Pyruvate kinase-deficient Escherichia coli exhibits increased plasmid copy number and cyclic AMP levels

Abstract

Previously established consequences of abolishing pyruvate kinase (Pyk) activity in Escherichia coli during aerobic growth on glucose include reduced acetate production, elevated hexose monophosphate (HMP) pathway flux, elevated phosphoenolpyruvate carboxylase (Ppc) flux, and an increased ratio of phosphoenolpyruvate (PEP) to pyruvate. These traits inspired two hypotheses. First, the mutant (PB25) may maintain more plasmid than the wild type (JM101) by combining traits reported to facilitate plasmid DNA synthesis (i.e., decreased Pyk flux and increased HMP pathway and Ppc fluxes). Second, PB25 likely possesses a higher level of cyclic AMP (cAMP) than JM101. This is based on reports that connect elevated PEP/pyruvate ratios to phosphotransferase system signaling and adenylate cyclase activation. To test the first hypothesis, the strains were transformed with a pUC-based, high-copy-number plasmid (pGFPuv), and copy numbers were measured. PB25 exhibited a fourfold-higher copy number than JM101 when grown at 37 degrees C. At 42 degrees C, its plasmid content was ninefold higher than JM101 at 37 degrees C. To test the second hypothesis, cAMP was measured, and the results confirmed it to be higher in PB25 than JM101. This elevation was not enough to elicit a strong regulatory effect, however, as indicated by the comparative expression of the pGFPuv-based reporter gene, gfp(uv), under the control of the cAMP-responsive lac promoter. The elevated cAMP in PB25 suggests that Pyk may participate in glucose catabolite repression by serving among all of the factors that tighten gene expression.

FIG. 1.

Elimination of pyruvate kinase as an acetate reduction strategy. Increased PEP leads to (i) feedback inhibition of Pfk and elevated HMP pathway flux and/or (ii) increased flux through the anaplerotic phosphoenolpyruvate carboxylase (Ppc). Dotted and fat arrows depict lessened and increased fluxes, respectively, relative to the wild type. Key precursors for plasmid DNA synthesis (boxed) are also shown. Additional abbreviations: 3PG, 3-phosphoglycerate; αKG, α-ketoglutarate; AcCoA, acetyl coenzyme A; F6P, fructose-6-phosphate; G6P, glucose-6-phosphate; MAL, malate; OAA, oxaloacetate; Pgi, phosphoglucose isomerase; PTS, phosphotransferase system; PykA, AMP-activated pyruvate kinase; PykF, FBP-activated pyruvate kinase; PYR, pyruvate; R5P, ribose-5-phosphate; Zwf, glucose-6-phosphate dehydrogenase.

FIG. 2.

Typical growth and GFPuv induction data used for intrinsic protein synthesis rate determinations. Data are shown for JM101 grown on glucose. Use of a low-cell-density inoculum and early GFPuv induction results in a postinduction near-linear accumulation phase followed by a leveling off during which multiple doublings occur at a constant μ. Using equation 2, an intrinsic GFPuv synthesis rate (RFU/OD/h) can be determined. Data from stationary phase are not used in r_s determinations, as μ goes to zero and growth becomes unbalanced.

FIG. 3.

Extracellular cAMP accumulation. The indicated strains were grown in M9 minimal medium supplemented with the indicated carbon sources (5.0 g/liter glucose or 5.1 g/liter glycerol). Determinations were made from cultures during exponential growth at an OD of 2.4.

FIG. 4.

Verification that GFPuv resists degradation. (A) GFPuv dynamics following the cessation of induction. GFPuv production in JM101 was induced (1 mM IPTG at 1 h) and allowed to proceed until mid-exponential phase. Cells from this induced culture were harvested, washed of the inducer, and used to inoculate a fresh culture whose growth kinetics and loss of GFPuv are shown. No inducer was added to this culture. (B) Comparison of GFPuv decay from the experimental data in panel A with the projected case (solid line) where loss occurs solely due to cytoplasmic dilution. Equation 1 was used to obtain the projection, assuming r_s = k_d = 0 and μ = 0.65 h⁻¹, based on the growth curve in panel A.

FIG. 5.

Verification that GFPuv expression reports on cAMP levels. (A) The effect of exogenous cAMP addition. Growth (filled symbols) and induced GFPuv production through stationary phase (open symbols) are shown for three different cases of exogenous cAMP addition: 0 mM (squares), 1 mM (triangles), and 5 mM (circles). (B) The effect of hierarchical carbon source utilization. Growth (filled squares) and induced GFPuv production (open squares) are shown for JM101 grown in M9 plus 5.1 g/liter glycerol (high intracellular cAMP) with glucose added to 2 g/liter at 3.5 h (low intracellular cAMP). Carbon source usage switches from glycerol to glucose at this point but switches back to glycerol upon glucose exhaustion at 7 h, with the remainder of the glycerol consumed by 10 h as the culture becomes stationary. The yield of JM101 on glucose is 1 OD unit per 1 g/liter of glucose; hence, glucose exhaustion occurs after the OD has increased by 2 units from the OD at 3.5 h.

FIG. 6.

Proposed involvement of pyruvate kinase in catabolite repression. Pyk provides a thermodynamic irreversibility that maintains the PEP/pyruvate ratio at a level that results in low glucose-specific PTS enzyme IIA (IIA^Glc) phosphorylation, low AC activity, and tight cAMP-mediated catabolite repression. Without Pyk, there is more phosphorylated IIA^Glc and correspondingly leakier catabolite repression. Abbreviations, in addition to those also used in Fig. 1, are the following: EI, PTS enzyme I; HPr, PTS heat-stable, histidyl-phosphorylatable protein; IIBC^Glc, glucose-specific PTS enzyme IIBC^Glc.