Unlike animals, plants synthesize isoprenoids via two pathways, the cytosolic mevalonate (MVA) pathway and the plastidial 2-C-methyl-d-erythritol-4-phosphate (MEP) pathway. Little information is known about the mechanisms that regulate these complex biosynthetic networks over multiple organelles. To understand such regulatory mechanisms of the biosynthesis of isoprenoids in plants, we previously characterized the Arabidopsis mutant, lovastatin insensitive 1 (loi1), which is resistant to lovastatin and clomazone, specific inhibitors of the MVA and MEP pathways, respectively. LOI1 encodes a pentatricopeptide repeat (PPR) protein localized in mitochondria that is thought to have RNA binding ability and function in post-transcriptional regulation of mitochondrial gene expression. LOI1 belongs to the DYW subclass of PPR proteins, which is hypothesized to be correlated with RNA editing. As a result of analysis of RNA editing of mitochondrial genes in loi1, a defect in RNA editing of three genes, nad4, ccb203 and cox3, was identified in loi1. These genes are related to the respiratory chain. Wild type (WT) treated with some respiration inhibitors mimicked the loi1 phenotype. Interestingly, HMG-CoA reductase activity of WT treated with lovastatin combined with antimycin A, an inhibitor of complex III in the respiratory chain, was higher than that of WT treated with only lovastatin, despite the lack of alteration of transcript or protein levels of HMGR. These results suggest that HMGR enzyme activity is regulated through the respiratory cytochrome pathway. Although various mechanisms exist for isoprenoid biosynthesis, our studies demonstrate the novel possibility that mitochondrial respiration plays potentially regulatory roles in isoprenoid biosynthesis.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Isoprenoids are indispensable compounds for all organisms because they are used as structural components of plasma membranes, in signalling molecules such as hormones, in pigments and in protein prenylation. Isoprenoids are biosynthesized by condensation of common five-carbon intermediates, isopentenyl diphosphate (IPP) and its isomer dimethylallyl diphosphate (DMAPP). Unlike animals, plants are characterized by the biosynthesis of tens of thousands of isoprenoids, which help them adapt to multiple environmental stressors. Plants biosynthesize IPP and DMAPP via two pathways, the mevalonate (MVA) pathway in the cytosol and the 2-C-methyl-d-erythritol 4-phosphate (MEP) pathway in plastids (Figure 1). Although metabolites flow between them (Hemmerlin et al., 2003; Kasahara et al., 2002; Nagata et al., 2002), these two pathways are compartmentalized under normal physiological conditions. Cytoplasmic isoprenoids (e.g., sterols and brassinosteroids) and mitochondrial isoprenoids (e.g., side chains of ubiquinone) are synthesized via the MVA pathway, while the MEP pathway is used to synthesize plastidic isoprenoids (e.g., carotenoids and the side chains of chlorophyll and plastoquinones) and some isoprenoid-type phytohormones (e.g., gibberellins, abscisic acid and cytokinins).
HMGR is highly regulated at transcriptional and post-transcriptional levels in mammalian cells. HMGR transcription is up-regulated through the action of transcription factors binding to the sterol regulatory elements (SRE). Membrane-bound transcription factors, i.e., sterol regulatory element binding proteins (SREBPs), recognize these sequences and function as central regulators of cellular lipid homeostasis (Goldstein et al., 2006). In the presence of the oxysterol, Insig, an endoplasmic reticulum (ER) membrane protein, blocks movement of SREBPs from the ER. Accumulation of lanosterol stimulates binding of HMGR to Insig, resulting in the ubiquitinylation of HMGR on two cytosolic lysine residues and subsequent degradation by the 26S proteasome (Song et al., 2005). Mammalian HMGR is regulated not only at transcript and protein levels but also at the enzyme activity level. HMGR is phosphorylated and inactivated by an AMP-activated protein kinase (AMPK) (Clarke and Hardie, 1990).
The regulatory mechanisms of plant HMGR have also been investigated using physiological, biochemical and genetic approaches. Inhibition of squalene synthase and squalene epoxidase has been shown to result in up-regulation of HMGR in tobacco BY-2 cells (Wentzinger et al., 2002) and in Arabidopsis (Nieto et al., 2009). These reports indicate that, like mammalian HMGR, plant HMGR undergoes feedback regulation. On the other hand, no SRE sequence or SREBP transcription factor has been identified in the plant genome, suggesting that novel regulatory mechanisms exist in plants. SNF1-related protein kinase 1 (SnRK1s), a homologue of the α subunit of AMPK purified from Brassica oleracea (Dale et al., 1995) and spinach (Douglas et al., 1997), reportedly phosphorylates and inactivates the catalytic domain of bacterially expressed Arabidopsis HMGR1 in vitro. However, it remains to be resolved whether this regulatory mechanism of HMGR enzyme activity occurs in plant cells. Genetic analyses have also been used to identify novel regulatory factors of HMGR. Two Arabidopsis mutants, rim1 (Rodríguez-Concepción et al., 2004) and loi1 (Kobayashi et al., 2007), have been reported as showing resistance to the inhibitor of HMGR. One of these, rim1, is an allelic mutant of phyB. Transcription of HMG1 is up-regulated in the dark (Learned, 1996; Learned and Connolly, 1997) and Rodríguez-Concepción et al. 2004 have shown that transcription of HMGR is up-regulated in photoreceptor and photosignalling mutants.
We previously isolated a lovastatin insensitive 1 (loi1) mutant with a resistant phenotype against both lovastatin and clomazone. Like plants grown with squalestatin and terbinafine, the inhibitors of squalene synthase and squalene epoxidase, respectively (Nieto et al., 2009), plants grown with lovastatin showed higher HMGR activity, indicating that lovastatin induces feedback up-regulation of HMGR in plant cells. We found that this feedback regulation of HMGR was activated in loi1 (Kobayashi et al., 2007). Surprisingly, the LOI1 gene encodes a novel pentatricopeptide repeat (PPR) protein that is localized in mitochondria (Kobayashi et al., 2007). The PPR protein family is characterized by the presence of tandem arrays of a degenerate 35-amino acid repeat motif (Small and Peeters, 2000). The family is greatly expanded in higher plants, with approximately 450 members in A. thaliana and even more members in other land plants in which genome sequences have been analysed. PPR proteins are considered to be imported into organelles, taking part in post-transcriptional processes of gene expression (Schmitz-Linneweber and Small, 2008). Some PPR proteins are reportedly involved in photosynthesis and plastid development through regulation of plastid gene expression (de Longevialle et al., 2008; Schmitz-Linneweber and Small, 2008). Also, in mitochondria, PPRs are considered to play an important role as regulators of mitochondrial gene expression. Molecular functions of a few mitochondrial PPRs have been reported, such as splicing (OTP43; de Longevialle et al., 2007), editing (MEF1; Zehrmann et al., 2009) and translation (PPR336; Uyttewaal et al., 2008). Although some PPRs are known to be restorative factors of cytoplasmic male sterility (CMS) through post-transcriptional regulation of mitochondrial genes (Fujii and Toriyama, 2008), the physiological functions of mostly mitochondrial PPRs, other than PPRs related to fertility, have not yet been characterized.
Here, we report that LOI1 is an essential factor for RNA editing of three mitochondrial genes, nad4, ccb203 and cox3, which constitute the respiratory chain. This finding suggests that the activity of the respiratory chain of loi1 is lower than that of WT. WT seedlings treated with inhibitors of the respiratory chain complex demonstrated lovastatin- and clomazone-resistant phenotypes and feedback regulation of HMGR activity was activated by antimycin A, one of these inhibitors. These results show that a novel regulatory mechanism of HMGR through the mitochondrial respiratory chain exists in plants.
Isolation of specific cellular RNA substrates bound to LOI1-FLAG
To identify the native target RNA of LOI1, we generated 35S promoter-driven LOI1-FLAG over-expressed transgenic Arabidopsis and hooked the target RNA bound to LOI1-FLAG to the anti-FLAG antibody. The RNAs bound to LOI1 from 2-week-old seedlings were separated as the RNA:LOI1-FLAG complex. The RNAs were purified, reverse-transcribed and cloned. A total of 76 clones were sequenced and 28 independent potential LOI1 binding RNAs were obtained. As LOI1 protein is predicted to localize in mitochondria (Kobayashi et al., 2007), we aimed at targets with mitochondrial RNAs. BLAST analyses demonstrated that two independent clones originated from mitochondrial RNAs. One was the transcript of the cox3 gene, encoding a component of cytochrome c oxidase (COX) and the other was atp1, a gene for subunit 1 of F1-ATPase in mitochondria. One clone of cox3 and two clones of atp1 were included in the 76 clones.
Defects in RNA editing of cox3 in loi1
PPR proteins are thought to take part in post-transcriptional regulation of organellar gene expression (Schmitz-Linneweber and Small, 2008). LOI1 may also be involved in post-transcriptional regulation of mitochondrial RNA, such as RNA splicing, RNA processing, RNA editing from C to U and RNA stability and translation. The conserved motifs, E and DYW, exist in the C-terminal domain of many PPR proteins. LOI1 belongs to the DYW subclass of PPR protein (containing E and DYW motifs), which may be correlated with RNA editing, based on phylogenetic analysis (Salone et al., 2007). Recently, many DYW subclass (Cai et al., 2009; Kim et al., 2009; Okuda et al., 2009; Robbins et al., 2009; Yu et al., 2009; Zehrmann et al., 2009; Zhou et al., 2008) and some E subclass (containing the E motif; Chateigner-Boutin et al., 2008; Okuda et al., 2006, 2007) PPR proteins have been shown to be involved in RNA editing. Therefore the editing of cox3 and atp1 RNAs was examined by sequencing RT-PCR products from WT and loi1-1. Seven editing sites have been reported to exist in cox3 RNA (Giegé and Brennicke, 1999). Sequence analyses showed that editing occurred at six sites in both WT and loi1-1. In contrast, the nucleotide of the seventh editing site (+422nt) in loi1-1 was not detectably edited, while that in WT was normally edited. The amino acid residue containing the edited nucleotide, the141st residue of cox3, was altered from Leu (CTT) in WT to Pro (CCT) in loi1-1 (Figure 2). This deficiency of editing in loi1-1 was also observed in another loi1 mutant allele, loi1-2 (Figure S1) and was complemented by the introduction of the gene cassette of the LOI1 promoter::LOI1 gene (Figure 2). On the other hand, no difference in the RNA editing of atp1 was observed between WT and loi1-1 (Table S1). These results demonstrate that LOI1 is required for editing at the +422nt site of the cox3 gene.
Reduced COX activity in loi1
COX3 is a component of the complex IV cytochrome c oxidase (COX) complex in the mitochondrial respiratory chain. To investigate the influence of a deficiency in editing of cox3 RNA in loi1, we examined COX activity using cytoplasmic extracts of 2-week-old seedlings of WT and loi1-1. As shown in Figure 3, COX activities of loi1-1 were approximately 34% of those in WT. This implies that the 141st proline substituted by leucine in cox3 may improve enzyme activity or protein stability of COX.
The other two defects of RNA editing in loi1
To identify whether LOI1 is involved in editing events of mitochondrial RNA other than cox3, we examined RNA editing in loi1-1 at 414 sites, which can be examined using primers for sequencing analysis in ORFs, out of 441 sites known to be edited in mitochondria of Arabidopsis, using Columbia cell suspension culture (Giegé and Brennicke, 1999). In 2-week-old WT seedlings, 295 sites were edited and 119 sites were not edited (Table S1). However, in 2-week-old loi1-1 seedlings, RNA editing was found in 292 sites, but not in 122 sites containing the +422nt of the cox3 gene. We determined that the two additional sites edited in WT but not in loi1-1 were located in nad4, encoding a gene for a component of complex I, NADH dehydrogenase and in ccb203, encoding a gene for cytochrome assembly protein, independently (Figure 2). These defects of RNA editing in loi1-1 were also identified in another allelic mutant of loi1-2 (Figure S1), but not in the complementation line (Figure 2). While editing observed in the +124nt of nad4 (TTG to CTG) is silent at the protein level, the defect in RNA editing at the +344nt of ccb203 in loi1-1 alters an amino acid from leucine to proline. These results suggest that LOI1 is responsible for three editing events on three distinct mitochondrial transcripts, nad4, ccb203 and cox3.
Inhibitors of the respiratory chain complex mimic the loi1 phenotype
To ascertain whether the phenotypes described for loi1 seedlings (lower sensitivity to lovastatin and clomazone) were caused by impaired enzyme activity or protein stability of the respiratory complex, as in COX, we examined lovastatin and clomazone sensitivity of WT seedlings grown with respiration inhibitors, such as rotenone, antimycin A, salicylhydroxamic acid (SHAM) and oligomycin. Rotenone is an inhibitor of complex I (NADH dehydrogenase), antimycin A is an inhibitor of complex III (ubiquinone:cytochrome c oxidoreductase) of the cytochrome pathway, oligomycin is an inhibitor of complex V (F0F1–ATP synthase) and SHAM inhibits the alternative oxidase (AOX). The inhibition of complex I by rotenone resulted in a loi1-like clomazone insensitive phenotype, but with a WT-like lovastatin sensitivity. Inhibition of complex III by antimycin A resulted in a loi1-like lovastatin insensitive phenotype and a weak clomazone-resistant phenotype. Inhibition of complex V by oligomycin resulted in a loi1-like lovastatin and clomazone insensitive phenotype (Figures 4, S3 and S4). In contrast, inhibition of AOX by SHAM did not have any influence on lovastatin and clomazone sensitivity in a WT seedling (Figures S2 and S3). These data suggest that inhibition of the mitochondrial respiratory pathway affects isoprenoid biosynthesis.
We previously reported that HMGR activity in loi1-1 grown with lovastatin was higher than that in WT grown under the same conditions, while transcription of HMGR did not differ between these two plants, suggesting that protein accumulation and/or enzyme activity of HMGR is up-regulated by feedback regulatory systems (Kobayashi et al., 2007). We then measured HMGR protein accumulation and activity in WT grown with lovastatin combined with antimycin A and rotenone. Protein accumulation of HMGR was up-regulated by lovastatin treatment, but not affected by either antimycin A or rotenone (Figure 5a). HMGR activity in WT grown with lovastatin and antimycin A was higher than that in WT grown only with lovastatin, while that in WT grown with lovastatin and rotenone was similar to that in WT grown only with lovastatin (Figure 5a). Lovastatin and antimycin A did not affect transcription of HMG1 and HMG2 (Figure 5b). These data suggest that inhibition of complex III up-regulates relative activity of HMGR.
Taken together, our results suggest that defects in the respiratory chain affect the isoprenoid biosynthesis pathway, especially post-translational up-regulation of HMGR activity, eventually resulting in lovastatin insensitivity.
We found that RNA editing was not observed in loi1 in three editing sites of transcripts in nad4 (a gene for a component of Complex I), ccb203 (a gene for cytochrome assembly protein) and cox3 (a gene for a component of complex IV), all involved in the respiratory chain (Figure 2) and that inhibition of complexes III and I of WT seedlings resulted in lovastatin- and clomazone-resistant phenotypes, respectively (Figures 4, S3 and S4). Inhibition of complex III of WT seedlings by antimycin A also resulted in up-regulation of HMGR activity (Figure 5a). On the other hand, inhibition of the AOX pathway by SHAM had no effect on either lovastatin or clomazone sensitivity. Based on these results, we concluded that inhibition of the cytochrome pathway results in up-regulation of HMGR activity and inhibition of complex I affects metabolic activity of the MEP pathway (Figure 6). As we could not determine whether inhibition of complex III or complex IV results in up-regulation of HMGR activity, further analyses using other mitochondrial PPR mutants may be informative. Inhibition of complex V by oligomycin caused a resistant phenotype against both lovastatin and clomazone in WT seedlings (Figures 4, S3 and S4). Because inhibition of complex V by oligomycin may affect total respiratory chain activity, including complex I and complex III, WT seedlings grown with oligomycin should show lovastatin- and clomazone-resistant phenotypes.
The feedback regulatory mechanisms of HMGR are well studied in mammals and feedback regulation of HMGR is also known in plants. Almost all regulatory factors found in mammals, such as SRE, SREBP and Insig, are not conserved in plants; hence, plants must have different regulatory mechanisms. HMGR activity is up-regulated when Arabidopsis is grown in lovastatin (Kobayashi et al., 2007), squalestatin and terbinafin (Nieto et al., 2009) and when tobacco BY-2 cells are cultured in squalestatin and terbinafin (Wentzinger et al., 2002). The up-regulation of HMGR occurs with no increase in either HMGR transcription or protein accumulation when Arabidopsis is grown with squalestatin or terbinafine (Nieto et al., 2009), whereas HMGR activity is up-regulated with an increase of protein accumulation but without any increase of transcription when Arabidopsis is grown with lovastatin (Figure 5). This difference suggests at least two independent feedback regulatory systems: one is up-regulation of HMGR enzyme activity in response to a deficiency of metabolites downstream of squalene and the other is up-regulation of HMGR protein accumulation in response to a deficiency of metabolites upstream of squalene. Antimycin A enhanced the up-regulation of HMGR in Arabidopsis grown with lovastatin (Figure 5). As no further increase of protein accumulation was observed in this case, this enhancement must be an activation of HMGR enzyme activity. The increased HMGR protein accumulation in Arabidopsis grown with lovastatin was much greater than the increase of HMGR activity in the same plants, suggesting that lovastatin treatment strongly induced HMGR protein accumulation, but that a part of HMGR in this plant is inactivated. The inhibition of complex III by antimycin A and reduced COX activity in the loi1 mutant may cancel this inactivation of HMGR enzyme activity.
What is the signal from the cytochrome pathway of mitochondria that regulates HMGR enzyme activity? It is well known that the nucleus controls organelle gene expression, development and function. However, recent studies have demonstrated that organelles send signals to the nucleus to control nuclear gene expression, a process called retrograde signalling. Plastid-to-nucleus retrograde signalling is currently well studied in Arabidopsis (Pogson et al., 2008). Although analysis of CMS events and the use of respiration inhibitors have shown that mitochondrial retrograde signalling also exists in higher plants, a large part of this signalling still remains unknown (Fujii and Toriyama, 2008).
One candidate would be signalling through AMPK/SnRK1. HMGR enzyme activity is regulated by phosphorylation in mammals and is inactivated by AMPK. This inactivation occurs in response to high AMP concentrations and low ATP/ADP ratios (Hardie et al., 1997). It has been proposed that AMPK does not play a role in end-product feedback regulation of HMGR but, rather, comes into play when cellular ATP levels are depleted, thereby lowering the rate of cholesterol synthesis and preserving energy stores of the cell (Sato et al., 1993). Arabidopsis HMGR1 is inactivated by phosphorylation at Ser577 by cauliflower HMGR kinase in vitro (Dale et al., 1995). AMPK homologous genes, designated SNF1-related protein kinase 1 (SnRK1), have been identified from many plant species, such as Arabidopsis (Le Guen et al., 1992) and tobacco (Muranaka et al., 1994). Moreover, the level of seed sterols of transgenic tobacco plants expressing modified HMGRs lacking the SnRK1 target site was up to 2.5-fold higher than that of WT plants (Hey et al., 2006). These findings raise the possibility that phosphorylation of HMGR by SnRK1 may contribute to inactivation of HMGR enzyme activity in Arabidopsis grown with lovastatin. We are now characterizing the physiological meaning of the phosphorylation and inactivation of HMGR by AMPK homologous kinase(s).
What is the signal from the cytochrome pathway of mitochondria to SnRK1s? Orthologues of AMPK/SNF1 are a conserved family and are activated in response to energy limitation and starvation of the carbon sources in mammals and yeast. In higher plants, it has been suggested that SnRK1s are also inactivated by sugars and share central roles in energy signalling (Baena-González et al., 2007). On the other hand, the activity of Arabidopsis SnRK1s in light-grown plants is stimulated by sucrose (Bhalerao et al., 1999). The data provided here demonstrate that inhibition of the respiratory chain, which may equate to ATP depletion, results in up-regulation of HMGR. Although it remains to be resolved whether the regulatory mechanisms of SnRK1 involve energy limitation and/or sugar starvation in higher plants, we consider that SnRK1 may play a role as a signal transduction factor to the cytosol and/or nucleus in response to energy limitation caused by abnormality of the respiratory chain.
Another candidate that mediates mitochondrial retrograde signal is the redox signal. Redox-based signalling may be a key component in mitochondria–nucleus communication (Noctor et al., 2007; Rhoads and Subbaiah, 2007). The inhibition of complex III by antimycin A (Maxwell et al., 1999) and the ppr40 mutation (Zsigmond et al., 2008) have been shown to induce an increase in cellular ROS accumulation, leading to oxidative stress in Arabidopsis cells. AOX is thought to be up-regulated by ROS. Inhibition of complex I by rotenone has been shown to induce AOX and an increase of ROS (Garmier et al., 2008). AOX expression was also induced in the loi1 mutant (Figure S5). It is possible that ROS in mitochondria might affect isoprenoid biosynthesis in both the cytoplasm and plastid. The role of redox signalling in isoprenoid biosynthesis should be further studied.
Indeed, it has been reported that AMPK and ROS play an important role in mitochondrial retrograde signalling in Drosophila (Owusu-Ansah et al., 2008). Furthermore, mitochondrial complex I deficiency causes ROS accumulation and activates a signal for progressing the cell cycle from the G1 phase to the S phase in Drosophila (Owusu-Ansah et al., 2008). We propose the possibility that the abnormality of mitochondrial respiration affects isoprenoid biosynthesis in the cytosol and plastid through SnRK1 and/or mitochondrial ROS. Further analysis of these phenomena may reveal a novel role of mitochondrial signalling in higher plants.
Inhibition of complex I by rotenone resulted in a clomazone-resistant phenotype, namely up-regulation of the MEP pathway (Figure 4b). The clomazone-resistant phenotype of loi1 may be derived from a deficiency of RNA editing of +124nt in the nad4-encoded component of complex I (Figure 2). However, as this editing site was ‘silent’, it is difficult to conclude whether this RNA editing deficiency in nad4 inhibits complex I activity. Inhibition of complex III by antimycin A was also induced in a weak clomazone-resistant phenotype (Figure 4b). It is possible that inhibition of the cytochrome pathway by treatment with antimycin A and by a defect in editing of cox3 in the loi1 mutant, affects the activity of complex I, similar to treatment with oligomycin. The existence of a regulatory system based on protein stability of MEP pathway enzymes was demonstrated using the fosmidomycin-resistant mutants, rif1 (Flores-Pérez et al., 2008) and rif10 (Sauret-Güeto et al., 2006). Protein accumulation of DXS and DXR is not affected in loi1 (Kobayashi et al., 2007). Inhibition of complex I by rotenone and by impaired splicing of nad4 in the css1 mutant, induces significant remodelling of metabolic pathways involving the mitochondria and other compartments (Garmier et al., 2008; Nakagawa and Sakurai, 2006). Regulation of the MEP pathway by inhibition of complex I may demonstrate the possibility of a novel metabolic regulatory mechanism between mitochondria and plastids.
LOI1 is a member of the PPR protein family targeted in mitochondria. The PPR protein family, characterized by the presence of tandem arrays of a degenerated 35-amino acid repeat motif (Small and Peeters, 2000), is divided into two subfamilies, P and PLS. In the PLS subfamily, tandem arrays of PPR and PPR-like motifs are usually followed by several conserved motifs, E and DYW, arranged in this order. Based on the different appearance of C-terminal motifs, the PLS subfamily is divided into three subgroups, PLS (without any C-terminal motif), E (with an E motif) and DYW (with E and DYW motifs) (Schmitz-Linneweber and Small, 2008). LOI1 belongs to the DYW subgroup. As a result of comprehensive RNA editing, defects in RNA editing in three genes, nad4, ccb203 and cox3, were identified (Figure 2). Although DYW class PPR proteins involved in RNA cleavage, such as CRR2, have been reported (Nakamura and Sugita, 2008), we could not find abnormal cleavage of cox3 in the loi1 mutant in RNA gel blot analysis (data not shown). Thus, the possibility that LOI1 is involved in RNA cleavage may be low. Very recently, it was reported that the DYW subclass PPR proteins play a role in RNA editing in the chloroplast (Cai et al., 2009; Okuda et al., 2009; Robbins et al., 2009; Zhou et al., 2008) and in mitochondria (Kim et al., 2009; Zehrmann et al., 2009). Some of these PPR proteins (CRR22, CRR28, MEF1, OGR1) have been reported to be involved in multiple sites of RNA editing. The mechanisms of how PPR domains of a PPR protein recognize multiple RNA sequences are very interesting. It has been proposed that several RNA editing sites may be recognized by a single trans factor, depending on the conserved cis sequence in the 5′ region of the edited C (Chateigner-Boutin and Hanson, 2002). Similar sequences were found near three independent RNA editing sites in which LOI1 is involved (Figure S6). Biochemical and molecular biological analyses of LOI1 will provide more information on RNA editing in mitochondria.
Our studies using loi1 and inhibitors of the respiratory chain demonstrated that metabolic activity is regulated over multiple organelles: cytoplasm, plastids and mitochondria. What is the molecular mechanism of the signalling system in the inter-organellar network? Further studies of isoprenoid regulatory mechanisms found in loi1 will be useful to metabolic engineering and improve our understanding of the signalling system in the inter-organellar network.
Plant growth conditions
WT Arabidopsis (ecotype Col-0), loi1-1 and loi1-2 (Kobayashi et al., 2007) were germinated and grown in 1× MS plates containing 3% sucrose and 1.2% agar. After stratification for at least 2 days at 4°C, plates were incubated at 23°C under long-day conditions (16 h under fluorescent white light, 8 h dark). Lovastatin (Calbiochem, http://www.calbiochem.com), clomazone (provided by Dr Tadao Asami), antimycin A (Sigma, http://www.sigmaaldrich.com), rotenone (Sigma) and oligomycin (Calbiochem) were dissolved in dimethylsulfoxide. SHAM (Sigma) was dissolved in water. These inhibitors were diluted to different final concentrations in growth medium (300 nm lovastatin, 300 nm clomazone, 2 μm antimycin A, 5 μm rotenone, 5 μg ml−1 oligomycin and 100 μm SHAM).
For complementation of the loi1-1 mutant, the genomic DNA fragment containing the LOI1 promoter and the LOI1 gene was amplified with 5′-CACCTTCCTGCGATAGATAAACACCCAAA-3′ and 5′-TTACCAATAATCCTTACAAGAACATATCCCAT-3′ and cloned into pENTR Directional-TOPO (Invitrogen, http://www.invitrogen.com). The resultant entry clone was integrated into the GATEWAY™ converted binary vector pBCR-112 (Seki et al. unpublished data) using the attL X attR (LR) recombination reaction.
Target RNA isolation
The 35S promoter-driven LOI1-FLAG over-expressed transgenic Arabidopsis and Arabidopsis containing the empty destination vector (negative control) were used in this experiment. To prepare organelle fractions, 2-week-old seedlings were homogenized in ice-cold Solution A (250 mm sucrose, 250 mm tricine, 2 mm EDTA, 25 mm Na2S2O5, 100 mm ascorbic acid, 5 mm 1,4-dithioerythritol, 2.5% poly (vinylpolypyrrolidone), 1 mm phenylmethylsulfonyl fluoride (PMSF), pH 8.0) and the homogenate was centrifuged at 200 g for 5 min at 4°C. The supernatant was subsequently centrifuged at 10 000 g for 15 min at 4°C. The pellet (organelle fraction) was resuspended in solution B (250 mm sucrose, 100 mm tricine, 2 mm EDTA, 50 mm NaCl, 1 mm PMSF, pH 7.5) containing 1% Triton X-100, rotated for 60 min at 4°C and then glycerol was added to a final concentration of 10%. This organelle fraction was diluted with four volumes of Solution B to reduce the Triton X-100 to 0.2% and then centrifuged at 15 000 g for 15 min at 4°C. The RNA:LOI1–FLAG complex was purified using anti-FLAG M2 agarose resin (Sigma). The purified LOI1–FLAG was assessed with silver staining and western blotting using the anti-LOI1 antibody. The eluate (RNAs bound to LOI1–FLAG) was extracted with phenol:chloroform (1:1) and chloroform extraction, followed by ethanol precipitation. First-strand 5′-RACE cDNA for the co-purified isolated RNA was synthesized using a BD SMART™ RACE cDNA Amplification kit (BD Bioscience Clontech, http://www.clontech.com) and then used in PCR amplification. Compared with the PCR product from the negative control, the specific products for LOI1–FLAG were purified, cloned and sequenced. The sequences of these clones were analysed using BLAST.
Analysis of RNA editing
Total RNA was isolated from 2-week-old seedlings using the RNAeasy Plant Mini kit (Qiagen, http://www1.qiagen.com), treated with RNase-free DNase to remove genomic DNA and reverse-transcribed using a RT-PCR kit (AMV) with random primers (Takara Bio, http://www.takara-bio.com). The sequencing search of mitochondrial RNA editing sites was carried out as described by Okuda et al. (2007). The sequences of PCR primers used for amplifying the mitochondrial genes are presented in Table S1. The resultant RT-PCR products were sequenced directly.
COX enzyme assay
Two-week-old leaves were ground in liquid N2 and extracted into 100 mm KPi (pH 7.5) buffer containing 250 mm sorbitol and 0.2 mm EDTA. After filtration with a Falcon 70-μm filter, the homogenate was centrifuged at 13 000 g for 10 min at 4°C. The pellet was resuspended in 50 mm KPi (pH 7.5) buffer containing 1% Triton X-100, rotated for 30 min at 4°C and then centrifuged at 5000 g for 5 min at 4°C. Glycerol was added to the supernatant to a final concentration of 10% and this was then stored at −80°C. COX activity was measured as described previously (Yoshida et al., 2007). The protein concentration of the extract was measured using the Bradford method.
Gene expression analyses
Total RNA was extracted and purified using an RNeasy Plant Mini Kit (Qiagen) from 1-week-old seedlings of WT and loi1-1 grown on 1× MS agar plates containing 3% sucrose and 1.2% agar, with or without inhibitors. The relative quantification of HMG1 and HMG2 was performed with an Applied Biosystems 7500 real-time PCR system, using the following primers and probes: HMG1-forward, 5′-GGCGTGACAAGATCCGTTACA-3′; HMG1-reverse, 5′-GCAATAATGGCGCCGAGTT-3′; HMG10-probe, 5′-CACGTCGTCACTATCACA-3′; HMG2-forward, 5′-TACTCGGTGTGAAAGGATCAAACA-3′; HMG2-reverse, 5′-CCGAACCAGCCACTATTCTTG-3′; HMG2-probe, 5′-AGAAACCTGGCTCGAACGCACAGC-3′; ADPDH-forward, 5′-CTTCCGATGAGTTCGTTTCCA-3′; ADPDH-reverse, 5′-CCACCACTGCTTCCCATTG-3′; and ADPDH-probe, 5′-CTCCTTCCAGACTTCT-3′.
Anti-HMG1cd antibody and anti-LOI1 antibody
His-tagged-AtHMG1cd (HMGR1 catalytic domain, 166–592 aa) and His-tagged-LOI1 expressed in E. coli were purified. Their antibodies were generated in rabbits. For immunoblot analysis, rabbit antiserum raised against HMG1cd was used after 100 000-fold dilution in TBS containing 10 mm MgCl2, 0.1% (v/v) Tween-20 and 3% skimmed milk. Detection was performed with the chemiluminescent system (Immobilon Western; Millipore, VersaDov; Bio-Rad) using Protein A conjugated to horseradish peroxidase (GE Healthcare, http://www1.gelifesciences.com) as a secondary antibody. Anti-HMG1cd antibodies recognize HMGR2 as well as HMGR1 (Leivar et al., 2005).
We thank Dr Hikaru Seki (Yokohama City University, Japan) for providing the binary vector pBCR112 and Dr Tadao Asami (Tokyo University, Japan) for providing clomazone. We thank Dr Takahiro Nakamura (Kyushu University, Japan) for helpful discussions concerning PPR proteins and Drs Ko Noguchi (Tokyo University, Japan) and Chihiro Watanabe (Tokyo University, Japan) for technical advice on measurement of mitochondrial respiration in plants. We also thank Mr Mitsutaka Araki for his contribution to DNA sequencing and Ms Mariko Ishikawa for her contribution to plant growth. This work was supported in part by a Grant-in-Aid for Scientific Research (no. 21024009 to T.M.) from the Ministry of Education, Culture, Sports, Science and Technology, Japan and a grant from the Japan Society for the Promotion of Science (no. 21580080 to T.M.).