An enquiry into metabolite domains

Abstract

It is currently assumed that two or more pools of the same metabolite can coexist in the cytosolic compartment of mammalian cells. These pools are thought to be generated by the differential subcellular location of enzymes and transporters, much in the way calcium microdomains arise by the combined workings of channels, buffers, and pumps. With the aim of estimating the amplitude and spatial dimensions of these metabolite pools, we developed an analytical tool based on Brownian diffusion and the turnover numbers of the proteins involved. The outcome of the analysis is that ATP, glucose, pyruvate, lactate, and glutamate cannot be concentrated at their sources to an extent that would affect their downstream targets. For these metabolites, and others produced by slow enzymes or transporters and present at micromolar concentrations or higher, the cytosol behaves as a well-mixed, homogenous compartment. In contrast, the analysis showed microdomains known to be generated by calcium channels and revealed that calcium and pH nanodomains are to be found in the vicinity of slow enzymes and transporters in the steady state. The analysis can be readily applied to any other molecule, provided knowledge is available about rate of production, average concentration, and diffusion coefficient. Our main conclusion is that the notion of cytosolic compartmentation of metabolites needs reevaluation, as it seems to be in conflict with the underlying physical chemistry.

FIGURE 1

Geometry of the model and definition of parameters. (A) A mammalian cell is pictured as a sphere of radius b. A metabolite is produced by a single source of radius a, located at the center of the cell, and diffuses freely, to be cleared at the cell surface (sink). (B) A metabolite domain generated in the steady state. The concentration along the radius of the cell was calculated using Eq. 4. In this example, the concentration at the source was 2.7 mM and the average was 0.1 mM, Amplitude was therefore 27. At 25 nm from the source (arrow), a distance that is referred to as the domain Extension, the concentration falls to below a value equivalent to twice the cell average.

FIGURE 2

3-PGK as a source of ATP. (A) The three-dimensional crystal structure of yeast 3-PGK is outlined in shading, together with the projection of its products Mg-ATP (solid) and 3-phosphoglycerate (hatched), modified from Watson et al. (18). Note that the bulky structure of the enzyme provides a lower limit for the size of the ATP source. (B) A hypothetical cluster of all 3-PGK copies contained in an astrocyte would occupy a sphere of ∼100 nm diameter. The lack of domain generation by either arrangement is discussed in the text.

FIGURE 3

Unrealistic conditions are needed to form ATP domains. Domain Amplitude was estimated using Eq. 10 for a source of 1 nm diameter in a cell of 10 μm diameter. (A) After fixing the diffusion coefficient D at 500 μm²/s and the average concentration ū at 5 mM, Amplitude was determined at increasing turnover numbers (q). The arrow points to the place in the curve where 3-PGK is located. (B) With a ū of 5 mM and a turnover number (q) of 1000 s⁻¹, Amplitude was determined at increasing values of D. The arrow indicates the position of ATP, given by its measured value of D in cytosol. (C) Amplitude was calculated at increasing values of ū, fixing D at 500 μm²/s and q at 1000 s⁻¹. Note that five orders of magnitude separate the physiological condition from that required by a domain.