The same protein that brings it in will carry it right back out again, but not if its structure has been changed. One enzyme rearranges atoms in the glucose molecule to turn it into fructose. Then the phosphofructokinase or PFK enzyme joins a phosphate group to the fructose molecule. That readies it for the next step in glycolysis and also prevents the transport protein from taking the sugar back out of the cell. In the same way you don't need to slice another apple if you aren't hungry and you have plenty of slices lying around, PFK doesn't need to act if there's plenty of ATP and lots of citrate; high levels of those compounds will reduce glycolysis.
Some of the steps of glycolysis require the intermediate products to get rid of a hydrogen atom so they can continue to break up and provide more energy. If there's no other molecule to accept the hydrogen atom, then glycolysis will stop.
The rate of glycolysis is also modified depending upon the amount of glucose around. If no glucose molecules are transported into the cell, then glycolysis will stop. First published in , Richard Gaughan has contributed to publications such as "Photonics Spectra," "The Scientist" and other magazines. Glycolysis is the first step in the energy metabolism of cells. Competing interests: The authors have declared that no competing interests exist. In many organisms, glycolysis is an essential pathway in energy metabolism that converts glucose to pyruvate with net production of two ATP molecules per glucose molecule [ 1 ].
In yeast, such a phenotype is usually associated with mutants of the trehalose metabolism [ 2 , 4 , 5 ]. A yeast cell trapped in the imbalanced state will restore metabolic balance and resume growth if glucose is removed within several hours, but the cell will die if the imbalance continues [ 3 ].
This is an example of substrate-accelerated death, a wider phenomenon observed in prokaryotes and eukaryotes when cells are unable to grow when exposed to excess substrate that previously limited growth [ 6 — 8 ]. The co-occurrence of a balanced and an imbalanced state in yeast glycolysis is well captured by a generalized core glycolysis model Fig 1A developed previously by van Heerden et al.
The key factor determining the fate of the system is the dynamics of inorganic phosphate P i during the start-up of glycolysis. According to the model, the transition to excess glucose proceeds as follows [ 3 ]. For a balanced steady state, v lo has to accelerate and catch up with v up. This challenge is more difficult if UG activity v up is higher, e. Accumulation of FBP binds P i and will cause a drop in its concentration in the cytosol.
Which of these two effects is dominant determines the trajectory of the system. If P i concentration remains sufficiently high, v lo will increase to become equal with v up , and a balanced steady state will be established. Otherwise, P i will become a limiting factor for LG and v lo will not accelerate fast enough to catch up with v up , causing the system to collapse into the imbalanced state.
Once caught in the imbalanced state, cells are trapped, because P i mobilized from the vacuole maintains the imbalance: at low concentrations of P i and ATP, an imported P i molecule enhances v lo , but the concomitant production of 2 ATP molecules increases v up twice as much as v lo. Given that imbalanced cells exist in an alternative stable metabolic state, random initial variation in enzyme and metabolite concentrations can be enough to drive a subpopulation of cells into the imbalanced state.
This explains why both balanced and imbalanced cells can be present in an isogenic population upon transition to excess glucose after starvation [ 3 ]. A A generalized core model of yeast glycolysis [ 3 ] considers the intracellular concentrations of the glycolytic intermediate fructose-1,6-bisphosphate FBP , ATP and inorganic phosphate P i , and four reactions arrows : i a lumped upper glycolysis reaction that produces FBP from extracellular glucose with rate v up , ii a lumped lower glycolysis reaction that generates ATP and the waste product ethanol EtOH at rate v lo , iii ATPase reaction reflecting general ATP demand in the cell at rate v atp , and iv reversible phosphate transport between the cytosol and the vacuole at rate v p.
PolyP refers to the vacuolar store of polyphosphate. B The flux through the ATPase reaction v atp and the associated cell growth rate can be visualized by water flow through a tank with two outlets.
The coupling is mediated via cellular reserves, quantified by a variable H cell health, depicted by the water level in the tank. A cell grows increases in volume when its health H is at the maximal value and v atp is larger than obligatory cellular maintenance cost v atp,c. When v atp is not sufficient to cover cellular maintenance cost, cell health decreases. The cell dies if its health drops to zero.
However, their analysis of the full kinetic glycolysis model shows that quicker liberation of P i by enhancing ATPase activity, activation of the glycerol formation branch, futile trehalose cycling, or quicker import of P i into the cytosol from the vacuole can all markedly decrease the probability of reaching the imbalanced state [ 3 ]. Furthermore, the presence of trehalose cycling combined with experimentally observed trehalosephosphate mediated inhibition of hexokinase [ 9 ] can remove the existence of metabolic imbalance in the model altogether.
Why have WT cells not evolved such mechanisms to completely eliminate the risk of imbalance? One possible evolutionary explanation is that although imbalanced v up and v lo are dangerous to the cell, regulatory mechanisms to keep them tightly balanced, or constitutive higher expression of LG enzymes are just too costly relative to the fitness benefit of avoiding substrate-accelerated death.
An alternative hypothesis that we propose here is that imbalanced v up and v lo are not always detrimental to the cell, but may, in fact, be adaptive under a range of natural conditions. In particular, allocating a larger fraction of enzymatic capacity to UG at the expense of LG would allow cells to acquire glucose from the environment faster, increasing their competitive advantage under conditions of low resource availability.
Moreover, in variable environments, glucose may normally run out before metabolic imbalance becomes irreversible, so that periods of starvation would restore normal levels of glycolytic intermediates and cells would be protected from substrate-accelerated death. From this perspective, the vulnerability of cells to fall into the imbalanced state in rich and constant environments e. In other words, we suggest that imbalanced dynamics in WT yeast cells are observed because cells are adapted to a different glucose availability regime than the one used in the experiments [ 3 ].
To investigate this idea, we performed evolutionary simulations of a population of yeast cells with the simplified yeast glycolysis pathway shown in Fig 1A , subject to different glucose availability regimes. Variation in the population was introduced by mutating maximum reaction rate constants of the pathway, reflecting changing expression levels of the glycolytic enzymes.
Cells contributed to future generations in proportion to their growth rate, so that natural selection acted on the simulated populations to improve the functionality of the pathway in the current environment. We then quantified the likelihood of cells with the evolved pathways to fall into the imbalanced state upon transitioning to excess glucose. Our results demonstrate that the regime of glucose availability that cells have previously adapted to has a marked effect on their measure of balancedness.
The model also predicts a range of environmental conditions where balanced and imbalanced cells can stably coexist in the population, and where such polymorphism drives periodic crashes and recoveries of the population. We discuss these results in relation to the tragedy of the commons and evolutionary suicide to illustrate how eco-evolutionary mechanisms can shed new light on seemingly maladaptive aspects of cellular metabolism. We model a population of yeast cells that metabolize glucose in a chemostat under anaerobic conditions.
To model glycolysis, we employ a generalized core model [ 3 ] comprising four reactions Fig 1A and S1 Model , with the addition of explicit glucose dynamics and phosphate depletion from the yeast vacuole:.
Metabolite concentrations in the cell are affected not only by glycolysis reactions, but also by dilution due to the increase in volume V of a growing cell. While this effect is often ignored in metabolic models, we included it here, because we track cells over entire generations, involving a substantial change of cell volume and accumulation of metabolites. The decrease in metabolite concentration c due to dilution can be found from the conservation of the amount of metabolite cV in the cell as its volume increases: 6 where the derivatives denoted by the prime symbol are with respect to time.
Four of the parameters of the metabolic pathway, v max,up , v max,lo , k atp and k p , reflect the expression levels of respective enzymes i. Since we aim to study evolutionary adaptation of the metabolic pathway, we must next specify its connection to growth and survival, the two key components of cellular fitness.
These fitness characteristics, together with the nature of metabolite dynamics balanced, imbalanced or intermediate constitute the phenotype of the cell. In our model, glycolysis is coupled to growth and survival by the general ATPase reaction. It is assumed that all generated ATP is used up by a cell.
This assumption is valid not only under energy glucose limitation conditions, but also under energy surplus, because our model includes an element of control of the glycolytic flux by demand, i. As a result, the growth rate of a cell with balanced metabolism is coupled to its rate of glucose uptake.
We also assume that the flux through the ATPase reaction v atp is first allocated to cover cellular maintenance costs v atp,c and that any remaining flux v atp,g is invested in cell growth, i. The maintenance costs are further decomposed into v atp,e , the ATP demand required for expressing the glycolytic enzymes, which therefore may vary between cell genotypes, and the ATP required by other transcription and general cell maintenance processes v atp,m , which is assumed to be equal between genotypes.
Hence, 8. This occurs at low intracellular ATP concentration during periods of starvation or glycolytic imbalance. We assume that cells can buffer short periods of negative energy balance by drawing on internal reserves, but that they eventually die when starvation or imbalance persists. To model the deteriorating condition of a starving cell, we introduce a variable H that reflects cell health. Cell health decreases when ATP production falls short of meeting the energy demands for maintenance i.
In fact, when v atp,g becomes positive after a period of starvation, it is first invested into replenishing the internal reserves modeled as an increase in H. We assume that the amount of ATP needed to produce a unit of new cell volume is constant and independent of cell volume or genotype.
In other words, the increase in cell volume is proportional to the total amount of ATP converted by the flux v atp,g in the cell, 9 i. This will yield an exponential increase in cell volume at constant v atp,g , consistent with experimental measurements of yeast cell growth [ 14 ]. To find the proportionality constant, we assume that a cell with the reference genotype reported by Heerden et al.
Since under these conditions the ATP demand is , the expression cost of the reference genotype is chosen to yield a positive reference growth flux Table 1.
The dynamics of cell growth in our model is therefore 10 where. Under these conditions, a cell with a reference genotype has a small ATPase flux and thus a negative reference growth flux. Cell health dynamics is therefore 11 where. Given that parameters v max,up , v max,lo , k atp and k p , which constitute the genotype of the cell, are proportional to the expression levels of glycolytic enzymes, we utilize these parameters to quantify the cost of expression, v atp,e.
The rationale to consider similar weights for multi-step, multi-enzyme pathways of UG and LG, and single-protein ATP demand and phosphate transport reactions stems from the fact that a multi-step pathway can be sped up by increasing the rate of one rate-limiting reaction e.
The nonlinearity in the cost function ensures that glucose flux through the pathway, and therefore ATP production, cannot be increased infinitely by the cell by increasing the total level of glycolytic enzyme expression. Once the cell volume has increased to twice the standard cell volume V c , the cell divides. To prevent clonal subpopulations from dividing or dying synchronously, we introduce individual variability in the initial cell volume and the parameter H max.
A division event without mutation does not alter the metabolite, nor the enzyme concentrations in the daughter cells, consistent with a simple physical division.
Daughter cells may be exposed to mutations of the genotype, which implies changing expression levels of corresponding glycolytic enzymes. The final component of the model concerns the interaction between cells and the environment. A straightforward approach is to assume that the population of cells take up glucose, grow and divide in a chemostat chamber [ 17 ]. Cells are washed out from the chamber at a rate proportional to population size N , This is equivalent to cell removal rate per capita being independent of the population size and equal to D.
A chemostat is suitable to study a population of cells that compete for nutrient, because cells take up the nutrient and thus affect its concentration in the growth chamber. A mathematical analysis of the chemostat model shows that the nutrient concentration and the population size at steady-state depend on the maximum reproduction rate of cells [ 17 ]. Cells that reproduce faster, take up the nutrient faster, and thus, at steady-state, reach a larger population size and leave less nutrient in the growth chamber.
As nutrient uptake and reproduction rates of cells evolve during evolutionary simulations, the steady-state nutrient concentration will also shift, making it difficult to determine the optimal evolutionary response of the metabolic pathway to a particular glucose availability regime. Therefore, we also considered an alternative model, where cell density is assumed to be so low that the consumption of glucose by cells has no noticeable effect on the glucose concentration in the chamber.
In this version of the model, individual cell growth is still limited by glucose availability, but cells do not compete for glucose, i. The NCG conditions can arise as a limiting case of the chemostat model where cells are attached to the substratum in a large chamber with a high flow rate, i. The glucose uptake term in Eq 15 then vanishes and cells no longer affect glucose concentration in the chamber, i. However, glucose concentration in the chamber is still affected by the medium inflow and outflow, allowing us to impose a particular glucose dynamics regime by adjusting D and [Glc] 0.
In this alternative model, cell loss rate from the chamber is 17 where d is the removal rate constant. Here, in contrast to Eq 16 , the per-capita cell loss rate increases with population size, reflecting the effect of competition for limited space. Integration was carried out with the Dormand-Prince fifth-order Runge-Kutta method [ 19 ] modified with a non-negativity constraint for the metabolite concentrations, i. A simulation was started with a population of N 0 cells.
To ensure that initial cell genotypes were sufficiently fit to survive and reproduce in a given glucose regime, initial values of the evolving parameters of each cell were drawn from a uniform distribution U 0. Simulation time was divided into three segments: i from simulation start time t s to mutation start time t ms the mutation process was disabled, allowing the establishment of a viable steady-state population from the genetic variation created at the start of the run; this was necessary because many initial random genotypes were not viable under a given glucose availability regime; ii from t ms to mutation end time t me cells were exposed to mutations enabling a gradual evolution of the metabolic pathway, iii from t me to simulation end time t e the mutation process was disabled once again to allow only the fittest genotypes to remain in the population.
To achieve that, we first performed a pre-simulation without mutation of the same duration as segment i with provisional values that regulate population size, i. V ch,0 for the standard chemostat model or d 0 for the NCG scenario.
After determining the steady-state size population size N p , full simulations with adjusted parameters or were run. We also define an indicator to quantify the balancedness of the dynamics of the model glycolysis pathway in a fluctuating environment. In a balanced cell, high external glucose coincides with high intracellular ATP, whereas in an imbalanced cell, high external glucose coincides with low intracellular ATP. The phenotypic balancedness of a cell, B p,cov is therefore defined as covariance between the external glucose concentration and intracellular ATP concentration during an integral number of cycles in a periodic environment where the cell lives, 18 B p,cov will have positive values if the dynamics of glycolysis is balanced and negative values if it is imbalanced.
This measure is appropriate if glucose and ATP values oscillate regularly around their means; it is more difficult to interpret when the dynamics is irregular, as in the case of catastrophic dynamics see Section Evolution of increased imbalancedness… below. Under the NCG scenario studied here, the external glucose concentration in the chamber changes abruptly between a high value during the ON phase and a low value during the OFF phase because of a high dilution rate D.
Therefore, a simpler and more easily interpretable measure of phenotypic balancedness, B p,phs , can be used by comparing the average ATP concentrations in the cell during ON and OFF phases: Also here, positive values indicate balanced dynamics more ATP is produced and a cell grows faster during the ON phase , whereas negative values indicate imbalanced dynamics more ATP is produced and a cell grows faster during the OFF phase.
Balanced and imbalanced glycolysis can exist as alternative steady-states for the same genotype. Therefore, we define the balancedness of a genotype B g as its propensity to exhibit balanced dynamics.
Genotypic balancedness is determined by the following procedure, in part similar to the one described by van Heerden at al. We repeat the procedure for a range of glucose values, 2. Thus, higher B g indicates a more balanced genotype. To determine whether the metabolism is balanced or not in each of these simulations, we apply the following criterion. In a balanced phenotype under constant [Glc], [FBP] reaches a steady-state value. Eq 7 shows that at the steady-state Taking into account that the cell needs time to reach metabolic steady state, the phenotype is considered balanced if Eq 20 holds true within 0.
We first investigate the evolution of the core glycolysis pathway in an environment with a constant concentration of glucose NCG scenario, see Model and methods. In each of these simulations, the pathway evolves to optimize its performance in the particular glucose availability regime that is imposed externally.
Next, we consider the evolution of the pathway in populations subject to competition for glucose in a chemostat, where adaptation of the pathway alters the ecological conditions experienced by the population. Due to this eco-evolutionary feedback, no single strategy may be optimal in a given environment, creating the possibility of more complex eco-evolutionary dynamics. Under NCG conditions, the core glycolysis pathway adapts to different constant levels of glucose availability by optimizing the expression levels of glycolytic enzymes Fig 2A.
The evolved expression pattern optimizes the balance between three selective forces. One component of selection favors an increase in v max,up , v max,lo and k atp , because the increasing flux of glucose through the pathway enhances ATP production and cell growth rate. Next, there is a pressure to lower these genotype parameters in order to reduce the cost of expression. Finally, selection acts against v max,up becoming too large compared to v max,lo to avoid the loss of fitness due to cells falling into the imbalanced state.
Each dot represents the average of an evolving genotype parameter v max,up , v max,lo , k atp and k p or a measure of balancedness at the end of an evolutionary simulation t e. Genotype parameter averages were computed over the entire population of cells; balancedness values B g,1 , B g,2 were averaged over a randomly selected subpopulation of cells that were tracked individually, and B p,phs was calculated for the subset of tracked cells that survived through at least one ON and one OFF phase.
Results of 5 replicate simulations are shown for each of the studied [Glc] and T value. As a consequence, these cells become more vulnerable to imbalanced dynamics at high glucose concentration Fig 2A , B g,1 and B g,2. Throughout, we observe low values of k p and a relatively high level of variation in this parameter, indicating that k p is under weak selection.
Since cells at constant glucose must show a balanced phenotype to survive, and phosphate transport is of little importance for balanced cells, k p likely evolves to low values solely in response to weak selection for a reduction in the cost of enzyme expression.
In this case, high expression of UG enzymes can evolve to enhance glucose uptake at low glucose without major costs to the cell, while the expression of LG enzymes cannot be increased to the same level without the cell incurring prohibitive costs.
Interestingly, cells of various balancedness coexisted in the population to form a continuum of strategies of nearly identical fitness, from strongly balanced to strongly imbalanced Figs 3 and 4 at time t me , and S1 — S3 Videos. The two extremes of this balancedness continuum illustrates two radically different strategies of survival under varying glucose. The polymorphism in the population was only observed during the mutation-on segment of the simulation i.
After mutations were stopped, only one strategy with the highest fitness survived at the final time point Fig 4 , time t e. In the v up , v lo plot, v up is shown in blue and v lo in orange. The mutation-on segment of the simulation starts at time t ms with surviving genotypes sampled from random standing genetic variation introduced at the beginning of the simulation. Mutations with small phenotypic effects then allow for a gradual evolution of the reaction rates between times t ms and t me.
Mutation is switched off again during the final segment of the simulation between t me and t e so as to allow suboptimal genotypes to be purged from the population. Points with higher r values reflect higher cell growth rates; cells with the highest r values at time t ms are the ones to survive at the end of the simulation time t e. Therefore, r appears to be a good proxy for cell fitness i. S1 — S3 Videos show the dynamics of these plots during the whole length of simulation.
Evolved k p values have smaller variation than at constant glucose and are largest at cycle lengths with the most strongly imbalanced cells Fig 2B. This is consistent with a selective pressure to keep phosphate transport at an optimal level, because it plays a crucial role for FBP accumulation in ICs. Why does the optimal strategy in a periodically fluctuating environment depend on the cycle length T?
ICs would be expected to have higher fitness than BCs irrespective of how rapidly glucose fluctuates, because they could sustain larger v max,up at the expense of v max,lo , and thus would be able to take up glucose faster than BCs. Consistent with this idea, ICs that evolved at short cycles indeed take up glucose faster than BCs that evolved at long cycles, and therefore have higher v atp flux S4 Fig.
One obvious explanation of the success of BCs in slowly varying environments could be that FBP accumulates to high concentrations in the cytosol of ICs during long ON phases, depleting phosphate reserves from the vacuole. As a result, subsequent accumulation of FBP becomes less efficient, limiting the potential of growth of ICs. FBP accumulation does indeed slow down with increasing FBP in the cytosol see Section Evolution of increased imbalancedness… below , but further analysis shows that this effect is not solely responsible for the fact that BCs outcompete ICs in slowly varying environments.
Instead, the advantage of BCs during long cycles arises because the glucose uptake rate per cell is proportional to cell volume, which changes differently through time for the two cell types.
Within this complex, intermediate products are passed directly from one enzyme to the next. Cells are expert recyclers. They disassemble large molecules into simpler building blocks and then use those building blocks to create the new components they require. The breaking down of complex organic molecules occurs via catabolic pathways and usually involves the release of energy.
Through catabolic pathways, polymers such as proteins, nucleic acids, and polysaccharides are reduced to their constituent parts: amino acids, nucleotides, and sugars, respectively.
In contrast, the synthesis of new macromolecules occurs via anabolic pathways that require energy input Figure 4. Cells must balance their catabolic and anabolic pathways in order to control their levels of critical metabolites — those molecules created by enzymatic activity — and ensure that sufficient energy is available.
For example, if supplies of glucose start to wane, as might happen in the case of starvation, cells will synthesize glucose from other materials or start sending fatty acids into the citric acid cycle to generate ATP. Conversely, in times of plenty, excess glucose is converted into storage forms, such as glycogen, starches, and fats.
Figure 4: Catabolic and anabolic pathways in cell metabolism Catabolic pathways involve the breakdown of nutrient molecules Food: A, B, C into usable forms building blocks. In this process, energy is either stored in energy molecules for later use, or released as heat. Anabolic pathways then build new molecules out of the products of catabolism, and these pathways typically use energy. The new molecules built via anabolic pathways macromolecules are useful for building cell structures and maintaining the cell.
Figure 5: Feedback inhibition When there is enough product at the end of a reaction pathway red macromolecule , it can inhibit its own synthesis by interacting with enzymes in the synthesis pathway red arrow. Figure Detail Not only do cells need to balance catabolic and anabolic pathways, but they must also monitor the needs and surpluses of all their different metabolic pathways. In order to bolster a particular pathway, cells can increase the amount of a necessary rate-limiting enzyme or use activators to convert that enzyme into an active conformation.
Conversely, to slow down or halt a pathway, cells can decrease the amount of an enzyme or use inhibitors to make the enzyme inactive. Such up- and down-regulation of metabolic pathways is often a response to changes in concentrations of key metabolites in the cell.
For example, a cell may take stock of its levels of intermediate metabolites and tune the glycolytic pathway and the synthesis of glucose accordingly. In some instances, the products of a metabolic pathway actually serve as inhibitors of their own synthesis, in a process known as feedback inhibition Figure 5. For example, the first intermediate in glycolysis, glucosephosphate, inhibits the very enzyme that produces it, hexokinase.
This page appears in the following eBook. Aa Aa Aa. Cell Metabolism. What Do Enzymes Do? Figure 1: Glycolysis. Energy is used to convert glucose to a 6 carbon form.
Figure 2: Activation and inactivation of of enzyme reaction. Enzymes are proteins that can change shape and therefore become active or inactive. What Are Metabolic Pathways? Figure 3: Reaction pathway.
0コメント