Clinical epigenetics-seizing opportunities for translation.pdf

For the past two decades,the study of genetic variants has held the most promise inpersonalized medicine.The Human Genome Project was at the forefront of several international efforts to fully describe the genome in both normal and diseased tissue1.It is indisputable that the knowledge gleaned from these genetic studies has strongly contributed to our understanding of the molec-ular basis of several human disorders and has even provided a few cases that are translated into clinical practice.However,through this long process,we have learned that the phenotype of complex disease states cannot be fully explained by single genetic variants and that the roles of specific genes may have been overes-timated2.For complex traits such as common diseases,the risk,progression and response to treatment are dif-ficult to model and leverage in clinical decision-making as they do not follow classic genetic heritability.Despite the development of computational methods to infer risk prediction models on the basis of the interaction between genetic variants and environmental factors3,in most cases common genetic traits do not produce a con-sistent phenotype.As the combined effects of known trait-associated genetic variants typically explain only part of the heritability of complex traits,the term miss-ing heritability was coined to highlight the polygenic nature and the response to the environment of complex phenotypes4.Epigenetics provides a molecular explanation to bridge the gap between the genome and environmental signals during development and can be associated with lifestyle and environmental conditions either during intra-uterine or postnatal development(Box1).The flexibility of theepigenome represents an enticing opportunity to understand disease variation,and several initiatives to obtain the epigenome of human disorders were rapidly developed,particularly in cancer2,5.Despite limitations in methodologies and/or sample size,epigenomic maps of human health and disease are now available and form a complex network with genetic and environmental data.In this Review,we provide a comprehensive state-of-the-art overview of clinical epigenetics in major human disorders.Oncology is the main focus in trans-lational epigenetics,with biomarkers approved by the US Food and Drug Administration(FDA)for diagnosis,prognosis or response to therapy(here termed pharma-coepigenetics)(Fig.1),as well as epigenetic-based ther-apy(here termed epidrugs)(Fig.2).Extending from the paradigm of cancer,we are close to the implementation of clinical epigenetics in a plethora of complex disor-ders,as demonstrated by the large number of clinical trials currently performed for neurological,immuno-logical,metabolic(such as obesity and type 2 diabetes)and infectious diseases.We provide brief comments on human disorders that are under preclinical evaluation as they represent a major challenge in epigenetic research,including cardiovascular diseases,ophthalmic condi-tions and human infertility.Beyond providing clues to the deregulation of specific molecular pathways,we emphasize the most advanced application of epigenetics in the clinical setting for each pathology specifically,tests that analyseCpg methylation or histone modifications.Discussion of the role of non-coding RNAs(ncRNAs)is largely outside the scope of this article,and reviews have been published elsewhere6,7.Finally,we provide insights Personalized medicine(Also known as precision medicine).A form of medicine that uses specific information about a person(for example,genes,proteins or environment)to tailor preventive care,disease prognosis or drug therapy.Clinical epigenetics:seizing opportunities for translationMaraBerdasco 1*and ManelEsteller 1,2,3,4,5*Abstract|Biomarker discovery and validation are necessary for improving the prediction of clinical outcomes and patient monitoring.Despite considerable interest in biomarker discovery and development,improvements in the range and quality of biomarkers are still needed.The main challenge is how to integrate preclinical data to obtain a reliable biomarker that can be measured with acceptable costs in routine clinical practice.Epigenetic alterations are already being incorporated as valuable candidates in the biomarker field.Furthermore,their reversible nature offers a promising opportunity to ameliorate disease symptoms by using epigenetic-based therapy.Thus,beyond helping to understand disease biology,clinical epigenetics is being incorporated into patient management in oncology,as well as being explored for clinical applicability for other human pathologies such as neurological and infectious diseases and immune system disorders.1Cancer Epigenetics and Biology Program(PEBC),Bellvitge Biomedical Biomedical Research Institute(IDIBELL),Barcelona,Catalonia,Spain.2Centro de Investigacin Biomdica en Red Cncer(CIBERONC),Madrid,Spain.3Josep Carreras Leukaemia Research Institute,Badalona,Barcelona,Catalonia,Spain.4Physiological Sciences Department,School of Medicine and Health Sciences,University of Barcelona,Barcelona,Catalonia,Spain.5Instituci Catalana de Recerca i Estudis Avanats(ICREA),Barcelona,Catalonia,Spain.*e-mail:mberdasco idibell.cat;mestelleridibell.cathttps:/doi.org/10.1038/s41576-018-0074-2 EPIGENETICSREVIEWSNature reviews|Geneticsinto the current limitations and future challenges for fully translating epigenetics to the clinic.Epigenetic biomarkers:principles and methodsEpigenetic biomarkers are highly attractive options in clin-ical practice for several reasons8.First,they represent a simple method for associating molecular markers with contributions of lifestyle and environment in disease.This added layer of information represents an advan-tage over genetic biomarkers based exclusively on DNA sequence alterations9.Second,the technical stability of the epigenetic marks,especially DNA methylation,is also remarkable10.DNA-based biomarkers are more stable than RNA-based tests,and they can be studied without special handling requirements.However,epi-genetic biomarkers are presumably intermediate in stability between DNA-sequence-based biomarkers and RNA-based biomarkers.Furthermore,epigenetic modifications are quite stable in both fluids and tis-sue specimens commonly used in clinical practice.Interestingly,although methods for nucleic acid(DNA or RNA)extraction are continuously being improved,owing to greater amplifiability and stability the starting amount of sample required for the DNA-based analysis is frequently orders of magnitude lower than RNA or protein analysis11.Third,epigenetic modifications can be detected in all genomic contexts,not only in coding regions,and at early stages of diseases.This capability is not the case for RNA-based or protein-based biomarkers because RNAs or proteins expressed at low levels are dif-ficult to detect,and in most cases they are quantified in later stages of disease progression.In any case,it should also be noted that DNA detection in low-abundance populations of specific cells or as circulating cell-free DNA(cfDNA)is also complex.Furthermore,epigenetic marks have been shown to be robust,sensitive and measurable across individuals and populations.This aspect confers an advantage to epigenetic biomarkers with respect to RNA-based biomarkers,as epigenetic-based panels may be attainable with relatively small sets of well-defined sequences,whereas RNA-based panels are focused on large sets of transcript signatures11.As occurs with many molecular biomarkers(for example,those based on genomic,proteomic or metabo-lomic analysis),progress from preclinical observations to clinical translation is a slow process,with only a few biomarkers overcoming the transition.Studies with sam-ples that are limited in number or heterogeneous,as well as a lack of validation among cohorts,are principal pit-falls in biomarker designs.Epigenetic studies are not an exception and,despite the potential of epigenetic-based biomarkers,some challenges need to be solved.First,the biological dynamism of epigenetic alterations is a double-edged sword:it confers flexible adaptation of the genome to environmental signals,but it implicates strong cell-type dependence.Epigenetic states appear to systematically change during normal development and differentiation12,and this variation should be taken into account when samples from different sources are com-pared.Second,cost-effectiveness needs to be maintained to a practical margin.Detection methods should be simple without the requirement of complex handling or high specialization of personnel.Laborious techniques such as those used for histone modification profiling(for example,chromatin immunoprecipitation(ChIP),radio-active assays or enzyme-coupled assays)or for chromatin conformation studies(assay for transposase-accessible chromatin using sequencing(ATAC-seq)or chromatin conformation capture(3C)-derived assays)cannot easily be incorporated into clinical practice and thus are used only for research13.Thus far,detection of CpG methylation at specific loci is the most successful epigenetic biomarker.An inter-esting paper published by the BLUEPRINT consortium,an initiative of the International Human Epigenome Consortium(IHEC),compared the performance of all widely used assays for locus-specific DNA methylation analysis(n=21)across 48 genomic regions,as well as global DNA methylation levels(n=6)14.In the study,32 reference DNA samples from diverse cancer and normal tissue types were analysed in 18 laboratories.The main conclusion was that absolute DNA methyl-ation assays,which provide quantitative measures of DNA methylation at single-CpG resolution,are robust methods to validate DNA methylation differences found in large cohorts(and thus these assays have potential utility for DNA methylation biomarker detection).Specifically,amplicon bisulfite sequencing and bisulfite pyrosequencing showed the best all-round performance,although selection of the specific method depends on the number of regions studied in each sample14.Detection of changes in CpG methylation in sam-ples obtained from non-invasiveliquid biopsy samples,including blood cells,saliva,semen or urine,is a power-ful potential application of epigenetics in clinical prac-tice8(Fig.1).The detection of microRNAs(miRNAs)and exosomes in circulating blood is also particularly Box 1|epigenetic changes in the intrauterine environment and postnatal healthLifestyle and environmental conditions of the mother during early pregnancy may have long-term effects on the health of the offspring191.the molecular mechanisms of this maternal influence are thought to be mediated by fetal programming involving epigenetic changes that alter the expression of genes linked to diseases,a rationale in accordance with the Developmental Origins of Health and Disease(DOHaD)hypothesis192.Most of the evidence has been generated using animal models,as scarce epidemiological studies on humans can be found.the pioneering work linking human epigenetic deregulation with maternal lifestyle was performed on a cohort of individuals prenatally exposed to famine during the Dutch Hunger winter,who six decades later showed less DNa methylation of the imprinted insulin-like growth factor 2(IGF2)gene compared with non-famine intrauterine exposed controls193.Diet was only the tip of the iceberg,and soon several studies on the maternal exposure during pregnancy to chemical stressors(for example,arsenic,mercury or lead),endocrine disruptors(for example,bisphenol a)and non-healthy habits such as smoking or alcohol consumption demonstrated changes in the CpG methylation levels and histone modifications of the offspring191.Most interestingly,this intrauterine exposure can be associated with specific epigenetic patterns and increased disease risk in postnatal and adult life.as some examples,maternal smoking can alter the epigenetic state of genes related to immune function and be associated with later asthma or allergies in offspring194,maternal alcohol consumption is associated with increased risk of mental disorders in offspring195 and maternal high-calorie diet is associated with the development of metabolic diseases in offspring196.the great challenge is in deciphering how lifestyle and environmental conditions of the mother during pregnancy affect the epigenome of the offspring,and such knowledge may contribute to devising disease prevention strategies to reduce the risk of postnatal diseases.EpigeneticsThe study of heritable changes in gene function that do not involve changes in the DNA sequence.Epigenetic mechanisms include the covalent modifications of DNA and histones.EpigenomeThe complete set of epigenetic modifications across an individuals entire genome.EpidrugsSmall-molecule inhibitors that target either the epigenome or an enzyme with epigenetic activity.They are classified according to theirrespective target enzymesand include the following:DNA methyltransferase inhibitors(DNMTi),histone acetyltransferase inhibitors(HATi),histone methyltransferase inhibitors(HMTi),histone demethylase inhibitors(HDMi),histone deacetylase inhibitors(HDACi)and bromodomain but is not within the scope of this Review.Despite their current usefulness,the sensitivity and repro-ducibility of the liquid-biopsy-based methods still need to be improved,especially in those applications working with low cell numbers.Another challenge in epigenetic profiling is to deci-pher the specific patterns of a particular cell type in samples with heterogeneous cellular composition.This challenge is substantial in diseases such as cancer,which features mixed populations of normal and tumour cells17,or in immune-associated diseases,in which blood cell composition varies widely among patients18.Circadian fluctuations can also add variability in cellular composi-tion19.An additional caveat in the use of epigenetic bio-markers is the existence of recurrent epigenetic changes at specific loci.These epigenetic hot spots could lead to LymphocytescfDNAcfnucleosomesFrozenFreshFFPEStoolSalivaUrineNon-invasive sampleAdipose tissueSkeletal muscleBiomarker useEpigenetic measurementtechniquesBMP3 and NDRG4 methylation(CRC)30VIM methylation(CRC)33TWIST1 and OTX1 methylation(bladder cancer)37CACACACACADDDEPIMMUNE(methylation panel)(lung cancer)42SHOX2 and PTGER4 methylation(lung cancer)34miRNA expression(AD)69DPPGC1A methylation(T2D)125Histone marks at mitocondrial-function genes(T2D)129RRCerebrospinalfluidPleuraleffusionInvasive liquid biopsyBloodPrimary tissue biopsy(Surgical surplus)CACAGWAGWACAGWACACAInvasive biopsyMMGWACAGWACAGWACAGWACATSNCA methylation(PD)78APP and BACE1 methylation(AD)83,84BDNF methylation(depression,autism)69,209MBP methylation(MS)113SEPT9 methylation(CRC)31,32DDIL10 and IL1R2 methylation(SLE)108DiagnosisPrognosisResponse to therapyDPRisk of diseaseTRMonitoring lifestyle interventionGenome-wide approachesCandidate approachesMPTPDPEPICUP(methylation panel)(CUP)48GSTP1 methylation(prostate cancer)36MGMT me