Figure 6.

Predictions of the model for the GS-A levels of two computational strains. (a) In the first strain, AT/AR is regulated only by GlnB but not by GlnK, yet GlnK still regulates AmtB and competes with GlnB for UT/UR. Before starvation, the levels of GS-A are not affected by GlnK due to low GlnK abundance. If GlnK is assumed not to regulate AT/AR, the level of GS-A after ammonium upshift is reduced, implying increased ammonium assimilation. Hence the high level of GlnK induced during starvation in the WT acts as a buffer to prevent overshoots in ammonium assimilation by regulating GS activity. (b) In the second strain, the expression of glnK (and conjointly of amtB) is constitutive (unregulated). To avoid the overshoot due to ammonium upshift, a high level of GlnK is necessary before starvation. This overabundance of GlnK prestarvation leads to higher GS-A levels, and hence suboptimal ammonium uptake. In each figure, the gray-shaded area indicates the ammonium starvation period; the dashed line shows the original model for the WT data (same as in Fig. 5); and the solid lines are predictions of the model for the respective computational strains.