Is the kinetome conserved?

Abstract

Computational biologists have labored for decades to produce kinetic models to mechanistically explain complex metabolic phenomena. The estimation of numerical values for the large number of kinetic parameters required for constructing large-scale models has been a major challenge. This collection of kinetic constants has recently been termed the kinetome (Nilsson et al, 2017). In this Commentary, we discuss the recent advances in the field that suggest that the kinetome may be more conserved than expected. A conserved kinetome will accelerate the development of future kinetic models of integrated cellular functions and expand their scope and usability in many fields of biology and biomedicine. Mol Syst Biol. (2022) 18: e10782 Genome-scale models O nce whole-genome sequences became available in the mid-1990s, the reconstruction of metabolic networks reached the genome-scale. Over the past two decades, genome-scale metabolic models (GEMs) built from these network reconstructions have contributed to our understanding of systems biology of metabolism, with applications ranging from metabolic engineering to bacterial evolution to infectious disease. Genetic variation can now be incorporated into genome-scale models of metabolism, transcription, translation , proteostasis, and cellular stresses. GEMs are often simulated using flux balance analysis that requires a minimal number of parameters estimated from empirical data. Kinetic models, on the other hand, have traditionally been limited by the need for extensive parameterization. If the kinetome can be estimated from high-throughput data, then we can develop large-scale-even genome-scale-kinetic models. Since GEMs have a direct genetic basis, a new generation of kinetic models can be directly rooted in protein structures and sequence variation. The kinetome of such models would be large, requiring the estimation of many unmeasured parameters. However, if segments of the kine-tome are conserved, then parameterization will be simplified. Reference kinetomes can be estimated for well-characterized strains and applied to less well-known strains. Estimating the kinetome using omic data As omic technologies advanced, it was recognized that multi-omic data combined with GEMs could lead to the estimation of a large set of enzyme turnover rates, the most important kinetic parameters in the kinetome. Indeed, in vivo turnover rates of bacterial enzymes have been characterized using ratios of proteomic and fluxomic data (Davidi et al, 2016). This landmark study showed that in vitro enzyme assays concur with maximal in vivo rates for many enzymes, and that in vivo estimated parameters could be used to fill in some of the scarcity in the parameterization of large-scale kinetic models (Fig 1A). Enzyme turnover rates are largely conserved A few recent studies have suggested that the kinetome is more conserved than previously thought. The abundance of available whole-genome sequences allows for large-scale allelic comparison across metabolic genes, and the initial analysis of such data shows that most metabolic genes have a low amino acid substitution rate (Norsigian et al, 2020) (Fig 1B). Thus, only a small number of metabolic genes seem to face selection pressures , suggesting that their estimated enzyme turnover rates may have broad applicability. This potential broad applicability of estimated turnover rates has been further supported by two recent adaptive laboratory evolution (ALE) studies. ALE allows for the generation of strains that have adapted to high growth rates following the deletion of genes that encode specific metabolic enzymes. This approach results in the generation of "metabolic specialist" strains whose pathway usage has been rewired by ALE following the loss of a key metabolic enzyme. A large study resulted in an estimation of turnover rates for the same enzyme in multiple metabolic specialists (Fig 1C). These estimates were consistent among the specialists and with the wild type. Consistently, only relatively few structural mutations were identified, but regulatory mechanisms altered the abundance of metabolic enzymes, resulting in the required alteration of metabolic fluxes (McCloskey et al, 2018). A second ALE study swapped glycolytic genes in E. coli with orthogenes from a diverse range of other species, from hyper-thermophilic archaea to humans (Sandberg et al, 2020). Following ALE, many E. coli lineages adapted to use the orthogenes to replace their own. Adaptive mutations were rarely found in orthogene coding sequences, with the majority of mutations falling within regulatory regions that altered enzyme expression levels. These two ALE studies suggest that optimal flux levels in vivo are more often impacted by the adjustment of an enzyme's abundance rather than an alteration to its turnover rate (Fig 1D). Again, these results suggest that the kinetome may exhibit a notable degree of conservation.