The cyclobutane pyrimidine dimer (CPD), which represents a major type of DNA damage induced by ultraviolet-B (UVB) radiation, is a principal cause of UVB-induced growth inhibition in plants. CPD photolyase is the primary enzyme for repairing CPDs and is crucial for determining the sensitivity of Oryza sativa (rice) to UVB radiation. CPD photolyase is widely distributed among species ranging from eubacteria to eukaryotes, and is classified into class I or II based on its primary structure. We previously demonstrated that rice CPD photolyase (OsPHR), which belongs to class II and is encoded by a single-copy gene, is a unique nuclear/mitochondrial/chloroplast triple-targeting protein; however, the location and nature of the organellar targeting information contained within OsPHR are unknown. Here, the nuclear and mitochondrial targeting signal sequences of OsPHR were identified by systematic deletion analysis. The nuclear and mitochondrial targeting sequences are harbored within residues 487–489 and 391–401 in the C-terminal region of OsPHR (506 amino acid residues), respectively. The mitochondrial targeting signal represents a distinct topogenic sequence that differs structurally and functionally from classical N-terminal pre-sequences, and this region, in addition to its role in localization to the mitochondria, is essential for the proper functioning of the CPD photolyase. Furthermore, the mitochondrial targeting sequence, which is characteristic of class-II CPD photolyases, was acquired before the divergence of class-II CPD photolyases in eukaryotes. These results indicate that rice plants have evolved a CPD photolyase that functions in mitochondria to protect cells from the harmful effects of UVB radiation.
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.
The cyclobutane pyrimidine dimer (CPD) represents a major type of DNA damage induced by ultraviolet-B (UVB; 280–320 nm) radiation (Britt, 1996). Such DNA damage can be mutagenic or lethal by impeding replication and transcription. In fact, CPD is a principal cause of UVB-induced growth inhibition in Oryza sativa (rice) plants grown under supplementary UVB radiation (Hidema et al., 2007). Plants possess mechanisms to cope with the UVB-induced formation of CPDs, including nucleotide excision repair and photoreactivation (Britt, 1996). In nucleotide excision repair, the dimers are replaced via de novo DNA synthesis, where the undamaged complementary strand is employed as the template. Photoreactivation is mediated by the enzyme CPD photolyase, which absorbs blue/UVA radiation, and uses the energy to monomerize dimers. In higher plants, photoreactivation is the primary mechanism for repairing UVB-induced CPDs, as the rate of nucleotide excision repair is slower than that of photoreactivation (Quaite et al., 1994; Sutherland et al., 1996; Hidema et al., 1997). We previously demonstrated that CPD photolyase is a crucial factor for determining the sensitivity of rice to UVB radiation, and increasing CPD photolyase activity can significantly alleviate UVB-caused growth inhibition in rice plants (Hidema et al., 2005, 2007). Thus, CPD photolyase is an essential protein for plants grown in sunlight, including UVB radiation.
CPD photolyase is widely distributed among species, ranging from eubacteria and archaebacteria to eukaryotes, except for eutherian mammals. CPD photolyases are classified into two classes based on their primary structure, i.e. class I and class II. Class-I CPD photolyase genes have been isolated from prokaryotes and eukaryotes (fungi), and class-II CPD photolyase genes have been isolated from a wide variety of organisms, including eubacteria, archaebacteria and higher eukaryotes (Yasui et al., 1994; Kanai et al., 1997). Eukaryotic cells have at least two DNA-containing organelles, i.e. nuclei and mitochondria. Furthermore, in addition to nuclear and mitochondrial genomes, the photosynthetic eukaryotic cells contain an additional genome, which is located in the chloroplast. Yeast efficiently photoreactivates CPDs using CPD photolyase in both nuclei and mitochondria, and yeast CPD photolyase, which belongs to class I, is transported into both organelles via signal sequences present in its amino (N)-terminal region (Yasui et al., 1992). Much less is known about the subcellular localization of class-II CPD photolyase than is known about class I. Using biochemical and subcellular localization (immunogold electron microscopy) analyses, we previously demonstrated that rice CPD photolyase, which belongs to class II and is encoded by a single-copy gene and not a splice variant, is expressed and targeted to the mitochondria and chloroplasts as well as the nucleus. This protein repairs UVB-induced CPDs in all three genomes (Takahashi et al., 2011). Therefore, rice CPD photolyase is a nuclear, mitochondrial and chloroplast triple-targeting protein, but the location and nature of the organellar targeting information contained within rice CPD photolyase remain unknown. The subcellular localization predictor program targetp (http://www.cbs.dtu.dk/services/TargetP; Emanuelsson et al., 2007) failed to identify a mitochondrial targeting sequence in the deduced amino acid sequence of rice CPD photolyase, whereas a putative chloroplast-targeting signal sequence (transit peptide) was identified in its 80 N-terminal amino acids. Thus, the organellar targeting signal sequences of plant class-II CPD photolyases may differ structurally and functionally from those of yeast class-I CPD photolyase.
Here, we describe the identification of the nuclear and mitochondrial targeting signal (MTS) sequences of rice triple-targeting CPD photolyase by systematic deletion analysis. Both sequences are located in the C-terminal region of rice CPD photolyase. The MTS represents a distinct topogenic sequence that differs structurally and functionally from classical N-terminal presequences and, in addition to its role in localization to the mitochondria, this region is essential for the proper functioning of this protein. Furthermore, alignment analysis of class-I and class-II CPD photolyases revealed that the mitochondrial targeting sequence is characteristic of class-II CPD photolyases, and was acquired at an early stage of evolution, prior to the divergence of class-II CPD photolyases in eukaryotes.
Transgenic rice plants expressing a fusion protein comprising rice CPD photolyase and green fluorescent protein (GFP)
In general, DNA deletion analysis using GFP fusion proteins is a useful tool for predicting the organelle-targeting sequences of proteins (Millar et al., 2009). First, to determine whether this approach would be useful for investigating rice CPD photolyase (OsPHR) protein targeting within cells, we generated transgenic rice plants stably expressing GFP fused to the carboxy (C) terminus of OsPHR (OsPHR-GFP) under the control of the cauliflower mosaic virus (CaMV) 35S promoter. Compared with the localization of GFP using a control vector, OsPHR-GFP fluorescence was clearly detected in the nuclei and chloroplasts of rice cells (Figures 1a and S1). This result was confirmed by observing the localization of OsPHR-GFP on three-dimensional (Figure 1b) and two-dimensional (Figure S2) images. In addition, OsPHR-GFP fluorescence was obvious in cellular compartments other than the nucleus and chloroplast (Figure 1a, arrowheads; Figures 1b, S1 and S2). To confirm the fluorescence in rice cells was in mitochondria, OsPHR-GFP overexpressing rice cells were stained with MitoTracker Orange, a mitochondrial-specific stain. In this experiment, we used cells with immature chloroplasts because the intensity of fluorescence emitted by chloroplasts impedes the observation of mitochondrial fluorescence in rice cells. OsPHR-GFP fluorescence in cellular compartments other than the nucleus and chloroplasts significantly merged with the fluorescence of MitoTracker Orange (Figure 1c). Thus, we confirmed that the triple targeting protein OsPHR-GFP was targeted to the nuclei, mitochondria, and chloroplasts in rice cells. Next, we tried to identify the organelle-targeting sequences of OsPHR using the GFP fusion protein.
Transient expression assay of OsPHR-GFP in rice cells
To predict the organelle-targeting signal sequences, we constructed a transient expression vector expressing OsPHR under the control of the CaMV 35S promoter and fused to GFP at the C terminus. In general, when transient expression assays are performed, onion epidermal cells, cell suspensions, protoplasts or other systems with high transient transfection efficiency are employed. By contrast, rice leaves are seldom used as plant material for transient expression assays because they are difficult to transform and their transient transfection efficiency is low. We initially tried to introduce the expression vectors transiently into rice leaf cells using the particle bombardment method. To visualize mitochondria, cells were co-bombarded with a construct that expresses mtDsRed, a fusion protein consisting of the Arabidopsis mitochondrial F0F1ATPase γ-subunit and DsRed. When GFP alone as a control vector was introduced into rice cells, the GFP fluorescence was observed not only in cytoplasm but also in nuclei (Figure 2a). It is well known that the relatively small proteins, such as GFP, often enter the nucleus passively. The localization of GFP into nuclei would thus be caused by passive entry. Next, OsPHR-GFP vector was transiently introduced into rice cells. Consequently, GFP fluorescence was detected in all of the organelles, although the transient expression efficiency of the OsPHR-GFP vector was lower in rice cells (Figure 2b); however, the results were ambiguous, with some experiments showing nuclear and mitochondrial targeting (Figure 2b-1), nuclear and chloroplast targeting (Figure 2b-2) and targeting to other regions (Figure 2b-3), as was previously observed for the localization of dual-targeted proteins by transient expression analysis (Abdelnoor et al., 2003, 2006). These ambiguous results may have been caused by differences in the expression levels of OsPHR-GFP within each cell or by differences in the status of the cells (Carrie et al., 2009). Nonetheless, OsPHR-GFP proteins were also expressed and targeted to the nuclei, mitochondria and chloroplasts in some transiently transfected rice cells.
Identification of the nuclear localization signal (NLS) of OsPHR
Takahashi et al. (2002) reported that Cucumis sativus (cucumber) CPD photolyase [CsPHR; Genbank/EMBL/DDBJ accession number (GED number): AF053365] has a putative NLS sequence located at positions 471–475 (KKRKP) in its C terminus, as determined using the Prediction of Protein Sorting Signal and Localization Sites in Amino Acid Sequences (psort) program (Nakai and Kanehisa, 1992), whereas the predictor failed to identify the NLS of rice CPD photolyase. We compared the amino acid sequences of OsPHR and CsPHR. Sequence homology analysis revealed three conserved amino acids, KKR, at positions 487–489 in the C terminus of OsPHR (Figure 2c). To determine whether these amino acids are related to the NLS of OsPHR, we constructed a transient expression vector encoding GFP fused to the C terminus of OsPHR, in which residues K487, K488 and R489 were replaced with alanine residues (A) [OsPHR(487AAA489)-GFP], and transiently introduced this expression vector into rice leaf cells. OsPHR(487AAA489)-GFP localized to not only cytoplasm but also nuclei (Figure 2d), which was similar to the localization pattern of GFP alone (Figure 2a). Furthermore, we constructed transient expression vectors encoding β-glucuronidase (GUS) fused to the C terminus of OsPHR(487AAA489)-GFP [OsPHR(487AAA489)-GFP-GUS] or of OsPHR-GFP with single alanine substitution at K487, K488 or R489 [OsPHR(K487A)-GFP-GUS, OsPHR(K488A)-GFP-GUS or OsPHR(R489A)-GFP-GUS], to suppress passive entry into the nucleus (Grebenok et al., 1997). These vectors were transiently introduced into Allium cepa (onion) epidermal cells by particle bombardment because onion cells are useful for transient expression analysis, and their transient expression efficiency is high. Passive entry into the nucleus of expression control vector GFP-GUS (Figure 3) was significantly suppressed, compared with localization of GFP alone (Figure 2a). OsPHR(487AAA489)-GFP-GUS was primarily localized to the cytoplasm, whereas OsPHR-GFP-GUS was localized exclusively to the nuclei (Figure 3). In addition, the fluorescence of each alanine-substituted OsPHR-GFP-GUS [OsPHR(K487A)-GFP-GUS, OsPHR(K488A)]-GFP-GUS or OsPHR(R489A)-GFP-GUS] was suppressed in the nucleus, and the fluorescence pattern was nearly indistinguishable from that of GFP-GUS alone (Figure 3). Thus, these results strongly suggest that three amino acids, KKR, located at positions 487–489 in the C terminus of OsPHR, are related to the NLS of OsPHR, and each amino acid residue at K487, K488 or R489 of OsPHR plays an important role in its import into the nucleus.
Identification of the MTS of OsPHR
Dual targeting to chloroplasts and mitochondria has been described for several proteins. Most of these proteins have an organelle-targeting sequence located at their N termini, although there are several examples of proteins that have internal or C-terminal targeting signals (Peeters and Small, 2001; Hirakawa et al., 2009). Yeast photolyase is transported into mitochondria by signal sequences present in the N-terminal region (Yasui et al., 1992); however, an MTS was not identified in the amino acid sequence of OsPHR using the prediction program targetp (Emanuelsson et al., 2007). To map the regions of OsPHR that are associated with chloroplast or mitochondrial targeting, we constructed expression vectors encoding protein chimeras of GFP and a fragment of the N terminus (amino acids 1–444, N1–444-OsPHR-GFP; amino acids 1–228, N1–228-OsPHR-GFP) or C terminus (amino acids 385–506, C385–506-PHR-GFP) of OsPHR (Figure 4a, numbers 2–4). First, the expression vectors were introduced into onion epidermal cells. The OsPHR-GFP as a control was detected in nuclei and mitochondria, although this was not clearly detected in plastids (Figure 4b,c, no. 1). N1–444-OsPHR-GFP and C385–506-OsPHR-GFP were detected in mitochondria (Figure 4b,c, nos 2 and 4), whereas N1–228-OsPHR-GFP was not clearly detected in plastids or mitochondria (Figure 4b,c, no. 3). These results indicate that the C terminus of rice CPD photolyase (amino acids 385–444) contains a functional targeting signal for transport into mitochondria. On the other hand, despite the probability prediction of a chloroplast-targeting signal in the N-terminus of rice CPD photolyase, we found no evidence of a chloroplast-targeting signal in this region. Therefore, we next focused on the MTS of OsPHR.
To narrow down the region of OsPHR that contains the MTS, we constructed transient expression vectors expressing GFP fused to a fragment of the C-terminal region of OsPHR (amino acids 385–444), and co-transformed these constructs with mtDsRed into onion epidermal cells. C385–394-OsPHR-GFP and C395–403-OsPHR-GFP were not detected in mitochondria (Figure 4b,c, nos 7 and 8), which was also the case for GFP alone (Figure 4b,c, no. 10). By contrast, GFP fusion proteins fused to partial sequences that included 11 amino acids 391–401 (C385–437-OsPHR-GFP, C385–403-OsPHR-GFP and C391–401-OsPHR-GFP) were detected in mitochondria, suggesting that these amino acid residues are related to the MTS of OsPHR (Figure 4b,c, nos 5, 6 and 9). Furthermore, we investigated the localization of GFP fusion proteins fused to partial sequences of deleted amino acid residues (K401, A400 and W399) in the 11 amino acids at positions 391–401 (C391–400-OsPHR-GFP, C391–399-OsPHR-GFP and C391–398-OsPHR-GFP). The GFP fluorescence of these GFP fusion proteins was not detected in mitochondria (Figure S3). These results thus suggest that amino acid residues at 391–401 of OsPHR are minimally needed for transport into mitochondria.
To determine whether these amino acid residues of OsPHR comprise a functional targeting signal for transport into mitochondria in rice cells, we introduced transient expression vectors harboring GFP fused to partial sequences with or without these 11 amino acid residues into rice leaf cells. As described above, although the transient expression efficiency of the OsPHR-GFP vector was lower in rice cells, and the localization of the OsPHR-GFP was non-constant (Figure 2b), C385–506-PHR-GFP and C391–401-OsPHR-GFP were significantly localized to mitochondria (Figure 5b, nos 4 and 9). On the other hand, N1–228-OsPHR-GFP was not clearly detected in chloroplasts or mitochondria (Figure 5b, no. 3), similar to the results obtained in the experiments using onion epidermal cells (Figure 4b,c, no. 3). In addition, as these amino acid residues are within the C terminus of OsPHR, we constructed a transient expression vector expressing only these 11 amino acid residues (amino acids 391–401) fused to the C terminus of GFP (GFP-C391–401-OsPHR) and transiently introduced this construct into rice cells. GFP-C391–401-OsPHR also localized to mitochondria (Figure 5b, no. 11), similar to the localization of C391–401-OsPHR-GFP (Figure 5b, no. 9). Furthermore, a transient expression vector expressing OsPHR lacking these 11 amino acid residues and fused to the N-terminus of GFP (Δ391–401-OsPHR-GFP) was transiently introduced into rice cells. Δ391–401-OsPHR-GFP fluorescence was obvious in nuclei and chloroplasts, but not in mitochondria (Figure 5b, no. 12).
Most nucleus-encoded mitochondrial proteins target to mitochondria using an N-terminal transit peptide (Neupert, 1997). The transit peptide is not transported into the organelle, but is cleaved by a peptidase. Therefore, if OsPHR uses an N-terminal sequence for transit into mitochondria, OsPHR localized in mitochondria should lack an N-terminal sequence. To determine whether OsPHR localized in the mitochondria has N-terminal amino acid residues, we prepared a mitochondria-enriched fraction from leaves of transgenic rice plant overexpressing rice CPD photolyase (Hidema et al., 2007), and analyzed the crude extract by western blot analysis using an anti-rice N-terminal OsPHR antiserum, which was raised against an N-terminal peptide containing residues 2–16. To confirm whether the mitochondria was enriched in the mitochondria extract prepared in this experiment, the concentrations of anti-voltage-dependent anion-selective channel protein 1 (VDAC1), as a mitochondrial marker, and of light-harvesting chlorophyll a/b protein of PSII (LHCII), as a chloroplast marker, in the mitochondria-enriched fraction were compared with those in the crude extract fraction and in the crude mitochondria fraction (Figure 6). The crude extract fraction was prepared from a homogenized solution of rice leaves (lane 1), the crude mitochondria fraction was prepared from crude extract before Percoll fractionation (lane 2) and the mitochondria-enriched fraction was prepared from isolated mitochondria extract after Percoll fractionation (lane 3). We prepared each extract solution subjected to SDS-PAGE with approximately similar concentrations of CPD photolyase. The concentration of VDAC1 in the mitochondria-enriched fraction was significantly higher than that of the crude extract fraction or crude mitochondria fraction. In addition, large quantities of LHCII in the crude mitochondria fraction, in which large quantities of thylakoid membrane were found, were removed by Percoll fractionation. Thus, the mitochondria were indeed enriched in the ‘mitochondria-enriched fraction’. When we performed western blot analysis using twice affinity-purified anti-rice N-terminal OsPHR antiserum, the 54-kDa polypeptide band of rice CPD photolyase was detected in the mitochondria-enriched fraction, and did not show a mobility shift with respect to the OsPHR band in the crude extract (Figure 6, lane 3). These results strongly suggest that OsPHR localized to mitochondria was not subjected to proteolytic processing, and that the predicted amino acids at positions 391–401 of OsPHR (MHGFMRMYWAK) are a functional targeting signal for its transport into the mitochondria of rice cells.
The predicted MTS of OsPHR is highly conserved in class-II CPD photolyases of various species
The predicted amino acid residues of OsPHR form a novel MTS sequence. To determine whether the predicted MTS of OsPHR is conserved in CPD photolyases among various species, we aligned the deduced amino acid sequences of several class-I and class-II CPD photolyases (Figures 7 and S4). The predicted MTS sequence of OsPHR is highly conserved in class-II CPD photolyases, except for that of Myxococcus xanthus, which belongs to eubacteria, whereas it is not conserved in class-I CPD photolyases. Therefore, the predicted MTS sequence identified in this study is characteristic of class-II CPD photolyases, but not of class-I CPD photolyases. In class-II CPD photolyases, the MTS sequence is located in the highly conserved helical domain, which plays important roles in the functioning of the enzyme (Figure S4; Todo, 1999; Hitomi et al., 2012). This alignment analysis suggests that class-II CPD photolyases might be targeted to mitochondria using these amino acid residues, which are essential for the proper functioning of the enzymes.
In this study, we identified the functional NLS and MTS of the nuclear, mitochondrial and chloroplast triple-targeting protein CPD photolyase of rice. The NLS of OsPHR comprises basic amino acid residues (amino acid residues 487–489; KKR; Figures 2d and 3), and is located at the end of the C terminus, similar to that of cucumber (Takahashi et al., 2002). In addition, each amino acid residue at positions 487–489 in the NLS of OsPHR plays an important role in its import into the nucleus (Figure 3). The predicted MTS is also located in the C-terminal region (amino acids 391–401, MHGFMRMYWAK), although targeting and translocation of most nucleus-encoded mitochondrial proteins depends on N-terminal extensions called mitochondrial targeting sequences or presequences (Neupert, 1997; Pfanner et al., 1997). The MTS of OsPHR represents a distinct topogenic sequence that differs structurally and functionally from classical N-terminal presequences. The sequence is located in the highly conserved helical domain of OsPHR. FAD, which acts as the active site co-factor, binds in this helical domain, and a region of this domain is considered to be a putative CPD-binding site (Kanai et al., 1997; Todo, 1999). Thus, in addition to its role in mitochondrial localization, this region is thought to be essential for the proper functioning of CPD photolyases. Indeed, the targeting signal sequence is both necessary and sufficient for import of OsPHR into mitochondria, and it represents a general topogenic signal because it can direct a non-mitochondrial protein, such as GFP, to mitochondria with high efficiency (Figure 5b, nos 9 and 11). We carried out a blast search using the minimum mitochondria import motif of OsPHR (MHGFMRMYWAK) as a query. The database analysis showed that no other proteins, except for class-II CPD photolyase, with an MTS highly similar to that of OsPHR exist. Thus, the MTS of OsPHR is a unique sequence with interesting properties.
Several properties distinguish the OsPHR MTS from classical mitochondrial pre-sequences. Such pre-sequences typically consist of 15–40 amino acid residues and are enriched with positively charged and hydroxylated residues (Schatz and Dobberstein, 1996; Neupert, 1997; Pfanner et al., 1997). The ability of most pre-sequence peptides to form amphipathic α-helices is thought to be important for their recognition by the translocation machinery in the outer (TOM complex) and inner (TIM complex) mitochondrial membranes (von-Heijne, 1986; Roise and Schatz, 1988). The targeting signal of OsPHR is highly hydrophilic, containing positively charged residues. Furthermore, examining the crystal structure of OsPHR revealed that the region containing the MTS sequence (residues 391–401) forms an α-helical structure (Hitomi et al., 2012). Furthermore, this sequence was identified as a putative mitochondrial targeting sequence using ipsort (http://ipsort.hgc.jp; Bannai et al., 2002), a subcellular localization prediction program. Thus, the chemical characteristics of the OsPHR MTS differ markedly from those of classical N-terminal pre-sequences.
Approximately 30% of mitochondrial proteins lack typical N-terminal pre-sequences (Diekert et al., 1999). In particular, most outer mitochondrial membrane proteins and some proteins of the intermembrane space and the inner mitochondrial membrane lack such signals (Hurt et al., 1985; McBride et al., 1992). There are several examples of proteins that have internal or C-terminal targeting signals (Peeters and Small, 2001). For example, mitochondria/nuclei dual-targeted human APE1 (apurinic/apyrimidinic endonuclease 1) protein, which plays a central role in cellular responses to oxidative stress, and functions as a DNA repair enzyme in DNA base excision repair, is targeted to mitochondria by 30 C-terminal amino acids (Li et al., 2010). Furthermore, the mitochondrial cytochrome heme lyases, which are localized in the mitochondrial intermembrane space, and are essential for the covalent attachment of heme to c-type cytochrome, also lack N-terminal targeting information (Lill et al., 1992). Diekert et al. (1999) reported that the MTS of heme lyase is located in the C-terminal region, in the third quarter of the heme lyase molecule, and comprises approximately 60 amino acid residues. This MTS is hydrophilic (30% charged residues), and secondary structure predictions for this sequence assign a helical structure. Thus, the chemical properties of the heme lyase targeting signal are also distinct from those of classical mitochondrial pre-sequences. Interestingly, the sequence contains two highly conserved motifs that are characteristic signatures of all known mitochondrial heme lyases. Therefore, the targeting signal sequences of heme lyase are essential for the function of this protein, in addition to their role in sorting to mitochondria, similar to the targeting sequences of OsPHR. Why is this sequence, which is essential for enzymatic function, used for mitochondrial targeting? Although these findings are interesting, their biological significance is unclear.
The predicted MTS of OsPHR, which significantly differs from that of yeast class-I CPD photolyase, is unique and characteristic of class-II CPD photolyases (Figures 7 and S4). To date, class-II CPD photolyase genes have been isolated from a wide variety of organisms, including eubacteria, archaebacteria and higher eukaryotes. Class-II CPD photolyases show relatively high mutual homology (25–80%), whereas much lower homology (10–17%) is found between class-I and class-II CPD photolyases (Yasui et al., 1994; Kanai et al., 1997). To help elucidate the evolutionary process leading to the functional divergence of the predicted MTS of OsPHR, an unrooted phylogenetic tree was constructed using the neighbor-joining method, based on 21 and five sets of class-II and class-I CPD photolyase sequences, respectively (Figure 7). As described by Yasui et al. (1992), the unrooted phylogenetic tree is composed of two clusters: cluster 1 only consists of class-II CPD photolyases, and cluster 2 only consists of class-I CPD photolyases. These two clusters are connected by the longest branch in the tree. When compared with the relationship between the evolutionary processes of class-II CPD photolyase and the sequence of the homologous region of the predicted MTS of OsPHR in the phylogenic tree, changes in the amino acid residues corresponded to the evolutionary processes of class-II CPD photolyase. The MTS sequence of OsPHR is highly conserved with those of eukaryotic class-II CPD photolyases. In addition, interestingly, the MTS of OsPHR is also highly conserved with those of class-II CPD photolyase of Methanobacterium and Clostridium, which has no mitochondria. Therefore, perhaps the root of the MTS of class-II CPD photolyases differs from that of class-I CPD photolyases, and the MTS of OsPHR may have been acquired before the divergence of class-II CPD photolyases in eukaryotes.
In this study, we identified the NLS and the MTS of the nuclear, mitochondrial and chloroplast triple-targeting protein CPD photolyase in rice; however, the mechanism by which this protein is transported into mitochondria is unknown and it is unclear whether the TOM or TIM complexes recognize the MTS sequence of OsPHR. Furthermore, the chloroplast-targeting signal of OsPHR has not been definitively identified. The fluorescence of neither N1–444-OsPHR-GFP nor N1–228-OsPHR-GFP, both of which contain a putative chloroplast-targeting sequence predicted by targetp, was detected in plastid or chloroplast (Figure 4b,c, nos 2 and 3, Figure 5b, nos 3). In addition, no fluorescence of GFP fusion proteins that contain a fragment of OsPHR were detected in chloroplasts or plastids, although OsPHR-GFP that contains full-length OsPHR or Δ391–401-OsPHR-GFP was localized in the chloroplasts of rice leaves (Figures 1, 2 and 5). Thus, higher-order structure of OsPHR might be required for transport into chloroplasts. In this situation, single alanine substitution analysis could be useful for identifying the chloroplast-targeting sequence. The study should be conducted in the future. Future studies may help reveal the strategy that plants grown in natural sunlight have used to adapt to UVB exposure. Nonetheless, the results of this study indicate that rice plants have evolved a CPD photolyase that functions in mitochondria to protect cells from the harmful effects of UVB radiation.
Construction of vectors and transgenic rice plant production
The cDNA clone of Sasanishiki CPD photolyase (GED number AB096003; Hirouchi et al., 2003) was amplified by PCR using the primers PHR-F1 (5′-CCGCCGACCTCAGTGAGCCCACCAAGAA-3′) and PHR-R1 (5′-TGCTGAGACTTGGAAAGC-3′). The PCR product was ligated to the 5′ end of the GFP gene of CaMV-35S-sGFP (S65T)-nos 3′ (pCaMV-35S-OsPHR-GFP). OsPHR-GFP was amplified by PCR using the primers PHR-GFP-Fw (5′-ATGCCGCCGACCTCAGTGAGCCCA-3′) and PHR-GFP-Rv (5′-TTACTTGTACAGCTCGTCCAT-3′). OsPHR-GFP was subcloned into the binary vector pPZP2Ha3 (Fuse et al., 2001) in the sense orientation, and expressed under the control of the CaMV 35S promoter. Agrobacterium-mediated transformation was performed using the rice cultivar ‘Nipponbare’ (Oryza sativa L. ssp. japonica) by Inplanta Innovations Inc. (http://www.inplanta.jp/en.html). The control transgenic plant, rice cultivar ‘Nipponbare’ expressing sGFP (S65T) under the control of the CaMV 35S promoter, was provided by Dr. S. Toki (National Institute of Agrobiological Sciences, Japan; Toki et al., 2006).
Plant materials and growth conditions
Transgenic rice plants were grown under visible light in a growth cabinet (Koito Ind. Ltd. Co., http://www.koito-ind.co.jp; 12-h photoperiod, day/night temperatures of 27 and 17°C, respectively), as described previously (Hidema et al., 2007). Rice cultivar ‘Sasanishiki’ (O. sativa L. ssp. japonica) was used in the transient expression assay. The seeds were soaked in water at 28°C for 5 days under visible white light (approximately 35 μmol m−2 sec−1).
The thin layers were peeled from young leaves of transgenic rice plants using a needle and forceps. To detect mitochondria, the thin layers of leaves were incubated with H2O containing 0.5 μm MitoTracker Orange (Life Technologies, http://www.lifetechnologies.com) for 15 min at 25°C, and then washed twice with H2O. To detect nuclei, the thin layers of leaves were incubated with H2O containing Hoechst 33342 (NucBlue Live Cell Stain; Life Technologies) for 20 min at 25°C, and then washed twice with H2O. Fluorescence was observed using a confocal laser scanning microscope (LSM710; Zeiss, http://www.zeiss.com).
Generation of expression constructs for transient expression analysis
The vector pCaMV-35S-OsPHR-GFP described above was used to express chimeric proteins comprising GFP and full-length rice CPD photolyase (OsPHR-GFP). To generate the construct, which encodes chimeric proteins comprising GUS and OsPHR-GFP (OsPHR-GFP-GUS), the GUS gene was amplified by PCR with pBI121 vector (GED number AF485783) as a template using the primers GUS-F1 (5′-GGACGAGCTGTACAAGATGTTACGTCCTGTAGAAACCCC-3′) and GUS-Rv1 (5′-GTTTGAACGATCTGCAGTCATTGTTTGCCTCCCTGCTGC-3′). The GFP expression vectors pCaMV-35S-GFP and pCaMV-35S-OsPHR-GFP were amplified by PCR using the primers VEC-F1 (5′-CTGCAGATCGTTCAAACATTTGGC-3′) and VEC-R1 (5′-CTTGTACAGCTCGTCCATGCCG-3′). PCR-amplified GUS gene and the GFP expression vectors were connected using the In-Fusion HD Cloning Kit (Clontech Laboratories, http://www.clontech.com). To generate the N1–444-OsPHR-GFP, N1–228-OsPHR-GFP and C385–506-OsPHR-GFP constructs, the cDNA clone of Sasanishiki CPD photolyase was amplified by PCR using the primers shown in Table S1. The PCR product was ligated to the 5′ end of the GFP gene of CaMV-35S-sGFP (S65T)-nos 3′. To make the other substituted or truncated constructs, an inverse PCR-based site-directed mutagenesis reaction was performed using a KOD-Plus-Mutagenesis Kit (Toyobo Co. Ltd, http://www.toyobo-global.com), according to the manufacturer's instructions. In brief, the plasmid shown in Table S1 was subjected to inverse PCR using the specific primers shown in Table S1. The template plasmid DNA was digested with DpnI. The PCR product was self-ligated using T4 polynucleotide kinase and ligase. An expression construct encoding mtDsRed, a fusion protein consisting of the Arabidopsis mitochondrial F0F1ATPase γ-subunit and DsRed (Plant Research International, http://www.pri.wur.nl) was used to visualize mitochondria.
Transient expression and detection of fluorescence
Thin layers were peeled from the first leaves of rice plants using a needle and forceps. The indicated constructs were introduced into rice leaf cells or Allium cepa (onion) epidermal cells using a PDS-1000/He particle delivery system (Bio-Rad, http://www.bio-rad.com). To detect nuclei, the epidermal cells of onion were incubated with H2O containing 10 μg ml−1 4′,6-diamidino-2-phenylindole (DAPI) for 30 min at room temperature, and then washed twice with H2O. The cells were observed using a fluorescence microscope (Axio Imager D1; Zeiss) or a confocal laser scanning microscope (LSM710; Zeiss; or FluoView FV1000; Olympus Co., http://www.olympus-global.com).
Isolation of mitochondria and western blot analysis
Mitochondria were isolated from fresh leaves of transgenic rice overexpressing rice CPD photolyase (S-C plant). The rice CPD photolyase transcript level in the S-C plant was 149-fold higher than in the wild-type (WT) rice plant (Sasanishiki rice plant; Hidema et al., 2007). In brief, 10-day-old seedlings of S-C plants were cut into pieces of 3–5 mm in length and homogenized with an UltraTurrax using 2 × 10-sec bursts in homogenization solution [0.3 m mannitol, 5 mm EDTA, 30 mm 3-(N-morpholino)propanesulphonic acid (MOPS), 1% (w/v) BSA, 5 mm DTT, 1% polyvinylpyrrolidone (PVP), pH 7.3]. The homogenate (crude extract) was strained through nylon cloth (120 μm) and then centrifuged at 1000 g for 5 min at 4°C to remove debris, including the nuclear fraction. Following centrifugation of the supernatant at 20 000 g for 25 min at 4°C, the pellet (crude mitochondria fraction) was resuspended in wash solution (0.3 m mannitol, 1 mm EDTA, 10 mm MOPS, pH 7.2). Crude mitochondria were layered on top of a Percoll step gradient consisting of 20, 28 and 40% Percoll in Percoll buffer (0.4 m mannitol, 5 mm MOPS, 1 mm EDTA, 0.4% PVP, pH 7.2) and centrifuged at 40 000 g for 30 min at 4°C. Mitochondria were recovered from the 28–40% Percoll interface, diluted >10 times in wash solution and then pelleted by centrifugation at 20 000 g for 20 min. The pellet was resuspended in wash solution (mitochondria-enriched fraction). SDS-PAGE was performed using 7 or 15% (w/v) SDS-polyacrylamide gels. The proteins were separated by SDS-PAGE, transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad), and probed with twice affinity-purified antiserum raised against a peptide containing the N-terminal sequence of OsPHR (PPTSVSPPRTAPGPA; amino acids 2–16 of OsPHR; Scrum Inc., http://www.scrum-net.co.jp), VDAC1 antibody (Agrisera, http://www.agrisera.com) to label mitochondria and LHCII antiserum to label chloroplasts (Hidema et al., 1992). The immune complex was detected using alkaline phosphatase-conjugated anti-rabbit IgG (Sigma-Aldrich, http://www.sigmaaldrich.com) and developed using premixed 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium solution (Bio-Rad).
Construction of a phylogenic tree and sequence alignment of CPD photolyase sequences
Unrooted neighbor-joining phylogenetic tree construction (Saitou and Nei, 1987) was performed using clustalx 2.0.12 (Thompson et al., 1997). Class-II CPD photolyases: O. sativa (Oryza; GED number AB096003), Zea mays [Zea; NCBI reference sequence number (NCBI number) NM_001137108], Triticum aestivum (Triticum; GED number AK330529), Arabidopsis thaliana (Arabidopsis; NCBI number CAA67683), Solanum lycopersicum (Solanum; NCBI number XM_004245655), Cucumis sativus (Cucumis; GED number AF053365), Glycine max (Glycine; NCBI number NM_001251781), Lotus japonicus (Lotus; GED number AK338540), Spinacia oleracea (Spinach; GED number AY267198), Physcomitrella patens (Physcomitrella; NCBI number XP_001764990), Selaginella moellendorffii (Selaginellaceae; NCBI number XM_002972541), Marchantia polymorpha (Marchantia; GED number AB853889), Pityrogramma austroamericana (Pityrogramma; GED number AY271662), Volvox carteri f. nagariensis (Volvox; NCBI number XM_002953258), Chlamydomonas reinhardtii (Chlamydomonas; NCBI number AAD39433), Drosophila melanogaster (Drosophila; UniProtKB/Swiss-Prot: Q24443), Oryzias latipes (Oryzias; NCBI number BAA05043), Potorous tridactylus (Potorous; NCBI number BAA05041), Clostridium sp. CAG:967 (Clostridium; NCBI number WP_021945639), Methanobacterium thermoautotrophicum (Methanobacterium; NCBI number NP_276041) and Myxococcus xanthus (Myxococcus; GED number AAC43723). Class-I CPD photolyase: Saccharomyces cerevisiae (Saccharomyces; NCBI number NP_015031), Escherichia coli (Escherichia; NCBI number NP_415236), Neurospora crassa (Neurospora; NCBI number XP_964834), Bacillus cereus (Bacillus; NCBI number YP_006597942) and Anacystis nidulans (Anacystis; NCBI number YP_172102).
We thank Dr K. Ishizaki and T. Kohchi for providing the CPD photolyase nucleotide sequence of M. polymorpha, Dr S. Toki for providing GFP-expressing transgenic rice, Dr Y. Niwa for providing the CaMV-35S-sGFP (S65T)-nos 3′ vector and Dr H. Takahashi for providing the mtDsRed vectors. This work was supported by JSPS KAKENHI (grant nos 23120502, 24241028, 25120702 and 24620001) and by JST, PRESTO.