X‐chromosome inactivation and telomere size in newborns resulting from intracytoplasmic sperm injection

In vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) are generally considered to be safe technologies for treatment of infertility. However, there is evidence for: (i) a small increase in chromosome abnormalities, especially involving the sex chromosomes, after ICSI [Bonduelle et al., 1998] (ii) an increase in pre-maturity and decrease in mean birthweight in singleton babies born after IVF or ICSI [Hansen et al., 2002; Winston and Hardy, 2002; Powell, 2003], and (iii) a slight increase in birth defects after both IVF and ICSI [Hansen et al., 2002; Winston and Hardy, 2002; Powell, 2003]. Recent reports of an increase in imprinting errors (loss of normal parent-of-origin specific gene expression) after IVF/ICSI have heightened concern for the safety of assisted reproduction procedures [Gosden et al., 2003; Niemitz and Feinberg, 2004]. All these findings may be a consequence of the selected population utilizing assisted reproductive technologies, if the underlying cause of sub-fertility is somehow related to increased abnormalities in eggs or sperm. However, one cannot exclude that the technology itself affects risk of birth defects by altering the normal environment during germ cell maturation or early embryo development. If that were the case, then subtle perturbations in development may be relatively common but only rarely yield a detectable phenotype such as intrauterine growth restriction or an imprinting disorder. We propose that assessment of developmentally regulated features of chromosomes, such as X-chromosome inactivation (XCI) and telomere length, can be used as additional measures for the assessment of early embryo development in the context of assisted reproduction. Both these changes are completed by the late blastocyst stage and thus may be sensitive to perturbations in epigenetic programming due to embryo culture. To assess this, we evaluated XCI skewing ratios and/or telomere length in DNA extracted from cord blood from a series of 48 female and 19 male newborns (including three sets of non-identical twins) conceived after ICSI. University of British Columbia Ethics Committee approval was obtained before initiating the study. A karyotype from umbilical cord blood was available for all studied cases. One of these infants was a reciprocal translocation carrier [46,X,t(X;20)] [Ma et al., 2003]. The androgen receptor (AR) PCR assay of XCI was used as described previously [Beever et al., 2003]. Informative results were obtained for the AR assay of XCI in 44 of the 48 cord blood samples. The remaining four samples were either homozygous or had alleles too close to accurately evaluate. Control values for XCI skewing in 74 blood samples from 0–19 year olds were taken from a previous study in our laboratory [Hatakeyama et al., 2004]. The observed distribution of skewing is presented in Figure 1. There was no significant difference between ICSI samples and controls in mean level of skewing (65.1 vs. 69.8), or in the frequency of skewing ≥75% (18% vs. 35%) or ≥90% (4.6% vs. 10.8%). If the X;20 translocation case is removed from the analysis (since skewed XCI is expected in such translocations independent of ICSI) then the frequency of extreme skewing in the ICSI group would be only 2.3% (n.s.) and the mean level of skewing of 64.3 is significantly less than observed in controls (P = 0.025; t-test). Our control sample consisted of cases aged 0–19, and while there is no significant change in skewing between ages 0–19 [Hatakeyama et al., 2004], a slight effect of age in this range cannot be excluded. The distribution of skewing values in the 44 ICSI cord blood samples is compared to 74 young controls. No difference in the distribution of skewing values is observed. The distribution of XCI skewing in the general population is thought to be largely due to chance deviations from 50:50 as a consequence of the limited number of embryonic precursor cells present (4–20) at the time an X-chromosome is committed to inactivation within each cell of the developing blastocyst [Puck et al., 1992; Monteiro et al., 1998]. For example, extremely skewed XCI defined as >90% inactivation of one X, is present in only about 7% of young females [Beever et al., 2003] but is present in over 50% of chromosomally normal fetuses or newborns associated with high levels of placental trisomy [Peñaherrera et al., 2000]. As the distribution of XCI skewing in the ICSI population was similar or less than that in the control population, we can assume that there is no major reduction in size of the embryonic precursor pool in the ICSI conceived blastocyst as compared to normally conceived blastocysts. Thus, we also infer that there are not likely to be increased levels of chromosome mosaicism in viable ICSI blastocysts as compared to viable normally-conceived blastocysts. Reduced skewing in the ICSI group could be explained by selection against slower growing embryos during the ICSI procedure (e.g., by selecting only the best growing embryos for transfer back to the mother's uterus). Clearly, the present sample size is too small to exclude that abnormal events may occur in a small number of cases, or have an effect that is mostly limited to non-viable embryos. Telomere regeneration is another programmed developmental change occurring primarily at the early embryo to blastocyst stage of development. Studies of bovine embryos have shown that telomerase activity is maintained throughout oocyte and early embryo development, but with a notable increase in telomerase activity from the eight-cell to blastocyst stage embryo [Xu and Yang, 2001]. Some cloned organisms have been found to have either shorter or longer telomere length then expected [Shi et al., 2003]. This may have been due to culture conditions affecting telomerase activity in the early developing embryo. To our knowledge, there have not been studies of telomere length in IVF or ICSI conceived humans. Due to insufficient DNA, only a subset of 24 of the female ICSI cases was tested for telomere size. In addition 19 cord blood samples from male ICSI cases were analyzed. Only newborn cord blood samples were used as controls since telomere length declines rapidly in the first year of life [Rufer et al., 2001]. An estimation of average telomere length was obtained using the Telo TAGGG Telomere Length Assay kit from Roche Diagnostics (Montreal, CA). This assay utilizes Southern analysis of the terminal restriction fragment (TRF) that is obtained by digestion of 0.5 µg of genomic DNA using frequently cutting restriction enzymes, HinfI and RsaI. The TRF includes the (TTAGGG)n repeat as well as a portion of the subtelomeric region or telomere associated repeats (TAR). Telomere length was quantified by measuring chemiluminescence using a Bio-Rad Fluor-S MultiImager, and analysis with Quantity One software. To assess the reproducibility of the telomere length assay, four samples were analyzed two times on separate Southern blots. They were well correlated (r = 0.96) and a one-way ANOVA for correlated samples determined that there was no significant difference between the two groups of measurements (F = 2.63; P = 0.20). Figure 2 shows the distribution of telomere length values for the 43 ICSI samples and 16 control samples. Average telomere length varied considerably in each group, with a mean telomere length of 12.3 (±1.4, range 7.2–21.2) in the ICSI samples and 11.56 (±1.2, range 7.7–15.2) in the controls. There was no significant difference between the male (mean = 13.3, n = 19) and female (mean = 11.5, n = 24) ICSI cases. There were several very large telomere lengths in the ICSI group. However, there is less accuracy in length designation for large sized alleles due to the shorter migration of fragments in the larger size range. The distribution of telomere sizes in 43 cord blood samples from ICSI newborns is compared to that for 16 control cord blood samples. No difference is observed. Much remains to be understood about the etiology of slightly elevated risk for birth defects and increase in rare imprinting disorders after IVF and ICSI. A concern is that epigenetic mutations may be common among IVF/ICSI conceived infants, but only rarely have an effect apparent at birth. Alternatively, the observed cases of Beckwith–Wiedemann syndrome and Angelman syndrome may be reflective of a susceptibility particular to a very small number of cases and not the IVF/ICSI group as a whole. Follow-up studies looking at neurodevelopment after ICSI/IVF have not shown any differences with each other or as compared to their naturally conceived peers [Bonduelle et al., 2003; Sutcliffe et al., 2003], thus suggesting gross changes in gene expression are unlikely to be common. Our present results are also consistent with normal early blastocyst development in the majority of viable embryos conceived by ICSI and should be reassuring. However, it is important to stress that we measured only two types of epigenetic changes occurring in early development and did not assay any measures of genomic imprinting. Further testing of other epigenetically marked regions (e.g. imprinted genes), along with XCI and telomere length, in a larger multi-institutional study will be necessary to assess the degree of perturbation, if any, after ICSI. We gratefully thank the families for donating the samples for this study. We thank the laboratory and clinical staff and physicians at the UBC IVF clinic, the physicians and nurses who aided in the collection of umbilical cord blood, and also Ruby Jiang and Luana Avila for technical support. This presented study was approved by UBC ethics (C98-0315).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.