The Plant Cell, Vol. 2, 195-206, March 1990 O 1990 American Society of Plant Physiologists

Acyl Carrier Protein (ACP) lmport into Chloroplasts Does not Require the Phosphopantetheine: Evidence for a Chloroplast Holo-ACP Synthase Michael D. Fernandez’ and Gayle K. Lamppab3’ Department of Biochemistry and Molecular Biology, University of Chicago, 920 East 58th Street, Chicago, lllinois 60637 Department of Molecular Genetics and Cell Biology, University of Chicago, 920 East 58th Street, Chicago, lllinois 60637

a

lmport of the acyl carrier protein (ACP) precursor into the chloroplast resulted in two products of about 14 kilodalton (kD) and 18 kD when analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Time, course experiments indicate that the latter is a modification derivative of the 14-kD peptide after the removal of the transit peptide. Substitution of serine 38 by alanine, eliminating the phosphopantetheine prosthetic group attachment site of ACP, produced a precursor mutant that gave rise to only the 14-kD peptide during import, showing that the modified form depends on the presence of serine 38. Furthermore, these results demonstrate that the prosthetic group is not essential for ACP translocation across the envelope or proteolytic processing. Analysis of the products of import by nondenaturing, conformationally sensitive gels showed reversal of the relative mobility of the 14-kD peptide and the modified form, raising the possibility that the modification is the addition of the phosphopantetheine. Proteolytic processing and the modification reaction were reconstituted in an organelle-free assay. The addition of coenzyme A to the organelle-free assay completely converted the 14-kD peptide to the modified form at 10 micromolar, and this only occurred with the wild-type substrate. Reciprocally, treatment of the products of a modification reaction with Escherichia coli phosphodiesterase converted the modified ACP form back to the 14-kD peptide. These results strongly support the conclusion that there is a holo-ACP synthase in the soluble compartment of the chloroplast capable of transferring the phosphopantetheine of coenzyme A to ACP.

INTRODUCTION Acyl carrier protein (ACP) is a key component of the fatty acid synthetase complex in all organisms studied. ACP functions as a carrier of growing fatty acid chains, which are attached to the protein through a 4‘-phosphopantetheine prosthetic group. ACP is an acidic protein but, in contrast to ACP of fungi and animals, which exist as part of a large multi-functional fatty acid synthetase polypeptide, the ACPs of plants and bacteria are small (9 kD to 1O kD), nonassociated soluble proteins. The amino acid sequences of Escherichia co// (Vanaman, Wakil, and Hill, 1968), spinach leaf (Kuo and Ohlrogge, 1984), and barley leaf [H0j and Svendsen, 1983) ACPs have been determined and exhibit extensive homology, particularly in the sequence surrounding the phosphopantetheine binding site, at a “special” serine residue that is 38 amino acids from the N terminus in spinach ACP. In higher plants there are at least two isoforms of the protein that are expressed in a tissue-specific manner (H0j and Svendsen, 1984; Ohlrogge and Kuo, 1985) and may have different functions (Guerra, Ohlrogge, and Frentzen, 1986).

’ To whom correspondence should be addressed

In higher plant leaves, ACP is localized in the chloroplast, where it appears that de novo fatty acid biosynthesis occurs (Ohlrogge, Kuhn, and Stumpf, 1979; Stumpf, 1980). lsolation and analysis of cDNAs generated from poly A’ selected RNA for a spinach ACP indicates that the genes for ACP are nuclear-encoded (Scherer and Knauf, 1987). Similar results have been presented for Brassica ACP (Safford et al., 1988). Nuclear-encoded chloroplast proteins are generally synthesized in the cytoplasm as precursor polypeptides with N-terminal transit peptides and post-translationally imported into the organelle, where they are proteolytically processed to their mature forms (Schmidt and Mishkind, 1986). The transit peptide of spinach ACP was tentatively identified as a 56-amino acid sequence (Scherer and Knauf, 1987) that is typically rich in serine, threonine, and the basic amino acids arginine and lysine. The study of the import of proteins into the chloroplasts of higher plants has focused primarily on proteins involved in the reactions of photosynthesis (for review, see Keegstra, 1988). The analysis of the import and processing of an essential protein not involved in photosynthesis, such

196

The Plant Cell

as ACP, might, therefore, add new insights to our understanding of these events, as well as confirm those more general features of the import pathway. Furthermore, posttranslational modification of ACP provides an opportunity to study another aspect of this complex process. A recent report (Elhussein, Miernyk, and Ohlrogge, 1988) indicates that the enzyme holo-ACP synthase, responsible for the transfer of the phosphopantetheine prosthetic group from COA to ACP, is found in the cytoplasm of spinach leaf cells. This observation raised the possibility that the addition of the phosphopantetheine is an essential step for ACP import, conferring an import competent conformation on the ACP precursor (pre-ACP). Only a minor amount of holo-ACP synthase activity was found associated with the chloroplast. On the other hand, apo-ACP has been found in the chloroplast upon overexpression of spinach ACP in transformed tobacco plants (Post-Beittenmiller, Schmid, and Ohlrogge, 1989). We have undertaken an analysis of the import and proteolytic processing of spinach ACP. Mutational analysis provides direct evidence that the addition of the prosthetic group is not required for the import of the precursor protein. Furthermore, our results show that, upon import, the processed protein is rapidly modified to a slower migrating form. The modification is dependent on the special serine at position 38. In the presence of a soluble chloroplast extract that is capable of processing pre-ACP, the modification is stimulated by the addition of COA and Mg*+. Treatment of the modified form with E. coli ACP phosphodiesterase regenerates the 14-kD peptide. These results strongly support the hypothesis that the chloroplast itself indeed contains a holo-ACP synthase activity responsible for the addition of the phosphopantetheine prosthetic group. The implications of this are discussed in terms of the function of ACP.

Figure 1. Structure of plBI-ACP500 Plasmid Used for in Vitro RNA Synthesis and Strategy To Produce the Mutant ACPASer.

The spinach ACP cDNA sequence was placed 3' to the T7 promoter as diagrammed. The plasmid was linearized with EcoRl for RNA synthesis with T7 RNA polymerase.The DNA sequence flanking the serine 38 codon (see asterisk) of the wild-type gene is shown, along with the corresponding amino acids for which it codes. The oligonucleotide sequence used to produce the serineto-alanine substitution is presented immediately below, and the nucleotides changed are indicated in bold face. The bottom line gives the sequence of the mutant precursor, pre-ACPASer. The smaller arrow indicates the position of the diagnostic Mlul restriction site created immediately5' to the new alanine codon.

RESULTS lmport of Pre-ACP into Chloroplasts Yields Two Processed Forms

A 500-bp Hindlll-EcoRI fragment of the plasmid pCGNl SOL (Scherer and Knauf, 1987) containing a cDNA sequence coding for the spinach ACP I protein as well as a putative transit peptide sequence of 56 amino acids was subcloned into the plasmid vector ~16131.Figure 1 shows the resulting construct, which was linearized with EcoRl and used as a template to synthesize pre-ACP RNA in vitro. As shown by SDS-PAGE in Figure 2A, lane 1, translation of this RNA by a reticulocyte lysate in the presence of "S-methionine produced a single band with an apparent molecular mass of about 21 kD, which is about 6 kD larger than expected for the ACP precursor based on amino acid composition, but consistent with previous SDS-PAGE es-

timates of ACP precursor protein translated from spinach mRNA (Ohlrogge and Kuo, 1984). AI1 acyl carrier proteins migrate with an anomalously high apparent molecular weight, presumably because of their small size and relative lack of SDS binding sites (Rock and Cronan, 1979). To investigate the ability of the putative transit peptide sequence to direct ACP into chloroplasts, radiolabeled precursor protein was incubated with intact chloroplasts isolated from young spinach leaves. After incubation, half of the organelles were treated with the protease thermolysin to remove any proteins on the exterior of the chloroplasts, and then both the protease-treated and nontreated chloroplasts were washed, lysed, and separated into their membrane and soluble components. As shown in Figure 2A, lane 5 , the putative transit peptide was functional and sufficient to direct ACP transport into the chloroplast. From the single precursor, there arose two products in the

Chloroplast Import of ACP

A Spinach

B pea

I

2

3

1

2

T.P.

M

M T

T.P.

M

3

4

5

M

S

S

T

197

was carried out in which spinach Chloroplasts were incubated with radiolabeled precursor protein for 1 min, 5 min, and 15 min, as illustrated in Figure 3. We argued that, if the upper band represented an intermediate form, it should appear first and then gradually give rise to the lower, fully processed approximately 14-kD form over time. We saw the opposite effect where the lower band appeared in greater abundance at the 1 min time point and then the ratio rapidly shifted so that by 15 min the upper band was the most prominent. This result indicated that the higher molecular weight product was a derivative of the lower molecular weight form, suggesting a covalent modification of the 14-kD product upon import of ACP into the Chloroplasts.

T

Figure 2. in Vitro Import of Pre-ACP into Chloroplasts Yields Two Products. (A) The reticulocyte lysate translation products of pre-ACP (T.P., lane 1) were incubated with spinach Chloroplasts, and the total membrane (M, lane 2), membrane after thermolysin treatment (M + T, lane 3), total soluble (S, lane 4), and soluble proteins after thermolysin treatment (S + T, lane 5) were analyzed. The open arrow indicates the position of the 14-kD product and the filled arrow indicates the position of the 18-kD product. (B) In vitro import of pre-ACP into pea Chloroplasts. The lanes are numbered the same as (A).

soluble fraction, both lower in molecular weight and resistant to thermolysin treatment, indicating that the precursor was proteolytically processed and sequestered to the stromal compartment of the chloroplast during the import reaction. The size of the lower band, which we estimated as about 14 kD, was consistent with the removal upon import of the 56-amino acid putative transit peptide by the chloroplast processing activity, i.e., "transit peptidase." The appearance of the upper band (about 18 kD), typically the more abundant of the two products, was unexpected. Significantly, the relative ratio of these two bands was variable. Although not yet systematically analyzed in detail, our results suggest that the ratio of these two products is sensitive to seasonal changes affecting the physiological/ developmental state of the plant. In contrast to spinach, Chloroplasts isolated from young pea leaves consistently generated more of the lower molecular weight product (see Figure 2B). The nature of these two products of preACP import is investigated in the experiments described below. The Larger Import Product Is not a Proteolytic Processing Intermediate To explore the possibility that the upper band was a precursor processing intermediate, a time course of import

Conversion of the "Special" Serine to Alanine Does not Prevent Import but Eliminates Production of the Larger Import Product Spinach ACP is known to be modified by the addition of a 4'-phosphopantetheine prosthetic group attached, via a phosphodiester bond, exclusively at a serine 38 amino acids from the start of the mature protein. Evidence has been presented that the major site of this addition is the cytoplasm (Elhussein et al., 1988), but it has not been excluded that there is also addition of the prosthetic group in the chloroplast itself. To determine whether the slower migrating, apparently higher molecular weight product that we observed upon import resulted from a modification of ACP through the special serine, site-directed mutagenesis

1

2

3

4

T.P.

M

M T

S

1 min

5

S T

2

M

3

M T

4

S

5 min

5

S + T

2

M

3

4

M + T

5

S +

15 min

Figure 3. Time Course of Import Indicates That the Larger Product Is a Derivative of the 14-kD Peptide. Reticulocyte lysate translation products are shown in lane 1, and lanes 2 to 5 are the same as those described in Figure 2 and as indicated below each lane. Translation products of pre-ACP were incubated with Chloroplasts for 1 min, 5 min, and 15 min as indicated below each reaction. The positions of the 14-kD and modified form are indicated by open and filled arrows, respectively.

198

The Plant Cell

was used to eliminate the phosphopantetheine binding site by converting serine 38 to alanine (see Figure 1 and Methods). In addition to the serine-to-alanine conversion, the mutation also included a silent single-nucleotide change immediately 5' to the new alanine codon that created a unique Mlul restriction site, providing a rapid method of screening for mutated plasmids. A restructured plasmid containing the desired substitution was transcribed and translated as previously described for wild-type pre-ACP. The radiolabeled mutant precursor, referred to as pre-ACPASer, was used in an in vitro import reaction. If imported and processed, it should, by definition, give rise to apo-ACP with a single residue change because the phosphopantetheine cannot be added. The results of this experiment are shown in Figure 4. The radiolabeled ACPASer precursor protein co-migrated with wild-type pre-ACP during SDS-PAGE (Figure 4, lanes 1). The products of the import reaction, however, were strikingly different. As before, the wild-type precursor gave rise to two thermolysin-resistant products in the soluble fraction. On the other hand, the pre-ACPASer mutant produced only one band, migrating slightly faster than the approximately 14-kD product of wild-type import. Thus, the serine-to-alanine conversion resulted in a loss of the higher molecular weight derivative, indicating that the

ACP wild type 1

T.P.

2

Organelle-Free Processing Yields Two Products That Co-Migrate with Those from Import Reactions

ACP ASer 3

M M

4

5

1

S

T.P.

T

2 3 4

M

smaller form of pre-ACP import is apo-ACP. Two important points emerged from this mutational analysis. First, the larger import product arises from a modification that depends on the presence of serine 38. Second, the phosphopantetheine prosthetic group, which cannot be added to pre-ACPASer, is not required for ACP translocation across the chloroplast envelope or for proteolytic processing upon import. The minor difference in mobility of the lower molecular weight product of wild-type ACP import and the product of pre-ACPASer import is most likely due to an altered conformation caused by the single serine-to-alanine mutation. A similar change in electrophoretic mobility has been identified for a mutant ACP that has Ser 38 converted to cysteine (Jaworski, Post-Beittenmiller, and Ohlrogge, 1989b). Predictions of E. coli ACP secondary and tertiary structure (Rock and Cronan, 1979; Mayo, Tyrell, and Prestegard, 1983) suggest that the special serine resides at the base of a central hydrophobic corridor formed by four a-helices. Predictions of spinach ACP secondary structure are consistent with this model (M. D. Fernandez, unpublished data; Kuo and Ohlrogge, 1984). The substitution of the nonpolar alanine residue for the polar serine in this critical region might, therefore, change the overall conformation and migration of the protein.

5

M T

Figure 4. In Vitro Import of the Pre-ACPASer Mutant Produces Only the 14-kD Form, Defined as Apo-ACP. Parallel in vitro import reactions were carried out using radiolabeled pre-ACP and pre-ACPASer proteins from reticulocyte lysates. Translation products are shown in lanes 1, and lanes 2 to 5 are as described in Figure 2 and as indicated below each lane. The positions of apo-ACP and the modified form are indicated by open and filled arrows, respectively.

We have recently developed an assay for organelle-free processing of the precursor of light-harvesting chlorophyll a/b-binding protein in which the transit peptide is removed (Lamppa and Abad, 1987; Abad, Clark, and Lamppa, 1989). In addition, the crude soluble extract processes the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase precursor, as has been observed in analogous experiments by Robinson and Ellis (1984). Using a similar protocol, we have prepared chloroplast soluble extracts from spinach that contain an active processing enzyme (see Methods). Our goal was to establish whether direct removal of the transit peptide of pre-ACP, independent of translocation across the chloroplast envelope, generated the same products as the in vitro import reactions. Figures 5A and 5B show the products of organelle-free processing reactions, using the wild-type and ACPASer precursors synthesized in a wheat germ lysate, compared with the soluble products of import reactions, analyzed on a denaturing SDS-polyacrylamide gel. The organelle-free assay (lanes 2) generated products of the same size as the import reactions (lanes 1) for both the wild-type and mutant substrates. That is, pre-ACP was cleaved and again gave two products, whereas preACPASer yielded only one peptide, the smaller form. In contrast to the import reactions, however, the soluble extracts always produced more of the 14-kD peptide than of the higher molecular weight form from the wild-type

Chloroplast Import of ACP

A

B 1 2 3 4

1 2 3 4

D

Figure 5. An Organelle-Free Assay Reconstitutes Removal of the Transit Peptide and the Modification Reaction. Analysis by denaturing and conformationally sensitive gels. (A) Analysis of the wild-type precursor products on denaturing gels. The soluble products of an import reaction using reticulocyte lysate translation products from thermolysin-treated chloroplasts (lane 1), and the products of organelle-free processing reactions using translation products from a wheat germ lysate and soluble chloroplast extracts prepared from pea (lane 2) and spinach (lane 3) leaves were analyzed on a 15% SDS-polyacrylamide gel. Lane 4 shows an organelle-free reaction to which CoA was added at a final concentration of 10 ^M. The positions of apo-ACP and the modified form are indicated by open and filled arrows, respectively. (B) Analysis of the pre-ACPASer products on denaturing gels. Reactions identical to those described for the wild-type precursor were carried out and the lanes are numbered the same. (C) Analysis of the wild-type precursor products on nondenaturing, conformationally sensitive gels. Samples from parallel reactions to (A) were analyzed on 20% conformationally sensitive gels. The lanes are numbered as above, and the positions of apoACP and the modified form are indicated by open and filled arrows, respectively. (D) Analysis of the pre-ACPASer products on conformationally sensitive gels. Reactions identical to those described for the wildtype precursor were carried out and the lanes are numbered the same.

199

substrate. These results support the conclusion that the 14-kD form is the primary product of the simple removal of the transit peptide. In addition, the organelle-free assay exhibits a modifying activity, yielding the 18-kD peptide, but it is less active than in intact organelles. Parallel reactions were carried out in the presence of a soluble extract from pea chloroplasts (Figures 5A and 5B, lanes 3), and we usually found that, with a pea extract, less of the higher molecular weight form was produced than upon addition of the spinach extract. This difference in the ability of the spinach and pea extracts to produce the modified form is consistent with the results obtained using intact organelles (see Figure 2B).

Conformationally Sensitive Gel Analysis of Import and Processing Reaction Products It has been well documented that modification of ACP through serine 38 leads to bands of altered mobility during electrophoresis. This phenomenon has been best characterized using E. coli ACP derivatives on nondenaturing or "conformationally sensitive" gels (Rock and Cronan, 1981; Rock, Cronan, and Armitage, 1981), and has also been applied to the analysis of spinach ACP derivatives (Jaworski, Clough, and Barnum, 1989a). ACP undergoes a pHinduced conformational change upon transition from the stacking gel (pH 6.8) to the separating gel (pH 9.0) that is characterized by the loss of «-helical content (Schulz, 1975) and an expansion of the Stokes radius (Rock and Cronan, 1979). Therefore, ACP derivatives that destabilize the protein moiety migrate more slowly because of this partial denaturation, and, conversely, ligands that stabilize the protein migrate faster. These studies have shown that the conformational changes induced by addition of the phosphopantetheine prosthetic group to the unmodified protein, apo-ACP, to produce holo-ACP and further addition of various acyl chains yielding acyl-ACP cause predictable shifts in mobility on nondenaturing gels. The order of relative mobility of these conformers in this system is acyl-ACP > holo-ACP > apo-ACP. To explore the nature of the ACP modification, the products of both the import and processing reactions were analyzed on conformationally sensitive gels. Figures 5C and 5D show an analysis of the same samples as in Figures 5A and 5B. As on the denaturing gels, we saw a single band of identical mobility in both organelle-free processing and import reactions using pre-ACPASer as the substrate (see Figure 5D, lanes 1 to 3). Two products were again detected from the wild-type precursor, but their relative mobility was reversed on the conformationally sensitive gel. The modified form of ACP, the slower migrating band on denaturing gels, which is the major product of import reactions (Figure 5C, lane 1) and the minor product of processing reactions (Figure 5C, lanes 2 and 3), was now seen to migrate ahead of apo-ACP, as represented by the

200

The Plant Cell

pre-ACPASer products (Figure 5D, lanes 1 to 3). That is, the modification that produces the upper band on denaturing gels causes a conformational change that enhances the mobility of the protein on nondenaturing gels. Importantly, the 14-kD peptide and apo-ACP (pre-ACPASer product) have not changed their relative positions, as would have been expected if the 14-kD peptide carried the phosphopantetheinegroup, i.e., was holo-ACP, supporting the conclusion that their different mobilities are due to the serine-to-alanine substitution. Hence, the enhancement of the mobility of the modified protein in relation to apo-ACP on conformationally sensitive gels, its slower migration during SDS-PAGE, and the observation that its presence is dependent on serine 38 strongly suggested that the modification of the 14-kD peptide is due to the addition of the phosphopantetheine prosthetic group. CoA-Dependent Modification of the 14-kD Peptide The enzyme holo-ACP synthase catalyzes the transfer of the phosphopantetheine group from CoA to apo-ACP in £. co// (Alberts and Vagelos, 1966; Elovson and Vagelos, 1968) and higher plants (Elhussein et al., 1988), although the properties of the latter reaction are not well understood. In addition to CoA as a substrate, the reaction also requires divalent cations such as Mg2+. We examined the effects of the addition of MgCI2 and CoA to the organellefree assays. MgCI2 stimulated production of the modified form at as little as 1 mM, with a parallel loss of the 14-kD peptide (data not shown). We initially observed complete conversion to the modified form upon the addition of CoA to a final concentration of 0.5 mM (data not shown), indicating that, although Mg2* ions stimulate modification, the level of CoA in the assay appears to be a limiting factor for this step. To establish more clearly the CoA-dependent modification of ACP in these reactions, CoA was added at a final concentration of 0.1 nM to 500 ^M. When pre-ACP was used as the substrate, the 14-kD peptide appeared as the primary product at low levels ( u

60 -

U

0.2

0.4

0.6

0.8

1.0

CoA concentration

B ACP wild type

0

0.1

0.5 1.0

ACPASer

10

0

0.1

0.5

1.0

10

[CoA] in u.M Figure 6. Modification Reaction Is Dependent on CoA in the Organelle-Free Assay. (A) A graphic representation of the results shown in (B), where percentage conversion to the modified form (as determined from densitometry scan of gel shown below) is plotted against the CoA concentration in the organelle-free assay. Pre-ACP (Q) and preACPASer (O) were analyzed. (B) Organelle-free reactions were carried out using wheat germ lysate translation products of pre-ACP and pre-ACPASer as substrates with the addition of CoA ranging from 0.1 uM to 10 ^M as indicated below each lane. The positions of apo-ACP (open arrow) and modified form (filled arrow) are indicated.

Chloroplast Import of ACP

We have noted that the wheat germ lysate possesses a low level of the Chloroplast processing enzyme capable of cleaving various precursors, suggesting that it may also contain other Chloroplast enzymes. In fact, mock reactions using 5 mM Hepes-KOH, pH 8.0, instead of the spinach extract have shown that the wheat germ lysate indeed contributes to the modification reaction converting the 14kD peptide to the 18-kD form. However, when the wheat germ lysate was boiled to inactivate its enzymatic activity, the CoA-dependent modification in the organelle-free assay still occurred with the spinach extract (data not shown). To establish unequivocally that the spinach extract is capable of carrying out the modification independent of the components of the wheat germ lysate, reactions with and without the addition of CoA were performed using the translation products of a reticulocyte lysate. Neither the processing nor the modification activities were found in mock reactions containing the reticulocyte products and 5 mM Hepes-KOH, pH 8.0, substituted for the spinach extract. When the spinach extract was added without CoA, pre-ACP gave rise to only the 14-kD peptide. Upon addition of CoA, however, at least 40% of the 14-kD peptide was converted to the modified form, as shown in Figure 7. Both the processing and modification activities were lost when the spinach extract was boiled (data not shown).

1

2

3

4

5

6

7

201

Figure 8. E. co/i Phosphodiesterase Reverses the CoADependent Modification of the 14-kD Peptide. An organelle-free assay was carried out using wheat germ translation products and 10 p.M CoA to convert all of the 14-kD peptide (open arrow) to the modified form (closed arrow). The products (lane 1) were then dialyzed, subdivided into four fractions, and employed in a phosphodiesterase reaction (see Methods and text). Lanes 2 through 5 show a mock reaction containing no enzyme (lane 2), and reactions using 20 units (lane 3), 40 units (lane 4), or 400 units (lane 5) of E. coli phosphodiesterase.

E. coli ACP Phosphodiesterase Regenerates the 14-kD peptide from the Modified Form

4

Figure 7. Organelle-Free Processing and CoA-Dependent Modification of Reticulocyte Translation Products. Translation products of pre-ACP (lane 1) were incubated in organelle-free reactions without (lanes 2, 4, and 6) and with (lanes 3, 5, and 7) the addition of CoA to 100 ^M, as indicated by minus and plus signs below each lane. The reactions contained either 5 mM Hepes-KOH, pH 8.0, instead of spinach Chloroplast extract (lanes 2 and 3), spinach Chloroplast extract clarified by ultracentrifugation (lanes 4 and 5), or crude soluble extract from lysed chloroplasts (lanes 6 and 7). The positions of apo-ACP and modified form are indicated by open and filled arrows, respectively.

To investigate further the identity of the modified form generated in the CoA-dependent reaction, the products of an organelle-free assay were treated with ACP phosphodiesterase purified from E. coli (Therisod and Kennedy, 1987). This enzyme is specific for holo-ACP, catalyzing the cleavage of the phosphodiester bond linking the phosphopantetheine to the special serine, and does not react with other substrates such as CoA (Vagelos and Larrabee, 1967). An organelle-free reaction containing pre-ACP synthesized in a wheat germ lysate and 10 nM CoA was performed as described above, in which the 14-kD peptide was completely converted to the 18 kD product, as illustrated in Figure 8, lane 1. The products of this reaction were dialyzed against 50 mM Tris-HCI, pH 8.5, to remove CoA and used as the substrate for the ACP phosphodiesterase enzyme. Phosphodiesterase reactions were carried out using 20 units (Figure 8, lane 3), 40 units (lane 4), and 400 units (lane 5) of the enzyme. The units of activity are based on the ability of this enzyme to remove the phosphopantetheine from the homologous substrate, bacterial holo-ACP

202

The Plant Cell

(Therisod and Kennedy, 1987), and are given to provide a relative measure of the efficiency of the bacterial enzyme using the spinach ACP substrate (see Methods). A mock reaction, containing no enzyme, was included to demonstrate that the phosphodiester bond remains stable under the reaction conditions in the absence of the phosphodiesterase (lane 2). In the reaction containing 400 units of the E. coli phosphodiesterase, approximately 30% of the modified ACP product was converted back to the 14-kD unmodified form. The sensitivity of the CoA-dependem modification to the phosphodiesterase, i.e., the conversion of the 18-kD peptide back to the 14-kD form, indicates that the modified form is holo-ACP. Hence, the results of the organelle-free assay, coupled with those from the import reactions, strongly support the conclusion that the chloroplast contains an enzyme capable of transferring the phosphopantetheine group of COA to ACP, producing hOlO-ACP.

DISCUSSION We have demonstrated in this study that the transport of ACP into the chloroplasts does not require the attachment of its phosphopantetheine prosthetic group. Substitution of serine 38, to which the phosphopantetheine is usually attached, does not prevent efficient import. Significantly, our results indicate that the chloroplast itself contains an enzyme, i.e., holo-ACP synthase, capable of adding the functional prosthetic group. Evidence has recently been presented that plant cells contain a cytoplasmic holo-ACP synthase that transfers the phosphopantetheine group from COA. However, it is not known where this reaction normally occurs in vivo, and a chloroplast enzyme could not be excluded (Elhussein et al., 1988). The addition of prosthetic groups to various proteins targeted to organelles can occur both before and after import. Among proteins destined for the mitochondria, for example, pyridoxal phosphate is added to aspartate aminotransferase in'the mitochondrial matrix after its translocation and proteolytic processing (Sharma and Gehring, 1986). Similarly, ferredoxin is imported and processed to its mature form by chloroplasts before incorporation of the iron-sulfur complex (Takahashi et al., 1986). On the other hand, heme appears to be added to cytochrome c at the inner face of the mitochondrial outer membrane before its localization to the intermembrane space (Nicholson and Neupert, 1989). and addition of FAD to ferredoxin-NADP+ oxidoreductase, which is destined for the chloroplast thylakoids, has also been reported to occur in the cytoplasm (Carrillo, 1985). One might predict that the addition of the prosthetic group to ACP confers on the precursor a conformation that favors efficient import into the chloroplast. Current evidence indicates that features that stabilize precursor tertiary structure, such as ligand binding, block protein transport into the mitochondria (Eilers and Schatz, 1986),

whereas destabilization by mutation or precursor unfolding by cytosolic factors (Vestweber and Schatz, 1988; Cheng et al., 1989) promote the import process. Although the phosphopantetheine apparently stabilizes the conformation of mature ACP (Rock et al., 1981), it is difficult to predict the structural changes that the precursor, containing a 56-amino acid transit peptide with a basic composition, would undergo upon phosphopantetheine addition. In any case, if the phosphopantetheine interacts with the protein moiety to confer a unique tertiary structure, the conformation is not essential to mediate pre-ACP import into the chloroplast. Pre-ACP is processed to two forms upon import into the chloroplast. Our results indicate that the 14-kD peptide arises from the removal of the ACP transit peptide by the chloroplast processing enzyme. The slower migrating peptide, however, is a derivative of the 14-kD peptide produced by a highly specific modification reaction. Four criteria indicate that the modification is due to the attachment of the phosphopantetheine to apo-ACP by a chloroplast holo-ACP synthase. First, the production of the modified form depends on the presence of serine 38, to which the prosthetic group is normally attached. Second, compared with denaturing gels, conformationally sensitive gels show a reverse in the relative mobility of the 14-kD peptide and modified form as predicted for the relationship between apo-ACP and holo-ACP. Third, COA rapidly stimulates the modification reaction in an organelle-free assay. Fourth, most convincingly, the modified peptide is converted back to the 14-kD peptide by E. coli phosphodiesterase. In addition, the conformationally sensitive gels indicate that the newly imported and proteolytically processed ACP does not carry the phosphopantetheine based on the mobilities of apo-ACP generated from pre-ACPASer and the 14-kD peptide cleaved from pre-ACP. That is, the 14-kD peptide does not migrate ahead of apo-ACP, as predicted if it carriad the prosthetic group, thereby maintaining a more stable conformation. Finally, it has been demonstrated that the transfer of acyl groups such as malonyl from malonyl-COA, for example, to holo-ACP by an ACP transacylase is not affected by the addition of COA at low concentrations (less than 10 gM), and, at higher concentrations, the transfer is inhibited (Guerra and Ohlrogge, 1986). It is, therefore, improbable that the modified form is the result of a two-step process requiring first the synthesis of malonyl-COA and then the transfer reaction. Based on these considerations, we conclude that a holoACP synthase performs the modification reaction. The question arises, what would be the function of two holo-ACP synthases-one cytoplasmic and the other chloroplastic-in the plant cell? What are their substrates and respective roles in terms of converting ACP to its active form? If the phosphopantetheine is usually added in the cytoplasm immediately upon pre-apo-ACP synthesis, then what is the function of the chloroplast enzyme? One possibility is that there is considerable turnover of the

Chloroplast lmport of ACP

prosthetic group regenerating pools of apo-ACP in the chloroplast under certain conditions, which would then be available as a substrate for holo-ACP synthase. This would be one way of regulating de novo fatty acid biosynthesis. In E. coli, the turnover of the phosphopantetheine appears to be low (Jackowski and Rock, 1984), although reported rates have differed by an order of magnitude (Powell, Bauza, and Larrabee, 1973), and the cells do not appear to maintain a steady-state amount of apo-ACP (Jackowski and Rock, 1983). Regulation of fatty acid biosynthesis by the alteration of the apo-ACP to holo-ACP ratio, then, may not be a strategy employed in E. coli (see Jackowski and Rock, 1984). However, it is interesting to note that recently a new function for ACP in the synthesis of membranederived oligosaccharides in E. coli has been described in which the phosphopantetheine is not required, suggesting an alternative physiological role for apo-ACP (Therisod and Kennedy, 1987). In higher plants, the levels of apo-ACP and holo-ACP have not been extensively examined, and the possibility that prosthetic group turnover represents another mechanism for control of fatty acid biosynthesis cannot be discounted. We have been able to reconstitute the COA-dependent modification reaction in an organelle-free assay. The chloroplast extract typically employed was clarified by ultracentrifugation to remove all membranes, and, thus, we conclude that the modifying activity, i.e., holo-ACP synthase, is most likely located in the soluble stromal phase of the chloroplast or, perhaps, the space separating the outer and inner membranes. The conditions of the organelle-free assay were originally optimized for removal of the transit peptide (Lamppa and Abad, 1987; Abad et al., 1989). Thus, apo-ACP, the 14-kD peptide, is the predominant product of the reaction when pre-ACP is the substrate. The COA concentration curve suggests that the 14kD peptide can be converted to the modified form, i.e., the mature form of ACP, lacking a transit peptide, may serve as a substrate for phosphopantetheine addition in the chloroplast. Therefore, the following pathway for the synthesis and import of ACP should be considered: (1) preapo-ACP is synthesized in the cytoplasm, (2) pre-apo-ACP is imported and proteolytically processed to apo-ACP, and (3) the prosthetic group is transferred from COA to apoACP to produce holo-ACP within the chloroplast. The import reactions show that ACP is found predominantly in the stroma of the chloroplast. In a recent report, using immunogold labeling and electron microscopy, ACP was localized to the thylakoid membranes (Slabas and Smith, 1988). We conclude from our results that, if ACP does become membrane-associated, it must be a very transient association that is easily disrupted because we have observed essentially no radiolabeled protein in the membrane fractions from thermolysin-treated organelles after import. By using the same methods in other studies, however, we have shown that mutant polypeptides of the light-harvesting chlorophyll a/b-binding protein that be-

203

come peripherally associated with the thylakoids, but cannot insert, still fractionate with the membranes after chloroplast lysis (Clark, Abad, and Lamppa, 1989). A comparison of the spinach and E. coli ACPs shows that about one-third of their residues are identical, i.e., 29 of 82 amino acids of the spinach ACP. The most conserved domain extends almost continuously for 17 amino acids surrounding serine 38. This degree of relatedness is apparently sufficient for the bacterial phosphopantetheine to recognize spinach ACP, albeit less efficiently than the homologous substrate, and remove the phosphopantetheine. Systematic substitution of the conserved residues could now be employed to identify the minimum determinants necessary for phosphodiesterase activity. In summary, we have identified a modification activity in the soluble fraction of the chloroplast that converts a 14kD peptide to a slower migrating form that is dependent on the addition of COA. This modification reaction is reversible by treatment with E. coli ACP phosphodiesterase. Taken together, our results provide convincing evidence that the enzyme involved is a holo-ACP synthase. By using the organelle-free assay, we can now characterize the properties of this enzyme, unequivocally establish its substrate specificity, and begin its purification.

METHODS

Plant Growth and Chloroplast lsolation

Spinach (Spinacea oleracea, Melody) plants were grown in a greenhouse (Department of Ecology and Evolution, University of Chicago) and transferred to a growth chamber (26OC, cool fluorescent lights) 3 weeks to 4 weeks after planting and 1 week to 2 weeks before harvesting.A typical chloroplast isolation used 25 g of leaf tissue, fresh weight, homogenized with a Polytron homogenizer at 4OC in 2 mM EDTA, 1 mM MgC12,1 mM MnCI2,50 mM Hepes-KOH, pH 8.0, 0.33 M sorbitol, 5 mM sodium ascorbate, 0.25% BSA at a ratio of 20 mL/g of tissue, fresh weight. lntact chloroplasts were isolated on Percoll gradients (Bartlett, Grossman, and Chua, 1982). Construction of Plasmids

500-bp EcoRI-Hindlll fragment of a cDNA sequence (pCGNlSOL, kindly provided by Vic Knauf, Calgene, Davis, CA) coding for spinach ACP I (Scherer and Knauf, 1987) was inserted into the transcription vector plB131 (InternationalBiotechnologies Inc., New Haven, CT) downstream from the T7 promoter, yielding the plasmid plBI-ACP500. The 500-bp insert contains the sequence coding for the mature ACP I protein as well as a 56-amino acid putative transit peptide and 36 bp of the 5'- and 53 bp of the 3'-nontranslated sequences.

A

Single-Stranded DNA lsolation

Single-stranded DNA was isolated by inoculating 3 mL of 2 X YT media (1.6% tryptone, 17'0 yeast extract, 8.5 mM NaCI) containing

204

The Plant Cell

150 pg/mL ampicillin with a single colony of E. coli strain MV1193 containing the plBI-ACP500 plasmid.After 2 hr of growth at 37OC, the culture was inoculated with the helper phage M13K07 at a multiplicityof infection of 20, and growth was continued at 37°C for 45 min, at which time kanamycin was added to 80 pg/mL final concentration. After continued incubation at 37OC for 45 min, 1O0 pL of the culture was used to inoculate 1O mL of 2 x YT media (75 pg/mL kanamycin, 150pg/mL ampicillin),which was incubated at 37°C overnight. The cells were removed by centrifugation at 30009 for 1O min, and the phage particles were precipitated out of the supernatant at room temperature for 2 hr by the addition of 3% PEG, 0.25 M NaCI. The precipitated phage particles were collected by centrifugation at 77009 for 20 min, resuspended in 10 mM Tris, 0.1 mM EDTA, and extracted with phenol and chloroform. Single-strandedDNA was collected by ethanol precipitation and resuspended in 30 pL of 10 mM Tris, 0.1 mM EDTA. In Vitro Mutagenesis

The ACPASer mutation was created using a protocol, based on the methodof Ecksteinand co-workers(Taylor, Ott, and Eckstein, 1985a), designed to give high yields of mutant plasmids (Amersham Corp., Arlington Heights, IL). Briefly, an oligonucleotide (sequence shown in Figure l), complementary to the ACP cDNA sequence surrounding the sequence coding for the special serine but containing three nucleotide changes, was hybridized to the single-stranded plBI-ACP500 DNA and used as a primer for second-strand synthesis by the Klenow fragment of DNA polymerase I. In the resulting heteroduplex, the newly synthesized mutant strand was chemically distinguished from the original wildtype strand by the incorporation of dCTP&. Certain restriction endonucleases (Ncil was used here) are unable to cut phosphorothioate DNA, allowing selective nicking of the wild-type strand (Taylor et al., 1985b). Exonuclease 111 was then used to excise through the sequence on the wild-type strand complementary to the mutated sites. The gaps were repaired with DNA polymerase I, using the mutant strand as a template, creating the doublestranded mutant plasmid plBI-ACPASer. The incorporation of a silent, single-nucleotidechange immediately5' to the new alanine codon created a unique Mlul restrictionsite, allowing for rapid and accurate isolation of positive plasmids.

translation products. The chloroplasts were incubated with translation products at 26°C with gentle shaking for 45 min (except where indicated in text), and then diluted fivefold with HSM, spun at 25009 for 1 min, and immediately resuspended in cold 1 mM phenylmethylsulfonylfluoride (PMSF) for lysis. When chloroplasts were treated with thermolysin, half of the incubationreaction was mixed with an equal volume of 200 pg/mL thermolysin in HSM, 4 mM CaCI,. The reaction was stopped by making it 40 mM EGTA, 2 mM PMSF, and diluted threefold with HSM; the chloroplasts were pelleted at 25009 for 1 min and immediately lysed in 1 mM PMSF. To separate the membrane and soluble phases of the chloroplast, the samples were spun for 20 min in a microcentrifuge at 16000g. For SDS-PAGE analysis, the pellet was resuspended in 0.5 M Tris-HCI, pH 6.8, 5% 6-mercaptoethanol, 2% SDS, 10% glycerol. The protein from the supernatant was precipitated with 10% TCA, washed with 80% acetone, and resuspended in the same buffer. For nondenaturing gels, the samples were resuspended in 0.5 M Tris-HCI, pH 6.8, 1O mM DTT, 10% glycerol. Organelle-Free Reactions

Chloroplast soluble extracts were prepared according to the protocol of Abad et al. (1989). lntact chloroplasts were resuspended in HSM and centrifuged for 1 min at 25009. The pellet was gently resuspended in a volume of ice-cold 5 mM HepesKOH, pH 8.0, equal to the original HSM volume, and kept at 4°C for 30 min for lysis. The lysed chloroplasts were centrifuged at 160009for 1O min to remove the bulk membranes,and membrane vesicles were removed by recentrifuging the supernatant at 142,OOOg. A typical 25-pl processing reaction contained 20 mM Tris-HCI, pH 8.0, 3 pg/mL chloramphenicol, 5 pL of a 30-pL translation reaction (40,000 cpm to 50,000 cpm), and 15 pL of clarified soluble chloroplast extract, and was incubated at 26OC for 90 min. Both wheat germ and reticulocyte lysate translation products have been used; the wheat germ lysate generally translated the pre-ACP RNA more efficiently. The reactions were stopped by bringing the products to 0.5 M Tris-HCI, pH 6.8, 5% p-mercaptoethanol, 2% SDS, 10% glycerol for SDS-PAGE analysis, or to 0.5 M Tris-HCI, pH 6.8, 10 mM DTT, 10% glycerol for analysis on nondenaturing polyacrylamide gels. For reactions containing COA(free acid; Sigma, St. Louis, MO), a sterile solution was added to a final concentration as indicated in the text.

In Vitro Transcription, Translation, and lmport PhosphodiesteraseTreatment

The plBI-ACP500 and plBI-ACPASer plasmids were linearized with EcoRl and transcribed with T7 RNA polymerase (New England Biolabs, Beverly, MA). The RNA generated from a 50-pL reaction was resuspended in 25 pL of sterile water, and 2 pL to 4 pL (estimated 0.5 pg to 1.O pg) were used in a standard 30-pL reticulocyte (Bethesda Research Laboratories, Bethesda, MD) or wheat germ (Amersham Corp.) translation reaction as prescribed by the vendor, using 35S-methionine(1O00 Ci/mmol) as the labeled amino acid. Optimal translation was achieved using 33 mM Kacetate in both wheat germ and reticulocyte translation systems. Procedures for the import reaction were essentially as described (Bartlett et al., 1982). lntact chloroplasts were resuspended in HSM (50 mM Hepes-KOH, pH 8.0, 0.33 M Sorbitol, 8 mM methionine) at a concentration of 800 pg to 1000 pg of chlorophyll/mL, and 50 pL was added to a 300-pL reaction containing HSM, plus 10 mM ATP, 10 mM MgCI?,and 50 pL of

After organelle-freereactions containing wheat germ lysate translation products and 10 pM COA, the samples were dialyzed against 50 mM Tris-HCI, pH 8.5, on microdialysis filters (Type SV, 0.025-pm pore size; Millipore Corp., Bedford, MA) for 1 hr at room temperature to remove COA.They were then treated with purified E. coli ACP phosphodiesterase (kindly provided by Eugene Kennedy and Anthony Fishl, Harvard University, Cambridge, MA) essentially as described. The reactions contained 50 mM TrisHCI, pH 8.5, 25 mM MgCI?, 20 pM MnCI,, 1 mM DTT, and were incubated at 35'C for 2 hr with either 20 units, 40 units, 400 units or no enzyme as described in the text. Units of activity were expressed as picomoles per minute per milliliter on the basis of releaseof the TCA-soluble radiolabeledphosphopantetheine from the TCA-insolubleE. coli holo-ACP(Therisodand Kennedy, 1987). The reactions were stopped by bringing the products to 0.5 M

Chloroplast lmport of ACP

Tris-HCI, pH 6.8, 5% P-mercaptoethanol,2% SDS, 10% glycerol, and analyzed by SDS-PAGE. Gel Analysis For SDS-PAGE analysis, samples were boiled for 3 min and immediately loaded on 15% polyacrylamide gels as described (Laemmli, 1970; Piccioni, Bellemere, and Chua, 1982). Nondenaturing gels were essentially as described (Rock and Cronan, 1981). The separating gel contained 20% acrylamide, 0.5% N,Nmethylenebisacrylamide, 0.5% N,N,N’,N’-tetramethylethyldiamine, 0.375 M Tris-HCI, pH 9.0. A stacking gel was poured over the separating gel and contained the same components except that the acrylamide concentrationwas 5% and the Tris-HCI buffer was 0.125 M, pH 6.8. The gels were run at 7 mA overnight at 37OC, using a running buffer of 50 mM Tris, 400 mM glycine, until the tracking dye reached the bottom of the gel. Under these conditions, the precursor proteins do not substantially enter the gel. After electrophoresis, gels were stained with Coomassie Blue in 50% methanol, 7% acetic acid, destained, and prepared for autoradiography by soaking in DMSO for 1 hr, 2,5-diphenyloxazole-DMSO for 2.5 hr, water for 30 min, and then dried.

ACKNOWLEDGMENTS We wish to thank Drs. Anthony Fischl and Eugene Kennedy for kindly providing the purified E. coli ACP phosphodiesterase.We are grateful to Roben Buell for her excellent technical assistance. This work was supportedby U.S. Department of Agriculture Grant 89-3761-4471 (awarded to G.K.L.) and a Sigma Xi Grant-in-Aidof Research (to M.D.F.).

Received October 27, 1989; revised December 29, 1989.

REFERENCES Abad, M.S., Clark, S.E., and Lamppa, G.K. (1989). Properties of a chloroplast enzyme that cleaves the chlorophyll a/b binding protein precursor. Plant Physiol. 90, 117-1 24. Alberts, A.W., and Vagelos, P.R. (1966). Acyl carrier protein VIII. Studies of acyl carrier protein and coenzyme A in Escherichia coli pantothenate or palanine auxotrophs. J. Biol. Chem. 241, 5201-5204. Bartlett, S., Grossman, A.R., and Chua, N.-H. (1982). In vitro synthesis and uptakeof cytoplasmicallysynthesizedchloroplast proteins. In Methods in Chloroplast Biology, M. Edelman, R. B. Hallick, and N.-H. Chua, eds (New York: Elsevier Biomedical), pp. 1081-1 091. Carrillo, N. (1985). Biosynthesisof ferredoxin-NADP+oxidoreductase. Eur. J. Biochem. 150, 469-474. Cheng, M.Y., Hartl, F.-U., Martin, J., Pollock, R.A., Kalousek, F., Neupert, W., Hallber, E.M., Hallberg, R.L., and Horwich,

205

S.L. (1989). Mitochondrial heat-shock protein hsp6O is essential for assembly of proteins imported into yeast mitochondria. Nature 337, 620-625. Clark, S.E., Abad, M.S., and Lamppa, G.K. (1989). Mutations at the transit peptide-mature protein junction separate two cleavage events during chloroplast import of the chlorophyll a/bbinding protein. J. Biol. Chem. 264, 17544-17550. Eilers, M., and Schatz, G. (1986). Binding of a specific ligand inhibits import of a purified precursor protein into mitochondria. Nature 322, 228-232. Elhussein, S.A., Miernyk, J.A., and Ohlrogge, J.B. (1988).Plant holo-(acylcarrier protein) synthase. Biochem. J. 252, 39-45. Elovson, J., and Vagelos, P.R. (1968). Acyl carrier protein X. Acyl carrier protein synthetase. J. Biol. Chem. 243, 3603-3611. Guerra, D.J., and Ohlrogge, J.B. (1986). Partia1 purification and characterization of two forms of malonyl-coenzyme A: acyl carrier protein transacylase from soybean leaf tissue. Arch. Biochem. Biophys. 246, 274-285. Guerra, D.J., Ohlrogge, J.B., and Frentzen, M. (1986). Activity of acyl carrier protein isoforms in reactions of plant fatty acid metabolism. Plant Physiol. 82, 448-453. Haj, P.B., and Svendsen, 1.8. (1983). Barley acyl carrier protein: Its amino acid sequence and assay using purified malonylCoA:ACP transacylase. Carlsberg Res. Comm. 48,285-305. H0j, P.B., and Svendsen, I.B. (1984). Barley chloroplasts contain two acyl carrier proteins coded for by different genes. Carlsberg Res. Comm. 49,483-492. Jackowski, S., and Rock, C.O. (1983). Ratio of active to inactive forms of acyl carrier protein in Escherichia coli. J. Biol. Chem. 258, 15186-1 5191. Jackowski, S., and Rock, C.O. (1984). Turnover of the 4’phosphopantetheine prosthetic group of acyl carrier protein. J. Biol. Chem. 259, 1891-1895. Jaworski, J.G., Clough, R.C., and Barnum, S.R. (1989a). A cerulenin insensitive short chain 3-ketoacyl-acyl carrier protein synthase in Spinacia oleracea leaves. Plant Physiol. 90,41-44. Jaworski, J.G., Post-Beittenmiller, M.A., and Ohlrogge, J.B. (1989b). Site-directed mutagenesis of the spinach acyl carrier protein I prosthetic group attachment site. Eur. J. Biochem. 184,603-609. Keegstra, K. (1988). Transport and routing of proteins into chloroplasts. Cell 56, 247-253. Kuo, T.M., and Ohlrogge, J.B. (1984). The primary structure of spinach acyl carrier protein. Arch. Biochem. Biophys. 234, 290-296. Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Lamppa, G.K., and Abad, M.S. (1987). Processing of a wheat light-harvesting chlorophyll a/b protein precursor by a soluble enzyme from higher plant chloroplasts. J. Cell Biol. 105, 2641-2648. Mayo, K.H., Tyrell, P.M., and Prestegard,J.H. (1983). Acyl carrier proteinfrom Escherichiacoli. 1. Aspects of the solutionstructure as evidencedby proton nuclear overhauser experiments at 500 MHz. Biochemistry 22, 4485-4493. Nicholson, D.W., and Neupert, W. (1989). lmport of cytochrome

206

The Plant Cell

c into mitochondria: Reduction of heme, mediated by NADH and flavin nucleotides, is obligatory for its covalent linkage to apocytochrome c. Proc. Natl. Acad. Sci. USA 86, 4340-4344. Ohlrogge, J.B., and Kuo, T.M. (1984). Spinach acyl carrier protein: Primary structure, mRNA translation and immunoelectrophoretic analysis. In Structure, Function and Metabolism of Plant Lipids, P.-A. Siegenthalerand W. Eichenberger,eds (Amsterdam: Elsevier Science Publishers, BV), pp. 63-67. Ohlrogge, J.B., and KUO,T.M. (1985). Plants have isoforms for acyl carrier protein that are expressed differently in different tissues. J. Biol. Chem. 260, 8032-8037. Ohlrogge, J.B., Kuhn, D.N., and Stumpf, P.K. (1979). Subcellular localization of acyl carrier protein in leaf protoplasts of Spinacia oleracea. Proc. Natl. Acad. Sci. USA 76, 1194-1198. Piccioni, R., Bellemere, G., and Chua, N.-H. (1982). Methods of polyacrylamide gel electrophoresis in the analysis and preparation of plant polypeptides. In Methods in Chloroplast Biology, M. Edelman, R. Hallick, and N.-H. Chua, eds (New York, Elsevier Biomedical), pp. 985-1 O14. Post-Beittenmiller, M.A., Schmid, K.M., and Ohlrogge, J.B.

(1989). Expressionof holo and apo forms of spinach acyl carrier protein-l in leaves of transgenic tobacco plants. Plant Cell 1, 889-899. Powell, G.L., Bauza, M., and Larrabee, A.R. (1973). The stability of acyl carrier protein in Escherichia coli. J. Biol. Chem. 248, 4461-4466. Robinson, C., and Ellis, R.J. (1984). Transport of proteins into chloroplasts: Partia1 purification of a chloroplast protease involved in the processing of imported precursor polypeptides. Eur. J. Biochem. 142,337-342. Rock, C.O., and Cronan, J.E., Jr. (1979). Re-evaluation of the solution structure of acyl carrier protein. J. Biol. Chem. 254, 9778-9785. Rock, C.O., and Cronan, J.E., Jr. (1981). Acyl carrier protein from Escherichia coli. Methods Enzymol. 71, 341-351. Rock, C.O., Cronan, J.E., Jr., and Armitage, I.M. (1981). Molecular properties of acyl carrier protein derivatives. J. Biol. Chem.

256,2669-2674. Safford, R., Windust, J.H.C., Lucas, C., desilva, J., James, C.M., Hellyer, A., Smith, C.G., Slabas, A.R., and Hughes, G.

(1988). Plastid-localized seed acyl-carrier protein of Brassica napus is encoded by a distinct, nuclear multigene family. Eur.

J. Biochem. 174,287-295. Scherer, D.E., and Knauf, V.C. (1987). lsolation of a cDNA clone for acyl carrier protein I of spinach. Plant MOI.Biol. 9,127-134. Schmidt, G.W., and Mishkind, M.L. (1986). The transport of

proteins into chloroplasts. Annu. Rev. Biochem. 55, 879-91 2. Schulz, H. (1975). On the structure-function relationship of acyl carrier protein of Escherichia coli. J. Biol. Chem. 250,

2299-2304. Sharma, C.P., and Gehring, H. (1986). The precursor of mito-

chondrial aspartate aminotransferase is translocated into mitochondria as apoprotein. J. Biol. Chem. 261, 11146-1 1149. Slabas, A.R., and Smith, C.G. (1988). lmmunogold localization of acyl carrier protein in plants and Escherichia coli: Evidence for membrane association in plants. Planta 175, 145-152. Stumpf, P.K. (1980). Biosynthesis of saturated and unsaturated fatty acids. In The Biochemistry of Plants: A Comprehensive Treatise. Vol. 4. Lipids: Structure and Function, P.K. Stumpf, ed (New York, Academic Press), pp. 177-204. Takahashi, Y., Mitsui, A., Hase, T., and Matsubara, H. (1986). Formation of the iron-sulfur cluster of ferredoxin in isolated chloroplasts. Proc. Natl. Acad. Sci. USA 83, 2434-2437. Taylor, J.W., Ott, J., and Eckstein, F. (1985a). The rapid generation of oligonucleotide-directed mutations at high frequency using phosphorothioate-modifiedDNA. Nucl. Acids Res. 13, 8765-8785. Taylor, J.W., Schmidt, W., Cosstick, R., Okrusrek, A., and Eckstein, F. (1985b). The use of phosphorothioate-modified

DNA in restriction enzyme reactions to prepare nicked DNA. Nucl. Acids Res. 13, 8749-8764. Therisod, H., and Kennedy, E.P. (1987). The function of acyl carrier protein in the synthesis of membrane-derived oligosaccharides does not require its phosphopantetheine prosthetic group. Proc. Natl. Acad. Sci. USA 84, 8235-8238. Vagelos, P.R., and Larrabee, A.R. (1967). Acyl carrier protein IX. Acyl carrier protein hydrolase. J. Biol. Chem. 242, 1776-1781. Vanaman, T.C., Wakil, S.J., and Hill, R.L. (1968). The complete amino acid sequence of the acyl carrier protein of Escherichia coU. J. Biol. Chem. 243, 6420-6431. Vestweber, D., and Schatz, G. (1988). Point mutations destabilizing a precursor protein enhance its post-translational import into mitochondria. EMBO J. 7, 1147-1 151.

Acyl carrier protein (ACP) import into chloroplasts does not require the phosphopantetheine: evidence for a chloroplast holo-ACP synthase. M D Fernandez and G K Lamppa Plant Cell 1990;2;195-206 DOI 10.1105/tpc.2.3.195 This information is current as of May 19, 2015 Permissions

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw15322 98X

eTOCs

Sign up for eTOCs at: http://www.plantcell.org/cgi/alerts/ctmain

CiteTrack Alerts

Sign up for CiteTrack Alerts at: http://www.plantcell.org/cgi/alerts/ctmain

Subscription Information

Subscription Information for The Plant Cell and Plant Physiology is available at: http://www.aspb.org/publications/subscriptions.cfm

© American Society of Plant Biologists ADVANCING THE SCIENCE OF PLANT BIOLOGY