Biology Reference
In-Depth Information
Often biochemical and physiological data for transport
reactions may not be available. In such cases, educated
assumptions can be made based on similar transport reac-
tions, and pathway-based algorithms can also be used to
suggest potential transport reactions [28] . However, for any
of these predictions, the lowest confidence score is assigned
to the reaction as well as a note explaining why the trans-
port reaction was added with a specific stoichiometry.
with the stoichiometric coefficients being organism
specific. Non-growth-associated maintenance (NGAM) is
a measure of the energy needed for maintaining homeo-
stasis (e.g., turgor pressure [36] ) in the cell and is included
as a separate reaction in the network with the form:
þ
þ
1P i þ
þ
1 ATP
1H 2 O
/
1 ADP
1H
The flux through the NGAM reaction must be experi-
mentally determined and constrained to the appropriate
value when using the reconstruction for modeling.
The Biomass 'Reaction'
The biomass objective function [29] or 'reaction' allows
a metabolic network model to simulate growth. The
biomass reaction is a mathematical representation of the
molar amounts of various metabolites needed to make all
cellular components for growth ( Box 12.5 ). Therefore, in
order to accurately model growth, it is necessary to know
the overall composition (e.g., protein, lipids, etc.) of the cell
and incorporate this information into the biomass reaction.
Such information can be gathered from primary literature
[30] or determined experimentally [31 e 34] . The amino
acid and nucleic acid contributions can be measured or
estimated from the genome, e.g., using the CMR database
[35] . Lipid content is often inferred from experimental data
by calculating the average molecular weight of fatty acid
chains and using this to compute phospholipid contribu-
tions to biomass for the three major phospholipid groups
[9] . Additionally, the content of the soluble pool (poly-
amines, vitamins, and cofactors) as well as the ion content
of the cell needs to be determined and incorporated. Finally,
the amount of energy the cell requires for replication, also
known as the growth-associated maintenance (GAM),
needs to be experimentally determined and incorporated
into the biomass function as follows:
Stage 3: Converting a Network to
a Mathematical Model
To be used as a predictive tool, a reconstruction must be
converted into a model. A reconstruction consists of a list
of the reactions taking place within the organism and
detailed information about each reaction and metabolite.
However, a metabolic model is the computable form of
the reconstruction. Therefore, to be a model, the network
must be (1) described mathematically and (2) provide
system boundaries with constraints on network inputs and
outputs.
To describe the reconstruction mathematically, each
reaction gathered during the reconstruction is included as
a column in a stoichiometric matrix, S. Each metabolite in
the network is included as a row. The elements of this
matrix represent the stoichiometric coefficients of each
metabolite in a given reaction. Metabolites that are
consumed in a reaction have negative values and those
produced have positive values. For example if in reaction j,
metabolite i is consumed and has a stoichiometric coeffi-
cient of 1, then S ij ¼
-1 ( Figure 12.2 , inset).
After the system boundaries are set, extracellular
metabolites that can be obtained from the media or secreted
are determined and set as inputs or outputs of the model,
respectively. For modeling purposes, the uptake or secre-
tion of these extracellular metabolites is facilitated by using
exchange reactions. These reactions allow the metabolites
to enter or leave the system, thus, simulating their avail-
ability in the environment or their excretion ( Box 12.6 ).
Constraints can be subsequently placed on exchange reac-
tions to provide upper limits on the uptake or secretion rates
of their associate metabolites.
xH þ
Since GAM is the energy needed by the cell to replicate, x
accounts for all the ATP requirements for macromolecule
synthesis and other growth-related processes. The inclusion
of GAM will result in a biomass function of the following
general form:
x ATP
þ
xH 2 O
x ADP
þ
xP i þ
/
dNTPs þ AAs þ lipid precursors þ ions þ soluble pool
þ
H þ
ATP
þ
H 2 O
ADP
þ
P i þ
/
BOX 12.5 Biomass
Many important biomass components (e.g., nucleic acids)
are not explicitly accounted for in core metabolism, so their
precursors (e.g., pyruvate and L -glutamate for L -alanine) were
substituted in the biomass reaction. The inclusion of the
biomass reaction increases the reaction count of the recon-
struction to 75.
BOX 12.6 Network Conversion and Exchange Reactions
A visualization of the conversion from reconstruction to stoi-
chiometric matrix is seen in Figure 12.2 , inset for a subset of
the TCA cycle. An additional 20 reactions were added as
exchange reactions to simulate known uptake and/or secre-
tion capabilities of E. coli (e.g., glucose, ethanol, and lactate).
This brings the final reaction count of the model to 95.
Search WWH ::




Custom Search