the constitutively expressed system of the yeast alcohol dehydrogenase gene (ADH1) promoter–terminator sequences and the URA3 gene for selection. Yeast transformation was performed using the lithium acetate method. The transformed yeast cells were selected on URA-free synthetic medium supplemented with 2% glucose. The heterologous functional complementation assay was performed as described previously (van Dyck et al., 1998).
Isolation of mitochondria
Intact coupled mitochondria were isolated from 12-d-old A. thaliana wild type (Col-0) and lon1 mutants according to Day et al. (1985). After disruption of the seedlings with three 5-s bursts in 300 ml of grinding medium (0.3 M sucrose, 25 mM tetrasodiumpyrophosphate, 1% (weight/volume (w/v)) bovine serum albumin, 1% (w/v) polyvinylpyrrolidone-40 (PVP-40), 2 mM Na2-ethylenediaminetetraacetic acid (EDTA), 10 mM KH2PO4, 20 mM ascorbate and 5 mM cysteine, pH 7.5) the extract was filtered and separated by differential centrifugation. Mitochondria were purified on a Percoll-PVP gradient, washed twice and resuspended in a buffer containing 0.3 M mannitol and 10 mM TES, pH 7.5.
Respiratory and TCA cycle enzyme activities of isolated mitochondria
The oxygen consumption of the mitochondrial respiratory chain complexes was measured in a Clark-type oxygen electrode according to Sweetlove et al. (2002). TCA cycle enzyme activities were assayed as described in Jenner et al. (2001).
Transmission electron microscopy
For transmission electron microscopy (TEM), hypocotyls of 5-d-old Col-0 and lon1-1 plants were fixed in a mixture of 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M sodium cacodylate buffer, pH 7.3, at 4°C for 2–3 h, post-fixed in 1% OsO4 for 2 h, washed in buffer, dehydrated in a series of ethanol, embedded in Spurr epoxy resin and polymerized at 70°C for 36 h. Ultrathin sections were cut with a Reichert OMU-3 ultramicrotome (C. Reichert, Vienna, Austria), stained with uranyl acetate and lead citrate and examined and photographed with a Zeiss 9S-2 transmission electron microscope (Carl Zeiss, Oberkochen, Germany).
Phenotypic defects of lon1 mutants
A genetic screen was set up in order to identify mutants with an impaired root growth pattern. We isolated a novel mutant, lon1-1, which appeared to have problematic root growth (Fig. 1b). Phenotypic analysis revealed that not only primary root elongation was affected but also post-embryonic growth as a whole. Hence, the primary root elongation pattern was actually delayed compared with wild-type plants, resulting in delayed seedling establishment soon after germination (Fig. 2a). This post-germinative growth retardation persisted during the entire biological cycle of lon1-1 plants. Thus, mature lon1-1 mutants were significantly shorter than the wild-type plants (Fig. 1f). The stem length of lon1-1 plants showed a reduction of approximately 40% compared with the wild type (Fig. 2b). In terms of the growth rate of other plant organs, the overall difference in size also persisted. Measurements of both leaf lamina area and silique length indicated that the difference in growth rate between wild-type and lon1 mutant plants persisted at maturity (Fig. 2c,d). In particular, the leaf lamina of lon1-1 plants showed a large reduction in area, the leaf lamina area being less than half that of wild-type plants (Fig. 2c).
In addition to the lon1-1 EMS mutant, a T-DNA-tagged lon1-2 allele was isolated. Every feature of lon1-2 development and growth examined showed a reduction pattern similar to that observed in lon1-1 plants (Figs 1, 2). However, the growth retardation of lon1-2 plants was less marked than that of lon1-1. For instance, the stem length (Figs 1f, 2b), leaf lamina area (Figs 2b, S1) and silique length (Figs 2d, S1) of lon1-2 plants were not as restricted as in lon1-1, albeit these differences were still significant compared with the wild type.
The mutants also showed a conditional germination efficiency phenotype that was apparent at elevated temperature (Table 1). Under normal growth conditions, the germination efficiencies of plants carrying the lon1 mutant alleles were more or less comparable and similar to that of the wild-type plants. However, the germination efficiency of plants carrying the lon1 mutant alleles was dramatically diminished when seeds were germinated at elevated temperature (32°C).
Table 1. Germination efficiency of Arabidopsis thaliana lon1 mutants under heat-shock conditions
Germinated seeds/total number of seeds measured.
Values are the mean ± SD of three independent experiments.Col-0, Columbia.
A positional cloning approach was applied to identify the AtLon1 gene. Recombination mapping placed the molecular locus responsible for the lon1-1 phenotype on the upper arm of chromosome V. Thus, novel markers were further generated restricting the AtLon1 locus in a zero-recombination interval of ~40 kb, flanked by proximal markers indicating a unique recombination event (Fig. 3a).
The eight candidate genes in the 40-kb interval with AGI annotation numbers At5g26840–At5g26910 were tested for their ability to complement the lon1-1 mutant phenotype. Phenotypic analysis of T1 and T2 stably transformed lon1-1 mutant plants revealed that At5g26860 restored the lon1-1 phenotype (Fig. 1d,f). In accordance with the genetic complementation, sequence analysis of the At5g26860 locus led to the identification of a premature termination codon in the lon1-1 mutant allele (Fig. 3b).
At5g26860 corresponds to a 2823-bp ORF containing 19 exons. It encodes a 941-amino acid polypeptide with a molecular mass of 104.1 kDa. A conserved Kozak consensus CCACCATG resides upstream of the initiation of translation, which marks the beginning of a putative mitochondrial transient peptide. The genoplante™predotar V1.03 program (http://urgi.versailles.inra.fr/tools/predotar) predicted that Lon1 protease is probably localized to the mitochondrial matrix. Having linked the At5g26860 molecular locus to the lon1-1 mutant allele, an additional T-DNA-tagged allele, lon1-2, was identified. As in the case of lon1-1, the AtLon1 gene restored the post-germinative growth retardation of plants carrying the lon1-2 mutant allele (Fig. 1e,f). Genotypic characterization of the lon1-2 allele confirmed the site of T-DNA insertion in exon 18 of the AtLon1 coding sequence (Fig. 3b,c).
Primary structure analysis of the A. thaliana Lon1 protease revealed the N-terminal (N) domain (amino acid residues 1–503), the AAA+ module adjacent to the sensor- and substrate-discrimination (SSD) domain (residues 504–809) and the proteolytic (P) domain (residues 810–941) (Fig. 3d). The AAA+ module consists of two fundamental domains: the nucleotide-binding (α/β) domain containing the conserved motifs Walker A and B and the helical (α) domain. The AAA+ module is involved in ATP hydrolysis and in protein substrate remodeling (Neuwald et al., 1999; Iyer et al., 2004).
The AtLon1 protein is localized to the mitochondria
To confirm the predicted localization of Lon1, a Lon1-YFP fusion was expressed in transgenic A. thaliana plants. The YFP reporter gene was inserted into the 11th exon of AtLon1 cDNA, proximal to the carboxy terminal of the Lon1 protease N-domain. The expression of the AtLon1-YFP transgene was driven by the constitutively expressed CaMV35S promoter. In root-hair cells expressing the AtLon1-YFP translational fusion, YFP fluorescence was restricted to structures of typical mitochondrial morphology, 0.5–1-µm round- or elliptical-shaped particles that were distributed throughout the cytoplasm, often forming conglomerates (Fig. 4b). To confirm that the YFP fluorescence observed indeed localized to the mitochondria, whole transgenic seedlings were stained with MitoTracker Orange CMTMRos, which is a mitochondrial-specific dye for living cells (Fig. 4c). The signals of the red mitochondrial-specific dye and green YFP fluorescence were compared within the same root-hair cell using epifluorescence microscopy. The overlap of the green and red signals resulted in yellow fluorescent mitochondria (Fig. 4d). This co-localization of the fluorescent dye that specifically stains mitochondria and the green fluorescence emanating from the AtLon1-YFP translational fusion revealed that the Lon1 protease is targeted in planta to mitochondria.
Functional complementation of yeast pim1 cells
Yeast cells lacking the PIM1 gene accumulate lesions in the mtDNA (Suzuki et al., 1994; van Dyck et al., 1994). As essential respiratory chain components are encoded by the mitochondrial genome (Unseld et al., 1997; Giegéet al., 2005), pim1 mutants are respiratory-deficient and fail to grow on nonfermentable carbon sources such as glycerol or ethanol. In order to investigate whether the A. thaliana Lon1 and yeast PIM1 proteases are functionally equivalent, we carried out a complementation analysis of S. cerevisiae. The AtLon1 gene heterologous functional complementation assay was performed by assessing the restoration of the respiratory-deficient Δpim1 yeast cells that carried intact mtDNA as a result of the presence of the SSS1 (Suppressor of sec61) gene (van Dyck et al., 1998). The SSS1 protein stabilizes mtDNA integrity in yeast cells when PIM1 protease is absent, despite the fact that the respiratory incompetence remains. Detailed information concerning the genotype of yeast cells used for this assay is presented in Table S4.
Functional analysis revealed that the A. thaliana Lon1 gene successfully complemented the respiratory-deficient phenotype of Δpim1 yeast cells (Fig. 5). When their mtDNA was stabilized, Δpim1 cells were able to grow on medium containing glycerol as the sole carbon source as a result of the activity of the A. thaliana homologous Lon1 protease. However, the expression of AtLon1 promoted minimal growth of Δpim1 cells on nonfermentable carbon sources at 36°C (Fig. 5). The results demonstrate the functional conservation between the mitochondrial localized yeast PIM1 and A. thaliana Lon1 proteases. Furthermore, they demonstrate that specific functions of PIM1, for example under heat stress, within yeast mitochondria cannot be performed by the plant homolog.
Pattern of Lon1 gene expression
Having confirmed the mitochondrial localization of Lon1, we were interested to determine whether the gene was predominantly expressed in nonphotosynthetic tissues or organs with high growth rates. These tissues or organs are highly dependent upon respiration to meet their energy requirements. The AtLon1 gene exhibited high levels of expression in germinating seedlings and organs of embryonic origin such as cotyledons and primary roots (Fig. 6a). However, in mature root tissues AtLon1 transcripts were barely detected, while the level of expression was higher in developing stems and inflorescences than in stems of adult plants.
A potential role for Lon proteases in thermal stress has been suggested by heat-shock transcriptional induction of the bacterial Lon (Chin et al., 1988) and yeast PIM1 homolog (van Dyck et al., 1994). To determine whether AtLon1 expression is affected by elevated temperatures, total RNA was isolated from 5-d-old seedlings grown under several heat-shock conditions (Fig. 6b). The AtLon1 transcript exhibited a moderate decrease when seedlings were grown either at 32°C overnight or at 42°C for 1 h. However, no significant decline of expression was observed when seedlings were grown at either 32°C or 37°C for 3 h. The same pattern of a moderate reduction in expression was found in several independent experiments. We confirmed that the seedlings were exposed to conditions that induced a heat-shock response by analyzing the transcription of the nuclear encoded cytoplasmic (AtHSP90.1) or mitochondrial (mtHSP90) heat-shock 90 proteins (Milioni & Hatzopoulos, 1997; Haralampidis et al., 2002). The results indicate that A. thaliana Lon1 transcription is down-regulated under either acute or prolonged heat-shock conditions. Similarly, the expression of the maize (Zea mays) Lon1 gene has also been reported to decline in response to thermal stress (Barakat et al., 1998). Hence, unlike Escherichia coli Lon and yeast PIM1, the expression of the homologous plant Lon1 genes from A. thaliana and maize declined in response to heat shock.
Role of Lon1 in maintenance of mitochondrial function
To investigate the role of AtLon1 in maintenance of mitochondrial function, we examined whether AtLon1 deficiency alters the transcription level of nuclear encoded genes associated with mitochondrial activities. Thus, the effects of a small increase in growth temperature (to 27°C) and the absence of light on the abundance of the measured transcripts were examined. Using a primer set that anneals downstream of the T-DNA insertion of lon1-2 mutant plants, no AtLon1 transcript was detected. Growth of plants in the dark or under mild heat-shock conditions did not alter the level of AtLon1 gene expression. There was no evidence that the expression pattern of nuclear genes encoding respiratory proteins, namely the cytochrome c 1 (Cytc-1) and Cytc-2 genes (Welchen & Gonzalez, 2005) and the alternative oxidase 1a (Aox1a) gene, was altered in the lon1 mutant in comparison with the wild type (Fig. 6c). Furthermore, the pattern of enolase 1 (Eno1) and alcohol dehydrogenase 1 (ADH1) expression in lon1 mutants, encoding the extramitochondrial respiratory pathways markers enolase and alcohol dehydrogenase, was the same as in wild-type plants (Fig. 6c).
In yeast, proper expression of mitochondria-encoded genes, such as the cytochrome b (Cob) and cytochrome c oxidase subunit I (Cox1) genes, requires PIM1 protease activity (van Dyck et al., 1998). Therefore, transcriptional analysis of the A. thaliana mitochondrial genes Cob and Cox1–Cox3, encoding cytochrome b of respiratory complex III and cytochrome c oxidase subunits I–III of respiratory complex IV, respectively, was performed. In contrast to yeast cells, transcription of A. thaliana mitochondrial genes in lon1 mutants was similar to that in wild-type plants, even for the intron containing the Cox2 gene (Fig. 6d). Consequently, we investigated whether such expression of the mitochondrial genome in lon1 mutants is attributable to changes in the abundance of transcripts encoding other members of the lon family. Nevertheless, the pattern of expression was unaltered in comparison to wild type (Fig. 6e).
We also assessed the respiratory capacity of mitochondria isolated from lon1 mutant seedlings to further elucidate the role of Lon1 in organelle function. Mitochondria isolated from lon1 mutants had similar outer-membrane integrity as assessed by latency to cytochrome c (data not shown), suggesting that overall mitochondrial integrity is not affected. However, both lon1 mutants showed evidence of reduced respiratory capacity, with a significant decrease in the respiration by isolated mitochondria of succinate and cytochrome c via complexes II and IV, respectively (Fig. 7a). The activities of other respiratory complexes and enzymes (complex I, the external NADH dehydrogenase and the alternative oxidase) were unaffected. Additionally, in the absence of AtLon1 the activities of at least five of seven measured TCA cycle enzymes, namely citrate synthase, aconitase, NAD- and NADP-dependent isocitrate dehydrogenase and 2-oxoglutarate dehydrogenase, were significantly decreased (Fig. 7b).
The mitochondrial morphology of the lon1-1 mutant is abnormal
Morphological examination of pim1 yeast mitochondria revealed that they had an abnormal morphology (Suzuki et al., 1994). The shape of the mitochondria is irregular, and electron-dense inclusion bodies, probably representing aggregates of mitochondrial proteins, accumulate in the matrix. Thus, we examined the mitochondrial morphology of lon1-1 hypocotyl tissue in ultrathin sections by transmission electron microscopy. The transmission electron micrographs indicated that lon1-1 mitochondria were swollen and nonreticulated with a poorly developed internal membrane system composed of few discernible cristae (Fig. 8b,d). In contrast to lon1-1, mitochondria from wild-type hypocotyl tissue sections had the expected appearance, that is, reticulated with numerous cristae and an electron-dense mitochondrial matrix (Fig. 8a,c). Notably, these ultrastructural features of lon1-1 mitochondria are reminiscent of the promitochondrial morphology of maize (Logan et al., 2001) and rice (Oryza sativa) (Howell et al., 2006) dry seeds, suggesting a potential role of the A. thaliana Lon1 protease in mitochondrial biogenesis during germination.
Lon1 is targeted to mitochondria
Arabidopsis thaliana has four genes encoding proteases of the LonA subfamily (Sinvany-Villalobo et al., 2004; Janska, 2005; Sakamoto, 2006) and, using prediction programs, AtLon1 has been characterized as mitochondrial. Using YFP imaging we have confirmed this localization of AtLon1 (Fig. 4). In accordance with these results, immunoassays (Sarria et al., 1998) and mitochondrial proteome analysis (Heazlewood et al., 2004; Ito et al., 2006) have also indicated that AtLon1 is mitochondrially targeted. AtLon2 was initially predicted to be chloroplast-located but recently it was characterized as peroxisomal, sharing high sequence similarity with the rat peroxisomal-specific isoform (Kikuchi et al., 2004). Although the AtLon3 protein has been characterized as mitochondrial, we could detect no gene transcripts (Fig. 6e) and a recent report has suggested that it is probably a pseudogene (Ostersetzer et al., 2007). Experimental evidence supports dual targeting of AtLon4 to both chloroplasts and mitochondria (Sakamoto, 2006; Ostersetzer et al., 2007). Beyond A. thaliana, two maize Lon orthologs, designated maize Lon1 and Lon2, have been identified (Barakat et al., 1998). The ZmLon1 gene probably encodes a peroxisomal protein (Janska, 2005). However, ZmLon1 was able to partially complement the yeast mitochondrial pim1 mutant phenotype, suggesting moderate functional similarities between the Lon isoforms located in different subcellular compartments (Barakat et al., 1998).
Lon1 is important for mitochondrial function and biogenesis
During seed imbibition, two fundamental biological processes are performed resulting in germination and seedling establishment (Howell et al., 2006). The first consists of mitochondrial assembly or biogenesis, which refers to the developmental transition of prοmitochondria to active mature mitochondria. The second process involves the mobilization of storage reserves. Mitochondrial biogenesis requires the coordinated synthesis of proteins encoded by both nuclear and mitochondrial genomes, and recent evidence suggests that this is under post-transcriptional control at the level of complex assembly (Giegéet al., 2005). Unassembled subunits must be turned over to prevent toxic aggregation. Based on our analyses of the A. thaliana lon1 mutants, we postulate that the ATP-dependent proteolytic activity of the Lon1 protease is essential for controlling the quality of newly imported or synthesized mitochondrial polypeptides and thereby permitting proper mitochondrial biogenesis. Several lines of experimental evidence support this hypothesis. First, A. thaliana Lon1 is able to complement the respiratory-deficient phenotype of yeast pim1 mutant cells (Fig. 5). Secondly, the activities of mitochondrial respiratory chain complexes and TCA cycle enzymes were decreased in the lon1 mutants compared with the wild type (Fig. 7). Thirdly, transmission electron micrographs indicated an aberrant mitochondrial morphology with some resemblance to promitochondria (Fig. 8) (Logan et al., 2001). This aberrant morphology could be a result of incomplete biogenesis, or the consequences of the accumulation of nonassembled protein complexes. Given the extreme nature of this phenotype, one might have expected a more severe deficiency in respiratory capacity than we observed. However, for practical reasons, the respiratory capacity assays were performed on mitochondria isolated from a much later developmental stage (12-d-old seedlings) and it is conceivable that some recovery could have occurred during development. Further systematic study would be required to establish the exact cause of the aberrant morphology and any link to the pre-biogenesis state of promitochondria. Our analysis suggests that, unlike its counterparts in yeast and E. coli, Lon1 is not essential for normal mitochondrial genome expression. It is worth noting that there may be some redundancy in Lon1 function in mitochondria as a result of the possible presence of a dual-targeted Lon4. Additionally, nuclear genes encoding subunits of mitochondrial respiratory complexes were not altered in expression. Thus, the reduction in activity of complexes II and IV observed in isolated mitochondria must be post-transcriptionally regulated.
Finally, we observed that AtLon1 gene expression was high in organ tissues of embryonic origin, including cotyledons and primary roots, and in rapidly developing organs with increased energy requirements such as developing stems and inflorescences (Fig. 6a). Although both the stem and the inflorescence are photosynthetically competent organs, their energy needs are considerably high in order to accomplish rapid development. Such energy requirements can be partially fulfilled by enhanced mitochondrial activity. Thus, the developmental and growth abnormalities that we observed in lon1 mutants are probably attributable to compromised mitochondrial function and capacity, and suggest a vital role for Lon1 in maintaining that capacity during these developmental phases.
A heat-conditional phenotype of lon1 mutants
Lon proteases have also been implicated in tolerance of oxidative stress and heat shock. In the first case, it has been shown that a lack of mammalian Lon1 leads to accumulation of oxidized proteins and is implicated in ageing (Lee et al., 1999; Bota & Davies, 2002). Although we did not investigate this explicitly, the expression of Aox1a, which is known to be induced by oxidative stress (Clifton et al., 2005), was not increased in the lon1 mutants (Fig. 6c), suggesting either an absence of mitochondrial oxidative stress or that the Aox1a signaling pathway is disrupted in the mutant. Further evidence in support of this view is the lack of induction of the other AtLon genes in the lon1 mutants, which have been reported to be induced by oxidative stress (Lister et al., 2004) (Fig. 6e).
The heat-inducible expression of the E. coli Lon (Chin et al., 1988) and yeast PIM1 genes (van Dyck et al., 1994) has led to the suggestion of a role for Lon proteases in thermotolerance. Heat shock is a kind of proteotoxic stress that severely increases the fraction of cellular proteins occurring in an unfolded and/or misfolded state, thereby enhancing the probability of intracellular aggregate formation. These aggregates have to be disposed of by targeted proteolysis, otherwise they may cause significant biological complications. However, the A. thaliana mitochondria-targeted AtLon1 orthologs were not heat-shock-inducible (Fig. 6b) and in fact their expression declined in elevated temperature conditions. This has also been observed for maize peroxisomal-specific ZmLon1 (Barakat et al., 1998). Nevertheless, we did observe a pronounced heat-conditional phenotype relating to seed germination. When seeds were germinated at elevated temperatures, the germination efficiency of lon1 mutants was dramatically diminished. Such a conditional phenotype reveals a substantial role of AtLon1 protease during germination and seedling establishment.
In summary, we have identified mutant alleles of the A. thaliana ATP-dependent Lon1 protease. As Lon proteases regulate protein abundance and degradation of nonfunctional polypeptides in plant organelles, the A. thaliana lon1 mutants represent the first such genetic platform identified in plant species. These mutants provide novel and fascinating insights into mitochondrial biogenesis, a prerequisite for post-embryonic growth and seedling establishment. Further analysis will unravel the sequence of the molecular events that control plant mitochondrial differentiation.
The authors thank NASC, UK for kindly providing plant materials; Vitaly Citovsky at the State University of New York, USA for citrine-YFP; and Thomas Langer and Mirko Koppen from the University of Cologne, Germany for yeast mutants and helpful advice. This work was supported by Pythagoras grants Ι and II from the Greek Ministry of Education to SR and PH and by a grant from the EU ERA-Plant Genomics programme to LJS. GD is indebted for funding to the Greek State Scholarships Foundation.