Systems Biology Approaches to Cancer Energy Metabolism - Systems Biology of Metabolic and Signaling Networks - page 216

To explore the effect of changing isoforms on the glycolytic flux and metabolite
concentrations, the kinetic model of HeLa cells previously reported (Fig. 9.1 ;
Mar ´ n-Hern ´ ndez et al. 2011 ) was used. The enzymes selected were those whose
genes can be upregulated by HIF-1 under hypoxic conditions (Mar ´ n-Hern ´ ndez

et al. 2009 ). Among them, the catalytically more efficient isoforms of the glucose
transporter (GLUT3; K m Glu out ¼

α

:

52mM) and hexokinase (HKI K m Glu in ¼

:

03mM),
together with the phosphofructokinase-1 isoform exhibiting the higher affinity for
the main activator F2,6BP (PFK-1-L; Ka F2,6BP ¼

0 0

M) (Rodr´guez-Enr´quez
et al. 2009 ; Moreno-S´nchez et al. 2012 ), were manipulated. The in silico results
were compared with those of the initial model built for the normoxic condition
(Mar´n-Hern´ndez et al. 2011 ) in which GLUT1 ( K m Glu out ¼

0.53 μ

9 : 3mM), HKII

( K m Glu in ¼ M) are preferentially expressed

(Rodr ´ guez-Enr ´ quez et al. 2009 ; Moreno-S ´ nchez et al. 2012 ) (Fig. 9.2 ).
First, only the K m values for each enzyme were individually modified. The most
important changes were attained when GLUT3 values were included, consisting of
a 39 % increased glycolytic flux and 1.5-3 times increased metabolite
concentrations (Table 9.1 ). When HK affinity was modified (replacing HKII by
HKI), the flux increased only by 7 %, whereas the metabolites did not significantly
change, except for intracellular glucose (Glu in ) and G6P, which as expected
decreased and increased by 63 and 9 %, respectively. When the PFK-1 kinetic
parameters were modified (by replacing PFK-1-C for PFK-1-L), no significant
variation in the flux was observed, in agreement with the low control, determined
by elasticity analysis, that this enzyme has on tumor glycolytic flux (Mar´n-
Hern´ndez et al. 2006 ). However, increases in the Glu in (16 %), G6P (21 %), and
F6P (44 %) concentrations were observed, indicating that PFK-1 indeed exerts
homeostatic control of glycolysis.
Regarding the flux control coefficients (Table 9.2 ), significant changes were
attained when GLUT3 or HKII replaced GLUT1 or HKII, respectively. In turn,
changing PFK1-C by PFK1-L induced negligible variation in the control, i.e., C PFK1
values. The decrease in the flux control of glycogen degradation and increase in that
of HK, elicited by GLUT3 “expression” (Table 9.2 ), both occur because the cells
become more sensitive to changes in the Glu out concentration due to the higher
affinity of GLUT3. Remarkably, the main flux-controlling steps induced by chang-
ing to GLUT3 K m Glu out or HKI K m Glu in did not vary (glycogen degradation, GLUT,
and HK), indicating the robust function of the glycolytic pathway under these
conditions. These modeling results also indicated overcapacity of the remaining
glycolytic enzymes, brought about by their over-expression, thus exerting negligi-
ble flux control (Table 9.2 ).
However, tumor cells exposed to hypoxia not only induce one isoform of one
specific glycolytic enzyme/transporter; instead, they simultaneously express differ-
ent isoforms of several enzymes and transporters. Therefore, the affinities were
changed in the following two combinations (GLUT3 + HKI or GLUT3 + HKI +
PFK-L). In both cases, the flux and metabolite concentrations increased by 65 %

0 : 1 mM), and PFK-1-C ( Ka F2,6BP ¼ 1 μ