Erroneous energy-generating cycles in published genome scale metabolic networks: Identification and removal

Abstract

Energy metabolism is central to cellular biology. Thus, genome-scale models of heterotrophic unicellular species must account appropriately for the utilization of external nutrients to synthesize energy metabolites such as ATP. However, metabolic models designed for flux-balance analysis (FBA) may contain thermodynamically impossible energy-generating cycles: without nutrient consumption, these models are still capable of charging energy metabolites (such as ADP→ATP or NADP+→NADPH). Here, we show that energy-generating cycles occur in over 85% of metabolic models without extensive manual curation, such as those contained in the ModelSEED and MetaNetX databases; in contrast, such cycles are rare in the manually curated models of the BiGG database. Energy generating cycles may represent model errors, e.g., erroneous assumptions on reaction reversibilities. Alternatively, part of the cycle may be thermodynamically feasible in one environment, while the remainder is thermodynamically feasible in another environment; as standard FBA does not account for thermodynamics, combining these into an FBA model allows erroneous energy generation. The presence of energy-generating cycles typically inflates maximal biomass production rates by 25%, and may lead to biases in evolutionary simulations. We present efficient computational methods (i) to identify energy generating cycles, using FBA, and (ii) to identify minimal sets of model changes that eliminate them, using a variant of the GlobalFit algorithm.

Fig 1. A futile cycle that consumes energy drawn from a cofactor pool (left) and an energy generating cycle (EGC) (right), which is thermodynamically impossible but occurs in some metabolic network models (figure extended from [12]).

We can convert the type-II pathways to type-III pathways by closing the cycles in the cofactor pools (dashed arrows).

Fig 2. A simple (hypothetical) example of an energy generating cycle (EGC).

A symporter that exports a metabolite and a proton acts together with a transporter that takes the same metabolite up without a proton. A combination of both reactions builds up a proton gradient that can then be utilized to generate energy (e.g., via an ATP synthase).

Fig 3. The majority of metabolic network reconstructions in two of the examined databases (ModelSEED and MetaNetX) contain erroneous internal EGCs that generate energy.

In contrast, most models in BiGG do not contain EGCs. Total bar size reflects the number of models contained in each database. Green: models without EGCs; purple: models with EGCs that could be corrected through GlobalFit; orange: models with EGCs that cannot be corrected through reaction removals.

Fig 4. Most erroneous models can be corrected by making up to 5 originally reversible reactions irreversible.

Purple: histogram of the number of irreversible reactions removed in each model to eliminate EGCs. Orange: histogram of the number of reversible reactions made irreversible to eliminate EGCs.

Fig 5. Removal of EGCs led to substantially reduced maximal biomass yield in most models.

Histogram of the ratio between maximal biomass production rate before and after EGC removal.

Fig 6. Examples of EGCs found in published genome-scale models.

Green/red: metabolites; blue: reactions, linking substrates and products; orange: direction of the energy gradient utilized by the energy dissipation reaction. (A) The simplest identified cycle, which links a Na+/proton antiporter (exporting Na+ in exchange for a single proton) and a Malate/proton symporter (importing Malate together with two protons) via a Malate/Na+ symporter. (B) A cycle involving two antiporters and one symporter, driven by a transporter that translocates tartrate from the periplasm to the cytosol. (C) A NADH:menaquinone oxidoreductase, which translocates protons in the process of transferring electrons from NADH to Menaquinone 8, driven by a chain of four enzymes. (D) rxn00379 creates Adenosine 5'-phosphosulfate from ATP and sulfate. The equivalent sulfate adenyltransferase rxn09240 catalyzes the backward reaction, but charges a GTP in addition to the ATP.