The phenotype of soluble starch synthase IV defective mutants of Arabidopsis thaliana suggests a novel function of elongation enzymes in the control of starch granule formation

The phenotype of soluble starch synthase IV defectivemutants of Arabidopsis thaliana suggests a novel functionof elongation enzymes in the control of starch granuleformation

Isaac Roldan1, Fabrice Wattebled2, M. Mercedes Lucas3, David Delvalle2, Veronique Planchot4, Sebastian Jimenez1, Ricardo

Perez5, Steven Ball2, Christophe D’Hulst2 and Angel Merida1,*

1Instituto de Bioquımica Vegetal y Fotosıntesis. CSIC-US. Avda Americo Vespucio 49. Isla de la Cartuja, 41092-Sevilla, Spain,2Unite de Glycobiologie Structurale et Fonctionnelle, UMR8576 CNRS/USTL, IFR 118, Universite des Sciences et Technologies

de Lille, 59655 Villeneuve d’Ascq Cedex, France,3Instituto de Recursos Naturales, Centro de Ciencias Medioambientales, CSIC. c/Serrano 115bis, 28006 Madrid, Spain,4Unite de Recherche Biopolymeres, Assemblages et Interactions, Centre INRA de Nantes, rue de la Geraudiere, B.P. 71627,

44316 Nantes Cedex 3, France, and5Instituto de Investigaciones Quımicas, CSIC-US, Avda Americo Vespucio 49, Isla de la Cartuja, 41092 Sevilla, Spain

Received 31 July 2006; revised 20 September 2006; accepted 28 September 2006.*For correspondence (fax þ34 95 446 0065; e-mail [email protected]).

Summary

All plants and green algae synthesize starch through the action of the same five classes of elongation enzymes:

the starch synthases. Arabidopsis mutants defective for the synthesis of the soluble starch synthase IV (SSIV)

type of elongation enzyme have now been characterized. The mutant plants displayed a severe growth defect

but nonetheless accumulated near to normal levels of polysaccharide storage. Detailed structural analysis has

failed to yield any change in starch granule structure. However, the number of granules per plastid has

dramatically decreased leading to a large increase in their size. These results, which distinguish the SSIV

mutants from all other mutants reported to date, suggest a specific function of this enzyme class in the control

of granule numbers. We speculate therefore that SSIV could be selectively involved in the priming of starch

granule formation.

Keywords: starch synthase, starch granule, starch granule initiation, amylopectin, Arabidopsis, knock-out

mutant.

Introduction

Although most eukaryotes bacteria and archea accumulate

a-(1 fi 4) linked and a-(1 fi 6) branched storage polysac-

charides in the form of glycogen, green algae and land

plants have resorted to the synthesis of starch, a far more

complex semi-crystalline structure that accumulates as large

insoluble granules within their plastids.

Amylopectin defines the major branched polysaccharide

fraction of starch. Its synthesis requires the concerted action

of different enzymatic activities: ADP-glucose pyrophosph-

orylases, starch synthases, branching enzymes and starch-

debranching enzymes (Ball and Morell, 2003; Myers et al.,

2000). Starch synthases (SSs) catalyze the transfer of the

glucosyl moiety of ADP-glucose (the activated glucosyl

donor) to a pre-existing a-(1 fi 4) glucan primer. In a

recent study it has been demonstrated that, despite its tiny

genome, Ostreococcus tauri, a prasinophyte alga thought to

have diverged at the earliest stage within the green linage1 ,

displays the same number and family types of SSs as those

documented in either the rice or the Arabidopsis genomes

(Ral et al., 2004). This high degree of conservation of the

pathway suggests that these enzymes play a conserved and

specific function in the building of starch. Five distinct SS

families have been reported in plants. Four of these (SSI–

SSIV) are predominantly found as soluble enzymes,

492 ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd

The Plant Journal (2007) 49, 492–504 doi: 10.1111/j.1365-313X.2006.02968.x

whereas the fifth enzyme is granule-bound (GBSSI). Mutants

for all the SSs have been obtained and analyzed in distinct

plant and algal systems with the noticeable exception of

SSIV. Predicted SSIV proteins show a C-terminal region

highly similar to other SSs, which include the catalytic and

starch-binding domains (Cao et al., 1999). On the contrary,

the N-terminal half of SSIV protein differs significantly from

other SS isoforms. Recently, Hirose and Terao (2004) have

shown that the two SSIV genes found in rice (SSIV-1 and

SSIV-2) are expressed throughout the plant and at relatively

constant levels during the grain filling. However, genetic

evidence for a role of this enzyme in starch biosynthesis is

still lacking.

The picture emerging from the studies performed with all

other SSs (SSI–III and GBSSI) in Chlamydomonas, pea,

potato, Arabidopsis and cereals is that each enzyme is

responsible for the synthesis of specific size classes of

glucans within the amylopectin structure (Craig et al., 1998;

Delvalle et al., 2005; Fontaine et al., 1993; James et al., 2003;

Maddelein et al., 1993; Morell et al., 2003; Zhang et al.,

2005). In addition, GBSSI seems to be the only elongation

enzyme involved in the biosynthesis of amylose, the second

unbranched and dispensable polysaccharide fraction found

within starch (Delrue et al., 1992; Klosgen et al., 1986).

Despite these specialized functions, some SSs display some

degree of functional overlap whereas others do not. We now

report on the characterization of a novel class of SS mutant

in Arabidopsis. Mutants of SSIV are shown to display a

modest but significant decrease in starch levels. At variance

with all other SS mutants, the SSIV-lacking plants display a

severe growth defect phenotype with either little or no

modification in either starch granule composition or chain

length (CL) distributions. The mutants, however, have

dramatically decreased the number of starch granules

synthesized within the plastids. Consequently they have

significantly increased the size of the latter. We believe these

results demonstrate an important and selective function of

SSIV in the control of starch granule numbers. We speculate

that SSIV is involved in the priming of starch granule

formation.

Results

Levels of SSIV mRNA in different organs

As a first step in the characterization of the function of the

SSIV isoform, the spatial pattern of expression of the AtSS4

gene (locus At4g18240) was established. Using quantitative

real-time RT-PCR, this pattern was compared with that of the

other classes of SSs (SSI, SSII and SSIII encoded by AtSS1,

AtSS2 and AtSS3 loci respectively). The four genes were

expressed in all organs studied (leaves, roots, flowers and

immature fruits). In all cases the steady state level of AtSS1

mRNA was one order of magnitude higher than that of the

other AtSS genes (Figure 1a). On the other hand, AtSS4

mRNA accumulated at similar levels in all organs analyzed

with values equivalent to those obtained for the AtSS3 gene

(Figure 1b).

Isolation of mutant lines defective in SSIV

The AtSS4 (At4g18240) gene is located in chromosome 4

and is composed of 16 exons and 15 introns. It encodes a

1040 amino acid protein with a predicted mass of

117 747 Da. This protein shows a high level of similarity with

the previously annotated SSIV proteins found in other spe-

cies such as Vigna unguiculata (71% identity, accession

number AJ006752), wheat (58.2% identity, accession num-

Figure 1.25 Expression profile of Arabidopsis starch synthase genes. The

absolute mRNA levels of all the Arabidopsis soluble starch synthase encoding

genes (AtSS1, At5g24300; AtSS2, At3g01180; AtSS3, At1g11720; AtSS4,

At4g18240) were determined by real-time quantitative RT-PCR as described in

Experimental procedures. Data shown in Panel B are the same as for Panel A,

but the AtSS1 gene expression was omitted to make a clearer comparison

between the other AtSS genes. Values are the average of three determina-

tions of at least two cDNA preparations from different experiments. Samples

were harvested at midday from 21-day-old plants.

Function of starch synthase type IV 493

ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 49, 492–504

ber AY044844) or rice (56.8% identity with SSIVa, accession

number AY373257; 58.3% identity with SSIVb, accession

number AY373258). Bioinformatic analysis predicted the

presence of a chloroplast-targeting signal comprising the

first 42 amino acids, rendering a mature protein of

112 997 Da (ChloroP at http://www.cbs.dtu.dk/services/

ChloroP/; Emanuelsson et al., 1999). Two independent mu-

tant alleles, Atss4-1 [Columbia-0 (Col-0) ecotype] and Atss4-

2 [Wassilewskija (WS) ecotype], designed to correspond to

T-DNA insertions in the AtSS4 gene were found in the GABI-

KAT (http://www.gabi-kat.de/; Rosso et al., 2003) and the

Genoplante (http://www.evry.inra.fr/public/projects/bioinfo/

flagdb.html2 ; Balzergue et al., 2001) mutant collections,

respectively. The T-DNA insertions are located in intron 11

and 2 (position þ3763 and þ227 bp with respect to the start

codon for Atss4-1 and Atss4-2, respectively; Figure 2).

Homozygous mutant plants were selected and expression

of the AtSS4 mutant alleles was analyzed by RT-PCR using

specific oligonucleotides for different regions of the gene.

This analysis indicated that a modified messenger corres-

ponding to AtSS4 was still present in both alleles (data not

shown). Western blot analysis was then performed to check

for the absence of SSIV protein in both mutant lines. The

rabbit antiserum used was raised against a 178 amino acids

polypeptide fragment of SSIV protein corresponding to a

region that displays no similarity with all other SS isoforms

(from Glu236 to Glu414) (see Experimental procedures).

Western blots showed the presence of two close bands with

a mass of approximately 112 kDa in both wild-type (WT)

ecotypes (Figure 2b). These bands match the size of the

expected mature SSIV protein and are both absent in the two

mutant alleles. Truncated versions of the SSIV protein were

not detected in the mutant lines either (smaller, unspecific

bands were found in both mutant and WT lines; Figure 2b).

The presence of these two bands in both WT ecotypes is not

yet understood but could be a result of a post-translational

modification of the protein.

Phenotypic characterization of Atss4 mutant alleles

The absence of the SSIV protein has a deleterious effect on

plant growth. Both mutant lines showed lower growth rates

under a 16-h day/8-h night photoregime when compared

with their respective WT ecotypes (Figure 3, panels b and d).

Rosette leaves of mutant alleles were smaller than those of

WT (Figure 3, panels a and c). In addition we recorded a

delay in flowering time: 31 � 3 days for Atss4-1 versus

25 days for Col-O, and 21 days for Atss4-2 versus

18 � 1 days for the WS ecotype. However the number of

rosette leaves at bolting was the same in mutant and WT

plants, indicating that the delay in flowering time comes as a

consequence of the reduced growth rate in the mutant lines.

Fruit size, number of seeds per silique and germination

ratios were not altered in the mutant lines.

The quantity of leaf starch was also determined in the

mutant lines. The starch content at the end of the illuminated

period was reduced in both cases: a 35% decrease for the

Atss4-1 and a 40% decrease for the Atss4-2 line with respect

to their WT genetic backgrounds. A more detailed analysis of

starch accumulation over the day/night cycle was carried out

on both mutant lines. As shown in Figure 4, starch turnover

along a diurnal cycle was lower in mutant plants than in WT

plants, with a clear reduction of both synthesis and degra-

dation rates. Starch levels at the beginning of the light

period were higher in the Atss4 alleles than in their

respective WT plants. However, the reduced rate of starch

synthesis in the mutant led to a lower starch level at the end

of the illuminated period.

Growth rate and levels of starch accumulation analogous

to those in WT plants were restored when AtSS4 protein was

expressed in the Atss4-1 mutant allele (Figure 5), indicating

that those phenotypic alterations are a consequence of the

absence of the AtSS4 protein.

Finally, the levels of water soluble polysaccharides (WSP)

and low molecular weight sugars closely related to starch

metabolism, such as maltose, sucrose, fructose and glucose,

Figure 2.25 Analysis of mutant lines in locus AtSS4.

(a) Genomic structure of the AtSS4 locus. Exons and introns are indicated as

thick and thin black bars, respectively. Insertion sites of T-DNA in mutant lines

Atss4-1 and Atss4-2 at introns 11 and 2 respectively are indicated by triangles.

(b) Western blot analysis of crude leaf extracts of Atss4-1 and Atss4-2 mutant

alleles and their respective wild-type ecotypes. Proteins (25 lg) were separ-

ated by SDS-PAGE electrophoresis, transferred to nitrocellulose filters and

immunolabelled with rabbit antiserum raised against a 178 amino acids

fragment of the N-terminal region of soluble starch synthase IV (SSIV) protein

(see Experimental procedures). The molecular markers (kDa) positions are

indicated. Bands of approximately 112 kDa matching the predicted SSIV mass

in Columbia-0 (Col-0) and Wassilewskija (WS) wild-type ecotypes are indica-

ted by an arrow.

494 Isaac Roldan et al.


were determined in both mutant strains (Figure 6). No

difference in the WSP and maltose levels were observed

between mutants and WT plants; however, the intracellular

levels of sucrose, fructose and glucose were higher in SSIV

mutant plants, and is likely to be a consequence of the lower

rate of starch synthesis in those plants, which would divert

the fixed carbon to soluble sugars.

Enzymological characterization of Atss43 mutant lines

The levels of enzyme activities involved in the synthesis and

the degradation of starch were determined in mutant lines

using both zymograms and in vitro assays (Experimental

procedures). No decrease in total soluble SS activity was

observed in vitro in both mutants using either rabbit liver

glycogen or amylopectin as primers (Table 2). Other starch

metabolizing enzymes were unaffected (Table 1; Figure S24 ),

except for the total activity of starch phosphorylase (PHS),

which was increased by 1.4–2-fold, depending on the sub-

strate used for the assay (Table 1). Zymograms analysis

showed that this induction was caused by an increase in the

activity of both the cytosolic (PHS2) and plastidial (PHS1)

isoforms of starch PHS (Figure 7a). In order to test whether

this increase could be attributable either to an interaction

between SSIV and starch PHS proteins, which was missed in

the mutant lines, or to another indirect effect, the mRNA

level of both PHS genes in Atss4-1 was determined by real-

time RT-PCR. Figure 7c shows that the expression of both

genes was increased in the mutant line with respect to the

WT ecotype. The extent of PHS mRNAs induction was

comparable with that found for PHS activity (Figure 7b,c),

suggesting that the absence of SSIV induced starch PHS

activity through a metabolic alteration that triggers the

induction of both plastidial and cytosolic PHS gene expres-

sion.

Effect of eliminating SSIV on the structure and composition

of starch

Interruption of the AtSS4 locus did not affect the total sol-

uble SS activity in vitro (Table 1). However starch synthesis

was clearly reduced in the mutants. In order to determine if

Atss4 mutations were altering the structure and composition

of starch, the amylose/amylopectin ratio and CL distribution

of amylopectin were analyzed in both mutant alleles.

Amylose and amylopectin polymers were separated using

size exclusion chromatography performed on sepharose CL-

2B columns, and subsequently quantified by the amyloglu-

cosidase assay. This analysis showed that the amylose/

amylopectin ratio was not affected in Atss4 mutants (data

not shown).

Purified amylopectin was then subjected to complete

enzymatic debranching and the CL distribution was deter-

mined by fluorophore-assisted capillary electrophoresis

(FACE) after coupling the resulting linear glucans with 8-

amino-1,3,6-pyrenetrisulfonic acid (APTS). Comparison of

CL distribution profiles of Atss4 mutant alleles and their

respective WT ecotypes indicated that the Atss4 mutation

had minor effects on the structure of amylopectin (Figure 8).

Indeed, only a slight reduction in the number of chains of

degree of polymerization (DP)5 ¼ 7–10 could be observed

(note the scale of the difference of plots in Figure 8).

Morphology and size of the starch granules were analyzed

using both scanning (SEM) and transmission (TEM) electron

microscopy. Starch granules were isolated from Atss4

mutant and WT leaves collected at midday and processed

Figure 3.25 Growth of Atss4 mutant alleles and

their respective wild-type ecotypes. Seeds of

mutant and wild-type plants were incubated in

water at 4�C for 3 days before being sowed in

soil. Plants were cultured in a growth cabinet

under a photoregime of 16-h light/8-h dark.

Pictures were taken 21 days after sowing. Panel

A: Atss4-1 mutant (left) and Columbia-0 (Col-0)

plant (right). Panel C: Atss4-2 mutant (left) and

Wassilewskija (WS) plant (right). Panels B and D:

the above-ground organs fresh weight (FW) in

mg per plant (y-axis) during the time course

experiment plotted against the number of days

after the germination of seeds (x-axis). •, wild

type plants (Col-0 and WS in panel B and D

respectively). m, mutant plants (Atss4-1 and

Atss4-2 in panel B and D, respectively).



as described in Experimental procedures. Two major alter-

ations were observed: first, a dramatic enlargement of

granule size was detected in both mutant alleles (Fig-

ure 9b,d); second, starch granules in mutants showed a less

electrodense zone in the granule core (Figure 9d), which

could indicate the presence of a cavity in the hilum (the

center of the starch granules) isolated from mutant plants. A

more detailed study was carried out by TEM (Figure 10) and

light microscopy (Figure 11) analysis on sections of leaves

collected at 4 and 12 h during the light phase. Two relevant

results arose from those analyses: the greatest difference in

size between Col-0 and Atss4-1 starch granules was

observed at the beginning of the day (after 4 h of light,

panels A and B of Figures 10 and 11). This result is in line

with the lower rate of starch degradation observed in mutant

plants (Figure 4), which is expected to yield larger starch

granules after the dark period.6 Secondly, most of the

chloroplasts in Atss4-1 mutant plants contain a single starch

granule with only a few exceptions (two granules per

chloroplast) (See panels C and D in Figures 10 and 11). This

observation comes in stark contrast with that obtained in WT

chloroplasts, where 4–5 starch granules per chloroplast

could be observed (Figure 11c). Different sections from

different leaf samples were analyzed yielding the same

results in all cases (data not shown). We therefore conclude

that a loss-of-function mutation at the AtSS4 locus affects

the number and size of starch granules synthesized in the

chloroplast.

Discussion

In this study we have analyzed the phenotypic alterations

produced by the specific loss of SSIV. The analysis of two

independent T-DNA insertion mutants obtained in two dis-

tinct genetic backgrounds, such as Col-0 and WS ecotypes,

indicates that the shared phenotypic alterations found in the

two mutants are specifically caused by the loss of the SSIV

protein. The T-DNA insertions were located in intron 2 and

11 of the AtSS4 genomic locus (At4g18240) in the two

Figure 4. Starch accumulation in leaves of wild-type and Atss4 mutant plants

during a day/night cycle. Plants were cultured under a 16-h light/8-h dark

photoperiod over 21 days and then one leaf from three plants for each line

was collected at the indicated time. The starch content in leaves was

determined by enzymatic assay as described in Experimental procedures.

Values are the average of three independent experiments (error bars, � SE).

Panel A: Columbia-0 ( ) and Atss4-1 (•). Panel B: Wassilewskija ( ) and

Atss4-2 (•). The bar at the top indicates day (gray bar) and night (black bar).

Figure 5. Complementation of Atss4 mutation. Atss4-1 mutant allele, Atss4-1

mutant expressing AtSS4 protein (KO::P35SAtSS4) and wild type Columbia-0

(Col-0) ecotype plants were cultured under a 16-h light/8-h dark photoperiod

and pictures were taken 15 days after sowing. A leaf from each plant was

collected at the end of the light period, was depigmented by treatment with

ethanol 80% and finally the starch was stained using Lugol solution.



mutant alleles studied in this work (Figure 2). These inser-

tions determined in both cases the synthesis of a modified

SSIV mRNA. However, western blot analysis indicated that

these messengers failed to produce a WT SSIV protein

(Figure 2). The antibody used in this analysis was raised

against a 178 amino acids fragment of the amino-terminal

region of the protein, upstream of the T-DNA insertion site,

so that the presence of truncated versions of SSIV protein in

the mutant alleles should be also ruled out, as no small

mutant-specific polypeptide was detected by western blot

(Figure 2).

The loss of function of all previously characterized classes

of soluble SSs have systematically lead to a clear alteration

of the amylopectin structure, thereby changing the CL

distribution. Those changes of CL distribution have allowed

the assignment of preferential contributions for each of the

three soluble SS isoform classes in the synthesis of different

chain subclasses within amylopectin7 (Craig et al., 1998;

Delvalle et al., 2005; Fontaine et al., 1993; Gao et al., 1998;

Morell et al., 2003; Zhang et al., 2004, 2005). At variance with

this behaviour, Atss4 mutants showed amylopectin CL

profiles similar to those found in the respective WT eco-

types. Only a weak, although reproducible, reduction of the

number of short chains (DP 7–10) could be observed

(Figure 8). The amylose/amylopectin ratio and levels of

WSPs remained unchanged in both mutant alleles, with

respect to values in their respective WT (data not shown and

Figure 6). Zymogram analysis did not reflect any change in

the activity levels of either SSI or SSIII isoforms (see

Figure S2), thereby ruling out the existence of compensation

mechanisms leading to the selective increase in the activity

of other SSs making up for the loss of SSIV. These results

taken together suggest that the major function of the SSIV

isoform might be different from the elongation of amylo-

pectin chains during the process of starch biogenesis.

The most visible effect of Atss4 mutations is the growth

inhibition observed in Atss4 mutant lines cultured under a

16-h day/8-h night cycle (Figure 3). Previous studies of

Arabidopsis mutants affected in starch degradation, such

as sex1 (Caspar et al., 1991) and sex4 (Zeeman et al., 1998),

have shown that an efficient mobilization of transitory starch

in a diurnal cycle is required for normal growth. Thus, the

growth alteration of Atss4 mutants may be caused by the

decreased rate of starch degradation during the night phase

(Figure 4). The finding that a normal growth phenotype is

restored when mutant plants are cultured under continuous

light corroborates this idea (Figure S1). The deficient starch

mobilization found in Atss4 lines may also be responsible

for the induction of expression of both plastidial (PHS1) and

cytosolic (PHS2) starch PHS genes (Figure 7). It was sug-

gested that these enzymes may play a role in the tolerance to

abiotic stress (Zeeman et al., 2004). Thus, the nutritional

starvation induced by a poor starch mobilization during the

night period could be the metabolic signal that triggers the

Figure 6. Sugar content in the leaves of Atss4 mutants and wild-type plants.

Leaves from 3-week-old plants cultured under a 16-h light/8-h dark photope-

riod were collected at midday and the contents of sucrose, glucose, fructose,

maltose and water soluble polysaccharides (WSP) were determined as

described in Experimental procedures. Black columns: wild-type plants; gray

columns, Atss4 mutant plants. Values represent the average of four

independent experiments. Error bars,� SE.



induction of the expression of both PHS isoforms. The same

behavior was observed in the case of double mutants of

Arabidopsis defective for both AtBE2 (locus At5g03650) and

AtBE3 (locus At2g36390) in which starch synthesis is com-

pletely abolished (Dumez et al., 2006).

The main alterations in starch metabolism in the Atss4

mutants consisted in the reduction of the starch synthesis

and degradation rates (Figure 4) that correlated with an

enlargement of starch granule size and a decrease in granule

numbers per chloroplast (Figures 9–11). The reduced rate of

starch synthesis cannot be attributable to a concomitant

degradation of starch during the light period as maltose

levels were unaffected in plant mutants (Figure 6), but is

rather attributable to a block in the synthesis of starch that

deviates the photosynthetically fixed carbon to soluble

sugars such as glucose and fructose8 (Figure 6). The

decreased rate of starch turnover along the diurnal cycle

cannot be explained by a reduction in activity of the starch

metabolizing enzymes, as no significant changes in such

activities could be detected by both in vitro and zymograms

analysis (Table 1 and Figure S2). A possible explanation for

this reduced rate of starch turnover could be the impairment

in the ability of Atss4 mutants to synthesize more than one

or two starch granules per chloroplast (Figure 11).9 In that

case, the overall surface area of starch granules accessible to

either the starch synthesis or the starch degradation

machinery would be dramatically reduced. Indeed, if we

consider that starch granules are true spheres, and if we

consider that there is only one granule in the Atss4 mutant

but four in the WT, then the available surface at the end of

the day in the WT is almost 10 times larger than in the

mutant. Hence the rates of both starch synthesis and

degradation could be reduced if we assume that both

operate at, and are limited by, the available mutant granule

surface.

The reduced number of starch granules in Atss4 mutants

would also be responsible for the size enlargement observed

in these granules (Figures 9 and 10). In this case, all the ADP-

glucose pool must be channeled to one or two granules,

leading to considerably bigger starch granules in the

mutants in comparison with WT plants.

The data reported here indicate that SSIV is required to

determine the correct number of starch granules per chlo-

roplasts, and that the area of polysaccharide surface is an

important determinant of the rates of starch synthesis and

degradation. The control of granule numbers by SSIV

immediately suggests that the latter may be involved in

the priming of starch granule synthesis. This idea is consid-

erably strengthened by the finding of abnormal hilum

structures, exemplified by the presence of cavities in the

centre of mutant starch granules (Figure 9d). We propose

that SSIV is necessary to establish an initial structure that

will nucleate the crystallization and the biogenesis of a new

starch granule. Priming of glycogen synthesis in mammals

and fungi (uridine 5’-diphosphate (UDP)10 -glucose dependent

systems) is carried out by an autoglucosylating protein,

known as glycogenin, which synthesizes an a-1,4-linked

glucan that is subsequently extended by the unique glyco-

gen-synthase (Cheng et al., 1995). Recently, it was proposed

that glycogenin-like proteins could play a role in the starch

initiation process similar to that described in the animal

glycogen synthesis (Chatterjee et al., 2005). In addition,

other elements such as isoamylases (Bustos et al., 2004)

seem to be operating in the control of the starch granule

initiation process. Those data, taken together with results

shown in this work, indicate that the process of starch

granule initiation could be more complex than that des-

cribed for mammal or fungal glycogen. Other elements,

such as the one described in this work (SSIV), are likely to be

involved. Indeed it has been recently demonstrated that

starch metabolism is a mosaic of two distinct storage

polysaccharide metabolism pathways that are likely to have

existed in the first plant cells consecutively to endosymbi-

osis11 (Coppin et al., 2005). These probably consisted of a

‘eukaryotic’ pathway active in the cytoplasm, somewhat

similar to fungal glycogen synthesis, and of a cyanobacterial

pathway similar to that documented for other bacteria such

as Escherichia coli. The priming mechanism used by plants

Table 1 In vitro assays of several starch metabolizing enzymes performed with crude extracts of leaves from AtSS4 mutant alleles (Atss4-1 andAtss4-2) and their wild-type ecotypes [Columbia-0 (Col-0) and Wassilewskija (WS), respectively]. Activities are expressed in nmol min)1 mg)1 ofproteins � SE (in each case n ¼ 3). 1Starch-phosphorylase activity was assayed in the sense of glucan degradation. Samples were harvested atmidday from 3-week-old plants

Activity Col-0 Atss4-1 WS Atss4-2

Soluble starch synthaseWith rabbit liver glycogen 7.86 � 0.38 8.50 � 0.47 6.35 � 0.15 6.36 � 0.10With amylopectin 6.31 � 0.25 6.42 � 0.13 5.54 � 0.23 5.39 � 0.18

AGPase (biosynthetic assay) 21.07 � 3.5 21.77 � 4.8 28.30 � 2.1 27.48 � 3.0Starch branching enzyme 617 � 32 582 � 26 378 � 47 447 � 38a-1,4-Glucanotransferase 0.028 � 0.002 0.029 � 0.001 0.12 � 0.002 0.11 � 0.003Starch phosphorylase1

With amylopectin 16.0 � 2 21.2 � 1.7 14.7 � 1.1 22 � 1.2With DP7 32.2 � 2.1 62.5 � 3.1 36.8 � 3.1 64.2 � 2.8



to initiate starch granules may reflect this complexity, and

therefore may be composed of both bacterial-like and

eukaryotic-like components. In line with this, a selective

function of glycogen-synthase in the priming of bacterial

glycogen synthesis has been recently suggested (Ugalde

et al., 2003). It must be stressed however that the loss of

SSIV does not eliminate the ability of chloroplasts to make

starch granules, thereby suggesting a certain degree of

redundancy in the priming function assumed by SSIV.

Further analysis of multiple mutants defective for SSIV and

other genes (either directly involved or not in starch meta-

bolism) will be necessary to ascertain if other proteins can to

some extent supply part of the specialized SSIV function.

In this respect, it is worth exploring both the ability of other

SSs to initiate polysaccharide synthesis and the possible

interaction of SSs with plant glycogenin-like proteins12 .

Experimental procedures

Arabidopsis lines, growth conditions and media

Mutants lines of Arabidopsis thaliana were obtained from the T-DNAmutant collections generated at URGV, INRA, Versailles13 (Bechtoldet al., 1993; Bouchez et al., 1993) and the GABI-KAT (Koln, Germany)14

mutant collection (Rosso et al., 2003). WT ecotypes (WS line WS-4,accession number N5390 in the Nottingham Arabidopsis StockCenter (NASC, Nottingham, UK)15 ; Col-0, accession number N1093in NASC) and mutant lines were grown in growth cabinets under a16-h light/8-h dark photoregime at 23�C (day)/20�C (night), 70%humidity and a light intensity at the plant levels of 120 lE m)2 sec)1

supplied by white fluorescent lamps. Seeds were sown in soil andirrigated with 0.5X MS medium (Murashige and Skoog, 1962).

RNA extraction and reverse transcription

Total RNA was isolated according to the method described byPrescott and Martin (1987). Prior to cDNA synthesis, and in order toremove contaminating genomic DNA, the RNA preparations wereincubated with 10 U of DNAse I FPLC16 Pure for 10 min at 37�C,extracted with phenol and chloroform, precipitated and then dis-solved in nuclease-free MilliQ-water. First-strand cDNA was syn-thesized from 10 lg of total RNA using Moloney Murine LeukemiaVirus (MMLV)17 -RT and oligo(dT)12))18 primer, according to themanufacturer’s instructions. The reaction was incubated at 37�C for2 h and stopped by adding 1 ml of nuclease-free MilliQ-water. Allthe reagents were from Amersham Biosciences (Uppsala, Sweden).

Production of polyclonal antibody against SSIV

Leaf total RNA was used to obtain cDNA as described above.Oligonucleotides SA215 (5¢-CATATGGAGACTGATGAAAGGATT-3¢)and SA216 (5¢-CTCGAGTTCTTTATAAACGTTGGC-3¢) were used toamplify a 521-bp fragment of SSIV cDNA encoding the section fromGlu236 to Glu414 of the SSIV amino acids sequence. Those oligo-nucleotides introduced restriction sites for NdeI and XhoI at the 5¢and 3¢ ends of the cDNA fragment, respectively, which were used toclone the cDNA fragment in the expression vector pGEX-4T (Amer-sham Biosciences) fused in frame to the 3¢-end of the glutathione-S-transferase (GST) gene. Construct was confirmed by DNAsequencing and transformed into E. coli BL21 (DE3) strain. Protein

Figure 7.25 (a) Zymogram of starch phosphorylase activities. Approximately

100 lg of proteins from leaf crude extract were loaded on starch-containing

polyacrylamide gel. After separation and incubation overnight at room

temperature in the presence of glucose-1-P at 20 mM, starch-phosphorylase

activities (in the sense of glucan synthesis) were revealed by iodine staining.

PHS2, cytosolic form of starch phosphorylase (At3g46970); PHS1, plastidial

form of starch phosphorylase (At3g29320).

(b) In vitro starch phosphorylase activities. Enzymatic assays were performed

in the crude extracts of leaves using amylopectin as a substrate (as described

in Experimental procedures). Activities in mutant lines are expressed as the

percentage of values obtained for their respective wild-type ecotypes, which

are considered to be 100%. Values are the average of three different

experiments (Vertical bars ¼ SE). Panel C) Expression of AtPHS1 and AtPHS2

genes in leaves of Atss4-1 and Col-0 plants. Levels of PHS1 and PHS2 mRNAs

were determined using real-time quantitative RT-PCR as described in

Experimental procedures. Data are normalized to values obtained for Col-0

plants, which are considered to be 100. Black columns: mRNA levels in Atss4-

1 plants. White columns: mRNA levels in Col-0 plants (error bars, � SE). In all

cases, 3-weeks-old plants were used and samples were harvested at midday.



expression, purification of the GST-SSIV fragment fusion proteinwith glutathione agarose and purification of the SSIV fragment bycleavage of the matrix-bound GST fusion protein with thrombinwere performed following the procedure described by Ausubel et al.(1987). Rabbit polyclonal antiserum was raised against the purifiedSSIV fragment. Finally, the immunoglobulin G fraction of antiserumwas purified by FPLC using a Protein A Sepharose column (Amer-sham Biosciences) following the manufacture’s instructions.

Figure 8. Amylopectin chain length (CL) distribution profiles for mutant alleles (Atss4-1 and Atss4-2) and their respective wild-type ecotypes [Columbia-0 (Col-0)

and Wassilewskija (WS)]. Amylopectin was purified using a CL-2B column and was subsequently debranched with a mix of isoamylase and pullulanase. The

resulting linear glucans were analyzed by fluorophore-assisted capillary electrophoresis (FACE) after coupling with a fluorescent molecule (8-amino-1,3,6-

pyrenetrisulfonic acid, APTS) to their non-reducing ends. The relative proportion for each glucan in the total population is expressed as a percentage of the total

number of chains. x-Axes represent the degree of polymerization (DP) of the chains; y-axes represent molar%. In the two bottom plots, the normalized value for each

wild type was subtracted from that of their respective mutant allele, and in these cases the y-axes represent the molar% difference. Values are the average of three

different experiments. The standard deviation was <�15% of the average values.

Figure 9. Scanning (a and b) and transmission (c and d) electron microscopy

analysis of starch from 21-day-old mutant and wild-type plants. Starch

granules were isolated by Percoll gradient as described in Experimental

procedures from leaves collected after 8 h of illumination (midday).

(a) Columbia-0 (Col-0).

(b) Atss4-1.

(c) Wassilewskija (WS).

(d) Atss4-2. Figure 10. Transmission electron microscopy analysis of sections leaves

from Atss4-1 mutant and wild-type plants. Leaves of plants cultured under a

16-h light/8-h dark photoperiod were collected at 4 and 12 h of the light phase

and subsequently fixed, embedded and sectioned as described in the

Experimental procedures.

(a) Columbia-0 (Col-0) chloroplast at 4 h.

(b) Atss4-1 chloroplast at 4 h.

(c) Col-0 chloroplasts at 12 h.

(d) Atss4-1 chloroplasts at 12 h. CW, cell wall; S, starch; V, vacuole.



Real-time RT- PCR analysis

Real-time quantitative RT-PCR assays were achieved using an iCy-cler instrument (Bio-Rad, Hercules, CA, USA). The PCR reactionmixture contained in a total volume of 25 ll, 5 ll of cDNA, 0.2 mM ofdeoxyribonucleotide triphosphate (dNTPs)18 , 2.5 mM of MgCl2, a1 : 100 000 dilution of SYBR� Green I nucleic gel stain (MolecularProbes, Eugene, OR, USA) : fluorescein calibration dye (Bio-Rad),0.3 U of Taq polymerase, 2.5 ll 10 · Taq polymerase buffer,and 0.2 lM of each primer. The specific oligonucleotides usedwere: SA198 (5¢-TGATGAGAAGAGGAATGACCCGAAA-3¢) andSA199 (5¢-CCATAGATTTTCGATAGCCGA-3¢) for AtSS1; SA126(5¢-GGAACCATTCCGGTGGTCCATGCCG-3¢) and SA127 (5¢-CTCAC-CAATGATACTTAGCAGCAACAAG-3¢) for AtSS2; SA200 (5¢-GTG-CAAGACGGTGATGGAGCAA-3¢) and SA201 (5¢-CACGTTTTTTA-TATTGCTTTGGGAA-3¢) for AtSS3; SA419 (5¢-CGTGACTTAA-GGGCTTTGGA-3¢) and SA420 (5¢-GCAGCTCGGCTAAAATACGA-3¢)for AtSS4; SA546 (5¢-TGGAAGGAAACGAAGGCTTTG-3¢) and SA547(5¢-TGTCTTTGGCGTATTCGTGGA-3¢) for AtPHS1; SA548 (5¢-ACA-GGTTTTGGACGTGGTGATT-3¢) and SA549 (5¢-ACAGGACAAGCCT-CAATGTTCCA-3¢) for AtPHS2; UBQF (5¢-GATCTTTGCCGGAAAA-CAATTGGAGGATGGT-3¢) and UBQR (5¢-CGACTTGTCATTAG-AAAGAAAGAGATAACAG-3¢) for UBQ10. Thermal cycling consistedof 94�C for 3 min; followed by 40 cycles of 10 sec at 94�C, 15 sec at61�C and 15 sec at 72�C. After that, a melting curve was generated tocheck the specificity of the amplified fragment. The efficiency of allthe primers at the above conditions was between 75% and 110% inall the tested samples, and product identity was confirmed by se-quence analysis. Arabidopsis Ubiquitin 10 (Sun and Callis, 1997)was used as a house-keeping gene in the expression analysis.Absolute quantification (Ginzinger, 2002) was performed by cloning

the amplified products in pGEM-T19 vector (Promega, Madison, WI,USA), and then using them as external calibration standards.

Complementation of the Atss420 mutation

The full-length cDNA of AtSS4 (At4g18240) was amplified by PCRusing primers 5¢-GGGGACAAGTTTGTACAAAAAAGCAGGCTT-CGAAGGAGATAGAACCATGGCGACGAAGCTATCGAGCTT-3¢ (for-ward) and 5¢-GGGGACCACTTTGTACAAGAAAGCTGGGTACGTG-CGATTAGGAACAGCTCTT-3¢ (reverse) designed to contain the attBsites for cloning using the GATEWAYTM system (Invitrogen, Carls-bad, CA, USA21 ). Amplified fragment was cloned into pGEM-T easy(Promega), mobilized to vector pDONR221 (Invitrogen) and finallycloned into the GATEWAYTM binary vector pCTAPi (Rohila et al.,2004), which fused a small polypeptide containing protein A andcalmodulin-binding protein domains at the C-terminal of the pro-tein, and allows the expression of AtSS4 under a 35S promoter. Theconstruct was introduced into Agrobacterium tumefaciens strainC58, which was used to transform Arabidopsis Atss4-1 mutantplants by the floral-dip method described by Clough and Bent (1998)Previously, Atss4-1 mutant plants were tested to be sensitive toBASTA herbicide22 . Twenty T3 progeny resulting from homozygousself-crosses were used for phenotypic characterization.

Extraction and purification of starch

For the analysis of the structure and composition of starch, Ara-bidopsis leaves were harvested at the end of the light period.Approximately 10 g of fresh material was homogenized using aTissue Tearor (Biospec Products Inc., Bartlesville, OK, USA) in 30 mlof the following buffer: 100 mM 3-(N-morpholino) propanesul-phonic acid (MOPS), pH 7.2, 5 mM EDTA and 10% (v/v) ethanediol.The homogenate was filtered through two layers of Miracloth23

(Millipore, MA, USA) and centrifuged for 15 min at 4�C and 4000 g.The pellet was resuspended in 30 ml Percoll 90% (v/v) and centri-fuged for 40 min at 4�C and 10 000 g. The starch pellet was washedsix times with distilled sterile water (10 min at 4�C and 10 000 gbetween each wash). Starch was finally stored at 4�C in 20% ethanol.For the analysis of starch content in leaves along the diurnal cycle,the method was scaled down and three leaves (approximately300 mg) from three different plants were used in each point. Materialwas frozen with liquid nitrogen, homogenized with a mortar andpestle and resuspended in 1 ml of buffer. Starch isolation was per-formed using Percoll gradient centrifugation as described above.

Extraction and determination of sugars

Leaf tissue (0.5 g aprox.) was harvested and frozen in liquid N2. Thematerial was powdered and extracted in 1.5 ml 0.7 M perchloricacid, as described in Critchley et al. (2001). Fructose, glucose andsucrose content in the buffered extract was determined by enzy-matic analysis as described by Stitt et al. (1989). Maltose levels weredetermined by enzymatic analysis as described by Shirokane et al.(2000).

Determination of starch and WSP contents and spectral

properties of the iodine–starch complex

Starch content in leaves was quantified enzymatically as describedpreviously by Lin et al. (1988). A full account of kmax (the maximalabsorbance wavelength of the iodine–polysaccharide complex)measures can be found in Delrue et al. (1992). Water-soluble glucan

Figure 11.25 Light microscopy analysis of sections of leaves from Atss4-1 and

wild-type plants. Sections of the same samples described in Figure 8 were

stained with toluidine-blue.

(a) Columbia-0 (Col-0) cells at 4 h.

(b) Atss4-1 cells at 4 h.

(c) Col-0 cells at 12 h.

(d) Atss4-1 cells at 12 h. Scale bars in A and B ¼ 25 lm; scale bars in C and

D ¼ 15 lm.



contents in leaves were determined as described in Zeeman et al.(1998).

Separation of starch polysaccharides by size exclusion

chromatography

Starch (1.5–2.0 mg) was dissolved in 500 ll of 10 mM NaOH andsubsequently applied to a Sepharose CL-2B column (inner diamater,0.5 cm; length, 65 cm), which was equilibrated and eluted with10 mM NaOH. Fractions of 300 ll were collected at a rate of onefraction per 1.5 min. Glucans in the fractions were detected by theirreaction with iodine and levels of amylopectin and amylose weredetermined by amyloglucosidase assays.

Chain length distribution of amylopectin

FACE of debranched amylopectin After purification on aSepharose CL-2B column, 500 mg of amylopectin was dialyzedagainst distilled water and subsequently lyophilized. The amylo-pectin pellet was resuspended in 1 ml of 55 mM sodium acetate,pH 3.5 buffer, and incubated overnight at 42�C with 20 U of isoa-mylase isolated from Pseudomonas amyloderamosa (HayashibaraBiochemical Laboratories, Okayama, Japan) and 1 U of pullulanasefrom Klebsiella pneumoniae (Sigma, St Louis, MO, USA). Salts weresubsequently removed by passage through an extract-clean carb-ograph column (Alltech, Deerfield, IL, USA).

Derivatization procedure. Glucans were derivatized with APTSaccording to the manufacturer’s recommendations (BeckmanCoulter, Fullerton, CA, USA). Briefly, 2 ml APTS in 15% acetic acidsolution and 2 ml of 1 M of NaBH3CN in tetrahydrofolate weremixed and the coupling reaction was allowed to proceed overnightat 37�C in the dark.

Capillary electrophoresis analysis. Separation and quantifica-tion of APTS-coupled glucans was carried out on a P/ACE System5000 (Beckman Coulter, Fullerton, CA, USA) equipped with alaser-induced fluorescence system using a 4-mW argon ion laser.The excitation wavelength was 488 nm and the fluorescenceemitted at 520 nm was recorded on the Beckman P/ACE stationsoftware system (version 1.0). Un-coated fused-silica capillaries of57 cm in length and 75 lm inner diameter were used. Runningbuffers were from Beckman Coulter. Samples were loaded intothe capillaries by electroinjection at 10 kV for 10 sec and a voltageof 30 kV was applied for 20 min at a constant temperature of25�C.

Zymograms techniques

A complete description of these techniques can be found in Delvalleet al. (2005).

In vitro assays of starch synthesis enzymes

ADP-glucose pyrophosphorylase was assayed in the synthesisdirection according to the procedure described by Crevillen et al.(2003). Starch synthase activity was assayed as described by Delv-alle et al. (2005) using either amylopectin or glycogen as primers.Branching enzymes, starch PHS and a-1,4-glucanotransferaseactivities were performed according to procedures described byZeeman et al. (1998).

Western blot analysis

Proteins were transferred from an SDS-polyacrylamide gel tonitrocellulose membrane by electroblotting in a Trans-Blot SDtransfer cell (Bio-Rad) according to the manufacturer’s instructions.Blots were probed with rabbit anti-SSIV followed by horseradishperoxidase-conjugated goat-anti-rabbit serum and detected usingECL Plus Advanced Western Blotting Reagent (Amersham Bio-sciences).

Microscopy analysis

Fully expanded leaves from plants cultured under a 16-h light/8-hdark regime were collected at the indicated times. Small pieces(2 mm2) of leaves were cut with a razor blade and immediatelyfixed in 1% paraformaldehyde and 0.5% glutaraldehyde in 0.05 M

Na-cacodylate buffer, pH 7.4, containing 25 mg of sucrose per ml(3.5 h at 4�C, under vacuum). After fixing and rinsing with the samebuffer, tissues were dehydrated in an ethanol series and progres-sively embedded in LR White resin (London Resin Co., Reading,UK). Resin was polymerized with UV light at )20�C (Fedorova et al.,1999). Alternatively, some samples were fixed in 3% glutaralde-hyde in the above buffer and embedded in Araldite Durcupan ACMas described by Lucas et al., 1998. Semi-thin (1-lm) and ultra-thin(60-nm) sections were cut with a Leika Ultracut microtome (Leika,Vienna, Austria) fitted with a diamond knife. Semi-thin sections forlight microscopy were stained with 1% (w/v) toluidine blue inaqueous 1% sodium borate for direct observation with a ZeissAxiophot photomicroscope (Zeiss, Oberkochen, Germany). Ultra-thin sections for transmission electron microscopy were contrastedwith 2% aqueous uranyl acetate and lead citrate (Reynolds, 1963).Observations were performed with a STEM LEO 910 electronmicroscope (Oberkochen, Germany) at 80 kV, equipped with aGatan Bioscan 792 digital camera (Gatan, Pleasanton, CA, USA).Different sections from at least three different leaves samples wereanalyzed.

For scanning electron microscopy analysis samples were sputter-coated with gold and viewed with a JEOL JSM-5400 microscope(JEOL, Tokyo, Japan). Transmission microscopy analysis wasperformed as described by Delvalle et al. (2005).

Acknowledgements

We are grateful to Cesar Morcillo and Fernando Pinto for technicalassistance in the microscopy analysis. This work was supported bythe Spanish Ministerio de Educacion y Ciencias (grant no. BIO2003-00431), the French Ministere delegue a la Recherche (ACI jeunes-chercheurs n�5145), the Genoplante consortium (grant n�Af2001030), the Region Nord Pas de Calais, the European Union-FEDER (ARCir projet en emergence), the Centre National de laRecherche Scientifique, and EMBO (IR was a recipient of an EMBOshort-term fellowship in the C. D’Hulst laboratory).

Supplementary Material

The following supplementary material is available for this articleonline:Figure S1. Growth of Atss4-1 mutant allele and Col-O wild typeplants. Seeds of mutant and wild type plants were incubated inwater at 4�C for 3 days before sowing in soil. Plants were cultured ingrowth cabinet under a short-day photoregime of 8 h light/16 h dark(SD, plants at the top) or continuous light (LL, plants at the bottom).



Pictures were taken 21 days after sowing. Plants at the left, Atss4-1mutants. Plants at the right, WT plants.Figure S2. (a) Zymogram of soluble starch synthases from leafextract performed under native conditions. After migration of150 lg of proteins on a glycogen-containing gel and incubationovernight at room temperature with 1 mM ADP-glucose, SS activ-ities were revealed by soaking the gel for 30 min in lugol solution (I2/KI) and washed several times with distilled water before the picturewas taken. Lanes WS and Col-0: wild-type references correspondingto ecotypes Wassilewskija and Columbia respectively. Lanes Atss4-1 and Atss4-2: SS4 mutant samples.(b) Zymogram of starch-modifying activities performed on a starch-containing polyacrylamide gel. After migration and incubationovernight at room temperature, starch-modifying enzymes wererevealed by iodine staining. In all cases, 3 weeks old plants wereused and samples were harvested at midday.This material is available as part of the online article from http://www.blackwell-synergy.com

References

Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G.,

Smith, J.A. and Struhl, K. (1987) Current Protocols in MolecularBiology. New York: Greene Publishing Associates and Willey-Interscience.

Ball, S. and Morell, M. (2003) From bacterial glycogen to starch:understanding the biogenesis of the plant starch granule. Ann.Rev. Plant Biol. 54, 207–233.

Balzergue, S., Dubreucq, B., Chauvin, S. et al. (2001) Improved PCR-walking for large-scale isolation of plant T-DNA borders. Bio-techniques., 30, 496–504.

Bechtold, N., Ellis, J. and Pelletier, G. (1993) In planta Agrobacteri-um mediated gene transfer by infiltration of adult Arabidopsisthaliana plants. C. R. Acad. Sci. Paris Life Sci, 316, 1194–1199.

Bouchez, D., Camilleri, C. and Caboche, M. (1993) A binary vectorbased on Basta resistance for in planta transformation ofArabidopsis thaliana. C. R. Acad. Sci. Paris Life Sci. 316, 1188–1193.

Bustos, R., Fahy, B., Hylton, C., Seale, R., Nebane, N.M., Edwards,

A., Martin, C. and Smith, A.M. (2004) Starch granule initiation iscontrolled by a heteromultimeric isoamylase in potato tubers.Proc. Natl Acad. Sci. USA, 101, 2215–2220.

Cao, H., Imparl-Radosevich, J., Guan, H., Keeling, P.L., James, M.G.

and Myers, A.M. (1999) Identification of the soluble starch syn-thase activities of maize endosperm. Plant Physiol. 120, 205–216.

Caspar, T., Lin, T.P., Kakefuda, G., Benbow, L., Preiss, J. and Som-

erville, C. (1991) Mutants of Arabidopsis with altered regulation ofstarch degradation. Plant Physiol. 95, 1181–1188.

Chatterjee, M., Berbezya, P., Vyasa, D., Coatesb, S. and Barsby, T.

(2005) Reduced expression of a protein homologous to glycoge-nin leads to reduction of starch content in Arabidopsis leaves.Plant Sci. 168, 501–509.

Cheng, C., Mu, J., Farkas, I., Huang, D., Goebl, M.G. and Roach, P.J.

(1995) Requirement of the self-glucosylating initiator proteinsGlg1p and Glg2p for glycogen accumulation in Saccharomycescerevisiae. Mol. Cell Biol. 15, 6632–6640.

Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method forAgrobacterium-mediated transformation of Arabidopsis thaliana.Plant J. 16, 735–743.

Coppin, A., Varre, J., Lienard, L., Dauvillee, D., Guerardel, Y., Soyer-

Gobillard, M., Buleon, A., Ball, S. and Tomavo, S. (2005) Evolutionof plant-like crystalline storage polysaccharide in the protozoanparasite Toxoplasma gondii argues for a red alga ancestry. J. Mol.Evol. 60, 257–267.

Craig, J., Lloyd, J.R., Tomlinson, K., Barber, L., Edwards, A.,

Wang, T.L., Martin, C., Hedley, C.L. and Smith, A.M. (1998)Mutations in the gene encoding starch synthase II profoundlyalter amylopectin structure in pea embryos. Plant Cell, 10, 413–426.

Crevillen, P., Ballicora, M.A., Merida, A., Preiss, J. and Romero, J.M.

(2003) The different large subunit isoforms of Arabidopsis thali-ana ADP-Glucose pyrophosphorylase confer distinct kinetic andregulatory properties to the heterotetrameric enzyme. J. Biol.Chem. 278, 28508–28515.

Critchley, J.H., Zeeman, S.C., Takaha, T., Smith, A.M. and Smith,

S.M. (2001) A critical role for disproportionating enzyme in starchbreakdown is revealed by a knock-out mutation in Arabidopsis.Plant J. 26, 89–100.

Delrue, B., Fontaine, T., Routier, F., Decq, A., Wieruszeski, J.M.,

Vandenkoornhuyse, N., Maddelein, M.L., Fournet, B. and Ball, S.

(1992) Waxy Chlamydomonas reinhardtii monocellular algalmutants defective in amylose biosynthesis and granule-boundstarch synthase activity accumulate a structurally modifiedamylopectin. J. Bacteriol. 174, 3612–3620.

Delvalle, D., Dumez, S., Wattebled, F. et al. (2005) Soluble starchsynthase I: A major determinant for the synthesis of amylopectinin Arabidopsis thaliana leaves. Plant J. 43, 398–412.

Dumez, S., Wattebled, F., Delvalle, D., Planchot, V., Ball, S.G. and

D’Hulst, C. (2006) Mutants of Arabidopsis lacking starch branch-ing enzyme II substitute plastidial starch synthesis by cytoplasmicmaltose accumulation. Plant Cell, 18, 2694–2709.

Emanuelsson, O., Nielsen, H. and von Heijne, G. (1999) ChloroP, aneural network-based method for predicting chloroplast transitpeptides and their cleavage sites. Protein Sci. 8, 978–984.

Fedorova, E., Thomson, R., Whitehead, L.F., Maudoux, O., Udvardi,

M.K. and Day, D.A. (1999) Localization of Hþ-ATPase in soybeanroot nodules. Planta, 209, 25–32.

Fontaine, T., Dhulst, C., Maddelein, M.L. et al. (1993) Toward anunderstanding of the biogenesis of the starch granule – evidencethat Chlamydomonas soluble starch synthase-Ii controls thesynthesis of intermediate size glucans of amylopectin. J. Biol.Chem. 268, 16223–16230.

Gao, M., Wanat, J., Stinard, P.S., James, M.G. and Myers, A.M.

(1998) Characterization of dull1, a maize gene encoding for anovel starch synthase. Plant Cell, 10, 399–412.

Ginzinger, D.G. (2002) Gene quantification using real-time quanti-tative PCR: An emerging technology hits the mainstream. Exper.Hematol. 30, 503–512.

Hirose, T. and Terao, T. (2004) A comprehensive expression analysisof the starch synthase gene family in rice (Oryza sativa L.). Planta,220, 9–16.

James, M.G., Denyer, K. and Myers, A.M. (2003) Starch synthesis inthe cereal endosperm. Curr. Opin. Plant Biol. 6, 215–222.

Klosgen, R.B., Gierl, A., Schwarz-Sommer, Z. and Saedler, H. (1986)Molecular analysis of the waxy locus of Zea mays. Mol. Gen.Genet. 203, 237–244.

Lin, T.P., Caspar, T., Somerville, C. and Preiss, J. (1988) Isolation andcharacterization of a starchless mutant of Arabidopsis thaliana(L.) Heynh lacking ADP-glucose pyrophosphorylase activity. PlantPlysiol. 86, 1131–1135.

Lucas, M.M., Van de Sype, G., Herouart, D., Hernandez, M.J.,

Puppo, A. and de Felipe, M.R. (1998) Immnulocalization of ferritinin determinate and indeterminate root nodules. Protoplasma,204, 61–70.

Maddelein, M.L., Libessart, N., Delrue, B., Dhulst, C.D.,

Vandenkoornhuyse, N., Decq, A. and Ball, S.G. (1993) Amylo-pectin type-I and type-II define 2 independent steps for thebuilding of the starch granule. Plant Physiol. 102, 27.



Morell, M.K., Kosar-Hashemi, B., Cmiel, M., Samuel, M.S., Chandler,

P., Rahman, S., Buleon, A., Batey, I.L. and Li, Z. (2003) Barley sex6mutants lack starch synthase IIa activity and contain a starch withnovel properties. Plant J. 34, 173–185.

Murashige, T. and Skoog, F. (1962) A revised medium for rapidgrowth and bioassays with tobacco tissue cultures. Plant Physiol.15, 473–497.

Myers, A.M., Morell, M.K., James, M.G. and Ball, S.G. (2000) Recentprogress toward understanding biosynthesis of the amylopectincrystal. Plant Physiol. 122, 989–997.

Prescott, A. and Martin, A. (1987) A rapid method for quantitativeassessment of levels of specific mRNAs in plants. Plant Mol. Biol.Rep. 4, 219–224.

Ral, J.P., Derelle, E., Ferraz, C. et al. (2004) Starch division andpartitioning. A mechanism for granule propagation and main-tenance in the picophytoplanktonic green alga Ostreococcustauri. Plant Physiol. 136, 3333–3340.

Reynolds, E.S. (1963) The use of lead citrate at high pH as an elec-tron opaque stain in electron microscopy. J. Cell Biol. 17, 208–213.

Rohila, J.S., Chen, M., Cerny, R. and Fromm, M.E. (2004) Improvedtandem affinity purification tag and methods for isolation ofprotein heterocomplexes from plants. Plant J. 38, 172–181.

Rosso, M.G., Li, Y., Strizhov, N., Reiss, B., Dekker, K. and Weisshaar,

B. (2003) Arabidopsis thaliana T-DNA mutagenized population(GABI-Kat) for flanking sequence tag-based reverse genetics.Plant Mol. Biol. 53, 247–259.

Shirokane, Y., Ichikawa, K. and Suzuki, M. (2000) A novel enzymicdetermination of maltose. Carbohydr. Res. 329, 699–702.

Stitt, M., McC. Lilley, R., Gerhardt, R. and Heldt, H.W. (1989)Determination of metabolite levels in specific cells and subcel-lular compartments of plants leaves. Methods Enzymol. 174, 518–552.

Sun, C.W. and Callis, J. (1997) Independent modulation of Arabid-opsis thaliana polyubiquitin mRNAs in different organs and inresponse to environmental changes. Plant J. 11, 1017–1027.

Ugalde, J.E., Parodi, A.J. and Ugalde, R.A. (2003) De novo synthesisof bacterial glycogen: Agrobacterium tumefaciens glycogensynthase is involved in glucan initiation and elongation. Proc.Natl Acad. Sci. USA, 100, 10659–10663.

Zeeman, S.C., Northrop, F., Smith, A.M. and Ap Rees, T. (1998)A starch-accumulating mutant of Arabidopsis thaliana deficientin a chloroplastic starch-hydrolysing enzyme. Plant J. 15, 357–385.

Zeeman, S.C., Thorneycroft, D., Schupp, N., Chapple, A., Weck, M.,

Dunstan, H., Haldimann, P., Bechtold, N., Smith, A.M. and Smith,

S.M. (2004) Plastidial a-glucan phosphorylase is not required forstarch degradation in Arabidopsis leaves but has a role in thetolerance of abiotic stress. Plant Physiol. 135, 849–858.

Zhang, X., Colleoni, C., Ratushna, V., Sirghie-colleoni, M., James,

M. and Myers, A. (2004) Molecular characterization demonstratesthat the Zea mays gene sugary2 codes for the starch synthaseisoform SSIIa. Plant Mol. Biol. 54, 865–879.

Zhang, X., Myers, A.M. and James, M.G. (2005) Mutations affectingstarch synthase III in Arabidopsis alter leaf starch structureand increase the rate of starch synthesis. Plant Physiol. 138, 663–674.



The phenotype of soluble starch synthase IV defective mutants of Arabidopsis thaliana suggests a novel function of elongation enzymes in the control of starch granule formation

Documents