Slide 1 - Science and Mathematics Academy

breakfastcorrieBiotechnology

Feb 22, 2013 (4 years and 3 months ago)

150 views

Gal6P

Ga
l

T6P

Gal1P

T16BP

GlP

Extracellular Gal

F6P

F16BP

DHAP

G3P

BPG

3PG

2PG

PEP

Pyr

Computational modeling of
Clostridium acetobutylicum

Jonathan
Smeton

Mentored by Dr. Margaret Hurley

Since the advent of high
-
throughput genomic sequencing, the integration
of genomic, proteomic,
metabalomic
, and
transcriptomic

data in the
computational analysis of microorganisms has become increasingly
common (de Lorenzo, 2008). As of March 2010, thirty
-
five metabolic
models of various bacteria were available to the public in online databases
with exponentially more models becoming obtainable each year (
Orth
,
Theile
, &
Palsson
, 2010). Creation of these models is becoming increasingly
easy as new modeling techniques, software, and technologies are becoming
available to construct and analyze metabolic networks: steady
-
state,
stoichiometric
, time course, metabolic control, and flux
-
balance (Hoops et
al., 2006). Due to these recent advancements and increasing interest in
eliminating national dependence on foreign oil, the
butanol
-
producing
Clostridium acetobutylicum

has been examined thoroughly (Desai & Harris,
1999). The pathway that accomplishes this feat is dependent upon the flux
of
pyruvate
, a product of the transport and breakdown of multiple carbon
sources. Understanding the utilization of differential carbon sources in
C
.
acetobutylicum

is important in order to better understand
butanol
-
producing
capabilities. Creation of mathematical and visual computation
representations of the
in vivo

metabolic pathways of
C. acetobutylicum
is
feasible, due to advances in computational modeling and quantitative
analysis of organic systems, and useful for commercial and environmental
applications.



pairs, and NADH/NAD
+

and NADPH/NADP
+

oxidation
-
reduction
couples and ATP/ADP were included in all applicable reactions. At this
point, if
-
then switches for protein concentrations were changed to ordinary
differential equations (ODE). The half
-
time for transcription was set to six
hours and the maximum kinetic constant was kept as described previously.
Time course analysis was performed on the model for each carbon source.

Previous models of the acid/solvent formation pathways in
C.
acetobutylicum

have been capable of replicating the behavior of specific batch
fermentations (Desai & Harris, 1999). Unfortunately, these models, if
created with the intention of predicting the behavior of
C. acetobutylicum
in
all environments, often come short of the anticipated goal. In the previous
experiment, the inaccuracy in the current models relies on the lack of
experimental data published in regards to the kinetics of each individual
enzyme in the pathway and the
transcriptomics

of the genes that code for
them. Nonetheless, these models furnish a useful starting point for future
predictive work given the advent of more experimental detail.

Nevertheless, the results of this experiment indicate that the current
understanding of carbohydrate utilization pathways of
C. acetobutylicum
is
extensive enough to qualitatively model the transport and degradation of
thirteen carbon sources. The pool of
pyruvate

and the sequential buildup of
intermediate metabolites that results from time course analysis on each
carbon source verifies the accuracy of this model. However, the buildup of
secondary metabolites, such as galactose
-
1
-
phosphate in the case of
galactose

utilization, is indicative of mathematical errors that result from the
same lack of experimental data as the acid/solvent formation pathways.
Fortunately, computational modeling allows for a large degree of
malleability and, given applicable future publications, these errors can be
resolved. In the future, models will include more carbon sources, such as
cellulose and
mannitol
, as well as carbon
-
catabolite

response elements .
These models, as well as the current one, will help researchers better
understand and modify
butanol
-
production in
C. acetobutylicum
.

0
1
2
3
4
5
6
7
8
9
10
0
4
8
12
16
20
24
28
Concentration (mmol)

Time (h)

Time Course Simulation
-

Galactose

[Galactose_ext]
[Galactose-6-P_int]
[Galactose-1-P_int]
[Tagatose-6-P_int]
[Tagatose-1,6-BP_int]
[Dihydroxyacetone-P_int]
[3-Phosphoglycerate_int]
[2-Phosphoglycerate_int]
[Pyruvate_int]
Figure 1 (
left
): Galactose utilization pathway in
C.
acetobutylicum
(excluding ATP and
redox

couples). 10 mmol of
galactose

is utilized to form 9.525 mmol of
pyruvate
.

Graph 1(right): Time course simulation of
galactose

utilization pathway.


A hierarchy of genomic
-
metabolic models of
C. acetobutylicum

were
constructed in
COmplex

PAthway

SImulation

(COPASI) (version 4.6.
www.copasi.org), a freely available software for analysis and simulation of
biochemical networks (Hoops et al., 2006). Initial models centered on the
acid/solvent formation pathways in
C. acetobutylicum
based upon prior work
by Desai & Harris (1999). More comprehensive models were developed,
attempting to incorporate published research on
transcriptomic

data:
carbohydrate utilization as described in
Servinksy
,
Keil
,
Dupuy
, &
Sund

(2010). The preliminary form of this model excluded oxidation
-
reduction
couples and coenzymes. Mass action kinetics was assumed, with kinetic
constants proportional to the appropriate
tRNA

concentration of genes
coding for each enzyme or transport protein. Basal values for kinetic
parameters were set to one, with a simple carbohydrate
-
based switch
deciding whether reaction rates would increase to relative rates described by
supplementary microarray data (
Servinsky
,
Keil
,
Dupuy
, &
Sund
, 2010). A
subsequent model was then developed utilizing all reactions involved in
standard
Embden
-
Meyerhoff
-
Parnasas

and
Entner
-
Dodoroff

glycolysis
, and
the pentose phosphate pathway. Finally, more attention was paid to
redox



Materials and Methods


Introduction


Present models of the acid/solvent formation pathways do not correctly
model
butanol
-
producing behavior of
C. acetobutylicum
. Time course analysis
was correctly performed on the carbohydrate utilization model, while grown
on thirteen different carbon sources:
arabinogalactans
,
arabinose
,
cellobiose
,
fructose,
galactose
, glucose, lactose, maltose, mannose, pectin, starch,
sucrose, and
xylose
. Qualitative analysis indicates the model’s capability of
degrading each carbohydrate to
pyruvate

correctly mirrors the behavior of
C. acetobutylicum

grown on each carbon source. An example of a pathway
(Figure 1 ) and the corresponding time course simulation (Graph 1) is
shown below. As expected, peaks of each intermediate metabolite appear in
order of their appearance in the metabolic network (Figure 1). Resulting
steady state concentration of 9.525 mmol for
pyruvate

indicates most
galactose

breakdown was distributed to
pyruvate

production and
glycolytic

breakdown as expected with a small quantity unexpectedly forming
galactose
-
1
-
phosphate. Other models repeat similar behaviors, with most
carbohydrate forming
pyruvate

and minor concentrations of secondary
metabolites.


Materials and Methods (cont.)

Results

Conclusion

de Lorenzo, V. (2008). Systems biology approach to bioremediation.
Current Opinions in
Biotechnology, 19,
(579
-
589).

Desai, R. P., Harris, L. M., (1999). Metabolic flux analysis elucidates the importance of the
acid
-
formation pathways in regulating solvent production by
Clostridium acetobutylicum
.
Metabolic Engineering, 1
(3), 206
-
213.

Hoops. S.,
Sahle
, S., Gauges, R., Lee, C.,
Jürgen
, P.,
Simus
, N., …
Kummer
, U. (2006).
COPASI

a
COmplex

PAthway

SImulator
.
Systems biology. 22
(24)
,
2067
-
3074.

Orth
, J.D., Thiele, I., &
Palsson
, B. Ø. (2010). What is flux balance analysis?
Nature
Biotechnology , 28
(3), 245
-
248.

Servinksy
, M., D.,
Keil
, J. T.,
Dupuy
, N.F., &
Sund
, C. J. (2010). Transcriptional analysis of
differential carbohydrate utilization by
Clostridium acetobutylicum
.
Microbiology, 156
.
doi

10.1099/mic.0.03785
-
0.

References

Acknowledgements


I would like to thank Dr. Margaret Hurley, Mrs. Linda McDonough, and
Mr. Jacob Rosenthal for all their assistance and guidance.