The Plant Cell, Vol. 7, 1 1 17-1 127, August 1995 0 1995 American Society of Plant Physiologists

RESEARCH ARTICLE

Storage, Photosynthesis, and Nature of Mutations Affecting Structure in Chlamydomonas

Growth: The Conditional Starch Synthesis and

Nathalie Libessart,a Marie-Lise Maddelein,b Nathalie Van den Koornhuyse,b André Decq,b Brigitte Delrue,b Gregory Mouille,b Christophe D’Hulst,b and Steven Ballb>’ a Roquette Frères, F62136 Lestrem, France

Universite des Sciences et Technologies de Lille FlandresArtois, 59655 Villeneuve d’Ascq Cedex, France Laboratoire de Chimie Biologique, Unite Mixte de Recherche du Centre National de Ia Recherche Scientifique NOlll,

Growth-arrested Chlamydomonas cells accumulate a storage polysaccharide that bears strong structural and functional resemblance to higher plant storage starch. It is synthesized by similar enzymes and responds in an identical fashion to the presence of mutations affecting these activities. We found that log-phase photosynthetically active algae accumu- late granular a(l--ll)-linked, a(l+)-branched glucans whose shape, cellular location, and structure differ markedly from those of storage starch. That synthesis of these two types of polysaccharides is controlled by both a common and a specific set of genes was evidenced by the identification of a new Chlamydomonas (STA4) locus specifically involved in the biosynthesis of storage starch. Mutants defective in STA4 accumulated a new type of high-amylose storage starch displaying an altered amylopectin chain size distribution. It is expected that the dual nature and functions of starch syn- thesis in unicellular green algae will yield new insights into the biological reasons for the emergence of starch in the eukaryotic plant cell.

INTRODUCTION

Starch accumulates as a complex granular structure made of a glucans in the leaf cell chloroplast (transient starch) and in the amyloplast of the plant storage tissue cell (storage starch) (for review, see Preiss, 1991). To date, all published structural characterizations of starches deal with storage starch, and very little remains to be known about the fine structure and composi- tion of leaf (transient) starch. That both kinds of polysaccharides are synthesized in plastids by similar enzymes from ADP- glucose is evident from biochemical experiments performed mainly with spinach leaves (Ghosh and Preiss, 1966; Ozbun et al., 1972) and from the genetic and biochemical work more recently performed with Arabidopsis (Caspar et al., 1985; Lin et al., 1988), Nicofiana sylvesfris (Hanson et al., 1988), and potato (Müller-Rober et al., 1992). The storage polysaccharide is usually defined as a mix of two distinct fractions: amylopec- tin and amylose. Amylopectin is by far the major compound. It is composed of intermediate-sized a(l+4)-linked glucans that are clustered together by a(l-6) linkages (for review, see Manners, 1989). Segments of these chains intertwine to form parallel arrays of double helices responsible for the crystallinity of starch. These arrays are separated one from another by

To whom correspondence should be addressed.

longer spacer glucans. The exact conservation in the case of storage starch of the amylopectin cluster size (9 nm) throughout the plant kingdom suggests the existence of a highly ordered, precise, and well-conserved biosynthetic pathway (Jenkins et al., 1993). Amylose is often referred to as a smaller linear mol- ecule with very few a(1-6) branches, whose association with amylopectin inside the granule remains to be determined.

In growth-arrested Chlamydomonas cells, we have been able to show that starch adopts, by many criteria, a structure reminis- cent of maize endosperm storage starch (Delrue et al., 1992; Fontaine et al., 1993; Maddelein et al., 1994). Nitrogen starva- tion in particular has enabled us to screen very effectively for mutants affected in starch structure or amounts. Under these particular physiological conditions, the destruction of chlo- rophylls and massive accumulation of starch allowed us to visualize the pure iodine polysaccharide interaction directly on colonies. Moreover, the large amounts of starch synthesized in these cells made routine structural characterizations of starches feasible for a single-cell organism. This genetic ap- proach, coupled with carbohydrate biochemistry, has enabled us to assign specific functions to granule-bound and soluble starch synthases in the building of different size classes of glucans of the amylopectin clusters (Maddelein et al., 1994).

1118 The Plant Cell

Here, we turn our attention to starch synthesis and struc-ture in actively photosynthesizing and dividing algal cells. Weshow not only that the amounts, shape, and cellular locationof starch are changed but also that the polysaccharide isdramatically impoverished or devoid of a distinct amylose frac-tion, despite the presence of massive granule-bound starchsynthase (GBSS) activities. We further show that the struc-ture of amylopectin is modified, that the balance of starchsynthases is affected, and that at least one specific additionalgene (STA4) is needed for normal storage starch synthesis,which is not required for building starch during growth. Thisnew locus is involved in building normal storage amylopectinclusters. We speculate that different structural needs for starchfunction in photosynthesis and storage may help to explainthe emergence of starch as a distinct entity of the photosyn-thetic eukaryotic plant cell.

RESULTS

Figure 1. Electron Microscopy of Undepleted and Nitrogen-StarvedCells.

(A) Strain IJ2 grown in TAP medium (undepleted). The pyrenoid (P)and its starch sheath show their classic shapes in the single hugeChlamydomonas chloroplast with normal thylakoid membranes. Oc-casional starch granules can also be found in the stroma.(B) Strain IR16 in TAP-N medium. Nitrogen-starved cells are virtuallyfilled with starch and lipid bodies (L) after 4 days of incubation in mediumwithout nitrogen in the presence of acetate and light. The Chlamydo-monas eyespot (E) responsible for light perception in phototaxis isprominent in these cells. A small-sized pyrenoid-like (P) structure isalso visible. The size of the visible pyrenoids is reduced in mediumwithout nitrogen.Although spectacular decreases in starch amounts with respect to thewild-type strain were reported during storage (Maddelein et al., 1994),the double mutant strain IJ2 in (A) (sta2-29::ARG7sta3-1) shows onlya modest decrease (two- to fourfold) in undepleted cultures. Moreover,although the morphology and the size distribution of the double mutantgranules were strongly altered in nitrogen-starved cultures (Maddeleinet al., 1994), few if any modifications were seen during growth.Bars = 1 |im.

Log-Phase Wild-Type Chlamydomonas Starch Is Devoidof Amylose

The starch content of Chlamydomonas under unrestrictedgrowth conditions ranges from 0.5 to 4 u,g per 106 cells. Thiscan be compared with the 30 to 80 u,g of starch accumulatedunder growth arrest (in nitrogen-, phosphate-, or sulfur-starvedmedia). Most but not all of the starch typically surrounds thepyrenoid (Figure 1) and adopts a morphology dictated by theshape of this ribulose-1,5-biphosphate carboxylase/oxygen-ase-containing cell structure. In nutrient-starved cells, starchaccumulates mostly in the stroma during 48 hr. This accumu-lation correlates with a scavenging mechanism that yields anonphotosynthetic cell with disorganized thylakoid membranes(Bulte and Wollman, 1992). Moreover, pyrenoid-like structures,when visible, often lack their typical surrounding starch sheath.The disappearance or rearrangements of both the pyrenoidand chloroplast membranes coincide with the appearanceof massive lipid droplets. Our preliminary lipid compositionstudies pointed to the presence of galactolipids, suggestingthat these bodies originate from preexisting thylakoid mem-branes. These growth-arrested cells are virtually filled withstarch whose shape and cell distribution are altered. The com-position and structure of the polysaccharide from the growth-arrested cells were compared with those of the actively grow-ing and photosynthesizing algae. Figure 2 clearly shows thatlog-phase photosynthesizing algae build starches with eithera drastic decrease in or disappearance of amylose. The amylo-pectin also seems to be modified, as shown by a 20-nmincrease (from 550 ± 5 nm to 570 ± 5 nm) in the Xmax of theiodine-polysaccharide complex of the Sepharose CL2B chro-matograms (Figure 2).

Similar results were obtained using TSK HW-75(S) columnsin 10% DMSO. Different researchers have named intermedi-ate material a series of starch fractions of variable structures

Starch Synthesis in Chlamydomonas 11 19

i 1.5

60 70 80 90 100 110 120

FRACTION (NO)

Figure 2. Separation of Amylopectin and Amylose by Sepharose CL2B C h romatography.

The optical density (O) of the iodine-polysaccharide complex was mea- sured for each 3-mL fraction at I,,,, where I,,, is displayed as an unbroken thin lhe. All samples were loaded on the same column setup as described previously (Delrue et al., 1992). NO, number. (A) Wild-type haploid 137C starch purified from nitrogen-starved cultures. (E) Wild-type maize endosperm starch. (C) Starch composition of the wild-type Chlamydomonas strain (137C) grown in undepleted medium and harvested during log phase.

and sizes whose branching levels are intermediate between those of amylose and amylopectin. In Chlamydomonas, we have named amylopectin type II such an intermediate but ho- mogeneous high molecular mass fraction that contains >3% branches (Delrue et al., 1992). Interestingly, in this case, al- though starch-storing cells never fail to separate the previously characterized type I and II amylopectins, the latter was very poorly separated from type I polysaccharide in log-phase Chlamydomonas. This suggests either the disappearance of type II material or, more likely, a significant increase in molec- ular mass of this fraction. We designated storage starch as the polysaccharide that accumulated during nitrogen starva- tion. The polysaccharide synthesized by light-grown log-phase

(unstarved) Chlamydomonas cells was called photosynthetic (rather than transient) starch. Figure 2 also shows that Chlamy- domonas storage starch has a composition and a h,, of the purified amylopectin and amylose fractions that are very simi- lar to those of maize endosperm starch. This analogy extends further to the distribution of chain length sizes that typifies Chlamydomonas storage starch.

GBSS 1s Present and Active under Unrestricted Growth but Remains Unable To Synthesize Amylose

The absence of the low molecular mass amylose fraction in storage starch has been associated with defects in GBSS ac- tivity in Chlamydomonas and in many higher plant systems (Nelson and Rines, 1962; Visser et al., 1991; Delrue et al., 1992). To ascertain whether the enzyme responsible for amylose synthesis was present, we assayed both GBSS and proteins extracted from the starch granule. Results shown in Figure 3 indicate that the absence or decrease of amylose in the nitro- gen-supplied cultures was not doe to a decrease in GBSS activity and protein. However, the in vivo contribution of GBSS was still visible, but solely in the amylopectin fraction, con- firming the involvement of GBSS in amylopectin biosynthesis (Maddelein et al., 1994; see later discussion).

Conditional and Unconditional Expression of Mutants Defective for Starch Biosynthesis

Three genes (STA7, STA2, and STA3) to date have been shown to control starch amounts andlor structure in Chlamydomonas. Strains carrying a sfal-7 defect harbor reduced ADP-glucose pyrophosphorylase activity through desensitization of the en- zyme to 3-phosphoglycerate activation (Ball et al., 1991). We confirmed that the severe low-starch phenotype of this mu- tant was unconditional and expressed itself in both storage and photosynthetic starches in a similar if not identical fash- ion (Ball et al., 1991). STA2 is the GBSS structural gene (Delrue et al., 1992). We have shown that this enzyme is not only in- volved in amylose synthesis, as is generally believed, but is also responsible for synthesis of long chains in amylopectin (Maddelein et al., 1994). The starch chromatogram of a sfa2 disruption under unrestricted growth also lacked the amylose fraction. However, the amylopectin displayed a 25-nm drop in the I, , of the iodine-polysaccharide complex (Table 1). This correlates with a significant decrease in the amount of long chains we detected after debranching the purified amylopectin. These observations add additional support to the involvement of GBSS in the synthesis of the long glucans of amylopectin, as was previously demonstrated (Maddelein et al., 1994).

staScarrying mutants are defective for the major soluble starch synthase (SSS) activity (Fontaine et al., 1993). They dis- played a very significant reduction (60%) in storage starch accumulation and a specific decrease in amylopectin of those chains whose size ranged from 8 to 40 glucose residues in

1120 The Plant Cell

+N

Bi

kD94 *._

67 **"*

43 MM*-

30

-N +N +N

Figure 3. GBSS Activities and Proteins during Storage or Growth.

(A) Histogram of GBSS-specific activity of wild-type strain 137C ex-pressed in nanomoles of ADP-glucose incorporated into glucan permilligram of starch per minute. Means are given for three separatemeasurements.(B) A Coomassie Brilliant Blue R 250-stained 5 to 7.5% SDS-acryl-amide gel of starch-bound proteins. Lane 2 contains the starch-boundprotein fraction from the nitrogen-starved wild-type 137C strain; lane3 contains starch-bound proteins extracted from the same strain grownin undepleted medium. Lane 4 contains proteins extracted from un-depleted cultures of BAFR1, a strain carrying a gene disruption(sta2-29::arg7) for the GBSS gene. The molecular size standards inlane 1 are given in kilodaltons. Proteins simultaneously extracted fromequal amounts of polysaccharide (1 mg) were loaded on the gel. Themajor 76-kD band corresponds to the GBSS protein and has the typi-cal GBSS N-terminal sequence (Delrue et al., 1992).+N, nitrogen supplemented; -N, nitrogen starved.

length (Fontaine et al., 1993; Maddelein et al., 1994). TheseSSS-defective strains displayed their structural deficiencieson both storage and photosynthetic starches (Table 1). Thedouble mutants defective for both GBSS and SSSII (sfa2 andsfa3) displayed similar behavior. They are characterized in bothcases (storage and photosynthetic starch) by the presence ofvery small amounts of a polysaccharide whose structure isintermediate between those of amylopectin and glycogen. Yet,in log-phase cells, the deficiency in starch amounts of all sfa3sfa2 double mutants was much less pronounced relative tothe wild-type or mutant strains (Table 1). Also, although in allwild-type or single mutant genotypes there was a fourfold (forSSSII-defective strains) to 20-fold increase (for wild-type orGBSS-defective strains) in starch during storage, the strainscontaining SSSI only (sfa3 sta2 double mutants) were unableto trigger the increase in polysaccharide synthesis that normallytakes place under these conditions. Storage starch synthesisthus requires either SSSII, GBSS, or both. These observationscould be explained if one assumes a change in the ratio be-tween the two types of soluble starch synthases activities duringstorage. Although we were able to detect some modificationsin both our purification (by anion exchange) and zymogram(on native PAGE) assays, these changes in balance, when de-tectable, never exceeded a twofold relative decrease in SSSI.The SSSII/SSSI activity ratio measured in the presence of gly-cogen decreased from 0.6 to 0.3 when the cells switched tostorage. However, the total SSS activity also decreased innitrogen-deprived medium from 40 (during growth) to 9 (dur-ing storage) nmol of Glc incorporated into starch per hour per106 cells. It remains to be shown how these small shifts de-tected in soluble extracts affect the precise balance of starchsynthase activities at the very surface of the granule wheresynthesis is occurring.

STA4, a Novel Chlamydomonas Locus Necessary forStorage Amylopectin Synthesis, Is Not Required forStarch Synthesis during Growth

To determine whether modifications in structure localizationand amounts could be solely explained either by physiologi-cal modifications of carbon fluxes or by the selective actionof specific genes, we undertook a systematic screen for mu-tants expressing a clear-cut conditional phenotype. The latterconsisted of strong alterations in both structure and amountsof storage starch that would not affect polysaccharide synthe-sis during growth. Of a total of 2 x 104 cells surviving afterx-ray treatment, we found seven mutants altered for storagestarch synthesis. Six of these carried mutations that mappedto the previously characterized STA2 and STA3 genes. As men-tioned previously, they expressed their phenotypes on starchstructure under all growth conditions. Strain I73 exhibited asevere high-amylose phenotype that is in fact due to a defectin amylopectin content and structure (see later discussion).

Starch Synthesis in Chlamydomonas 1121

Table 1. Starch Amounts, Composition, and kmax Values of the lodine-Amylopectin Complex during Growth or Storage for Various Combinations of Mutations Decreasing Soluble and Granule-Bound Starch Synthases from Chlamydomonas

Growth

+ N - N + N - N + N - N + N - N

Genotype

( + +) ( + +) (sta2::Al + ) ( s ta2: :Al +) (+ sta3-7) ( + sta3-1) ( s r a 2 : : ~ 1 sta3-7) Ista2::AI sta3-7)

AP ( O 4

>95

>95 >95 >90

>95 >95

65 to 85

40 to 60

Am (010)

<5 570 15 to 35 <5 540 <5 540

5 to 10 590 60 to 40 590 <5 515 <5 51 5

kmax of AP (nm)

555 (Ap I)

Starch Amounts

3.7

3.5

2.5

0.8 1.5

31

29

14

To estimate starch amounts, three mutant meiotic segregants for each genotype class were selected from crosses involving parents carrying sfa3-7 and s ta2: :Al . s ta2: :Al is an abbreviation for sta2-29::ARG7, a strain whose STA2 locus has been disrupted by a functional ARG7 gene. Means of starch accumulation were calculated for each class using the amyloglucosidase assay for three distinct meiotic products. The amount of starch is expressed as micrograms of starch per 106 cells. Starch composition is given in weight percentages as assayed by the amyloglucosi- dase assay. Levels of amylose below 5% cannot be detected on starch chromatograms. Am, amylose; Ap, amylopectin; + N, during growth; - N, during storage.

The starch chromatograms displayed in Figure 4 show that the composition and h,, values of the fractions are altered with respect to the wild type. However, these defects could not be scored during growth, during which the starch chromato- gram and h,, values of the froctions were identical to those displayed in Figure 2C. In addition, a significant decrease (60%) in starch amounts was scored only during storage.

Characterization of the STA4 Defects

The sta4-1 defect segregated as a single Mendelian defect through meiosis. Typically, it is incompletely dominant on amy- lose content and to a lesser extent on the hmax values of the amylopectin fraction. Genetic analysis clearly demonstrated that this locus segregates independent from both the STAP and STA3 defects. As with the previously characterized SSS-defec- tive (sta3), high-amylose strains, we found significant decreases in storage starch amounts in the meiotic mutant progeny. By many other criteria, sta4-7-carrying strains behaved like sta3 mutants. They were dramatically modified in the chain length distribution (Figure 5) of the purified amylopectin and showed no change in that of amylose. As was the case for the sta3 defect, we found no decrease in the branching for the mutant amylopectin (which amounted precisely to 5%, as esti- mated by both methylation analysis and proton nuclear magnetic resonance [NMR]; Figure 6D). However, in this case the distri- bution of small glucans could be distinguished from both the wild-type and SSS-defective mutants and did not- show as clearly the maximal frequency of chains containing six glucose residues (degree of polymerization 6) that was seen in the lat- ter (Figure 6). The combination of heterozygous defects in both genes added up in diploids to a point at which the starch chro- matograms and the h,,, values of the purified amylopectin

670 T

h

v

x 620

E w cl 9 570

c $

8 520 5 2 2 ' 620

[ 570

520

-4-l-

1.5

1 F 5; F3

2 W 0.5

4 i2

O E 9 8 %

' 2

o. 5

0 O 50 60 70 80 90 100 110

FRACTION (NO)

Figure 4. Separation of Wild-Type and Mutant Amylopectin and Amy- lose by TSK HW-75(S) Chromatography.

The optical density (O) of the iodine-polysaccharide complex was mea- sured for each 3-mL fraction at I,,,=, where I,, is displayed as an unbroken thin line. All samples were loaded on the same column setup as described previously (Delrue et al., 1992). NO, number. (A) Wild-type haploid 137C starch purified from nitrogen-starved cultures. (8) Starch from nitrogen-starved 173 cells carrying the sta4-7 mutation.

11 22 The Plant Cell

3

2.5

2

1.5

1

0.5

O O

3

2.5

2

1.5

1

0.5

O O

A

150

1 O0

50

- Il- 60 80 1 O0

B

150

1 O0

50

- I& 60 80 100

800

700

600

500

400

800

700

600

500

400

Figure 5. Separation of lsoamylase Debranched Glucans by TSK H W-50(F) C h romatograp hy.

Four milligrams of starch fractions purified by gel filtration was loaded on each column after debranching. The optical density (O) of the iodine-polysaccharide complex was measured for each 2-mL fraction at hma. The optical density scale is shown at left on the y-axis. The amount (micrograms) of glucose per milliliter fraction is scaled on the inner side of the y-axis at left. The y-axis at right represents the wave- length (nanometers) scale. The x-axis shows the elution volume scale (milliliters). h,, values are displayed for all fractions for which it could be determined (broken line). The column setup and debranching con- ditions are described in Methods. The degree of polymerization (DP) scale (O) was generated by using the h,, values of the debranched glucans as interna1 standards according to Banks et al. (1971). (A) Amylopectin type I (Ap I) from wild-type strain 137C. (B) Amylopectin from strain 173 carrying sta4-7.

were indistinguishable from those of homozygous sta3 or sta4 diploid mutants (Figures 78 to 7F). They showed a different kind of genetic interaction when combined with the GBSS (sta2) defects (Figure 7A).

In both cases, the double mutant starch is lacking amylose, and the amylopectin becomes more highly branched (6 to 7%, according to proton NMR). However, and most importantly, sfa4 sta2 double mutants did not show the spectacular decrease in polysaccharide content that was seen in sra3 sfa2 mutants during storage. However, the decrease in starch amounts

remained significant, as revealed by the yellow color of the iodine-stained cell patches of strains carrying both mutations (Figure 8). Decreases of 50 and 75% were measured in the single mutant (50% decrease) and the wild-type (75Vo de- crease) strains (Figure 7G). This correlates with the absence of SSS defects that could be scored in sra4 strains. In fact, despite intensive attempts, we were unable to detecta modifi- cation in the amount of activity or kinetics either in crude extracts or after both anion exchange chromatography or na- tive PAGE of all enzymes that could be scored and that are known to be involved in starch biosynthesis. We therefore con- cluded that the STA4 gene encodes an as yet unidentified product that is necessary for normal amylopectin cluster bio- synthesis in storage starch only.

DISCUSSION

The Decrease or Absence of Amylose in Starch from Growing Cells

Wild-type starch is usually defined as a mix of two distinct frac- tions: amylose and amylopectin. Despite many attempts, we were able to score only trace amounts of a distinct amylose fraction in growing algae. It has been shown previously that GBSS is present in rate-limiting amounts for storage starch synthesis in higher plants (Tsai, 1974; Kuipers et al., 1994). Al- though this remains true for growth-arrested starch-storing Chlamydomonas, it is clearly not the case for growing algae. On the contrary, the specific activity of GBSS in its natural environment (the starch granule) displays a very spectacular increase. Moreover, under these conditions, GBSS seems to be fully active in vivo. Indeed, the contribution of this activity to the structure of amylopectin can be easily monitored in strains carrying deletions in the GBSS structural gene.

There are two simple explanations for these seemingly con- tradictory observations. The first isto hypothesize two distinct populations of GBSS encoded by a single locus. One of these would be surface bound on the granule and would have full access to the soluble enzymes necessary to obtain amylopec- tin. The second would be less accessible and involved in amylose biosynthesis. It is well known that GBSS is charac- terized by a Michaelis constant for ADP-glucose that is well over those displayed by the SSSs. However, when GBSS is solu- bilized, this Michaelis constant drops to a leve1 comparable to the soluble enzymes (Macdonald and Preiss, 1985). It is thus reasonable to suppose that starch itself is responsible for these differences and that, according to its position, GBSS can display variable affinities for the substrate. We would thus assume quite simply that the availability of ADP-glucose could be more critical to the synthesis of amylose deeper inside than to that of amylopectin at the very surface of the granule. A sec- ond mechanism that could explain our results would involve the presence of a critical balance between elongation and

Starch Synthesis in Chlamydomonas 1123

2 0 i A 16

/ B I 16 f -

16 1 l2 i

2 4 6 8 10 12 14

8

* l 1

I

11, 5.0

Figure 6. High-Performance Anion Exchange with Pulsed Amperome- tric Detection Chromatography of Water-Soluble Debranched Glucans.

Glucans differing in length by only one glucose residue are clearly separated up to degree of polymerization 25 (DP 25). Except in (D), results are given in relative frequencies histograms of chains ranging from DP 3 to DP 15. These frequencies were computed from the chro- matogram peak surfaces, with the total amount of chains from 3 to 15 adjusted to 100%.

branching activities at the surface of the granule where syn- thesis is taking place. Any event that would affect that balance toward branching would impede amylose formation by GBSS. As has been observed for the relative expression of both SSS activities, a modification in enzyme balance is likely to occur when the plant cell switches from growth to storage. Both of these explanations can be put to the test by studying the struc- ture and composition of starch in low ADP-glucose-containing mutants and by setting up cell-free starch-synthesizing sys- tems from purified native granules.

Different Structure-Function Relationships during Storage and Growth Might Explain the Appearance of Starch in Photosynthetic Eukaryotes

Amylose synthesis is not the only event that distinguishes stor- age from recurrent starch synthesis coupled to photosynthesis and cell division. The fine structure of purified amylopectin from the growing algae is also different from that characteriz- ing storage starch. This correlates with a modification in the cellular location of starch synthesis that occurs predominantly around the pyrenoid in low-C02 growing cultures and that is confined to the stroma during storage.

Starch synthesis in growing cells has also been reported to respond to the availability of C02 by switching location (Ramazanov et al., 1994). This could reflect changes in the distribution of the active Calvin cycle enzymes or the building of distinct multienzyme complexes. On the other hand, starch in its pyrenoidal form may be involved as an active compo- nent of the C02 concentration mechanism. It is striking that SSSl on its own is unable to deal with storage yet remains perfectly able to yield what appears to be a normal pyrenoidal sheath of polysaccharide. Thus, different biosynthetic enzymes might have evolved to deal with many distinct physiological constraints in plants. This might explain the multiplicity of en- zymes catalyzing the same biochemical reaction. It is likely that specific enzyme forms would assume a predominant func- tion under quite different circumstances. The latter include synthesis not only during photosynthesis in high and low COp but also storage after growth arrest.

Storage in plants seems to have favored the appearence of a polysaccharide of high glucose-storing capacity with very

(A) Wild-type type I amylopectin. (B) Amylopectin from the SSSII-deficient strain 1152. (C) Amylopectin from the high-amylose 173 strain. (D) An example of a chromatogram for strain 173 harboring the sta4-7 defect. The degrees of polymerization of some of the chains are given above the chromatogram. Part of the proton NMR spectrum is also displayed, showing the signals used to quantitate the branching of the purified 173 amylopectin.

:::I A

0.9

670

1 6 2 0

30

20

10

O

O 85 100 115 130 145 160

1.2 .'

O 85 100 115 130 145 160

FRACTION (NO)

30 G +N

20

10

- 0 TB1 TB2 TB3 TB4

H

TB1 TB2 TB3 TB4

Figure 7. Genetic Analysis of the sta4-7 Defect.

In (A) to (F) are TSK HW-75(S) (in 100/0 DMSO) starch chromatograms from nitrogen-starved cells. The optical density (O) of the iodine-poly- saccharide complex was measured for each 3-mL fraction at I,,, where I,,,, is displayed as an unbroken thin lhe. All samples were loaded on the same column setup as described previously (Delrue et al., 1992). 60th heterozygotes ([C] and [O]) contain more amylose and have a slight significant increase (10 nm) in the Lm, value of the major amylopectin species. NO, number. (A) Starch chromatogram from the haploid recombinant strain TB3 (sta4-7 sta2-6) whose phenotype is shown in Figure 8. (e) Starch chromatogram from a wild-type diploid reference (obtained by crossing 137C with strain 37). (C) Starch chromatogram of a diploid heterozygous for STA3. (D) Starch chromatogram of a diploid heterozygous for STA4. (E) Starch chromatogram of a diploid simultaneously heterozygous for STA3 and STA4. (F) Starch chromatogram of a homozygous mutant (sta3-7 sta3-7) diploid. (G) and (H) Amounts of starch accumulated during growth (+N) or storage (-N) for the four genotype classes. The y-axis is expressed in micro- grams of starch per 106 cells. Values are means (n = 3) from three separate experiments.

Starch Synthesis in Chlamydomonas 1125

TB1

TB2 (+sta4-l)

TB3 (sta2-6 sta4-l)

TB4 (sta2-6

Figure 8. Phenotype of Wild-Type and Mutant Strains.

Cell patches from a tetratype tetrad generated from a cross betweenstrains I73 and 37E-8J. The genotype corresponding to each recom-binant is shown in parentheses. Cells were incubated for 5 days onsolid nitrogen-deprived medium and sprayed twice with iodine vapors.

low osmotic pressure (lower than that of animal, fungal, or bac-terial glycogen). This in turn yields a very tight crystallinepacking of glucan chains that is difficult to degrade. On a shorttime scale, such as that required during growth, the crystalpacking probably limits the availability of glucose stores to thecatabolic enzymes known to be present in the chloroplast. Onthe other hand, specific processes making use of storagestarch, such as tuber, seed, or zygospore germination, all re-quire the synthesis of a new set of catabolic enzymes. It wouldbe of great interest to know the precise structural differencesbetween the two kinds of starches, including the amount andtype of crystals found in both cases. Although it is clear to usthat the starch pathway is common for most steps for both kindsof polysaccharide, it also involves a specific set of genes andfunctions. The STA4 locus encodes what is likely to be a goodcandidate for a function required for storage starch synthesisonly. The profound modifications of the amylopectin structureand amylose content due to the sta4 mutations are reminis-cent of the maize dulll (du1), amylose extender! (ae1), and,even more so, sugary2 (su2) mutations. DU1 was reported tobe a regulatory locus affecting SSSII and branching enzymetype lla (Preiss and Boyer, 1980). The gene product of SU2has eluded characterization. AE1 is now known to be the struc-tural gene of the maize type II branching enzyme (Stinard etal., 1993). In Chlamydomonas, novel gene-tagging proceduresshould enable us to screen effectively for high-amylose sta4mutants and thus allow us to determine the nature of the geneproduct.

METHODS

Materials

Glucose 1-phosphate-U-"C and D-glucose-U-14C-ADP-glucose werepurchased from Amersham. ADP-glucose, maize amylopectin, andPseudomonas amyloderamosa isoamylase were purchased fromSigma. Glucose 1-phosphate, rabbit muscle glycogen, and rabbit mus-cle phosphorylase were obtained from Boehringer Mannheim.

Chlamydomonas reinhardtii Strains, Growth Conditions,Cytological Observations, and Media

The wild-type reference Chlamydomonas strain used in this study was137C (mf~ nitl nit2). Diploids were selected by complementation onminimal medium after crossing with either strain 37 (mt * pab2 ac14),strain B9 (mt+ pab2 ac14 sta3-1), or 37E-17 (mt ~ pab2 ac14 sta3-1).IR16 and IJ1 strains are meiotic segregants of a cross between BAFR1(mf+ nitl nit2 cw15 an?7-7 sfa2-29: :ARG7) and strain 37E-17. IR16 con-tains the sfa2-29::/\flG7gene disruption; IJ1 contains both sta2-29:'ARG7and sfa3-7. I73 was obtained from strain 137C by x-ray mutagenesisat 104 R, leading to 4% survival, and defined the sfa4-7 defect.

To examine genetic interactions between the STA2 and STA4 loci,tetrads of a cross between I73 and 37E-8J (mf+ pab2 ac14 sta2-6)were dissected. TB1 (mf ~ nitl and/or n/f2), TB2 (mf ~ sfa4-7 pab2 ac14nitl and/or n/f2), TB3 (mf+ sfa4-7 sfa2-6 ac74 nitl and/or n/f2), and TB4(mf+ sfa2-6 pa£>2) are derived from the same tetratype tetrad.

All experiments were performed in continuous light (40 nE m~2

sec~1) in the presence of acetate at 24°C in liquid cultures that wereshaken vigorously without air or CO2 bubbling. Late log-phase cul-tures were inoculated at 105 cells mL~' and harvested at 2 x 106 cellsmL~1. Nitrogen-starved cultures were inoculated at 5.105 cells mL~1

and harvested after 4 days at a final density of 1 to 2 x 106 cells mL~1.Genetic techniques are described by Harris (1989a). Standard Tris-acetate-phosphate (TAP) medium is fully detailed by Harris (1989b);nitrogen-starved medium (TAP-N) and diploid clone selection are de-scribed by Ball et al. (1990, 1991) and Delrue et al. (1992). Fixationand embedding protocols are as described by Harris (1989c).

Measures of Starch Levels, Starch Purification, and SpectralProperties of the Iodine-Starch Complex

A full account of amyloglucosidase assays, starch purification on Percollgradients, and Xmax measurements is provided by Delrue et al. (1992).

Crude Extract Preparation, Enzyme Assays, PartialPurification of Enzyme Activities, and Zymograms

Soluble crude extracts were always prepared from late log-phase cells(2 x 106 cells mL~1) grown in high-salts acetate medium under con-tinuous light (80 n£ m~2 sec~1). The detailed description of thedifferential (NH4)2SO4 precipitation of soluble starch synthase I (SSSI)and SSSII, together with the anion exchange purification on DEAETrisacryl type M (IBF Biotechnics, Villeneuve la Garenne, France) ofthose enzyme activities, is provided by Fontaine et al. (1993). SSS ac-tivity was assayed in a 0.1-mL final volume of 50 mM glycine-NaOH,pH 9, 100 mM (NH4)2SO4, 5 mM (3-mercaptoethanol, 5 mM MgCI2, 0.5

11 26 The Plant Cell

mg mL-l BSA, 10 mg mL-l rabbit liver glycogen, and 4 mM ADP- glucose containing 1 nmol of o-glu~ose-U-~~C-ADP-glucose (specific activity 200 WCi pmol-l). After a 15-min incubation at 3OoC, the reac- tion was stopped by adding 2 mL of ice-cold ethanol. Granule-bound starch synthase (GBSS) was assayed as described by Delrue et al. (1992). Branching enzymes were always assayed on the same DEAE chromatograms as used for the SSSs by incubating up to a 40-pL sam- ple in a0.2-mLfinal volume of 0.1 M sodium citrate, pH 7.0,l mM AMP, 40 Wg of rabbit liver phosphorylase containing 50 mM glucose l-phos- phate-4-14C (final specific activity of 0.22 pCi pmol-l). After a 30-min incubation at 3OoC, the reaction was stopped by addition of 10% tri- chloroacetic acid.

The resulting precipitate was filtered, rinsed, dried, and counted in a liquid scintillation counter (assay A). Amylase, phosphoglucomutase, ADP-glucose pyrophosphorylase, and phosphorylase activities were monitored by using the standard assays described by Ball et al. (1991). For SSSs, the analysis was completed by zymograms as described by Maddelein et al. (1994).

Starch Fractionation, Methylation, Nuclear Magnetic Resonance, and Debranching Analyses

Separation of starch fractions on TSK HW-75(S) columns (Merck) was performed as detailed previously (Delrue et al., 1992). In this study, we also used Sepharose CL2B chromatography with 10 mM NaOH as solvent on the same column setups. In thiscase, thestarch sample (10 mg) was first dissolved at 100°C in 90% DMSO and precipitated with four volumes of pure ethanol for 48 hr at room temperature. The precipitate was harvested by spinning at 50009 for 20 min and redis- solved without overdrying in 5 mL of 10 mM NaOH. Methylation of total starch or of pooled fractions dialyzed and freeze-dried after TSK HW-75(S) chromatography was performed according to Paz Parente et al. (1985) and adapted to starch analysis (Delrue et al., 1992). The branching percentage was assayed as the ratio of methyl ether deriv- atives of a(l+4)-linked glucose either to those of a(1+4)- and a(l+)-linked glucose or to those of glucose in terminal nonreducing position (Delrue et al., 1992; Fontaine et al., 1993). Nuclear magnetic resonance (NMR) analysis was performed as described by Fontaine et al. (1993). The leve1 of branching was estimated by integration of the same regions of proton resonances of the monosubstituted and disubstituted glucose ( 8 ~ 5 . 2 and 4.85 parts per million, respectively; Gidley, 1985). Isoamylase-mediated debranching of fractions purified by gel filtration was achieved as previously described (Maddelein et al., 1994).

for 30 sec in a microcentrifuge. After cooling, the supernatant was directly transferred to a 1-mL spectrophotometer cuvette containing 0.1 mLof 1% KI, 0.1% iodine. The full (between 400 to700 nm) iodine- polysaccharide complex spectrum was recorded. Provided the absorb- ance taken at h,, ranged between 0.3 to 1, each spectrum recorded from a cross between strain 173 (carrying sfa4-7 ) and the wild-type reference strain 37 fel1 into two very distinct phenotype classes. The first (high-amylose) class displayed a typical shoulder above 600 nm and had a h,, value generally (but not always) above 615 nm. The second class had a more regularly shaped spectrum with no shoulder above 600 nm and a hmax value generally below 605 nm. Selected members of the first class (even those with a Iow 1,, value) never failed to display a starch chromatogram on TSK HW-75(S) columns identical to that of strain 173, whereas all members tested of the sec- ond class displayed starch chromatograms similar to those of wild-type strains. More than 200 colonies were analyzed using this technique and yielded a clean single-gene inheritance pattern. The same analy- sis was performed on the meiotic progeny (122 colonies analyzed) of a cross performed between strains 173 and B9 (carrying sfa3-I). In this case, 27 wild-type recombinants were scored, and double checked by TSK HW-75(S) chromatography, demonstrating independent segre- gation of STA3 and STA4. lndependent segregation of STA2 and STA4 was demonstrated by similar techniques using both random spores and tetrad analysis after crossing strain 173 with 37E-8J.

ACKNOWLEDGMENTS

This work was supported by the Université des Sciences et Technolo- gies de Lille, the Ministère de I'Education Nationale, the Centre National de Ia Recherche Scientifique (lJnit6 Mixte de Recherche du CNRS nO111; Andr6 Verbert, director), and special grants from the starch pro- cessing company Roquette Frbres (Lestrem, France) and from the Région Nord Pas-de-Calais (Réseau dEnseignement et de Recher- che des Universit6s du Grand-Nord). We thank André Dhainaut and Claude Defives (Unité de Formation et de Recherches de Biologie, Université des Sciences et Technologies de Lille) for helping with the electron microscopy. We also thank Jocelyn Celen for helping us with the artwork.

Received February 10, 1995; accepted May 3, 1995.

REFERENCES Genetic Analysis ot the sta4-1 Defect

The segregation of the high-amylose sta3 mutations in crosses can be easily followed by simply scoring the color of nitrogen-starved cell patches sprayed with iodine vapors. The sta4-7 mutation, on the other hand, yields a phenotype that systematically requires double check- ing by recording the full spectrum of the iodine-polysaccharide complex. Briefly, individual patches inoculated as 10-pL drops on solid TAP-N were grown for 4 to 7 days under continuous light. The cell patches were then sprayed twice at 24-hr intervals with iodine vapors. This treatment ensured complete bleaching and fixation. The equivalent of one bleached cell patch was scraped off the plate and suspended in 0.4 mL of 0.1 M HCI. The suspension was incubated for 1 min at 100°C, and the cell pellet was discarded after spinning

Ball, S.G., Dirick, L., Decq, A., Martiat, J.C., and Matagne R.F. (1990). Physiology of starch storage in the monocellular alga Chlamydo- monas reinhardfii. Plant Sci. 66, 1-9.

Ball, S., Marianne, T., Dirick, L., Fresnoy, M., Delrue, B., and Decq, A. (1991). A Chlamydomonas reinhardfii low-starch mutant is defec- tive for 3-phosphoglycerate activation and orthophosphate inhibition of ADP-glucose pyrophosphorylase. Planta 185, 17-26.

Banks, W., Greenwood, C.T., and Khan, K.M. (1971). The interaction of linear amylose oligomers with iodine. Carbohydr. Res. 17,25-33.

BultB, L., and Wollman, F.A. (1992). Evidence for a selective de- stabilization of an integral membrane protein, the cytochrome bdf

Starch Synthesis in Chlamydomonas 1127

Macdonald, F.D., and Preiss, J. (1985). Partia1 purification and char- acterization of granule-bound starch synthases from normal and waxy maize. Plant Physiol. 78, 849-852.

Maddelein, M.-L., Libessart, N., Bellanger, F., Delrue, B., DHulst, C., Van Den Koornhuyse, N., Fontaine, T., Wieruszeski, J.M., Decq, A., and Ball, S.G. (1994). Toward an understanding of the biogenesis of the starch granule: Determination of granule-bound and soluble starch synthase functions in amylopectin synthesis. J. Biol. Chem. 269, 25150-25157.

Manners, D.J. (1989). Recent developments in our understanding of amylopectin structure. Carbohydr. Polymers 11, 87-112.

Müller-Rober, B., Sonnewald, U., and Willmitzer, L. (1992). Inhibi- tion of the ADP-glucose pyrophosphorylase in transgenic potatoes leads to sugar-storing tubers and influences tuber formation and expression of tuber storage protein genes. EMBO J. 11, 1229-1238.

Nelson, O.E., and Rines, H.W. (1962). The enzymatic deficiency in the waxy mutant of maize. Biochem. Biophys. Res. Commun. 9,297-300.

Ozbun, J.L., Hawker, J.S., and Preiss, J. (1972). Soluble adenosine diphosphate glucose-a-lP-glucan a-4-glucosyltransferases from spinach leaves. Biochem. J. 126, 953-963.

Paz Parente, J., Cardon, P., Leroy, Y., Montreuil, J., Fournet, B., and Ricart, G. (1985). A convenient method for methylation of gly- coprotein glycans in small amounts by using lithium methyl sulfonyl carbanion. Carbohydr. Res. 141, 41-47.

Preiss, J. (1991). Biology and moiecular biology of starch synthesis and its regulation. In Oxford Surveys of Plant Molecular and Cell Biology, Vol. 7, B.J. Miflin, ed (Oxford: Oxford University Press), pp.

Preiss, J., and Boyer, C.D. (1980). Evidence for independent genetic control of the multiple forms of maize endosperm branching enzymes and starch synthases. In Mechanisms of Saccharide Polymeriza- tion and Depolymerization, J.J. Marshall, ed (New York: Academic Press), pp. 161-174.

Ramazanov, Z., Rawat, M., Henk, M.C., Mason, C.B., Matthews, S.W., and Momney, J.V. (1994). The induction of the COp-concen- trating mechanism is correlated with the formation of the starch sheath around the pyrenoid of Chlamydomonas reinhardfii. Planta

Stinard, P.S., Robertson, D.S., and Schnable, P.S. (1993). Genetic isolation, cloning, and analysis of a Murafor-induced, dominant antimorph of the maize amylose extenderl locus. Plant Cell 5, 1555-1566.

Tsai, C.-Y. (1974). The function of the waxy locus in starch synthesis in maize endosperm. Biochem. Genet. 11, 83-96.

Visser, R.G.F., Somhorst, I., Kuipers, G.J., Ruys, N.J., Feenstra, W.J., and Jacobsen, E. (1991). lnhibition of the expression of the gene for granule-bound starch synthase in potato by antisense con- structs. MOI. Gen. Genet. 225, 289-296.

59-1 14.

195, 210-216.

complex, during gametogenesis in Chlamydomonas reinhardfii, Eur. J. Biochem. 204, 327-336.

Caspar, T., Huber, S.C., and Somerville, C. (1985). Alterations in growth, photosynthesis, and respiration in a starchless mutant of Arabidopsis fhaliana (L.) deficient in chloroplast phosphoglucomu- tase activity. Plant Physiol. 79, 11-17.

Delrue, B., Fontaine, T., Routier, F., Decq, A., Wieruszeski, J.M., Van Den Koornhuyse, N., Maddelein, M.-L., Fournet, B., and Ball, S. (1992). Waxy Chlamydomonas minhardfii: Monocellular alga1 mu- tants defective in amylose biosynthesis and granule-bound starch synthase accumulate a structurally modified amylopectin. J. Bac- teriol. 174, 3612-3620.

Fontaine, T., DHulst, C., Maddelein, M.-L., Routier, F., Marianne- Pepin, T., Decq, A., Wieruszeski, J.M., Delrue, B., Van Den Koornhuyse, N., Bossu, J.P., Fournet, B., and Ball, S.G. (1993). Toward an understanding of the biogenesis of the starch granule: Evidence that Chlamydomonas soluble starch synthase II controls the synthesis of intermediate size glucans of amylopectin. J. Biol. Chem. 268, 16223-16230.

Ghosh, H.P., and Preiss, J. (1966). Adenosine diphosphate glucose pyrophosphorylase: A regulatory enzyme in the biosynthesis of starch in spinach chloroplasts. J. Biol. Chem. 241, 4491-4504.

Gidley, M.J. (1985). Quantification of the structural fsatures of starch polysaccharides by N.M.R. spectroscopy. Carbohydr. Res. 139, 85-93.

Hanson, K.R., and McHale, N.A. (1988). A starchless mutant of Nico- fiana sylvesfris containing a modified plastid phosphoglucomutase. Plant Physiol. 88, 838-844.

Harris, E.H. (1989a). Genetic analysis. in The Chlamydomonas Source- book: A Comprehensive Guide to Biology and Laboratory Use, E. Harris, ed (San Diego: Academic Press), pp. 399-446.

Harris, E.H. (1989b). Culture and storage methods. In The Chlamydomo- nas Sourcebook: A Comprehensive Guide to Biology and Laboratory Use, E. Harris, ed (San Diego: Academic Press), pp. 25-63.

Harris, E.H. (1989~). Histological techniques for Chlamydomonas. In The Chlamydomonas Sourcebook: A Comprehensive Guide to Bi- ology and Laboratory Use, E. Harris, ed (San Diego: Academic Press), pp. 581-586.

Jenkins, P.J., Cameron, R.E., and Donald, A.M. (1993). A universal feature in the starch granules from different botanical sources. Starke

Kuipers, A.G.J., Jacobsen, E., and Visser, R.G.F. (1994). Formation and deposition of amylose in the potato tuber starch granule are affected by the reduction of granule-bound starch synthase gene expression. Plant Cell 6, 43-52.

Lin, T.-P., Caspar, T., Somerville, C., and Preiss, J. (1988). lsolation and characterization of a starchless mutant of Arabidopsis thaliana (L.) Heynh. lacking ADPglucose pyrophosphorylase activity. Plant Physiol. 86, 1131-1135.

45, 417-420.