A kinetic study of sugarcane sucrose synthase
Wolfgang E. Scha¨ fer
1
, Johann M. Rohwer
2
and Frederik C. Botha
3
1
Institute for Plant Biotechnology and
2
Department of Biochemistry, University of Stellenbosch, South Africa;
3
South African
Sugarcane Research Institute, Mount Edgecombe, South Africa
The kinetic data on sugarcane (Saccharum spp. hybrids)
sucrose s ynthase ( SuSy, UDP-glucose:
D
-fructose 2-a-
D
-
glucosyltransferase, EC 2.4.1 .13) are limited. W e c haracter-
ized kinetically a SuSy activity partially purified from
sugarcane variety N19 leaf roll t issue. Primary p lot analysis
and product i nhibition studies showed that a compulsory
order ternary complex mechanism is followed, with UDP
binding first and UDP-glucose dissociating last from the
enzyme. Product inhibition studies showed that UDP-glu-
cose is a competitive inh ibitor w ith respect to UDP and a
mixed inhibitor with r espect to sucrose. Fructose is a mixed
inhibitor with r egard to both s ucrose and UDP. Kinetic
constants a re as follows: K

Keywords: metabolic control analysis; sugarcane; sucrose
synthase; kinetic modelling.
The kinetic parameters of enzymes provide important
information about their interactions with substrates, prod-
ucts and effectors. Typically, substrate K
m
values are
interpreted to g ive an indication of the affinity of enzymes
for their substrates, and conclusions about enzymes’ phy-
siological roles are often based on these values. However,
the kinetic parameters of individual enzymes do not by
themselves provide much insight into the b ehaviour of an
intact, functioning metabolic pathway. Cellular network
models, such as those applied in the approach of compu-
tational systems biology, extend the usefulness of k inetic
data on individual enzymes immensely and can have both
explanatory and predictive value.
Several papers that give a n overview of different approa-
ches for studying and modelling metabolism, such as
metabolic flux analysis, metabolic control analysis ( MCA)
and positional isotopic labelling combined with NMR or
MS, have been published recently [1–3]. Of these approaches,
MCA [4,5] is particularly useful in studies of metabolic
pathways, as it quantifies the d egree of c ontrol of individu al
reaction steps o n the steady-state pathway flux or metabolite
concentrations. Hence, MCA can be a great help in
determining potential target steps for metabolic engineering,
because the reactions in the pathway that have the most
potential of modifying a target flux or metabolite c oncen-
tration can be identified. For example, MCA has been used

Africa. Fax: +27 21 8083835, Tel.: +27 21 8083834,
E-mail:
Abbreviations: MCA, metabolic control analysis; SuS y, s ucrose
synthase.
Enzyme: sucrose sy nthase (EC 2.4.1.13).
(Received 10 June 2004, revised 7 J uly 2004, accepted 13 July 2004)
Eur. J. Biochem. 271, 3971–3977 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04288.x
to calculate the control coefficients of enzymes in t he sucrose
synthesis pathway for sucrose futile cycling (cleavage and
resynthesis of sucrose), with a view to determining which
reactions control this energetically wasteful process. Like
any kinetic model, it requires the rate equations of all
reactions in the pathway and therefore the kinetic param-
eters of every enzyme. Typically the rate equations require
more information than simply K
m
values for the substrates,
which are the only kinetic parameters reported in most
studies not focusing exclusively on kinetics. For sugarcane
SuSy (SuSy, UDP-glucose:
D
-fructose 2-a-
D
-glucosyltrans-
ferase, E C 2 .4.1.13), s ubstrate K
m
values have been reported
[12], but not other important parameters, such as substrate
K
i

M
EDTA and Roche Complete
TM
pro-
tease inhibitor. The homogenate was filtered through a
double-layered nylon cloth, centrifuged at 10 000 g for
10 min, and the pellets discarded. The proteins in the
supernatant were precipitated by 80% saturation with
ammonium sulfate and recovered by centrifugation at
10 000 g for 10 m in. The pellets were resuspended in
100 m
M
Tris/HCl (pH 7.5) buffer c ontaining 2 m
M
MgCl
2
,
2m
M
dithiothreitol and 2 m
M
EDTA (buffer A). The
protein extract was then desalted by passage through a
Pharmacia PD-10 (Sephadex G25) column and the eluant
was diluted two times with buffer A. The desalted extract
wasappliedtoa5mLAmersham/PharmaciaHi-trapQ
anion exchange column that had previously been equili-
brated with buffer A. The protein was eluted with a linear
KCl gradient at a flow speed of 1 mLÆmin
)1

M
NADH, 1 m
M
phosphoenolpyru-
vate
2
, and appropriate concentrations of UDP-glucose and
fructose. Pyruvate kinase and lactate dehydrogenase were
each added to a final activity of 4 UÆmL
)1
.NADH
oxidation was monitored at 340 nm wavelength.
Activity in the sucrose breakdown direction was rou-
tinely measured in an assay containing 100 m
M
Tris/HCl
(pH 7.0), 2 m
M
MgCl
2
,2m
M
NAD
+
,1m
M
pyrophos-
phate and appropriate concentrations of sucrose and UDP.
UDP-glucose pyrophosphorylase ( UDPGlcPP), phospho-
glucomutase (PGM) and Leuconostoc glucose-6-phosphate

GRAFIT
TM
version 4 for
Windows
TM
( Initial estimates
were calculated automatically by the program based on
linear r egression of r earranged data. U niform weighting was
used for all data points.
Kinetic p arameters o ther than the substrate K
m
values
were taken as the median values calculated from the
experimental data. To calculate the product inhibition
constants, kinetic experiments were performed at the
product inhibitor and substrate concentrations as indicated
in Figs 2 and 3.
The program
WINSCAMP
v1.2 [15] was used for kinetic
modelling, using a published model of sucrose accumula-
tion [11]. This model can be viewed and interrogated at
.
Results
The purpose of the kinetic experiments reported in this
paper was to establish the reaction mechanism of
sugarcane SuSy and also determine kinetic parameters
needed for metabolic modelling. As far as the SuSy
reaction mechanism is concerned, there are conflicting
reports in the literature; some of these results do not

m
and V
max
values that were derived from nonlinear fit (n ¼ 6) to the Michaelis–Menten equation as described in
Materials and methods. Kinetic assays were performed as describ ed in Materials and methods. s, Sub strate concentration; s/v, substrate con-
centration divided by reaction rate.
Table 1. Inhibition types and kinetic parameters for SuSyC. P arameters were determ ined as described i n Materials and methods.; w.r.t., with
respect to.
6
Kinetic parameter
type
Substrate
Sucrose
(m
M
)
UDP
(m
M
)
UDP-glucose
(m
M
)
Fructose
(m
M
)
K
i

Inhibition types and inhibition constants derived from
Dixon and Cornish–Bowden plots for UDP-glucose
(Fig. 2 ) and fructose product inhibition (Fig. 3) are shown
in Table 1 . Competitive inhibition is characterized by a
series of parallel lines in the C ornish–Bowden p lot, while the
Dixon plot shows t he lines intersecting to the left of the
y-axis. Mixed inhibition shows the lines intersecting to
theleftofthey-axis i n both plots. The inhibition patterns
indicate an ordered m echanism with UDP binding firs t and
UDP-glucose dissociating last. Product inhibition patterns
for both fructose a nd UDP-glucose agreed fully with the
predicted patterns for an ordered ternary complex mechan-
ism [16], with UDP-glucose a competitive inhibitor with
regard to UDP and a mixed inhibitor with regard to
sucrose. Fructose was a mixed inhibitor with r egard to both
UDP and sucrose. Although only three data points were
obtained for each concentration of the variable substrate,
the inhibition patterns for both UDP-Glc and fructose are
nonetheless clear.
The ordered ternary complex mechanism, with UDP
binding first and UDP-glucose dissociating last, agrees with
that proposed for Helianthus tuberosus SuSy [17] and
validates the assumption made in a kinetic model of sucrose
accumulation [11], although the substrate K
i
values obtained
experimentally differ substantially from those used in the
model. The data obtained from the kinetic experiments were
then incorporated in the model of sucrose accumulation, to
investigate the effect of changes in SuSy kinetic parameters

M
.1/v,Reciprocal
reaction rate; i, inhibitor concentration; s/v, substrate concentration divided by reaction rate.
3974 W. E. Scha
¨
fer et al.(Eur. J. Biochem. 271) Ó FEBS 2004
levels. The model ÔbehavesÕ like a sugarcane storage
parenchyma cell, in that it accumulates s ucrose, with other
metabolite levels fairly close to experimentally measured
values.
Variable outputs from the model are shown in Fig. 4.
Outputs from the original model a re shown a s the first bar
in every panel. For all the other model variants, the
equilibrium constant for the SuSy reaction was changed to
0.50 (the published model used an equilibrium constant of
five in the sucrose breakdown direction [18], but this is
incorrect; reported values range from 0.15 to 0.56 [19]).
Also, the SuSy parameters which were input in the origin al
model did not obey t he two Haldane relationships, which
relate the K
eq
to the V
f
/V
r
ratio, K
m
and K
i
values [16]. The

ÁK
mA
Þð2Þ
whereAisUDP;B,sucrose;P,fructose;Q,UDP-glucose;
and V
f
and V
r
refer to maximal reaction rates in the s ucrose
breakdown and synthesis directions, respectively.
For the corrected model (Fig. 4, model variant 2) all
kinetic p arameters were kept the same as the values used in
the published m odel, except the K
i
value for UDP ( K
iA
)was
Fig. 3. Fructose product inhibition. Dixon(A,C)andCornish–Bowdenplots(B,D)withsucrose (A,B) and UDP (C,D) a s t he variable s ubstrates.
For (A) and (B), U DP was kept constant at 0.020 m
M
, while f or ( C) and (D) sucrose was ke pt c onstant at 40 m
M
.1/v,Reciprocalreactionrate;i,
inhibitor concentration; s/v, substrate concentration divided by reaction rate.
Fig. 4.
WINSCAMP
kinetic model variable o utputs. Mo del v ariants are as follows: or., original pub lished m ode l; c orr., m odel with K
eq
and K
i

as shown in Table 1. Note that the modified K
i
values f or
fructose and sucrose are both in the same range as the
experimentally determined values, while the values for
UDP-glucose and UDP are extremely close to the experi-
mentally determined values.
The output variables differed appreciably b etween mod-
els c ontaining two d ifferent SuSy isoforms. Sucrose, g lucose,
Fru-6P and UDP-glucose concentrations were all h igher in
model variant C than in 2. Fructose was the variable most
affected by changes in the SuSy isoform in the model or
changes in SuSy activity (see Discussion), although s ucrose
concentration a lso increased by about 41% in m odel variant
C. Sucrose content was positively correlated with SuSy
activity, but these changes were qu ite small compared with
the c hanges in enzyme activity, at about a 4% increase and
9% d ecrease in sucrose for a doubling and halving of
activity, respectively. Sucrose futile cycling was about 7%
higher in the models containing the SuSyC isoform,
compared with the model (variant 2) w ith the ÔgenericÕ
SuSy. Notably, percentage conversion of hexoses t o sucrose
increased from 84.4 to 87.0%, and percentage carbon to
glycolysis decreased from 1 5.6 t o 13.0% in model variant C,
compared with 2.
Discussion
It is interesting to compare the results obtained in this
study with those for maize [20] and Helianthus t uberosus
SuSy [17]. UDP-glucose is a competitive inhibitor with
regard to UDP, and fructose a competitive inhibitor with

programmes.
Changes in SuSy activity also impacted the model
variables. The biggest changes were in fructose c oncentra-
tion, which decreased by 42% when a ctivity was doubled,
and increased by 140% when activity was halved. Incor-
poration of the SuSyC isoform in the model dramatically
reduced the steady-state concentration of fructose com-
pared w ith t he model with estimated SuSy parameters, from
22.6 to 3.04 m
M
. This may seem alarming when compared
with experimentally reported values of about 30 m
M
for
fructose in internode five [22], but it has to be kept in mind
that these experimental values assume equal distribution of
fructose between the cytosol and vacuole. Up to 99% of
glucose a nd fructose in this tissue might actually be present
in the vacuole [23], and hence the low value for cytosolic
fructose obtained with the modified model is not necessarily
incorrect. On the other hand, on e would expect the glucose
and fructose values to be more or less equal, but this is not
so in the modified model. Only metabolite measurement
methods that can distinguish between the cytosolic and
vacuolar compartments can resolve this issue.
Next, the model was expanded so that i n a ddition to the
SuSy isoform w ith generic kinetic parameters, it included a
second SuSy isoform, with experimentally determined
kinetic parameters. Total SuSy b reakdown a ctivity was
kept the same as i n the models with only one SuSy isoform.

¨
fer et al.(Eur. J. Biochem. 271) Ó FEBS 2004
results indicate that, at least in a fairly young internode,
sucrose futile cycling is not greatly affected by specific SuSy
isoforms. This may not be the case in a mature internode;
therefore mature tissue should also be m odelled in order to
answer this question.
In conclusion, kinetic modelling can be use d not only to
predict the effects of variation in the activity or kinetic
parameters of enzymes catalysing different reactions, but
can also yield information a bout the metabolic effects of
the presence of more than one isoenzyme, such as SuSy
isoforms in sugarcane. This makes possible much more
informed decisions on manipulation strategies for yield
improvement in any system that can be m odelled this way.
Obtaining the reaction mechanisms and kinetic parameters
of all enzymes involved in such a system is an essential step
in this approach.
Acknowledgements
Support from the South African Sugar Association and the South
African National Research Foundation is gratefully acknowledged.
References
1. Giersch, C. (2000) Mathematical modelling of metabolism. Curr.
Opin. Plant Biol. 3, 249–253.
2. Wiechert, W . (2001) Modeli ng and s imulation: tools f or metabolic
engineering. J. Biotechnol. 94, 37–63.
3. Morgan, J.A. & Rhodes, D. (2002) Mathematical modeling of
plant metabolic pathways. Metab. Eng. 4, 80–89.
4. Kacser, H. & Burns, J.A. (1973) The control of flux. Symp. Soc.
Exp. Biol. 27, 64–105.

14. Scha
¨
fer, W.E., R ohwer, J.M. & Botha, F.C. (2004) Partial puri-
fication and characterization of the s ucrose synthase in sugarcane.
J. Plant Phys. doi: 10.1016/j.jplph.2004.04.010.
15. Sauro, H.M. (1993)
SCAMP
:ageneralpurposesimulatorand
metabolic control analysis p rogram. CABIOS 9, 441–450.
16. Segel, I.H. (1975) Enzyme Kinetics – Behaviour and Analysis of
Rapid Equilibrium and Steady-State Enzyme Systems,1stedn.
John Wiley & Sons,
5
New York, USA.
17. Wolosiuk, R.A. & Pontis, H.G. (1974) Studies on sucrose syn-
thase. Arch. Biochem. Biophys. 165, 140–145.
18. Kruger, N.J. (1990) Carbohydrate synthesis and degradation. In
Plant Physiology, Biochemistry and Molecular Biology (Dennis,
D.T. & Turpin, D.H., eds), pp. 59–76. Longman Scientific &
Technical publishers, Harlow, UK.
19. Geigenberger,P.&Stitt,M.(1993) Sucrose synthase catalyses a
readily reversible reaction in vivo in developin g potato tubers an d
other plant tissues. Planta 189, 329–339.
20. Nguyen-Quoc, B., Krivitzky, M., Huber, S.C. & Lecharny, A.
(1990) Sucrose synthase in devel oping maize l eaves. Plant Physiol.
94, 516–523.
21. Delmer, D.P. (1972) The purification and properties of sucrose
synthase from etiolated Pha seolus aureus seedlings. J . Biol. C hem.
247, 3822–3828.
22. Whittaker, A. & Botha, F.C. (1997) C arbon partitioning during