Cloning and characterization of chloroplast and cytosolic forms of cyclophilin from Arabidopsis thaliana.

Cyclophilin (CyP), a protein with peptidyl-prolyl cis-trans isomerase (rotamase) activity, is the specific cellular target of cyclosporin A. We have isolated cDNA clones of two genes (designated ROC1 and ROC4) encoding CyP homologs from Arabidopsis thaliana (L.). The protein products of these genes are distinct from a previously identified Arabidopsis CyP. ROC1 is expressed in all tested plant organs and encodes a protein which is highly similar to previously described cytosolic CyP isoforms of other plants. In contrast, ROC4 is expressed only in photosynthetic organs and encodes a protein which includes an amino-terminal extension with properties of known chloroplast transit peptides. In vitro import experiments using the putative precursor protein to ROC4 showed that the protein is imported into chloroplasts where it is processed to the predicted mature size. Rotamase assays and immunoblot analysis of subcellular fractions indicate the presence of a CyP isoform in the stroma of chloroplasts but not in the thylakoid membranes or thylakoid lumen. Together, these data show that ROC4 is a novel CyP isoform which is located in the stroma of chloroplasts. In vitro chloroplast import of precursors of other chloroplast proteins was unaffected by concentrations of cyclosporin A which completely inhibit rotamase activity of chloroplast stromal CyP. Thus, this activity is not essential for protein import into chloroplasts.

Cyclophilin (CyP)' is an abundant, highly conserved protein present in virtually all organisms (Koletsky et al., 1986). CyP was first identified as a high affinity binding protein for the immunosuppressive drug cyclosporin A (CsA) (Handschumacher et al., 1984). It has now been shown that the CyP.CsA complex suppresses the immune response through inhibition of :k This research was supported by National Science Foundation grants NSF90-58284 (to C. S. G.) and DCB88-17371 (to S. M. T.1, The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore he hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequencels) reported in this paper has been submitted to the GenBankTMfEMBL Data Bank with accession numberlsi L14844 (ROCI) and L14845 (ROC4).
5 Supported by the Molecular Biology of the Plant Cell Training Program.  the phosphoprotein phosphatase calcineurin (reviewed in Schreiber (1992) and Walsh et al. (1992)). A similar mechanism is responsible for CsAinhibition of recovery from a-factor arrest in yeast (Foor et al., 1992) demonstrating that this property of CyP is conserved in widely diverged organisms. Other naturally occurring ligands of CyP include a protein of unknown function identified in a mouse bone marrow-derived stromal cell line (Friedman et al., 1993), and the GAG protein of human immunodeficiency virus (Luban et al., 1993). However, the functional significance of these additional interactions remains unknown.
CyP has also been shown to be a peptidyl-prolyl cis-trans isomerase ("rotamase") whose activity can be inhibited by CsA (Fischer et al., 1989;Takahashi et al., 1989). In uitro folding experiments have shown that isomerization around Xaa-Pro bonds is one of the slow, rate-limiting steps in the folding of many proteins (Brandts et al., 1975;Schmid and Baldwin, 1978;Langet al., 1987) and that this process can be accelerated by the rotamase activity of CyP (Bachinger, 1987;Lang et al., 1987;Davis et al.. 1989;Fischer et al., 1989). In vivo experiments support a physiological role of CyP rotamases in protein folding. Treatment of cultured fibroblasts with CsA slows the formation of collagen triple helix (Steinmann et al., 1991). Studies in transgenic flies show that the Drosophila NinaA protein (a membrane bound CyP homolog) is required for the folding of a subset of rhodopsin isoforms (Stamnes et al., 1991). CsA also slows an early step in transferrin folding in human hepatoma cells, and high concentrations (10 p~) block exit of this protein from the rough endoplasmic reticulum, suggesting that at this high concentration CsA inhibits correct folding of transferrin (Lodish and Kong, 1991).
Studies in higher plants have revealed the presence of cytosolic CyPs (Gasser et al., 1990;Marivet et al., 1992). Recently, Breiman et al. (1992) reported evidence of the presence of CsAsensitive rotamase activity in plant subcellular fractions enriched in components of chloroplasts and mitochondria. However, the isolation of neither the corresponding proteins nor their genes has yet been reported. Because of the numerous advantages of Arabidopsis thaliana (L.) as an experimental organism (Meyerowitz, 1989) we have chosen to focus our efforts on the CyP genes of this plant. Here we describe the isolation of cDNA clones deriving from two different genes encoding cytosolic and chloroplast stromal forms of CyP. We additionally provide evidence that the rotamase activity of the stromal form is not required for efficient import of proteins into chloroplasts.
RNA Isolation and Characterization-Total RNA isolation and Northern blotting experiments were performed as described by Gasser et al. (1989).
Isolation of Arabidopsis CyP Clones-Approximately 50,000 phage from a AZAP (Stratagene) cDNAlibrary (a gift of J. Callis) of leaves from 3-week-old Arabidopsis (ecotype Columbia) were screened for cDNA clones encoding CyP. The screening was performed a t low stringency (final wash = 2 x SSPE, 0.1% SDS a t 55 "C; 1 x SSPE = 150 m M NaC1, 10 m M NaH,PO,, 1 mM EDTA, pH 7.4) using the coding region of the Brassica napus CyP cDNA (Gasser et al., 1990) which had been szPlabeled by random oligonucleotide priming (Feinberg and Vogelstein, 1983). Standard plaque hybridization procedures were used (Davis et al., 1980). Plasmid subclones were made from the isolated phage by coinfection with the helper phage M13K07 (Vieira and Messing, 1987) according to instructions supplied with AZAP by the manufacturer. The resulting plasmids carry the inserts in the Eco-RI site of the vector pBluescript SK(-) (Stratagene). Three independent clones containing cDNA sequences encoding ROCl were isolated. The clone with the longest insert, pIC1, was used for all subsequent experiments.
The ROCl coding sequence was used to screen plates representing 2.4 x lo5 plaques from a n independently amplified aliquot of the same Arabidopsis leaf cDNAlibrary (see above) a t high stringency (final wash = 0.3 x SSPE, 0.1% SDS at 65 "C). Twenty-seven independent clones were isolated and all contained the same size insert encoding ROC4. Plasmid subclones were prepared as described above, and one of the clones, designated pIC43, was selected for use in further experiments.
DNA Manipulation and Sequencing-Construction of expression vectors and subcloning for sequencing were done by standard techniques (Crouse et al., 1983). Single-stranded templates were produced from plasmid subclones (Vieira and Messing, 1987) and sequenced by the dideoxynucleotide-termination method (Sanger et al., 1977) with the Sequenase kit (U. S. Biochemical Corp.). Both strands of each cDNA were completely sequenced.
Preparation of Total Plant Extracts and Chloroplast Fractionation-Whole plant extracts were prepared from 26-day-old Arabidopsis ecotype Landsberg erecta and 10-day-old pea seedlings. Plants were ground to powder under liquid nitrogen with a mortar and pestle, and extraction buffer (50 m M Tris.HC1, pH 7.8, 20 p~ leupeptin, 1 p~ pepstatin, 1 m M phenylmethylsulfonyl fluoride) was added to the pulverized material in a ratio of 2 mug of fresh weight. Pea chloroplasts, chloroplast stroma, thylakoid membranes, and thylakoid lumen were purified as described (Ettinger and Theg, 1991).
Rotamase Activity Assay-Rotamase activity was assayed as described by Liu et al. (19901, except that N-succinyl-Ala-Ala-Pro-Phe-pnitroanilide was initially dissolved in 100% methanol, and the assays were performed at 5 "C for 100 s. The assays were initiated by addition of u-chymotrypsin (10 pg/ml final concentration), and cleavage of Nsuccinyl-Ala-Ala-Pro-Phe-p-nitroanilide was monitored at 390 nm with a Hewlett Packard 8452A spectrophotometer. First-order rate constants (k,,,,,) were calculated starting a t 20 s using the first-order curve fit of Enzfitter software (Biosoft, Cambridge, UK). For CsAinhibition experiments, CsA was added prior to reaction initiation. Production of Recombinant Arabidopsis CyP Proteins-The EcoRI fragment containing the ROC4 coding sequence was transferred from pIC43 into the EcoRI site of pBluescript Sk(+) (Stratagene), resulting in pVL7. A fortuitous NcoI site at the putative junction between the leader sequence and mature ROC4 protein allowed the isolation of the coding sequence of the mature protein on an NcoI-BamHI fragment (using the BamHI site in the multi-linker of the cloning vector). This NcoI-BamHI fragment was inserted into pET8c (Studier et al., 1990) using these same sites. The resulting plasmid, pVL8, was introduced into Escherichia coli BL21 (DE3) pLysS (Studier et al., 1990) for production of recombinant ROC4 as described (Gasser et al., 1990). Insoluble inclusion bodies containing ROC4 were pelleted by centrifugation and washed with 20 m M Tris.HC1 (pH 7.8), 0.5% Triton X-100.
An NdeI site was engineered at the translation start site of the ROCl coding sequence by oligonucleotide mediated mutagenesis (Kunkel, 1985), using the oligonucleotide CGTTTCAAACAACATATGGCGTTCC.
This plasmid, pRJB3, was introduced into E. coli XA90 ) for production of recombinant ROC1. The procedure for expression and purification of ROCl was essentially as described , except that the proteins that precipitated in the 60-80% saturated ammonium sulfate solution were used in subsequent steps. This scheme yielded essentially pure ROCl with minor contaminants.
Chlorophyll and Protein Assays-Chlorophyll was assayed according to Arnon (1949). Protein assays were determined using the bicinchonic acid method (Smith et al., 1985).

SDS-Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel
electrophoresis (SDS-PAGE) was performed using 15% acrylamide gels and the Laemmli (1970) buffer system. Antibody Production and Immunoblot Analysis-For production of ROCl antibodies, minor contaminants in purified ROCl (see above) were removed by preparative SDS-PAGE as described by Harlow and Lane (1988). Approximately 150 pg of ROCl were injected intradermally with complete Freund's adjuvant into a male New Zealand White rabbit. Boosts were performed in incomplete Freund's adjuvant every 3-10 weeks.
Immunoblots were prepared as described by Beers et al. (19921, except that blots were not autoclaved. Incubations were performed for 2 h with a 1:5000 dilution of anti-ROC1 serum and for 1 h with a 1:lOOO dilution of goat anti-rabbit IgG alkaline-phosphatase conjugate (Kerkegaard & Perry Laboratories Inc.). Blots were typically developed for 5-30 min (Harlow and Lane, 1988).
In Vitro Protein Synthesis and Chloroplast Import-Plasmids containing either a cDNA clone encoding the putative precursor of ROC4 (pre-ROC4), or the precursor of the small subunit of ribulose-1,5-bisphosphate carboxylase (pre-rbcS) were CsCl purified and linearized by restriction endonuclease digestion at sites 3' of the coding regions. Transcriptions from the linearized plasmids were performed using T3 (Stratagene Cloning Systems) or SP6 (Promega) RNA polymerase as described by Titus (1991). Precursor proteins were synthesized in the presence of L3H1leucine (Du Pont) in a wheat germ cell-free lysate (Cline et al., 1985). Chloroplast import reactions were performed for 20 min (unless otherwise indicated) at 25 "C as described by Ettinger and Theg (1991). except that 2 x lo5 dpm of each radiolabeled precursor protein was used per 60 pl of reaction. When no further treatments were required following import, chloroplasts were repurified by spinning through silicone oil into 1.5 M perchloric acid (Theget al., 19891. In some cases, after import and prior to repurification, the chloroplasts were treated with 200 pg/ml proteinase K (Boehringer Mannheim) for 25 min on ice, followed by inhibition of the protease reaction with 1 m M phenylmethylsulfonyl fluoride. Samples which were treated with Triton X-100 were first washed once with import buffer (IB = 50 m M K-Tricine, pH 8.0, 0.33 M sorbitol) following the import reaction, and were then resuspended in 60 pl of IB containing 0.1% Triton X-100, 200 pg/ml proteinase K for 25 min on ice. These reactions were stopped by addition of 1 m M phenylmethylsulfonyl fluoride and an equal volume of 2 x SDS-PAGE sample buffer. Samples were separated by SDS-PAGE and visualized by fluorography (Theg et al., 1989).

Cloning cDNAs Encoding Arabidopsis CyP
Proteins-Screening of a n Arabidopsis leaf cDNA library with the coding region of a B. napus CyP cDNA clone (Gasser et al., 1990) resulted in the isolation of three independent clones encoding identical proteins. The gene from which these cDNAs derive was designated ROCl. The predicted sequence of the 172amino acid ROCl protein is compared to previously reported plant CyPs and to human cytosolic CyP in Fig. 1. ROCl exhibits a high degree of sequence identity to the cytosolic CyPs of other plants, showing 81% identity to that of tomato (another dicot) and 78% identity to that of maize (a monocot). ROCl is also 71% identical to human cytosolic CyP. Like these known cytosolic proteins, the ROC1 cDNA clones do not encode presequences that would provide information necessary for translocation of the protein out of the cytosol. As seen in the previously described plant CyPs, ROCl includes a seven amino acid insertion beginning at position 47 relative to the human cytosolic and most other known CyP proteins ( Fig.  1) (Gasser et al., 1990). On the basis of the sequence similarities and lack of apparent targeting information we conclude that ROCl is a cytosolic form of CyP.
Subsequent screening resulted in isolation of cDNA clones encoding a unique CyP isoform. The gene from which these cDNAs derive was designated ROC4. The predicted protein product of the ROC4 gene includes a n extension of 91 amino acids amino-terminal to the region homologous to known cytosolic CyPs (Fig. 1). The sequence comprised of the first 80 FIG. 1. Alignment of CyP protein sequences. The amino acid sequences of ROCl and ROC4 deduced from the corresponding cDNA sequences are aligned with the published sequences of tomato and maize CyP (Gasser et al., 1990), the previously isolated Arabidopsis CyP (ATHCYC) (Bartlinget al., 19921, and human CyP A (Haendler et al., 1987). Only amino acids that differ from ROC4 are shown. Dots indicate amino acids which are identical to the ROC4 sequence, and dashes indicate gaps introduced to allow for optimal alignment of the sequences. Ala residues (20 and lo%, respectively), relatively rich in basic amino acids (11% Lys + Arg), and deficient in acidic amino acids (0% Asp + Glu). The region of the amino-terminal extension between amino acid 82 and the beginning of the alignment with known CyPs contains five Asp residues. This region would be strongly negatively charged at physiological pH, inconsistent with it being part of a chloroplast transit peptide (von Heijne et al., 1989). Thus, the putative cleavage site for production of a mature protein would be expected to be upstream of this region. On the basis of this sequence analysis and data presented below, we hypothesized that the initial product of ROC4 is a precursor protein (pre-ROC41 which is cleaved after chloroplast import near amino acid 80 to yield the mature protein designated ROC4. In addition to the presence of the amino-terminal extension, the product of the ROC4 gene shows other significant differences from previously described plant CyPs. ROC4 lacks the seven amino acid insertion characteristic of the other plant CyPs, and in this respect is more similar to human cytosolic CyP ( Fig. 1 and see above). However, ROC4 shares only 66% sequence identity with human cytosolic CyP, and is thus less similar to this protein than are the plant cytosolic CyPs. On the basis of the presence and nature of its amino-terminal extension and the sequence of the putative mature protein, pre-ROC4 represents a novel class of plant CyP which has not been previously described.

Spatial Distribution of ROCl and ROC4 Gene Expression-
To determine the pattern of expression o f R O C l and ROC4 in mature plants, northern blots of root, leaf, and flower total RNA were hybridized with sequences from each of the cDNA clones (Fig. 2). Analysis of previously identified higher plant cytosolic CyP genes showed that these genes are expressed at relatively high levels in all tested parts of plants (Gasser et al., 1990;Marivet et al., 1992). mRNA from ROCl was readily detected in all tested organs of Arabidopsis plants (Fig. 2 A ) , consistent with ROCl being the Arabidopsis homolog of the previously reported plant cytosolic CyPs. In contrast, while ROC4 mRNA was readily detected in RNA from leaves and flowers, no mRNA was detected in root RNA (Fig. 2B). Leaves are the primary photosynthetic organs ofdrabidopsis, and Arabidopsis flowers also include leaflike photosynthetic sepals. Thus, the observed expression pattern is consistent with a role for ROC4 in chloroplasts. The fact that the ROCl and ROC4 probes detected different sets of bands demonstrates that the conditions used provided for gene specific hybridization.
CyP Is Present in the Stroma of Chloroplasts-The presence of a putative chloroplast transit peptide in the predicted ROC4 protein product prompted us to investigate the presence and localization of CyP in chloroplasts. Rotamase activity in pea chloroplast fractions was assayed using a coupled assay with chymotrypsin first developed by Fischer et al. (1984a, 198413). Chloroplast stromal extracts were shown to contain detectable rotamase activity which showed a linear relationship with the amount of stroma added (Fig. 3). The rotamase activities of all  Rotamase activity assays of the stroma of chloroplasts were performed as described under "Experimental Procedures." Rotamase activity is shown as a function of stromal protein quantity (pg) in the presence (0) or absence (0) of 10 p~ CsA. Assays were performed a t least in duplicate for each concentration tested. known eukaryotic CyPs and some prokaryotic CyPs are inhibited by CsA (Herrler et al., 1992;Walsh et al., 1992). Addition of CsA (10 PM) to the stromal extract abolished the rotamase activity dropping the rate into the range of the uncatalyzed reaction (Fig. 3). These results demonstrate the presence of a CyP homolog in the chloroplast stroma in agreement with previous findings (Breiman et al., 1992).
Material from the thylakoid lumen was also analyzed for rotamase activity, but no activity significantly above background was detected (data not shown). We were unable to assay for rotamase activity in the thylakoid membrane fraction because chlorophyll associated with these membranes interferes with the rotamase assay.
To further determine the intracellular localization of CyP, immunoblot analysis was performed using antiserum raised against ROC1. Anti-ROC1 serum reacted strongly with purified ROCl (Fig. 4A, lane 2 ) and cross-reacted with recombinant ROC4 obtained from solubilized inclusion bodies (Fig. 4A, lane   3 ). I t also detected recombinant cytosolic tomato CyP (data not shown; Gasser et al. (1990)) and even reacted weakly with E. coli CyP (data not shown; ) indicating cross-reactivity to a broad range of CyP isoforms. Immuno- staining of total Arabidopsis (Fig. 4A, lane 1 ) and total pea (Fig.  4 A , lane 4 ) extracts showed a t least three cross-reacting bands ranging from 18 to 20 kDa, consistent with the sizes of known CyP proteins. Pea chloroplast extract contained a single immunoreactive band (Fig. 4A, lane 5) which comigrated with recombinant ROC4 and the 20-kDa immunoreactive band in both total pea and total Arabidopsis extracts.
To determine the intraorganellar location of ROC4, chloroplast stroma, thylakoid membrane, and thylakoid lumen fractions representing equal amounts of chloroplasts were reacted with anti-ROC1 serum (Fig. 4A, lanes 6-8). An immunoreactive protein was detected in the stromal fraction (Fig. 4A, lane 6 ) , but not in thylakoid membrane or thylakoid lumen fractions ( Fig. 4 A , lanes 7 and 8). The chloroplast stromal CyP comigrated with the immunoreactive band in total extract of pea chloroplasts and was of equal intensity. Thus, the CyP protein detected in pea chloroplasts is located in the stroma.

ROC4 Precursor Protein Is Imported into Chloroplasts-
Transit peptides of chloroplast proteins are necessary and sufficient for targeting these proteins to chloroplasts. Since this targeting reaction is organelle-specific, import of pre-ROC4 into chloroplasts would prove that ROC4 is a plastid-resident protein. Incubation of radiolabeled pre-ROC4 (Fig. 5, lane 1 ) with freshly isolated pea chloroplasts followed by repurification of the chloroplasts resulted in a decrease in size of the majority of the pre-ROC4 band to that predicted for the mature form of this protein (Fig. 5, lane 2 ) . Control experiments carried out with the well characterized precursor of the small subunit of ribulose-l,5-bisphosphate carboxylase (pre-rbcS) showed identical results (Fig. 5, lanes 5 and 6).
Protease digestions were performed to demonstrate that the processed form of ROC4 was inside the chloroplasts. Following import, all residual precursor was eliminated by digestion with proteinase K, but the processed protein was protected from digestion (Fig. 5, lane 3). In contrast, when Triton X-100, which lyses chloroplasts, was included during the protease digestion both the residual precursor and processed ROC4 were digested  (lanes 2 and 6), or treated with 200 pg/ml proteinase K in the absence (lanes 3 and 7), or in the presence (lane 4 ) of 0.1%-Triton X-100 for 25 min on ice and then processed as described under "Experimental Procedures." Proteins were resolved on 15% SDS-polyacrylamide gels and visualized by fluorography. The positions of the precursor ( p r ) and mature ( m ) forms of ROC4 (left) and rbcS (right) are indicated. Molecular masses (in kDa) and relative positions of protein standards are indicated on the right .
( Fig. 5, compare lanes 3 and 4 ). Thus, the mature-sized ROC4 was protected inside the chloroplasts. We conclude that pre-ROC4 was competent for import into chloroplasts.
Rotamase Activity Is Not Required for Chloroplast Protein Import-To investigate whether rotamase activity is necessary for efficient import into chloroplasts, in vitro import reactions were carried out in the presence of CsA, which is known to penetrate readily through membranes (Schreiber, 1992). 10 p~ CsA added to an import reaction containing pre-rbcS did not significantly decrease the rate of protein import (Fig. 6). This concentration of CsA was sufficient to eliminate activity of the chloroplast stromal CyP (Fig. 3), as well a s activity of cytosolic CyP rotamases carried over from the wheat germ translation system used to produce the precursor proteins (data not shown). Similar results were obtained with the precursor to the 33-kDa protein of the oxygen evolving complex, a thylakoid lumen protein (data not shown). This indicates that CyP rotamase activity is not essential for efficient import of these two proteins. DISCUSSION CyP homologs have been identified in the cytoplasm of bacteria (Hayano et al., 1991) and in the cytosol of animals (Handschumacher et al., 1984), fungi (Tropschug et al., 1988;Haendler et al., 1989) and higher plants (Gasser et al., 1990;Marivet et al., 1992). Toward our goal of defining the rotamase activities and the in vivo roles of CyPs in plants, we have isolated cDNAs for two CyPs from Arabidopsis. Based on amino acid identity, presence of a seven amino acid insertion that is found in other plant cytosolic CyPs (Fig. l ) , absence of NH2-or COOH-terminal extensions which are usually necessary for translocation of proteins out of the cytosol, and constitutive high level expression, we conclude that ROCl is the major cytosolic CyP in this species. Bartling et al. (1992) recently described the isolation of a cDNA clone, designated ATHCYC, encoding a CyP from Arabidopsis. While ROCl has -80% amino acid identity to other plant cytosolic CyPs, ATHCYC has only -70% identity to ROCl or to the previously described plant CyPs (Fig.  1). Thus, ATHCYC is more diverged from all previously described monocot and dicot cytosolic CyPs (including ROCl), than the previously described proteins are from each other (Gasser et al., 1990). In fact, ATHCYC is as different in sequence from ROCl and the other known plant cytosolic CyPs as is human CyP A CSA -+ FIG. 6. Effect of CsA on in vitro import into chloroplast. Import of pre-rbcS into isolated chloroplasts was performed essentially as in Fig. 5 in the absence (-) and presence (+) of 10 p~ CsA. Assays were performed for 8 min, a time point that is still in the linear range of the import reaction (data not shown). Similar results were obtained in three replications of the experiment (data not shown). p r and m indicate the precursor and mature-sized polypeptides, respectively.
(-70% amino acid identity). ATHCYC, therefore, appears to represent a member of a class of plant CyPs that has not previously been described. No information on the pattern of expression of this gene is yet available. However, because ATHCYC lacks apparent targeting information i t may represent an additional cytosolic CyP in Arabidopsis.
Protein folding occurs in several subcellular compartments in addition to the cytosol. Consistent with an important role for CyP in this process, members of the CyP family have been found in the endoplasmic reticulum (Friedman and Weissman, 1991;Hasel et al., 1991;Price et al., 1991;Stamnes et al., 1991) and mitochondria (Tropschuget al., 1988;Walsh et al., 1992) of different eukaryotic species, and in the periplasm ofE. coli . The chloroplast is a unique subcellular compartment which is present in all photosynthetic eukaryotes. While chloroplasts have their own genomes, most chloroplast proteins are encoded in the nucleus, synthesized in the cytosol, and imported post-translationally into the organelle (Theg and Scott, 1993). Targeting information is contained on an aminoterminal extension called a transit peptide.
ROC4 is expressed only in parts of the plant containing photosynthetic tissues, and encodes an NH2-terminal extension having an amino acid composition consistent with chloroplast transit peptides. We have demonstrated the presence of CyP in the stroma of chloroplasts by enzyme activity assays and immunoblot analysis. These results support the previous observation of CsA-sensitive rotamase activity in chloroplast extracts (Breiman et al., 1992). Our results differ from those of Breiman et al. in that we were unable to detect immunoreactive CyP in the thylakoid fraction. Because the purity of their fractions was not assessed, stromal contamination could account for the observed thylakoid activity. Regarding our immunoblot experiments, the possibility remains that a thylakoid CyP may be too diverged in sequence to be recognized by our CyP antiserum. The ability of pre-ROC4 to enter the chloroplasts in in vitro import assays, together with the finding that the major immunoreactive CyP in chloroplasts is in the stroma, demonstrate that ROC4 is a novel CyP which functions in the stroma of chloroplasts.
della-Cioppa and Kishore (1988) obtained evidence indicating that proteins must undergo a change in conformation to be translocated across the chloroplast envelope membranes. Molecular chaperones are likely to act as catalysts for import (Waegemann et al., 1990;Theg and Scott, 1993). Given the known function of CyP in the protein folding process, we questioned whether, like other chaperones, the rotamase activity of CyP might facilitate the import process. Addition to import reactions of CsA sufficient to inhibit both cytosolic and chloroplast stromal CyPs did not block uptake of the tested chloro-plast precursors, demonstrating that the rotamase activity is not essential for efficient translocation of these proteins across the chloroplast envelope. However, these experiments do not preclude an in vivo role of CyP rotamase activity in the chloroplast stroma. The rotamase activity may be required for import of other proteins not tested here, or for proper folding and assembly of proteins inside the plastid.