Targeted gene insertion methodology was used to study the effect of perturbing ot-aminoadipic acid precursor flux on the overall production rate of ,(-lactam biosynthesis in Streptomyces clavuligerus. A high-copy-number plasmid containing the lysine E-aminotransferase gene (lat) was constructed and used to transform S. clavuligerus. The resulting recombinant strain (LHM100) contained an additional complete copy of tat located adjacent to the corresponding wild-type gene in the chromosome. Biological activity and production levels of I-lactam antibiotics were two to five times greater than in wild-type S. clavuligerus. Although levels of lysine r-aminotransferase were elevated fourfold in LHM100, the level of ACV synthetase, whose gene is located just downstream of lat, remained unchanged. These data strongly support the notion that direct perturbation of ox-aminoadipic acid precursor flux resulted in increased antibiotic production. This strategy represents a successful application of metabolic engineering based on theoretical predictions of precursor flux in a secondary metabolic pathway.A critical component of understanding the basic biological processes of secondary metabolism involves dissecting the molecular mechanisms that control carbon flow from primary to secondary metabolic pathways (7). Although the biosynthesis of P-lactams represents one of the most thoroughly studied antibiotic pathways, there remains a fundamental lack of knowledge about the control of carbon flow into the synthesis of these mctabolites. Our approach to understanding the control of r-lactam biosynthesis has been to use a two-stage strategy in which (i) a kinetic model is constructed to predict an essential rate-limiting enzymatic step and how this step is controlled by precursor flux and (ii) the model is tested by engineering a novel biosynthetic pathway that enhances precursor flux or kinetic parameters of the predicted key ratelimiting enzyme.The wealth of information available on specific biosynthetic steps, including enzyme kinetic data, has provided unique opportunities to analyze rate-limiting reactions in construction of f-lactam antibiotics from i-(L-cx-aminoadipyl)-L-cysteinyl-Dvaline (ACV) tripeptide and its amino acid precursors (5) (Fig.

A kinetic model describing the cephalosporin biosynthesis in Streptomyces clavuligerus was developed. Using previously reported kinetic data of biosynthetic enzymes, we examined the kinetics of cephalosporin production. The predicted time profile of the specific production rate during a batch culture parallels that of experimental observation. Sensitivity analysis reveals that &(L-cY-aminoadipyl)-Lcysteinyl-o-valine (ACV) synthetase is the rate-limiting enzyme. The effect of amplifying ACV synthetase on the specific production rate was analyzed theoretically. Increasing ACV synthetase enhances the production rate initially until deacetoxycephalosporin C hydroxylase becomes ratelimiting. Such kinetic analysis can provide a rational basis for modifying the biosynthetic machinery of cephalosporin through gene cloning.

A recombinant strain of S. clavuligerus (LHM100) that contains an additional copy of the gene (lat) encoding lysine epsilon-aminotransferase (LAT) was analyzed and compared to the wild-type for intracellular concentrations of primary metabolites involved in cephamycin C biosynthesis. This strain had been shown previously to produce higher levels of the antibiotic because of increased levels of LAT, a rate-limiting enzyme involved in the production of alpha-aminoadipic acid. The results showed that the overall growth kinetics of the two strains were comparable, including the intracellular concentrations of cysteine, valine and lysine. In contrast, 60% higher antibiotic production was observed in LHM100, which reflected a significant temporal variation in specific metabolite production rate. The time profile of LAT activity was consistently higher in LHM100; however, alpha-aminoadipic acid levels showed unexpected variation during the growth cycle. These results support the proposal that rate-limiting enzymes in cephamycin C biosynthesis are temporally controlled, and indicate that optimization of metabolite production will require differential overexpression of several biosynthetic genes.

The mercury resistance (mer) operon of plasmid R100 was cloned onto various plasmid vectors to study the effect of mer gene amplification on the rate of Hg2+ reduction by Escherichia coli cells. The plasmids were maintained at copy numbers ranging from 3 to 140 copies per cell. The overall Hg2' reduction rate of intact cells increased only 2.4-fold for the 47-fold gene amplification. In contrast, the rate of the cytoplasmic reduction reaction, measured in permeabilized cells, increased linearly with increasing gene copy number, resulting in a 6.8-fold overall amplification. RNA hybridizations indicated that mRNA of the cytoplasmic mercuric reductase (merA gene product) increased 11-fold with the 47-fold gene amplification, while mRNA of the transport protein (merT gene product) increased only 5.4-fold. Radiolabeled proteins produced in maxicells were used to correlate the expression levels of the mer polypeptides with the measured reduction rates. The results indicated that, with increasing gene copy number, there was an approximately 5-fold increase in the merA gene product compared with a 2.5-fold increase in the merT gene product. These data demonstrate a parallel increase of Hg2+ reduction activity and transport protein expression in intact cells with plasmids with different copy numbers. In contrast, the expression level of the mercuric reductase gene underwent higher amplification than that of the transport genes at both the RNA and protein levels as plasmid copy number increased.