Plant Biotechnology Journal (2007) 5, pp. 240–253 doi: 10.1111/j.1467-7652.2006.00235.x

240 © 2007 Blackwell Publishing Ltd

SummaryAn efficient in planta sugarcane-based production system may be realized by coupling the

synthesis of alternative products to the metabolic intermediates of sucrose metabolism, thus

taking advantage of the sucrose-producing capability of the plant. This was evaluated by

synthesizing sorbitol in sugarcane (Saccharum hybrids) using the Malus domestica sorbitol-

6-phosphate dehydrogenase gene (mds6pdh). Mature transgenic sugarcane plants were

compared with untransformed sugarcane variety Q117 by evaluation of the growth,

metabolite levels and extractable activity of relevant enzymes. The average amounts of

sorbitol detected in the most productive line were 120 mg/g dry weight (equivalent to 61%

of the soluble sugars) in the leaf lamina and 10 mg/g dry weight in the stalk pith. The levels

of enzymes involved in sucrose synthesis and cleavage were elevated in the leaves of plants

accumulating sorbitol, but this did not affect sucrose accumulation in the culm. The activity

of oxidative reactions in the pentose phosphate pathway and the non-reversible

glyceraldehyde-3-phosphate dehydrogenase reaction were elevated to replenish the

reducing power consumed by sorbitol synthesis. Sorbitol-producing sugarcane generated

30%−40% less aerial biomass and was 10%−30% shorter than control lines. Leaves

developed necrosis in a pattern characteristic of early senescence, and the severity was

related to the relative quantity of sorbitol accumulated. When the Zymomonas mobilis

glucokinase (zmglk) gene was co-expressed with mds6pdh to increase the production of

glucose-6-phosphate, the plants were again smaller, indicating that glucose-6-phosphate

deficiency was not responsible for the reduced growth. In summary, sorbitol

hyperaccumulation affected sugarcane growth and metabolism, but the outcome was not

lethal for the plant. This work also demonstrated that impressive yields of alternative products

can be generated from the intermediates of sucrose metabolism in Saccharum spp.

Blackwell Publishing, Ltd.Oxford, UKPBIPlant Biotechnology Journal1467-7644© 2007 Blackwell Publishing Ltd? 2007??Original ArticleSorbitol synthesis in sugarcaneBarrie Fong Chong et al.

Growth and metabolism in sugarcane are altered by the creation of a new hexose-phosphate sinkBarrie Fong Chong1,2,*, Graham D. Bonnett2,3, Donna Glassop2,3, Michael G. O’Shea1,2 and Stevens M. Brumbley1,2,*,†1David North Plant Research Centre, BSES Limited, PO Box 86, Indooroopilly, Qld 4068, Australia 2Cooperative Research Centre for Sugar Industry Innovation Through Biotechnology, The University of Queensland, St Lucia, Qld 4072, Australia 3CSIRO, Plant Industry, Queensland Bioscience Precinct, 306 Carmody Road, St Lucia, Qld 4067, Australia

Received 22 September 2006;

revised 12 November 2006;

accepted 17 November 2006.

*Correspondence (fax +61 7 3871 0383;

e-mail bfongchong@bses.org.au) or

(fax +61 7 3871 0383;

e-mail s.brumbley1@uq.edu.au)

†Present address: Australian Institute for

Bioengineering and Nanotechnology,

The University of Queensland, St Lucia,

Qld 4072, Australia

Keywords: glucose-6-phosphate,

Saccharum spp., sorbitol, sorbitol-6-

phosphate dehydrogenase.

Introduction

Sugarcane (Saccharum hybrids) is a very efficient biomass

producer as a result of its high photosynthetic rate and

specialized C4 pathway for assimilating atmospheric CO2. A large

proportion of the biomass is stored in the stem as sucrose.

This has led to the concept that this grass would be an

efficient in planta biofactory for the synthesis of specific industrial

chemicals (McQualter et al., 2005; Ming et al., 2006). With

the exception of high-value, low-volume speciality products

(e.g. pharmaceuticals), the viability of in planta production

for numerous chemicals depends on the generation of

substantial yields without a deleterious impact on the host.

A logical approach is to match the class of product with the

type of crop to be used as a biofactory. For example, oilseed

crops are ideally suited for the production of designer oils (Alonso

and Maroto, 2000) because they efficiently produce large

quantities of fatty acids. A similar rationale proposes that

sugarcane has a marked advantage over other plants for the

production of chemicals that can be derived from sugar

Sorbitol synthesis in sugarcane 241

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

building blocks via a few enzymatic steps. This theory has

already been demonstrated in sugar beet (Sevenier et al., 1998),

a major source of the caloric sweetener sucrose. Sugar beet

transformed with the Helianthus tuberosus fructosyltransferase

converts sucrose into low-molecular-weight fructans, a low-

caloric sweetener. The oligosaccharides accumulate to an

astonishing 40% of the tap root dry weight without

detriment to the host. Sugar manipulation in sugarcane may

be similarly productive.

This work examines sugar manipulation in sugarcane by

engineering a new carbon sink into sugarcane, specifically

the six-carbon sugar alcohol sorbitol. Sorbitol is an ideal

candidate for production in sugarcane because its synthesis

requires only a single enzyme that utilizes glucose-6-

phosphate directly. Moreover, high concentrations of this

compound are likely to be tolerated by the plant because it

is a compatible solute. Sorbitol itself has intrinsic value as a

non-caloric sweetener and is also used to manufacture

ascorbic acid and personal care products (Kirschner, 2004).

The sorbitol-6-phosphate dehydrogenase (S6PDH) gene

(mds6pdh), derived from Malus domestica (Kanayama

et al., 1992), was expressed in sugarcane to catalyse the reduced

nicotinamide adenine dinucleotide phosphate (NADPH)-

assisted reduction of glucose-6-phosphate to sorbitol-6-

phosphate (EC 1.1.1.200). This phosphorylated intermediate

is subsequently cleaved by nonspecific endogenous

phosphatases to liberate sorbitol (Tao et al., 1995).

Sorbitol has been produced in tobacco by the expression

of the mds6pdh gene. Transgenic tobacco plants produced

0.2−130 µmol sorbitol/g fresh weight. Although normal

growth was observed in transgenic tobacco plants producing

less than 3 µmol sorbitol/g fresh weight, aberrant growth

and development (lesions, stunted growth, infertility and

no root growth) was observed with increasing sorbitol con-

centrations (Sheveleva et al., 1998). The altered growth char-

acteristics were attributed to altered carbohydrate allocation

and the depletion of glucose-6-phosphate, affecting cell

wall biosynthesis. A working hypothesis in the present inquiry

presumes that non-trivial quantities of sorbitol can be

synthesized in sugarcane without agronomic penalty. This is

a reasonable expectation as glucose-6-phosphate is rapidly

turned over during sucrose synthesis. Furthermore, it is

replenished from sucrose reserves via a futile cycle (Figure 1)

involving sucrose phosphate synthase (SPS), sucrose synthase

(SuSy) and neutral invertase (INV) (Sacher et al., 1963).

Nonetheless, to protect against the possibility of substrate

depletion, transgenic sugarcane co-expressing mds6pdh and

the Zymomonas mobilis glucokinase (GLK) gene (zmglk, EC

2.7.1.2) (Barnell et al., 1990) was produced. NADPH is also

required for the production of sorbitol-6-phosphate (Figure 1).

It is not considered limiting because there are two main

sources of reducing power in the cell. The first is the two

oxidative reactions of the pentose phosphate pathway, and

the second is the non-reversible glyceraldehyde-3-phosphate

dehydrogenase (GAPDH). The operation of the former

consumes glucose-6-phosphate, whereas the latter is linked

to the photosynthetic oxidation of water. Transgenic sugarcane

plants transformed with mds6pdh only, mds6pdh and zmglk

Figure 1 Metabolic reactions in the cytosol that are involved in the synthesis and breakdown of sorbitol and sucrose in sorbitolcane. FBPase, fructose-1,6-bisphosphatase; FK, fructokinase; GAPDH, non-reversible glyceraldehyde-3-phosphate dehydrogenase; GLK, glucokinase; G6PDH, glucose-6-phosphate dehydrogenase; INV, neutral invertase; NADP, oxidized nicotinamide adenine dinucleotide phosphate; NADPH, reduced nicotinamide adenine dinucleotide phosphate; 6PGDH, 6-phosphogluconate dehydrogenase; PGI, phosphoglucoisomerase; PGM, phosphoglucomutase; SDH, sorbitol dehydrogenase; S6PDH, sorbitol-6-phosphate dehydrogenase; SorPP, sorbitol-6-phosphate phosphatase; SPP, sucrose phosphate phosphatase; SPS, sucrose phosphate synthase; SuSy, sucrose synthase; UGPase, UDP-glucose pyrophosphorylase.

242 Barrie Fong Chong et al.

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

and control plants were produced for this study. Throughout

this article, the sorbitol-producing sugarcane plants are called

‘sorbitolcane’.

The objective of this study was to assess the changes that

occur in sucrose metabolism and photoassimilate partitioning

in sugarcane leaf and culm tissue following the introduction

of a new hexose-phosphate sink. This entailed the measurement

of the plant growth characteristics, tissue-specific carbohydrate

distribution and activities of the enzymes and metabolite

levels in key pathways associated with the provision of glucose-

6-phosphate and NADPH (Figure 1). This study reveals numerous

changes in the sugarcane plant resulting from the synthesis

and accumulation of sorbitol, and also provides a new

perspective into the application of sugarcane as an in planta

biofactory.

Results

Five transgenic sugarcane lines were created for this investi-

gation. Three S-lines (S-10, S-34 and S-76) were engineered

to express mds6pdh and the neomycin phosphotransferase

(nptII) selection marker. Two GS-lines (GS-4 and GS-90) expressed

mds6pdh, zmglk and nptII. Two control plants (C-4 and C-

17), expressing only nptII, were also included in the study.

The introduced enzymes were not targeted to a particular

subcellular compartment and therefore probably localized to

the cytosol of sugarcane cells.

Growth and morphological phenotype of sorbitolcane

Necrosis was evident in the expanding leaves before the

dewlap was visible. This was most pronounced in S-76, the

plant producing the highest quantity of sorbitol (Figure 2).

Necrosis originated at the leaf apex and advanced towards

the leaf base as the leaf matured. Regions of healthy, bleached

and necrotic tissue were excised from the leaf lamina and

analysed for sorbitol content. The measurements indicated

that the sorbitol concentration (mean ± standard error) was

higher in the necrotic zone (38.5 ± 0.4 mg sorbitol/g dry weight)

than in the bleached (3.7 ± 0.3 mg sorbitol/g dry weight) or

healthy (0.9 ± 0.03 mg sorbitol/g dry weight) zones.

Plant biomass yield and size were evaluated by comparing

two control lines with the sorbitolcane lines S-10, S-34 and

S-76 (Table 1). S-10 and S-34 stalk yield, stalk length and

number of internodes were less than those of the controls. Of

the three S-lines tested, S-76 was larger (in terms of stalk

length), but still produced less stalk biomass than the control

C-4. The plants engineered with enhanced GLK activity

(GS-4 and GS-90) were shorter and yielded less stalk biomass

than the control C-4.

Photosynthate partitioning

Soluble carbohydrates accumulating in the leaf were shown

to be sorbitol, sucrose, glucose and fructose. Sorbitol was

detected in sorbitolcane lines, but was absent in the controls.

The authenticity of the sugars was confirmed by gas

Table 1 Size and biomass yield of 11-month-old sorbitolcane plants. The data presented are the mean ± standard error of determinations from six individual plants per independent line (n = 6). The data were tested by analysis of variance and the means were compared by Tukey’s honestly significant difference (HSD) test. Means with the same letter were not significantly different at P < 0.05

C-4 C-17 S-10 S-34 S-76 GS-4 GS-90

Leaf mass (g FW) 566 ± 29a 590 ± 16a 591 ± 53a 471 ± 30a 550 ± 31a 631 ± 58a 616 ± 53a

Stalk mass (g FW) 1687 ± 63a 1528 ± 138ab 917 ± 126c 788 ± 92c 1016 ± 111bc 1038 ± 136bc 1113 ± 158bc

Stalk length (cm) 301 ± 7a 269 ± 11ab 206 ± 14c 204 ± 12c 260 ± 17abc 241 ± 12bc 230 ± 15bc

No. internodes 34 ± 1a 33 ± 1a 27 ± 1b 25 ± 1b 29 ± 2ab 30 ± 1ab 30 ± 1ab

Tiller mass (g FW) 416 ± 258a 334 ± 185a 149 ± 99a 124 ± 108a 143 ± 84a 47 ± 35a 286 ± 149a

No. tillers 1.5 ± 0.7a 1.2 ± 0.7a 0.3 ± 0.2a 0.3 ± 0.2a 0.5 ± 0.2a 0.7 ± 0.2a 1 ± 0.4a

FW, fresh weight.

Figure 2 Visual appearance of a section of the last fully expanded leaf from an 8-month-old S-76 plant (bottom leaf) and a control plant (top leaf). Necrosis developed at the S-76 leaf apex and margins.

Sorbitol synthesis in sugarcane 243

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

chromatography-mass spectrometry (GC-MS) (results not

shown). Another compound, eluting immediately after

fructose, was observed in sorbitolcane samples only. Its

identity could not be found in available GC-MS libraries. Acid

hydrolysis was used to break down the compound into its

constituents. An equimolar mixture of glucose and sorbitol

was produced by the hydrolysis reaction. This suggested that

the original compound was a sorbitol glucoside. However, it

did not correspond to any commercially available sorbitol

glucosides, such as maltitol, isomaltitol or cellobiitol. Maltitol

was employed as the calibration standard in subsequent

analyses because its high-performance liquid chromatography

(HPLC) response factor was similar to the sorbitol glucoside

found in sorbitolcane.

Sugar concentrations represent the average for the entire

leaf lamina or stalk internode pith (Table 2). The line with the

most sugars was S-76, for which the combined mass of

sugars was 20% of the lamina dry weight, compared with

4% in the controls. Total sorbitol was defined as the free

sorbitol plus the component that was converted into sorbitol

glucoside (fixed sorbitol). The high total sugar content in S-

76 was chiefly attributed to the total sorbitol component,

with free sorbitol representing 61% w/w of the total soluble

sugars and fixed sorbitol contributing a further 9% w/w,

resulting in a potential total of 70% w/w. The amount of

sorbitol glucoside produced generally increased with the

amount of free sorbitol present, but its percentage contribution

to the total sorbitol content varied between the lines. The

other notable change in the leaf lamina sugar composition

was the higher glucose content in S-10, S-34 and S-76. The

presence of sorbitol was the only distinguishing feature in the

pith of the 10th internode of 11-month-old S-76 plants

(Table 2). Sorbitol glucoside was either absent or below detection

limits in the stalk. The concentrations of sucrose, glucose

and fructose were comparable with those of the equivalent

internode of the control.

Spatial distribution of sorbitol

Young leaves from 6-month-old S-34 and S-76 plants were

dissected transversely into four approximately equal length

segments, and the free sorbitol and sorbitol glucoside

contents in the lamina of each segment were analysed (Table 3).

The free sorbitol and sorbitol glucoside profiles along the

leaves were not uniform, with the highest concentrations

appearing in the apical segment and the lowest in the basal

segment. The ratio of sorbitol glucoside to free sorbitol was

lower at the leaf apex than in the middle segments. A large

difference in the total amount of free sorbitol was observed

between the apical and adjacent (pre-medial) segments.

This difference was substantially less for sorbitol glucoside. To

put this into perspective, the apical segment represented

approximately one-quarter (S-76) or less (S-34) of the total

lamina dry weight, but accrued more than half of the total

amount of free sorbitol found in the lamina.

The average sugar concentrations in different leaves and

internodes of S-76 and control plants at 11 months were

also compared. A whole-plant sugar profile was obtained by

sampling leaves and internodes down the stem. No statistically

significant difference was found between the sucrose,

glucose and fructose profiles of S-76 and the control (results

not shown). The free sorbitol concentration was highest in

Table 2 Concentration of sugars extracted from the last fully expanded leaf (lamina) of 6-month-old plants and the 10th internode (pith) from the top of 11-month-old plants. Measurements were taken from the same samples as used to measure enzyme activities in Table 5. The data presented are the mean ± standard error of determinations from six individual plants per independent line (n = 6). The data were tested by analysis of variance and the means were compared by Tukey’s honestly significant difference (HSD) test. Means with the same letter (lowercase, leaf; uppercase, stalk) were not significantly different at P < 0.05

Sugar

Concentration (mg/g dry weight)

Leaf lamina Stalk pith

C-4 C-17 S-10 S-34 S-76 GS-4 GS-90 C-4 S-76

Sucrose 31 ± 3a 33 ± 4a 35 ± 3a 34 ± 2a 33 ± 6a 32 ± 3a 26 ± 3a 531 ± 21A 512 ± 32A

Glucose 1.9 ± 0.2e 2.4 ± 0.3de 5.2 ± 0.5ab 4.2 ± 0.4abc 5.6 ± 0.7a 2.9 ± 0.3cde 3.0 ± 0.1bcd 54 ± 10A 55 ± 17A

Fructose 5.6 ± 0.3b 6.3 ± 0.3b 9.3 ± 0.3a 5.7 ± 0.2b 5.4 ± 0.3b 5.6 ± 0.1b 6.0 ± 0.1b 45 ± 8A 48 ± 13A

Sorbitol 0 ± 0d 0 ± 0d 2.6 ± 0.5c 5.9 ± 1.1bc 120 ± 16a 12 ± 3b 2.5 ± 0.5c 0 ± 0B 9.8 ± 0.7A

Sorbitol glucoside 0 ± 0c 0 ± 0c 5 ± 0.5b 8.7 ± 2.1b 34 ± 2a 9.2 ± 2.5b 2.2 ± 0.7b 0 ± 0A 0 ± 0A

Total sorbitol 0 ± 0e 0 ± 0e 5.3 ± 0.5cd (51)* 11 ± 2bc (46) 138 ± 16a (13) 17 ± 4b (29) 3.7 ± 0.8d (32) 0 ± 0B 9.8 ± 0.7A

Total sugar 38 ± 3c 41 ± 4bc 57 ± 3b 57 ± 6bc 198 ± 20a 61 ± 6b 40 ± 3bc 631 ± 12A 625 ± 16A

*Percentage of total sorbitol derived from sorbitol glucoside.

244 Barrie Fong Chong et al.

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

younger leaves and lower in older leaves (Figure 3a), with leaf

20 being the exception. Because it was presumed that

sorbitol was converted into the glucose conjugate as the

leaf matured, it was anticipated that there would be more

sorbitol glucoside in the older than in the younger leaves.

However, this was not corroborated by the data because of

the large variation between replicates. Similarly, the total

sorbitol content did not change, implying that there was no

net sorbitol accumulation in the older leaves. The sorbitol

concentration was constant along the length of the stalk and

no sorbitol glucoside was detected (Figure 3b).

Effect on sugar intermediates

The behaviour of glucose-6-phosphate and sorbitol-6-phosphate

is pivotal in this investigation because they are the substrate

and product of the enzyme S6PDH. In the leaves, sorbitol-6-

phosphate was only detected in sorbitolcane lines, as expected

(Table 4). In S-34 and S-76, the glucose-6-phosphate

concentration was significantly lower than the levels measured

in the controls, but not in the case of S-10, which, coincidently,

produced less sorbitol (Table 4). The decrease in the glucose-

1-phosphate level in S-34 and S-76 was probably a direct

Table 3 Free sorbitol and sorbitol glucoside concentration profiles along the lamina of the last (1) and second last (2) fully expanded leaves of two 6-month-old sorbitolcane plants. The leaves were divided transversely into four segments (apical, pre-medial, post-medial and basal segments). The concentrations described are the mean ± standard error of three independent samples (n = 3) taken from each segment. The data were tested by analysis of variance and the means were compared by Tukey’s honestly significant difference (HSD) test. Means with the same letter were not significantly different at P < 0.05

Line Leaf Segment

Percentage of

total lamina

dry weight

Concentration (mg/g dry weight)

Sorbitol glucoside/

sorbitol ratio

Percentage of total

amount in lamina

Sorbitol

Sorbitol

glucoside Sorbitol

Sorbitol

glucoside

S-34 1 Apical 13 104 ± 18a 42 ± 1.8a 0.41 ± 0.05d 56 31

Pre-medial 26 14 ± 0.7b 22 ± 0.2b 1.6 ± 0.07a 16 34

Post-medial 31 11 ± 0.1b 12 ± 0.2c 1.1 ± 0.02b 15 23

Basal 30 9.8 ± 0.1b 6.4 ± 0.2d 0.66 ± 0.03c 13 12

S-76 1 Apical 23 241 ± 24a 32 ± 2a 0.14 ± 0.02b 54 38

Pre-medial 29 70 ± 1b 19 ± 1b 0.27 ± 0.01a 20 28

Post-medial 28 72 ± 3b 21 ± 0.5b 0.29 ± 0.01a 19 29

Basal 20 37 ± 1c 4.9 ± 0.2c 0.13 ± 0.01b 7 5

S-76 2 Apical 25 328 ± 10a 53 ± 1a 0.16 ± 0.01c 51 32

Pre-medial 28 139 ± 5b 41 ± 0.5b 0.29 ± 0.01b 25 28

Post-medial 28 85 ± 1c 37 ± 0.6c 0.43 ± 0.01a 15 25

Basal 19 77 ± 7c 32 ± 1d 0.42 ± 0.03a 9 15

Table 4 Concentration of sugar intermediates extracted from the last fully expanded leaf (lamina) of 6-month-old plants and the 10th internode (pith) from the top of 11-month-old plants. Measurements were taken from the same samples as used to measure enzyme activities in Table 5. The presented data are the mean ± standard error of determinations from six individual plants per independent line (n = 6). The data were tested by analysis of variance and the means were compared by Tukey’s honestly significant difference (HSD) test. Means with the same letter (lowercase, leaf; uppercase, stalk) were not significantly different at P < 0.05

Concentration (nmol/g fresh weight)

Leaf lamina Stalk pith

Metabolite C-4 C-17 S-10 S-34 S-76 GS-4 GS-90 C-4 S-76

Fructose-6-P 317 ± 29a 278 ± 10a 313 ± 26a 362 ± 60a 307 ± 15a 353 ± 19a 364 ± 34a 5.7 ± 1.7A 2.0 ± 0.5A

Glucose-6-P 148 ± 11bc 144 ± 7bc 136 ± 13cd 59 ± 5e 82 ± 7de 226 ± 16a 192 ± 19ab 27 ± 3A 21 ± 1A

Glucose-1-P 88 ± 9a 84 ± 10a 34 ± 5b 47 ± 9b 20 ± 8b 45 ± 6b 32 ± 8b 94 ± 10A 69 ± 11A

Sorbitol-6-P 0 ± 0c 0 ± 0c 31 ± 3b 75 ± 6a 112 ± 14a 94 ± 9a 113 ± 17a 0 ± 0B 17 ± 5A

UDP-glucose 34 ± 3a 26 ± 2ab 27 ± 3ab 21 ± 2b 24 ± 2ab 26 ± 2ab 24 ± 3ab 5.9 ± 1.4A 5 ± 1.5A

Total phosphorylated intermediates* 553 ± 28bc 506 ± 17c 507 ± 29c 542 ± 58c 521 ± 20c 718 ± 29a 701 ± 44ab 127 ± 11A 109 ± 9A

P, phosphate; UDP, uridine 5′-diphosphate.

*Sum of fructose-6-P, glucose-6-P, glucose-1-P and sorbitol-6-P.

Sorbitol synthesis in sugarcane 245

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

result of the decrease in the glucose-6-phosphate level, as

glucose-6-phosphate, fructose-6-phosphate and glucose-1-

phosphate are interconvertible through the reactions catalysed

by phosphoglucoisomerase (PGI) and phosphoglucomutase

(PGM) (Figure 1). However, fructose-6-phosphate levels were

not affected in a similar manner. There was a significant

increase in leaf glucose-6-phosphate concentration in the

GS-lines, particularly GS-4. Despite the higher glucose-6-

phosphate, glucose-1-phosphate and fructose-6-phosphate

were not higher in GS-4.

Uridine 5′-diphosphate (UDP)-glucose pyrophosphorylase

(UGPase) catalyses a theoretically reversible reaction that, in

one direction, transforms glucose-1-phosphate into UDP-

glucose (Figure 1). In practice, pyrophosphatase-mediated

breakdown of the pyrophosphate liberated during UDP-

glucose synthesis renders the reaction irreversible. This

effectively buffers the UDP-glucose pool from small perturbations

originating in the sugar phosphate pool. Consequently, this

might explain why the leaf UDP-glucose levels remained

relatively uniform across all lines.

Although significant changes were observed in the

concentration of some sugar intermediates in the leaves, the

production of sorbitol in the stalk pith had no impact on

sugar intermediates in this tissue, apart from the appearance

of sorbitol-6-phosphate (Table 4).

Activity of key enzymes related to sorbitol and sucrose

synthesis and breakdown

The activities of the enzymes mediating the synthesis and

interconversion of phosphorylated hexoses, the synthesis and

breakdown of sucrose and the provision of reducing power

were investigated (Table 5). Two enzymes involved in the

synthesis and breakdown of sucrose (SPS and SuSy) were

elevated in S-34 and S-76. Interestingly, the SuSy level in

S-76 stalk pith was also fourfold higher than in the control.

The activities of the three enzymes that generate reducing

power [glucose-6-phosphate dehydrogenase (G6PDH), 6-

phosphogluconate dehydrogenase (6PGDH) and GAPDH] were

generally higher in the S-lines. The GLK activity in GS-90 was

especially high because of the introduced transgene, zmglk.

Intriguingly, only GAPDH was elevated in GS-4, whereas GAPDH,

G6PDH and 6PGDH were all generally higher in GS-90. Because

only a moderate quantity of sorbitol was produced in the stalk

pith, only pith enzyme activities for S-76 and C-4 were evaluated.

The enzymes UGPase, SuSy and GAPDH were significantly higher

in S-76 than in C-4 stalk tissue, whereas the remaining

enzymes exhibited no significant change.

The notion that sugarcane may possess the capability to

break down sorbitol via the enzyme sorbitol dehydrogenase

(SDH, EC 1.1.1.14) was considered. There was detectable

SDH activity in all leaves and stalk (Table 5), but only S-76 leaves

displayed significantly higher SDH activity than the controls.

S6PDH activity and mds6pdh expression in selected lines

There is evidence that some Brazilian sugarcane cultivars can

synthesize sorbitol naturally (Marino et al., 2003). Therefore,

the S6PDH-like enzyme activity detected in control plants

(Table 5) may be caused by a native S6PDH. However, the

absence of sorbitol in the control plants contradicts this

Figure 3 Sorbitol and sorbitol glucoside concentration profiles in the leaf lamina (a) and stalk pith (b) of 11-month-old S-76 plants. Leaves and internodes are numbered consecutively from the top to bottom of the plant. Leaf 1 is the last fully expanded leaf and internode 1 is the internode enclosed by it. �, sorbitol; �, sorbitol glucoside; �, total sorbitol. The data plotted are the mean ± standard error of measurements collected from six individual plants per line (n = 6). Repeated measures analysis of variance (ANOVA) was applied to the leaf and internode data. There was no significant concentration change for leaf sorbitol glucoside (P = 0.1732), leaf total sorbitol (P = 0.3189) and internode sorbitol (P = 0.6373). However, changes in leaf sorbitol concentration were significant (P = 0.0182). Leaf sorbitol concentration means assigned the same letter are not significantly different as determined by Tukey’s honestly significant difference (HSD) test (P < 0.05).

246Barrie Fong C

hong et al.

© Blackw

ell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

Table 5 Enzyme activities in leaf and stalk tissue. Measurements were taken from the last fully expanded leaf (lamina) of 6-month-old plants and the 10th internode (pith) from the top of 11-month-old plants. The data presented are the mean ± standard error of determinations from six individual plants per independent line (n = 6). The data were tested by analysis of variance and the means were compared by Tukey’s honestly significant difference (HSD) test. Means with the same letter (lowercase, leaf; uppercase, stalk) were not significantly different at P < 0.05

Enzyme

Enzyme activity (pkat/mg protein)

Leaf lamina Stalk pith

C-4 C-17 S-10 S-34 S-76 GS-4 GS-90 C-4 S-76

S6PDH 212 ± 13b 168 ± 21b 220 ± 20b 300 ± 46ab 527 ± 138a 269 ± 17ab 356 ± 84ab 631 ± 86A 677 ± 60A

SDH 23 ± 7b 16 ± 4b 29 ± 9ab 27 ± 7ab 84 ± 23a 17 ± 2b 22 ± 5ab 130 ± 16A 124 ± 19A

FBPase 289 ± 42b 447 ± 43ab 450 ± 73ab 466 ± 27ab 549 ± 28a 274 ± 21b 595 ± 64a 412 ± 51A 536 ± 44A

PGI 19223 ± 666a 19191 ± 630a 17570 ± 496a 19093 ± 553a 17367 ± 311a 18048 ± 751a 17636 ± 84a 19755 ± 980A 17547 ± 862A

GLK 140 ± 5c 185 ± 17bc 158 ± 12c 177 ± 19bc 189 ± 13bc 286 ± 47b 589 ± 95a 808 ± 92A 754 ± 63A

FK 580 ± 42ab 531 ± 35b 632 ± 62ab 578 ± 56ab 777 ± 39a 550 ± 26b 630 ± 59ab 1058 ± 115A 1232 ± 53A

PGM 29076 ± 2209bc 39257 ± 3703ab 33249 ± 2599abc 42979 ± 2873a 35895 ± 1342abc 25129 ± 1314c 37020 ± 3167ab 33865 ± 2533A 34885 ± 1846A

UGPase 3128 ± 160a 3950 ± 436a 4177 ± 228a 4521 ± 420a 3409 ± 220a 3168 ± 233a 3168 ± 233a 8646 ± 949B 44344 ± 11 940A

SPS 335 ± 25c 385 ± 16bc 488 ± 37abc 531 ± 77ab 609 ± 53a 433 ± 47abc 496 ± 26abc 885 ± 156A 1122 ± 139A

SuSy 69 ± 9bc 71 ± 7bc 102 ± 3ab 114 ± 10a 111 ± 5a 58 ± 6c 90 ± 10abc 300 ± 23B 1284 ± 200A

INV 199 ± 24ab 181 ± 29b 305 ± 35a 244 ± 24ab 256 ± 20ab 163 ± 29b 226 ± 16ab 466 ± 44A 492 ± 35A

G6PDH 532 ± 55c 1173 ± 142a 1080 ± 104ab 1454 ± 142a 1323 ± 121a 680 ± 30bc 1435 ± 116a 5958 ± 745A 6081 ± 300A

6PGDH 684 ± 32b 727 ± 14b 711 ± 31b 1013 ± 83a 1058 ± 38a 655 ± 20b 966 ± 55a 2702 ± 240A 3072 ± 174A

GAPDH 250 ± 14c 195 ± 4d 363 ± 20b 356 ± 16b 380 ± 22ab 451 ± 27a 332 ± 8b 54 ± 9B 250 ± 8A

FBPase, fructose-1,6-bisphosphatase; FK, fructokinase; GAPDH, non-reversible glyceraldehyde-3-phosphate dehydrogenase; GLK, glucokinase; G6PDH, glucose-6-phosphate dehydrogenase; INV, neutral invertase;

6PGDH, 6-phosphogluconate dehydrogenase; PGI, phosphoglucoisomerase; PGM, phosphoglucomutase; SDH, sorbitol dehydrogenase; S6PDH, sorbitol-6-phosphate dehydrogenase; SPS, sucrose phosphate synthase;

SuSy, sucrose synthase; UGPase, UDP-glucose pyrophosphorylase.

Sorbitol synthesis in sugarcane 247

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

suggestion. The S6PDH assay employed in this study was an

indirect measure of the sorbitol-synthesizing capability in

tissue extracts because it was based on the reverse reaction,

i.e. the oxidation of sorbitol-6-phosphate. Consequently, the

S6PDH-like activity detected in the controls may not be the

result of a legitimate S6PDH enzyme. Thus, a more accurate

representation of the true MdS6PDH activity in the S- and GS-

lines may be the difference between the measured values

and the value obtained for the controls.

An inconsistency between S6PDH enzyme activity and the

total amount of sorbitol produced (free sorbitol plus fixed

sorbitol) was also evident (cf. Table 2 and Table 5). After

subtracting the average background S6PDH activity

measured in the controls, the 3.1-fold increase in enzyme

activity between S-34 and S-76 corresponded to a 12.5-fold

increase in the amount of total sorbitol accrued. In contrast,

the 3.7-fold increase in enzyme activity between S-10 and S-

34 only generated a 2.1-fold increase in the amount of total

sorbitol. To investigate this result further, the mds6pdh

transcript abundance was measured using real-time

quantitative polymerase chain reaction (PCR). The sugarcane

25S ribosomal RNA was chosen as the reference gene

because its suitability as a reference gene has been dem-

onstrated previously (Iskandar et al., 2004). Although

transcript abundance increased with the amount of sorbitol

produced (Figure 4), the result did not quantitatively

explain the discrepancy between the enzyme activity and

total sorbitol. This disparity suggests that factors other than

the transcript level or extractable enzyme activity may be

influencing the accumulation of sorbitol in these plants.

Discussion

This study examined the impact of diverting hexose-

phosphate equivalents in sugarcane to an alternative carbon

sink, namely sorbitol. The quantity of sorbitol produced in

the leaf tissue of sugarcane transformed with mds6pdh was

substantial. In some leaves, total sorbitol was comparable

with the levels reported for native sorbitol producers of

the Rosaceae family (Loescher, 1987). Compared with

conventional sugarcane, sorbitolcane was distinguished by

leaf necrosis, less overall biomass and altered metabolism.

Accumulation of sorbitol in sugarcane cells affects

growth and development

The synthesis of large amounts of sorbitol in tobacco using

mds6pdh resulted in stunted and infertile plants (Sheveleva

et al., 1998). Some of the effects in tobacco were attributed

to myo-inositol deficiency caused by the depletion of its

precursor, glucose-6-phosphate. Despite the nature of the

sorbitolcane phenotype, two observations do not support

the notion that the maintenance of glucose-6-phosphate in

sorbitolcane was inadequate. Firstly, the expression of mds6pdh

was under the control of a promoter that is usually regarded

as constitutive, although it may not be constitutive in all

cell and tissue types, as found in rice (Cornejo et al., 1993).

This implies that a shortfall in glucose-6-phosphate should

lead to systemic symptoms, such as generalized necrosis

throughout the leaf. Instead, an apical to basal progression

of leaf necrosis was observed, similar in appearance to

age-mediated leaf senescence. A second observation that

does not support the theory of inadequate glucose-6-

phosphate supply is the fact that the mds6pdh/zmglk co-

transformants that produced glucose-6-phosphate at a

faster rate failed to grow as large as the control plants and

still exhibited leaf necrosis.

Factors other than substrate depletion could be responsible

for the sorbitolcane growth phenotype. The changes observed

may be associated with the sugar composition. Disruption of

the source–sink relations in the cell has been suggested as the

cause of growth inhibition and lesions in tobacco engineered

to over-express the yeast invertase cDNA (von Schaewen et al.,

1990; Sonnewald et al., 1991). Perhaps the higher glucose

concentration that was frequently observed in sorbitolcane

disrupted normal sugar sensing and eventually caused

premature leaf senescence. In Arabidopsis, it has been proven

Figure 4 Amount of Malus domestica sorbitol-6-phosphate dehydrogenase gene (mds6pdh) transcript in the leaves of three sorbitolcane lines (S-10, S-34, S-76) and a control line (C-17). No transcript was detected in control plants. The amount of mds6pdh transcript is normalized to the amount of sugarcane 25S rRNA detected in the same RNA sample. The data plotted are the mean ± standard error of measurements collected from five individual plants per line. Means assigned the same letter were not significantly different as determined by Tukey’s honestly significant difference (HSD) test (P < 0.05).

248 Barrie Fong Chong et al.

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

that glucose induces the expression of phospholipase D

(Xiao et al., 2000), which is a promoter of leaf senescence

(Fan et al., 1997). Glucose is also implicated in the dis-

mantling of the photosynthetic apparatus during senescence

(Wingler et al., 1998).

The quantity of sorbitol that accumulated at the leaf apex

may not be entirely age dependent. It is possible that sorbitol

diffused into xylem vessels and was subsequently propelled

to the leaf apex by guttation. A high osmolality resulting from

the extreme concentration of sorbitol may be problematic,

because it can lead to metabolic imbalances between

subcellular compartments, which may cause aberrant growth

(Heineke et al., 1992). The generally lower sorbitol glucoside/

sorbitol ratio at the leaf apex conflicts with the contention

that sorbitol is translocated in the guttation fluid. However,

the result is consistent in the presence of a competing process,

such as the translocation of sorbitol glucoside into vacuoles.

Many plants employ glucose as a tag to target secondary

metabolites and endogenous/exogenous toxins into vacuoles

(Bartholomew et al., 2002).

The possibility that the reduced growth observed in

sorbitolcane may be related to a reduced photosynthetic rate

was also explored by measuring the in situ CO2 assimilation

rate from a healthy region of the first fully expanded leaf on

each plant. The average CO2 assimilation rates were not

significantly different under the test conditions (results not

shown). Although the photosynthetic rates on an area basis

were similar, a diminished green leaf area caused by necrosis

may reduce the overall CO2 uptake, thus restricting stalk growth.

Osmolyte tolerance varies between plant species. The

synthesis of low to moderate levels of mannitol (Karakas

et al., 1997) or sorbitol (Sheveleva et al., 1998) in tobacco

had a limited impact, but high sorbitol levels (> 15 µmol sorbitol/

g fresh weight) led to major metabolic imbalances and severely

disrupted growth and development. Persimmon engineered

with high levels of sorbitol (49 µmol sorbitol/g fresh weight)

exhibited severe dwarfism but, surprisingly, no necrosis

(Deguchi et al., 2004). In this respect, sugarcane demonstrates

resilience compared with many other plants. Although the

reduced growth and leaf necrosis observed in sorbitol-

producing sugarcane are undesirable, it is still possible to

propagate sorbitolcane vegetatively, and re-growth also occurs

when the previous shoot is removed.

Enzymes that synthesize and cleave sucrose are more

abundant in sorbitolcane leaves

Sucrose concentration in the leaves was largely unaffected by

the sorbitol sink. Glucose-6-phosphate is an allosteric activator

of SPS in many plants (Huber and Huber, 1996). The lower

glucose-6-phosphate concentration in sorbitolcane did not

seem to impede sucrose synthesis. In addition, fructose-6-

phosphate and UDP-glucose concentrations were unaffected

by the production of sorbitol; hence, these probably did not

affect the underlying catalytic rate of SPS either.

In some sorbitolcane lines, SPS and SuSy levels were higher

in leaf tissue. The activity of these enzymes is often greater in

tissue under osmotic stress. Sweet potato grown on medium

containing a high concentration of sorbitol showed increased

SPS, SuSy and INV enzyme activities and SPS transcription

(Wang et al., 2000). The up-regulation of both sucrose

synthesis and cleavage enzymes may permit the cell to respond

more rapidly to changes in osmotic stress by adjusting the

rate of carbon storage. The sucrose futile cycle in sorbitolcane

leaves did not lead to an increase in sucrose synthesis and,

hence, the sucrose concentration in the stalk was unchanged.

The oxidative pentose phosphate pathway and the

non-reversible GAPDH are more active in sorbitolcane

The reduction of glucose-6-phosphate to sorbitol-6-phosphate

is reliant on NADPH. Two sources of reducing power were

examined in this study: the oxidative reactions of the pentose

phosphate pathway and the non-reversible GAPDH. A mate-

rial disadvantage is associated with the operation of G6PDH

because it competes with S6PDH for glucose-6-phosphate.

The non-reversible version of GAPDH is restricted to plants

(Dennis and Blakeley, 2000) and, in the leaves, it is located in

the cytosol, where it is a component of a redox shuttle that

utilizes triose phosphates to transfer reducing equivalents

from the chloroplast to the cytosol (Kelly and Gibbs, 1973).

In contrast with G6PDH, carbon is conserved in this process

by recycling the triose phosphate redox carrier between the

two compartments.

As expected, the activities of G6PDH, 6PGDH and GAPDH

were generally higher in sorbitolcane leaves that accumulated

appreciable amounts of sorbitol. The exception was GS-4,

which was principally reliant on GAPDH. The higher activity

of these enzymes underscores their importance in the supply

of NADPH for the S6PDH reaction in photosynthetic tissue.

Moreover, there is no evidence of a preferred route for obtaining

the extra NADPH.

It was envisaged that reducing power in the stalk pith

would originate chiefly via G6PDH and 6PGDH. The effectiveness

of GAPDH in the stalk would be diminished, because there is

no photosynthetic avenue to regenerate the electron donor,

glyceraldehyde-3-phosphate. The lower GAPDH activity in

the pith compared with the leaves of the control plants

Sorbitol synthesis in sugarcane 249

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

supports this notion. Surprisingly, however, G6PDH and 6PGDH

activities were not higher in S-76 stalk pith compared with

the control. Instead, the GAPDH activity was almost fivefold

higher. Therefore, it seems that the reducing power used to

synthesize sorbitol in the pith may be derived from the

GAPDH reaction operating in a mode that leads to the net

consumption of glyceraldehyde-3-phosphate.

In conclusion, a number of metabolic changes accompany

the formation of sorbitol in sugarcane. Most notable are the

increased activity of enzymes that can increase the futile

cycling of sucrose and the increased demand on the

machinery that generates reducing power. The synthesis

and accumulation of sorbitol appear to impede growth, but

the plants can still be propagated vegetatively. There is no

evidence linking the reduced growth to a deficit in glucose-

6-phosphate. Reduced growth may be caused by factors

related to sorbitol-mediated stress, the precise nature of

which requires further elucidation. Two items of circumstantial

evidence suggest that the sugarcane plant may possess

measures to lessen the effects of sorbitol hyperaccumulation.

Firstly, sorbitol may be re-assimilated via SDH, and secondly,

conjugation of sorbitol with glucose may diminish its impact.

This study has demonstrated that sugarcane metabolic

intermediates that are normally used to synthesize sucrose

can be readily diverted to an alternative sugar sink.

Experimental procedures

Cloning of mds6pdh and zmglk expression cassettes

Plasmid constructs were cloned into Escherichia coli DH10B

(Invitrogen Corp., Carlsbad, CA, USA) and subsequently grown

on Luria–Bertani (LB) medium containing 100 µg/mL of

ampicillin. Plant expression cassettes were constructed in the

plasmid pU3z-ubi-nos, a version of pAHC20 (Christensen and

Quail, 1996), which contains the constitutive maize ubi-1

promoter and nopaline synthase (nos) terminator. The

M. domestica mds6pdh (GENBANK accession number D11080)

was cloned from cDNA derived from young leaves of the

apple variety Royal Gala. The 933-bp mds6pdh cDNA, with

SpeI-KpnI restriction sites engineered at the ends, was cloned

into the vector pU3z-ubi-nos immediately downstream of the

ubi-1 promoter. The bacterial strain Z. mobilis ATCC 29191

(Australian Collection of Microorganisms, The University of

Queensland) was grown anaerobically on peptone–yeast

extract medium at pH 7.2 and 28 °C. The Z. mobilis zmglk

(GENBANK accession number M60615) was amplified by PCR

from genomic DNA isolated from the bacterial culture. The

984-bp zmglk gene, with SpeI-KpnI sites engineered at the

ends, was cloned downstream of the ubi-1 promoter in the

vector pU3z-ubi-nos. A consensus monocot ribosomal binding

site, CACC (Joshi et al., 1997), was also included immediately

upstream of the initiator codon of the cloned genes.

Sugarcane transformation

Embryogenic callus was prepared from sugarcane cultivar

Q117, as described previously (Franks and Birch, 1991). The

callus was propagated in the dark at 25 ± 1 °C on Murashige–

Skoog (MS) agar (Murashige and Skoog, 1962) supplemented

with 3 mg/L of 2,4-dichlorophenoxy acetic acid (2,4-D). The

antibiotic selection plasmid pUKN (Joyce et al., 1998), containing

the ubi-1 promoter, the nptII gene and the nos terminator, was

used to select transgenic plants. The expression cassette(s)

containing the gene(s) of interest and pUKN was co-

transformed into calli by microprojectile bombardment

(Bower et al., 1996). Following a 10-day post-transformation

recovery period in the dark, the calli were transferred to

MS−2,4-D agar that was supplemented with 150 mg/L of

paramomycin. Two weeks later, antibiotic-resistant calli

were transferred to MS agar that was supplemented with

150 mg/L of paramomycin, and incubated at 25 ± 1 °C with

a 16-h photoperiod provided by cool white fluorescent

tubes (photon flux density, 30 µmol/m2/s). Individual callus

clumps were segregated throughout the selection and

regeneration period and subcultured on to fresh agar every

2 weeks. After 8 weeks of regeneration, healthy shoots and

roots developed and the plants were transferred to a

glasshouse. 100 S-lines, 125 GS-lines and 10 controls were

regenerated from the callus under these conditions.

Plant cultivation and sampling

All sugarcane plants that were transferred to the glasshouse

were propagated individually in pots (diameter, 20 cm)

containing a soil substitute comprising equal parts of perlite

and vermiculite (Chillagoe Perlite, Mareeba, Qld, Australia).

The soil was supplemented at three-monthly intervals with

Osmocote slow-release nitrogen fertilizer granules (Scotts

Australia Pty Ltd, Baulkam Hills, NSW, Australia). Plants were

watered for 20 min at 06.00 h and 15.00 h daily at a rate of

2 L/h by drippers connected to an automatic irrigation

system (Hunter SRC Irrigation Controller, Hunter Industries

Incorporated, San Marcos, CA, USA).

Genomic integration of the desired genes was confirmed

by PCR testing for the presence of mds6pdh and zmglk in

crude DNA extracts prepared from the leaf tissue of 1-month-old

plants. Seventy per cent of the S-lines contained mds6pdh and

250 Barrie Fong Chong et al.

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

47% of the GS-lines contained both mds6pdh and zmglk.

Plants that were positive for the desired genes were sub-

sequently tested for sorbitol synthesis.

Three independent S-lines, two independent GS-lines and

two controls were grown for a further 9 months, when

vegetative cuttings were re-propagated to generate biological

replicates. The re-propagated lines were grown for 11 months

and sampled twice during this period. At 6 months, the last

fully expanded leaf from each plant was collected, immediately

frozen in liquid nitrogen and stored at −80 °C until analysis.

At 11 months, the first, fifth, 10th, 15th and 20th leaves

(leaf 1, last fully expanded leaf) and the internode enclosed by

it were harvested. The pith tissue in the internodes was

collected, frozen in liquid nitrogen and stored at −80 °C until

analysis.

Real-time quantitative PCR

Total RNA was extracted from 80 mg of sugarcane leaf

lamina using the Qiagen RNeasy Plant Kit (Qiagen Pty Ltd,

Doncaster, Vic., Australia). cDNA was generated from 100 ng

of DNase I-treated RNA using random hexamer primers and

the ImProm-II reverse transcription system (Promega Corp.,

Madison, WI, USA).

The real-time quantitative PCR probes carried a 5′ FAM

reporter and 3′ BHQ1 quencher. Primers and probes were

designed to align with target DNA derived from the M. domestica

s6pdh (GENBANK accession number D11080) (forward primer,

TTCAAGATGTCCACCGTCAC; reverse primer, AAGCTCGTC-

CTTCTCCAGAC; probe, CGGTCATCGGTCTCGGCCTT) and

reference DNA from the sugarcane 25S ribosomal RNA gene

(GENBANK accession number BQ536525) (forward primer,

CGGCTCTTCCTATCATTGTG; reverse primer, CCTGTCTCA-

CGACGGTCTAA; probe, TTCACCCACCAATAGGGAACGTGA).

Transcript levels were calculated from real-time PCR curves

using the relative standard curve method (Applied Biosystems

Sequence Detection Systems Chemistry Guide, 2003). A

reference gene and target gene PCR was set up for each sample

and comprised 125 pg cDNA, 0.9 µM of each primer, 0.2 µM

of probe and Taqman Universal PCR Master Mix (Applied

Biosystems, Foster City, CA, USA). Relative standards prepared

by serial dilution of the samples were also included in each

run. A CAS-1200 liquid handling robot (Corbett Research Pty

Ltd, Mortlake, NSW, Australia) was employed to assemble

reaction mixtures and perform serial dilutions. The PCRs were

run on a Rotagene 3000 (Corbett Research Pty Ltd) using the

universal PCR cycling conditions recommended by Applied

Biosystems (95 °C for 10 min, followed by 50 cycles of 95 °Cfor 15 s and 60 °C for 1 min).

Determination of enzyme activities

Cell-free leaf extracts were prepared from approximately 1 g

of leaf lamina; 5% w/w polyvinylpolypyrrolidone (PVPP) was

added to the tissue before homogenization in liquid nitrogen.

The homogenate was re-suspended in 2 mL of extraction

buffer comprising 50 mM N-2-hydroxyethylpiperazine-N′-2-

ethanesulphonic acid (HEPES)–KOH (pH 7.5), 10 mM MgCl2, 1 mM

ethylenediaminetetraacetic acid (EDTA), 1 mM ethylene glycol-

bis(β-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA), 5%

v/v glycerol, 0.1% v/v Triton X-100, 10 mM dithiothreitol, 1 mM

benzamidine–HCl, 1 mM benzamide, 1 mM phenylmethyl-

sulphonyl fluoride, 5 mM aminocapronate, 2 µM antipain and

2 µM leupeptin. The slurry was then centrifuged for 15 min at

3100 g and 4 °C, and the supernatant was desalted using a

3-mL Sephadex G-25-M column (Amersham Biosciences,

Uppsala, Sweden). The column was pre-equilibrated with

extraction buffer, loaded with 900 µL of supernatant and

then eluted with 1.2 mL of extraction buffer. Extracts were

prepared and maintained at 4 °C until assay. Cell-free stem

extracts were prepared analogously, except for the following

modifications. Approximately 4 g of pith tissue was used and

the resultant homogenate was re-suspended in 8 mL of

extraction buffer. The volume of supernatant recovered from

the slurry was reduced to approximately 1 mL using Amicon

Ultra-4, 10 000 NMWL centrifugal filters (Millipore, Bedford,

MA, USA). This was then desalted as described for the leaf

extracts and subsequently assayed.

Fructokinase (EC 2.7.1.4) (Renz et al., 1993), cytosolic fructose

1,6-bisphosphatase (EC 3.1.3.11) (Hurry et al., 1995), GAPDH

(EC 1.2.1.9) (Gao and Loescher, 2000), GLK (EC 2.7.1.2) (Scopes

et al., 1985), G6PDH (EC 1.1.1.49) (Wright et al., 1997), 6PGDH

(EC 1.1.1.43) (Magel et al., 2001), PGI (EC 5.3.1.9) (Burrell

et al., 1994), PGM (EC 5.4.2.2) (Manjunath et al., 1998), SDH

(EC 1.1.1.14) (Yamaguchi et al., 1994), S6PDH (EC 1.1.1.200)

(Tao et al., 1995) and UGPase (EC 2.7.7.9) (Ciereszko et al.,

2001) assays were assembled as described, with the exception

that the reaction buffer employed was Tris-HCl and the

optimum pH was determined for each reaction. The assays

were conducted at 30 °C and quantified by tracking the

change in absorbance in a spectrophotometer (Helios γ,Thermo Electron Corporation, Cambridge, Cambridgeshire,

UK). Neutral INV (EC 3.2.1.26) (Albertson et al., 2001), SPS

(EC 2.4.1.14) (Grof et al., 1998) and SuSy (EC 2.4.1.13) (Zhu

et al., 1997) assays were assembled as described. The assays

were conducted at 37 °C and quantified by measuring the

reaction products using the HPLC method outlined in the

next section. The cell-free extract protein concentration was

determined by the Bradford method (Bradford, 1976) using

Sorbitol synthesis in sugarcane 251

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

bovine serum albumin as the protein standard. Specific enzyme

activities were expressed in units of katals per milligram of

total protein.

Carbohydrate analysis

Approximately 120 mg of plant tissue was vacuum dried and

subsequently homogenized in an FP120 FastPrep cell disruptor

(Savant Instruments, Inc., Holbrook, NY, USA); 750 µL of

0.02% w/v sodium azide was added to the tissue powder.

After 3 h of incubation at 80 °C, the supernatant was collected

and the extraction step was repeated on the pellet. The

pooled extracts were filtered through a 0.2-µm syringe filter

and diluted as required with 0.02% w/v sodium azide.

The sugars in the extract were analysed by isocratic HPLC.

The instrument system consisted of an LC-10AT VP pump

(Shimadzu Corporation, Kyoto, Japan), DGU-12 A inline degasser

(Shimadzu Corporation), Sugar-Pak Guard-Pak guard column

(Waters, Milford, MA, USA), 7.8 × 300-mm Aminex HPX-87P

analytical column (Bio-Rad, Hercules, CA, USA) and RID-10 A

VP refractive index detector (Shimadzu Corporation). The

aqueous mobile phase was pumped at a flow rate of 0.8 mL/

min and the column was operated at 80 °C (CTO-10 A VP

column oven, Shimadzu Corporation); 20 µL of sample was

injected per run (SIL-10AD VP autosampler, Shimadzu Corpo-

ration). The resolved sugars were quantified by refractive

index measurement, and converted to concentration units

by comparison with calibration standards prepared from

analytical grade reagents obtained from Sigma-Aldrich

(St. Louis, MO, USA).

Measurement of sugar intermediates

Fructose-6-phosphate, glucose-6-phosphate, glucose-1-

phosphate, sorbitol-6-phosphate and UDP-glucose were

extracted as described in McLaughlin (2005), with the

following modifications. Approximately 100 mg of tissue was

homogenized in liquid nitrogen; 1 mL of chilled extraction

buffer (91% v/v methanol, 6.1 mM HEPES, pH 7) was then

added to the tissue homogenate. After incubating for 10 min

at 4 °C, the supernatant was recovered by centrifugation.

The pellet was washed four times by repeating the previous

step using 1 mL of wash buffer (75% v/v methanol, 5 mM

HEPES, pH 7) in each wash step. The sugar phosphates and

UDP-glucose were isolated from the methanol extract on a

strong anion exchange (SAX) solid phase extraction (SPE)

column (Isolute, International Sorbent Technology, Hengoed,

Mid-Glamorgan, UK). The SPE steps were carried out as

described in McLaughlin (2005).

The analytes were separated on a Dionex DX500 ion

chromatography system (Dionex Corporation, Sunnyvale,

CA, USA) consisting of a GP50 microbore pump, Dionex

AminoTrap guard column (2 × 50 mm), Dionex CarboPac

PA10 anion exchange column (2 × 250 mm) and ED40

electrochemical detector. The column was operated at 30 °C(Dionex AS50 thermal compartment). The electrochemical

detector was equipped with a gold working electrode and

Ag/AgCl reference electrode. Eluents were prepared from

helium-degassed water and carbonate-free liquid sodium

hydroxide, and stored under a nitrogen atmosphere. The

mobile phase, comprising a gradient mixture of sodium

acetate and sodium hydroxide, was pumped at a flow rate of

0.25 mL/min; 25 µL of sample was injected per run (Dionex

AS50 autosampler). The resolved analytes were quantified by

integrated amperometry, and converted to concentration

units by comparison with calibration standards prepared

from analytical grade reagents obtained from Sigma-Aldrich.

Three different gradient conditions were needed to

separate glucose-6-phosphate/fructose-6-phosphate, UDP-

glucose and glucose-1-phosphate/sorbitol-6-phosphate. The

gradient composition and profiles used to separate glucose-6-

phosphate/fructose-6-phosphate and UDP-glucose are described

in McLaughlin (2005). The eluents used to separate glucose-

1-phosphate and sorbitol-6-phosphate were as follows: A,

1 mM sodium acetate in 75 mM sodium hydroxide; B, 75 mM

sodium hydroxide. The gradient profile applied during the

run was as follows: a linear gradient from 5% A and 95% B

to 15% A and 85% B between 0 and 30 min; a concave gra-

dient to 50% A and 50% B at 35 min; a linear gradient to

100% A at 35.1 min; a linear gradient to 5% A and 95% B

from 40 to 40.1 min; finally, 5% A and 95% B maintained to

50 min.

The recovery of this method was determined by extracting

sugarcane tissue that had been spiked with analytical

standards. The recoveries obtained for glucose-6-phosphate,

glucose-1-phosphate, fructose-6-phosphate, sorbitol-6-

phosphate and UDP-glucose were 95%, 88%, 75%, 79%

and 85%, respectively.

Photosynthetic rate

The photosynthetic rate was determined by measuring the

CO2 exchange in sugarcane leaves in situ using an LI-6400

portable photosynthesis system (LI-COR Biosciences, Inc.,

Lincoln, NE, USA). Measurements were taken from a leaf area

of 6 cm2 at a photon flux of 2000 µmol/m2/s, which was set

by the LED light source in the leaf chamber. The air flow

rate through the leaf chamber was set to 400 µmol/s and

252 Barrie Fong Chong et al.

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

reference CO2 was controlled at 400 µmol CO2/mol by the

CO2 mixer.

Statistical analysis of the data

All statistical analysis was performed using the software

Statistix 8 for Windows (Analytical Software, Tallahassee, FL,

USA) or SAS/STAT (SAS Institute Inc., Cary, NC, USA).

Acknowledgements

We are grateful to Niall Masel (BSES Limited, Indooroopilly,

Qld, Australia) and Peter Abeydeera (Department of Chemical

Engineering, The University of Queensland, St Lucia, Qld,

Australia) for their assistance with the sugar and sugar inter-

mediate chromatography analysis. We are also indebted to

Sooknam Patterson who performed the tissue culture. This

work was funded through the Cooperative Research Centre

for Sugar Industry Innovation Through Biotechnology

(CRCSIIB).

References

Albertson, P.L., Peters, K.F. and Grof, C.P.L. (2001) An improvedmethod for the measurement of cell wall invertase activity in sug-arcane tissue. Aust. J. Plant Physiol. 28, 323–328.

Alonso, D.L. and Maroto, F.G. (2000) Plants as ‘chemical factories’for the production of polyunsaturated fatty acids. Biotechnol.Adv. 18, 481–497.

Barnell, W.O., Yi, K.C. and Conway, T. (1990) Sequence and geneticorganization of a Zymomonas mobilis gene cluster that encodesseveral enzymes of glucose metabolism. J. Bacteriol. 172, 7227–7240.

Bartholomew, D.M., Van Dyk, D.E., Lau, S.-M.C., O’Keefe, D.P.,Rea, P.A. and Viitanen, P.V. (2002) Alternate energy-dependentpathways for the vacuolar uptake of glucose and glutathione con-jugates. Plant Physiol. 130, 1562–1572.

Bower, R., Elliott, A.R., Potier, B.A.M. and Birch, R.G. (1996) High-efficiency, microprojectile-mediated cotransformation of sugar-cane using visible or selectable markers. Mol. Breed. 2, 239–249.

Bradford, M. (1976) A rapid and sensitive method for the quantita-tion of microgram quantities of protein utilizing the principle ofprotein-dye binding. Anal. Biochem. 72, 248–254.

Burrell, M.M., Mooney, P.J., Blundy, M., Carter, D., Wilson, F.,Green, J., Blundy, K.S. and ap Rees, T. (1994) Genetic manipulationof 6-phosphofructokinase in potato tubers. Planta, 194, 95–101.

Christensen, A.H. and Quail, P.H. (1996) Ubiquitin promoter-basedvectors for high-level expression of selectable and/or screenablemarker genes in monocotyledonous plants. Transgenic Res. 5,213–218.

Ciereszko, I., Johansson, H., Hurry, V. and Kleczkowski, L.A. (2001)Phosphate status affects the gene expression, protein content andenzymatic activity of UDP-glucose pyrophosphorylase in wild-typeand pho mutants of Arabidopsis. Planta, 212, 598–605.

Cornejo, M.-J., Luth, D., Blankenship, K.M., Anderson, O.D. andBlechl, A.E. (1993) Activity of a maize ubiquitin promoter intransgenic rice. Plant Mol. Biol. 23, 567–581.

Deguchi, M., Koshita, Y., Gao, M., Tao, R., Tetsumura, T., Yamaki,S. and Kanayama, Y. (2004) Engineered sorbitol accumulationinduces dwarfism in Japanese persimmon. J. Plant Physiol. 161,1177–1184.

Dennis, D.T. and Blakeley, S.D. (2000) Carbohydrate metabolism. InBiochemistry and Molecular Biology of Plants (Buchanan, B.B.,Gruissem, W. and Jones, R.L., eds), pp. 630–675. Rockville, MD:American Society of Plant Biologists.

Fan, L., Zheng, S. and Wang, X. (1997) Antisense suppression ofphospholipase Dα retards abscisic acid- and ethylene-promotedsenescence of postharvest Arabidopsis leaves. Plant Cell, 9, 2183–2196.

Franks, T. and Birch, R.G. (1991) Gene transfer into intact sugarcanecells using microprojectile bombardment. Aust. J. Plant Physiol.18, 471–480.

Gao, Z. and Loescher, W.H. (2000) NADPH supply and mannitolbiosynthesis. Characterization, cloning, and regulation of the non-reversible glyceraldehyde-3-phosphate dehydrogenase in celeryleaves. Plant Physiol. 124, 321–330.

Grof, C.P.L., Knight, D.P., McNeil, S.D., Lunn, J.E. and Campbell, J.A.(1998) A modified assay method shows leaf sucrose-phosphatesynthase activity is correlated with leaf sucrose content across arange of sugarcane varieties. Aust. J. Plant Physiol. 25, 499–502.

Heineke, D., Sonnewald, U., Bussis, D., Gunter, G., Leidreiter, K.,Wilke, I., Raschke, K., Willmitzer, L. and Heldt, H.W. (1992)Apoplastic expression of yeast-derived invertase in potato. Effectson photosynthesis, leaf solute composition, water relations, andtuber composition. Plant Physiol. 100, 301–308.

Huber, S.C. and Huber, J.L. (1996) Role and regulation of sucrose-phosphate synthase in higher plants. Annu. Rev. Plant Physiol.Plant Mol. Biol. 47, 431–444.

Hurry, V.M., Strand, A., Tobiaeson, M., Gardestrom, P. and Oquist,G. (1995) Cold hardening of spring and winter wheat and raperesults in differential effects on growth, carbon metabolism, andcarbohydrate content. Plant Physiol. 109, 697–706.

Iskandar, H.M., Simpson, R.S., Casu, R.E., Bonnett, G.D., Maclean,D.J. and Manners, J.M. (2004) Comparison of reference genes forquantitative real-time polymerase chain reaction analysis of geneexpression in sugarcane. Plant Mol. Rep. 22, 325–337.

Joshi, C.P., Zhou, H., Huang, X. and Chiang, V.L. (1997) Contextsequences of translation initiation codons in plants. Plant Mol.Biol. 35, 993–1001.

Joyce, P.A., McQualter, R.B., Handley, J.A., Dale, J.L., Harding, R.M.and Smith, G.R. (1998) Transgenic sugarcane resistant to sugarcanemosaic virus. Proc. Aust. Soc. Sugar Cane Technol. 20, 204–210.

Kanayama, Y., Mori, H., Imaseki, H. and Yamaki, S. (1992) Nucleo-tide sequence of a cDNA encoding NADP-sorbitol-6-phosphatedehydrogenase from apple. Plant Physiol. 100, 1607–1608.

Karakas, B., Ozias-Akins, P., Stushnoff, C., Suefferheld, M. and Rieger,M. (1997) Salinity and drought tolerance of mannitol-accumulatingtransgenic tobacco. Plant Cell Environ. 20, 609–616.

Kelly, G.J. and Gibbs, M. (1973) A mechanism for the indirecttransfer of photosynthetically reduced nicotinamide adeninedinucleotide phosphate from chloroplast to the cytoplasm. PlantPhysiol. 52, 674–676.

Kirschner, M. (2004) Chemical profile: sorbitol. Chem. Market Rep.266, 43.

Sorbitol synthesis in sugarcane 253

© Blackwell Publishing Ltd, Plant Biotechnology Journal (2007), 5, 240–253

Loescher, W.H. (1987) Physiology and metabolism of sugar alcoholsin higher plants. Physiol. Plantarum, 70, 553–557.

Magel, E.A., Hillinger, C., Wagner, T. and Holl, W. (2001) Oxidativepentose phosphate pathway and pyridine nucleotides in relation toheartwood formation in Robinia pseudoacacia L. Phytochemistry,57, 1061–1068.

Manjunath, S., Lee, C.-H.K., VanWinkle, P. and Bailey-Serres, J.(1998) Molecular and biochemical characterization of cytosolicphosphoglucomutase in maize. Expression during developmentand in response to oxygen deprivation. Plant Physiol. 117, 997–1006.

Marino, C.L., Leite, S.M.M., Farro, A.P.C., Sassaki, F.T., de Campos,H.L.V. and Coscrato, V.E. (2003) Putative metabolic pathway ofmannitol and sorbitol in sugarcane. Sci. Agric. 60, 723–728.

McLaughlin, R.L. (2005) Hyaluronic acid production in continuouscultures of Streptococcus zooepidemicus. PhD Thesis. The Univer-sity of Queensland, Brisbane.

McQualter, R.B., Fong Chong, B., Meyer, K., Van Dyk, D.E., O’Shea,M.G., Walton, N.J., Viitanen, P.V. and Brumbley, S.M. (2005)Initial evaluation of sugarcane as a production platform forp-hydroxybenzoic acid. Plant Biotechnol. J. 3, 29–41.

Ming, R., Moore, P.H., Wu, K.-K., D’Hont, A., Glaszmann, J.C., Tew,T.L., Mirkov, T.E., da Silva, J., Jifon, J., Rai, M., Schnell, R.J., Brum-bley, S.M., Lakshmanan, P., Comstock, J. and Paterson, A.H.(2006) Sugarcane improvement through breeding and biotech-nology. In Plant Breeding Reviews (Janick, J., ed.), 27, 18–118.New York: John Wiley.

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

Renz, A., Merlo, L. and Stitt, M. (1993) Partial purification frompotato tubers of three fructokinases and three hexokinases whichshow differing organ and developmental specificity. Planta, 190,156–165.

Sacher, J.A., Hatch, M.D. and Glasziou, K.T. (1963) Sugaraccumulation cycle in sugarcane. III. Physical and metabolicaspects of cycle in immature storage tissues. Plant Physiol. 38,348–354.

von Schaewen, A., Stitt, M., Schmidt, R., Sonnewald, U. andWillmitzer, L. (1990) Expression of a yeast-derived invertase in thecell wall of tobacco and Arabidopsis plants leads to accumulationof carbohydrate and inhibition of photosynthesis and strongly

influences growth and phenotype of transgenic tobacco plants.EMBO J. 9, 3033–3044.

Scopes, R.K., Testolin, V., Stoter, A., Griffiths-Smith, K. and Algar,E.M. (1985) Simultaneous purification and characterization ofglucokinase, fructokinase and glucose-6-phosphate dehydro-genase from Zymomonas mobilis. Biochem. J. 228, 627–634.

Sevenier, R., Hall, R.D., van der Meer, I.M., Hakkert, H.J.C., vanTunen, A.J. and Koops, A.J. (1998) High level fructan accumulationin a transgenic sugar beet. Nat. Biotechnol. 16, 843–846.

Sheveleva, E.V., Marquez, S., Chmara, W., Zegeer, A., Jensen, R.G.and Bohnert, H.J. (1998) Sorbitol-6-phosphate dehydrogenaseexpression in transgenic tobacco. High amounts of sorbitol lead tonecrotic lesions. Plant Physiol. 117, 831–839.

Sonnewald, U., Brauer, M., von Schaewen, A., Stitt, M. andWillmitzer, L. (1991) Transgenic tobacco plants expressing yeast-derived invertase in either the cytosol, vacuole or apoplast: apowerful tool for studying sucrose metabolism and sink/sourceinteractions. Plant J. 1, 95–106.

Tao, R., Uratsu, S.L. and Dandekar, A.M. (1995) Sorbitol synthesis intransgenic tobacco with apple cDNA encoding NADP-dependentsorbitol-6-phosphate dehydrogenase. Plant Cell Physiol. 36, 525–532.

Wang, H.-L., Lee, P.-D., Chen, W.-L., Huang, D.-J. and Su, J.-C.(2000) Osmotic stress-induced changes of sucrose metabolism incultured sweet potato cells. J. Exp. Bot. 51, 1991–1999.

Wingler, A., von Schaewen, A., Leegood, R.C., Lea, P.J. and Quick,W.P. (1998) Regulation of leaf senescence by cytokinin, sugars,and light. Plant Physiol. 116, 329–335.

Wright, D.P., Huppe, H.C. and Turpin, D.H. (1997) In vivo and invitro studies of glucose-6-phosphate dehydrogenase from barleyroot plastids in relation to reductant supply for assimilation.Plant Physiol. 114, 1413–1419.

Xiao, W., Sheen, J. and Jang, J.-C. (2000) The role of hexokinase inplant sugar signal transduction and growth and development.Plant Mol. Biol. 44, 451–461.

Yamaguchi, H., Kanayama, Y. and Yamaki, S. (1994) Purificationand properties of NAD-dependent sorbitol dehydrogenase fromapple fruit. Plant Cell Physiol. 35, 887–892.

Zhu, Y.J., Komor, E. and Moore, P.H. (1997) Sucrose accumulationin the sugarcane stem is regulated by the difference between theactivities of soluble acid invertase and sucrose phosphate syn-thase. Plant Physiol. 115, 609–616.

NO2−