ARCHIVES

OF BIOCHEMISTRY

Vol. 276, No. 1, January,

AND

BIOPHYSICS

pp. 38-46,199O

Concentrations of Long-Chain Acyl-Acyl Carrier Proteins during Fatty Acid Synthesis by Chloroplasts Isolated from Pea (Pisum sativum), Safflower (Carthamus tinctoris), and Amaranthus (Amaranthus lividus) Leaves Grattan

Roughan*”

and Ikuo Nishidat

*Division of Horticulture and Processing, DSIR, Mt. Albert Research Centre, Private Bag, Auckland, New Zealand, and tNationa1 Institute for Basic Biology, 38 Nishigonaka, Okazaki 444, Japan

Received April 17,1989, and in revised form August 5, 1989

Fatty acid synthesis from [1-14C]acetate by chloroplasts isolated from peas and amaranthus was linear for at least 15 min, whereas incorporation of the tracer into long-chain acyl-acyl carrier protein (ACP) did not increase after 2-3 min. When reactions were transferred to the dark after 3-5 min, long-chain acyl-ACPs lost about 90% of their radioactivity and total fatty acids retained all of theirs. Half-lives of the long-chain acyl-ACPs were estimated to be lo-15 s. Concentrations of palmitoyl-, stearoyl-, and oleoyl-ACP as indicated by equilibrium labeling during steady-state fatty acid synthesis, ranged from 0.6-1.1,0.2-0.7, and 0.4for peas and from 1.6-1.9, 1.31.6 MM, respectively, 2.6, and 0.6-1.4 MM, respectively, for amaranthus. These values are based on a chloroplast volume of 47 /.d/ mg chlorophyll and varied according to the mode of the incubation. A slow increase in activity of the fatty acid synthetase in safflower chloroplasts resulted in longchain acyl-ACPs continuing to incorporate labeled acetate for 10 min. Upon re-illumination following a dark break, however, both fatty acid synthetase activity and acyl-ACP concentrations increased very rapidly. Palmitoyl-ACP was present at concentrations up to 2.5 PM in safflower chloroplasts, whereas those of stearoyland oleoyl-ACPs were in the lower ranges measured for peas. Acyl-ACPs were routinely separated from extracts of chloroplasts that had been synthesising longchain fatty acids from labeled acetate by a minor modification of the method of Mancha et al. (Anal. Biochem., 1975, 68, 600-608). The results compared favorably with those obtained using alternative analytical methods such as adsorption to filter paper and partition chromatography on silicic acid columns. The acyl-ACP which coprecipitated with ammonium sulfate was not ’ To whom correspondence 38

should be addressed.

affected by treatments with neutral hydroxylamine or borohydride, whereas that eluted from silicic acid was relatively easily derivatized. A single radioactive polypeptide of M, 11,500 from pea and amaranthus chloroplasts was revealed by autoradiography of gels from sodium dodecyl sulfate-polyacrylamide gel electrophoo isso Academic resis analysis of the silicic acid eluates. Press,

Inc.

Acyl-acyl carrier proteins (acyl-ACP)2 are the direct products of long-chain fatty acid synthesis in cyanobacteria (1) and chloroplasts (2) and are the preferred substrates for acyltransferases in pea and spinach chloroplasts (3). They are also hydrolyzed by specific hydrolases (4). Current hypotheses suggest that competition between acyltransferases and thioesterases for available acyl-ACP will determine whether a fatty acid molecule is destined to be incorporated into a prokaryotic or eukaryotic glycerolipid (5). Should hydrolysis occur, then the liberated fatty acid cannot be reactivated for incorporation directly into chloroplast glycerolipids (6) but must be exported from the organelle and be incorporated into eukaryotic glycerolipids in the endoplasmic reticulum. The localization of acyl-CoA synthetase within the outer membrane (7,8) of the chloroplast envelope probably ensures a vectorial movement of newly synthesized fatty acids out of the chloroplasts. Therefore, particularly in 16:3 plants, but also in l&3 plants, concentrations of acyl-ACP may regulate the flow of acyl ’ Abbreviations used: ACP, acyl carrier protein; BAW, butan-l-ol/ acetic acid/water; FAME, fatty acid methyl esters; G3P, x-glycerol 3phosphate; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; TLC, thin-layer chromatography. 0003-9861/90 53.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

ACYL-ACYL

CARRIER

PROTEIN

CONCENTRATIONS

carbon between the prokaryotic and eukaryotic pathways of glycerolipid synthesis (5), other factors such as glycerol 3-phosphate (G3P) concentrations being equal (9). The fatty acid compositions of prokaryotic glycerolipids could also be controlled by acyl-ACP concentrations. Although the G3P acyltransferase of pea and spinach chloroplasts shows a decided preference for oleoyl-ACP (3), no such selectivity was evident in the reaction catalyzed by the enzyme in amaranthus chloroplasts (10). Such differential selectivity neatly explains the different fatty acid compositions of, e.g., chloroplast phosphatidylglycerol from the different plants, providing that the acyl-ACP concentrations occurring in vivo are compatible with those used in vitro (3,lO). To date, however, there are only two reports of acylACP concentrations within isolated chloroplasts during steady-state fatty acid synthesis from labeled acetate. Palmitoyl-, stearoyl-, and oleoyl-ACP in spinach chloroplast were 0.5-2.1, 0.1-1.0, and 0.5-0.8 PM, respectively (ll), and were briefly reported in amaranthus chloroplasts to be 0.5-0.7,0.5-O& and 0.5-0.8 PM, respectively (12). The values in spinach chloroplasts were markedly affected by additions (CoA, Triton X-100, glycerol 3phosphate) to incubation media which stimulated higher rates of either fatty acid synthesis or glycerolipid synthesis (11). Probably the prime limitation to obtaining meaningful measurements of acyl-ACP concentrations during chloroplast lipid synthesis in uitro has been that, because they are present in such low concentrations, very high rates of fatty acid synthesis are necessary to enable their detection, and the routine isolation of chloroplasts with suitably high biosynthetic activities from plants other than spinach has long eluded workers in the field. Recently, however, a number of plants have been identified which, when grown in controlled environments, yield preparations of chloroplasts with very high biosynthetic activities (12, 13). Acyl-ACP concentrations during steady-state fatty acid synthesis in chloroplasts isolated from leaves of three of those plants, pea, safflower, and amaranthus, are reported here. In addition we have employed different analytical methods in an attempt to confirm that the fatty acid methyl esters finally analyzed do indeed originate from acyl-ACP in the chloroplasts. MATERIALS

AND

METHODS

Sodium [1-‘“Clacetate, 61.6 Ci/mol, was from Amersham, CoA and sn-glycerol 3-phosphate were from Sigma, Triton X-100 was from British Drug Houses, and tetramethylammonium hydroxide and 2,2dimethoxypropane were from Aldrich. Acyl carrier protein labeled with /3-[“Hlalanine was a gift from Professor J. E. Cronan, Jr. Pea, safflower, and amaranthus plants were grown as before (12, 13) and chloroplasts were isolated from about 15 g of expanding leaves using low ionic strength buffers (14) for pea (pH 7.8) and safflower (pH 6.5), and the high-EDTA media (15) for amaranthus. Basal incubation media, which were either 0.25 or 0.5 ml final volume in 20 x 125-mm

IN CHLOROPLASTS

39

screw-capped culture tubes and contained about 0.2 mg chlorophyll/ ml, for the different chloroplast types were as described previously (12, 13). Additions were as indicated in the legends to the figures and tables. Incubations were initiated in the light by adding chloroplasts and were stopped at the times indicated by adding an equal volume of 5% (v/v) glacial acetic acid in propan-2-01 in such a manner as to cause instantaneous mixing. The method of choice for the routine recovery of acyl-ACP was a modification of that of Mancha et al. (16). Saturated (NH,),SO, (25 ~1) was added to each terminated reaction, now 1 ml in 50% propan2-01, followed by 4 ml of methanol/chloroform (2/l, v/v) containing 1% (v/v) glacial acetic acid, slowly and with vortex mixing (16). The copious precipitate was allowed to form for at least 20 min at room temperature before centrifugation (9OOg max, 5 min). The supernatant was decanted and the pellet washed once more in the same solvent mixture before being given a final wash in 4 ml of methanol/chloroform containing 1% (v/v) dimethoxypropane. The combined supernatants (13 ml) were partitioned against 7 ml of 0.2 M HnPO, in M KC1 and the chloroform layer was recovered for analyses of labeled lipids (13). The acyl-ACP was routinely analyzed by transmethylation and recovery of labeled fatty acid methyl esters (FAME). Egg phosphatidylcholine (0.2 Fmol in CHCl,) was added to the coprecipitate followed (95/5 v/v). by 1 ml of 0.5 M NaOCH,, in methanol/dimethoxypropane After standing for 20 min at room temperature, 2 ml of 125 IIIM H,SO., was added and fatty acid methyl esters were extracted into 3 + 2 ml of petroleum ether. Transmethylations employing 0.2 M tetramethylammonium hydroxide in methanol resulted in lower recoveries of FAME compared with those achieved using NaOCH,,. In comparing recoveries of different acyl derivatives from the acyl-ACP, the coprecipitates were: (i) supplied with phosphatidylcholine as above, heated with 8% (w/v) KOH in 80% (v/ v ) methanol at 80°C for 30 min so that, after cooling and acidifying, fatty acids were recovered in petroleum ether; (ii) lyophilized and treated with 0.1 M NH,OH. HCl, pH 8, at room temperature for up to 24 h, and the labeled hydroxamates extracted into petroleum ether (21); and (iii) lyophilized and treated with excess NaBH, in 30% (v/v) tetrahydrafuran at 37°C for up to 24 h, and the labeled fatty alcohols extracted into petroleum ether (21). Unlabeled fatty acyl hydroxamates or fatty alcohols prepared from egg phosphatidylcholine (17,18) were added to ii and iii, respectively, prior to extracting with petroleum ether. For measuring total acyl-ACP, the petroleum ether extracts were dried under N, at 35-40°C and the different acyl derivatives were purified by thin-layer chromatography (TLC) on silica gel; FAME by developing with benzene; fatty acids and fatty alcohols by developing with petroleum ether/diethylether/acetic acid (60/40/l, by vol); and fatty hydroxamates by developing with chloroform/methanol/acetic acid (95/5/l, by vol). Appropriate bands were detected using I:, vapor and the lipids plus adsorbent were transferred to vials for scintillation counting. To measure radioactivity in individual acyl-ACPs, FAME were separated by a combination of argentation and reverse-phase TLC essentially as described in (12), except that methyl eicosoenoate, oleate, linoleate, linolenate, and hexadecatrienoate were used as carriers and markers for methyl stearate, palmitate, myristate, laurate, and decanoate, respectively, in reverse-phase TLC. Methods other than coprecipitation with (NH&SO, were also used to separate acyl-ACP from labeled lipids. In the filter paper method, reactions (0.25 ml) were stopped with an equal volume of propan-2. al/acetic acid and individually transferred to 20 cm” of Whatman 3M paper. Papers were then washed once in acetone (10 ml per “disk”), thrice with chloroform/methanol/acetic acid (l/2/0.3, by vol) and, finally, once with diethyl ether (19). The washed papers were individually cut into slivers, phosphatidlylcholine was added, and FAME was prepared and analyzed as above. A third method took advantage of the mobility of ACP on silica gel (20) m mixtures of butan-l-01, acetic acid, and water (BAW). Reactions (0.5 ml) were stopped by adding 0.75 ml of butanol/acetic acid (2/l, v / v ) , and the resulting monophasic

40

ROUGHAN TABLE

Recovery

of Acyl-ACP

FAME

Light

I

Using Different ACP from Lipids

Precipitation

Methods

Filter paper

Dark

AND

Light

Dark

to Separate

Silicic acid Light

Dark

89 189 108 8

39 14 18

(dpm X 10-“/mg chlorophyll) Oleate Stearate Palmitate Myristate

62 147 85 6

8 12 9 2

74 162

42 29

108 8

32 3

2

Note. Amaranthus chloroplasts equivalent to 53 pg chlorophyll were incubated in minimal buffer (12) containing 0.13 mM Triton X-100 at 25°C in the light. Reactions were stopped using propan-2-01 or butanol/acetic acid, either after 5 min in the light or after a further 5min dark incubation. Acyl-ACPs were separated by the three different methods described in the text and analyzed after transmethylation with NaOCH3 by a combination of AgN03 TLC and reverse-phase TLC. The rate of long-chain fatty acid synthesis was equivalent to 2.23 rmol of acetate incorporated per hour per milligram chlorophyll. solution was cooled on ice and centrifuged. The supernatant was adjusted with additional butanol and acetic acid to give a BAW ratio of 4:l:l and was loaded onto a column of silicic acid (1 g, void vol ca. 2 ml) in a 9-mm-diam glass tube and preequilibrated with the same solvent mix. The column was then washed with 7 ml of degassed BAW (4:l:l by vol) before the acyl-ACP was eluted, usually in 5 X 2-ml portions of degassed BAW (5:2:4 by vol). The fractions containing acylACP were lyophilized and then analyzed by procedures described above. The acyl-ACP in the lyophilized column fractions was also redissolved in buffered sodium dodecyl sulfate (SDS), but without P-mercaptomethanol, and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (22). Attempts to separate acyl-ACP from lipids by column chromatography using DEAE-cellulose, octyl-sepharose, or LH-Sephadex were unsuccessful. RESULTS

Measuring

Chloroplast Acyl-ACP

Labeled ACP from Escherichia coli, when added to unlabeled chloroplasts and then coprecipitated from methanol/chloroform with (NH&SO, (above), was recovered in yields greater than 90%. However, the recovery of endogenous [14C]acyl-ACP from pea, spinach, and safflower chloroplasts by the same method was reduced by 25-50% when acetic acid was omitted from the propan-2-01 used to stop reactions. No such losses occurred with chloroplasts from amaranthus. The reason for the loss of acyl-ACP is unknown although it seems more likely to be due to deacylation, rather than incomplete precipitation, of existing acyl-ACP. As an added precaution in the present work the concentration of acetic acid in the stopping solution, where required, was raised from 2.5% (11, 16) to 5% (v/v). Recovery of FAME from transmethylations with NaOCH3 was used as the basis for comparing three different methods of separating acyl-ACP from reaction mixtures (Table I). Adsorption to filter paper, which has

NISHIDA

been used to assay acyl-ACP synthetase from E. coli (19), produced slightly higher recoveries of long-chain acyl-ACP with slightly differing proportions of constituent fatty acids compared with coprecipitation. However, in those reactions that had been transferred to the dark, residual “oleoyl- and palmitoyl-ACP” were three to five times greater on the filter papers than in the coprecipitates (Table I). When these “dark controls” were considered as backgrounds, recoveries were very similar by both methods. This residual radioactivity may indicate that filter papers are more likely than precipitates to retain some labeled lipid, e.g., lysophosphatidate. Chromatography on columns of silicic acid as an alternative to coprecipitation for separating acyl-ACP from lipids was suggested by the high mobility of both ACP and acylACP in butanol/acetic acid/water on thin layers of silica gel (20). In the present work, labeled acyl-ACP was eluted along with other proteins (Figs. 1 and 2) from small columns of silicic acid in BAW (5:2:4 by vol) following absorption from a less polar solvent mixture (BAW, 4:l:l by vol). The performance of the columns was highly reproducible. Acyl-labeled material eluting just behind the second solvent front, coincident with tritium-labeled ACP from E. coli, contained labeled FAME in the same proportions as those recovered from coprecipitates but in slightly higher amounts (Table I), even when corrected for “dark controls.” Glycerolipids, along with sorbitol and salts, seemed to be completely eluted in the loading solvent (Fig. 1) and had a palmitate: stearate:oleate molar ratio of 2l:l:ll compared with a ratio of 1:1.6:1 in the acyl-ACP fractions (amaranthus chloroplasts). Compared with coprecipitation, however, more acyl-label was recovered in the acyl-ACP fraction from darkened chloroplasts. As with the filter paper method, therefore, material having a high proportion of oleate (acyl-CoA?) was apparently retained on the column (Fig. 1) during washing with BAW (4:l:l by vol) and eluted with the more polar solvent mixture. Although there was satisfactory agreement among the three methods, and particular advantages in recovery from silicic acid columns (below), coprecipitation from methanol/chloroform provided the most convenient means for routine analysis of acyl-ACPs in, for example, time course experiments. The question of which acyl derivative would prove the most appropriate in the analysis of acyl-labeled ACP in coprecipitates was considered. Hydroxamate formation in aqueous solution at neutral pH would be indicative of acyl-ACP (21) and would provide the most suitable derivatives for the recovery of short- and medium-chain fatty acids (18). However, hydroxamates were not formed from coprecipitated acyl-ACP under the conditions described, despite the ease with which they may be derived from pure acyl-ACP (21). Treating the precipitates 50%

with

ethanolic

NHsOH

(18)

yield of labeled hydroxamates

resulted

in less

than

compared with that

ACYL-ACYL

CARRIER

PROTEIN

CONCENTRATIONS

16

8

1

2

3

4 5 FRACTION

6 7 8 NUMBER

9

10

FIG. 1. Silicic acid column chromatographic separation of lipids and acyl-ACP from amaranthus chloroplasts incubated in the minimal medium with Triton X-100, as in Table III, for 5 min in the light (A), and after a further 5 min in the dark (B). Columns (2 ml column vol) were equilibrated in BAW (4:l:l by vol) and samples, equivalent to 0.5-ml reactions, were loaded in 3 ml of the same solvent. Lipids were eluted with a further 7 ml of that solvent, and acyl-ACP with 10 ml of BAW (5:2:4 by vol). Fractions (2 ml) were lyophilized prior to treatment with NaOCH, and fatty acid methyl esters were recovered and analyzed as described in the text. The bars represent total radioactivity per reaction in palmitate (left), stearate (middle), and oleate (right). Note the lo-fold change of scale in they axis between fractions l-5 and 6-10. The rate of fatty acid synthesis in this experiment was equivalent to 1.95 wmol of acetate incorporated per hour per milligram chlorophyll.

of methyl esters; the remaining label was recovered predominantly in fatty acids. Nor were FAME in petroleum ether extracts quantitatively converted to hydroxamates using the latter reagent. Medium- and long-chain fatty acids are only slightly less volatile than their corresponding methyl esters, and their recovery should not be significantly greater than that of the methyl esters. Although recovery of radioactivity from the precipitates as fatty acids was invariably slightly higher than recovery as FAME, the increased recovery was attributable to oleic acid rather than to medium-chain fatty acids and probably represented contamination of the acyl-ACP fraction by labeled fatty acids (16). Recovery of the label from coprecipitates as fatty alcohols under conditions that ensure quantitative conversion of pure acyl-ACP from E. coli (21) was about 30% of that which could be

IN CHLOROPLASTS

41

recovered as FAME. Similar results were obtained when attempts were made to derivatize acyl-ACP adsorbed to filter paper, and eluting filter papers with 10 InM bisTris, pH 6, resulted in very low recoveries of radioactivity. The insolubility of the acyl-labeled material of coprecipitates in aqueous buffers (but see Ref. (16)) possibly accounted for the difficulty in preparing hydroxamates and alcohols, and has also precluded the use of ion-exchange and reverse-phase column chromatography (21) to confirm its identity as acyl-ACP. Quantitative analysis of the acyl composition of coprecipitated acyl-ACP, therefore, depended upon recovery of fatty acids or FAME. Since contaminating fatty acids would be eliminated by transmethylating acylACP with NaOCH,, and since FAME chromatographed more compactly than fatty acids, the results reported here for coprecipitated acyl-ACP were all obtained using FAME derived by transmethylation using NaOCH,. That the radioactivity in these FAME declined to very low levels within 1 min of incubations being transferred to darkness (see Fig. 2) is taken to indicate that any contribution to these FAME from labeled glycerolipids was minimal. Similarly, the decreased labeling of long-chain FAME from acyl-ACP when glycerolipid synthesis was stimulated by the addition of G3P to incubations (see also Ref. (ll)), suggests that glycerolipids did not interfere significantly with the estimations of acyl-ACP following coprecipitation from methanol/chloroform. The putative acyl-ACP eluted from silicic acid in BAW (5:2:4 by vol) was more readily derivatized by both borohydride and by neutral hydroxylamine than was the coprecipitated material; from a bulked preparation and using reaction times of 12-24 h at room temperature, recovery of label as fatty alcohols was 94%, and as hydroxamates was 65%, of that recovered as FAME. No radioactivity was solubilized into 10 mM bis-Tris, pH 6, following lyophilization of the column eluates in glass tubes. However, the radioactivity was soluble in boiling SDS-buffer solution (22) and was thus amenable to analysis by SDS-PAGE (22). Stained gels (Fig. 2) revealed that a large number of chloroplast proteins were soluble in BAW (2:1:2 by vol), adsorbed to silicic acid from BAW (4:l:l by vol), and subsequently eluted in the more polar solvent (BAW, 5:2:4 by vol). This was consistent with there being barely discernible pellets from incubations stopped with butanol/acetic acid and then centrifuged (see Materials and Methods). A single labeled polypeptide of M, 11,500 in both pea and amaranthus chloroplasts was detected by autoradiography of developed gels. Significantly less radioactivity was present in the IV, 11,500 polypeptide when reactions had been placed in the dark for 2 min. The presence of acyl chain lengths shorter than Cl6 should be diagnostic of acyl-ACP compared with glycerolipids, but analyses of short-chain FAME were unreliable because of low and variable recoveries: whereas

42

ROUGHAN

A

A.ND 1

lHIDA

6

FIG. 2. Identification of the acyl-labeled material from amaranthus chloroplasts as a polypeptide ofM, 11,500 by SDS-PAGE and autoradiography. Chloroplasts were incubated in the light for 5 min in the minimal medium (12) containing Triton X-100 and 50 pg of BSA before reactions were stopped with 1.5 vol of butanol/acetic acid (2:l). The fraction eluting from silicic acid columns in 2-8 ml following the change to BAW (5:2:4) was lyophilizedand the proteins were redissolved by boiling in buffered SDS, (22) but omitting the fl-mercaptoethanol. Conditions for SDS-PAGE were as in (22). Molecular weight markers were boiled in buffered SDS containing P-mercaptoethanol. (A) Dried, stained gel. Lanes 1 and 10 are ubiquitin, 8.5 kDa; lanes 2 and 9 are proteins of M, (top to bottom) 97.4,66.2,47.7,31, and 14.4 kDa, respectively; lanes 3-8, chloroplast proteins eluted from silicic acid columns. Arrowheads show the position of the acyl label. (B) Autoradiogram prepared from the dried gel.

methyl palmitate added to a transmethylation mix was better than 90% recovered following extraction, AgNOB TLC, and reverse-phase TLC, methyl laurate recovery was only 50%. Methyl myristate recovery, therefore, is expected to be about 75% and, although radioactivity was detected in saturated FAME down to ClO, those measurements are not reported here. Whereas the agreement between replicate measurements of individual acyl-ACPs separated by coprecipitation was almost always better than lo%, there was normally greater variation in the values obtained for similar treatments in different experiments. However, trends observed in one experiment were always reproduced in subsequent experiments. Any variation between experiments may be due to the very high turnover rate of longchain acyl-ACPs (the half-life of endogenous acyl-ACP is estimated to be less than 15 s from the data shown in Figs. 3 and 5, and in Ref. (11)) so that absolute concentrations will be dependent upon relatively small differences in rates of formation and/or utilization of acylACPs in a particular preparation. Acyl-ACP

in Pea Chloroplasts

Incorporation of acetate into the fatty acids of pea chloroplasts became linear after a lag of l-2 min and ceased completely when reactions were transferred to the dark (Fig. 3). However, a linear rate of incorporation was established immediately upon re-illumination. There was a less pronounced lag in the labeling of the

acyl-ACP fraction, which apparently became saturated with radioactivity (11) within 4-5 min but then lost more than 90% of its label within 1 min of incubations being transferred to the dark (Fig. 3). The subsequent recovery of label by acyl-ACP upon re-illumination was very rapid and was essentially complete within 1 min. This very rapid decrease in the labeling of acyl-ACP, particularly of oleoyl-ACP, following transferring reactions to the dark, while total fatty acids retained their radioactivity, indicates a low level of contamination of pea chloroplast acyl-ACP either by glycerides or by fatty acids. It also suggests that the rate of acyl-ACP turnover was sufficiently high so that the specific radioactivity of the carbon within the acyl chains of ACP would be the same as that within the supplied [1-14C]acetate within 4-5 min of the onset of fatty acid synthesis (11). In support of this, methyl oleate recovered from the acyl-ACP fraction was shown by oxidation with permanganatel periodate (23) to be uniformly labeled within 2 min of initiating reactions (results not shown). The concentration of oleoyl-ACP can, therefore, be calculated from Fig. 1 to be 2.2 PM, assuming a chloroplast volume of 47 pl/mg chlorophyll (24), during fatty acid synthesis by pea chloroplasts in the presence of Triton X-100. The effects of various additions to incubation media, which either stimulated total acetate incorporation into chloroplast lipids or altered the proportions of the lipid products synthesized, on concentrations of acyl-ACPs were similar to those reported for spinach chloroplasts

ACYL-ACYL

CARRIER

PROTEIN

CONCENTRATIONS

I

.2

.l

-1

2

4 6 TIME hid

a

10

FIG. 3. Incorporation of [1-“‘Clacetate into acyl-ACP (solid lines) and total fatty acids (dotted lines) by isolated pea chloroplasts when fatty acid synthesis was interrupted by a dark break; the lights were turned off at 5 min, and on again at 8 min. Reactions contained 0.13 mM Triton X-100,0.17 mM [1-‘“Clacetate (5 &i), and 100 +g chlorophyll in 0.5 ml of the basal buffer. Acyl-ACP was recovered by coprecipitation from methanol/chloroform. The maximum rate of acetate incorporation into fatty acids was 1.5 pmol h-’ per milligram chlorophyll. Saturated (@) and monounsaturated (m) fatty acid methyl esters were separated by AgNOzj TLC.

(11). Results of a typical experiment are shown in Table II. However, the effects in pea chloroplasts appeared to be concentrated upon stearoyl- and oleoyl-ACPs. In the presence of Triton X-100, for example, stearoyl-ACP was increased 246% and oleoyl-ACP was increased by 64% over control values whereas palmitoyl-ACP was reduced by 20% (Table II). This disproportionate increase in stearoyl-ACP in response to the presence of Triton X-100 in the medium was also reported for spinach chloroplasts (11). Similarly, adding glycerol 3-phosphate to incubation media depressed total acyl-ACP concentrations even though pea chloroplasts are not as active as those from spinach in glycerolipid synthesis (13). Oleoyl-ACP showed by far the largest reduction in concentration (2.6-fold) in response to added glycerol 3phosphate (Table I). Exogenous CoA increased acylACP concentrations to about the same extent (20%) as it increased acetate incorporation into total lipids. After a further 5 min in the dark, palmitoyl-, stearoyl-, and oleoyl-ACP were reduced to 5, 4, and 8%, respectively,

43

IN CHLOROPLASTS

of their former concentrations in the CoA-stimulated reaction (Table II). Maximum and minimum concentrations of acyl-ACPs during fatty acid synthesis by pea chloroplasts calculated from Table I are 0.6-1.1 PM palmitoyl-ACP, 0.2-0.7 PM stearoyl-ACP, and 0.4-1.6 FM oleoyl-ACP. Acyl-ACP

in Safflower Chloroplasts

There was a lag of at least 6 min before acetate incorporation into the lipids of safflower chloroplasts became linear (Fig. 4). However, following 10 min incubation in the light and 5 min in the dark, a linear rate was established immediately upon re-illumination (results not shown). The initial increasing rate of fatty acid synthesis from acetate was a result of a slow rise in fatty acid synthase activity in isolated chloroplasts, as indicated by a continuing increase in acyl-ACP concentrations for at least 8 min after initiating reactions (Fig. 4). Upon re-illumination following a short dark break, acyl-ACP concentrations probably increased as rapidly in safflower as in spinach and pea chloroplasts (result not shown). Concentrations of palmitoyl-, stearoyl-, and oleoyl-ACP measured 12 min after initiating CoA-stimulated reactions were 2.4, 0.5, and 1.1 PM, respectively, assuming a chloroplast volume of 47 pl/mg chlorophyll, and these concentrations decreased six- to eightfold within 2 min of reactions being transferred to the dark. Acyl-ACP concentrations in safflower chloroplasts were scarcely affected by treatments which increased or decreased the concentrations in spinach and pea chloroplasts (Ref. (11); Table I). This may have been a consequence of the slow equilibration of the label of safflower

TABLE

II

Effect of Various Additions to Incubation Media upon AcylACP Concentrations during Steady-State Fatty Acid Synthesis in Pea Chloroplasts Acyl group

Addition None Glycerol3-phosphate Triton X-100 CoA CoA + 5 min dark

18:l 18:O 16:0 (pmol/mg chl) 45 17 74 54 4

13 9 32 19 1

45 30 36 52 3

Rate of fatty acid synthesis (nmol acetateh-‘/mg chl) 941 966 1300 1207 1215

Note. Chloroplasts equivalent to 106 pg chlorophyll were incubated for 5 min in the light at 25°C and in 0.5 ml of the basal medium (14) containing the additions shown: 0.4 mM sn-glycerol Y-phosphate; 0.5 mM CoA; 0.13 mM Triton X-100. In one pair of tubes, the incubation was continued a further 5 min in the dark. Acyl-ACP was recovered by coprecipitation with ammonium sulfate and analyzed as in Table I.

ROUGHAN

AND

NISHIDA TABLE

III

Effect of Various Additions to Incubation Media upon AcylACP Concentrations during Steady-State Fatty Acid Synthesis in Amaranthus Chloroplasts Acyl group 18:l l&O 16:O (pmol/mg chl)

Addition None GlycerolS-phosphate Triton X-100 CoA

2

4

6

8

41 29 64 39

80 62 120 90

Rate of fatty acid synthesis (nmol acetate h-‘/mg chl)

88 75 74 82

1940 1930 2090 2050

Note. Chloroplasts equivalent to 42 pg of chlorophyll were incubated for 5 min at 25°C in the light and in 0.25 ml of minimal medium (12) containing the additions shown; 0.8 mM DL-glycerol 3-phosphate; 0.5 mM CoA; 0.13 mM Triton X-100. Acyl-ACP was recovered and analyzed as in Table II.

TIME (mid FIG. 4. Incorporation of [1-“Clacetate into acyl-ACP (solid lines) and total fatty acids (dotted lines) of safflower chloroplasts in continuous light. Reactions contained 0.17 mM [ 1-‘%]acetate (5 &i), 0.13 mM Triton X-100, and 106 pg chlorophyll in 0.5 ml of basal buffer. The maximum rate of fatty acid synthesis (i.e., between 6 and 8 min) was equivalent to 0.8 pmol acetate h-’ per milligram chlorophyll. Saturated (0) and monounsaturated (w) fatty acid methyl esters were separated by AgNOR TLC. Acyl-ACP was separated by coprecipitation.

acyl-ACP with that of the supplied acetate, since the additions had the expected effects upon rates and products of acetate incorporation into different lipid classes (results not shown, but see Ref. (13)). Acyl-ACP

in Amaranthus

Chloroplasts

Very high rates of fatty acid synthesis, equivalent to 2 pmol of acetate. h-l per milligram chlorophyll, were routinely achieved by chloroplasts isolated from expanding leaves on young amaranthus plants (Tables I and III, Fig. 5). Linear rates of acetate incorporation were established within 2 min and maintained for at least 10 min, but incorporation ceased when incubations were transferred to the dark (Fig. 5). Long-chain acyl-ACPs became saturated with radioactivity within 2-3 min of initiating reactions and then lost up to 95% of that label when transferred to the dark for 1 min (Fig. 5). However, the label was completely recovered within l-2 min of reillumination. As with pea chloroplasts, addition of glycerol 3-phosphate to incubation media significantly reduced the concentrations of oleoyl- and stearoyl-ACP, while addition of Triton X-100 increased their concentrations (Table III). The high proportion of stearate accumulated within the acyl-ACP fraction of amaranthus chloroplasts in the presence of Triton X-100 (Table III, Fig. 5), provided a useful marker for tracking acyl-ACP in alternative methods of isolation (above). Added COA

had little influence either on rates of acetate incorporation or on concentrations of acyl-ACPs (Table III). Maximum and minimum concentrations of palmitoyl-, stearoyl-, and oleoyl-ACP in amaranthus chloroplasts as calculated from Table III are 1.6-1.9, 1.3-2.6, and 0.61.4 PM, respectively. The respective values calculated from Fig. 5 are 1.5, 2.4, and 1.1 PM. These are based on

10 T-

8

‘;

6

0

10

ii 6;

2

4

6

8

INCUBATION

2 TIME

4

6

x

a

(mid

FIG. 5. Incorporation of [1-‘“Clacetate into fatty acids, glycerides, and acyl-ACPs of amaranthus chloroplasts when fatty acid synthesis was interrupted by a dark break. Reactions in continuous light are represented by solid lines and those transferred to the dark at 4 min and back into light at 6 min are represented by broken lines. (A) Total unesterified fatty acids (a), glycerides (V). (B) Stearoyl-ACP. (C) Palmitoyl-ACP. (D) Oleoyl-ACP. Incubations contained 0.2 mM [l-‘“Clacetate (2.5 PCi), 0.13 mM Triton X-100, and 75 ~g chlorophyll in 0.25 ml of minimal buffer (12). Radioactivity is in total counts per reaction. Acyl-ACP was separated by coprecipitation.

ACYL-ACYL

CARRIER

PROTEIN

CONCENTRATIONS

the dubious assumption that mesophyll chloroplasts from amaranthus and spinach chloroplasts have the same volume per unit of chlorophyll. DISCUSSION

There seems little doubt that the material isolated for analysis of its content of long-chain acyl groups, although present in very small amounts (e.g., 100-200 ng per measurement), was indeed acyl-ACP. Its protein nature was demonstrated both by precipitation from chloroform/methanol (16) and by mobility in SDS-PAGE. The apparent M, 11,500 for the mixed acyl-ACPs from amaranthus or pea chloroplasts was considerably higher than the M, approximately 7000 reported for palmitoylACP from spinach (22), but closer to the value of 10,000 recently reported for palmitoyl-ACP from Spirodela oligorrhiza (25). Also, the material was retained on filter paper under conditions reported to elute acyl-CoA (19) and was susceptible to neutral hydroxylamine following elution from silicic acid columns. The very high turnover rate of constituent acyl groups is consistent with its role as an acyl donor. However, we are still trying to optimize the recovery of acyl-ACP in these experiments. For example, it is not yet clear why recoveries of acyl-ACP from silicic acid columns are almost invariably higher than those by coprecipitation from chloroform/methanol, when recovery of exogenous ACP by coprecipitation is demonstrably quantitative. Since chloroplast acylACP has such a very high turnover rate, more work is required to establish the most appropriate methods of stopping reactions to stabilize acyl-ACP concentrations. Long-chain acyl-ACPs of chloroplasts apparently have half-lives measured in seconds so that their concentrations will be influenced by relatively small changes in rates of formation and/or rates of utilization in individual preparations. In spite of this however, there was remarkably good agreement between duplicated experiments using different chloroplast preparations. Originally it had been anticipated (23) that factors increasing rates of long-chain fatty acid synthesis de nouo in isolated chloroplasts would also decrease endogenous concentrations of acyl-ACPs. However, the opposite was true in spinach chloroplasts (11) and has now been confirmed, at least for pea chloroplasts; exogenous CoA and Triton X-100 increased rates of both fatty acid synthesis and endogenous concentrations of acyl-ACPs. A stimulation of fatty acid synthetase activity was the likely cause of the increased acyl-ACP. Triton X-100 may stimulate fatty acid synthetase activity in isolated chloroplasts to the extent that stearoyl-ACP desaturase activity becomes rate limiting and the concentration of stearoyl-ACP disproportionately enhanced. Added G3P, on the other hand, had no effect on rates of total fatty acid synthesis yet increased rates of glycerolipid synthesis, and significantly decreased endogenous concentra-

IN CHLOROPLASTS

45

tions of long-chain acyl-ACP in both the presence and the absence of CoA or Triton X-100. Therefore, for any given rate of fatty acid synthesis, turnover of acyl-ACP was significantly increased when acyltransferase activity was not limited by low G3P concentrations. The physiological relevance of acyl lipid metabolism by isolated chloroplasts was previously discussed (26) in relation to spinach. Many important advances have been made since that time and more plant species have yielded chloroplasts amenable to study. The outcome has been that the basic tenets of the earlier discussion have been reinforced. Isolated chloroplasts utilize exogenous acetate to synthesize unesterified fatty acids and fatty acid moieties of phosphatidate and diacylglycerol. These products are precursors of microsomal phosphatidylcholine and of chloroplast phosphatidylglycerol and diacylgalactosylglycerol, respectively, in uitro. They are synthesized in about the same proportions and with the same fatty acid compositions as those glycerolipids labeled in viva with acetate. Additionally, rates of fatty acid synthesis by isolated chloroplasts (1-2 pmol acetate. h-l per milligram chlorophyll) are similar to those measured in viuo. Therefore, there is good reason to believe that the acyl-ACP measurements reported here accurately represent values likely to occur in uiuo. Previous reports of very low concentrations (27), and of acyl compositions (28), of long-chain acyl-ACP in leaves “in Go” must be interpreted with caution in view of the very high, light-dependent turnover rates reported here. Palmitate, stearate, and oleate are the only long-chain acyl groups initially incorporated in significant amounts into plant glycerolipids. Stearate generally appears in minor proportions in the final products of chloroplast lipid synthesis even though stearoyl-ACP may be present in isolated chloroplasts in concentrations approaching those of palmitoyl- and oleoyl-ACPs. This tends to confirm that stearoyl-ACP is a poor substrate for chloroplast acyltransferase and thioesterase in uiuo so that its concentration should normally have little influence on glycerolipid synthesis. On the other hand, palmitoyland oleoyl-ACPs are maintained between 18 and 88, and 16 and 98 pmol per milligram chlorophyll, respectively, during active lipid synthesis in all four plant species examined thus far. Possibly the incubation mode most closely simulating in viva conditions is that where both CoA and G3P are supplied exogenously so that both fatty acid synthesis and glycerolipid synthesis proceed at maximum rates (23). Concentrations of acyl donors under those conditions are estimated to be 0.4-1.0 PM oleoyl-ACP and 0.7-1.5 pM palmitoyl-ACP in chloroplasts from spinach, pea, safflower, and amaranthus leaves, assuming a chloroplast volume of 47 pl/mg chlorophyll for all species. These concentrations appear to be sufficient to drive the different reactions utilizing the acyl donors to near maximum rates. The K, for oleoylACP in the thioesterase reaction was estimated to be

46

ROUGHAN

lower than 0.2 PM for the enzyme from avocado mesocarp (4), and was measured to be 0.5 PM for the enzyme from developing safflower seeds (29). The acyl-ACP: glycerol-3-phosphate acyltransferase from spinach and pea chloroplasts had a K, of 0.3 PM for oleoyl-ACP (3). On the other hand, stearoyl-ACP concentrations, at 0.20.5 PM, may often limit the rate of stearoyl-ACP desaturation in spinach, pea, and safflower chloroplasts since the purified desaturase from safhower seeds had a K,,, of 0.4 PM for stearoyl-ACP (29). A Km for palmitoyl-ACP in the elongase reaction has not yet been reported. Different isoforms of plant ACP have been shown to exhibit different preferences in the chloroplast acyltransferase and thioesterase reactions (27), and may control glycerolipid synthesis by directing newly synthesized fatty acids into prokaryotic or eukaryotic glycerolipids, respectively. However, other work (9) suggests that rates of synthesis of prokaryotic glycerolipids in viuo are strongly influenced by cellular concentrations of G3P. Similarly, and particularly in chloroplasts isolated from 16:3 plants, exogenous G3P strongly stimulates the synthesis of prokaryotic lipid precursors at the expense of eukaryotic lipid precursors (13). Proposals for physiological roles for the different ACP isoforms should, therefore, take account of the demonstrable effect of G3P in partitioning fatty acids between the two pathways of glycerolipid synthesis in plants (5). Some form of regulatory control of the reactions utilizing long-chain acyl-ACPs within chloroplasts may be inferred, however, from the apparently high proportion of total ACP which is acylated in isolated chloroplasts. It was concluded, for instance, that fatty acid synthesis de nouo rather than the G3P acyltransferase of E. coli was under regulatory control when acyl-ACPs of chain lengths suitable for the acyltransferase reaction were found to constitute just 1% of the total ACP of exponentially growing cells (30). We have consistently found concentrations of long-chain acyl-ACP in isolated chloroplasts with high rates of fatty acid synthesis to be 3-4 pM whereas a reasonable estimate of total ACP in spinach chloroplasts is B-10 pM (31), i.e., 30-40% of the total ACP existed in the form of acyl donors for glycerolipid synthesis. However, a regulatory role for the G3P acyltransferase in chloroplast glycerolipid synthesis would need to be confirmed by in vivo studies similar to those carried out with E. coli.

AND

NISHIDA

REFERENCES 1. Stapleton, S. R., and Jaworski, J. G. (1984) Biochim. Biophys. Actu 794,249-255. 2. Stumpf, P. K. (1984) in Fatty Acid Metabolism and Its Regulation (Numa, S., Ed.), pp. 1555180, Elsevier Science, Amsterdam. 3. Frentzen, M., Heinz, E., McKoen, T. A., and Stumpf, P. K. (1983) Eur. J. Biochem. 129,629-636. 4. Ohlrogge, J. B., Shine, W. E., and Stumpf, P. K. (1978) Arch. Biothem. Biophys. 189,382-391. 5. Roughan, P. G., and Slack, C. R. (1984) Trends Biochem. Sci. 9, 383-386. 6. Roughan, P. G., Holland, R., and Slack, C. R. (1980) Biochem. J. 188,17-24. 7. Andrews, J., and Keegstra, K. (1983) Plant Physiol. 72,735-740. 8. Block, M. A., Dorne, A-J., Joyard, J., and Deuce, R. (1983) FEBS L&t. 153,377-381. 9. Gardiner, S. E., Roughan, P. G., and Slack, C. R. (1982) Plant Physiol. 70,1316-1320. 10. Cronan, J. E., Jr. and Roughan, P. G. (1987) Plant Physiol. 83, 676-680.

11. Soll, J., and Roughan, P. G. (1982) FEBS I&t. 146,189-192. 12. Roughan, P. G. (1986) Biochim. Biophys. Actu 878,371-379. 13. Gardiner, S. E., Heinz, E., and Roughan, P. G. (1984) Plant Physiol. 74,890-896. 14. Roughan, P. G. (1987) in Methods in Enzymology (Packer, L., and Deuce, R., Eds.), Vol. 148, pp. 327-337, Academic Press, San Diego. 15. Jenkins, C. L. D., and Russ, V. J. (1984) Plant Sci. Lett. 35, 1924.

16. Mancha, M., Stokes, G. B., and Stumpf, P. K. (1975) Anal. Biothem. 68,600-608. 17. Nichols, B. W., and Safford, R. (1973) Chem. Phys. Lipids 11, 222-227.

18. Rosenfeld, I. S., D’Agnolo, G., and Vagelos, P. R. (1975) Anal. Biothem. 64,221-228. 19. Ray, T. K., and Cronan, J. E., Jr. (1976) Proc. Nutl. Acud. Sci. USA 73,4374-4378. 20. Jackowski, S., and Rock, C. 0. (1981) J. Bucteriol. 148,926-932. 21. Rock, C. O., Garwin, J. C., and Cronan, J. E., Jr. (1981) in Methods in Enzymology (Lowenstein, J. M., Ed.), Vol. 72, pp. 397-403, Academic Press, San Diego. 22. Kuo, T. M., and Ohlrogge, J. B. (1984) Arch. Biochem. Biophys. 230,110-116.

23. Roughan, P. G., Holland, R., and Slack, C. R. (1979) Biochem. J. 184,193-202. 24. Wirtz, W., Stitt, M., and Heldt, H. W. (1980) Plant Physiol. 66, 1877193. 25. Mattoo, A. K., Callahan, F. E., Mehta, R. A., and Ohlrogge, J. B. (1989) Plant Physiol. 89,707-‘711. 26. Roughan, P. G., and Slack, C. R. (1982) Annu. Reu. Plant Physiol. 33,977132. 27. Guerra, D. J., Ohlrogge, J. B., and Frentzen, iol. 82, 448-453.

ACKNOWLEDGMENTS Part of this work was performed under the auspices of the International Joint Research Project of the Japan Society for the Promotion of Science. Ikuo Nishida thanks the JSPS for supporting his visit to New Zealand. Dr. Ian Ferguson’s assistance with the SDSPAGE analyses is also gratefully acknowledged.

M. (1986) Plant Phys-

28. Sanchez, J., and Mancha, M. (1980) Phytochemistry 19,817-820. 29. McKeon, T. A., and Stumpf, P. K. (1982) J. Biol. Chem. 257, 12,141-12,147. 30. Rock, C. O., and Jackowski, S. (1982) J. Biol. Chem. 257,10,75910,765. 31. Ohlrogge, J. B., Kuhn, D. N., and Stumpf, P. K. (1979) Proc. Nutl. Acud. Sci. USA 76,1194-1198.