THE JOUFNAL OF CHEMISTRY Vol. 269, No. 41, Issue of 14, pp ... · THE JOUFNAL OF BIOWCICAL CHEMISTRY 0 1994 hy The American Society for Biochemistry and Molecular Biology, Inc. Vol.

THE JOUFNAL OF BIOWCICAL CHEMISTRY 0 1994 hy The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 41, Issue of October 14, pp. 25754-25760, 1994 Printed in U.S.A.

Porins from Plants MOLECULAR CLONING AND FUNCTIONAL CHARACTERIZATION OF TWO NEW MEMBERS OF THE PORIN FAMILY*

(Received for publication, January 31, 1994, and in revised form, June 23, 1994)

Karsten FischerS, Andreas Weber$, Susanne Brink$, Bettina ArbingerS, Danja SchunemannP, Sieglinde BorchertP, Hans W. Heldtg, Birgit Poppn, Roland Bena, Thomas A Linkll, Christoph Eckerskorn**, and Ulf-Ingo FluggeS S$ From the $Julius-von-Sachs-Znstitut fur Biowissenschaften der Universitat Wurzburg, Mittlerer Dallenbergweg 64, 0-97082 Wurzburg, the SZnstitut fur Biochemie der Pflanze der Universitat Gottingen, Untere Karspule 2, 0-37073 Gottingen, the ITheodor-Boveri-Znstitut fur Biowissenschaften, Am Hubland, 0-97074 Wurzburg, the IjGustav-Embden-Zentrum der Biologischen Chemie der Universitat Frankfurt, Theodor-Stern-Kai 7, Haus 25B, 0-60596 Frankfurt, and the **Max-Planck-Znstitut fur Biochemie, Am Klopferspitz 18, 0-82152 Martinsried, Federal Republic of Germany

Porins are voltage-gated diffusion pores found in all eukaryotic kingdoms. Here we describe, for the first time, the identification and characterization of two cDNAs encoding porins from plants. Peptide sequences obtained from a 30-kDa protein of envelope membranes from pea root plastids allowed the isolation of two cDNA clones from pea and maize. On the protein level, both proteins are homologous by 58%. Sequence comparison against the Swiss-Prot sequence data base revealed a homology of about 25% to mitochondrial porins from fungi and human. Computer-aided predictions of the secondary structure of the plant porins revealed the presence of 16 antiparallel /3-strands that are also found in mitochondrial porins. Porins from non-green plastids and from the outer mitochondrial membrane were re- constituted into planar lipid bilayers. The proteins showed high pore-forming activities and similar single- channel conductances. In uitro translated porin was preferentially imported only into non-green plastids but not into chloroplasts. To our knowledge, this is the first example of selective import of a plastidial protein into different types of plastids. This finding is in line with the observation that an immunoreactive 30-kDa band was only found in non-green plastids and mitochondria but not in chloroplasts. We conclude that mitochondria and non-green plastids possess homologous porin proteins, whereas chloroplasts are characterized by a different type of porin.

Plastids and mitochondria are both surrounded by two mem- branes that separate the organelles from the cytosol and that are mainly involved in the exchange of metabolites between these compartments and in the import of nuclear-encoded pro- teins. The mitochondrial outer membrane shows a permeabil- ity for hydrophilic molecules up to a molecular mass of 4-5 kDa due to the presence of general diffusion pores called voltage- dependent anion-selective channels or porins in analogy to the pore-forming proteins in the outer membrane of Gram-negative

* This research was supported by Grants F1 126/3-2, 3-3, SFB 176, Graduiertenkolleg Ka 456/5-1, and Ha 1232/8-1 from the Deutsche For- schungsgemeinschaft and by the Fonds der Chemischen Industrie. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adver- tisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Fax: 49-931-71446. $$ To whom correspondence should be addressed. Tel.: 49-931-73085;

bacteria (Colombini, 1979; Benz, 1985) (for a recent review see Benz (1994)). Isolation and reconstitution of the mitochondrial porins from Paramecium (Schein et al., 19761, Dictyostelium Cl'l"I'rll et al., 19921, yeast (Forte et a l . , 1987; Ludwig et al., 1988), corn (Smack and Colombini, 1985; Aljamal et a l . , 1993), and pea (Schmid et al., 1992) led to a detailed analysis of their biochemi- cal and biophysical properties. At low membrane potentials, all porins are weakly anion-selective in the "open" state. At volt- ages higher than 20 mV, the pore switches to the cation-selec- tive "closed" state. The large, water-filled pore is probably built up by 16 membrane-spanning antiparallel P-strands. The N terminus of the protein and the large extramembrane loops are located at the cytosolic side of the membrane (De Pinto et al., 1991). The primary structures of mitochondrial porins from Saccharomyces cerevisiae (Mihara and Sato, 19851, Neurospora crassa (Kleene et al., 1987), Dictyostelium discoideum (Troll et al . , 19921, and human (Kayser et al., 1989; Blachly-Dyson et al., 1993) have been determined. Surprisingly, the sequences ex- hibited only limited homology among each other (24-29%) de- spite the striking structural and functional similarities of all voltage-dependent anion-selective channels (Benz, 1994). The mitochondrial porins that are encoded in the nucleus without any cleavable N-terminal extensions have molecular masses of about 30-35 kDa. In contrast to mitochondria, little is known yet about pore-forming proteins in the outer membrane of plas- tids. The pore of the porin from chloroplast outer envelope membranes has a diameter of about 3 nm and has been found to be permeable to hydrophilic solutes up to a molecular mass of 10 kDa (Flugge and Benz, 1984), a value which is much higher than the exclusion limits of bacterial or mitochondrial porins (1-2 and 4-5 kDa, respectively). No data are available so far for porins from other plastids.

Here we describe the isolation and reconstitution of a porin from pea root plastids. Peptide sequences obtained from a 30- kDa outer membrane protein allowed the isolation of cDNA clones from pea and maize. On the protein level; both proteins are homologous by 58% but showed homologies of only about 25% to mitochondrial porins from fungi and man. In vitro syn- thesized porin was analyzed in import studies into different plastids and found to be specifically imported into non-green plastids but not into chloroplasts. The identity of the clones and the relationship between porins from mitochondria, chloro- plasts, and non-green plastids are discussed.

MATERIALS AND METHODS Cloning and Sequencing Procedures

Plastids from pea roots were isolated according to Borchert et al. (1989, 1993). Envelope membranes were prepared as described below.

25754

Porins from Plants 25755

After separation of envelope membrane proteins on preparative SDS- polyacrylamide gels (Laemmli, 19701, proteins were blotted onto polypropylene membranes (Selex 20, Schleicher and Schiill, Germany), and the N terminus of the 30-kDa protein (the major protein component of this membrane (Schunemann et al., 1993)) was subjected to auto- mated Edman degradation (Eckerskorn et al., 1988). In addition, the 30-kDa band was excised from the gel and directly digested with endo- proteinase Lys-C while still in the polyacrylamide matrix. The resulting peptides were eluted from the gel, separated by reverse-phase high performance liquid chromatography, and then sequenced in a gas-phase sequencer (Eckerskorn and Lottspeich, 1989). Four mixed oligonucle- otides modeled from these sequences were synthesized: P1, 5'-yTTIA- CRAARTTIACIGCIGCRTT-3'; P2,5'-RTGIGCIACYTCIGCICCIACNG- CNGT-3'; P3, 5'-GGNTITAYACIGAYATIGGIAARAARG-3'; and P4, 5'- AARAARGCIIIIGAYYTNTITAYAARGAYTA-3', where R is A or G and Y is C or T. The first two oligonucleotides (Pl, P2), based on the internal peptides 2 and 3, respectively, were used as reverse primers in polym- erase chain reactions, whereas P3 and P4 were derived from the N- terminal peptide and were used as forward primers. The polymerase chain reaction contained 30 ng of cDNA synthesized from pea root poly(A+) RNA, 10 mM Tris/HCl (pH 8.3),50 mM KC1,2 mM MgCl,, 0.2 mM of each dNTP, 0.01% gelatin, and 2 units of Taq-DNA polymerase. Po- lymerase chain reaction amplification was for 35 cycles of l min at 95 "C, 2 min at 45 "C, and 2 min at 72 "C. The 470- and 440-bp' DNA fragments produced were treated with T4-DNA polymerase (Ausubel et al., 1987) to get blunt ends, phosphorylated, subcloned into the bacterial vector pBSC, and then sequenced from both ends (Chen and Seeburg, 1985). The longer fragment was used as a hybridization probe for screening a pea root and a maize root cDNA library in A g t l O (Benton and Davis, 1977). Positive plaques were purified, their inserts were subcloned into pBSC, and both strands were sequenced.

Isolation of poly(A+) RNAfrom pea roots and synthesis of cDNAs was performed according to standard procedures (Ausubel et al., 1987).

Fractionation of Plastid Preparation and Enzyme Assays Pea root plastids were resuspended in 3.5 ml of 10 mM TricineXOH

(pH 7.9), 0.6 mM MgCl,, 0.33 M sorbitol, centrifuged (5 min, 100 x g) to remove starch granules, and the suspension was then adjusted to 0.7 M

sorbitol. The plastids were lysed by three freezekhaw cycles (1 h a t -20 "C; thawing at room temperature). The suspension was loaded on a sucrose step gradient (0.4, 1.1, 1.24, and 1.8 M) and centrifuged for 13 h at 80,000 x g. The gradient fractions were diluted to 16 ml with 10 m~ Tricine (pH 7.9),0.6 m~ MgCl, and centrifuged twice a t 100,000 x g for 60 and 30 min, respectively. The resulting membrane pellets were re- suspended in 10 mM Tricine (pH 7.91, 0.6 m~ MgCl,.

Measurements of Marker Enzyme Activities NADH-Ferricyanide Reductase (Mitochondrial Outer Membrane)"

The activity was determined by measuring the reduction of ferricyanide a t 436 nm (25 "C). The assay contained 10 mM TricineKOH (pH 7.9),0.6 mM MgCl,, 1 mM qFe(CN),, 1 mM KCN, 0.05 mM antimycin A, and 1 mM NADH in a final volume of 700 pl (Douce et al., 1973).

Cytochrome c Oxidase (Mitochondrial Inner Membrane)-The activ- ity was determined by measuring the oxidation of cytochrome c at 546 nm. The assay contained 100 mM KH,P04/K2HP0, (pH 7.0),70 mM KC1, 0.02% Triton X-100, and 200 p~ reduced cytochrome c in a final volume of 700 pl (Douce et al., 1973).

Acyl-CoA Synthetase (Plastidial Outer Membrane)-The assay con- tained 100 mM Tricine/NaOH (pH 8.01, 0.5 mM ['4Cloleate (12.7 pCi/ pmol), 5 mM MgCl,, 5 mM ATP, 5 m~ phosphoenolpyruvate, 0.5 mM Coenzyme A, 5 mM dithiothreitol, 0.02% Triton X-100, and 167 nano- katal pyruvate kinase (Andrews and Keegstra, 1983). The reaction was started by the addition of 10-30 pg of membrane protein and incubated for 2 min at 30 "C. After termination of the reaction by the addition of 1 ml of H,O/isopropyl alcohoUacetic acid (33:33:1, v/v/v), the samples were centrifuged for 5 min. The supernatant was extracted three times with 600 pl of petroleum etheddiethyl ether ( L l , v/v), and aliquots of the water phase were used for liquid scintillation counting.

Acyl-CoA Thioesterase (Plastidial Inner Membrane)-The activity was determined in a medium containing 10 mM TricineKOH (pH 9.0), 4 p~ [14Clpalmitoyl-CoA (36 pCi/pmol), and 0.02% Triton X-100 (Joyard and Stumpf, 1980). The reaction was started by the addition of 20-40 pg of protein (final volume, 67 111). The samples were incubated for 4 min

' The abbreviations used are: bp, base pair(s); PAGE, polyacrylamide gel electrophoresis; HTP, hydroxylapatite; SSU, the small subunit of ribulose-l,5-bisphosphate carboxylase/oxygenase; Tricine; N-[2-hy- droxy-1,l-bis(hydroxymethy1)ethyllglycine; nS, nanosiemens.

at 37 "C and the reaction was stopped by the addition of 0.6 ml of H,O/isopropyl alcohollacetic acid (33:33:1, v/v/v). The fatty acids were extracted three times with petroleum etheddiethyl ether (l:l, v/v, satu- rated with 50% isopropyl alcohol). Aliquots of the collected organic phases were used for liquid scintillation counting.

Isolation of Envelope Membranes Isolation of envelope membranes from Percoll-purified intact pea root

plastids was performed according to Mourioux and Douce (1981). The plastids (12 mg of protein) were osmotically lysed in 16 ml of 10 mM TricineKOH (pH 7.8), 0.6 mM MgC1, (buffer A). The suspension was centrifuged (5 min, 100 xg ) t o remove starch granules. The supernatant was adjusted to 10% (w/v) sorbitol, loaded onto discontinuous sucrose gradients (17% (w/v), 5 ml; 32% (w/v), 4.5 ml) and centrifuged at 80,000 x g for 2 h. The envelope membranes at the 17%/32% interphase were diluted to 16 ml of buffer A and centrifuged at 100,000 x g for 1 h. Marker enzymes were measured as described above; ADPglucose pyro-

Borchert et al. (1993). phosphorylase and citrate synthase were measured according to

Lipid Bilayer Experiments The methods used for the lipid bilayer experiments have already

been described (Benz et al., 1978). The experimental set up consisted of a Teflon cell with two water-filled compartments connected by a small circular hole. The hole had an area of about 0.5 mm'. Black lipid bilayer membranes were obtained by painting onto the hole a 1% (wh) solution of diphytanoyl phosphatidylcholine or phosphatidylserine (Avanti Polar Lipids, Alabaster, A L ) in n-decane. The temperature was maintained at 20 "C during the experiments. The single channel records were per- formed using calomel electrodes (with salt bridges) connected in series to a voltage source and a current-to-voltage converter (Keithley 427). The amplified signal was monitored on a storage oscilloscope (Tektronix 7633) and recorded on a strip chart or tape recorder.

Protein Import Assay cDNA of pea porin, of omp24, a chloroplast outer envelope protein

(Joyard et al., 1983; Fischer et al., 19941, and of the small subunit of ribulose-bisphosphate carboxylase/oxygenase (Lubben and Keegstra, 1986) were transcribed according to the instructions given by the man- ufacturer of the RNA polymerases (Boehringer, Mannheim, Germany). Translation was performed in a reticulocyte lysate system (Promega, Heidelberg, Germany) using l3H1leucine or [35Slmethionine (Amer- sham, Braunschweig, Germany). Cauliflower bud plastids were isolated from cauliflower inflorescences essentially as described previously (Journet and Douce, 1985; Alban et al., 1988). The in vitro synthesized labeled proteins were used for uptake studies into spinach chloroplasts and cauliflower bud plastids, respectively. Protein import assays con- tained 250 m~ sorbitol, 10 mM methionine (or leucine), 25 mM potassium gluconate, 0.2% (w/v) bovine serum albumin, 2 mM MgSO,, 2 mMATP, 50 mM HepedKOH (pH 8.01, and purified organelles equivalent to 200 pg of chlorophyll or 1.7 mg of protein, respectively, in a final volume of 300 pl. The import reaction was allowed to proceed for 20 min a t 25 "C. Isolation of chloroplast envelope membranes was performed as de- scribed by Flugge et al. (1989). Cauliflower bud plastids were disrupted by osmotic shock in 10 mM Tricine/NaOH (pH 7.8), 4 mM MgCl, and homogenized with a pestle. Membranes were sedimented by centrifu- gation for 15 min at 20,000 x g. All fractions were analyzed by SDS- PAGE and fluorography.

RESULTS AND DISCUSSION

Molecular Cloning and Sequence Analysis of TILIO Por in Clones from Pea and Maize-Envelope membranes were puri- fied from pea root plastids by discontinuous sucrose density centrifugation. Compared with plastid preparations (Borchert e t al., 1993) (see below, Fig. 61, mitochondrial contamination of the envelope membranes was further decreased (Table I). Whereas 37% of the outer envelope marker was recovered in the envelope fraction, the purified membranes contained only 0.2% of citrate synthase (marker for matrix enzymes), 0.5% of cytochrome c oxidase (marker for the inner mitochondrial membrane), and 2% of NADH-ferricyanide reductase (marker for the outer mitochondrial membrane). The 30-kDa protein from pea root plastids was isolated from these purified enve- lope membranes. The N terminus and, in addition, internal peptides obtained by digestion of the protein with endoprotein-

25756 Porins from Plants

TABLE I Activities and recoveries of marker enzymes in the envelope fraction of pea root plastids

Envelope membranes were purified from pea root plastids as described under “Materials and Methods.” The recovery is related to the total activity in the pea root plastid fraction. The sum of the recoveries in the various fractions was between 85 and 97% for all enzymes listed. n, number of independent preparations.

Enzyme Activity Recovery of total activity in the envelope fraction n

nmol . Plastidial stroma, ADP-glucose

pyrophosphorylase Outer envelope membrane, acyl-CoA

Mitochondrial matrix, citrate synthetase

synthase Inner mitochondrial membrane,

cytochrome c oxidase Outer mitochondrial membrane,

NADH-ferricyanide reductase

img proteinPmin” f S.D. 6.6 e 5

91 * 45

6.8 * 6.3

410 * 400

490 * 350

% f S.D. 0.1 * 0.1

37 * 21

0.2 * 0.1

0.5 e 0.3

2.0 * 1.0

ase Lys-C were subjected to amino acid sequence analysis. The N-terminal sequence VKGPGLYTDIGKKAXDLLYKDY (Pepl) and the sequences of two internal peptides, SNAAVNFVK (Pep2) and TAVGAEVAHRFSTK (Pep3), were obtained.

Two sense oligonucleotides (P3 and P4) based on the N-ter- minal amino acid sequences GLYTDIGKK and KKAXDL- LYKDY and two antisense oligonucleotides (P1 and P2), mod- eled from the internal peptide sequences NAAVNFVK and TAVGAEVAH, respectively, were synthesized and used as prim- ers in a polymerase chain reaction with cDNAfrom pea roots as a template. The primer combinations P3P2 and P4P2 yielded no product, whereas the forward primers used in combination with P1 generated two DNA fragments, 470 and 440 bp in length, respectively. Since P4 is located 21 bp downstream of P3, a reduction in length of 20-30 bp was expected for the correct product. Sequence analysis of the 470-bp fragment re- vealed that it contained both primers and nucleotide sequences coding for that part of the peptides that was not used for mod- eling of the oligonucleotides. This indicates that the DNA frag- ment is generated from a cDNA coding for the 30-kDa envelope protein. The polymerase chain reaction fragment was then used to screen a pea root cDNA library for a full-length clone. Four clones (pspor2, -3, -8, and -9) could be isolated, each of them containing inserts of - 1200 bp in length that all showed identical restriction patterns. The insert of clone pspor8 con- sists of a 51-bp 5”untranslated sequence region, an open read- ing frame of 828 bp, and 270 bp at the noncoding 3”region. The nucleotide sequence of clone pspor8 (porin from pea) has been submitted to the GenBanWEMBL data bank and is available under accession number 225540. The length of the coding re- gion corresponds to a protein of 276 amino acids and a molec- ular mass of 29,596 Da, which is quite similar to the value determined by SDS-PAGE. As revealed by N-terminal sequenc- ing, the start methionine is not present in the 30-kDa protein indicating its removal in vivo by a methionine aminopeptidase (Lucas-Lenard and Lipman, 1971). The N-terminal peptide (Pepl) and one of the internal peptides (Pep2) are completely encoded by the clone, but the second internal peptide (Pep31 showed four amino acid exchanges (TAVGAEVAHRFSTK) com- pared with the sequence encoded by the clone pspor8 (TAVGVDISHRFSTK).

The cDNA insert of clone pspor8 was used as a probe for screening a cDNA library from maize roots in A g t l O (Strat- agene, La Jolla, CA). Four clones were obtained, three of them having -1000-bp EcoRI inserts, but one (zmporl) containing two EcoRI inserts -350 and - 1500 bp in length. The inserts of the clones zmporl and -3 were subcloned and sequenced. The longer DNA fragment of clone zmporl starts at an internal EcoRI site at position 328, while the shorter fragment com-

prised the 5’-end of the cDNA up to the EcoRI site. The nucle- otide sequence of clone zmporl (porin from maize) has been submitted to the GenBanWEMBL data bank and is available under accession number X73429. The open reading frame of the zmporl cDNA codes for a polypeptide of 277 amino acid resi- dues corresponding to a molecular mass of 29,977 Da. The amino acid sequences of the proteins from pea and maize are homologous by 58%. Sequence comparison against the Swiss- Prot sequence data base revealed a low but significant homol- ogy (20-25%) to mitochondrial voltage-dependent anion chan- nels (or porins) from yeast (Mihara and Sato, 1985), N. crassa (Kleene et al., 1987), D. discoideum (Troll et al., 1992), and human (Kayser et al., 1989; Blachly-Dyson et al., 1993). A se- quence alignment of these porin proteins is shown in Fig. 1. Remarkably, only 9 out of about 280 amino acid residues (3 glycine, 3 aspartate, and 3 lysine residues, respectively) are conserved in all porins; 3 residues are located at the N terminus and 2 in the tripeptide GL(V)K. The hydrophobicity distribu- tion analysis (not shown) indicates the hydrophilic nature of the porins (51% hydrophilic residues) lacking any hydrophobic stretches long enough to span the membrane as an a-helix. Computer-aided predictions (De Pinto et al., 1991) of the sec- ondary structure of the maize porin revealed a pattern of one amphiphilic a-helix at the N terminus and 16 antiparallel transmembrane p-strands (Fig. 2 A ) . Except for p-sheets 9 and 10 these p-strands are found at almost the same positions in the maize protein as compared with the human porin 31 HL (Fig. 2B) (De Pinto et al., 1991) despite their low homology on the primary sequence level. Thus, the secondary structure of plant porins shows a striking similarity to mitochondrial porins from fungi and human. The spinach chloroplast porin appar- ently does not belong to this porin family since it forms a pore with an exclusion limit of 10 kDa, a value that is significantly higher than that determined for mitochondrial porins (Fliigge and Benz, 1984) and since the chloroplast porin, in contrast to the porin from pea root plastids, is not recognized by an anti- body raised against yeast mitochondrial porin (see below, Fig. 6B).

Reconstitution of Porin from Pea Root Plastids into Lipid Bilayer Membranes-To study the pore formation by the porin from non-green plastids, the plastids were subjected to sucrose density centrifugation (see “Materials and Methods”). Fraction 2 of the sucrose gradient contained a high percentage of outer plastidial membranes and almost no mitochondrial contamina- tion (Fig. 6). This fraction was treated with the detergent Ge- napol X-080 (l%, w/v), and hydroxylapatite (HTP) was then added to the detergent-solubilized proteins. Using this batch procedure, most of the proteins but not porin were adsorbed to the HTP material (Schmid et al., 1992). The supernatant was

Porins from Plants 25757

FIG. 1. Alignment of amino acid sequences of porin from pea (pspor) with sequences from maize (zmpor), yeast (ncporin). N. crmsa (ncporin), human (hnporinl and -2), D. dincoideum (ddporin), and a short peptide from maize mitochondrial porin (maize mito). The positions of the three peptides obt:lined I)! protein sequencing of the pea porin are indicated by bars ( I - I I f ). Identitics of amino acids are boxed in black and conservative amino acid suhstitu- tions are boxed in gray, respectively.

subsequently added to black lipid bilayer membranes made of diphytanoyl phosphatidylcholineln-decane. The addition of the detergent-solubilized envelopes resulted in a stepwise increase of the membrane conductance in single-channel recordings (Fig. 3A) . The conductances of these steps were not uniform in size but were distributed across a certain conductance range as shown in Fig. 4.4. Maxima for the single-channel distribution were found around 1.5 and 3.7 nS (1 31 KCI). 4 nS is a typical value for the single-channel conductance of mitochondrial por- ins under the conditions of Fig. 4.4. It may be noted that also smaller single-channel conductances have been observed for a variety of different mitochondrial porins (Benz, 1994).

We also studied the pore-forming activity of fraction 4 of the sucrose gradient. This fraction was highly enriched in both mitochondrial outer and inner membranes (Fig. 6). Reconsti- tution of the proteins that were not adsorbed by hydroxylapa- tite into planar bilayer membranes resulted in a stepwise in- crease of the membrane current (Fig. 3B). A histogram of the conductance fluctuations of the mitochondrial porin also re- vealed maxima a t 1.5 and 3.7 nS (Fig. 4R). Obviously, plastids and mitochondria from non-green tissues contain porins with very similar characteristics.

Import of Pea Porin into Chloroplasts and Non-green Plastids-Recent studies on protein transport into different types of plastids, i.e. chloroplasts and non-green plastids, showed that, despite their different function, the mechanism of protein import into these organelles is quite similar (Klosgen et al., 1989). Chloroplast outer envelope membrane proteins do not possess N-terminal extensions (transit peptides), and their insertion into the envelope membrane apparently proceeds without the participation of a receptor or the utilization of energy (Salomon et al., 1990; Li et al., 1991). Thus, a t least two different protein import pathways exist in chloroplasts, one for targeting of proteins to the outer envelope and the second used by proteins located to other chloroplastic compartments, i.e. the

I I I I I 1 50 100 150 200 250 300

Amino acid residue

B

I 1

50 100 150 200 250 300 Amino acid residue

FIG. 2. Sided hydropathy profiles of the porin from maize (A) and averaged sided hydropathy plot of the four different mito- chondrial porins ( R ) . The sided hydropathy values for sided C(-sheets and the corresponding averaged hydropathy values were obtained by summing up the hydropathy indices of the respective residues (Kyte and Doolittle, 1982) using the following weight9 (given in parentheses):

be detected in the sided p-sheet profile by an alteration between hydro- residues n (1.0). n = 2 (1.0). and n = 4 (0.5). Possible sided p-sheets can

philic and hydrophobic values every second residue. Tentative am- phipathic membrane-spanning pstrands are indicated by bars (2-16).

inner envelope membrane and the stroma (Keegstra et a!., 1989). Here we asked whether the porin encoded by clone pspor8 could be imported into chloroplasts and non-green plas- tids. In vitro translated pea porin was added to either isolated chloroplasts from spinach or non-green plastids from cauli- flower buds. For comparison, in vitro synthesized omp24, a chloroplast outer envelope membrane protein (Joyard et al., 1983; Fischer et al., 1994), and the precursor of the small sub- unit of ribulose-bisphosphate carboxylaseloxygenase (SSU) (Lubben and Keegstra, 1986), a chloroplast stromal protein, were used as controls in the same assays. The results of the import experiments are shown in Fig. 5. SSU and omp24 were imported into both types of plastids in a protease-resistant manner (lanes 2 and 6). Furthermore, the precursor of SSU was processed to its mature form albeit import and processing of both chloroplast proteins in non-green plastids were not as eficient as in chloroplasts. In contrast to the chloroplast pro- teins, pea porin was only imported into non-green plastids (lane 2 ) , but not into chloroplasts (lane 6). All the protein bound to chloroplasts was degraded by the externally added protease thermolysin (lane 6). Porin protein appears to be deeply embedded in the outer membrane of non-green plastids as imported porin (and omp24) was not digested by thermolysin (lane 2 ) and, moreover, was not affected by carbonate treat- ment of envelope membranes (lane 3) whereas SSU was fully extracted by this procedure (lane 4) . Such treatment at alka-

25758

A Porins from Plants

B A ri

i Y

J i

I00 pA 4y i jIO " S 100 pA

FIG. 3. Single-channel recordings of a diphytanoyl phosphati- dylcholineln-decane membrane in the presence of 0.1 pg/ml de- tergent-solubilized and HTP-purified enriched outer plastidial membranes (fraction 2 of Fig. 6) (A) and 0.1 pg/ml detergent- solubilized and HTP-purified mitochondrial membranes (frac- tion 4 of Fig. 6) (B) . The aqueous phase contained 1 M KC1 (pH 6.0). The applied membrane potential was 10 mV, and the temperature was 20 "C.

line pH is known to extract only soluble and peripheral mem- brane proteins, while integral membrane proteins remain as- sociated with the membrane sheets (Fujiki et al., 1982).

It is currently not known how protein targeting into the appropriate kind of plastid is achieved. To our knowledge, all plastid precursor proteins tested so far are imported into both chloroplasts and non-green plastids indicating the similarity between the different import machineries (Nosgen et al., 1989). Surprisingly, the outer membrane protein porin is, as a first example, preferentially imported only into non-green plas- tids. This finding is corroborated by the observation that cross- reactivity with an antibody raised against the yeast porin was only found with the porin from non-green plastids but not with that from chloroplasts (Fig. 6B). Since other outer envelope membrane proteins (E6.7 (Salomon et al., 1990), OM14 (Li et al., 1991), and omp24 (Fischer et al., 1994)) are imported with- out participation of receptor proteins, the molecular basis of this discrimination is unknown. Work is now in progress to elucidate whether or not porin is imported via a receptor- dependent pathway.

Are Mitochondrial and Non-green Plastidial Porins Identical?-Considering the sequence homology of both plant porins to the mitochondrial voltage-dependent anion-selective channels from fungi and human and their similar single-chan- ne1 conductances and voltage dependences, the question arises whether both non-green plastids and mitochondria possess al- most identical porins. On the other hand, the plant porins characterized in this paper could actually be of mitochondrial rather than of plastidial origin due to a mitochondrial contami- nation of the plastidial envelope from which membrane the protein was isolated. Non-green plastids that served as a source for the preparation of envelope membranes are indeed contaminated by mitochondrial marker enzymes (Borchert et al., 1993). To measure the extent of mitochondrial contamina- tion of the plastidial envelope preparation, isolated pea root plastids were separated on a sucrose density step gradient (0.4, 1.1,1.24, and 1.8 M). Five fractions were obtained and analyzed

1 4 1 2 3

B

- 5 6 7

G [nSI

1 2 3 4 5 6 7

G [nSl FIG. 4. Histograms of the probabilityp(G) for the occurrence of

a given conductivity unit observed with membranes formed of diphytanoyl phosphatidylcholineln-decane in the presence of 0.1 pg/ml detergent-solubilized and HTP-purified enriched outer plastidial membranes (fraction 2 of Fig. 6) (A) and 0.1 pg/ml detergent-solubilized and HTP-purified mitochondrial membranes (fraction 4 of Fig. 6) (B) . The average single-channel conductances were 1.5 nS for the left side and 3.7 nS for the right side maxima and were calculated from n = 333 and n = 326 current steps, respectively. p ( G ) is the probability that a given conductance increment G is observed in the single-channel experiments. It was calculated by

by the total number of conductance fluctuations. dividing the number of fluctuations with a given conductance increment

for the presence of mitochondrial and plastidial outer and inner membranes, respectively, and for the localization of the immu- noreactive 30-kDa protein. Fig. 6A shows the distribution of the activity of marker enzymes (NFO, NADH-ferricyanide reduc- tase, outer mitochondrial membrane; COX, cytochrome c oxi- dase, inner mitochondrial membrane; ACS, acyl-CoA synthe- tase, outer plastidial membranes; ACT, acyl-CoA thioesterase, inner plastidial membranes). The marker for the outer plas- tidial membrane was mainly found in fractions 2 and 3, whereas the outer mitochondrial membrane was almost exclu- sively found at a higher density (84% in fraction 4). Remark- ably, the marker for the outer mitochondrial membrane was almost absent in fraction 2, which contained, however, the highest percentage of outer plastidial membranes. It thus ap- pears possible to separate the outer and inner organelle mem- branes on the basis of their different densities. Fig. 6B shows that the antibody directed against the yeast mitochondrial porin strongly cross-reacted with the porin from pea root plas- tids (Fig. 6B, lane P ) and, albeit with lower sensitivity, also

Porins from Plants 25759

1 2 3 4 5 6

porin 4

omp24 -+ ssu p-+

SSU m+

- porin

- omp24

- p s u

protease - + + i - + , sed. sup., Na2C03

I - cauliflower bud plastids chloroplasts

SSU into non-green plastids and chloroplasts. Intact cauliflower FIG. 5. Import of pea porin, omp24, and the precursor ( p ) of

bud plastids (5 mg of proteidml) or spinach chloroplasts (0.6 mg of chlorophylllml) were incubated for 20 min in import buffer in the light and in the presence of the in vitro synthesized proteins (see “Materials and Methods”). After import, samples 2,3,4, and 6 (lanes 2,3,4, and 6 ) were further treated with thermolysin (100 pg/ml) as described earlier (Fliigge et al., 1989). Membranes were separated and analyzed as de- scribed under “Materials and Methods.” Membrane pellets containing imported pea porin, omp24, and SSU were resuspended in 100 mM NalCO,, incubated for 30 min a t 0 “C, and afterward centrifuged a t 166,000 x g for 60 min. The membrane pellets (sed. ) and the superna- tants (sup. ), which had been precipitated by trichloroacetic acid (final concentration, lo%, v/v) were analyzed by SDS-PAGE and fluorography m. mature.

with the pea mitochondrial porin (lane M). In addition to the functional similarity of the plastidial and the mitochondrial porins (see above), this result further leads one to assume that non-green plastids and mitochondria possess (a set of) closely related porin proteins. Remarkably, no cross-reactivity was found with envelope membranes from chloroplasts (Fig. 6B, lane C ) . When fractions 2-4 of the sucrose density gradient were subjected to Western blot analysis, an immunoreactive 30-kDa band was detected in all fractions with the highest signal found in fraction 2. Since mitochondrial membranes were almost absent from this fraction, the Western blot anal- ysis provides definitive evidence for an organelle-specific porin present in plastids of pea roots. Although this result does not address the issue of subcellular localization of the porins en- coded by the cDNAs characterized it may be supposed that these porins are of plastidial origin since (i) purified envelope membranes from which the 30-kDa protein was isolated were contaminated by mitochondria to a very low extent (Table I), (ii) a peptide obtained from the maize mitochondrial porin (Alja- mal et al., 1993) does not perfectly fit the deduced amino acid sequence of clone zmporl from maize (Fig. 1, exchange of 7 out of 22 amino acid residues), and (iii) the N terminus of the mitochondrial porin from pea was found to be blocked (Schmid et al., 1992) whereas the N terminus of the porin from pea root plastids could be sequenced. Finally, the porin protein is efi- ciently imported into non-green plastids and deeply embedded into the outer envelope membrane (Fig. 5).

The observation that the third peptide (Pep31 that was ob- tained from Lys-C cleavage of the protein from pea root plastids does not perfectly match the deduced amino acid sequence of pspor8 indicates that the 30-kDa protein band that had been subjected to enzymatic cleavage contains a t least two different porin homologues with similar molecular masses. This further supports the idea that mitochondria and non-green plastids (at least from pea and maize) contain a set of homologous porins. On the other hand, non-green plastids and chloroplasts contain obviously totally different porins and transport machineries for importing the respective porin proteins. Work is now in pro- gress to elucidate the underlying mechanism for this develop- mental diversity.

A

100 ,

1 2 3 4 5

fraction number

B Fraction No.

C P M 2 3 4 M, (k) 66-

45-

3o-D - -” 24+

FIG. 6. A, histogram of the distribution of marker enzymes of mito- chondrial and plastidial envelope membranes. Isolated pea root plastids that routinely contain a recovery of 30% of plastidial marker enzymes compared with a recovery of mitochondrial marker enzymes of 4% (Borchert et al., 1993) were separated on a sucrose density step gradient (0.4, 1.1.1.24, and 1.8 M). Five fractions were obtained and analyzed for

branes, respectively (see ”Materials and Methods”). NFO, NADH-ferri- the presence of mitochondrial and plastidial outer and inner mem-

cyanide reductase, outer mitochondrial membrane; COX, cytochrome c oxidase, inner mitochondrial membrane; ACS, acyl-CoA synthetase, outer plastidial membranes; ACT, acyl-CoA thioesterase, inner plas- tidial membranes. B, Western blot analysis of envelope membrane pro- teins from chloroplasts (C), pea root plastids (P), pea leaf mitochondria ( M ) , and of fractions 2-4 of the sucrose gradient (A) using an antibody raised against the yeast mitochondrial porin. Envelope membranes from chloroplasts (C, 20 pg) and pea root plastids (P, 20 pg), and mem- branes from pea leaf mitochondria (M, 20 pg), and aliquots (20 pg) of fractions 2-4 (A) were separated by SDS-PAGE and subsequently im- munodecorated using an antibody raised against the yeast mitochon- drial porin.

REFERENCES

Alban, C., Joyard, J., and Douce, R. (1988) Plant Physiol. (Rochuille) 88,709-717 Aljamal. J. A,, Genchi, G., De Pinto, V., Stefanazzi, L., De Santis, A., Benz, R., and

Andrews, J., and Keegstra, K. (1983) Plant Physiol. (Rochuille) 72,735-740 Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D.. Snuth, J. A,, Seidman, J.

G., and Struhl, K. (1987) Current Protocols in Molecular Biology, Greene Pub- lishing Associates and Wiley-Interscience, New York

Palmieri, F. (1993) Plant Physiol. (Rockuille) 102, 615421

Benton, D. W., and Davis, R. W. (1977) Science 196, 180-182 Benz, R. (1985) Crit. Rev. Biochem. 19, 145-190 Benz, R. (1994) Biochim. Biophys. Acta 1197, 167-196 Renz, R., Janko, K., Boos, W., and Liiuger, P. (1978) Riochim. Biophys. Acto 511,

Blachly-Dyson, E., Zambronicz, E. B., Yu, W. H., Adams, V., McCabe, E. R. B., 305-319

Borchert, S., Grosse, H., and Heldt, H. W. (1989) FERS Left. 253, 183-18fi Adelman, J., Colombini, M., and Forte, M. (1993) J. niol. Chem. 268,1835-1841

Borchert, S., Harborth, J., Schunemann, D.. Hoferichter, P., and Heldt, H. W.

Chen, E. Y., and Seeburg, P. H. (1985) DNA (N. Y ) 4, 165-170 Colombini, M. (1979) Nature 279,643-645 De Pinto, V., Prezioso, G., Thinnes, F., Link, T. A,, and Palmieri, F. (1991) Bio-

Douce, R., Mannella, C. A,, and Bonner, W. D., Jr. (1973) Biochim. Biophys. Acta

Eckerskon, C., and Lottspeich, F. (1989) Chromatographia 28,92-94 Eckerskon, C., Mewes, W., Goretzki, H., and Lottspeich, F. (1988)Eur. J. Biochem.

(1993) Plant Pl~ysiol. (Rockuille) 101, 303-312

chemistry 30,10191-10200

292, 105-116

176,509-519

25760 Porins from Plants Fischer, K., Weber, A,, Arbinger, B., Brink, S., Eckerskorn, C., and Fliigge, U. I.

Fliigge, U. I., and Benz, R. (1984) FEES Lett. 169,8549 Fliigge, U. I., Fischer, K., Gross, A,, Sebald, W., Lottspeich, F., and Eckerskorn, C.

Forte, M., Adelsberger-Mangan, D., and Colombini, M. (1987) J. Membr: Eiol. 99,

Fujiki, Y., Hubbard, A. L., Fowler, S., and Lazarow, P. B. (1982) J. Cell Eiol. 93,

Journet, E . 2 , and Douce, R. (1985) Plant Physiol. (Rockuille) 79,45%467 Joyard, J., and Stumpf, P. K. (1980) Plant Physiol. (Rockuille) 65, 1039-1043 Joyard, J., Billecocq, A,, Bartlett, S. G., Block, M. A,, Chua, N. H., and Douce, R.

Kayser, H., Kratzin, H. D., Thinnes, F. P., a t z , H., Schmidt, W. E., Eckart, K., and

Keegstra, K., Olsen, L. J., and Theg, S. M. (1989) Annu. Reu. Plant Physiol. Plant

Kleene, R., Pfanner, N., Pfaller, R., Link, T. A,, Sebald, W., Neupert, W., and

Klbsgen, R. B., Saedler, H., and Weil, J.-H. (1989) Mol. & Gen. Genet. 217,155-161

(1994) Plant Mol. Eiol. 25, 167-177

(1989) EMEO J. 8,3%46

65-72

97-102

(1983) J. Eiol. Chem. 258, 10000-10006

Hilschmann, N. (1989) Biol. Chem. Hoppe-Seyler 370,1265-1278

Mol. Eiol. 40, 471-501

Tropschug, M. (1987) EMEO J. 6,2627-2633

Laemmli, U. K. (1970) Nature 227,680-685 Li, H., Moore, T., and Keegstra, K. (1991) Plant Cell 3, 709-717 Lubben, T. H., and Keegstra, K. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 5502-

Lucas-Lenard, J., and Lipmann, F. (1971) Annu. Reu. Biochern. 40, 409448 Ludwig, O., Krause, J., Hay, R., and Benz, R. (1988) Eur: Eiophys. J. 15,269-276 Mihara, K., and Sato, R. (1985) EMEO J. 4, 769-774 Mourioux, G., and Douce, R. (1981) Plant Physiol. (Rockuille) 67,470-473 Salomon, M., Fischer, K., Fliigge, U. I., and Soll, J. (1990) Proc. Natl. Acad. Sei.

Schein, S. J., Colombini, M., and Finkelstein, A. (1976) J. Mernbr: Eiol. 30,99-120 Schmid, A,, Kromer, S., Heldt, H. W., and Benz, R. (1992) Eiochim. Ewphys. Acta

Schiinemann, D., Borchert, S. , Fliigge, U. I., and Heldt, H. W. (1993)PlantPhysiol.

Smack, D., and Colombini, M. (1985) Plant Physiol. (Rockville) 79, 1094-1097 Troll, H., Malchow, D., Miiller-Taubenberger, A,, Humbel, B., Lottspeich, F., Ecke,

M., Gerisch, G., Schmid, A,, and Benz, R. (1992) J. Eiol. Chem. 267, 21072- 21079

5506

U. S. A. 87,6778-6782

1112, 17P180

(Rockuille) 103, 131-137