第二章 研究設計與方法
第十節 統計學方法
Statistical Analysis
The mean with standard deviation for the continuous variables were calculated and the between-group differences were analyzed by Student’s t-test. Categorical variables were presented as raw frequencies with corresponding percentages, and the between-group differences were analyzed by Chi-square test with Yates correction if indicated. The results were considered statistically significant at p < 0.05. The strength of correlation between two independent variables was calculated with Pearson product-moment correlation coefficient. All the continuous and categorical variable analysis were performed with GraphPad QuickCalcs, and all the figures and tables were created with Microsoft Excel 2016.
第三章 研究結果
第一節 胚胎染色體套數與粒線體 DNA 含量之關係
The relationship between mtDNA score to embryo ploidy
In term of embryo ploidy, a significance difference in mean mtDNA score was observed between all euploid and aneuploidy blastocysts, irrespective of development rate (0.00098966347 ± 0.00071762096 vs. 0.00113364289 ± 0.00129129378; p = 0.023). When further analyzing only blastocysts of identical development rate, day 5 euploid blastocysts showed a significantly lower mean mtDNA score compared to day 5 aneuploid blastocysts (0.001072905 ± 0.000714348 vs. 0.001239826 ± 0.001475416; p = 0.03). On the contrary, though day 6 euploid blastocysts also showed a trend of lower mean mtDNA score compared to day 6 aneuploid blastocysts, the difference was not statistically significant (0.000819607 ± 0.000696004 vs. 0.000963201 ± 0.00089825; p = 0.05). The results from the above analysis are outlined in Figure 1.
第二節 胚胎發育速度與粒線體 DNA 含量之關係
The relationship between mtDNA score to embryo development rate
In term of embryo development rate, a significance difference in mean mtDNA score was observed between all day 5 and day 6 blastocysts, irrespective of ploidy (0.001172775 ± 0.00122991626 vs. 0.000913612 ± 0.00083600924; p = 0.001). When sub-analyzing only embryos of identical ploidy, day 6 euploid blastocysts showed a significantly lower mean mtDNA score compared to day 5 euploid blastocysts (0.000819607 ± 0.000714348 vs.
0.001072905 ± 0.000696004; p = 0.022). Similarly, day 6 aneuploid blastocysts also showed a significantly lower mean mtDNA score compared to day 5 aneuploid
blastocysts (0.001239826 ± 0.001475416 vs. 0.000963201 ± 0.00089825; p = 0.001). The results from the above analysis are outlined in Figure 2.
第三節 母體年齡與粒線體 DNA 含量之關係
The relationship between mtDNA score to maternal age
A significant correlation failed to exist between mtDNA score and maternal age (Pearson product-moment correlation coefficient, r = -0.002). In order to further investigate the effects of maternal age, the difference of mtDNA score between patients <40 year-old and ≥ 40 year-old at the time of oocyte retrieval was also analyzed. No significant difference in mean mtDNA score was found between the two maternal age groups.
Furthermore, when considering the blastocysts of similar ploidy and development rate, a significant correlation also failed to exist between maternal age and mtDNA score. The results from the above analysis are summarized in Figure 3.
第四節 第五天和第六天囊胚之比較
The outcomes of Day 5 euploid blastocysts vs. Day 6 euploid blastocysts
The embryo that developed into mature blastocyst on day 5 had a significantly higher proportion being euploid compared to day 6 blastocysts (40.1% vs. 34.5%, p = 0.044).
When comparing the clinical pregnancy rate of day 5 euploid blastocysts versus day 6 euploid blastocysts, a significant difference was noted favoring the day 5 embryos (69.7%
vs. 58.1%; p = 0.038). The euploid blastocysts that resulted in clinical pregnancy have significant lower mean mtDNA score than the euploid blastocysts that did not result in clinical pregnancy (0.00069757754 ± 0.00034627866 vs. 0.00159407610 ±
0.00092511726; p < 0.0001). When inspecting blastocysts of different development rate, the day 5 euploid blastocysts that resulted in clinical pregnancy have significant lower mean mtDNA score than the day 5 euploid blastocysts that did not result in clinical pregnancy (0.00075879001 ± 0.00037345518 vs. 0.00168701665 ± 0.00079820584; p <
0.0001). The same phenomenon was also observed when comparing the mean mtDNA score between Day 6 euploid blastocysts that resulted in clinical pregnancy and the one that did not (0.00057267100 ± 0.00024040394 vs. 0.00139528658 ± 0.00113822518; p <
0.0001). The day 5 euploid blastocysts that resulted in clinical pregnancy after transfer had significantly higher mean mtDNA level than day 6 euploid blastocysts that resulted in clinical pregnancy (0.00075879001 ± 0.00037345518 vs. 0.00057267100 ± 0.00024040394; p < 0.0001). Theresults from the above analysis are outlined in Figure 4 and Figure 5, respectively.
第五節 粒線體 DNA 含量用於預測胚胎之懷孕率
The optimal cut-off mtDNA score for predicting clinical pregnancy of euploid embryos
The absolute mtDNA threshold score for the prediction of clinical pregnancy after transfer of an euploid embryo was 0.002, above which no clinical pregnancy was attained.
(100% negative predictive rate). Among all the transferred euploid embryos (n = 338), none with mtDNA score greater than the absolute threshold (n = 25) achieved clinical pregnancy. On the other hand, the transferred embryos with mtDNA score below the absolute threshold value (n = 313), 71.8% achieved clinical pregnancy (n = 225).
In order to determine the optimal upper and lower threshold of mtDNA score that predict clinical pregnancy for euploid embryo, we performed an analysis of the mtDNA score for all the transferred euploid embryos. Specifically, individual mtDNA score for all the
transferred euploid embryos and the euploid embryos that established successful clinical pregnancy were ranked and stratified into respective centiles and represented as box plots.
A sub-analysis was also performed for transferred euploid embryos of different development rates (Day 5 and Day 6 blastocysts). The results from the above analysis are summarize in Table 1.
When all the transferred euploid embryos were analyzed together, the inter-quartile range (IQR) of mtDNA score for embryos that resulted in clinical pregnancy (Q1 = 0.000443845, Q3: 0.000850386; range = 0.000406541) corresponded approximately to the range of mtDNA score between the median and the 1st decile for all transferred euploid embryos (1st decile = 0.0003513795, median = 0.0008071195; range = 0.00045574).
Similar trends were also observed in the sub-analysis for embryo of different development rate. For transferred Day 5 euploid embryo, the inter-quartile range (IQR) of mtDNA score for embryos that resulted in clinical pregnancy (Q1 = 0.000468681, Q3: 0.00095547;
range = 0.000486789) corresponded approximately to the range of mtDNA score between the median and the 1st decile for all transferred Day 5 euploid embryos (1st decile = 0.00038259, median = 0.0008634155; range = 0.00038259). For transferred Day 6 euploid embryo, the inter-quartile range (IQR) of mtDNA score for embryos that resulted in clinical pregnancy (Q1 = 0.000379062, Q3 = 0.000729541; range = 0.000350479) corresponded approximately to the range of mtDNA score between the median and the 1st decile for all transferred Day 6 euploid embryos (1st decile = 0.000313369, median = 0.0006550315; range = 0.0003416625). Theresults from the above analysis are outlined in Figure 6.
研究討論
The results of the present study collaborated, and at the same time contradicted numerous findings from the past studies on mitochondria and IVF outcomes. First of all, our data provided another robust evidence, that aneuploid embryos contained significantly greater quantity of mtDNA compared to eupoid embryos. Secondly, when considering the embryo development rate, day 5 blastocysts had a significantly higher mtDNA quantity than day 6 blastocysts, independent of embryo ploidy. Thirdly, when factoring the effects of maternal age, no differences in the mtDNA quantity was noted when analyzing blastocysts pooled from subjects aged < 40 and >40 years-old, irrespective of embryo ploidy and development rate. The comparison of embryo euploidy rates, clinical pregnancy rates, and mtDNA score of euploid embryos that resulted in clinical pregnancy between day 5 blastocysts and day 6 blastocysts also showed significant difference favoring day 5 blastocysts. Lastly, the calculated optimal range of mtDNA scores that predicted clinical pregnancy of euploid blastocysts were the values between median and the 1st decile of all the transferred euploid embryos, respectively. The same upper and lower cutoff range were also observed when blastocysts of different development rate were analyzed separatety.
第一節 粒線體於卵子及胚胎發育時期所扮演之角色
The role of mitochondria in oogenesis and embryogenesis
In order to truly appreciate the complex dynamics of mitochondria biogenesis during embryonic development, one must also consider their roles in oogenesis. At early phase of female embryogenesis, the mtDNA content of primordial germ cells (PGCs) first undergoes a marked reduction in quantity, followed by gradual increase during ensuing
stages of oogenesis, peaking at late folliculogenesis when the oocyte reached metaphase II (MII) stage [24]. Following fertilization, due to absence or minimal levels of mtDNA replications factors such as POLGA, POLGB and TFAM, the mtDNA copy number in mammalian embryos either remained constant [25] or reduced [26-27] throughout the cleavage stage up until morula stage, which roughly coincided with the commencement of embryonic genome activation (EGA) / maternal to zygote transition (MZT) [28].
Consequently, the number of mtDNA copies in each blastomere would progressively dwindle after each cell division, limiting individual blastomere’s capacity to generate ATP through OXPHOS, as the organelles become more reliant on anaerobic respiration as energy source [29-30]. Morphology analysis at this stage revealed immature spherical mitochondria with few cristae and low membrane potentials [31]. When the embryo reached the blastocyst stage, the mtDNA replication factors would once again be upregulated, leading to reactivation of mtDNA replication [25-27]. Morphology wise, mature mitochondria at this time would appear as elongated organelles containing swollen cristae, with increased mitochondrial membrane potential and oxygen consumption that facilitate OXPHOS activity and ATP production [31]. This process of initial reduction and subsequent amplification of mtDNA at the end of folliculogenesis constitute a genetic bottleneck effect that allows the preferential passage of only a fraction of mtDNA content to oocytes [32]. It is probable that the founding mitochondria were selected based on their mtDNA integrity and OXOPHOS functions, thus ensuring only the healthiest mitochondria would be preserved, and at the same time minimizing the transmission of abnormal mitochondria genome.
第二節 於染色體套數異常或品質不佳胚胎中 粒線體 DNA 含量增加之原因與意義
The significance of increased mtDNA quantity in embryo with aneuploidy or reduced viability
The vital roles of mitochondria in oocyte and embryo development have been long recognized [33-34]. Therefore, one may wonder, if mitochondria have such important functions, how can increased mitochondria quantity in a cell be detrimental? The possible explanations for such paradoxical relationship may be viewed in context of suboptimal mitochondria quality per se, or the overall dysfunction of embryo development.
In the first scenario, when the embryo contained excessive number of functionally deficient mitochondria, a compensatory mechanism of mitochondria pool expansion would be activated as an effort to restore normal ATP production. However, if most of the newly replicated mitochondria still harbor the same defect, the defective energy state of the embryo would still be unrectified thus affecting embryo viability. Indeed, the above theory merits consideration since mtDNA is highly prone to mutation due to its proximity to the DNA-damaging reactive oxygen species (ROS) generated by the electron transport chain, as well as the absence of protective histone that shields DNA against hydroxyl radical-induced DNA strand breaks [35]. Therefore, the mutation rate of mtDNA has been estimated to be almost 25 times higher than that of nuclear DNA [36]. Tzeng et al. have reported decreased expressions of mitochondrial oxidative phosphorylation genes ND2, CO I, CO II, ATPase 6, CO III, ND3, ND6 and Cyt b in both unfertilized human oocytes and arrested embryos [37]. In addition, the same group also reported increased frequency of mitochondrial gene deletions in unfertilized human oocytes and arrested embryos, with the 4977-bp deletion being the most common [38]. The evidence of compensatory increment of mtDNA content in response to mutation has also been observed in the
oocytes and embryos from patients of myopathy, encephalopathy, lactic acidosis and stroke-like syndrome (MELAS) [39]. In spite of the above findings, a definitive link between mutation of the mitochondrial genome and the increased mtDNA levels associated with reduced implantation has not been demonstrated.
Another plausible theory considers the quality of embryo as a whole, rather than the functional status of mitochondria alone. In the “quiet embryo hypothesis” by Leese et al., the authors proposed that embryos with increased metabolic activity are in fact suffering from certain extrinsic stress (i.e. culture condition) or intrinsic structural defects (i.e., aneuploidy) [40]. Thus greater energy expenditures are required as an effort to restore proper embryonic development, which resulted in compensatory mitochondria replication and a net gain of mtDNA. Since blastocyst stage is the period when significant mtDNA synthesis takes place, which also coincide with the timing of embryo biopsy for PGS, therefore this might explain the phenomenon of embryos with elevated mtDNA levels detected at this stage of development. In theory, the excessive amount of ATP generated could lead to increased free radical production, causing damage to DNA and other organelles, eventually impairing embryo viability [41-42]
第三節 粒線體異質性與試管嬰兒成功率之關係
The role of mitochondrial heteorplasmy in relation to IVF outcomes
The extent of heteroplasmy, defined as the presence of more than one mtDNA variant in the mitochondria of an individual, was often used to determine the quality of mitochondria gene pool. It has been estimated that each person on average carry one heteroplasmy, with one in eight carry disease-associated heteroplasmic variant [43]. Even though mitochondrial heteroplasmy has been extensively implicated in numerous human disease states [44-45] and cancer progression [45-46], its association with the viability of
embryos derived from ART is less clear. In the study by Fraguli et al., varying degrees of heteroplasmy were observed in all of the embryo biopsy samples, irrespective of embryo mtDNA levels and transfer outcomes [10]. On the other hand, Lledo et al. reported a significant positive association between the degree of hetroplasmy and mtDNA copy number [14]. Though a lower trend of implantation rate and ongoing pregnancy rate, along with a higher miscarriage rate were observed for embryos that carried at least one heteroplasmic variant, the differences did not reach statistical significance. In essence, since each cell can contain up to thousands of mitochondria molecules, and each mitochondrion also has multiple copies of mtDNA, therefore a threshold of hetroplasmy, or “mutational load” must be exceed for the phenotypic manifestation of the defect to be evident [47]. For human embryos, it has been estimated that a heteroplasmic level up to 18% could still bear 95% or greater chance of having unaffected offsprings [48]. Base on the above observations, we opted not to incorporate the degree of heteroplasmy in our present study.
第四節 “越少越好?” 粒線體 DNA 含量多寡與胚胎發育速度之關係
“The less the better?” The significance of increased mtDNA quantity and embryo developmental rate
An interesting discovery was made when comparing the mtDNA score from blastocysts with different development rate. The day 5 blastocysts consistently demonstrated a higher mtDNA score than the day 6 blastocysts, even when euploid and aneuploid embryos were considered separately. Similar finding was observed in the study by Lledo et al., and the authors proposed a “dilution effect” as the explanation for such phenomenon [14]. The theory is based on the observation from a mouse model study that found the quantity of mtDNA remained relatively constant between fertilization and the early
post-implantation stage, and no additional mtDNA replication would occur during that period of time. Therefore, with increased rounds of cell divisions in an embryo, the “same”
amount of mtDNA would be distributed among growing number of embryonic cells, and thus by day 6 of development each cell would contain fewer copies of mtDNA than at day 5. The above theory is questionable since no comparison between the transfer outcomes of day 5 and day 6 euploid embryos was performed. Such distinction is crucial, since in order for the above-mentioned “dilution effects” to be valid, the authors would be making the assumption that day 5 and day 6 blastocysts possessed similar developmental competency, therefore it did not matter when biopsy was performed as the mere difference would be just the number of cell divisions.
We hereby propose a “energy threshold” theory for the discrepancy. In our study, all of the blastocysts that reached expanded stage of development (Gardner’s grading ≥ 3CC) on day 5 were biopsied on the same day, while the ones that had a slower growth rate were either biopsied on day 6 when the same criteria were met or not biopsied. The slower blastocyst formation rate could be viewed as inferior developmental competency due to lower number of mitochondria organelles (thus mtDNA) presented, making embryos less energetically favorable to sustain normal blastocyst formation. This theory is backed by the results of the present study that demonstrated significantly higher embryo euploid rate and clinical pregnancy rate for day 5 blastocysts than for day 6 blastocysts. Other studies also collaborated our theory, that embryos that developed into mature blastocysts on day 5 almost always had higher euploidy rate [49-51], implantation rate [52-53], and live birth rate [53-56] when compared day 6 blastocysts, with or without PGS. Therefore, according to the above findings, the lower quantity of mtDNA in an embryo does not always imply increased competency, rather the mtDNA quantity above a certain threshold must also be maintained in order for optimal embryo development.
第五節 粒線體 DNA 量與母體年齡無顯著關聯之原因
The absence of maternal age effect on mtDNA quantity – Why?
In line with previous reports, the present study did not find a significant correlation between maternal age and mtDNA score, regardless of ploidy status. The possible rationale for this observation could be explained by the bias of our sample selection process. Since all of the study’s samples were from embryos that developed successfully either on day 5 or day 6 post fertilization into matured blastocysts, in essence all the embryos which exhibited slower growth rate, poor cellular components, and growth arrest were excluded from the final analysis. Therefore, only the mtDNA content from the embryo that demonstrated adequate degree of developmental competence were analyzed, thus diluting the adverse effects of advancing maternal age. Indeed, similar phenomenon has been observed in past studies with PGS, whereas the clinical pregnancy rate and live birth rate after the transfer of euploid embryos did not demonstrated significant differences across all age groups, even for women of advanced age (>40 year-old) [57-59]. These results are in accordance with the ultimate goal of embryo selection, which is to minimize the transfer of developmentally compromised embryos.
第五章 研究限制
Study Limitations
The main limitation of the present study was its retrospective single-center design. In addition, the mtDNA was evaluated exclusively in quantitative manner, while the degree heteroplasmy, as well as other important mitochondrial function such as oxygen consumption rate (OCR) and ATP synthesis capacity, were not evaluated. The method mitochondrial quantification used in the present study also have several drawbacks. First of all, as stated previously, the number of trophectoderm cells obtained during biopsy cannot be controlled. Therefore, the quantity of mitochondria DNA must always be presented in a relative manner instead of the actual amount. In theory, the most ideal study design would be the analysis of individual trophectoderm separately. That way, the actual amount of mitochondrial DNA presented would be specifically for a single trophectoderm cell. However, such strategy faces numerous technical challenges under the original study setting. Presently there is no known effective method for separating the individual cell in a given trophectoderm biopsy sample without compromising the integrity the cellular contents. Secondly, with single cell analysis, each cell must undergo its own round of WGA, which would result in variable degree of ADO and preferential amplification among cells within a given sample, making the NGS results more susceptible to technical biases. Thirdly, the overall cost of PGS procedures would be significantly increased with single cell analysis, thus inflicting more financial burden upon patients. For example, with a five cell biopsy sample, a total of five rounds of WGA would need be performed, and more rounds of multiplexing would also be required for processing increased number of WGA samples. It is also technically difficult to infer the number of cells analyzed from the NGS data, since the sample has already undergone WGA. Furthermore, whether a single trophectoderm biopsy can adequately represent the chromosomal and
mitochondrial composition of an entire embryo still merits careful consideration. Lastly, all the evaluated embryos were derived from stimulated IVF cycles from a very specific population. Therefore, whether these findings are translatable in general population warrant further studies.
第六章 研究結論
Conclusions
The results of the present study have proven that the mtDNA quantity alone is a useful indicator that accurately predicts the implantation potential of euploid embryos. With the prevalence of high throughput NGS platforms used in PGS, scientists now are allowed to
The results of the present study have proven that the mtDNA quantity alone is a useful indicator that accurately predicts the implantation potential of euploid embryos. With the prevalence of high throughput NGS platforms used in PGS, scientists now are allowed to