Abstract : Current prenatal diagnosis depends on invasive procedures and is thus offered only to high-risk pregnancies. Development of non-invasive prenatal diagnosis (NIPD) would change the risk-benefit ratio and make it likely that more women would benefit from prenatal testing. Scientists have documented the presence of rare fetal cells in maternal blood and envisioned targeting them with specific markers and their use in NIPD. Considering their extremely low frequency in maternal blood, fetal cells have been difficult to retrieve and use in clinical practice. Therefore, there is a pressing need for systematic sequential studies to evaluate their feasibility in NIPD. Generally, detection of rare cells within a large cell population carries great potentialities for the prospects of cancer management and NIPD. Manual scanning is very cumbersome and time-consuming Therefore; the first part of our project was, dedicated to the optimization of an effective strategy to evaluate retrieval of rare cells. We have developed a way of distributing a controlled number of target cells among hundreds of thousands of other cells on microscope slides. This strategy allows the precise evaluation of the retrieval of rare events and the comparizon of the efficacy of different techniques and enrichment approaches by knowing the definite number and locations of target cells on the slides. Furthermore, it allows the evaluation of hybridization of missed events. We have also developed a robust custom-made detection algorithm for rare cells using the MetaSystems automated platform and have used this strategy in the validation of manual and automatic scanning of 60 slides with a pre-defined number of rare male cells among a pure population of female cells using XY-FISH. Consequently, we tested the developed classifier for the detection of real fetal cells from maternal blood in both normal and aneuploid pregnancies with Down syndrome. We further evaluated the number of fetal cells with different methods of enrichments in the first and second trimesters. The data collected confirmed the early presence of fetal cells in all of the pregnancies tested and their frequencies were higher in cases of aneuploidies. Fetal cells are in a state of dynamic change throughout the pregnancy. Higher numbers of these cells can be obtained by optimizing the harvest time and methods of enrichment. We found that automatic scanning is more sensitive and reliable than manual detection. Furthermore, it alleviates the burden of scanning large numbers of cells and thus is more suitable for clinical application. We also demonstrated the feasibility of using rare cells in NIPD. Five microdissected amniotic fetal cells from 26 cases of normal and aneuploid pregnancies were quite enough to provide accurate NIPD through using whole genome amplification coupled with QF-PCR. Our findings laid the ground for the use of rare fetal cells in maternal blood for NIPD. // Résumé : Le diagnostic prénatal résulte encore aujourd’hui de procédures invasives, qui présentent des risques pour la grossesse. Le développement du diagnostic prénatal non-invasif (DPNI) changerait le rapport risque : bénéfice, rendant le diagnostic prénatal plus intéressant pour les femmes enceintes. Plusieurs chercheurs ont montré la présence de cellules fœtales dans le sang maternel et des travaux ont été entrepris afin de les cibler et de les utiliser éventuellement en DPNI. Toutefois, la faible concentration des cellules fœtales dans le sang maternel réduit les possibilités d’isolement ainsi que celles de leur utilisation en clinique. Un autre aspect technique du DPNI, le balayage manuel, est très laborieux, surtout en terme de temps technique. Il y a donc un besoin certain pour des études approfondies afin d’évaluer et d’améliorer la faisabilité du DPNI. La détection d’évènements rares dans une grande population cellulaire offre un potentiel pour le diagnostic en oncologie mais aussi en diagnostic prénatal. Dans cette thèse, la première étude était dédiée à l’optimisation d’une stratégie pour détecter les cellules rares. Nous avons développé une méthode d’étalement sur lame d’un nombre précis de cellules cibles parmi des centaines de milliers de cellules. Cette stratégie a permis d’évaluer le taux de détection d’évènements rares et de comparer l’efficacité des techniques d’enrichissement en connaissant le nombre exact et la localisation de cellules cibles sur les lames. De plus, il a été possible d’évaluer les problèmes d’hybridation des évènements manqués. Nous avons, par la suite, développé un algorithme robuste pour la détection de cellules rares en utilisant la plateforme de microscopie automatisée MetaSystems et utilisé cette approche dans la validation des balayages manuel et automatique d’un nombre précis de cellules mâles parmi une large population de cellules femelles marquées avec la technique FISH. Nous avons testé ce classificateur avec des échantillons de sang de femmes enceintes de grossesses normales et aneuploïdes et évalué la fréquence de cellules fœtales isolées par différentes méthodes d’enrichissement au cours des premier et second trimestres de grossesse. Les données accumulées ont confirmé la présence de cellules fœtales chez toutes les grossesses et leur fréquence plus élevée dans les grossesses aneuploïdes. Le nombre de cellules fœtales est dynamique tout au long de la grossesse. De plus, un nombre plus élevé de cellules fœtales peut être obtenu en optimisant le moment du prélèvement et les méthodes d’enrichissement. De plus, le balayage automatique s’est avéré plus sensible et constant que le balayage manuel, ce qui permet de balayer un grand nombre de cellules et devient plus approprié pour une application clinique. Nous avons aussi montré la faisabilité d’utiliser des cellules fœtales dans le cadre du DPNI. Cinq cellules amniotiques microdisséquées, provenant de grossesses normales et aneuploïdes, ont suffi pour poser un diagnostic prénatal par une combinaison de l’amplification du génome complet et de la technique QF-PCR (réaction quantitative en fluorescence d’amplification entraînée par une polymérase) permettant la détection d’anomalies chromosomiques. Nos résultats ouvrent la voie à l’utilisation de cellules fœtales dans le sang maternel pour le DPNI.

1997年，胎兒DNA被首次證實存在於母體血漿中。這一發現促進了無創產前診斷技術的發展。由於孕婦血漿中含大量來自母體的背景DNA，這給針對胎兒特異性DNA序列以外的產前診斷變得有挑戰性。近期開發的大規模平行測序技術把DNA定量精度提升到了一個空前的水平。本團隊已證實這一技術可應用於對胎兒染色體非整倍體和對胎兒全基因組的檢測。由於目前平行測序的費用仍相當昂貴，目標性測序技術可提高目標區域的數據比例從而降低測序成本。 / 在論文第一部分，本人論述了目標性測序在母體血漿DNA應用的可行性。本實驗採用雜交型富集技術對X染色體的外顯子進行富集。我們用平行測序比較了經由和未經富集處理的樣本。對比發現，經富集處理的樣本在目標區域的平均測序深度是未經富集處理樣本的213倍。目標區域的母體和胎兒DNA分子的富集程度相當。經富集處理後，目標區域的胎兒特異性等位基因的檢測率從3.5%提升至95.9%。 / 在論文第二部分，本人論述了目標性測序對胎兒21三體無創產前診斷的應用價值。我們對7，13，18和21號染色體上的單核苷酸多態性位點進行目標性測序。目標性測序數據顯示，在父源性21三體的樣本中，21號染色體上的胎兒特異性等位基因與共有性等位基因的比值上升約2倍。而在母源性21三體的樣本中，這一比值則下降約11%。本人採用電腦模擬實驗探討胎兒DNA濃度，有效等位基因數量和測序深度對檢測準確率的影響。 / 在論文第三部分，本人論述了目標性測序對胎兒單基因疾病無創產前診斷的應用。針對兩個需進行β地中海貧血產前診斷的家庭，我們對其β球蛋白基因進行目標性測序。我們用數字PCR技術推導了父母親β球蛋白基因區域的單倍型。經過相對性單倍型劑量分析，兩個胎兒的β地中海貧血遺傳狀況均得到了正確的推斷。其中一對夫婦位於致病區域的單倍型結構相近。 / 鑒於目標性測序技術可降低測序成本和提高目標序列的通量，其在血漿DNA的應用將有助於平行測序技術在無創性產前診斷、癌症診斷和移植監控等分子診斷學領域的發展。 / The presence of fetal DNA in the cell-free plasma of pregnant women was first reported in 1997. This discovery has facilitated the development of noninvasive prenatal diagnosis. The coexistence in maternal plasma of a minor population of fetal DNA among a major background of maternal DNA has posed challenges for extending noninvasive prenatal diagnostic applications that require analytical information beyond the detection of fetus-specific DNA sequences. The recent availability of massively parallel sequencing has enhanced the precision of DNA quantification to an unprecedented level. Our group has demonstrated the application of massively parallel sequencing in noninvasive prenatal diagnosis of chromosomal aneuploidies, as well as genome-wide fetal whole genome sequencing and mutational profiling. While the current costs of massively parallel sequencing are relatively expensive, targeted massively parallel sequencing may enhance the cost-effectiveness compared with the non-targeted approach because it increases the proportion of informative data from the regions-of-interest. / In the first part of this thesis, I have demonstrated the feasibility of targeted massively parallel sequencing in maternal plasma DNA. In this proof-of-principle study, hybridization-based target enrichment was used to enrich exons on chromosome X. Plasma DNA libraries with and without target enrichment were analyzed by massively parallel sequencing. For the targeted regions, the mean sequencing depth of the enriched samples was 213-fold higher than that of the non-enriched samples. Maternal and fetal DNA molecules were enriched to similar extents within the targeted regions. With target enrichment, the detection rate of fetus-specific alleles within the targeted regions increased from 3.5% to 95.9%. / In the second part of this thesis, I have demonstrated the potential application of targeted massively parallel sequencing of plasma DNA for noninvasive prenatal diagnosis of trisomy 21 using an allelic ratio approach. Targeted sequencing was used to enrich single nucleotide polymorphism loci on chr7, chr13, chr18 and chr21. The targeted sequencing data showed that the ratio between fetus-specific and shared alleles increased by approximately 2-fold on chr21 in a paternally-derived trisomy 21 case, and decreased by approximately 11% on chr21 for maternally-derived trisomy 21 cases. I have also used computer simulation to determine the impact of fractional fetal DNA concentration, number of informative alleles and sequencing depth on the detection accuracy. / In the third part of this thesis, I have demonstrated the feasibility of targeted massively parallel sequencing of maternal plasma DNA for noninvasive prenatal diagnosis of monogenic diseases. Targeted sequencing was used to enrich the β-globin gene region in two families undergoing prenatal diagnosis for β-thalassemia. Parental haplotypes of the β-globin gene region were deduced via digital polymerase chain reaction. Relative haplotype dosage analysis was performed successfully to determine the β-thalassemic status for the fetuses, including one family in which the parents had similar haplotype structures in the disease-causing region. / Because of its potential to save costs and increase throughput, targeted sequencing may catalyse the translation of massively parallel sequencing based molecular diagnostics into many fields, including noninvasive prenatal diagnosis, cancer diagnostics and transplantation monitoring. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Liao, Jiawei. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 147-155). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts also in Chinese. / ABSTRACT --- p.i / ACKNOWLEDGEMENTS --- p.vi / PUBLICATIONS --- p.vii / CONTRIBUTORS --- p.viii / TABLE OF CONTENTS --- p.ix / LIST OF TABLES --- p.xiii / LIST OF FIGURES --- p.xiv / LIST OF ABBREVIATIONS --- p.xvi / Chapter SECTION I : --- BACKGROUND --- p.1 / Chapter CHAPTER 1: --- Cell-free fetal DNA and targeted massively parallel sequencing --- p.2 / Chapter 1.1 --- Cell-free fetal DNA in maternal plasma --- p.2 / Chapter 1.2 --- NIPD by qualitative analysis --- p.3 / Chapter 1.2.1. --- Fetal sex assessment --- p.4 / Chapter 1.2.2. --- RHD genotyping --- p.5 / Chapter 1.2.3. --- Other implementations --- p.5 / Chapter 1.3 --- NIPD by quantitative analysis --- p.6 / Chapter 1.3.1. --- NIPD of chromosomal aneuploidies --- p.6 / Chapter 1.3.2. --- NIPD of autosomal recessive monogenic diseases --- p.8 / Chapter 1.4 --- Massively parallel sequencing of maternal plasma --- p.9 / Chapter 1.4.1. --- Massively parallel sequencing --- p.9 / Chapter 1.4.2. --- MPS-based NIPD of chromosomal aneuploidies --- p.11 / Chapter 1.4.3. --- MPS-based prenatal fetal whole-genome scanning --- p.15 / Chapter 1.5 --- Targeted massively parallel sequencing of maternal plasma --- p.19 / Chapter 1.5.1. --- Microdroplet-based PCR --- p.19 / Chapter 1.5.2. --- Molecular inversion probe --- p.20 / Chapter 1.5.3. --- On-array capture --- p.21 / Chapter 1.5.4. --- In-solution capture --- p.21 / Chapter 1.5.5. --- Implementation on plasma DNA --- p.22 / Chapter 1.6 --- Aims of this thesis --- p.29 / Chapter SECTION II : --- Feasibility of targeted MPS in maternal plasma DNA --- p.30 / Chapter CHAPTER 2: --- Targeted MPS of maternal plasma DNA permits efficient and unbiased detection of fetal alleles --- p.31 / Chapter 2.1 --- Introduction --- p.31 / Chapter 2.2 --- Methods --- p.34 / Chapter 2.2.1 --- Study participants and sample collection --- p.34 / Chapter 2.2.2 --- Sample processing and DNA extraction --- p.34 / Chapter 2.2.3 --- DNA quantification --- p.36 / Chapter 2.2.4 --- Microarray genotyping --- p.39 / Chapter 2.2.5 --- Plasma DNA library preparation --- p.39 / Chapter 2.2.6 --- Plasma DNA library validation --- p.40 / Chapter 2.2.7 --- Target enrichment --- p.44 / Chapter 2.2.8 --- Massively parallel sequencing and alignment --- p.45 / Chapter 2.3 --- Results --- p.48 / Chapter 2.3.1 --- Efficiency of target enrichment --- p.48 / Chapter 2.3.2 --- Sequence coverage within the targeted region --- p.53 / Chapter 2.3.3 --- Fetus-specific allele detection --- p.59 / Chapter 2.3.4 --- Fractional fetal DNA concentrations before and after enrichment --- p.63 / Chapter 2.4 --- Discussion --- p.66 / Chapter SECTION III : --- NIPD of trisomy 21 by targeted MPS of maternal plasma DNA --- p.71 / Chapter CHAPTER 3: --- NIPD of fetal trisomy 21 by allelic ratio analysis using targeted MPS of maternal plasma DNA --- p.72 / Chapter 3.1 --- Introduction --- p.72 / Chapter 3.2 --- Methods --- p.74 / Chapter 3.2.1 --- Study participants and sample collection --- p.74 / Chapter 3.2.2 --- Sample processing and DNA extraction --- p.74 / Chapter 3.2.3 --- Targeted MPS of plasma DNA libraries --- p.74 / Chapter 3.2.4 --- F-S ratio calculation --- p.76 / Chapter 3.2.5 --- Microarray genotyping --- p.78 / Chapter 3.2.6 --- Computer simulation --- p.78 / Chapter 3.3 --- Results --- p.80 / Chapter 3.3.1 --- Efficiency of target enrichment --- p.80 / Chapter 3.3.2 --- Estimation of fractional fetal DNA concentrations --- p.83 / Chapter 3.3.3 --- F-S ratio calculation using non-targeted sequencing data --- p.83 / Chapter 3.3.4 --- F-S ratio calculation using targeted sequencing data --- p.85 / Chapter 3.3.5 --- Computer simulation --- p.85 / Chapter 3.4 --- Discussion --- p.90 / Chapter SECTION IV : --- NIPD of monogenic diseases by targeted MPS of maternal plasma DNA --- p.94 / Chapter CHAPTER 4: --- NIPD of monogenic diseases by targeted MPS of maternal plasma: application to Beta-thalassemia --- p.95 / Chapter 4.1 --- Introduction --- p.95 / Chapter 4.2 --- Methods --- p.98 / Chapter 4.2.1 --- Sample collection and DNA extraction --- p.98 / Chapter 4.2.2 --- Microarray-based genotyping --- p.100 / Chapter 4.2.3 --- Targeted MPS of plasma DNA libraries --- p.100 / Chapter 4.2.4 --- Genotyping by Sanger sequencing --- p.103 / Chapter 4.2.5 --- Haplotyping by digital PCR --- p.105 / Chapter 4.2.6 --- RHDO analysis --- p.105 / Chapter 4.3 --- Results --- p.110 / Chapter 4.3.1 --- Effectiveness of targeted sequencing --- p.110 / Chapter 4.3.2 --- Determination of fetal HBB genotype in the first family --- p.112 / Chapter 4.3.3 --- Determination of fetal HBB genotype in the second family --- p.113 / Chapter 4.4 --- Discussion --- p.115 / Chapter SECTION V : --- CONCLUDING REMARKS --- p.120 / Chapter CHAPTER 5: --- Conclusion and future perspectives --- p.121 / Chapter 5.1 --- Targeted MPS in plasma DNA --- p.121 / Chapter 5.2 --- Targeted MPS in NIPD of chromosomal aneuploidies --- p.124 / Chapter 5.3 --- Targeted MPS in NIPD of monogenic diseases --- p.126 / Chapter Appendix I: --- Primer names and sequences for genotyping and haplotyping of βeta-globin gene cluster region --- p.128 / Chapter Appendix II: --- Primers used in PCRs and sequencing for the parents in the first family --- p.132 / Chapter Appendix III: --- Primers used in PCRs and sequencing for the parents in the second family --- p.138 / Chapter Appendix IV: --- RHDO analysis on maternal plasma DNA in the first family --- p.140 / Chapter Appendix V: --- RHDO analysis on maternal plasma DNA in the second family --- p.145 / References --- p.147