| Cancer as malignant tumor can affect any part of the body,and is currently the second leading cause of death in the world.Rapid and uncontrolled proliferation is the key hallmark of cancer.The proliferation of normal cells is generally regulated by external or internal signals.Specifically,it will start when receiving sufficiently strong signals and stop when such signals diminish.In comparison,the driving signals for cancer cell proliferation remain largely unknown although various proposals have been made,including(1)signals produce by themselves and(2)constitutive activation of certain oncogenes due to genomic mutations.KRAS,MYC and SRC are among such genes.However,mutation-centric cancer-driver theory faces one issue: drugs targeted at such mutations tend to lose their efficacy within months as the rapidly evolving cancer cells tend to develop new survival pathways utilizing other mutations.From an evolutionary perspective,cancer cell proliferation may prove to be a survival mechanism for cells under certain,currently unknown stresses present in their microenvironments.In addition to cell proliferation,cancer cells also have other distinguishing characteristics such as reversal of their intracellular and extracellular p H compared with normal tissue cells and utilization of glycolysis instead of respiration as the main pathway for ATP production regardless of the level of O2 availability – an observation made by German biochemist Otto Warburg in 1927,hence widely referred to as the Warburg effect.This has perplexed generations of cancer researchers,knowing that ATP production by glycolysis is considerably less efficient compared to the normally used ATP-generation pathway by oxygenic respiration.Inspired by the idea that cancer is generally accompanied with chronic inflammation during the entire development of cancer,which tends to accumulate H2O2 and iron ions,we have studied the possible relationship between cancer proliferation and this particular combination of endogenous most molecules.Specifically,we have studied the possible persistent occurrence of the Fenton reaction: Fe2+ + H2O2 ? Fe3+ + ·OH + OH-,an inorganic reaction that does not need any enzymes whose reaction rate is solely dependent on the concentrations of Fe2+ and H2O2.To accomplish this,we have systematically examined the transcriptomic data of cancer tissue samples in a few public databases such as TCGA and GEO,to address a number of key questions related to possible drivers for cancer initiation for multiple cancer types.To accomplish this,we have developed and applied a number of data analysis and mining methods for information discovery from the tissue-based transcriptomic data.To addition,we have integrated such information with information collected from the literature covering cancer biology,chronic inflammation,wound healing among others,to predict models for drivers of cancer cell proliferation of multiple types.It has two major parts as followings:The first part of the work used transcriptomic data of more than 7,000 tissues of 14 cancer types from the TCGA and GEO databases,and microarray data of 693 chronic inflammatory diseases from GEO for validation purposes.We have predicted the occurrence of Fenton reactions in three subcellular localizations: mitochondria,cytoplasm,and extracellular matrix of cells based on gene-expression data.Specifically,we have estimated the level of the levels of the reactants and products of the reaction using gene-expression;and assessed the statistical correlation between the two sides of an Michaelis-Menten equation representing the reaction.The rationale is that it has been established that hydroxyl radical can be produced only by Fenton reactions;and the two sides have little or no correlation on gene-expression data collected from cell samples known to have no Fenton reaction.We have developed a reliable predictor for the occurrence of Fenton reactions in specific subcellular location,using expression data of genes whose protein products are known to appear in the location.We have then applied correlation analysis techniques to search for biological processes who expression levels might be affected by Fenton reactions.This is done through finding genes whose expression levels strongly correlate with our predicted level of Fenton reaction in a specific subcellular location and by pathway enrichment analyses by such genes.Our analysis results have showed that continuous Fenton reactions take place in at least three subcellular locations in cancer cells of all the cancer types we examined.Furthermore,we predicted:(i)continuous cytosolic Fenton reactions of cancer cells produce a large amount of OH-,which will increase the intracellular p H,hence casting a severe stress to the affected cells.A few metabolic processes are induced as response to the increasing intracellular p H,including induction of multiple acid-loading transporters and inhibition of acid-extruding transporters except for SLC16A1,3.As the disease advances,the levels of Fenton reactions go up,which requires increased level of acidification to keep acid-base homeostasis.One of the most dominating acidification process is the Warburg effect,i.e.,using glycolysis to produce ATP and consume the ATP by nucleotide synthesis,which together release one net proton for each ATP produced while it is p H neutral for respiration generated ATP;(ii)OHproduced by the mitochondrial Fenton reactions will change the potential difference between the mitochondrial inner membrane,thereby continuously generating ATPs using electrons donated by H2O2 and superoxide released from macrophages and neutrophils;and(iii).OH produced by Fenton reaction in the extracellular matrix will damage the structure of collagen and proteoglycan,hence providing extra growth signals required for cell proliferation.The Fenton reactions-based model links several major factors of cell proliferation in cancer,in which the rate of cell proliferation is determined by the intensity of cytosolic Fenton reactions.Under the combined effect of p H-imbalance and oxidative stresses,some genetic mutations would be created and selected,resulting in irreversible consequences,such as unlimited,uncontrolled cell proliferation.In essence,our model provides novel information,in addition to the idea of mutations driving cancer development.Specifically,there are deeper reasons for persistent cell divisions,and certain mutations are selected to enable divisions at rates comparable with those of Fenton reactions.Currently,very little research on demonstration of the occurrence of Fenton reactions in cellular locations by using computational techniques.Our study here provides a novel perspective about how cancer might have been formed.It has been previously observed that both cancer and normal proliferating cells have the Warburg effect.In the second part of our study,we have compared the causes of the Warburg effect in cancer cells and in normal proliferating cells(NPCs),and statistically showed that there are fundamental differences in the cause of the effect in cancer vs.in normal proliferating cells.Our results indicate that(i)there is no or low level of Fenton reactions in NPC;(ii)they regulate the intracellular p H in different ways;specifically,NPCs increase the alkalization of the intracellular space to drive the intracellular p H go up while cancer cells increase the acidification of the intracellular space to drive the intracellular p H go down;(iii)cancer cells synthesize glycolytic ATP for producing proton to neutralize Fenton reactions produced OH-,while NPC accumulates ATPs via aerobic respiration to increase the intracellular p H before proliferation,and the Warburg effect is used to maintain the p H at a high level by secreting lactic acids.In comparison,secretion of lactic acids by cancer cells is to weaken attacks by activated immune cells.In sum,our study strongly demonstrated that the reasons for Warburg effect in cancer cells and NPC are fundamentally different.Overall,our study,through mining cancer omic data and modeling the observed events,developed a model for cancer drivers,which starts from iron and H2O2 accumulation,largely due to chronic inflammation.The model improves our knowledge about the possible causes of cancer and complements the existing mutation centric theory of cancer.The new knowledge should enable new ways for drug target selection and improved techniques for biomarker prediction for cancer detection,prognosis and possibly treatment assessment. |
PDF Full Text Request