Page 164 - Read Online

Page 528 Belizario et al. Cancer Drug Resist 2019;2:527-38 I http://dx.doi.org/10.20517/cdr.2018.009

Nonetheless, different genetic and epigenetic variations within patients to patients can lead to a same
disease phenotype, e.g., uncontrolled cell growth. Recent advances in whole genome throughput DNA
sequencing, systems biology and machine learning algorithms have produced promise information on
[2-4]
the evolutionary dynamics of tumors, from a single cell to a cell population . The complex landscape
of somatic modifications (copy number variations, mutations, gene rearrangements) observed in breast
tumors are typically the result of a relatively small number of functional driver oncogenic alterations and a
large number of non-functional passage alterations. The functional consequence of most cancer mutations
could be characterized in phenotypic readouts such as growth assays or drug response screens using tumor
[5-7]
cell lines . The impact of mutations on kinase binding specificity has been explored in vitro protein
binding microarrays, or in vivo, for example, by measuring kinase target phosphorylation after kinase
[8]
mutation . However, what substantially contribute to oncogenesis and progression of tumors is intra and
inter-clonal heterogeneity, which is determined by a stochastic mutational process in cancer cells [9,10] .

Most common factors taken into account for classification of tumors into distinct groups include: degree
of local invasion, degree of remote invasion, histological types of cancer and specific grading based on
[2,3]
various tumor markers, and general status of the patient . However, cancers with similar morphological
and histopathological features reveal very distinct patterns of progression and response to therapy. With
growing number of database of tumor and single cell genomic, epigenomic, and pharmacogenomics
reporting on differences and similarities between clinical outcomes of patient cohort to the same therapies,
we have a rewarding opportunity to modeling and predicting outcomes to traditional chemotherapy and
[11]
targeted therapies .

Epigenetic alterations participate in all steps of cancer, from tumor initiation to cancer progression and
metastasis [1,12] . It is now well understood that both losses and gains of DNA methylation occurring in
CpGs motifs, as well as aberrant chromatin organization, contribute significantly to cancer associated
[12]
phenotypes . Epigenetic inactivation of tumor suppressor genes (TSGs) is a well-established event in
cancer progression. Oncogenic activating mutations are now known to occur in a number of epigenetic
modifiers such as isocitrate dehydrogenase 1 and 2, histone-lysine N-methyltransferase enzyme (EZH2),
[13]
DNA methyl transferases (DNMTs) . In cancer cells, global DNA hypomethylation is frequently
concomitant with both local hypo and hypermethylations (epigenetic mosaicism). This may contribute
to cancer progression through chromosomal instability, reactivation of transposable elements and loss of
imprinting. The DNMTs: DNMT1, DNMT2, and the DNMT3A/3B/3L are main components methylation
machineries whereas demethylases Ten-eleven translocation proteins TET1, TET2 and TET3, and
repressive histone-modifying enzymes of the Polycomb repressive complex 1 (PRC1) and PRC2, specifically
[12]
named, “writers” and “erasers” are modulators of epigenetic patterns . Defects in DNA methylation
could induce TSG silencing and resistance to classical chemotherapeutic agents. Interplay between DNA
methylation and DNA repair machineries may also occasionally provoke de novo methylation and
aberrant gene silencing [12-14] . Methylome profiles provides a molecular signature of cancers and might serve
as potential diagnostic and predictive biomarkers [15-17] . Some new epigenetic modifier drugs combined with
conventional chemotherapies and immunotherapies may reverse more permanent changes that affect the
[18]
epigenetic processes such as histone hypermethylation and transcriptional gene silencing .

Genomic and post-genomic research technologies have shifted to focus on biomarker discovery for
[6]
diagnosis, prognosis and prediction of treatment response to targeted therapies . Continuous discovery
of cancer subtypes has proven that tumor cell surface markers used in traditional pathology-based
[2-4]
classifications and clinical phenotyping cannot capture the full complexity of tumors . In tumors,
different cell types interact with each other in the tumor microenvironment (TME). Immune and stroma
cells of TME play an important role in tumorigenesis and development of metastasis [9,10] . TME studies
[19]
revealed that tumors can be classified into inflamed and non-inflamed tumors . The first type contains