A network-based approach to cell metabolism: from structure to flux balances

Abstract

[cat] La visió completa del metabolisme cel·lular, és a dir, tenint en compte totes les reaccions que el componen, permet descobrir nous mecanismes i respostes que són impossibles d’obtenir amb els mètodes reduccionistes tradicionals. L’estudi d’una xarxa metabòlica completa requereix eines que pertanyen a la Biologia de Sistemes i a la Ciència de les Xarxes Complexes. La present tesi mostra com la combinació de les eines que pertanyen a aquests dos camps es pot aplicar per a descobrir noves propietats de les xarxes metabòliques. D'aquesta manera, s’han estudiat les xarxes metabòliques de tres bacteris amb les següents eines: (1) algoritme de cascada, que es pot usar per estudiar si les xarxes metabòliques poden sobreviure a inactivacions de determinades reaccions, (2) Flux Balance Analysis, que s’usa per a calcular els fluxos a través de les reaccions que composen la xarxa metabòlica suposant que l’objectiu biològic de l’organisme a estudiar és maximitzar la velocitat de creixement, (3) Disparity Filter, que permet obtenir versions reduïdes de xarxes metabòliques, cosa que facilita el seu estudi i anàlisi, i (4) Hit-And-Run, que permet obtenir totes les solucions metabòliques independentment de que maximitzin el creixement de l’organisme. En aquesta tesi es demostra que el metabolisme de les cèl·lules dels organismes vius ha evolucionat de forma que aconsegueix sobreviure a les inactivacions de les reaccions que el componen. Addicionalment, s’identifiquen les rutes metabòliques responsables dels processos evolutius i adaptatius que es donen en les xarxes metabòliques. A més, també es demostra que la tècnica Flux Balance Analysis dóna una solució de fluxos que no es representativa de totes les possibles solucions. Cal remarcar que això no invalida la tècnica, sinó que les assumpcions que usa donen una solució concreta que té sentit biològic però que és molt diferent de la resta de solucions. És important recalcar que els resultats obtinguts en aquesta tesi podrien emprar-se en aplicacions mèdiques, per exemple estudiar el metabolisme de les cèl·lules cancerígenes, que podia utilitzar-se per a que aquestes cèl·lules no proliferin en el cos dels humans, un fet que causa moltes problemes en l'ésser humà.

[eng] The thesis called “A network-based approach to cell metabolism: from structure to flux balances” shows how the vision of cell metabolism as a whole allows to unveil new mechanisms and responses impossible to reach by traditional reductionist approaches. Different lines of research have been used, and each one has allowed extracting new insights about the properties of cell metabolism of three organisms, Mycoplasma pneumoniae, Escherichia coli, and Staphylococcus aureus. To do so, tools that belong to the complex network science and Systems Biology have been used. The first line of study analyzes how the structure of the metabolic networks of the three mentioned organisms respond when their metabolic networks are affected by perturbations, i.e., when a reaction or a set of them are forced to be non-operative. To do this, the applied algorithm spreads a structural cascade when an initial reaction is forced to be non-operative. This study determines that evolutionary pressure favors the ability of efficient metabolic regulation at the expense of losing robustness to reaction failures. The second line of study focuses on the application of the technique called Flux Balance Analysis (FBA), which is able to compute the fluxes of all reactions composing a metabolic network, assuming that the biological target of the organism is to maximize maximizes the growth rate. The study of synthetic lethal pairs in E. coli and M. pneumoniae with FBA allows identifying two protection mechanisms called plasticity and redundancy. Plasticity sets up as a backup mechanism that is able to reorganize metabolic fluxes turning on inactive reactions when coessential counterparts are removed in order to maintain viability in a specific medium. Redundancy corresponds to a simultaneous use of different flux channels that ensures viability and besides increases growth. The third part combines FBA and the technique called Disparity Filter in E. coli and M. pneumoniae to obtain metabolic backbones, which are reduced versions of metabolic networks composed by the most relevant connections, this relevancy being determined by the importance of the chemical fluxes. One finds that the disparity filter recognizes metabolic connections that are important for long-term evolution, these connections being related to ancestral pathways. In addition, the disparity filter identifies metabolic connections that are important for short-term adaptation. These connections are related to pathways whose reactions quickly adapt to external stimuli. The last line of study studies whether the assumption of maximizing the growth rate leads to a representative solution or not. Although FBA gives a single solution, there exist a number of other possible solutions that are chemically feasible but that do not maximize growth, and that form part of the whole flux space. In this way, the third line of study computes all the possible solutions, obtaining in this way the whole space of flux solutions of E. coli. The information content in the whole space of solutions provides with an entire map of phenotypes to evaluate behavior and capabilities of metabolism. Therefore, it is found that the FBA solution is eccentric compared to the mean of solutions. In addition, the whole flux solution map can be used to calibrate the deviation of FBA from experimental observations. To finish, in the map it is possible to find solutions that perform aerobic fermentation, a process which is impossible to recover with FBA computations unless extra constraints are used. The obtained results could be applied in medical applications, for example to study the metabolism of cancer cells. Thus, it could be a way to study how to force that these cells do not proliferate in the human body, a fact that causes many problems in humans.