Arabidopsis - Plant Physiology

Plant Physiol. (1987) 84, 353-360 0032-0889/87/84/0353/08/$O1 .00/0

The Effects of Reduced Amounts of Lipid Unsaturation on Chloroplast Ultrastructure and Photosynthesis in a Mutant of

Arabidopsis' Received for publication October 23, 1986 and in revised form January 21, 1987

PETER MCCOURT, LJERKA KUNST2, JOHN BROWSE, AND CHRIS R. SOMERVILLE* MSU-DOE Plant Research Laboratory, Michigan State University, East Lansing, Michigan 48824 (P.M., L.K., C.R.S.); and DSIR, Plant Physiology Division, Private Bag, Palmerston North, New Zealand (J.B.) ABSTRACT A mutant of Arabidopsis thaliawa with reduced content of C,53 and CI3 faty acids in membrane lipids exhibited a 45% reduction in the

cross-sectional area of chloroplasts and had a decrease of similar magnitude in the amount of chloroplast lamellar membranes. The reduction in chloroplast size was partially compensated by a 45% increase in the number of chloroplasts per cell in the mutant. When expressed on a chlorophyll basis the rates of COrfixation and photosynthetic electron trnsport were not affected by these changes. Fluorescence polarization measurements indicated that the fluidity of the thylakoid membranes was not significantly altered by the mutation. Similarly, on the basis of temperature-induced fluorescence yield enhancement measurements, there was no significant effect on the thermal stability of chlorophyllprotein complexes in the mutant. These observations suggest that the high content of trienoic fatty acids in chloroplast lipids may be an important factor regulating organelle biogenesis but is not required to support normal levels of the photosynthetic activities associated with the thylakoid membranes.

The lipids of chloroplast membranes contain an unusually high proportion of polyunsaturated fatty acids. Depending on the plant species, linolenic (C8:3) or a combination of linolenic and hexadecatrienoic acid (C16:3) may account for more than 80% of the total fatty acids of this organelle (10). However, the functional significance of the high trienoic acid content is unknown (20). Frequent allusions to a possible role in determining membrane fluidity seem inadequate to explain the high trienoic fatty acid content since, in model systems, the presence of a single cis double bond in the hydrocarbon chain is sufficient to ensure a high degree of fluidity which is not significantly affected by the presence of additional double bonds (25). An alternate view is based on a consideration of the unusual properties of

MGD,3 the most abundant lipid in the chloroplast (8). Small changes in the degree of unsaturation of MGD may cause a conversion from a lamellar to a nonlamellar configuration designated the hex,, type phase (20). The importance of the degree of unsaturation of the fatty acyl residues on the phase properties of MGD has been demonstrated by subjecting aqueous suspensions of lipids or membranes to different extents of catalytic hydrogenation (9). These studies have indicated that, depending on the temperature, an average reduction ofless than one double bond per lipid is sufficient to induce conversion from hex,, to lamellar phase. Furthermore, very slight changes in the degree of unsaturation were required to cause MGD to undergo a liquidto-gel phase transition at room temperature (9). In an attempt to relate the in vitro studies on the role of acyl group unsaturation to the in vivo situation, the effects of catalytic hydrogenation on chloroplast membranes have been extensively examined (9, 12, 22, 29, 30). This approach has provided evidence that large decreases in the proportion of trienoic acids do not lead to significant decreases in the rates of photosynthetic electron transport, but has not permitted an assessment of other possible functions, such as a role in membrane biogenesis, which cannot be assayed in vitro. This question has been approached by exploiting the fact that the growth of barley seedlings in sublethal concentrations of the herbicide SAN9785 results in seedlings with large decreases in the amount of C183 in leaf lipids (14). Such treatment results in an increased ratio of granal to stromal lamellae and an altered ratio of PSII to PSI activity (13, 14). However, because the herbicide causes a direct inhibition of photosynthetic electron transport (11) it has not been possible to attribute these effects to alterations in the amount of trienoic acids. As a complementary approach to examining the role of trienoic acids on chloroplast structure and photosynthetic function

we have employed the fadD mutant of Arabidopsis thaliana (L.) which has large reductions of both C8:3 and C16:3 acyl groups in all membrane lipids, and has concomitant increases in C18:2 and C16:2 (4). This mutant, which has no obvious phenotype, was isolated by randomly screening a population of plants by gas chromatographic analysis of leaf samples (3). On the basis of detailed analysis of the lipid composition of this mutant it appears to be specifically deficient in an n-3 desaturase which

3Abbreviations: MGD, diacylgalactosylglycerol; Cx:y, a fatty acyl group containing X carbons with Y double bonds: DGD, diacyldigalac'This work was supported in part by grants from the United States tosyldiglycerol; PG, phosphatidylglycerol; SL, sulfoquinovosyldiacylglyDepartment of Energy (DE-AC02-76ER01338), The McKnight Foun- cerol (sulfolipid); fadD, symbol for a gene controlling activity of an n-3 dation and the U.S.-N.Z. Agreement for Scientific and Technological desaturase; F0, initial fluorescence; F, variable fluorescence; Fm, maximum fluorescence; DPH, 1,6-diphenyl-1,3,5-hexatriene; fwt, fresh Cooperation. 2 Supported by a fellowship from Monsanto. weight; LHCP, light-harvesting Chl-protein complex. 353

McCOURT TR ETAL. 3MC E APlant Physiol. Vol. 84, 1987 354 normally desaturates Ct8:2 and C 6:2 (4, 19). We describe here the Chl Fluorescence Measurements. Fluorescence measurements effects of the fadD mutation on chloroplast ultrastructure and were made on thylakoids and dark-adapted leaves as previously photosynthetic function. Our results indicate that trienoic acids described(18). For phosphorylation experiments, thylakoids are not required in high concentration for the maintenance of were isolated as described above, resuspended at 200 Mg Chl ml-' normal rates of photosynthesis but that they may play a role in in assay medium and dark adapted4°C at for 2 h before use. the development of normal chloroplast ultrastructure. Phosphorylation was initiated by the addition of ATP to a final concentration of 200uM and illumination was provided at 300 MATERIALS AND METHODS MLE m-2 s-'. NaF(10 mM) was also added to inhibit phosphatases. At specific times an aliquot was removed, diluted to10 Mg Chl Plant Material and Growth Conditions. The lines of Arabidopin 60% glycerol (v/v), and frozen in capillary tubes (0.5 sis thaliana used in this study were descended from the Columbia ml-' mm inner diameter) in liquid N2. Fluorescence was measured at wild type. We have previously described (4) the derivation and on an SLM 4048 spectrofluorometer(18). biochemical properties of the twice backcrossed mutant line 77°K Measurements of Relative Fluidity. The relative fluidity of JB 101, which carries a single recessive nuclear mutation at a isolated chloroplast membranes was determined by steady state locus designatedfadD. Plants were grown in continuous fluores- fluorescence polarization measurements of membranes containm- 2s-') in a mixture composed of ing the hydrophobic cent illumination (200 ME fluroprobe DPH (Aldrich) (1). DPH was equal parts of vermiculite, perlite, and sphagnum irrigated with added from a 3 mm stock solution in tetrahydrofuran directly to mineral nutrients (4). Except as noted in the text, the plants were a suspension of chloroplast membranes (50 Mg ml-' Chl) to a grown at27°C. final concentration of 5 After in darkness for 40 incubation Measurements of Growth Rate. Seeds were germinated at19°C at room temperature, the membranes were centrifuged at min on 100 x 25 mm Petri plates containing 25 ml of mineral for 5 min, the supernatant was discarded, and the pellet 2500gresuspended nutrients solidified with 0.7% (w/v) agar at a density of 16 evenly was in 10 mM Tricine (pH 7.9), 10 mM NaCl, 100 2 m spaced plants/dish. The light intensity was 120uE s-'. After 1M sorbitol to final a concentration of10 Mg ml1' Chl and, mM Fluorescence polarization 7 d the temperature was adjusted as noted in the text. Because DPH. were carried out measurements of restricted airflow in the Petri plates, actual leaf temperature on an SLM 4048 spectrofluorometer in a T-format. Excitation may have been slightly higher than the air temperature in the was provided by light at 360 nm with a half-bandwidth of 16 chamber. At 3 d intervals for the next 21 d plants were removed nm. Fluorescence was monitored at 460 nm with a half-bandand the fresh weight ofthe aerial portion determined. The relative width of 8 nm. storage, and calculation of the degree growth rate (&-') was determined as the slope of the log of the of polarization Acquisition, was performed by an on line Hewlett-Packard (P) average fwt of 4 plants (in mg) plotted against days since the 9825 computer. temperature adjustment. Electron Microscopy. All stages of preparative electron miChloroplast Membrane Isolation and Electron Transport croscopy were carried out at The leaf tissue was fixed in 2% Measurements. Thylakoid membranes were isolated by grinding (v/v) glutaraldehyde in sodium cacodylate buffer (pH 7.2) M 0.1 washed leaves in cold 20 mM Tricine (pH 8.4), 10 mM for h and postfixed in1% (w/v) osmium tetroxide in the same 1 NaCl, 10 mm EDTA, and 450 mM sorbitol. The homogenate buffer. After dehydration in a graded ethanol series, speciwas passed through four layers of cheesecloth and centrifuged at mens were embedded in Spurr's epoxy resin (26). the Ultrathin 3000g for 5 min at The pellet was washed in cold 10 mm sections were stained with uranyl acetate and lead citrate (23) Hepes (pH 7.9), 10 mm NaCl, 5 mm EDTA, dispersed in cold and examined in a JEOL 100 CX electron microscope. Quantiassay medium containing 20 mm Hepes (pH 7.9), 10 mM NaCl, tative measurements of membrane profiles on electron micro300 mm sorbitol, and 2mM MgCl2 and stored on ice until use. were made on sections of 20 chloroplasts from both wild For some experiments concerning the effect of Mg2' on room graphs and mutant lines. type temperature fluorescence, the preparation was subdivided into Measurement of Chloroplast Copy Number. Protoplasts were two aliquots, one of which was resuspended in assay medium prepared by placing leaves from which the lower epidermis had lacking MgC12. stripped, on a digestion medium containing 10 mm MES Electron transport was measured in a Rank oxygen electrode been 500 mm sorbitol, 1% (w/v) Macerozyme (pH 5.5), mm CaCl2, illuminated with 1000 ME m-2 s-' PAR. The standard assay RIO, 1% 1(w/v) Onozuka RIO Cellulase for 1 h at 23C. The medium contained 0.1 M sorbitol, 25 mM Tricine (pH 7.8), 10 protoplasts were diluted in the same lacking enzymes, 10 5mM MgC12, 0.1 uM gramicidin, and 1.0 mM NH4Cl. to 20 of suspension was placed on buffer NaCI, mM slides, and the microscope ,l Whole chain electron transport was assayed in the presence of protoplasts were flattened by application of a coverslip so that 02 consumption by 0.1 mm methyl the chloroplasts formed a monolayer which could be readily 0.1 mM NaN3 by measuring evolution counted. viologen. PSII activity was assayed by measuring 02electron acusing 0.25 mm 2,6-dimethyl-p-benzoquinone as an ceptor and 1.0 gM 2,5-dibromo-3-methyl-6-isopropyl-p-benzoRESULTS quinone to inhibit PSI activity. PSI activity was measured after differential inhibition of PSII by 1.0 MM DCMU. Ascorbate (2.5 Growth Rate of Mutant and Wild-type Plants. Because of the and 0.5 mm N-phenyl-N,N,N-trimethyl-p-phenylenedi- widespread interest in the possible role of membrane lipid commM) amine were used to supply electrons directly to PSI and 02 position in thermal adaptation (2, 15, 21, 24), we examined the consumption by 0.1 mm methyl viologen was measured in the effect of temperature on growth of the fadD mutant line JBI0 1 presence of 0.1 mm NaN3 and 70 ,g ml-' superoxide dismutase and the wild type. Under the conditions of this experiment both (Sigma). the mutant and the wild type had maximal growth rates at Extraction and Analysis of Chl and Lipids. Leaves were har- approximately and had similar growth rates at all tempervested at the rosette stage (3 weeks) and either fwt or leaf areas atures measured (Fig. 1). At temperatures greater than about were determined. Leaves were ground in ethanol to determine 20°C the growth rate of the mutant was slightly higher than that Chl concentration (31) and the extracts were dried under nitro- of the wild type. This observation may be considered complegen. Fatty acid methyl esters were made from the dried samples mentary to observations that acclimation of some species to by transesterification in hot methanolic-HCl and quantitated by growth at elevated temperature is associated with a decrease in detection (5). Chl a/b trienoic fatty acid content (2, 21). Additional studies involving a gas chromatography and flame ionization (16).

,M.

4°C.

(4°C) 4°C.

27°C

ratios were measured in 80% acetone

broader range of environmental conditions are required to sub-

PHYSIOLOGICAL ROLE OF MEMBRANE LIPID UNSATURATION

355

the alteration in fatty acid composition with reduced Chl levels indicates that the two biochemical changes are due to the same mutation. Additional evidence of the link between fatty acid composition 0.4and Chl deficiency was indicated by a pronounced temperature effect on Chl levels in the mutant. The fadD mutation affects fatty acid composition at temperatures above about 20°C but has a relatively minor effect at lower temperatures (4). Thus, the 3 0.3 effect of temperature on Chl content was tested by growing plants .40 at 27°C to allow full expression of the fadD phenotype. Plants S were then shifted to 19°C to allow recovery of normal trienoic acid levels. During this permissive period, the amount of Chl per 3 0.2 unit of leaf area slowly increased to wild-type levels (Fig. 2A) 0 and the amount of trienoic acids began to approach wild-type levels (FIg. 2C). By contrast, plants maintained at 27°C showed no change in fatty acid composition (Fig. 2D) or Chl per unit X 0.1 leaf area (Fig. 2B) during the same period. The observation that 06 changes in both trienoic acid and Chl content are responsive to temperature provides an independent line of evidence that both are caused by the same mutation. n%v 20 25 15 10 35 30 From the results presented in Table I it is apparent that the Temp (°C) fadD mutation has a relatively minor effect on the Chl/lipid ratio in the membranes, suggesting that the stoichiometry FIG. 1. Effect of temperature on the relative grovwth rate of wild-type of thesechloroplast is maintained. Since the total amount components two (-) and mutant (0) Arabidopsis. of Chl per unit fwt is reduced, it may be inferred that the mutant has approximately a 20% reduction in the amount of chloroplast Table I. Relative Amounts of Lipid and Chl in Alutant JBIOI and membrane lipid per cell. and' Thylakoids Leaves Arabidopsis (WT) Wild-Type Chloroplast Ultrastructure and Number. In previous studies WT in which trienoic acid content of barley seedlings was reduced JBIOI with the herbicide SAN9785, changes in chloroplast ultrastruc3.65 ± 0.12 3.42 ± 0.04 Chl a/b ratio ture were observed (14). We therefore carried out a comparative 1.39 ± 0.04 1.62 ± 0.10 Chl/fwt (mg g-') morphometric analysis of electron micrographs of chloroplasts ± 0.11 3.47 ± 0.08 3.42 g-') (g Total lipid/Chl from the mutant line JB101 and wild type. The fadD mutation 2.73 ± 0.05 2.93 ± 0.08 Thylakoid lipid/Chl (g g') increased the granal width about 20% and caused a reduction in the total length of stromal thylakoids and in the number of grana stantiate this effect. However, the results in Fi igure 1 emphasize per plastid (Table II). These effects can be clearly seen in the the fact that there are no major deleterious e ffects of the fadD electron micrographs presented in Figure 3. In cross-sections of mutation, or any other background mutations i on the growth of chloroplasts from the mutant, grana and stroma membranes the mutant line JB I()l. Since the JaaD mutation nas very uiue effect on lipid composition at temperatures less than about 20C (4), this experiment did not permit an assessment of the effect of reduced trienoic acid composition at temperatures below this threshold. Effects of Trienoic Acid Composition on Chl and Lipid ConE tent. When grown at 27°C the leaves of the mutant line JB1O1 0 were slightly lighter in color than those of the wild type due to a -o 15% reduction in Chl per unit fwt (Table I). Since all Chl in higher plants is believed to be associated with thylakoid proteins E (17), a decrease in Chl content is indicative of a reduction in the amount of one or more of the Chl-protein complexes. The observation that the Chl a/b ratio is not significantly different between the mutant and wild type (Table I) suggests that the stoichiometry of the various Chl-protein complexes is not altered but that all Chl-protein complexes are reduced equally. In support of this, no significant differences were observed in the 0 C protein patterns in SDS-polyacrylamide gels of polypeptides 0cv extracted from chloroplasts from wild-type and mutant leaves 0+ (results not presented). I._ To determine if the reduction in Chl was related to the fadD mutation, 44 F2 plants from a cross of wild type x JB1O1 were 40 am 0. _ tested for cosegregation of low trienoic acid and reduced Chl. Of IL in and levels, C16:3 reductions C183 the 10 plants which showed all had a 10 to 15 % reduction in the amount of Chl per unit fwt Time (days) relative to the wild type. The 34 plants with normal or intermeFIG. 2. Change in Chl and trienoic acid content following a shift of diate levels of trienoic acids had normal levels of Chl. Furthermore, all of the 10 segregants with reduced Chl content produced wild type (0) and mutant (0) plants from 27'C to 19C. Plants grown at progeny with the same phenotype indicating that they were 27°C were transferred to 19°C (A and C) or maintained at 27°C (B and homozygous for the relevant mutation. The co-segregation of D). Each point represents the mean ± SD (n = 3).

2k

_

n tx

._

356

MCCOURT ET AL.

Table II. Morphometric Analysis ofChloroplasts from Mutant JB1OJ and Wild-Type (WT) Arabidopsis Measurements were made on 20 chloroplasts from each line. JB1I1 WT 54.1 ± 13.0 37.2 ± 8.2 Grana/plastid 4.8 ± 2.5 5.5 ± 3.3 Thylakoids/granum 0.40 ± 0.03 0.48 ± 0.03 Granal width (jam) 55.7 ± 14.2 67.0 ± 20.6 Stroma thylakoids/plastid 0.20 ± 0.03 0.26 ± 0.03 Stroma thylakoid length (;tm) 85.7 119.0 Total grana (m/plastidr 11.1 17.4 Total stroma (,um/plastidr 96.8 136.4 Total thylakoids (,gm/plastidr 7.7 6.8 Grana/stromaa 5.3 ± 1.6 9.7 ± 2.0 Surface area (Am2/plastid) a These values were derived from others in the table.

Plant Physiol. Vol. 84, 1987

measuring photosynthetic gas exchange in mutant and wild-type plants at saturating light intensities. C02-fixation rates of mutant and wild type were indistinguishable when expressed on a Chl basis (Table III). Although the mutant showed a 20% reduction in C02-fixation rates when expressed on the basis of fwt (Table III) it seems likely that this is due to the reduction of similar magnitude in the amount of chloroplast membranes per unit

fwt.

In order to measure the effect of lipid composition on the light reactions catalyzed by isolated thylakoids, plants were grown at 27°C to ensure maximum reduction of trienoic acid content in the mutant, and then assayed at various temperatures. Wholechain electron transport rates were very similar in mutant and wild type at 6, 14, and 25°C (Table III). There were also no major differences in the rates of PSI or PSII activity between mutant and wild type, indicating that the high levels of trienoic acids found in the wild type are not required to support maximal levels of these activities. The mutant actually had a slightly higher rate were reduced in amount to approximately 73 and 64% of wildof PSI activity. This could reflect a small decrease in the amount type levels, respectively. The most striking effect of the mutation of Chl associated with PSI due to the reduced amount of stromal was a 45% reduction in the cross-sectional area of plastids in the thylakoid lamellae (Table II). In any case the difference was not mutant. This unusual effect was observed in each of the inde- considered large enough to merit further investigation pendent preparations of thin sections from three different batches Fluorescence Measurements. Fluorescence emitted from chloof plants. This appears to be the first example of a genetic effect roplast membranes excited with 440 nm light at 77°K can be on chloroplast size. into three major peaks at 685, 695, and 734 nm (18). Since the decrease in chloroplast size in the mutant was not resolved been attributed to LHCP associated with PSII, peaks These accompanied by a proportional decrease in the amount of Chl PSIl reactionhave centers, and PSI, respectively (6). Barley seedlings With per unit fwt we measured the number of chloroplasts. sublethal doses of SAN9785 show large enhancewith treated plants grown at 19°C there was no difference between the number ments of low temperature fluorescence at 685 and 695 nm in ± wild type and of chloroplasts per cell the mutant (32.4 10) This is thought to reflect an altered ratio of 734 nm. to relative (34.6 ± 1 1) (Fig. 4). By contrast, when grown at 27°C the mutant appressed to nonappressed membranes which was observed in ± ± 15 57.9 25 40.1 and and wild type had an average of treated seedlings (14). chloroplasts from of micrographs electron chloroplasts per cell, respectively (Fig. 4). The fact that chloromembranes ratio in chloroplast of the comparison A F6s5/F734 the in with temperature plast number varies disproportionately any change in mutant and wild type did not reveal the from mutant and wild type establishes a common link between the IV). of MgCl2 (Table or absence the presence in ratio PSII/PSI and on acid composition fatty effects of the fadD mutation a due to is of MgCl2 the presence in ratio The higher F685/F73 chloroplast number. which protein complexes of reorganization cation-dependent fatty of trienoic effects the estimate Membrane Fluidity. To PSII. The fact acid composition on the fluidity of thylakoid membranes, fluo- causes an increase in energy distribution towards two genotypes the between any difference not observe we did that freshly out carried on rescence polarization measurements were complexes the Chl-protein that implies condition either under isolated thylakoids from mutant and wild-type leaves. The principle of the method is that the fluorophore DPH readily partitions are structurally similar, that they are present in the same proporinto thylakoid membranes (1), and upon excitation with polar- tion as in the wild type, and that the efficiency of energy transfer ized light an appropriately oriented subset of DPH molecules between the major complexes is normal. This conclusion was substantiated by measuring variable flureemit polarized light. The plane at which the polarized fluorescence is reemitted is dependent on the rotational diffusion of the orescence at room temperature in the mutant and wild type. probe during its excitation lifetime. If the probe does not move Room temperature fluorescence is primarily emitted from PSII, between excitation and emission, the fluorescence polarization and since it is a kinetic measure it is a sensitive indicator of the value (P) will be maximal. The diffusive motion of the probe is efficiency of light capture and of PSII photochemistry. However, in turn dependent on the viscosity of the membrane. Thus, the the magnitude of the variable fluorescence (expressed as FVIFo) greater the fluidity of a membrane, the faster the rotational was indistinguishable between mutant and wild type in both the diffusion of the probe during its excitation lifetime and hence presence and absence of MgC12 (Table IV). There was also no difference in the shape of the fluorescence induction curves the lower the fluorescence polarization value. The P values obtained for Arabidopsis thylakoids are very obtained from mutant and wild type (results not presented). similar to those of other species (1) and, in general, indicate a Thus, by these criteria, we were unable to observe a difference highly fluid membrane (Fig. 5). Although the mutant exhibited in the photochemistry of the mutant and wild type. Effects of Trienoic Acid Composition on Protein Diffusion. It slightly higher polarization values than the wild type at all temperatures, the difference was below the limit of statistical has been proposed that light harvesting antenna complexes can significance. Since Barber et al. (1) have shown by similar criteria undergo reversible lateral diffiusion between appressed and nonthat grana have a higher P value than stroma, the slight difference appressed regions of thylakoid membranes under the appropriate between the mutant and the wild type may be accounted for by environmental conditions (27). This phosphorylation-dependent the increased ratio of granal to stromal membranes in the mutant process, which regulates the light energy delivered to the two (Table II). These observations suggest that considerable changes photosystems, can be artificially induced in isolated thylakoids in the level of trienoic acyl groups result in relatively small by the provision ofATP and light, and is thought to be controlled decreases in membrane fluidity and are, in this respect, consistent by a protein kinase (27). Because phosporylation and subsequent movement of the antenna complexes from appressed to nonapwith the results of studies with model membrane systems (28). Photosynthetic Studies. A gross test of the effects of the re- pressed regions of the chloroplast lamellae is a relatively slow duced trienoic acid content on photosynthesis was performed by process, it can be monitored by a change in the ratio of fluores-


A.

357

I..

Io

C- .at 'I

-,@i.

-.

&% 4

ft..

___________

4¾ ..4?

FIG. 3. Transmission electron micrographs of ultrathin (80-100 nm) sections of whole leaves from wild-type (A) and mutant (B) Arabidopsis at x26,000 magnification. The bar inserted in the figure indicates I um. The magnification of the inset is x54,000.

358


McCOURT ET AL.

Table III. Comparison ofPhotosynthetic Activities in Mutant JBJ01 and Wild-Type (WT) Arabidopsis WT JBIOI CO2 fixation 116.0 ± 2.5 114.1 ± 2.5 ,gmol mg-' Chl h-' mg mg fwt-' h-' 6.8 ± 0.6 5.5 ± 0.2 Electron transport rates Amol 02 mg Chl-' h-' Whole chain 25°C 261.4± 7.1 270.7± 19.1 14°C 202.3 ± 5.9 200.0 ± 9.3 60C 132.9±7.5 123.8± 11.3 PSII 25°C 261.8 ± 14.5 265.7 ± 4.3 PSI 25°C 375.4 ± 9.5 422.4 ± 16.7

15

=10

E 5 z

0 20

27C

*WT

E0

15 0

JBIOI

Table IV. Room Temperature (FV/F) and Low Temperature (F685/ F73) Fluorescence of Isolated Thylakoids from Mutant JBIOJ and

Wild-Type Arabidopsis F,/FO Wild type (+Mg2+) 2.73 ± 0.11 JBIO (+Mg2+) 2.61 ± 0.05 Wild type (-Mg2+) 1.27 ± 0.02 JBIOI (-Mg2+) 1.26 ± 0.16

10

E z

I~~~~~~~~~~~~~~

5

LO

0

F68/F734 1.49 ± 0.08 1.52 ± 0.08 0.87 ± 0.01 0.88 ± 0.07

F

>

0

10 20 30 40 50 60 70 80 90 100 Number of chloroplasts per cell FIG. 4. The effect of the fadD mutation on the number of chloroplasts in 50 leaf cells of each genotype.

I

(0

t-C2

0 c

0.8

0.7 F

00

0

C,

C0

c

*° 0.6;H

.2 0.3.N

0 Co

0 0

~.0.2-

10

15

20

Time (min) FIG. 6. Rate of ATP-induced change in ratio of 77°K fluorescence (F734/F685) from thylakoid membranes of wild-type (0) and mutant (0) Arabidopsis. Each point represents the mean of three samples.

4)0.1 0

W%f

5

0

V.V-

40 50 20 30 Temp (°C) FIG. 5. Effect of temperature on DPH fluorescencx polarization by thylakoid membranes from wild-type (-) and mutant (0) Arabidopsis.

0

10

from PSI (F734) relative to PSII (F685) at 7T 70K. Analysis of low temperature fluorescence emiission of chloroplast membranes following illumination in the p presence of ATP indicated that Arabidopsis has a mechanism, similar to that observed in other species, which leads to a redistribution of absorbed energy between the two photosystem;s. However, the kinetics of the change in the F734/F685 ratio was not distinguishable in mutant and wild-type samples (Fig. 6). TIhus, the reduced degree of unsaturation in the mutant did not aalter the rate or extent of thylakoid membrane reorganization to) any significant extent. This indicates, by an additional criterion that the structural changes in the chloroplast membranes of the mutant are cence

L,

not accompanied by any apparent functional defect. Effects of Hieh Temnerature on Fluorescence. A heat-induced increase in

room

as a breakdown

temperature

fluorescence has been interpreted

of energy transfer from LHCP antenna pigments

to PSII centers (21, 24). Because of this separation of LHCP from the PSII core, excitation energy transfer is inhibited, causing reemission of the energy as fluorescence. In this respect, Chl fluorescence may be considered an intrinsic probe of lipidprotein interaction and an indicator of thermal stability of PSII Chl-protein complexes. The temperature at which enhanced fluorescence occurs may vary in response to environmental acclimation and has been suggested to be affected by the composition of the lipid matrix in which proteins are embedded (15, 21, 29). The effect of temperature on Chl fluorescence in intact leaves of mutant and wild type was measured by slowly heating detached leaves at a rate of 1°C min-' up to 550C and continuously monitoring the level of fluorescence (Fig. 7). There was no significant difference in the temperature at which fluorescence


00

000

0.000

0 00 00 0 00 00

0 0 c

00

0 0

00

cn 0

s-

00 0

0

o

a:

0

0@

00 0000000000000000~

0~00

000000000000000000 o .00000 I

30

I

I

I

I

50

40 Temp

I

(OC)

FIG. 7. Temperature-induced fluorescence enhancement yield (F0) of wild-type (0) and mutant (0) leaves. Plants were grown at 27C. Each point represents the mean of four independent measurements.

started to increase in the mutant as compared to wild type, or in the magnitude of the responses. Thus, the stability of LHCPPSII association does not appear to be significantly affected by the reduced C18:3 and C16:3 content of the membranes from the mutant.

DISCUSSION On the basis of a detailed comparison of the lipid composition of the mutant and wild type we previously suggested that the mutant is specifically deficient in activity of a chloroplast-located n-3 desaturase (4). When plants are grown at 27C this results in an approximately 45% reduction in trienoic acid content of both chloroplast and extrachloroplast lipids and a corresponding increase in dienoic acids (4). Since the mutation is maximally expressed when plants are grown at temperatures above 27C but is of minor effect in plants grown below 19°C, we consider it likely that the fadD mutation rendered the n-3 desaturase or a regulatory factor thermolabile. As noted below, this characteristic of the mutation was very useful in establishing that the altered fatty acid composition and reduced Chl content of the mutant were due to the same mutation. The fadD mutation had no major effect on CO2 fixation, electron transport (Table III), low temperature fluorescence, or room temperature fluorescence (Table IV). Net CO2 fixation was reduced when expressed on a fwt basis but this probably reflects the reduction in chloroplast membranes since this difference was not apparent when rates were expressed on a Chl basis. These results complement studies in which trienoic acid content was reduced to varying degrees by catalytic hydrogenation of chloroplast membranes. Although there was some inhibition of electron transport due to the catalyst, there was no additional effect, relative to the nonhydrogenated control, when up to 40% of the lipids were fully saturated (22). Thus, in agreement with these studies we conclude that at at least 50% of the trienoic acid content is not required for normal photosynthetic function in the range of light intensity and temperature used here. Sublethal concentrations of SAN9785 have also been used to

359

reduce trienoic acid content in barley seedlings (13, 14). Since SAN9785 inhibits photosynthetic electron transport directly (1 1), there are significant technical problems associated with efforts to examine the effects of reduced trienoic acid content caused by sublethal concentrations of the herbicide on photosynthesis. Where this has been attempted (13) it was concluded that a trienoic/dienoic acid ratio of 2 was the threshold, below which photosynthesis was progressively disrupted. However, since the trienoic/dienoic acid ratio of the fadD mutant is reduced from 3.4 to 0.45 without any effect on electron transport we infer that the reduced photosynthetic activity in the SAN9785-treated material is not due to the altered lipid composition. Because of the association between lipid polyunsaturation and membrane fluidity (28), it might be assumed that the high trienoic fatty acid content of thylakoids is important in maintaining an extremely fluid matrix for lateral movement of photosynthetic components. Indeed, as indicated by fluorescence polarization measurements, the chloroplast membranes from the fadD mutant were slightly less fluid than those of the wild type. However, since the rate of electron transport was not affected we infer that fluidity does not normally limit electron transport or the rate of membrane reorganization associated with the state I/ state II transition (27). Since thefadD mutation was not expressed at low temperature we could not assess the role of trienoic fatty acid composition on growth at low temperature. On the other hand, the slightly enhanced growth rate of the mutant relative to the wild type at elevated temperatures merits further study. The implication, that a reduction in the amount of trienoic fatty acids enhanced growth at high temperature, is consistent with the conclusions from a study in which the degree of lipid unsaturation declined during acclimation of Nerium oleander to growth at high temperature (21). However, in N. oleander the acclimation was also accompanied by an increase in the temperature at which Chl fluorescence was enhanced. We did not observe any difference between mutant and wild-type by this criterion. Since the magnitude of the changes in trienoic fatty acid composition in thefadD mutant were greater than those which accompanied acclimation in N. oleander (21) we infer that fatty acid composition is not the critical factor which determines the temperature at which enhanced Chl fluorescence occurs. Lynch and Thompson (15) have also failed to establish a correlation between gross fatty acid polyunsaturation and the thermal stability of the photosynthetic apparatus in cold-acclimated Dunaliella. The effects of thefadD mutation on chloroplast ultrastructure were similar, though less pronounced, than those obtained by treatment of barley with sublethal amounts of SAN9785 (14). In both cases the grana were wider than in the controls, with fewer thylakoids per granum and less granal membrane. The most striking effect of the fadD mutation was a 45% reduction in the cross-sectional area of chloroplasts and in the amount and distribution of lammelar membranes within the chloroplasts. Because volume is not linearly related to cross-sectional area, the amount of thylakoid membranes was not reduced proportionately to the reduction in chloroplast cross-sectional area. Also, because the number of chloroplasts was increased about 45% in the mutant, the amount of Chl per cell was reduced by only about 15%. It was not possible to measure the segregation of chloroplast size. However, since the decreased Chl content was expressed at 27°C but not at 19°C, and since chloroplast number was normal at 19°C but not at 270C, we infer that the reduced chloroplast size is due to thefadD mutation. Also, since Chl and trienoic fatty acid content co-segregated in an F2 population derived from a wild-type x JB 101 cross we believe that the fadD gene product regulates both phenomena. However, relatively minor effects on chloroplast size were observed in studies in which trienoic acid content of barley seedlings was decreased by

360

MCCOURT ET AL.


Biochemistry of Plants, Vol 4. Academic Press, New York, pp 1-55 growth on SAN9785 (14). Thus, it is not clear that it is the JL, AL SCHAREN, JB ST JOHN, DE MORELAND, KH NoRRIS 1969 decrease in trienoic fatty acid composition per se which is re- 11. HILTON Modes of action of pyridazinone herbicides. Weed Sci 17: 541-547 sponsible for the reduced chloroplast size. 12. HORVATH I, AR MANSOURIAN, L VIGH, PG THOMAS, F Joo, PJ QUINN 1986 There seem to be two distinct explanations for the co-regulaHomogeneous catalytic hydrogenation of the polar lipids of pea chloroplasts in situ and the effects on lipid polymorphism. Chem Phys Lipids 39: 251tion of trienoic fatty acid content and chloroplast size by the 264 fadD mutation. First, we may suppose that there is no mecha- 13. LASKAY G, E LEHOCZKI 1986 Correlation between linolenic acid deficiency in nistic connection between chloroplast size and lipid composition. chloroplast membrane lipidsanddecreasingphotosyntheticactivityin barley. The fadD gene product could be a factor which controls chloroBiochim Biophys Acta 849: 77-84 plast size and also independently and specifically regulates the n- 14. LEECH RM, CA WALTON, NR BAKER 1985 Some effects of 4-chloro-5-(dimethylamino)-2-phenyl-3(2H)-pyridazinone (SAN9785) on development of 3 desaturase (since this is the only fatty acid desaturase affected). thylakoid membranes in Hordeum vulgare L. Planta 165: 277-283 We find this less attractive than other possibilities since there 15. LYNCH D, GA THOMPSON 1984 Chloroplast phospholipid molecular species seems to be no obvious link between the two effects. A simpler alterations during low temperature acclimation in Dunalialla. Plant Physiol 74: 198-203 hypothesis is that the fadD gene product controls only the n-3 G 1941 Absorption of light by chlorophyll solutions. Biol Chem desaturase and that the amount of trienoic acid directly regulates 16. MACKINNEY 140: 315-322 chloroplast volume. The increased number of chloroplasts in the 17. MARKWELL JP, JP THORNBER, RT BoGGs 1979 Higher plant chloroplasts: mutant implies that the chloroplasts are smaller because they evidence that all the chlorophyll exists as chlorophyll-protein complexes. Proc Natl Acad Sci USA 76: 1233-1235 divide more frequently. Since the mechanisms and physical P, BROWSE, WATSON, CJ ARNTZEN, CR SOMERVILLE 1985 forces responsible for determining chloroplast size and organi- 18. MCCOURT of photosynthetic antenna function in a mutant of Arabidopsis Analysis zation are entirely obscure, it is not currently possible to evaluate thaliana (L.) lacking trans-hexadecenoic acid. Plant Physiol 78: 853-858 this possibility relative to other evidence concerning chloroplast 19. NORMAN H, JB ST JOHN 1986 Metabolism of unsaturated monogalactosyldiacylgylcerol molecular species in Arabidopsis thaliana reveals different sites morphogenesis. However, since changes in membrane fatty acid and substrates forlinolenic acid synthesis. Plant Physiol 81: 731-736 composition have been shown to dramatically alter erythrocyte 20. QUINN PJ, WP WILLIAMS 1983 The structural role of lipids in photosynthetic morphology, there is a precedent for such a concept (7). Studies membranes. Biochim Biophys Acta 737: 223-266 of the effects ofSAN9785 on chloroplast morphology of A. 21. RAIsONJK,JKM ROBERTS, JA BERRY 1982 Correlations between the thermal stability of chloroplast (thylakoid) membranes and the composition and thaliana and the characterization of additional mutants with fluidity of their polar lipids upon acclimation of the higher plant Nerium reduced trienoic acid composition would be useful in further oleander to growth temperature. Biochim Biophys Acta 688: 218-228 evaluating the relationship between chloroplast structure and 22. RESTALL CJ, P WILLIAMS, MP PERCIVAL, PJ QuINN, D CHAPMAN 1979 The trienoic acid content indicated by the properties of the fadD modification of membrane fluidity by hydrogenation processes. III. The hydrogenation of biomembranes of spinach chloroplasts and a study of the mutant. -

J

J

LITERATURE CITED

23.

1. BARBER J, RC FORD, RACMITCHELL, PA MILLNER 1984 Chloroplastthylakoid membrane fluidity and its sensitivity to temperature. Planta 161: 375-380 2. BERRYJ, 0 1980 Photosynthetic response and adaptation to temperature in higher plants. Annu Rev Plant Physiol 31: 491-543 3. BROWSE J, P MCCOURT, CR SOMERVILLE 1985 A mutant ofArabidopsislacking a chloroplast-specific lipid. Science 227: 763-765 4. BROWSE J, P MCCOURT, CR SOMERVILLE 1986 A mutant of Arabidopsis deficientC,8:3 in andC16:3 leaf lipids. Plant Physiol 81: 859-864 5. BROWSE J, P MCCOURT, CR SOMERVILLE 1985 Overall fatty acid composition ofleaf lipids determined after combined digestion and fatty acid methyl ester formation from fresh tissue. Anal Biochem 152: 141-146 6. BUTLER WL 1978 Energy distribution in photochemical apparatus of photosynthesis. Annu Rev Plant Physiol 29: 345-378 7. CHRISTIANSSON A, FAKUYPERS, BROELOFSEN, JAF DEN KAMP, LLM VAN DEENEN 1985 Lipid molecular shape affects erythrocyte morphology. Cell Biol 101: 1455-1462 8.GOUNARIS K, J BARBER 1983 Monogalactosyldiacylglycerol: the most abundant polar lipid in nature. Trends Biochem Sci 8: 378-381 9. GOUNARIS K, DA MANNOCK, A SEN, APR BRAIN, WP WILLIAMS, PJ QUINN 1983 Polyunsaturated fatty acyl residues of galactolipids are involved in the control of bilayer/non-bilayer lipid transitions in higher plant chloroplasts. Biochim Biophys Acta 732: 229-242 10. HARWOOD JL 1982 Plant acyl lipids. In PK Stumpf, EE Conn, eds, The

24.

J

effects of this on photosynthetic electron transport. Biochim Biophys Acta 555: 119-130 REYNOLDS ES 1963 The use of lead citrate at high pH as an electron opaque stain in electron microscopy. Cell Biol 17: 208-212 SCHREIBER U, JA BERRY 1977 Heat induced changes in chlorophyll fluorescence in intact leaves correlated with damage in the photosynthetic apparatus. Planta 136: 233-238 SMALL DM 1986 The Physical Chemistry of Lipids: From Alkanes to Phospholipids. Plenum, New York SPURR AR 1969 A low viscosity epoxy embedding medium for electron J microscopy. Ultrastruct Res 26: 31-43 STAEHELIN LA,CJ ARNTZEN 1983 Regulation of chloroplast membrane function: protein phosphorylation changes the spatial organization of membrane J

BJ1RKMAN

OP

J

25. 26. 27.

J Cell Biol 97: 1327-1337 components.

SMITH

28. STUBaS CD, AD 1984 The modification of mammalian polyunsaturated fatty acid composition in relation to membrane fluidity and function. Biochim Biophys Acta 779: 89-137 29. THOMAN PG, P3 DomINY, L VIGH, AR MANSOURIAN, P3 QUINN, WP WILLIAMS 1986 Increased thermal stability of pigment protein complexes of pea thylakoids catalytic hydrogenation of membrane lipids. Biochim

following

Biophys Acta 849: 131-140 30. VIGH L, FJoo, M DROPPA, LI HORVATH, G HoRVATH 1985 Modulation of chloroplast membrane lipids by homogeneous catalytic hydrogenation. Eur 3 Biochem 147: 477-481 31. WINTERMANS JFGM, A DEMOTS 1965 Spectrophotometric characteistics of chlorophylls a and b and their pheophytins in ethanol. Biochim Biophys Acta 109: 448-453