OEP37 Is a New Member of the Chloroplast Outer Membrane Ion Channels*

The chloroplast outer envelope protein OEP37 is a member of the growing β-barrel protein family of the outer chloroplast membrane. The reconstituted recombinant protein OEP37 from pea forms a rectifying high conductance channel with a main conductance (λ) of Λ= 500 picosiemens (symmetrical 250 mm KCl). The OEP37 channel is cation-selective (PK+/PK– = 14:1) with a voltage-dependent open probability maximal at Vmem = 0 mV. The channel pore reveals an hourglass-shaped form with different diameters for the vestibule and restriction zone. The diameters of the vestibule at the high conductance side were estimated by d = 3.0 nm and the restriction zone by d = 1.5 nm. The OEP37 channel displayed a nanomolar affinity for the precursor of the chloroplast inner membrane protein Tic32, which is imported into the chloroplast through a yet unknown pathway. Pre-proteins imported through the usual Toc pathway and synthetic control peptides, however, did not show a comparable block of the OEP37 channel. In addition to the electrophysiological characterization, we studied the gene expression of OEP37 in the model plant Arabidopsis thaliana. Here, transcripts of AtOEP37 are ubiquitously expressed throughout plant development and accumulate in early germinating seedlings as well as in late embryogenesis. The plastid intrinsic protein could be detected in isolated chloroplasts of cotyledons and rosette leaves. However, the knock-out mutant oep37-1 shows that the proper function of this single copy gene is not essential for development of the mature plant. Moreover, import of Tic32 into chloroplasts of oep37-1 was not impaired when compared with wild type. Thus, OEP37 may constitute a novel peptide-sensitive ion channel in the outer envelope of plastids with function during embryogenesis and germination.

The chloroplast outer envelope protein OEP37 is a member of the growing ␤-barrel protein family of the outer chloroplast membrane. The reconstituted recombinant protein OEP37 from pea forms a rectifying high conductance channel with a main conductance () of ⌳ ‫؍‬ 500 picosiemens (symmetrical 250 mM KCl). The OEP37 channel is cation-selective (P K ؉/P Cl ؊ ‫؍‬ 14:1) with a voltage-dependent open probability maximal at V mem ‫؍‬ 0 mV. The channel pore reveals an hourglass-shaped form with different diameters for the vestibule and restriction zone. The diameters of the vestibule at the high conductance side were estimated by d ‫؍‬ 3.0 nm and the restriction zone by d ‫؍‬ 1.5 nm. The OEP37 channel displayed a nanomolar affinity for the precursor of the chloroplast inner membrane protein Tic32, which is imported into the chloroplast through a yet unknown pathway. Pre-proteins imported through the usual Toc pathway and synthetic control peptides, however, did not show a comparable block of the OEP37 channel. In addition to the electrophysiological characterization, we studied the gene expression of OEP37 in the model plant Arabidopsis thaliana. Here, transcripts of AtOEP37 are ubiquitously expressed throughout plant development and accumulate in early germinating seedlings as well as in late embryogenesis. The plastid intrinsic protein could be detected in isolated chloroplasts of cotyledons and rosette leaves. However, the knock-out mutant oep37-1 shows that the proper function of this single copy gene is not essential for development of the mature plant. Moreover, import of Tic32 into chloroplasts of oep37-1 was not impaired when compared with wild type. Thus, OEP37 may constitute a novel peptide-sensitive ion channel in the outer envelope of plastids with function during embryogenesis and germination.
The plastid organelle family conducts vital biosynthetic functions in every plant cell. Chloroplasts carry out photosynthesis, which converts atmospheric carbon dioxide to carbohydrates. Furthermore, plastids are the site for fatty acid, amino acid, and porphyrin biosynthesis as well as sulfate and nitrate reduction. Thus, all biosynthetic pathway products are steadily exchanged with the surrounding cell by the assistance of specific transport proteins, localized in the plastid envelope membranes. Although the inner envelope carriers, e.g. the triose phosphatephosphate translocator, the dicarboxylic acid translocator, or the hexose phosphate carrier, show a distinct substrate specificity, until now it remained elusive to what extent transport through outer membrane channels was regulated (for review of plastid envelope transporters compare with Ref. 1).
The outer envelope membrane has long been assumed to be freely permeable for small molecular mass solutes of up to 10 kDa (2). Correspondingly, it is believed that the osmotic barrier against the cytosol is formed exclusively by the inner envelope. In Gram-negative bacteria, however, several different types of high conductance channels exist in the outer membrane (3) as follows. (i) One type is the so-called porins that allow the downhill diffusion of solutes, provided that the size of the solutes does not exceed the exclusion limit (Ͻ600 Da), e.g. OmpF. Modulation of some of these channels by ATP and other effectors has been reported, questioning their classification as generally open diffusion pores. (ii) Other porin-like channels include LamB from Escherichia coli, for example, which carries specific sites that facilitate selective diffusion processes. (iii) Specific ligand-gated pores, e.g. E. coli ferric enterobactin channels (FepA), support energy-dependent uptake of iron into bacteria. The ancestral relation between plastids and Gram-negative bacteria therefore suggests the presence of multiple channel proteins in the chloroplast outer membrane.
In fact, three specific pore-forming proteins in the outer envelope of pea chloroplasts have already been discovered. They were named according to their location (outer envelope protein) and their molecular weight as OEP16, OEP21, and OEP24 (4 -9). In addition to the preprotein-conducting channel Toc75 (10,11), these outer envelope proteins have been functionally characterized by electrophysiological measurements. They represent high conductance solute channels with the highest open probability at 0 mV. Their distinct substrate specificities indicate separate roles in different metabolic processes, challenging the notion that they are general diffusion pores.
The ␣-helical OEP16 forms a cation-selective high conductance channel with a remarkable permeability for amino acids and amines (4 -6). OEP16 shows similarities to components of the protein translocase of the inner mitochondrial membrane and to a lesser extent to LivH, an amino acid transporter in E. coli (12). In contrast, OEP21 has been shown to form an anion channel, which is regulated by ATP and triose phosphates from the intramembrane space (7,13). The channel properties of OEP24, however, closely resemble those described for general diffusion pores (3). OEP24 allows the passage of triose phosphates, ATP, PP i , dicarboxylate, and positively or negatively charged amino acids (8) and can functionally replace the yeast mitochondrial VDAC 2 (9). Secondary structure predictions suggest that OEP21 and OEP24 have a ␤-barrel-like structure containing 8 and 12 membrane-spanning ␤-strands, respectively (13,14). In summary, these findings indicate that the inter-membrane space of chloroplasts is not freely accessible to low molecular weight solutes (15,16).
Here we describe the functional characterization of OEP37, a new member of the ␤-barrel solute channel proteins in the outer chloroplast membrane. OEP37 forms a rectifying, cation-selective, high conductance channel, which is sensitive to peptides and may function during seed development and germination of Arabidopsis plants.

EXPERIMENTAL PROCEDURES
Plant Material and Growth Conditions-All experiments were performed on Arabidopsis thaliana (L.) Heynh. Columbia plants (cv. Col-0; Lehle Seeds, Round Rock, TX). The t-DNA insertion line 722C01 was purchased from GABI-Kat, MPI for Plant Breeding Research, Cologne, Germany. Prior to sowing, seeds were surface-sterilized with ethanol and 5% hypochloride. To synchronize germination, all seeds were kept at 4°C for 3 days. Plants were grown on soil or on medium containing 1% D-sucrose and 0.5ϫ MS salts (Murashige and Skoog medium), pH 5.7. Plant growth occurred in growth chambers with a 16-h light (21°C, photon-flux density of 100 mol m Ϫ2 s Ϫ1 ) and 8-h dark cycle (16°C).
Isolation of AtOEP37 cDNA and Expression Level Profiling-Total RNA from A. thaliana tissues was isolated using the plant RNeasy extraction kit (Qiagen, Hilden, Germany). The cDNA of AtOEP37 was amplified by RT-PCR (Superscript II One-Step RT-PCR system with platinum Taq DNA polymerase; Invitrogen) on 20 ng of RNA, from 4-week-old Arabidopsis leaves. The following gene-specific primers were used: O37araXhoIs (5Ј-ccgctcgagatggcggatccatcttctca-3Ј) and O37araNcoIr (5Ј-catgccatggtcaaatgtcccatcttttc-3Ј). The resulting PCR product was subcloned by restriction digestion (XhoI/NcoI) into the pRSET A vector (Invitrogen). The open reading frame of AtOEP37 was verified by sequencing.
The expression level profiling by Affymetrix gene chip analysis in leaves and roots of 4-week-old Arabidopsis plants was performed as described previously (17). Data used to create digital Northern blots were obtained from the AtGenExpress experiments "effect of abscisic acid during seed imbibition." "expression profiling of early germinating seeds," and "developmental series (siliques and seeds)" at the NASCArrays website.
For promoter-GUS analysis, a 1499-bp region upstream of the designated start codon of AtOEP37 was PCR-amplified on genomic DNA by the gene-specific primers 37GUSgate_fw (5Ј-tagtggacaagattaagac-3Ј) and 37GUSgate_rev (5Ј-tggaattggatttgaga gaata-3Ј). The resulting product was subcloned into the binary pKGWFS7 vector (18) using the GATE-WAY cloning system (Invitrogen). Plasmids were transformed into chemically competent Agrobacterium tumefaciens cells, strain GV 3103. Transformation of Arabidopsis plants, selection of positive transformants, and histochemical GUS analysis were performed as described previously (19).
Antiserum Production and Western Blot Analysis of atOEP37-An antiserum against the purified recombinant atOEP37 protein was raised in rabbit (Pineda Antibody Service, Berlin, Germany). Intact mitochondria and chloroplasts were isolated from leaf material of 3-week-old Arabidopsis plants according to Refs. 20 and 21, respectively. Total organellar proteins were separated by SDS-PAGE and subsequently transferred to nitrocellulose filters. The antisera were used in a 1:1000 dilution, and blots were stained using the alkaline phosphatase reaction in the presence of nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate as substrate.
Characterization of the t-DNA Mutant oep37-1-Genomic DNA of the T2 and T3 progeny of the heterozygous t-DNA insertion line 722C01 (GABI-Kat (22)) was PCR-screened using the AtOEP37 gene-specific primer O37-GABI722_rev (5Ј-gacgtttgaggatcccattaacagataaagttc-3Ј) in antisense orientation in combination with the t-DNA-specific left border primer GABI-8409 (5Ј-atattgaccatcatactcattgc-3Ј). Both primers generated a 372-bp DNA fragment on hetero-and homozygous plants. By sequencing this PCR product, we identified the t-DNA insertion at position 925 in the AtOEP37 gene. To identify plants with the t-DNA insertion in both alleles of AtOEP37, we used the gene-specific sense primer O37-GABI722_fw (5Ј-ggatttcggttgtatgctagtgattatgttgg-3Ј) in combination with O37-GABI722_rev. DNA from homozygous oep37-1 plants gave no amplification product, whereas the amplified region on wild-type and heterozygous DNA was 412 bp long. In all plants tested (n ϭ 35), the oep37-1 mutant allele segregated in a Mendelian fashion (26% homozygous, 63% heterozygous, and 11% wild-type).
To monitor transcript abundance in oep37-1 mutant lines, total RNA was extracted from leaves of 4-week-old plants, DNase-digested, and reverse-transcribed into cDNA as described previously (17). Detection of transcripts by real time RT-PCR was performed as depicted before (23), using a LightCycler (Roche Applied Science). For AtOEP37 we constructed the following gene-specific primers: OEP37 LCnt_fw (5Ј-gcgagatttcacagcc-3Ј) and OEP37 LCnt_rev (5Ј-gagcggtacaaacacca-3Ј). The primer pair was designed to amplify a product of 366 bp, spanning the site of t-DNA insertion in oep37-1.
To prevent amplification of contaminating genomic DNA, the primers were selected to flank intron regions (compare Fig. 7A).
Protein Import Experiments-Intact chloroplasts for import experiments were isolated from 8-day-old seedlings of wild-type and oep37-1 according to Ref. 21. In vitro transcription, translation, and protein import of the precursor protein Tic32 (IEP32) from pea was performed as described previously (24). The standard import reaction contained Arabidopsis chloroplasts equivalent to 10 g of chlorophyll in 100 l of import buffer and 3% in vitro translation product.
Reconstitution with Mega 9-Small unilamellar liposomes were obtained from purified azolectin (type IV-S, Sigma) as described previously (11), and Mega9 was added to a concentration of 80 mM. Purified OEP37 at a concentration of 0.5-1.2 mg/ml in 6 M urea, 10 mM Tris/ HCl, pH 6.9, was also supplemented with Mega9 up to a final concentration of 80 mM. Both solutions were mixed, adjusting the lipid to protein ratio to 0.5-0.8 mg of protein/10 mg of lipid. The mixture was then dialyzed overnight against 5 liters of 10 mM Mops/Tris, pH 7.0, as described previously (25).
Planar Lipid Bilayers-Planar lipid bilayers were produced by the painting technique (26). A solution of 80 mg/ml purified azolectin (type IV-S, Sigma) in n-decan (analytical grade; Merck) was applied to a hole (50 -100 m diameter) in a Teflon septum, separating two bath chambers (volume 3 ml each). Both chambers were equipped with magnetic stirrers. By continuously lowering and then re-raising of the solution level, the lipid layer across the hole was gradually thinned out until a bilayer was formed. This formation was monitored optically and by capacitance measurements. The resulting bilayers had a typical capacitance of 0.5 microfarad/cm 2 and a resistance of Ͼ100 gigohms. The noise was 3 pA (root mean square) at 5-kHz bandwidth. After the formation of a stable bilayer in 20 mM KCl, 10 mM Mops/Tris, pH 7.0 (in both chambers ϭ symmetrical conditions), the solutions were changed by perfusion to asymmetrical conditions 250/20 mM KCl, 10 mM Mops/Tris, pH 7.0, cis/trans. An osmotic gradient of a channelpermeant solute and a channel in the open state (27) are prerequisites for fusion of proteoliposomes with the bilayer (28). Proteoliposomes were then added to the cis chamber directly below the bilayer to cause a slow flow of proteoliposomes along the bilayer surface. After fusion the electrolytes were changed to the final composition. We observed asymmetric response (rectification) of the OEP37 channel incorporated into the planar bilayer providing single fusion of a single OP37 channel occurred. However, the orientation of the channel was randomly distributed. The Ag/AgCl electrodes were connected to the chambers through 2 M KCl-agar bridges. The electrode of the trans-compartment was directly connected to the head stage of a current amplifier (Axon Gene Clamp 500B, Axon Instruments, Foster City, CA). Reported membrane potentials are always referred to as the trans compartment. The amplified currents were digitized at a sampling interval of 0.2 ms and fed into a Digidata1302 A/D converter (Axon Instruments) for storage on a personal computer. For analysis, a Windows TM -based analysis software (single channel investigation program) developed in our laboratory was used in combination with Origin 7.0 (Microcal Software Inc.).
Calculation of the Pore Size-We employed the PEG method in order to determine the pore size of the OEP37 channel (29 -31). The effect of the presence of 20% polyethylene glycol (PEG) of different molecular weights on the channel conductance was measured. PEG particles are sphere-like, neutral polymers, and their hydrodynamic radii are given in Ref. 32. The principle of the measurement is as follows. The electric conductivity of the bulk solution is lowered in the presence of PEG. At a low hydrodynamic radius, PEG can enter the channel and lower the conductance of the channel by the same factor as the bulk conductivity, but as their radius increases the particles are progressively excluded from the channel interior, and the conductance begins to recover. The conductance ratio ⌳ PEG /⌳ noPEG is plotted against the hydrodynamic radius of PEG, and the data are fitted with the logistic dose-response where A 1 and A 2 denote the stationary levels of the sigmoidal developing level; x 0 corresponds to the inflection point, and p is an exponential fitting parameter. The minimum and the maximum of the second derivative of the fit are assumed to give the characteristic radius of the pore for the restriction zone and the vestibule, respectively (32).

RESULTS
OEP37 Forms a Rectifying Cation-selective Channel with Subconductant States-The recombinant OEP37 protein from pea was purified to homogeneity (Fig. 1A) and reconstituted into liposomes as described in detail previously (32). After fusion of OEP37 liposomes with planar lipid bilayers, voltage-dependent single channel currents could be well resolved. Single channel activity was observed at both positive and negative membrane potentials (Fig. 1B). The OEP37 channel displayed complex gating with multiple open channel amplitudes. Two main (chord) conductance levels with conductances of ⌳ m1 ϭ 500 pS, ⌳ m2 ϭ 315 pS, and two less frequent open states with ⌳ s1 ϭ 125 pS and ⌳ s2 ϭ 75 pS were detected (pH 7, 250 mM KCl symmetrical; see Fig. 1C). The conductance of OEP37 was pH-dependent decreasing with decreasing pH values (compare Fig. 1, C and D). Moreover, in bilayers containing only a single active OEP37 channel, a rectifying current voltage relation as shown in Fig. 1E was observed. The rectification ratio at pH 7, 250 mM KCl, symmetrical, was ⌳ m1 /⌳ m2 ϭ 1.6 referring to the main conductance states. Decreasing salt conditions to 20 mM or changing to pH 6 did not affect this ratio. In differently concentrated electrolytes the conductance of the OEP37 channel exhibits saturation with ⌳ 1max ϭ 6.7 nS, K1 M ϭ 2.3 M, and ⌳ 2max ϭ 2.5 nS, K2 M ϭ 1.2 M for both main conductance states, respectively (Fig. 1F ).
In asymmetric 250/20 mM KCl solutions at pH 7, the OEP37 channel revealed a reversal potential (V rev ) ϭ ϩ49 mV ( Fig. 2A). By using the constant field Goldmann-Hodgkin-Katz approach, this corresponds to a selectivity of P K ϩ/P Cl Ϫ ϭ 14:1. Lowering the pH to a value of pH 6 decreases the reversal potential by roughly 5 mV (Fig. 2B), which corresponds to a reduction of the cation selectivity to P K ϩ/P Cl Ϫ ϭ 10:1. The selectivity of the OEP37 channel was strongly dependent on the ionic strength of the medium. This is documented in Fig. 2C, where reversal potential is shown at a constant cis/trans concentration gradient of 5 but in dependence of different absolute concentrations. The cation selectivity, which is expressed by the permeability ratio of the Goldmann-Hodgkin-Katz approach, increases about 5-fold when decreasing the absolute concentration by a factor of 10.
Our results revealed that for the OEP37 channel the permeation properties (conductance and selectivity) of physiologically not relevant Na ϩ ions were in the same order of magnitude as those for K ϩ ions (details not shown). To determine the nature of selectivity for physiologically relevant cations like K ϩ , Mg 2ϩ , and Ca 2ϩ , they were compared in reference to Cl Ϫ , respectively. The relative permeabilities of the OEP37 channel for these ions can be summarized to P K ϩ Ͼ Ͼ P Ca 2ϩ Х P Mg 2ϩ Х P Cl Ϫ ϭ 14:0.9:0.8:1 (see Table 1).
The open probability (P open ) of the OEP37 channel was determined at different pH values. A representative measurement (pH 7) is shown in Fig. 2D. P open reveals a voltage dependence typically observed for porins, being maximal at V mem ϭ 0 mV and decreasing with increasing voltage.
The OEP37 Channel Reveals an Hourglass-shaped Topology and Is Reversibly Blocked by Cu 2ϩ -The pore dimensions of the OEP37 channel were estimated using the polymer exclusion method (33). The pore size was independently determined for the high and low conductance side of the channel using the conductance of the fully open state (main conductance). Fig. 3A shows an example of measurements for the high conductance side, and the inset shows the determination of the pore , and 500 l of the purified OEP37 eluted from the TALON column (E) were separated by SDS-PAGE and Coomassie-stained. Note that psOEP37 is purified to homogeneity and runs at a molecular mass of 37 kDa. B, single channel current recording from a diameter for the low conductance side (⌳ m2 ). A summary of the results is given in Table 2.
We also tested the effect of CuCl 2 , a known inhibitor of the protein import pores Tom40 and Toc75 from outer membranes of mitochondria and chloroplasts (25,32). As shown in Fig. 3B, 2 mM CuCl 2 is sufficient to completely block the OEP37 channel. However, repeated perfusion with buffer solutions lacking CuCl 2 recovered the channel activity to some extent. For comparison, we also tested the reconsti-tuted Toc75 channel sensitivity toward CuCl 2 under identical conditions (Table 3).
OEP37 Constitutes a Peptide-sensitive Ion Channel-In order to gain information on the possible physiological role of the OEP37 channel, we compared its sensitivity for different prepeptides, which are known to induce a current block in the Toc75 channel, the pore of the general import pathway for prepeptides (10,11,32). The OEP37 channel was almost completely blocked by micromolar concentrations of Tic32 (Fig.  4A), a protein that is not imported into the chloroplast through the usual Toc complex rather than through a yet unknown alternative pathway (24). The current block was time-dependent (Fig. 4, A and B). In contrast to this, the current block of Tic32 in the Toc75 channel was of an overall lower degree (Fig. 4F). The sensitivity of the OEP37 channel for the chloroplast transit peptide TrOE33, previously tested with Toc75 (32), was significantly lower than for Tic32 (Fig. 4, C-E). The artificial peptide SynB2 described previously (32) is reported to plug several different channels in a nonspecific voltage-driven manner (34).
bilayer containing a single active OEP37 rec channel at a holding potential of V m ϭ ϩ100 and Ϫ140 mV. The bath solution contained 250 mM KCl, 10 mM Mops/Tris, pH 7 (symmetrical cis/trans). C, conductance histogram from n ϭ 15 independent bilayers at holding potentials of ͉Ϯ160͉ mV Յ V m Ն ͉Ϯ100͉ mV. The bath solution contained 250 mM KCl, 10 mm Mops/Tris, pH 7 (symmetrical cis/trans). The occurrence of two main conductance states reflects a rectifying property of the channel (see E ). D, conductance histogram from n ϭ 3 independent bilayers at holding potentials of ͉Ϯ160͉ mV Յ V m Ն ͉Ϯ100͉ mV. The bath solution contained 250 mM KCl, 10 mM BisTris/HCl, pH 6 (symmetrical cis/trans); data were smoothed by adjacent averaging (5 points). At pH 6, the conductance is reduced by about 20% compared with pH 7 (see D). E, current voltage relation of the main conductance of two OEP37 rec channels with identical orientation deduced from gating events in 250 mM KCl, 10 mM Mops/Tris, pH 7 (symmetrical cis/trans). F, both main conductances against the activity of the solution, fitting data with binding isotherm gives saturated conductances of 6.7 and 2.5 nS, and K m values of 2.3 and 1.2 M, respectively.  Along with Tic32, it was the only substance tested that affected the current significantly at nanomolar concentrations. A Tic32 deletion mutant, which lacks the first 10 N-terminal amino acids (⌬N 10 ) and is no longer imported into chloroplasts (24), revealed a significant lower current block (at 1.75 M ⌬N 10 -Tic32, 30% reduction) when compared with the complete Tic32 protein (at 1.75 M Tic32, 70% reduction) (Fig.  4E). This shows that most of the specific interaction is achieved by contacts between the OEP37 channel and the N terminus of Tic32. In summary, a high affinity of Tic32 toward the OEP37 channel can be distinguished from nonspecific or low current block of control peptides carrying multiple positive charges (32).

OEP37 in Arabidopsis Is Ubiquitously Expressed and Accumulates in Germinating Seeds and Developing
Embryos-The OEP37 protein we used for electrophysiological characterization was isolated previously in a study of the ␤-barrel proteome of the outer envelope from pea chloroplasts (14). In A. thaliana, the orthologous gene to PsOEP37 is encoded by the gene At2g43950. To study the function of this outer envelope pore in the model plant, we thus isolated the corresponding full-length cDNA of AtOEP37 by RT-PCR on the RNA of 4-week-old Arabidopsis leaves. Translation of the predicted open reading frame resulted in a protein with a molecular mass of 38.8 kDa (343 amino acids). In Arabidopsis, AtOEP37 represents a single copy gene with no further homologs identified. The genomic sequence consists of 6 exons (compare Fig. 7A), spliced into an mRNA molecule represented by the NCBI reference sequence NM_129957 and the RAFL cDNA, GenBank TM accession number AY045937, respectively. When compared with the ortholog from pea, atOEP37 shows 61% identity on the amino acid level. By screening the public available sequence data, we could identify another OEP37 ortholog in the rice genome only. Schleiff et al. (14) predicted a ␤-barrel topology with 12 ␤-strands and large intermembrane loops for atOEP37. Because the amino acid sequence between the orthologs is highly conserved, we propose a similar secondary structure and conserved functional characteristics for the OEP37 proteins from pea and Arabidopsis.
When we probed the Affymetrix full genome microarray for expression of AtOEP37, it demonstrated that transcripts are present in leaves and roots of 4-week-old plants (Fig. 5A). Screening the array data of the AtGenExpress consortium (L. Nover, University of Frankfurt, Germany, Arabidopsis Functional Genomics Network) revealed that apart from mature pollen, AtOEP37 is ubiquitously expressed on a low level throughout all developmental stages of Arabidopsis. High transcript abundance, however, was found in early germinating seeds and during late embryogenesis in developing seeds (Fig. 5B). This expression profile could be underlined by promoter-GUS analysis (Fig. 5C). In germinating seedlings, GUS signals were detected in cotyledons and the root hair zone of the primary root. In older seedlings, however, GUS stain was restricted to cotyledons in etiolated as well as light-grown plantlets. During seed development, we detected OEP37 promoter activity in the green cotyledons of mature embryos. Reflecting the low transcript abundance in true leaves, the promoter of AtOEP37 was not able to induce GUS expression either in leaves of seedlings or in mature plants (data not shown).
atOEP37 Is a Chloroplast-localized Protein-By using antibodies directed toward atOEP37, we could detect the protein in chloroplasts, isolated from 3-week-old Arabidopsis rosette leaves (Fig. 6). In contrast, mitochondrial proteins, isolated from the same plant tissue, did not react with ␣-atOEP37. For controls, we used antibodies against the chloroplast outer envelope protein psToc159 (35) and the mitochondrial diffusion pore psVDAC from pea (17).
An OEP37 Knock-out Line Displays No Apparent Phenotype and Is Not Impaired in Tic32 Import-To understand the function of OEP37 within the plant, we characterized the t-DNA insertion line 722C01 FIGURE 3. Estimation of the OEP37 rec pore diameters by the PEG method and blockage of the OEP37 rec channel by CuCl 2 . A, effect of the presence of 20% (w/v) of PEG of different molecular weights on the channel conductance was measured. The conductance ration is plotted against the hydrodynamic radius of the employed PEG. The experimental data were fitted using the following logistic equation: A shows the data for the high conductance side, and the inset shows the data obtained for the low conductance side of the channel. B, current recordings in response to applied voltage ramps from Ϫ120 mV Յ V m Յ ϩ120 mV from a bilayer containing several active OEP37 rec channels. The bath solution contained 250 mM KCl, 50 mM Mops/Tris, pH 7 (symmetrical cis/trans). The lower trace was obtained in the presence of 2 mM CuCl 2 (after 5 min of incubation symmetrical cis/trans). The middle trace shows the current response after 2ϫ exchange of the bath medium by perfusion.   The areas below the current recordings in response to a voltage ramp in the absence and presence of a given effector were determined by integration. The relative block was determined as the ratio of these areas. The effectors used were as follows: TrOE33, ⌬10N Tic32, SynB2, and Tic32. F, comparison of the Tic320-induced channel block in OEP37 rec and Toc75. Relative current block was calculated as described previously (E ).
(GABI-Kat (22)), subsequently designated oep37-1. The t-DNA inserts at the border of exon2/intron2 of the AtOEP37 gene, disrupting the open reading frame at amino acid 167 (Fig. 7A). The oep37-1 mutant allele segregated in a Mendelian fashion, allowing the identification of homozygous progeny (T3, T4 generation). Within these lines, the t-DNA insertion evokes a knock-out of AtOEP37, leading to the absence of transcripts and proteins in oep37-1/oep37-1 (Fig. 7B). However, until now we could not associate any phenotype with homozygous oep37-1 mutants. The mutant plants displayed normal development when compared with wild-type under standard growth conditions (data not shown), indicating that OEP37 function is not essential for the plant. Because the protein Tic32, which is imported into the chloroplast through a yet unknown pathway (24), shows specific affinity to the recombinant OEP37 channel (compare Fig. 4), we tested whether Tic32 import is impaired in the homozygous oep37-1 mutant. However, Tic32 import into chloroplasts of oep37-1/oep37-1 was as efficient as into wild type (Fig. 7C), providing evidence that OEP37 does not represent the alternative import pore for Tic32.

DISCUSSION
OEP37 is a member of the ␤-barrel proteins in the outer chloroplast membrane (14). Our results show that recombinant psOEP37 forms a rectifying, high conductance channel with significant cation selectivity. The ion selectivity of the OEP37 channel was dependent on the ionic strength, which indicates that saturable binding of cations to negatively charged groups in the channel pore might occur (36). Gillespie and Eisenberg (37) allocate two different contributions to an experimentally observed selectivity. The first one is unequal diffusion for different ion species within the pore which is given by their specific (pore) diffusion coefficient. The second effect is disparate partitioning of different ion species to the pore. This can be described by the activity coefficient ratio of the bulk electrolyte and the solution within the chan- shown. B, digital Northern blots of AtOEP37 expression (arbitrary units) in dry and germinating seeds after 0, 1, 3, and 24 h imbibition in water (left) and during embryo/seed development (right). The seed stages 3-10 are defined according to embryo development as follows: column 3, mid-globular to early heart; column 4, early heart to late heart; column 5, late heart to mid-torpedo; column 6, mid-torpedo to late torpedo; column 7, late torpedo to early walking stick; column 8, walking stick to early curled cotyledons; column 9, curled cotyledons to early green cotyledons; column 10, green cotyledons. Note that the seed stages 3-5 include tissue of siliques. Data used to create the digital Northern blots were obtained from AtGenExpress experiments at the NASCArrays website. Signal intensities were averaged from 2 to 3 technical replicates. C, histochemical localization of GUS expression under the control of the AtOEP37 promoter in germinating Arabidopsis seedlings and mature embryos. Left, in germinating seedlings, GUS expression is induced in cotyledons, the primary root, and root hairs. Note that no GUS signals appear in the root tip. Middle, in 7-day-old etiolated and light-grown seedlings, the GUS expression is restricted to the cotyledons. No signals can be detected in roots. Circled inset, magnification of the stained cotyledons. Right, GUS signals accumulate in the green cotyledons of mature embryos. FIGURE 6. Western blot analysis of atOEP37. Western blot on proteins of isolated chloroplasts and mitochondria from 3-week-old Arabidopsis leaves. 27 g of chloroplast (C ) and mitochondrial proteins (M) were subjected to SDS-PAGE followed by immunodecoration with antiserum against atOEP37 (middle). The ␣-atOEP37 stain gives a single band around 38 kDa. Antisera against the marker proteins psToc159 (outer envelope of chloroplasts, top) and psVDAC (outer membrane of mitochondria, bottom) were used as controls. Numbers indicate the molecular mass of proteins in kDa.
nel. In particular, the latter may be influenced by charges at the entrance to the channel. Increased shielding of these protein charges by higher concentrated solutions is an obvious explanation of our observations. Alcaraz and co-workers (38) present a similar dependence of ionic strength and selectivity for the bacterial outer membrane protein OmpF, which is comparable in pore size.
Schirmer and Phale (39) calculated Brownian dynamics simulations of ion flow through OmpF and discovered separated diffusion pathways for cations and anions as the origin of selectivity. This is because of the spacious segregation of basic and acidic residues of inner pore wall of the OmpF. The observation that the selectivity of the OEP37 channel is decreased by lowering the pH values encourages a model of negatively charged and slightly acidic residues controlling the permeability of ions.
Concerning the topology model given by Schleiff et al. (14), a long, acidic intermembrane loop with a negative net charge located between ␤-sheet 5 and 6 features all the properties required for the observed selectivity behavior. It may be emphasized that in these large pores several (ϳ3-5) evenly hydrated ions fit into the restriction zone next to each other. Therefore, selectivity in large pore channels is achieved by completely different electrostatic mechanisms as compared with single file channels, e.g. potassium channels.
Similar to previously characterized chloroplast outer membrane channels (4,5,7,8,13), the open probability of the OEP37 channel was highest in the absence of a membrane potential and decreased drastically at membrane potentials above Ϯ80 mV. The relative permeability of the OEP37 channel for divalent cations was surprisingly low when . The positions of the gene-specific primers O37-GABI722 forward (fw) and reverse (rev) of the t-DNA-specific left border primer GABI-8409 as well as of the primers used for quantitative RT-PCR (OEP37 LCnt_fw, _rev) are indicated by arrows. B, left, detection of AtOEP37 transcripts in 4-week-old rosette leaves of wild-type (wt) OEP37-1/OEP37-1) and homozygous (ho) oep37-1 plants. Transcripts were detected by real time RT-PCR with the primer pair OEP37 LCnt (see A) as described under "Experimental Procedures." End point PCR products of a representative reactions (n ϭ 3 independent experiments on T3 and T4 mutant generation) are shown. The primers amplified a specific product of 366 bp on wild-type plants, whereas no product could be detected in homozygous leaves. The black bar indicates the DNA size of 330 bp. Ϫ, negative control. Right, Western blot on proteins of isolated chloroplasts from cotyledons of 8-day-old seedlings of wild-type Col-0 (wt) and homozygous oep37-1 (ho) plants. 50 g of chloroplast proteins were subjected to SDS-PAGE followed by immunodecoration with antiserum against atOEP37. In wild-type chloroplasts the ␣-atOEP37 stain gives a specific band around 38 kDa, which is absent in the homozygous mutant. The black bar indicates the protein marker position of 36 kDa. C, import of the inner envelope protein psTic32 into the chloroplasts, used for Western blot analysis in B. Purified intact chloroplasts equivalent to 10 g of chlorophyll were incubated for 15 min at 25°C with the in vitro translated 35 S-labeled Tic32 from pea (TL). After the import reaction, chloroplasts were either not treated (Ϫ) or treated (ϩ) with the protease thermolysin for 30 min at 4°C. Intact chloroplasts were recovered by centrifugation, and import products were analyzed by SDS-PAGE and autoradiography. Because psTic32 is not processed, the translation product and the imported protein were run at an apparent molecular mass of 34 kDa. Please note that the fraction of imported and protease-protected Tic32 in the inner envelope of Col-0 wild-type (wt) and oep37-1 knock-out mutant (k.o.) is the same. compared with monovalent cations (see Table 1), indicating that the binding of the divalent cations to negatively charged groups in the channel pore occurred with high affinity, thereby decreasing the transport rate of the divalent cations in the channel pore (36).
The OEP37 channel revealed an hourglass-like shape with a diameter of 3.0 nm for the vestibules and 1.5 nm for the restriction zone. Despite this apparent geometrical symmetry, OEP37 features a significant current rectification. This shows that the interactions of the moving charges with fixed charges of the channel depends on the direction of the currents indicating that the density of fixed charges is asymmetrically distributed along the channel axis.
Interestingly, the OEP37 channel was blocked by low concentrations of CuCl 2 . This block could only be partially released by buffer exchange with solutions containing no CuCl 2 (Fig. 3B ad Table 3). This indicates that the CuCl 2 block is redox-mediated because of the oxidation of neighboring cysteines. Similar channel blocks have been observed for the protein import complex of the outer membrane of chloroplasts Toc75 (40) as well as for the amino acid-specific solute channel OEP16 (4,5).
Moreover, our results show that OEP37 interacts sensitively with the N terminus of the Tic32 protein. The stepwise closure induced by Tic32 occurred only at positive potentials (Fig. 4, A and B), indicating that positive voltages may enhance the channel block, which in addition occurred only from one side of the membrane.
In the model plant A. thaliana we were able to isolate the ortholog to OEP37 from pea (compare with Ref. 14). AtOEP37, encoded by At2g43950, represents a single copy gene in Arabidopsis. Except in mature pollen, AtOEP37 is ubiquitously expressed throughout all organs and developmental stages of the plant. Although transcript levels were comparatively low, we could detect the corresponding protein in isolated chloroplasts of cotyledons and rosette leaves. Thus, we assume that OEP37 represents a stable housekeeping protein in plastids of all plant organs. In early germinating seedlings and mature embryos, however, the OEP37 mRNA content peaked. In embryos and seedlings, AtOEP37 expression is restricted to cotyledons, although during germination transcripts can be found in the root hair zone of the primary root as well. Green cotyledons during late embryogenesis are loaded with solutes necessary for germination and thus display a strong sink characteristic. Plastids in this stage play a key role in storage protein and lipid synthesis (41). In contrast, cotyledons of the germinating seedling release the stored compounds. Thus, it is tempting to speculate that OEP37 mediates solute fluxes into and out of plastids during late embryogenesis and early germination of seedlings.
In vitro, the OEP37 channel pore is blocked by the Tic32 protein (see above), known to be imported into the inner envelope of chloroplasts by a pathway alternative to the classical import via the Toc translocon (24). This current block of OEP37 is specific for the first 10 N-terminal amino acids of Tic32, which are necessary for Tic32 import into chloroplasts. However, Tic32 import is insensitive to CuCl 2 (24), whereas OEP37 is reversibly blocked by Cu 2ϩ ions. Furthermore, Tic 32 in Arabidopsis has been shown to be essential for chloroplast biogenesis (42), because null mutants of this gene display an embryo lethal phenotype. In consequence, the yet unidentified import pore for Tic32 should play a crucial role in plant development as well. However, homozygous oep37-1 knock-out plants are alive and well, displaying no apparent phenotype under standard growth conditions. Furthermore, we could show that Tic32 import into chloroplasts lacking the OEP37 channel pore is not impaired when compared with wild type. Thus, we assume that OEP37 is a new member of the chloroplast outer membrane ion channels, which is sensitive to peptides and might play a role during embryo development and germination.