Development of Simultaneous Genomic and Transcriptomic Analysis at the Single Cell Level

Master's Thesis , 2018 , 43 Pages , Grade: 15/20

Medicine - Biomedical Engineering

In this master thesis, we want to develop and implement simultaneous genome and transcriptome analysis at the single cell level, through analyzing single cells of human embryonic stem cells and human fibroblasts. For the validation of the experiments we used (1) purified RNA & DNA mixes, (2) larger numbers of cells and (3) dilutions down to one single cell.

The ultimate goal of using such technique is to investigate each single blastomere of human embryos at the two- to eight-cell stage. We particularly want to investigate the hypothesis whether shortage of previously mentioned transcripts leads to an abnormal number of chromosomes during embryonal development.

Roughly half of the human preimplantation embryos obtained after in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) show a high frequency of chromosomal mosaicism, the event where not all cells of an embryo display an identical chromosomal composition. These abnormalities are mostly explained by anaphase lagging, non-disjunction and endoreduplication, of which anaphase lagging is the most common mechanism resulting in mosaicism.

They originate during postzygotic, mitotic divisions. Several authors hypothesize that the causes of the aberrations lie in the depletion of certain transcripts such as for the components of the chromosomal passenger complex (CPC) and the spindle attachment checkpoint (SAC) which are key regulators during cell division that correct chromosome attachment errors and prevent chromosome mis-segregations and aneuploidy.

Conversely, transcriptome analysis on whole embryos has shown the presence of these transcripts which argues against their role during the development of aneuploidies. Interestingly, different SAC and CPC expression levels were shown at the single cell level, but without correlation to chromosomal abnormalities.

TABLE OF CONTENTS

SUMMARY

SAMENVATTING

1. INTRODUCTION

1.1 ASSISTED REPRODUCTIVE TECHNOLOGIES
1.2 PREIMPLANTATION GENETIC TESTING
1.2.1 PGT FOR MONOGENIC/SINGLE GENE DEFECT (PGT-M)
1.2.2 PGT FOR ANEUPLOIDIES (PGT-A)/STRUCTURAL REARRANGEMENTS (PGT-SR)
1.2.2.1 FLUORESCENCE IN SITU HYBRIDIZATION FOR SINGLE CELL ANALYSIS
1.2.2.2 COMPREHENSIVE CHROMOSOME SCREENING
1.2.2.3 NEXT GENERATION SEQUENCING
1.3 THE ORIGIN OF CHROMOSOMAL ABNORMALITIES IN PREIMPLANTATION EMBRYOS
1.3.1 ORIGIN OF MITOTIC ANEUPLOIDIES
1.3.1.1 UNDERLYING PROCESSES LEADING TO ABNORMAL CHROMOSOME NUMBERS
1.3.1.2 MATERNAL FACTORS AFFECTING MITOTIC ANEUPLOIDIES
1.3.1.3 THE SPINDLE ASSEMBLY CHECKPOINT DURING MITOSIS
1.3.1.4 THE CHROMOSOMAL PASSENGER COMPLEX

2. OBJECTIVES OF THE MASTER THESIS 15

3. MATERIALS AND METHODS

3.1 CELL CULTURE
3.2 RNA/DNA EXTRACTION
3.3 SINGLE CELL PICK-UP
3.4 CDNA SYNTHESIS
3.5 CDNA AMPLIFICATION
3.6 PHYSICAL SEPARATION OF MRNA AND GDNA
3.7 MEASURING CDNA CONCENTRATIONS
3.8 GENOMIC DNA PRECIPITATION AND AMPLIFICATION
3.9 QUANTITATIVE REAL-TIME PCR (QRT-PCR)
3.10 PCR

4. RESULTS
4.1 VALIDATION OF THE OLIGO-DT 30VN PRIMER SET
4.1.1 BULK RNA FIBROBLASTS
4.1.2 BULK RNA HESC
4.2 VALIDATION OF THE TSO AND ISPCR PRIMER SET
4.3 IMPLEMENTATION OF OLIGO-DT 30VN LABELLED MAGNETIC BEADS
4.3.1 RNA DILUTION SERIES
4.3.2 SINGLE CELL LEVEL
4.4 VALIDATION OF THE CFTR PRIMER SET
4.4.1 DNA DILUTION
4.4.2 SINGLE CELL LEVEL
4.5 VALIDATION OF PHYSICAL SEPARATION OF MRNA AND GDNA
4.5.1 RNA/DNA MIXES
4.5.1.1 VALIDATION OF AMPLIFIED CDNA
4.5.1.2 VALIDATION OF AMPLIFIED GDNA AFTER MULTIPLE DISPLACEMENT AMPLIFICATION
4.5.2 SINGLE CELL LEVEL
4.5.2.1 VALIDATION OF AMPLIFIED CDNA
4.5.2.2 VALIDATION OF AMPLIFIED GDNA AFTER MULTIPLE DISPLACEMENT AMPLIFICATION
4.5.2.3 VALIDATION OF THE PICOPLEX WGA KIT
4.5.2.4 VALIDATION OF AMPLIFIED GDNA AFTER PICOPLEX
4.5.2.5 SHALLOW WHOLE GENOME SEQUENCING DATA

5. DISCUSSION

6. CONCLUSION

7. ACKNOWLEDGEMENTS

8. ABBREVIATIONS

9. REFERENCES

Summary

Introduction: Roughly half of the human preimplantation embryos obtained after in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) show a high frequency of chromosomal mosaicism, the event where not all cells of an embryo display an identical chromosomal composition. These abnormalities are mostly explained by anaphase lagging, non-disjunction and endoreduplication, of which anaphase lagging is the most common mechanism resulting in mosaicism. They originate during postzygotic, mitotic divisions. Several authors hypothesize that the causes of the aberrations lie in the depletion of certain transcripts such as for the components of the chromosomal passenger complex (CPC) and the spindle attachment checkpoint (SAC) which are key regulators during cell division that correct chromosome attachment errors and prevent chromosome mis-segregations and aneuploidy. Conversely, transcriptome analysis on whole embryos has shown the presence of these transcripts which argues against their role during the development of aneuploidies. Interestingly, different SAC and CPC expression levels were shown at the single cell level, but without correlation to chromosomal abnormalities.

Aim: In this master thesis, we want to develop and implement simultaneous genome and transcrip- tome analysis at the single cell level, through analyzing single cells of human embryonic stem cells and human fibroblasts. For the validation of the experiments we used (1) purified RNA & DNA mixes, (2) larger numbers of cells and (3) dilutions down to one single cell. The ultimate goal of using such technique is to investigate each single blastomere of human embryos at the two- to eight-cell stage. We particularly want to investigate the hypothesis whether shortage of previously mentioned tran- scripts leads to an abnormal number of chromosomes during embryonal development.

Materials and Methods: Single cells are collected manually into a lysis buffer. Oligo-dT-30VN labelled magnetic beads are prepared by conjugating a biotinylated oligo-dT primer to streptavidin-coupled magnetic beads. The mRNA and genomic DNA are physically separated through binding of the con- jugated beads to the poly-A tail of the mRNA. After placing the tube on a magnet plate, we collect the supernatant containing the genomic DNA while the mRNA is attached to the tube. Reverse tran- scription is performed with a SMART-seq2 based protocol. Subsequently we amplify the cDNA (IS Primers) and gDNA (Multiple Displacement Amplification or Picoplex). To test the presence of the cDNA we measure the concentrations and perform qRT-PCR. We check the presence of amplified gDNA using a Nanodrop and checking the presence of the CFTR gene by PCR. The Picoplex amplified gDNA is subsequently sequenced by shallow whole genome sequencing at the Centre for Medical Genetics (CMG). RNA-sequencing was not validated within this thesis.

Results: We validated the oligodT-primer, TSO and ISPCR primer by conversion of RNA from hESC and fibroblasts into cDNA and subsequent amplification. We observed a non-linear amplification pattern in our qRT-PCR results. We showed presence of GAPDH and pluripotency genes (SOX2 and NANOG) in hESC RNA. After validating the mRNA part of our protocol, we implemented the physical separation of mRNA and gDNA. When applying our protocol to single cells we observed different gene expression patterns in every single cell compared to the bulk RNA analysis that remains stable. We measured cDNA amounts of 12-40ng. For the gDNA part we performed either multiple displace- ment amplification (MDA) or Picoplex and measured gDNA amounts of >0.5µg. The presence of the CFTR gene after performing PCR was shown in 12 out of 18 (66.6%) samples. Shallow whole genome sequencing (WGS) after Picoplex was able to detect duplications and deletions of regions of chro- mosomes 1, 5 and 18 of a hESC line with known abnormalities.

Discussion and Conclusion: We successfully validated and implemented simultaneous genome and transcriptome analysis at the single cell level in our lab. The observed non-linear amplification pat- tern of our cDNA which is likely to occur in PCR-based methods. It is well known that bulk RNA analysis hides cell-to-cell variability, confirmed by our qRT-PCR results. We measured the expected amounts of cDNA and gDNA after amplification in our lab. PCR after MDA on the CFTR gene showed absence of amplification that is a fact in whole genome amplification methods. Shallow WGS after Picoplex detected the known chromosomal abnormalities in two single cells from an abnormal hESC line, however, one sample showed a high noise. The noise is explained by incomplete transfer of gDNA during the separation from mRNA. With this technique we can analyze each single blastomere from 2-to 8-cell cleavage stage embryos. We aim to identify novel markers for genetic health by comparing the genome with the transcriptome. This may increase the efficiency of IVF in the future.

Samenvatting

Inleiding: Ongeveer de helft van de menselijke pre-implantatie embryo's verkregen na in vitro fer- tilisatie (IVF) en intracytoplasmatische spermainjectie (ICSI) vertonen een hoge frequentie van chromosomaal mozaïcisme, waarbij niet alle cellen van een embryo dezelfde chromosomale samen- stelling vertonen. Deze afwijkingen worden meestal verklaard door anafase-lagging, non-disjunctie en endoreduplicatie, waarvan anafase lagging het meest voorkomende mechanisme is dat resulteert in mozaïcisme. Ze ontstaan tijdens postzygotische, mitotische celdelingen. Verschillende auteurs veronderstellen dat de oorzaken van de aberraties in de afwezigheid van bepaalde transcripten lig- gen zoals voor de componenten van het chromosomale passenger complex (CPC) en het spindle assembly checkpoint (SAC), die tijdens de celdeling belangrijke regulatoren zijn, verantwoordelijk voor het herstellen van fouten in de binding van chromosomen op de mitotische spindle en het voorkomen van chromosoom mis-segregatie en aneuploidie. Analyse van het transcriptoom van volledige embryo’s heeft daarentegen de aanwezigheid van deze transcripten aangetoond, wat tegen hun rol tijdens de ontwikkeling van aneuploïdieën spreekt. Verschillende SAC en CPC levels op het niveau van één enkele cel werden aangetoond, maar zonder correlatie met chromosomale afwijkin- gen.

Doel: In deze master thesis willen we simultane genoom- en transcriptoom analyse op het niveau van één enkele cel in ons laboratorium ontwikkelen en implementeren, door het analyseren van cellen van humane embryonale stamcellijnen en humane fibroblasten. Voor de validatie van de ex- perimenten gebruiken we (1) geëxtraheerde RNA/DNA mixes, (2) groter aantal cellen en (3) ver- dunningen tot één enkele cel. Het uiteindelijke doel van het gebruik van een dergelijke techniek is om elk blastomeer van het twee- tot achtcellig stadium van menselijke embryo's afzonderlijk te onderzoeken. We willen vooral de hypothese onderzoeken of een tekort aan eerder genoemde trans- cripten leidt tot een abnormaal aantal chromosomen tijdens de embryonale ontwikkeling.

Materiaal en Methoden: De cellen worden manueel in een lysis buffer gebracht. Oligo-dT-30VN ge- merkte magnetische beads worden bereid door een gebiotinyleerde oligo-dT-primer te conjugeren aan streptavidine-gekoppelde magnetische beads. Het mRNA en genomisch DNA worden fysisch gescheiden door binding van de geconjugeerde beads aan de poly-A staart van het mRNA. Nadat de buis op een magneet is geplaatst, collecteren we het supernatant dat het genomisch DNA bevat terwijl het mRNA aan de zijkant van het buis plakt. Reverse transcriptie wordt uitgevoerd met een op SMART-seq2 gebaseerd protocol. Vervolgens amplificeren we het cDNA (IS Primers) en gDNA (Multiple Displacement Amplification of Picoplex). Om de aanwezigheid van het cDNA te testen, meten we de concentraties en voeren we qRT-PCR uit. We controleren de aanwezigheid van gDNA met behulp van een Nanodrop en controleren de aanwezigheid van het CFTR-gen met PCR. Het met Picoplex geamplificeerde gDNA wordt vervolgens gesequenced door shallow sequencing van het ge- noom door het CMG. RNA-sequencing werd niet gevalideerd binnen dit proefschrift.

Resultaten: We valideerden de oligodT-primer, TSO en ISPCR-primer door omzetting van hESC en fibroblast RNA in cDNA en daaropvolgende amplificatie. We hebben een niet-lineair amplificatie pa- troon waargenomen in onze qRT-PCR-resultaten. We toonden aanwezigheid van GAPDH en pluripo- tentie genen (SOX2 en NANOG) in hESC aan. Na validatie van het mRNA deel hebben we de fysieke scheiding van mRNA en gDNA geïmplementeerd. Bij de toepassing van ons protocol op één enkele cel hebben we verschillende genexpressiepatronen in elke cel waargenomen in vergelijking met de bulk-RNA-analyse, die stabiel blijft. We maten cDNA-hoeveelheden van 12-40ng. Voor het gDNA deel voerden we ofwel MDA of Picoplex uit en hebben gDNA-hoeveelheden van >0,5µg gemeten. De aanwezigheid van het CFTR-gen na het uitvoeren van PCR werd getoond in 12 van de 18 (66,6%) stalen. Shallow sequencing van het genoom na Picoplex kon duplicaties en deleties van bepaalde regio's van chromosomen 1, 5 en 18 van een hESC lijn met gekende afwijkingen detecteren.

Discussie en Conclusie: We hebben succesvol het protocol voor simultane genoom en transcriptoom analyse op het niveau van één enkele cel geïmplementeerd. Het waargenomen niet-lineaire ampli- ficatie patroon van ons cDNA is gekend in PCR-gebaseerde methoden. Het is bekend dat bulk RNA- analyse de cel-tot-cel variabiliteit maskeert, wat onze qRT-PCR-resultaten bevestigen. We maten de verwachte hoeveelheden cDNA en gDNA na amplificatie in ons lab. PCR na MDA op het CFTR-gen toonde in zes gevallen afwezigheid van amplificatie, dat optreedt in gehele genoom amplificaties. Shallow sequencing na Picoplex kon gekende chromosomale afwijkingen detecteren. Echter was één sample van slechte kwaliteit, wat te wijten is aan onvolledige scheiding van gDNA. Met deze techniek kunnen we elk blastomeer analyseren van 2-tot 8-cellige embryo's. We willen nieuwe markers voor genetische gezondheid identificeren door het genoom te vergelijken met het transcriptoom. Dit kan de efficiëntie van IVF in de toekomst verhogen.

1. Introduction

Since 1993 it has been shown that about 50% of human preimplantation embryos obtained after in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) display chromosomal mosaicism, the event where not all cells of an embryo show an identical chromosome composition (1–3). These chromosomal abnormalities originate during postzygotic, mitotic divisions (4–7). Furthermore, re- searchers explain the mitotic chromosomal mis-segregations with anaphase-lagging, non-disjunc- tion and endoreduplication (5). Within the field of human reproduction, we need to elucidate the origin of chromosomal abnormalities in early human embryos, because it may hamper the IVF out- come substantially (8). This can lead to new genetic markers, which may improve the efficiency of IVF.

1.1 Assisted Reproductive Technologies

After the first baby was born after IVF in 1978, the number of births after assisted reproductive technologies (ART) increased, resulting in more than 5 million children worldwide (9, 10). While female infertility is an indication for IVF, male infertility is treated with ICSI that was first described in 1992 (11, 12). In women, infertility occurs due to ovulatory problems, absence of the uterus, obstruction or absence of the tubes, endometriosis, abnormalities in the level of sex hormones and age. Infertility in men is linked to poor quality of the sperm and as in women to endocrinological malfunctions. In some cases the infertility problem occurs in both partners (10).

To allow for successful ART treatment patients are treated by controlled ovarian hyperstimulation (COH). The growth of follicles is stimulated by 7-10 days injection of recombinant FSH (follicle stimulating hormone). From day 6 of stimulation a GnRH (gonadotrophin releasing hormone) antagonist isadministered daily resulting in the reversible inhibition of gonadotrophin release. This creates a neg- ative feedback loop preventing a premature LH (lu- teinizing hormone) peak. During COH a premature LH peak is likely to occur which avoids multiple fol- licle maturation (13).

During follicle stimulation, doctors check the hor- monal status and the state of the follicles. When an adequate follicle stimulation is reached the injection of hCG (human chorion gonadotrophin) completes follicle maturation. Then oocytes are picked-up through vaginal puncture and aspiration of follicles. After preparation of the sperm sample the oo- cytes are inseminated in vitro. When the sperm quality is inadequate due to severe oligozoospermia and/or severe asthenozoospermia and/or severe teratozoospermia ICSI is advised by injecting one spermatozoon into the oocyte using a micropipette (12). The embryos are transferred two to five days after IVF or ICSI into the uterus to start implantation in the endometrium. Figure 1 shows the different stages of in vitro embryonal development from day 1 to day 5. The invention of IVF also provides access to human embryos in the laboratory.

However, embryos can undergo preimplantation genetic testing (PGT) to avoid transfer of an embryo affected by a genetic disease and/or aneuploidies and/or chromosomal structural rearrangements.

1.2 Preimplantation Genetic Testing

Preimplantation Genetic Testing allows for the investigation of the DNA content of polar bodies (PB) removed from oocytes and biopsies from cleavage or blastocyst stage preimplantation embryos. PGT is divided into three subgroups: PGT for aneuploidies (PGT-A); PGT for monogenic/single gene de- fects (PGT-M); and PGT for chromosomal structural rearrangements (PGT-SR). The formerly known terms Preimplantation Genetic Diagnosis (PGD) and Preimplantation Genetic Screening (PGS) are now replaced by the term PGT (14).

The development of such PGT methods was started when Edwards and Gardner made the first effort to determine the sex of rabbit blastocysts in 1967 (15). However, PGT only became feasible when polymerase chain reaction (PCR) was invented. Handyside et al. (16) and Verlinsky et al. (17) then reported independently the implementation of PGT by analyzing single cells (blastomeres) or polar bodies of preimplantation embryos which originate from couples with a higher risk of transmitting X-linked and monogenetic diseases. Nowadays, trophectoderm (TE) biopsy followed by PGT is an established method in the IVF clinic with numerous healthy new-borns assuming that PGT is a safe and reliable technique without significant appearance of adverse effects (18, 19). PGT aimed to improve life birth rates (PGT-A) and to avoid the birth of children affected with a monogenetic dis- ease (PGT-M) or translocations (PGT-SR). If this aim in the clinic is reached, however, still needs to be proven.

1.2.1 PGT for monogenic/single gene defect (PGT-M)

PGT was initially developed for the detection of monogenic disorders present at birth such as cystic fibrosis (CF), beta-thalassemia, myotonic dystrophy, Huntington’s disease, fragile X syndrome and other X-linked diseases. Nowadays, monogenetic disorders are diagnosed with PCR-based methods that allow for the detection of over 400 different diseases (20).

Single cells for PGT are biopsied from three different stages: 1st and/or 2nd PB at day 0; one or two blastomeres at day 3; or five-ten cells of the TE at day 5. One specific problem during single-cell PCR is allele-drop-out (ADO). ADO is understood as a non-amplification of individual alleles during PCR that leads to a false homozygosity (figure 2). Heterozygous embryos then could be diagnosed as homozygous (21). Several protocols improved the technique using different cell lysis methods and increasing the denaturation temperature (22–24). To lower the impact of high ADO rates re- searchers can increase the number of loci and analyse several replicates of the same sample (25). However, working with embryos would restrict the use of replicates. Multiplex fluorescent PCR, how- ever, solved the problem of ADO (25). PCR fragments are labelled with a fluorochrome and can be detected with an automated sequencer. It combines for example the mutation and one linked marker, usually microsatellites close to the mutation. This allows for the detection of the ADO (26).

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: Allele drop-out. Heterozygous DNA samples undergo amplification. One allele is not amplified during PCR resulting into two possible homozygous outcomes: normal or affected. This leads to a wrong diagnosis of the sample.

Since ~6 picogram DNA is present in a single cell, samples contaminate rapidly. Hence, all single cell experiments should be performed under sterile conditions in a laminar flow hood as a dedicated pre-amplification area separated from a post-amplification area (27). To overcome the low amount of DNA researchers and embryologists perform whole genome amplification (WGA) before PCR to increase DNA levels. This resulted in a significant increase of the ADO ratio in all biopsied cells (24, 28, 29). Multiple displacement amplification (MDA) is a non-PCR based isothermal technique for WGA. First DNA is denatured into single strands. Subsequently f29-DNA polymerase and random hexamers are added by which a second complementary strand is synthesised. The enzyme then displaces the strand. Further polymerization of these displaced strands leads to a hyperbranched structure (25). However, after regular PCR on samples which underwent MDA researchers observed a rather high ADO ratio and preferential amplification of certain alleles (28–30). In this case ADO can also be avoided by subsequent multiplex PCR, but it is technically more challenging especially without pre-amplification.

1.2.2 PGT for aneuploidies (PGT-A)/structural rearrangements (PGT-SR)

1.2.2.1 Fluorescence in situ hybridization for single cell analysis

Structural rearrangements such as Robertsonian and reciprocal translocations form another indication for PGT. Robertsonian translocations are fusions between two acrocentric chromosomes (13, 14, 15, 21, 22 and Y) at the centromere resulting in the loss of the short arm. Reciprocal translocations develop through fusion of breaks in non-homologous chromosomes. Both translocations appear balanced or unbalanced. Balanced translocations do not affect the phenotype, whereas unbalanced translocations result in monosomies and trisomies. Carriers of balanced translocations have an increased risk of abnormal offspring, such as trisomy 21 when a Robertsonian translocation affects chromosome 21 (30, 31). More frequently, however, they are prone to subfertility resulting in repeated miscarriage or implantation failure.

With the invention of fluorescence in situ hybridization (FISH) it was possible to differ between balanced situations (normal or balanced carrier) and an unbalanced chromosome composition (figure 3). Diagnosis of every translocation case became possible with the use of both subtelomeric and centromeric probes (26).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Outcome of PGT with FISH (25). An example of a reciprocal translocation. Chromosomes are labelled with centromeric (yellow and green) and telomeric (red) probes. The blue circle shows an inter- phase nucleus from a blastomere. The first result shows a normal situation (2 dots from each color). The second is a balanced situation with exactly the same dots as the normal. This translocation causes no phenotypically changes. The third result shows an aneuploid cell. One red dot is missing, which is inter- preted as a monosomy for one part of chromosome B. The other part is trisomic for the blue chromosome. The fourth result shows another possibility for an unbalanced situation. This cell is trisomic for a part of chromosome B (3 red dots) and monosomic for a part of the blue chromosome.

For PGT-SR with FISH, polar bodies, single blastomeres from day 3 embryos and blastocysts can be analyzed (figure 3). Polar bodies, however, only provide genetic information inherited from the mother. For FISH, the single cell is spread and fixed on a microscope slide. The cell membrane and cytoplasm is removed and the nucleus is permeabilized. Denaturation of DNA leads to single strands that can hybridize with target-specific probes labelled with fluorochromes. The slides are analysed with an UV microscope. With FISH unbalanced embryos can be excluded for transfer in the IVF laboratory allowing for the birth of healthy children.

Women with advanced age (>37 years) carry embryos and oocytes with increased aneuploidy rates which is an indication for PGT-A. PGT for aneuploidies was initially done by FISH to choose only euploid embryos for transfer after ART. It was hypothezied that PGT-A with FISH increases pregnancy rates, but several RCTs rejected this (32–34). Analyzing one single blastomere is unrep- resentative for the whole embryo since cleavage stage embryos show high rates of mosaicism. Ad- ditionally, the number of chromosomes that can be analysed at the same time is limited because there are only five different fluorochromes available. That makes it not practical to analyse all chro- mosome pairs. This problem could be overcome by performing multiple rounds, but the accuracy seems to drop to 41% when five chromosomes or more are analysed (35). Another limitation is the required spreading and fixing of the nucleus on a microscope slide that could lead to chromosome loss, harm of the nucleus, overlying signals and failure during hybridization (36). Additionally, FISH is not suitable to detect segmental abnormalities (see 1.2) since the fluorochrome labelled probes only hybridise to the target (37).

Also, the use of single blastomeres from day 3 embryos has mainly shifted to TE biopsy since it was unclear how blastomere biopsy affects the embryo on short-term and children born after ART on

long-term (32, 33). Furthermore, the reliability increases through analysing 5 to 10 cells from circa 150 cells present at blastocyst stage and researchers proved that blastocyst biopsy is less harmful to the embryo than day 3 biopsy (34). In the meantime, FISH for the detection of aneuploidies has been completely replaced by next generation technologies (NGT) such as array-Comparative Genomic Hybridization (aCGH) and Next Generation Sequencing (NGS), but the detection of unbalanced chromosome translocations is still an indication for FISH (26).

1.2.2.2 Comprehensive Chromosome Screening

In contrast to FISH the use of comprehensive chromosome screening (CCS), with first CGH on met- aphase spreads of chromosomes and subsequent array CGH, allows for genome-wide aneuploidy screening. With CGH it is possible to investigate DNA copy number variations (CNV) through com- parison of the DNA content of two differently labelled samples, after hybridisation on metaphase spreads of a normal individual (38). However, CGH has a high workload and is time consuming making it less convenient for the clinic. Based on the same principle, researchers introduced array- CGH (6, 39). In contrast to CGH it compares equal amounts of the labelled test and reference sample after co-hybridization on numerous DNA probes coated in an array instead of metaphase spreads of an normal individual (40). These probes commonly consist of complementary DNA, such as oligonu- cleotides or genomic fragments generated in different types of vectors allowing for a higher resolu- tion than CGH (41). Array-CGH at the single cell level requires WGA since DNA amounts in the nanogram range are necessary.

Randomized controlled trials (RCTs) have demonstrated that pregnancy rates do not increase after PGT-A with FISH. However, CCS on every single cell of day 3 embryos revealed the entire scope of mosaicism (4–6). Certain day 3 embryos are shown to contain both normal and aneuploid cells whereas others contain only aneuploid cells with varying levels of abnormality in each cell. Capalbo et al. (33) however claim that the mosaicism levels in day 3 embryos are overvalued due to technical faults and deficient robustness of aCGH. The other authors (4–6) validated aCGH and demonstrated highly reliable results. Unravelling the exact biological consequences of mosaicism in human preim- plantation embryos will keep researchers occupied.

Single Nucleotide Polymorphisms (SNP) arrays form another approach for CCS. SNPs are variations of a single base pair in a DNA strand present in at least 1% of the population. Diseases can be caused by a single base pair change, but not all SNP lead to a pathological mutation. Every SNP consist of two alleles or base pairs. These variations are not randomly distributed, but divided as haplotype blocks. Thus, in certain regions of our DNA only one allele type is present. A section of for example five SNPs will then displays two or three of the 25 or 32 possibilities (26). There exist numerous SNPs in our genome from which 15 million are identified. However, conventional SNP arrays do not require a library for all of the 15 million SNPs due to the haplotype blocks. They use rather oligonucleotides of usually 25 base pairs complementary to the two alleles of a several hun- dred thousands of SNPs (26). It can be used for all types of PGT.

1.2.2.3 Next Generation Sequencing

NGS reaches a higher sensitivity than aCGH and SNP array for the detection of aneuploidies in human preimplantation embryos and its costs are decreasing quickly (26). This technique has a typical coverage (depth of sequence) of 30x, which means that one base pair is read 30 times. We have to mention that for PGT the use of shallow whole genome sequencing is applied with a depth of for instance 0,3x (42). This extremely reduces the costs.

Moreover, NGS sequences mutations at very low levels in contrast to first generation technologies as for instance Sanger Sequencing (43). Subsequently, several groups developed their own ap- proaches to NGS but with a common scheme: Library preparation is required before sequencing. The DNA to be sequenced is sheared into small bits allowing for binding of platform specific adaptor sequences on the 3’ and 5’ end (paired-end) (44). In numerous platforms, the signal to noise ratio is enlarged by local clonal amplification, generating clusters (45). One of these platforms is the Illumina technology (figure 4) working on a flow cell, a glass slide with different lanes: 1 lane (MiSeq), 2 lanes (HiSeq 2500) or 8 lanes (HiSeq 2000) (43). These lanes are composed of two types of oligos (forward and reverse) on which the adapter sequences bind. After hybridization on one type of the oligos, a DNA polymerase generates a complement of the bound fragment. Then the initial DNA template is denatured and washed out. Subsequently, multiple steps of clonal amplifica- tion by bridge amplification are performed. The strand folds over to the other complementary oligo, while a complementary bridge is created by DNA polymerase. When these strands are denatured, two single strands are tethered to the flow cell. For sequencing, different labelled dNTPs are incor- porated allowing for detection of the signal with an optical sensor. This process is called Sequencing- by-Synthesis. The huge data volume needs to be aligned to reference sequences by using software programs.

Zhang et al. (46) analysed single cells derived from blastocysts with massive parallel sequencing (MPS), a general term for whole genome sequencing (WGS) approaches, using the HiSeq 2000 platform. They detected aneuploidies and CNVs larger than 1Mb which was confirmed by SNP arrays. Furthermore, NGS reached 99.63% sensitivity and 97,71% specificity. Such high specificity and sensitivity rates after single cell NGS were also shown by others as well as the ability to diagnose segmental abnormalities (47). Following validation, several RCTs show that these techniques for PGT-A are reliable, safe and effective (34, 48, 49). In turn, some authors criticize the study size and the methods to investigate the study results (50, 51). Additionally, it has been shown that mosaic blastocysts are capable to implant leading to healthy offspring (8, 52).

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Illumina sequencing technology. (1) Adapter sequences attach to the flowcell. (2) After bridge amplification, forward and reverse strands are tethered to the flowcell. (3) Denaturation allows for further amplification and cluster formation. Sequencing by synthesis of the forward and reverse strands is performed by detecting the fluorescence signals of different dNTPs. Figure from: https://emea.illumina.com/science/technology/next-generation-sequencing/sequencing-technol- ogy.html?langsel=/be/

1.3 The origin of chromosomal abnormalities in preimplantation embryos

A meta-analysis of van Echten-Arends et al. (2) showed that 73% of the human preimplantation embryos generated after IVF are chromosomally mosaic of which diploid-aneuploid mosaicism is the most common state (59%) followed by aneuploid mosaicism (15%). These chromosomal abnormal- ities were shown to arise during early embryonic development (6, 7, 53). Mitotic errors occur during all stages of embryogenesis, but appear more likely within the first three divisions (5, 54, 55). The two key mechanisms leading to mitotic aneuploidies are anaphase lagging and non-disjunction (55–59). Meiotic aneuploidies were observed at the first meiotic division predominantly in females (60). The underlying mechanisms of meiotic errors are predivision of chromosomes at metaphase I and premature segregation of sister chromatids which are related to advanced maternal age and the number of crossing-overs. For detailed review see (60).

Beside numerical aberrations the use of CCS revealed that structural abnormalities also appear dur- ing the first cleavages of the embryo which results in mosaicism of specific chromosomal segments (4–6, 59, 67). Researchers assume that these segmental abnormalities occur due to breakage-fusion-bridge cycles (66). Failure during the repair of DNA double strand breaks (DSB) could lead to these abnormalities, but the origin of DSBs in human embryos remains unclear.

Only a few research groups investigated the origin of chromosomal abnormalities in preimplantation embryos. In this part, we will discuss the fundamental molecular mechanisms of this exceptional event, especially the role of the spindle assembly checkpoint (SAC) and the chromosomal passenger complex (CPC). These transcripts are targets for investigating the origin of aneuploidies in human preimplantation embryos.

1.3.1 Origin of mitotic aneuploidies

1.3.1.1 Underlying processes leading to abnormal chromosome numbers

A number of mechanisms could be the source of aneuploidy in human preimplantation embryos. Causes of abnormal chromosome segregation during mitosis are endoreplication, anaphase lagging, non-disjunction, premature cell division, chromosome demolition and/or breakage, cell fusion of blastomeres and error in cytokinesis (3) (figure 5). Several authors reported that anaphase lag and non-disjunction are the most common processes leading to numerical aberrations (5, 61–65). Ana- phase lag during mitosis is explained as a delayed motion of a single chromatid where it fails to attach to the spindle and thus is not included in the newly formed daughter cells. Since the lagging chromosome forms an anucleated fragment that is lost this mechanism is leading to one normal daughter cell and one carrying a monosomy. Mitotic non-disjunction describes the failure of sister chromatids to split correctly during metaphase resulting in one daughter cell with a monosomy and another with a trisomy. Coonen e t al. (58) examined blastocysts by performing FISH and concluded that two-thirds of the mosaicisms appear through anaphase lag in both sex-chromosomes and au- tosomes, while in one-third of the aneuploid blastocysts non-disjunction and anaphase-lag are pre- sent with non-disjunction foremost in the sex-chromosomes and anaphase-lag in the autosomes. Different from Coonen and colleagues another study performed three sequential rounds of FISH instead of one round on day 5 embryos. They reported that chromosome loss and therefore ana- phase-lag is the main mechanism leading to aneuploidy (59).

However, chromosome loss could be seen as artefacts due to failures as overlying signals or failure during hybridization. A combination of FISH, CGH and aCGH a few years later revealed that chro- mosome loss is not elevated relative to chromosome gains which contradicts the previous studies (62). Other mechanisms leading to aneuploidy are premature cell division and endoreplication. Premature cell division causes the formation of haploid cells, while endoreplication results in poly- ploid cells due to duplication of DNA without subsequent division of the cell. Triploidy due to en- doreplication was detected in human zygotes and tetraploidy has been observed in one two-cell embryo (5, 63, 64).

Abbildung in dieser Leseprobe nicht enthalten

The mechanism of chromosome demolition was initially introduced to explain trisomic zygote rescue. During chromosome demolition one of the three chromosomes undergoes removal and/or fragmen- tation resulting in two diploid daughter cells (65). If chromosome demolition however affects a dip- loid cell this will plausibly lead to a numerical chromosome aberration.

Mitotic aneuploidy can furthermore be correlated with errors in cytokinesis. Absent or distorted cy- tokinesis leads to cells with two nuclei, tetraploidy or spindle abnormalities and hence to a chaotic karyotype (66, 67). Xanthopoulou et al. (68) have found that blastomeres with two nuclei can also be the result from haploid mononucleated blastomeres which underwent endoreplication followed by karyokinesis.

Chromosome breakage is the last mechanism to be discussed. It causes partial chromosome loss and gain on almost each chromosome (53, 56, 61, 69). Embryo analysis using CGH have shown that embryos with partial aneuploidy are prone to an unstable karyotype caused by acentric and dicentric chromosomes (70). Since the resolution of CGH is lower than these of SNP array or aCGH (see 1.2.2.2), chromosome breakage resulting in smaller deletions or duplications were not detected. Vanneste et al. (6) showed that segmental deletions, duplications and amplification are highly pre- sent during early embryonal development including the appearance of breakage-fusion-bridge-cy- cles. Subsequently, Voet et al. (61) hypothesized that these aneuploidies occur through incorrect repair of DNA double strand breaks (DSB) in the zygote during the first two cleavages. There exist two key repair pathways for DSB: homologous recombination (HR) and non-homologous DNA end joining (NHEJ). Elucidating these pathways would go beyond the scope of this master thesis. Here, suffice it to mention that the presence of components of both pathways in human preimplantation embryos has been observed (71, 72).

Although the mechanisms behind these DSB during early embryonal development remain unclear, the existence and causes of analogous DSB in human embryonic stem cells (hESC) have been described. Jacobs et al. (73) hypothesized that DSBs are caused by replication stress and replication fork collapse.

1.3.1.2 Maternal factors affecting mitotic aneuploidies

Before going in-depth of molecular mechanisms, we will discuss briefly the maternal factors affecting aneuploidy. One key event during the first cleavages is embryonic genome activation (EGA). It oc- curs stepwise: first at the 2-cell stage (early maternal), second at 4-cell stage (late maternal) and the major activation between the 6-and the 8-to 10-cell stage (figure 6). Before major EGA, the development is dependent on the stored proteins and mRNA in the oocyte (74–77).

When a maternal protein product or mRNA is defective through exposure to environmental factors as radiation and chemicals or internal factors as poor vascularization of the antral follicle during oocyte maturation, the mechanisms that direct and control mitosis show a high risk to fail. Supposing such failure occurs before EGA, incorrect chromosome segregation could lead to aneuploidy since mechanisms as microtubule kinetics, cell cycle checkpoints, DNA repair proteins, chromosome co-hesion, telomere shortening and mitochondrial function are affected by maternal proteins (73). Examples of such protein products of maternal oocyte origin are the SAC and the CPC. Their role throughout mitosis and the consequences of depletion of certain components is clarified in the fol-lowing sections.

1.3.1.3 The spindle assembly checkpoint during mitosis

The SAC is a mitotic feedback-control system that prevents onset of anaphase and chromosome segregation, until all chromosomes are properly attached to the mitotic spindle (78). Thus, it avoids the development of aneuploidies by arresting the cell in mitosis and blocking transition into the ultimate steps of cell division (79). It consists of several proteins such as the sensor protein mitotic arrest deficient 1 (MAD1), budding uninhibited by benzimidazoles 1 (BUB1), and multipolar spindle 1 (MPS1). The signal converter mitotic checkpoint complex (MCC) is an organization of MAD2, BUB3, BUB related 1 (BUBR1) and cell division cycle protein 20 (CDC20) A second organization is the ubiquitin ligase anaphase–promoting complex/cyclosome (APC/C) (78–80).

During prometaphase, the kinetochores of the chromosomes attach to the spindle microtubules (fig- ure 7 A+B). The SAC is then active and monitors the state of kinetochore attachment. In case of improperly attached or unattached chromosomes the SAC signal is transduced. Subsequently, the SAC effector MCC, assembled at unattached kinetochores, binds and inhibits the APC/CCDC20-complex which regulates metaphase-anaphase transition. When all the chromosomes are properly attached to the mitotic spindle, the activation of the APC/CCDC20-complex enables Cyclin B and Securin ubiq- uitination and subsequent proteolysis leading to inactivation of Cyclin Dependent Kinase (CDK1) and activation of Separase. Inactivated CDK1 results in start of mitotic exit whereas active Separase cleaves the cohesin complex of the sister chromatids. This results in sister chromatid separation and finally to onset of anaphase. Moreover, other SAC key elements including MAD1, BUB1, MPS1, and Aurora-B increase the SAC signal and the amount of MCC assembly at the kinetochores (78, 79). Aurora-B is a serine/threonine (S/T) protein kinase functioning as a subunit of another complex named the chromosomal passenger complex which is discussed later.

The attachment process and the separation of the sister chromatids is shown in figure 7C. SAC proteins are recruited due to a signal produced by unattached kinetochores. This and the presence of a high level of MAD2 at the kinetochores lead to the concentration of Aurora-B kinase at the centromeres. Unattached kinetochores cause a lack of tension between the sister chromatids which possibly leads to Aurora-B activation.

Abbildung in dieser Leseprobe nicht enthalten

Figure 7 : The SAC mechanism and the kinetochore attachment process. (A) Mitosis is halted by active CDK1. The SAC is switched on during prometaphase. In case of improperly attached or unattached kinetochores, a SAC signal is transduced. The MCC amplifies this signal and inhibits after binding the APC/CCDC20-complex. Therefore metaphase-anaphase transition is avoided. When all of the chromosomes are properly attached to the kinetochores (metaphase), the APC/CCDC20-complex is activated. Subsequently, the ubiquitination of Cyclin B and Securin is promoted leading to the start of mitosis exit and the cleavage of cohesin of the sister chromatids, respectively. (B) Concurrent MCC production and inhibition even when the SAC is switched on, assure for a high level of MCC at any given time. (C) Unattached kinetochores constantly produce a ‘wait’ signal, expressing high levels of MAD2 (red). Under these circumstances and probably due to lack of tension, Aurora B kinase also assembles but at the centromeres (green). Increase of tension due to attached kinetochores in turn, depletes MAD2 and the SAC signal terminates. Then Separase activity cleaves the cohesin of the sister chromatids and anaphase ensues. At this point, also Aurora B kinase is depleted from the centromeres. (A) and (B) from (79), (C) from (78).

After the attachment of the chromosomes to the microtubules, MAD2 levels decrease 50-100-fold. When all chromosomes completed attachment the SAC signal is quenched. Separase is then acti- vated and cleaves the sister chromatids by proteolysis of cohesin (78).

The SAC is indispensable for normal mitotic progression since it prevents metaphase-anaphase tran- sition until not all chromosomes are properly attached to the mitotic spindle. Wei et al. (80) investigated the role of SAC elements (BUB3, BUBR1 and MAD2) in mouse preimplantation embryos by overexpression or knock-down. Overexpression of such components inhibited the separation of sister chromatids, confirming the role of monitoring the metaphase-anaphase transition during the first cleavages. Downregulation in turn hastens metaphase-anaphase transition and causes micronuclei formation, improper chromosome segregation and aneuploidy resulting in implantation failure and delayed development.

1.3.1.4 The chromosomal passenger complex

The CPC is one of the main controllers of mitosis and is present at different localizations during certain phases of mitotic cell division and cytokinesis. It regulates key mitotic events such as cor- rection of chromosome-microtubule attachment errors, activation of the SAC, chromosome conden- sation, spindle assembly and cytokinesis. The CPC is organized into four subunits: Aurora B, INCENP (inner centromere protein), Survivin and Borealin. Especially Aurora B as a protein kinase has mul- tiple functions such as signaling to the SAC through phosphorylation which leads to the assembly of MCC, the effector of the SAC. The INCENP is the scaffold where the CPC proteins assembly occurs. Aurora B is fully activated through binding on the IN-BOX of INCENP leading to phosphorylation (figure 8 A). During early mitosis, the INCENP/Borealin/Survivin complex localizes to histones at the inner centromere, whereas during late mitosis this complex is binding to microtubules at the spindle midzone. Increased CPC on these sites enables phosphorylation and therefore full activation of Aurora B (figure 8 B) (81). Van de Werken et al. (82) revealed that the assembly of CPC proteins at the INCENP during prometaphase of human zygotes is less confined than in later stages of pre- implantation development. They assume that the disordered CPC localization in the zygote causes an increase in chromosome mis-segregation, particularly of paternal origin.

It has been shown that the inhibition of Aurora B and depletion of Borealin increases kinetochore- microtubule attachment errors (83, 84). During normal conditions, however, the CPC destabilizes and repairs those fail- ures (85, 86) . Moreover, the CPC is es- sential for SAC activation since active Aurora B facilitates recruitment of the SAC elements MAD1, MAD2, BUB1, BUBR1 and MPS1 (87–89). In late mito- sis researchers observed a role of the CPC in central spindle formation, regu- lation of furrow ingression and abscis- sion (86, 90). These processes appear in cytokinesis which allows for a coordi- nated and accurate separation of the two daughter cells. The role of the mei- otic kinase Aurora C in the CPC during human preimplantation development is demonstrated by Avo Santos et al. (91). Aurora C appears in oocytes, zygotes and stays in early preimplantation em- bryos. Somatic Aurora B in turn is only present in blastocysts. The investigators hypothesize from these results that Au- rora C is involved during the develop- ment of aneuploidies.

Figure 8 : Aurora B activation and CPC localization (81). (A) Inactive Au- rora B (red) binds to the INCENP) yellow) and is fully activated after phos- phorylation by trans catalysation. (B) in vivo Aurora B activation requires CPC localization. INCENP, Borealin and Survivin target the CPC to the his- tones of the inner-centromere (early mitosis) and the spindle midzone (late mitosis). Assembly at this region enables auto-phosphorylation in trans, which fully activates Aurora B kinase.

2. Objectives of the master thesis

The main objectives of this master thesis are to develop and validate a protocol for simultaneous genome and transcriptome analysis at the single cell level. For that, we plan to analyze hESC (which have a transcriptome close to that of the preimplantation embryos) and fibroblast cells with normal and abnormal karyotype. At the start of the experiments we will use (1) purified RNA & DNA mixes, (2) larger numbers of cells and (3) dilutions down to one single cell equivalent for validation of the protocol described by Macaulay et al. (92).

The three sub-goals are validation and implementation of (1) Separation of RNA & DNA, (2) RNA amplification, (3) DNA amplification. We want to mimic the amount of RNA and DNA in a single cell (10-30pg total RNA of which <1pg mRNA and ~6pg total DNA) before we perform the experiments on single cells. To validate for separation of RNA and DNA we will test their presence by real-time PCR (GAPDH, MSH2, SOX2 and NANOG) and PCR (CFTR gene), respectively. When reaching these aims on time, we will include the validation of shallow whole genome sequencing.

By implementing this innovative technique in our lab, it will be possible to investigate each single blastomere of human embryos at the two- to eight-cell stage. This allows for studying the origin of chromosomal abnormalities in early embryos. The REIM/REGE group particularly wants to investi- gate the hypothesis that the depletion of mRNA and protein products of maternal oocyte origin and that are involved in mitosis -especially SAC and CPC- lead to aneuploidy before EGA. RNA-sequenc- ing of the whole transcriptome will also allow to investigate alternative pathways, such as the DNA repair pathways.

Several authors have previously analyzed the transcriptome of whole embryos and observed the presence of transcripts for SAC, CPC and DNA repair mechanisms (71, 75, 77, 93). These investiga- tions disagree with a role of aforementioned components during development of aneuploidies in early embryos. Nevertheless, experiments on whole embryos hide possible effects at the single cell level, which is also the level at which aneuploidies occur. RNA-sequencing of individual single blas- tomeres reveals different levels of transcripts including of genes for SAC, CPC and DNA repair path- ways (77). We have to mention that these researchers did not associate the transcriptome with the presence of chromosomal abnormalities.

[...]

• College: Vrije University Brussel (Reproduction Genetics and Regenerative Medicine)
• Grade: 15/20
• Author: Marius Regin (Author)
• Publication Year: 2018
• Pages: 43
• Catalog Number: V490849
• ISBN (eBook): 9783346020598
• Language: English
• Tags: single cell analysis