picrender1.pdf

Starch Synthesis in Potato Tubers Is Regulated by

Post-Translational Redox Modification of ADP-Glucose

Pyrophosphorylase: A Novel Regulatory Mechanism

Linking Starch Synthesis to the Sucrose Supply

Axel Tiessen, Janneke H. M. Hendriks, Mark Stitt, Anja Branscheid, Yves Gibon, Eva M. Farré,

and Peter Geigenberger

Max Planck Institute of Molecular Plant Physiology, Am Mühlenberg 1, 14476 Golm, Germany

Transcriptional and allosteric regulation of ADP-Glc pyrophosphorylase (AGPase) plays a major role in the regulation of

starch synthesis. Analysis of the response after detachment of growing potato tubers from the mother plant revealed

that this concept requires extension. Starch synthesis was inhibited within 24 h of tuber detachment, even though the

catalytic subunit of AGPase (AGPB) and overall AGPase activity remained high, the substrates ATP and Glc-1-P in-

creased, and the glycerate-3-phosphate/inorganic orthophosphate (the allosteric activator and inhibitor, respectively)

ratio increased. This inhibition was abolished in transformants in which a bacterial AGPase replaced the potato AGPase.

Measurements of the subcellular levels of each metabolite between Suc and starch established AGPase as the only

step whose substrates increase and mass action ratio decreases after detachment of wild-type tubers. Separation of

extracts on nonreducing SDS gels revealed that AGPB is present as a mixture of monomers and dimers in growing tu-

bers and becomes dimerized completely in detached tubers. Dimerization led to inactivation of the enzyme as a result

of a marked decrease of the substrate affinity and sensitivity to allosteric effectors. Dimerization could be reversed and

AGPase reactivated in vitro by incubating extracts with DTT. Incubation of tuber slices with DTT or high Suc levels re-

duced dimerization, increased AGPase activation, and stimulated starch synthesis in vivo. In intact tubers, the Suc con-

tent correlated strongly with AGPase activation across a range of treatments, including tuber detachment, aging of the

mother plant, heterologous overexpression of Suc phosphorylase, and antisense inhibition of endogenous AGPase ac-

tivity. Furthermore, activation of AGPase resulted in a stimulation of starch synthesis and decreased levels of glycolytic

intermediates.

INTRODUCTION

ADP-Glc pyrophosphorylase (AGPase) catalyzes the con-

version of Glc-1-P and ATP to ADP-Glc and inorganic pyro-

phosphate (PPi), which is the first committed step in the

pathway of starch synthesis (Figure 1) (Preiss, 1988; Martin

and Smith, 1995; Smith et al., 1997). The higher plant en-

zyme is a heterotetramer, consisting of two “regulatory”

subunits (AGPS; 51 kD) and two slightly smaller “catalytic”

subunits (AGPB; 50 kD) (Okita et al., 1990). AGPase plays a

major role in the regulation of starch synthesis. Studies with

an Arabidopsis

50% of the wild-type level

(Müller-Röber et al., 1992; Geigenberger et al., 1999a).

Two mechanisms are known to regulate AGPase activity.

First, AGPase is subject to transcriptional regulation, with ex-

pression being increased by sugars (Salanoubat and Belliard,

1989; Müller-Röber et al., 1990; Sokolov et al., 1998) and

decreased by nitrate (Scheible et al., 1997) and phosphate

(Nielsen et al., 1998). This may allow starch accumulation to

respond to changes in the carbon and nutritional status

(Scheible et al., 1997; Stitt and Krapp, 1999). Second, AGPase

is exquisitely sensitive to allosteric regulation, being acti-

vated by glycerate-3-phosphate (3PGA) and inhibited by Pi

(Sowokinos, 1981; Sowokinos and Preiss, 1982; Preiss, 1988).

Increasing levels of phosphorylated intermediates typi-

cally lead to a marked increase of the 3PGA/Pi ratio. There-

fore, activation of AGPase by an increasing 3PGA/Pi ratio

allows the rate of starch synthesis to be adjusted in re-

sponse to changes in the balance between photosynthesis

and Suc synthesis in leaves (Heldt et al., 1977; Herold,

mutant have demonstrated that AGPase

is a key site for the control of starch synthesis in leaves

(Neuhaus and Stitt, 1990). In potato tubers expressing an

antisense

AGPS

AGPB

construct, starch synthesis is decreased

To whom correspondence should be addressed. E-mail geigenberger

@mpimp-golm.mpg.de; fax 49-331-5678408.

Article, publication date, and citation information can be found at

www.plantcell.org/cgi/doi/10.1105/tpc.003640.

when activity is decreased to

2192

The Plant Cell

1980; Stitt et al., 1987) and to changes in the balance be-

tween Suc breakdown and respiration in nonphotosyn-

thetic tissues (Stark et al., 1992; Hajirezaei et al., 1994;

Geigenberger et al., 1997, 1998a, 2000; Jenner et al., 2001).

Several situations have been reported in which starch

synthesis changes independently of overall AGPase activity

and reciprocally to the levels of phosphorylated intermediates

(Geigenberger et al., 1994; Geiger et al., 1998; Trethewey et

al., 1998, 2001; Geigenberger and Stitt, 2000). This fact indi-

cates that there may be important gaps in our understand-

ing of the regulation of starch synthesis. One such situation

formed the starting point for the experiments described

here. Detachment of growing potato tubers from the mother

plant leads within 24 h to a threefold decrease of ADP-Glc

and a 50% inhibition of starch synthesis, even though

AGPase activity remains unaltered and the overall levels of

hexose phosphates and 3PGA increase (Geigenberger et

al., 1994).

Because only the overall metabolite levels were mea-

sured, there are various explanations for these observa-

tions. One is that AGPase is being inhibited by a novel

mechanism. Another is that the substrate supply for AGP-

ase may be changing as a result of regulation of the enve-

lope hexose phosphate:phosphate transporter (Kammerer

et al., 1998), the plastidic phosphoglucomutase (PGM)

(Tauberger et al., 2000), or the envelope adenylate translo-

cator (Tjaden et al., 1998; Geigenberger et al., 2001). These

are required to transport the carbon substrates and ATP

from the cytosol to the plastid, where AGPase is located

(Figure 1). Another possibility is that AGPase activity may be

responding to changes in plastid 3PGA/Pi ratio caused by

changes in the activity of the envelope triose phos-

phate:phosphate translocator. This transporter is required

to transmit changes of the 3PGA/Pi ratio from one compart-

ment to the other (Borchert et al., 1989; Schott et al., 1995)

(Figure 1).

Here, we show that the inhibition of starch synthesis after

tuber detachment occurs via a mechanism that depends on

the properties of the plant AGPase; additionally, we used

nonaqueous fractionation to investigate the response of cy-

tosolic and plastidic metabolite levels and define a unique

crossover point at AGPase. We also demonstrate that de-

tachment does not lead to a decrease of the plastid 3PGA/

Pi ratio and has no effect on AGPB protein or overall AGP-

ase activity; instead, it leads to post-translational inactiva-

tion of AGPase via a reversible mechanism that involves re-

dox-dependent dimerization of the ABPB subunits. Finally,

we present evidence that this novel mechanism contributes

to the regulation of starch synthesis in response to a range

of treatments that modify the Suc level in tubers.

RESULTS

Inhibition of Starch Synthesis in Response to Tuber

Detachment Is Abolished in Transgenic Tubers That

Express a Heterologous AGPase

We used two independent approaches to determine the

step or steps in the pathway of starch synthesis that are the

targets for the novel mechanism that inhibits starch synthe-

sis when tubers are detached from the mother plant. The

first approach asked whether specific regulatory features of

potato tuber AGPase are essential for the inhibition of starch

synthesis. To answer this question, the response was com-

pared in wild-type tubers and in transformants in which

endogenous AGPase was replaced largely by a nonplant

AGPase.

The introduced construct encodes a form of the mono-

meric

Figure 1. Pathway of Suc-to-Starch Conversion and Its Subcellular

Compartmentation in Potato Tubers.

1, Suc synthase; 2, UDP - Glc pyrophosphorylase; 3, fructokinase; 4,

cytosolic phosphoglucomutase; 5, phosphoglucoisomerase; 6, plas-

tidic phosphoglucomutase; 7, ADP-Glc pyrophosphorylase; 8, alka-

line pyrophosphatase; 9, granule-bound starch synthase; 10, solu-

ble starch synthase; 11, branching enzyme; 12, hexose phosphate

translocator; 13, triose phosphate translocator; 14, adenylate trans-

locator. TCA, tricarboxylic acid.

90% and remained dependent on 3PGA. AGPase activity

in the double-transformed lines AF1-28 and AF1-20 was

threefold higher than that in AGP93 and was independent of

3PGA (Figure 2A) and Pi (data not shown) (Lloyd et al.,

1999).

enzyme (glgC16) that is different in its

kinetic properties from the plant enzyme (Stark et al., 1992).

It was introduced (Lloyd et al., 1999) into the AGPase anti-

sense line AGP93, which has low activity of the higher plant

AGPase (Müller-Röber et al., 1992). AGPase activity in wild-

type tubers was strongly dependent on 3PGA (Figure 2A).

AGPase activity in the antisense line AGP93 was reduced by

Escherichia coli

Redox Regulation of Starch Synthesis

2193

C-Glc were injected into a fine borehole in growing

tubers that were attached to the mother plant or 1 and 3

days after detaching them from the mother plant by sever-

ing the stolon (for details, see Geigenberger et al., 1994). Af-

ter 1 h, the area around the injected label was removed and

analyzed to determine how much of the injected label had

been converted to starch (Figure 2B). Metabolite levels were

measured in the same material (Figures 2C and 2D). Similar

results were obtained in an independent experiment in

which

C-Glc was supplied for 30 min to discs cut from at-

tached or detached tubers (data not shown).

In wild-type tubers, 34% of the label was incorporated

into starch in attached tubers, decreasing to 21 and 15% in

tubers that had been detached for 1 and 3 days, respec-

tively (Figure 2B). The inhibition of starch synthesis was ac-

companied by an increase of hexose phosphates (Figure

2C) and a slight increase of 3PGA (data not shown). Suc de-

creased gradually, declining by 24 and 70% after 1 and 3

days, respectively (Figure 2D). This finding confirms the re-

sults described previously (Geigenberger et al., 1994).

In the antisense line AGP93, 19% of the injected label was

incorporated into starch in attached tubers, decreasing to

8% in detached tubers (Figure 2B). Hexose phosphates

were high in attached tubers and remained high after de-

tachment (Figure 2C), and Suc decreased even more slowly

than in wild-type tubers (Figure 2D).

In the double transformants AF1-20 and AF1-28, 24 to

26% of the label was incorporated into starch in attached

tubers (Figure 2B). Detachment did not lead to a rapid inhi-

bition of starch synthesis in these lines (Figure 2B) but in-

stead led to a marked decline of hexose phosphates (Figure

2C) and a rapid 70 to 80% decrease of Suc during the first

day after detachment (Figure 2D). There were no substantial

changes in the specific activities of the internal hexose phos-

phate pools in the various genotypes after detachment (data

not shown), demonstrating that the different response in the

Changes in the rate of starch synthesis and metabolite lev-

els after tuber detachment. Tubers from 8-week-old wild-type po-

tato plants, AGPase antisense plants (AGP93), and supertrans-

formed plants expressing

(D)

(AF1-28 and AF1-20) were

analyzed either directly (black bars) or 1 day (gray bars) or 3 days

(white bars) after detachment from the plant. To measure the rate of

starch synthesis

glgC16

Inhibition of Starch Synthesis in Response to Tuber

Detachment Is Abolished in Transgenic Tubers Expressing the

40 to 50

kBq per tuber) was injected into a fine borehole of an otherwise in-

tact tuber. The tubers then were incubated for 1 h, and a concentric

core of material around the borehole was extracted and analyzed to

determine

(B)

, U-

C-Glc of high specific activity (

glgC16

Gene in an AGPB Antisense Background.

AGPase activity in wild-type tubers (WT), the parental antisense

AGPB line (AGP93), and four independent transgenic AF1 lines ex-

pressing the

C incorporation into starch. The data are expressed as a

percentage of the total label injected. The same samples were used

to measure the levels of hexose phosphates (sum of Glc-6-P, Fru-

6-P, and Glc-1-P)

gene in an antisense AGPase (AGP93) back-

ground. AGPase activity was assayed using a standard protocol

(Müller-Röber et al., 1992) in the absence (black bars) or the pres-

ence of 1 mM (dark-gray bars), 2 mM (medium-gray bars), or 3 mM

(light-gray bars) 3PGA.

glgC16

(C)

and Suc

) by enzymatic analysis. Results are

of four tubers from different plants.

FW, fresh weight.

To measure the rate of starch synthesis, low concentra-

tions of

(B)

Figure 2.

(A)

means

2194

The Plant Cell

double transformants is not attributable to isotopic dilution

of the incoming label by internal pools. These results pro-

vide strong genetic evidence that starch synthesis is inhib-

ited after detachment by a regulatory mechanism that re-

quires the presence of native potato tuber AGPase.

cellular concentrations. This analysis yields estimates for

the vacuole, plastids, and cytosol. The estimated values

for the cytosol include the mitochondrial metabolites be-

cause the cytosol and mitochondria are not separated (see

above).

Separation of Tuber Material into

Subcellular Compartments

Changes in Subcellular Levels of Its Substrates Reveal

That AGPase Is Involved in the Inhibition of

Starch Synthesis

Regulated enzymes can be identified by perturbing the flux

through a metabolic pathway and measuring the resulting

changes in metabolite levels to identify the step or steps at

which the substrate concentration(s) changes reciprocally to

the flux through the pathway (Rolleston, 1972). To allow the

unbiased identification of the step(s) at which starch synthe-

sis is inhibited after the detachment of wild-type tubers, we

investigated changes in the subcellular levels of every me-

tabolite from Suc to starch.

This was performed by nonaqueous fractionation, a tech-

nique developed to analyze the subcellular compartmenta-

tion of metabolites in leaves (Gerhardt and Heldt, 1984; Stitt

et al., 1989) and adapted recently for use with potato tubers

(Farré et al., 2001). Tubers are frozen in liquid nitrogen to

quench metabolism, homogenized in liquid nitrogen, ly-

ophilized at low temperature, and resuspended in heptane.

Enzymic reactions are blocked for the reminder of the frac-

tionation procedure because water is absent. During lyoph-

ilization, metabolites and proteins from a particular region

of the cell aggregate. The suspension then is ultrasoni-

cated to generate particles that are partially enriched for

different material from subcellular compartments and that

can be separated by nonaqueous density gradient centri-

fugation.

Marker enzyme activities are measured to reveal how ma-

terials from different cellular compartments distribute across

the gradient (Figure 3A). The vacuolar marker mannosidase

(Boller and Kende, 1979) was highly enriched in the pellet,

the cytosolic markers UDP-Glc pyrophosphorylase (UGP-

ase; Kleczkowsky, 1994) and pyrophosphate:Fru-6-P1 phos-

photransferase (MacDonald and Preiss, 1986) and the mito-

chondrial marker citrate synthase were enriched in fractions

0 and 1, and the plastidic marker AGPase (Kim et al., 1989)

was enriched in the lightest fractions (fractions 2 and 3) (Fig-

ure 3A) (Farré et al., 2001). This effect resembles the distri-

bution in gradients of leaf material (Stitt et al., 1989); in that

earlier study, it also was shown that the distribution of en-

zymes tracks the distribution of metabolites that are known

to be restricted to a particular compartment.

Metabolites are measured in each fraction, and their

subcellular distribution are estimated by more-dimensional

linear regression. The estimated distribution can be com-

pared with the overall content to estimate the metabolite

content in each compartment (per gram of total dry weight)

and with empirically determined values for the volume of

each compartment (Farré et al., 2001) to estimate the sub-

M (Table 1).

ATP is imported from the cytosol via the envelope adenyl-

ate exchanger (Figure 1). A substantial proportion of the ad-

enine nucleotides are located in the plastid (Figures 3E and

3F), whereas uridine and guanidine nucleotides are located

mainly in the cytosol of tubers (data not shown) (Farré et al.,

2001), as is found for leaves (Dancer et al., 1990; Riens et

al., 1991). Detachment led to a 50% increase of ATP in the

plastid and a 50% decrease of ATP in the cytosol (Figure

3E). ADP levels were not affected substantially after detach-

ment (Figure 3F).

The ATP/ADP ratio was lower in the plastid than in the

cytosol in attached tubers (1.2 compared with 3.3; data

calculated from Figures 3E and 3F), again resembling the

ratio in leaves (Stitt et al., 1982). The value attributed to the

cytosol underestimates the actual cytosolic value, because

significant amounts of adenine nucleotides also are pres-

ent in the mitochondria, and the ATP/ADP ratio in the mito-

chondria is lower than that in the cytosol (Stitt et al., 1982).

After detachment, the ATP/ADP ratio in the plastid in-

creased to a value (1.9) similar to that estimated for the cy-

tosol plus the mitochondria (2.0). The estimated plastid

ATP concentration increased after detachment from 179 to

279

M (Table 1).

The estimated plastidic concentrations of Glc-1-P and

ATP were in the range of the values determined in vitro for

the substrate affinity S

0.5

Glc-1-P (40 to 140

M) and S

0.5

M) of AGPase (Sowokinos and Preiss,

1982; Ballicora et al., 1995). Detachment led to an increase

of the levels of both substrates, whereas the rate of starch

The immediate substrates for AGPase are the pools of Glc-

1-P and ATP in the plastid (Figure 1). Suc is degraded in the

cytosol via Suc synthase (SuSy), UGPase, and PGM to form

Glc-6-P, which is imported into the plastid via the Glc-6-P/

Pi transporter and converted back to Glc-1-P by the plas-

tidic PGM (Figure 1). Glc-6-P, Fru-6-P, and Glc-1-P were

distributed between the plastid and cytosol in attached tu-

bers (Figures 3B to 3D) (Farré et al., 2001). The vacuole con-

tained negligible hexose phosphates, except for traces of

Glc-1-P (Figure 3D) (Farré et al., 2001). Detachment led to

an increase of Glc-6-P, Fru-6-P, and Glc-1-P in the cytosol

and the plastid. The error bars are large for Glc-1-P because

this metabolite is present at low levels, leading to analytic

errors that sum during the calculations. The estimated aver-

age plastid Glc-1-P concentration increased fourfold from

17 to 64

ATP (120 to 190

Redox Regulation of Starch Synthesis

2195

synthesis decreased. These results show that the inhibition of

starch synthesis involves a mechanism that acts on AGPase.

Estimation of Mass-Action Ratios for Each

Enzyme-Catalyzed Reaction and Transport Step

between Suc and Starch Identifies AGPase as the

Unique Site Involved in the Inhibition of Suc-to-Starch

Interconversion after Tuber Detachment

The cytosolic concentrations of Suc, Fru, UDP-Glc, Glc-1-P,

Glc-6-P, Fru-6-P, ATP, ADP, UTP, UDP, PPi, Pi, and 3PGA,

as well as the plastidic concentrations of Glc-1-P, Glc-6-P,

ADP-Glc, ATP, ADP, PPi, Pi, and 3PGA, are summarized in

Table 1. The variation for some metabolites, including Glc-

1-P and PPi in the plastid and ADP in the cytosol, was high.

The variation for Glc-1-P was discussed above. The varia-

tion in PPi was caused by the low level of PPi and by the

fact that only a very small fraction of the total PPi was lo-

cated in the plastid.

These results (Table 1) were used to calculate the ratio

between the in vivo concentrations of the products and the

substrates (termed the mass-action ratio) for every step be-

tween Suc and ADP-Glc (Table 2). The theoretical equilib-

rium constant (K

for the transport steps is set at unity. In attached tubers, the

mass-action ratios of the reactions catalyzed by SuSy,

UGPase, and cytosolic PGM, the transport exchanges cata-

lyzed by the Glc-6-P/Pi and ATP/ADP transporters, and the

reaction catalyzed by plastidic PGM are close to their K

The mass-action ratios of fructokinase, AGPase, and inor-

ganic pyrophosphatase are displaced from their K

(Table

2). These results resemble those reported previously for cy-

tosol and plastids in leaves (Stitt et al., 1982, 1989) and for

phloem sap (Geigenberger et al., 1993). Thus, our results

provide evidence for the reliability of the fractionation tech-

nique and subsequent calculations.

The inhibition of starch synthesis after detachment was

accompanied by a 40-fold decrease of the mass-action ra-

tio for AGPase (Table 2). This was the result of an increase in

the concentration of the substrates Glc-1-P and ATP and a

decrease in the concentration of the products ADP-Glc and

PPi in the plastid (Table 1) (Geigenberger et al., 1994). The

mass-action ratios for all of the other steps were unaltered

Subcellular Analysis of Metabolite Levels in Attached and

1-Day-Detached Wild-Type Tubers.

Marker enzyme distribution in a typical nonaqueous gradient of

lyophilized wild-type tuber tissues. The values represent the enzyme

activity found in each fraction as a percentage of the total applied to

the gradient. The marker enzyme distribution in the fractions of each

gradient was used to calculate the distribution of metabolites in the

subcellular compartments. CS, citrate synthase; NAF, nonaqueous

fractionation; PFP, pyrophosphate:Fru-6-P1 phosphotransferase.

Metabolites in the fractions of the nonaqueous gradients were ex-

tracted with trichloroacetic acid before analysis of Glc-6-P

(B)

, Fru-

The subcellular compartmentation of the metabolites was calculated

by more-dimensional regression analysis (best-fit method). The val-

ues are standardized to tissue dry weight and give the mean of three

separate gradients (means

(C)

, Glc-1-P

(D)

, ATP

(E)

, ADP

(F)

, 3PGA

(G)

, Pi

(H)

, and Suc

(I)

Compartmentation of metabolites in potato tuber attached

to the plant (gray bars) or after 1 day of detachment (black bars).

(I)

3).

; the ratio of product and substrate con-

centrations at which the reaction is at its thermodynamic

equilibrium and net flux is zero) is listed for comparison. K

Figure 3.

(A)

6-P

(B)

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: