Proc. Nati. Acad. Sci. USA Vol. 84, pp. 3288-3292, May 1987 Cell Biology Protein import into requires a ATPase (cell-free translation/precursor of small subunit of pea ribulose-1,5-bisphosphate carboxylase/pea chloroplasts/ nonhydrolyzable ATP analogs/ionophores) DEBKUMAR PAIN AND GUNTER BLOBEL Laboratory of Cell Biology, Howard Hughes Medical Institute, The Rockefeller University, New York, NY 10021 Contributed by Gunter Blobel, January 29, 1987

ABSTRACT We have transcribed mRNA from a cDNA the rbcS-E9 gene (9) and contains the entire mature polypep- clone coding for pea ribulose-1,5-bisphosphate carboxylase, tide (S) sequence, including two introns. A portion of translated the mRNA in a wheat germ cell-free system, and pUC3AE9 containing the two introns but not the transit studied the energy requirement for posttranslational import of sequence was excised with Sph I and Kpn I and an identical the [35S]methionine-labeled protein into the of pea Sph I-Kpn I fragment from SS15, which lacked the two chloroplasts. We found that import depends on ATP hydrolysis introns, was substituted. The resulting clone contained the within the stroma. Import is not inhibited when HI, K+, Na+, complete uninterrupted coding region of pS. or divalent cation gradients across the chloroplast membranes The pS coding DNA was excised from the pUC vector with are dissipated by ionophores, as long as exogenously added Xba I and Cla I and ligated to pT7-1 cut with Xba I and Acc ATP is also present during the import reaction. Our data I. The resulting plasmid (pT7-pS) was purified on a sucrose suggest that protein import into the chloroplast stroma requires gradient and linearized downstream from the gene with Pst I a chloroplast ATPase that does not function to generate a before transcription. In vitro transcription using phage T7 membrane potential for driving the import reaction but that RNA polymerase (United States Biochemicals, Cleveland, exerts its effect in another, yet-to-be-determined, mode. We OH) was done essentially according to supplier's assay have carried out a preliminary characterization of this ATPase conditions except that the nucleoside triphosphates and the regarding its specificity and the effects of various cap analog diguanosine triphosphate [G(5')ppp(5')G] were ATPase inhibitors. used at 0.25 mM each. After transcription, the solution was extracted with phenol/chloroform and the mRNA was pre- The in vitro import of cytoplasmically synthesized proteins cipitated with LiCl. About 2 ,ug oftranscript was obtained per into chloroplasts has been previously demonstrated. Trans- ,ug of DNA template (as judged by A260). lation of mRNA for a cytoplasmically synthesized polypep- Cell-Free Translation. The previously described wheat tide of the chloroplast stroma, the small subunit (S) of germ cell-free translation system (10) was supplemented with ribulose-1,5-bisphosphate carboxylase, yielded a large pre- 10 ,uM S-adenosyl-L-methionine and contained 200 ng of cursor (pS) (1), and posttranslational incubation with isolated mRNA and 2.8 ,ul of wheat germ extract per 10-,ul translation chloroplasts resulted in import of pS into the chloroplast mixture. The final concentrations ofthe other components of stroma accompanied by conversion of pS into S (2, 3) by the translation mixture were adjusted to 43 mM Hepes-KOH means of cleavage of an NH2-terminal transit sequence (4). at pH 7.5, 112 mM KOAc, 2.1 mM Mg(OAc)2, 0.25 mM Import is an energy-dependent process (5). When the spermidine, 3 mM dithiothreitol, 0.5 mM ATP, 20 ,uM GTP, import reaction was carried out in the dark, addition of ATP 8 mM creatine , creatine kinase at 65 ,ug/ml, each stimulated uptake into the chloroplast. Likewise, illumina- of the 20 amino acids except for methionine at 26 ,uM, and tion of the import reaction stimulated uptake, and this [35S]methionine at 0.9-1.2 mCi/ml (1 Ci = 37 GBq). After stimulation was abolished when uncouplers were present, 1-hr incubation at 25°C, the translation mixture was centri- again suggesting that ATP is required for import (5). fuged for 22 min at 4°C in a Beckman Airfuge at 30 psi Recently, Cline et al. (6) succeeded in experimentally (140,000 x g), yielding a postribosomal supernatant (PRS) separating the import reaction into two steps. In the first that contained the newly synthesized pS. reaction precursors were bound to the surface of ATP- Posttranslational Import Assay. Intact chloroplasts were depleted chloroplasts. In a second reaction much of the isolated from the leaves of 2- to 3-wk-old pea (Pisum sativum, bound precursor was imported into the chloroplast, depend- Progress no. 9) seedlings and purified on a Percoll step ing on exogenously added ATP. gradient as described (11). Isolated chloroplasts were resus- In the present paper we have addressed the questions of pended in 50 mM Hepes-KOH, pH 7.7/0.33 M sorbitol at a how and where ATP is used to drive the import ofpS into the concentration of 2-4 mg of chlorophyll per ml. To deplete chloroplast stroma. chloroplasts of their endogenous ATP, we incubated them in the dark for 15 min at room temperature prior to use in the import assay. The import reaction was carried out essentially MATERIALS AND METHODS as in ref. 5 with some modifications. The basic import assay Subcloning and Phage T7 Transcription. A clone represent- mixture (300 ,u) contained 0.5 ul of PRS (see above), ing the complete coding sequence ofpS was constructed from preincubated chloroplasts equivalent to 100 ,g of chloro- phyll, bovine serum albumin at 100 ,ug/ml, and 50 mM a partial cDNA clone, SS15, that lacks part of the NH2- Hepes-KOH, pH 7.7/0.33 M sorbitol/40 mM KOAc/2 mM terminal transit sequence (7) and a genomic clone, pUC- Mg(OAc)2/1.5 mM dithiothreitol/10 mM methionine. In most 3AE9. The plasmid pUC3AE9 contains two S gene segments; instances, the import assay mixture was further modified as one is derived from the rbcS-3A gene (8) and contains the specified in the figure legends. The import reaction was complete transit peptide of pS, and the other is derived from Abbreviations: S, small subunit of ribulose-1,5-bisphosphate car- The publication costs of this article were defrayed in part by page charge boxylase; pS, precursor of S; PRS, postribosomal supernatant; CF1, payment. This article must therefore be hereby marked "advertisement" coupling factor; pmf, protonmotive force; CCCP, carbonylcyanide in accordance with 18 U.S.C. §1734 solely to indicate this fact. m-chlorophenylhydrazone. 3288 Downloaded by guest on September 26, 2021 Cell Biology: Pain and Blobel Proc. Natl. Acad. Sci. USA 84 (1987) 3289

carried out at room temperature for 30 min with occasional -<---- Dark - <- Room Light -O gentle mixing, either in the dark or in the room light as specified. Chloroplasts - + Quantitative Analysis of Import. After incubation, each ATP - _ I - import mixture was chilled on ice, diluted with 5 ml of cold Proteases - - - - - 50 mM Hepes-KOH, pH 7.7/0.33 M sorbitol and centrifuged X-100 + for 1 min at 4000 x g. The pelleted chloroplasts were lysed Triton ------in 0.8 ml of2 mM EDTA, pH 7.5, by vigorous mixing. Sodium chloride was added to a final concentration of 0.24 M and the lysate was centrifuged at 12,000 x g for 15 min. To the resulting supernatant, representing the stroma fraction, ice- cold trichloroacetic acid was added to a final concentration of 10%o. The precipitate was'analyzed by electrophoresis on 12% polyacrylamide gels in NaDodSO4 (12). The gels were fixed in 10% acetic acid/35% methanol (vol/vol), treated with Enlightning (New England Nuclear), dried, and exposed to S- Fuji'x-ray film for 6-24 hr at -80'C. An AMBIS ,8-scanner was used to quantitate the radioactivity in pS or S in dried gels. Protease Protection. The procedure for protease protection 1 2 3 4 5 6 7 was identical to that above for the given quantitative analysis FIG. 1. Protein import into the chloroplast stroma depends on of import, except that prior to lysis the pelleted chloroplasts ATP. Lane 1 shows the 3"S-labeled products present in a 0.5-,ui were gently resuspended in a final volume of 300 ,ul of aliquot of a PRS after translation. Lanes 2-7 show the products of a ice-cold 50 mM Hepes-KOH, pH 7.7/0.33 M sorbitol. Either 0.5-1.l aliquot of PRS imported into the chloroplast stroma fraction. 250 ,tg/ml each of trypsin and chymotrypsin or the same Where indicated, ATP was present at a final concentration of 1 mM. amount of proteases plus Triton X-100 to a final concentra- The data are from a fluorograph of a dried Na1DodSO4/polyacryl- tion of 0.3% was added. After incubation for 30 min on ice, amide gel. the reaction mixture was diluted with 0.8 ml of cold 2 mM EDTA, pH 7.5 containing protease inhibitors (2.5 mM p- ground" import (e.g., import in the dark without exogenously aminobenzamidine/10 mM e-amino-n-caproic acid/i mM added ATP) that was observed in the earlier studies (5), for phenylmethylsulfonyl fluoride), and mixed vigorously. which the ATP concentration contributed to the import reaction by the translation mixture was calculated to be 166 RESULTS AM instead of 0.8 AM in the present study. ATP Acts Within the Stroma. Chloroplasts are capable of ATP Is Required for Import. mRNA for S was obtained by synthesizing ATP in the dark when supplied with dihydroxy- in vitro transcription of cloned DNA (see Materials and acetone phosphate/oxalacetic acid/phosphate buffer (13). Methods) and was translated in a wheat germ cell-free Therefore, addition ofthese substrates to the import reaction system. The major translation product was a polypeptide of in the dark yielded efficient import (65-70%) (Fig. 2, lane 2). 20 kDa (Fig. 1, lane 1) that was immunoprecipitable with The effect of the ATP synthesized in the dark on the import antibody raised against purified S from pea (data not shown) could be exerted either in the stroma itself or (after ATP and that represented the precursor of S, pS. export via the ADP/ATP carrier located in the inner chlo- A small quantity (0.5 ,l) of PRS from the cell-free trans- roplast membrane) on the outside of the chloroplast [the lation was added posttranslationally to a 300-,ul import outer is presumably not a barrier for reaction mixture containing chloroplasts isolated from pea the rapid equilibration of ATP concentrations between the leaves. The maximal ATP concentration contributed by the cytosol (the "outside") and the intermembrane space (14)]. translation mixture in the import reaction was 0.8 ,uM. After To localize the effect of the ATP synthesized in the dark on incubation for import, chloroplasts were sedimented, lysed in import we used two external, nonpenetrating ATP-consum- 2 mM EDTA, and fractionated into a soluble (stroma) and a ing systems (12, 15, 16). One of the ATP-consuming systems membrane fraction. Upon import into the chloroplast stroma, consisted of glucose and hexokinase (Fig. 2, lanes 3 and 4), pS is converted into S by cleavage of its NH2-terminal transit the other one was apyrase (Fig. 2, lanes 6 and 7). We found sequence (4). that neither of these ATP-consuming systems inhibited im- When the import reaction was carried out in the dark and port that was stimulated by ATP synthesized in the dark. The without additional ATP (see above), less than 2% of pS was ATP-consuming systems were, in fact, active, since prein- found to be translocated (Fig. 1, lane 2). When the import cubation (30 min at room temperature) of exogenously added reaction was carried out either in the dark in the presence of ATP with either of the ATP-consuming systems strongly 1 mM ATP (lane 3) or in the room light in the absence of inhibited import in the dark in the absence of dihydroxyac- exogenously added ATP (lane 4), about 65% of pS was etone phosphate/oxalacetic acid/phosphate buffer (Fig. 2, translocated. In the room light and in the presence of lanes 5, 8, and 9). These data suggest that ATP exerts its exogenously added ATP, import of pS was 80% (lane 5). A effect on import within the chloroplast stroma, not on the control in which, after completion of import, the chloroplasts outside, and most likely not in the intermembrane space (see were incubated with proteases prior to lysis showed that above). virtually a11 of the S molecules in the stroma fraction were ATP Hydrolysis Is Required. To determine whether ATP resistant to protease digestion (lane 6) but were degraded hydrolysis is required for import we took advantage of the when Triton X-100 was present during the incubation with observation that exogenously added nonhydrolyzable ATP proteases (lane 7). Thus, the S molecules detected in the analogs can enter the chloroplast stroma and be exchanged stroma fraction were, in fact, imported. for stromal ATP via the ADP/ATP exchanger (17) and that The results shown in Fig. 1 confirm and amplify the data these nonhydrolyzable analogs can act as competitive inhib- reported by Grossman et al. (5). The virtually complete itors ofATP hydrolysis. Two non-hydrolyzable ATP analogs, dependence of import on ATP demonstrated by the data in 5'-adenylyl imidodiphosphate and 5'-[8,y-,methyl- Fig. 1 provides an explanation for the considerable "back- eneltriphosphate were tested in the light-induced import Downloaded by guest on September 26, 2021 3290 Cell Biology: Pain and Blobel Proc. Natl. Acad. Sci. USA 84 (1987)

ATP I 1 ATP - - 1 - 1 - 1

OAA/DHAP/Pi - + + + - + + - _ GTP - 10 10 - - - -

Glucose - - + + + - - - - dATP -- - 10 10 -

Hexokinase - - 2U 20U 20U - - - - ADP - - - - - 10 10

Apyrase - - - _ _U lOU IOU - ps-w- pS-

S-o l

So-_ 1 2 3 4 5 6 7

FIG. 4. Effect of some on import. The import reactions were carried out in the light. Numbers on top indicate the concentrations in mM. 1 2 3 4 5 6 7 8 9

FIG. 2. ATP exerts its effect within the chloroplasts and not from brane that, in turn, could drive import into the chloroplast the outside. The import reactions were carried out in the dark. Some stroma. A membrane potential across the inner chloroplast import reaction mixtures contained 1 mM dihydroxyacetone phos- membrane could be generated either indirectly, by the phate (DHAP)/1 mM oxalacetic acid (OAA)/10 mM potassium CF1-Fo ATPase (CF1, coupling factor) in the phosphate buffer, pH 7.5 (Pi) to induce light-independent ATP membrane, or directly, by a proton- or ion-translocating generation in chloroplast stroma. Other import reaction mixtures ATPase in the inner chloroplast membrane. contained in addition ATP-consuming systems (10 mM glucose and To determine whether ATP hydrolysis is employed to 2 or 20 units of hexokinase; 2 or 10 units of apyrase). Others generate a protonmotive force (pmf), we tested the effect of contained in addition 1 mM ATP. the protonophore carbonylcyanide m-chlorophenylhydra- zone (CCCP) (18), which dissipates the total pmf. CCCP, in reaction in the absence or presence oflow (1 mM) or high (10 fact, completely inhibited the light-induced import (Fig. 5, mM) concentrations of ATP (Fig. 3). In the absence of ATP lanes 2 and 3). Most importantly, however, addition of ATP (lanes 3 and 6) either of these two analogs caused consider- in the presence of CCCP restored import into the chloroplast able inhibition of the light-stimulated import (for control see (Fig. 5, lane 4). Thus, the exogenously supplied ATP can lane 1). This inhibition was partially or completely abolished substitute for the endogenously synthesized ATP. Exoge- at low or high ATP concentrations (lanes 4, 5, 7, and 8). nously added ATP presumably enters the stroma via the Effect of Other Nudeotides. We have tested the effect of ADP/ATP exchanger and, in the presence of CCCP-i.e., CTP, GTP, dATP, or ADP on the light-stimulated import in under conditions in which the pmf has been abolished- the absence or presence of ATP (Fig. 4). CTP (10 mM, data serves as a substrate for an ATPase other than the CFj-Fo not shown) or GTP (10 mM) strongly inhibited light-stimu- ATPase in the thylakoid membrane or a proton-translocating lated import in the absence ofATP (lane 2), but inhibition was ATPase in the inner chloroplast membrane to accomplish greatly relieved in the presence of 1 mM ATP (lane 3). The import. mechanism by which GTP or CTP inhibits the putative To investigate whether a K+- or Na'-translocating ATPase stroma ATPase involved in import remains to be determined. might be involved in import we tested nigericin and dATP (lanes 4 and 5) showed only a slight inhibition of monensin, respectively. Cline et al. (6) had already demon- light-induced import, suggesting that dATP might be able to strated that light-induced uptake into chloroplasts is inhibited substitute for ATP as a substrate. ADP (lanes 6 and 7) had no by nigericin and that uptake is restored by exogenously added effect on light-induced import, presumably because it can be

converted to ATP in the chloroplast stroma. ATP - - - 1- I - -1 - - - 1

A Membrane Potential Is Not Required for Import. ATP CCCP 100 100 - - ______gener- hydrolysis in the chloroplast stroma could be used to Nilgricin -- - 100 100 ------ate a membrane potential across the inner chloroplast mem- monen -.20 o - - - - -

A23 7 - 1.7 3.3 6.7 67 ATP 1 - 1 10 - 1 10

AMP-PNP - - 1 1 1 - - - Pus ---PS APPC14 - - - - - 11 1

pS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

FIG. 5. A membrane potential is not required tor import. The import reactions were carried out in the light. Chloroplasts (equiv- alent to 100 ug of chlorophyll in 200 gl) were first incubated with 1 2 3 4 5 6 7 8 various ionophores for 15 min on ice in the room light. The import reactions were started by adding 100 1.l of a master mixture FIG. 3. Nonhydrolyzable ATP analogs inhibit import. The import containing all the other components of the basic import reaction. reactions were carried out in the light. Numbers on top indicate Numbers on top indicate the final concentrations in the import concentrations in mM of ATP or of nonhydrolyzable ATP analogs. reaction of the ionophores in AM (CCCP, monensin, and A23187) or AMP-PNP, 5'-adenylyl imidodiphosphate; APPCH2P, adenosine nM (nigericin) and ofATP in mM. Fluorographs comprising lanes 1-9 5'-[a3,v-methylene]triphosphate. and 10-15 were from separate experiments. Downloaded by guest on September 26, 2021 Cell Biology: Pain and Blobel Proc. Natl. Acad. Sci. USA 84 (1987) 3291

ATP. The data shown in lanes 5-7 of Fig. 5 confirm the 20-25% inhibitory in the absence ofATP (lane 10) and had no findings of Cline et al. (6). Nigericin dissipates a K+ gradient effect in the presence of added ATP (lane 11). across the membrane by catalyzing an electroneutral ex- change of K+ and H' leading to a breakdown of the pH gradient (ApH) without affecting the membrane potential DISCUSSION (AT) (19, 20). Thus, in the presence of nigericin at two Our data here suggest the existence of an ATPase located in different concentrations (10 or 100 nM, lanes 5 and 6, (or facing) the chloroplast stroma that is required for protein respectively) there was very little or no import. However, import from the cytoplasm into the chloroplast stroma. when ATP was present in addition to 100 nM nigenrcin, We reached this conclusion after we first extended earlier import was restored (lane 7). Monensin (19, 20), which acts studies (5) and demonstrated that ATP is required for import like nigericin except that it is more selective for Na', also into the chloroplast stroma (Fig. 1). When the posttransla- inhibited import (lane 8). Again, import was restored when tional import reaction was carried out in the dark where no ATP was also present (lane 9). These data suggest that neither ATP was synthesized in the chloroplast stroma, import was a K+ nor a Na' gradient is involved in import. found to be dependent on exogenously added ATP. When the Similar results were obtained with the ionophore A23187, import reaction was carried out in the room light, apparently which is predominantly selective for divalent over monova- enough ATP was endogenously synthesized to support the lent cations (19, 20). In the absence of ATP, light-induced import reaction. However, light-induced import was further uptake of pS was inhibited by A23187 in a dose-dependent stimulated by exogenously added ATP, suggesting that the manner (Fig. 5, lanes 12-14). The inhibition was virtually ATP synthesized endogenously during illumination (in the complete at 6.7 AM A23187 (lane 14) and it could be relieved room light) might not have reached optimal levels for import. by 1 mM ATP (lane 15). Inhibition of light-induced import by We then proceeded to demonstrate that the required ATP A23187 is in good agreement with the ability ofthe ionophore acts from within the chloroplast, presumably in the chloro- to dissipate the pH gradient and thereby to interfere with plast stroma (Fig. 2). This was done by adding substrates endogenous ATP production (19). (dihydroxyacetone phosphate/oxalacetic acid/Po) so that Effect of Various ATPase Inhibitors. Since ATP hydrolysis chloroplasts could synthesize ATP in the dark (13). Sub- was obligatory for import, it was of interest to test various strate-stimulated chloroplasts were able to import. Addition inhibitors of (Fig. 6). NaF and Na3VO4, 10 mM and of nonpenetrating ATP-consuming systems (glucose plus 1 mM respectively, gave about 35-45% inhibition of light- hexokinase or apyrase) did not abolish import, strongly induced pS uptake (lanes 2 and 4). In both cases the inhibition suggesting that the ATP synthesized in the substrate-stimu- could be reversed by exogenously added ATP (lanes 3 and 5). lated chloroplasts was not required on the outside of the The light-induced import of pS was inhibited by =7% when chloroplast, where it could have been exported to by the 1 mM inorganic pyrophosphate (PPO) was present in the ADP/ATP exchanger of the inner chloroplast membrane. We import assay (lane 6). Higher concentration of PP1 (10 mM) cannot rule out the possibility that the ATP was exported to led to about 20-30% inhibition (data not shown). A prein- and acted within the space between the inner and outer cubation of PP1 with was able to restore chloroplast membrane. However, we consider this highly translocation to the normal level (data not shown). The unlikely as molecules of up to 8000 daltons are thought to slightly inhibitory effect of PPj was also abolished in the rapidly equilibrate between the intermembrane space and the presence of 1 mM ATP (lane 7). These results are consistent cytoplasm (14), presumably through large pores in the outer with the much lower affinity of the adenine nucleotide membrane. transporter for PPi than for ATP (21). Oligomycin (20 AM) Hydrolysis of ATP is required for import. Nonhydrolyz- was ineffective to inhibit light-induced translocation to an able ATP analogs were shown to inhibit light-induced import appreciable extent (<8%) (lane 8). Oligomycin is known to (Fig. 3). These analogs were most likely transported into exert only a weak uncoupling effect in chloroplasts (22). chloroplast stroma in exchange for endogenous adenine Quercetin, a powerful inhibitor of the purified ATPase nucleotides via the ADP/ATP exchanger (17) and then activity of the CF1 from spinach chloroplasts (23), was only competitively inhibited ATP hydrolysis in the stroma. ATP hydrolysis was not used to generate a H+, Na+, K+, ordivalent cation ATP - 1 I I I I (Ca2+, Mg2+, etc.) gradient across the inner chloroplast membrane (Fig. 5), suggesting that a cationmo- NF - 10 10 - - - _ _ _ _ _ tive ATPase is unlikely to be involved in the import reaction. NaeVO4 - - - In the presence of ionophores dissipating either the ApH

- - - - alone or both pp, - I - - - ApH and At components of the total pmf, - - exogenously added ATP was still able to drive the import almyi - - - - - 20 20 - --in reaction. Thus a membrane potential does not appear to be - - Qurcetn -.- - - - 25 2A required for import into the chloroplast stroma. Therefore, the proton-translocating activity of the CFj-F0 ATPase in the thylakoid membrane is not directly involved in import. It is also unlikely that the ATPase activity of the CF1-F0 is responsible for ATP hydrolysis obligatory for P-- protein import, since hydrolysis by the CF1-Fo ATPase would then have to be coupled to some reaction other than generating a pmf. Moreover, quercetin, a potent inhibitor of 1 2 3 4 5 6 7 9S X0 I U ATPase activity of spinach CF1 (23), failed to inhibit import in the presence of added ATP (Fig. 6). FIG. 6. Effect of various ATPase inhibitors on import. The Likewise, a recently characterized ATPase in the inner import reactions were carried out in the light. Chloroplasts (equiv- chloroplast alent to 100 ,ug of chlorophyll in 200 ul) were first incubated with membrane (24, 25) does not appear to be a likely various ATPase inhibitors for 15 min on ice in the room light. The candidate for the ATPase activity required for import. The import reactions were started by adding 100 a4 of a master mixture former is known to have a broad substrate specificity, containing all the other components of the basic import reaction. whereas the latter appears to be inhibited by GTP (Fig. 4) and Numbers on top indicate the concentrations in mM for NaF, CTP (data not shown). Moreover, the inner membrane Na3VO4, PP1, and ATP, or in AtM for oligomycin and quercetin. ATPase has been shown to be inhibited (60-70%) by fluoride Downloaded by guest on September 26, 2021 3292 Cell Biology: Pain and Blobel Proc. Natl. Acad. Sci. USA 84 (1987) or vanadate (25), which at a more or less similar concentra- 3. Highfield, P. E. & Ellis, R. J. (1978) Nature (London) 27i, tion had no effect on import stimulated by exogenously added 420-424. ATP (Fig. 6). 4. Schmidt, G. W., Devillers-Thiery, A., Desruisseaux, H., Blo- Thus, the ATPase activity required for must be a bel, G. & Chua, N.-H. (1979) J. Cell Biol. 83, 615-622. import 5. Grossman, A., Bartlett, S. & Chua, N.-H. (1980) Nature hitherto uncharacterized ATPase. Since this ATPase is (London) 285, 625-628. apparently not involved in generating a membrane potential, 6. Cline, K., Werner-Washburne, M., Lubben, T. H. & Keegstra, other modes of action have to be considered. There are K. (1985) J. Biol. Chem. 260, 3691-3696. several, certainly not mutually exclusive, possibilities. One 7. Coruzzi, G., Broglie, R., Cashmore, A. & Chua, N.-H. (1983) possibility (12) is that ATP is hydrolyzed by a protein (a J. Biol. Chem. 258, 1399-1402. "") that would act as a mechanochemical trans- 8. Fluhr, R., Moses, P., Morelli, G., Coruzzi, G. & Chua, N.-H. ducer by coupling the energy obtained from ATP hydrolysis (1986) EMBO J. 5, 2063-2071. to the movement ofproteins across the chloroplast envelope. 9. Coruzzi, G., Broglie, R., Edwards, C. & Chua, N.-H. (1984) ATP a EMBO J. 3, 1671-1679. Other possibilities are that is hydrolyzed by kinase, by 10. Erickson, A. H. & Blobel, G. (1983) Methods Enzymol. 96, an adenylyltransferase, or by adenylate cyclase. A kinase or 38-50. an adenylyltransferase could modify protein(s) necessary for 11. Orozco, E. M., Jr., Mullet, J. E., Hanley-Bowdoin, L. & translocation. Likewise, cAMP could be used to activate a Chua, N.-H. (1986) Methods Enzymol. 118, 232-253. cAMP-dependent kinase, which in turn could activate a 12. Waters, M. G. & Blobel, G. (1986) J. Cell Biol. 102, 1543-1550. protein(s) involved in translocation. Phosphorylation may 13. Inoue, Y., Kobayashi, Y., Shibata, K. & Heber, U. (1978) also occur on the chains to be translocated. Transit se- Biochim. Biophys. Acta 504, 142-152. quences contain a substantial number of serine and threonine 14. Foyer, C. H. (1984) (Wiley, New York), p. residues (7). Their phosphorylation could somehow be cou- 105. pled to active transport by analogy to the 15. Krishnan, P. S. (1949) Arch. Biochem. 20, 272-283. phosphotransferase 16. Chen, L. & Tai, P. C. (1985) Proc. Natl. Acad. Sci. USA 82, system for sugar transport in bacteria (26). 4384-4388. Recently, other cell-free translocation systems in which 17. Robinson, S. P. & Wiskich, J. T. (1977) Biochim. Biophys. translocation of proteins can occur posttranslationally have Acta 461, 131-140. been analyzed as to their energy requirement. Similar to our 18. Heytler, P. G. (1979) Methods Enzymol. 55, 462-472. results on in vitro import into chloroplast stroma, in vitro 19. Reed, P. W. (1979) Methods Enzymol. 55, 435-454. translocation across inverted vesicles of the bacterial plasma 20. Pressman, B. C. (1976) Annu. Rev. Biochem. 45, 501-530. membrane (16) and microsomal vesicles ofyeast endoplasmic 21. Robinson, S. P. & Wiskich, J. T. (1977) Plant Physiol. 59, reticulum (12, 27, 28) have been shown to require ATP but not 422-427. a membrane potential. Only in vitro import into mitochondria 22. Avron, M. & Shavit, N. (1965) Biochim. Biophys. Acta 109, appears to depend on both ATP and a membrane potential 317-331. 23. Deters, D. W., Racker, E., Nelson, N. & Nelson, H. (1975) J. (29). Biol. Chem. 250, 1041-1047. We thank Dr. N.-H. Chua for his gift of the clones pUC3AE9 and 24. McCarty, D. R., Keegstra, K. & Selman, B. R. (1984) Plant SS15, Dr. Randolph Addison for useful discussions, Ms. Elizabeth Physiol. 76, 584-588. Orr for constructing the plasmid pT7-pS, Ms. Gisele Nimic for 25. McCarty, D. R. & Selman, B. R. (1986) Plant Physiol. 80, helping with the illustrations, and Arup Mukherjee for typing the 908-912. manuscript. This work was supported by National Institutes of 26. Meadow, N. D., Kukuruzinska, M. A. & Roseman, S. (1984) Health Grant GM 27155 to G.B. and a Damon Runyon-Walter in The of Biological Membranes, ed. Martonosi, Winchell Fellowship to D.P. A. N. (Plenum, New York), Vol. 3, pp. 523-559. 1. Dobberstein, B., Blobel, G. & Chua, N.-H. (1977) Proc. Natl. 27. Hansen, W., Garcia, P. D. & Walter, P. (1986) Cell 45, Acad. Sci. USA 74, 1082-1085. 397-406. 2. Chua, N.-H. & Schmidt, G. W. (1978) Proc. Nati. Acad. Sci. 28. Rothblatt, J. A. & Meyer, D. I. (1986) EMBO J. 5, 1031-1036. USA 75, 6110-6114. 29. Pfanner, N. & Neupert, W. (1986) FEBS Lett. 209, 152-156. Downloaded by guest on September 26, 2021