Biomedical Engineering Reference
In-Depth Information
previous theme (p. 302), and this approach has
been widely adopted (Baltz 1998). However, novel
polyketides can also be generated by simply chang-
ing the
order
of the different activities in type I
synthases (McDaniel
et al.
1999).
H
O
Actinorhodin
O
CH
3
O
COOH
O
H
O
Engineering metabolic control over
recombinant pathways
Medermycin
N(CH
2
)
2
When a recombinant cell overproduces a protein or
the components of a biosynthetic pathway, there is a
marked reduction in metabolic activity, coupled with
a retardation of growth (Kurland & Dong 1996).
These phenotypic characteristics result from the fact
that the constant demands of production placed
upon the cell interfere with its changing require-
ments for growth. To alleviate this problem, Farmer
and Liao (2000) designed a dynamic control circuit
that is able to sense the metabolic state of the cell and
thereby regulate the expression of a recombinant
pathway. This approach is termed 'metabolic control
engineering'.
An essential component of a dynamic controller is
a signal that will reflect the metabolic state of the
cell. Acetyl phosphate was selected as the signal,
since it is known to be a regulator of various operons
influenced by nutrient availability. A component of
the Ntr regulon in
E. coli
, NRI, is capable of sensing
the acetyl phosphate level in the cell. When phos-
phorylated by acetyl phosphate, it is capable of
binding to the
glnAp2
promoter and activating
transcription (Fig. 14.20). To reconstruct this
control module, the NRI-binding site and the
glnAp2
promoter were inserted into a plasmid vector up-
stream of a cloning site.
When the
lacZ
gene was placed under the control
of the
glnAp2
promoter, there was no significant
H
O
HO
O
CH
3
H
3
C
O
O
H
O
O
O
actVA
gene
Mederrhodin
N(CH
2
)
2
H
O
HO
O
CH
3
H
3
C
O
O
O
H
O
O
O
Fig. 14.19
The formation of the new antibiotic mederrhodin
from medermycin by the actVA gene product.
with other polyketides (for review, see Baltz 1998)
and is known as
combinatorial biosynthesis
.
Once a number of polyketide biosynthetic gene
clusters had been cloned and sequenced, new
insights were gained on the mechanism of synthesis.
In particular, two enzymic modes of synthesis were
discovered. In particular, polyketide synthesis takes
place on an enzyme complex in a manner analogous
to fatty acid synthesis. Furthermore, there are two
types of complex. In type II complexes, the different
enzymic activities are encoded by separate subunits.
In contrast, in type I synthesis all the different
enzyme activities are encoded by a single, very large
gene. Clearly, the polyketide synthases are prime
candidates for DNA shuffling, as described in the
-
galactosidase synthesis until late in the exponential
phase of growth, just as expected. A similar result
was obtained when the
lacZ
gene was replaced
with a construct encoding two different metabolic
enzymes. Finally, as a real test of the system, an en-
gineered construct encoding a pathway for the
synthesis of the carotenoid lycopene was placed
under the control of the
glnAp2
promoter. This was
a particularly interesting test because one of the pre-
cursors of lycopene, pyruvate, is also an immediate
precursor of acetyl phosphate. Once again, product
β
