GENETICS IN ENDOCRINOLOGY Genetic Diagnosis Of . - Bioscientifica

ReviewL Persani and othersNGS in endocrine diseases179:3R114Table 1 Principal applications, advantages and limitations of the three different next-generation sequencing (NGS) strategies.NGS strategyPrincipal applicationWhole genomesequencing (WGS)Studies of population Identification of SNVs, indels, SV andgenomicsCNVs in coding and non-coding regions PCR amplification not required, reducingthe potential of GC bias More consistent and uniform coveragecompared to WES or Targeted NGS A lower average read depth is requiredto achieve the same coverage as WES No impact by sequencing read length Performance unaffected by capturing oramplification procedures Specific protocols of enrichment are notrequired Complete information on the geneticvariability of each sampleIdentification of new Identification of SNVs, indels, SV andcausative genesCNVs in coding regions Reduction of the cost in comparison withWGS Reduced number of identifiedgenetic variants with a reduction ofresources needed for their storage andinterpretation Increased number of samples analysed More suitable than WGS for clinicalapplicationsAnalysis of known Most suitable for clinical applicationscausative genes Higher coverage and sequencing depththan WES Customizable for different samples types,e.g. formalin-fixed paraffin-embeddedtissues, cell free or circulating tumouralDNA, degraded samples Highest number of samples Reduced computational and storageresources Lower testing costs than WGS and WES Genetic variability is determined only forselected genes thus reducing the risk ofincidental findingsEuropean Journal of EndocrinologyWhole exomesequencing (WES)Targeted NGSAdvantagesbe induced by different mechanisms. In such context,the identification of causal variants generally requireslarger sample sizes than classic monogenic diseasesor affected families with multiple generations (15). Incomplex diseases, there is an extreme heterogeneityin both the clinical spectrum and underlyingmechanisms: individuals with similar phenotypesmay involve different causal variants from the samegene or multiple variants in different genes actingwithin the disease pathway(s). Conversely, patientswith the same causal genetic factor may manifest avariable phenotype due to incomplete penetrance, asa consequence of the interaction with other genetic,epigenetic or environmental modifying factors. Inthis field, the characterization of genetic aetiology canbecome difficult.Limitations High cost Low number of samples Huge amount of genetic variantsidentified Greater effort in the datainterpretation Limited applications in routinediagnostics Assessment of genetic variabilityonly in exons Heterogeneous coverageinfluenced by library preparationprocedures High risk of genetic incidentalfindings Genetic analysis restricted toselected regions Variable and heterogeneouscoverage Problematic design of probes Biases of library preparation:false-positive variants due to PCRduplicates or false negative resultsdue to allelic drop out Difficult identification of CNVsA large survey of human genetic variation (16) showsthat rare variants represent the great part ( 70–80%) ofgenetic variability, and the UK10K project identifiedmore than 42 M SNVs over 3781 subjects (SNVs, 34.2 Mrare and 2.2 M low frequency). Statistical geneticsconsiderations of rare variant association analysis havebeen the focus of intensive method development over thelast few years in the field of complex traits studies. Evenif there is a substantial contribution from rare variants,it remains challenging to detect rare variant effects dueto low statistical power. Deep WGS of large numbersof individuals would represent the most informativestrategy for association studies of complex traits anddiseases. However, large-scale WGS is generally unfeasiblein the field of classical epidemiological designs, suchas case–control and cohort studies, because of the highwww.eje-online.orgDownloaded from Bioscientifica.com at 05/17/2022 12:43:34AMvia free access

European Journal of EndocrinologyReviewL Persani and otherscost. Several less costly sequencing strategies have beenproposed and used in common traits studies, such aslow-depth WGS, WES and targeted NGS. Many statisticalmethods have been then proposed to increase the signalor reduce the noise in testing variant-disease associationusing sequencing data. These statistical methods canbe classified into three categories: (i) the BURDEN test(17, 18), (ii) the sequence kernel association test (SKAT)(19) and (iii) the P values combination methods (20).Once a gene/region emerged as significantly associated,the next important step is to identify rare causal variantswithin these regions/genes. Identifying a small numberof rare causal variants that contribute to complexdiseases has become a major focus of investigation. Eachof the above gene-based methods reports a P value forthe association of multiple rare variants and a specificphenotypic trait. However, the following identification ofa small proportion of truly causal variants is an even moredifficult challenge and often need other approaches suchas in vitro or in vivo functional studies.Sequencing errors and confirmationof resultsMore than 200 000 genomes and an even higher numberof exomes have been sequenced to date. It is still widelyestablished that variants found using NGS should bevalidated with the current ‘gold standard’ for DNAsequencing, Sanger sequencing (21), though severalreports suggest that NGS results are at least as accurateor in some cases more accurate than Sanger sequencing(22, 23). Massively parallel sequencing technologies haverevolutionized medical genetics, however, also NGS isprone to both negative and positive results. Problems maybe generated during library preparation procedure, PCRartefacts for example, can introduce false-positive resultsin capturing-based libraries, while amplicon-based librariesare prone to allelic dropout problems due to presenceof variation in the sequence that produce the selectiveamplification of a single allele. Errors can also be introducedduring the bioinformatics analysis, extended insertions ordeletions for example can be missed. To date, there are novalidated procedures to detect sequencing errors.NGS for discovery of novelpathogenic mechanismsConventionally, the diagnostic approach to endocrinediseases was based on physical findings, biochemicalNGS in endocrine diseases179:3R115testing and imaging analysis, followed, secondly bymolecular analysis. Until recently, DNA samples wereanalysed following a phenotype-driven strategy andusing Sanger sequencing method of the coding regions ofone candidate gene at a time, an operation that requiresseveral weeks for a response.The scenario has dramatically changed with theintroduction of NGS. The outcomes, related consequencesand risks are summarized in Table 2. With the NGSintroduction, the time needed to systematically sequencea set of candidate genes has decreased from several weeksto few days. This finally has progressively brought downthe costs for a massive parallel high-throughput DNAsequencing and a comprehensive genetic diagnosis.However, this advancement becomes possible with theintroduction of new key-personnel for the design of NGSexperiment and bioinformatics data analysis.An increasing number of scientific reports ( 600)using NGS for endocrine disease investigations werepublished in peer-reviewed journals since 2009 (Fig. 2).Here, we selected some examples showing how NGSappeared successful in identifying new causative genes orpathogenic mechanisms.One of these studies is on familial central precociouspuberty (CPP) (24). The Brazilian authors performeda WES in 40 members of 15 families affected by CPP ofunknown aetiology. NGS analysis identified MKRN3 geneas a potential causative candidate gene for the disease. Theimportant role of MKRN3 in human puberty initiation wasreinforced by large genome-wide studies involving womenof European descent from 57 studies (25). Subsequently,several studies confirmed that mutations in the paternallyexpressed imprinted MKRN3 gene constitute a majorgenetic cause of heritable or apparently sporadic CPP(26, 27). Very recently, a genomic defect in DLK1, a geneencoding a ligand of the Notch receptors, was discoveredby NGS as being also associated with isolated familial CPP(28). Interestingly, MKRN3 and DLK1 are both paternallyexpressed imprinted genes. These findings suggest a roleof genomic imprinting in regulating the timing of humanpuberty.Importantly, MKRN3 falls in the locus on chromosome15 that is known to be associated with andneurological dysfunctions associated with severe obesitydue to a complete resistance to satiety signals), whereasDLK1 had been previously implicated in Temple syndrome(including precocious puberty with intrauterine growthretardation, postnatal short stature, hypotonia, smallhands and mild facial dysmorphisms) (28). Therefore,www.eje-online.orgDownloaded from Bioscientifica.com at 05/17/2022 12:43:34AMvia free access

ReviewL Persani and othersNGS in endocrine diseases179:3R116Table 2 Novel scenarios opened by the NGS approach offering massive parallel high-throughput DNA sequencing for researchor diagnosis of endocrine conditions.NGS aftereffects (ref.)Comments Novel personnel requirements bioinformatics forexperiment planning and interpretation of results Potential explanation of patients previously classified asphenocopiesIdentification of novel causative genes (24, 28, 40, 46, 71, 76) Expanded understanding of endocrine conditions Novel pathogenic mechanismsGenetic classification of endocrine conditions (44, 50, 51, 52, Precision medicine targeted therapies and management,53, 54, 57, 61, 63, 64, 67, 68, 72, 75)and accurate prognosisIdentification of multigenic involvement (30, 31, 32, 33, 34, Novel pathogenic mechanisms38, 41, 42) Possible explanation for the variable expressivity andpenetrance of certain candidate gene variationsGenetic heterogeneity of one condition (34, 36, 38, 42, 43, 44, Expanded understanding of endocrine conditions57, 58, 59, 75) Novel pathogenic mechanisms Allelic quantification more defined prognosis for endocrinetumoursOne gene associated to multiple clinical conditions Role of gene modifiers(26, 28, 29, 31, 37, 38, 40) Expanded understanding of gene function Novel pathogenic mechanisms Challenging genetic counsellingDefined frequency of gene variants (46, 57, 61, 62, 63, 64, 67, Classification of gene variants68, 77) Improved genetic counsellingUnexpected variants in genes unrelated to the investigated Ethical issuescondition (incidental findings) (81) Challenging genetic counselling Adequate information for patient’s consentIdentification of variants of unknown significance (78, 79, 80) Adequate bioinformatic and clinical classification Challenging genetic counsellingEuropean Journal of EndocrinologyReduced time and costs and high accuracy for geneticdiagnosis (1, 2, 3, 39, 78, 79)one additional consequence of the NGS approach is theunexpected discovery of the association between a singlegene and two apparently distinct disorders. Anotherexample of this kind concerns the JAG1 gene that encodesanother ligand of the Notch receptors and was originallylinked to Alagille syndrome (a disorder characterized bycholestasis, heart malformations, together with eye, facialand skeletal abnormalities, and by an extremely variableexpressivity and penetrance of the heterozygous JAG1mutations) (29). By an NGS approach, we observed thatmonoallelic JAG1 variants are frequently associated withrare variants affecting thyroid specific genes in severalnewborns affected with congenital hypothyroidism (CH),another disorder characterized by variable expressivityand penetrance of the candidate gene defects (30). Arole for JAG1 in thyroid development was supported byin vivo studies in the zebrafish model, and it is thereforepossible that JAG1 minor alleles could contribute toCH pathogenesis by acting as genetic modifiers andamplifying the partial loss of function associated withmonoallelic defects in thyroid-specific genes (31).Our NGS data obtained in a large Italian cohort ofCH patients and in an ethnicity-matched population areconsistent with an oligogenic model of CH pathogenesis(30). We found several variants of 11 genes with arelevant role in thyroid morphogenesis or function thathave a rare or low frequency in the general population,but a significant enrichment in the CH population.When expressed alone in the first-degree relatives of CHpatients, these rare variants were associated with minorFigure 2Timeline of articles in PubMed with the key words: Nextgeneration sequencing AND endocrine disease (updatedMarch 2018).www.eje-online.orgDownloaded from Bioscientifica.com at 05/17/2022 12:43:34AMvia free access

European Journal of EndocrinologyReviewL Persani and othersthyroid defects, whereas the variable combinations of 2minor alleles were unexpectedly found in about 25% ofCH patients. Other groups obtained similar NGS findingsin smaller CH cohorts (32, 33). Such an oligogenic originof CH provides a suitable explanation for the frequentsporadic appearance of CH in the population. In all theseNGS studies targeted to a panel of known CH candidategenes, the frequency of variations found by theirsystematic and unsupervised analyses was unexpectedlyhigher that the rate of positive findings in previousstudies performed by Sanger in phenotypically selectedCH cohorts (5–20% depending upon the investigatedphenotype). This might be explained both by an increasedsensitivity of NGS protocols in the genetic variationdetection as well as by the involvement of morphogeneticor functional genes independently of the observed CHphenotype. Noteworthy, the number of cases remainingunexplained after these targeted NGS analyses falls downbelow 40% thus representing a significant improvementfor the genetic counselling in clinical practice (34).In other cases, NGS allowed to link one gene to twoendocrine conditions that were previously consideredcompletely distinct. Izumi et al. identified by NGS a SOX3polyalanine deletion in a patient with normosmic isolatedhypogonadotropic hypogonadism (IHH) (35). This findingexpanded the phenotypic spectrum of SOX3 polyalaninedeletion to include IHH without other pituitary hormonedeficiencies or mental retardation (36). Moreover, theydetected a WDR11 splice-site mutation in a patientwith multiple pituitary hormone deficiency (MPHD)(35). These data indicate that WDR11, a gene known tobe involved in normosmic IHH, can also cause MPHDs.Similar results had been previously reported for otherIHH and Kallmann syndrome (KS: IHH with anosmia)genes, such as PROKR2 or FGF8 and FGFR1 that have beendescribed in MPHDs also (37). The clinical presentationof heterozygous defects in candidate genes for IHH/KS is known to be highly variable among and withinthe same families which, in combination with the 30known candidate genes (38), makes the phenotype-drivengenetic analyses very complicated and expensive, as wellas disturbing and time consuming for the patients (39).The panel of candidate genes was further expanded by anNGS study conducted on 261 genes (known to be involvedin hypothalamic, pituitary, and/or olfactory pathways, orsuggested by chromosome rearrangements) that identified18 new potential candidate genes for IHH/KS (40). Insuch heterogeneous-related conditions (IHH, KS, MPHDs)associated with a high number of candidate genes, theapplication of NGS protocols is now unavoidable. IndeedNGS in endocrine diseases179:3R117the systematic NGS analyses will help to define the exactfrequency of involvement of single genes and the clinicalimpact of oligogenic involvement (41).This picture is very similar to that now seen in differentclinical settings, such as the primary ovarian insufficiency(POI) or the disorders of sex development (DSD) (42, 43).The genetic diagnosis in POI patients may become usefulfor the preservation of fertility in the affected families orin the identification of patients with higher chances toobtain a fertilizable egg or at higher risk of extra-ovariandefects (44). The Vilain Lab conducted two studies onDSD patients (44, 45). In the first, they proposed a novelprocedure, reversing the order of the diagnostic endocrinesteps and starting from the genetic analysis. By thisapproach, they claim the possibility to eliminate nonindicated clinical tests, sparing the patient unnecessarystress and saving healthcare system’s resources. Theydesigned a targeted panel sequencing of 35 DSD genesthat revealed genetic defects in two out of seven patientsnot previously diagnosed and confirmed the diagnosis inanother seven patients with known genetic causes (45).In the second study, they performed a WES expandingthe list of genes to analyse 64 candidates. They reacheda diagnostic yield of 35% in 40 patients with 46,XY DSDwho had not previously received a genetic diagnosis (46).The application of NGS recently expanded the viewon the mechanisms underlying tumour formation.Until recently, activating mutations in the TSH receptor(TSHR) and in GNAS represented the principal causeof autonomous thyroid adenomas (ATAs) and wereconsidered oncogenes sufficient to induce autonomousfunction and growth. In a study conducted on severalATAs by WES, we found that a relevant fraction ofATAs carry, beyond the well-known TSHR and GNASvariants, also a recurrent hot-spot mutation in EZH1, akey gene involved in the epigenetic regulation of celldifferentiation and proliferation. Interestingly, the EZH1variant was found to be associated with the TSHR or GNASvariants suggesting a 2-hit model for the pathogenesis ofthese benign tumours, whereby constitutive activation ofthe cAMP pathway and EZH1 mutations may cooperate toinduce the hyperproliferation of thyroid cells (47).NGS in endocrine tumours: role indifferential diagnosis and predictionof outcomeIn the last 10 years, NGS technology led to a betterknowledge in all human cancers, including endocrine,www.eje-online.orgDownloaded from Bioscientifica.com at 05/17/2022 12:43:34AMvia free access

European Journal of EndocrinologyReviewL Persani and otherssporadic and familial tumours. More importantly than inother clinical conditions, these new genetic knowledgewill become more and more useful for clinicians in thepersonalized management of patients.Cancer is a heterogeneous disease harbouring differentsubclonal cell populations that can be discriminated bytheir DNA mutations. The genome study of a cancer canhelp to better identify its heterogeneity and to supportthe clinicians in the choice of a more effective treatmentfor each patient. For this reason, NGS applicationsare becoming an integral part of the clinical routinediagnostics in different endocrine cancers.In thyroid tumours, NGS represents the gold standardtechnique for the pre-surgical molecular diagnosis ofthyroid nodules and for the molecular profiling ofthyroid carcinoma. Fine-needle aspiration cytology is thestandard pre-operative tool for thyroid nodule diagnosis,however, up to 25–30% of the samples are classified asindeterminate (48). In these cases, the guidelines of theAmerican Thyroid Association recommend moleculartesting in order to better identify malignant samples andto plan patient’s management (49). In order to improvediagnosis and optimize the management of thyroidnodules with indeterminate cytological diagnosis,Nikiforov et al. developed a targeted NGS test, ThyroSeqv2, which includes analyses of point mutations, genefusions and abnormal gene expression in 56 thyroidrelated genes (50). Subsequently, various groups validatedthe test (51, 52) that was very recently expanded to include112 genes, the ThyroSeq v3 (53). This panel now includesalso several genes whose role in thyroid carcinogenesisremains elusive (e.g. TSHR), thus making indefinite theinterpretation of several potential variations. However,the integration of the genetic and cytology results iscruci

genomes are sequenced at low depth and variants are simultaneously called across all samples. Variant calls on individual low-depth genomes have a high false-positive rate, but this problem is overcome by combining results across a high number of samples (7, 8, 9). When the aim of the study is to identify new genes