While there have been significant advances in the treatment of infertility, concerns have been raised due to the increased prevalence of certain defects in babies born by ART. In particular, two very rare disorders, Beckwith-Wiedemann syndrome (BWS) and Angelman Syndrome (AS), occur more frequently in babies born by ART than would be expected, based on the normal incidence of these disorders. Some case studies suggest as much as a nine-fold increase in the risk of BWS after ART. What is intriguing about these diseases is that they are both 'imprinting disorders'. Imprinting is a remarkable phenomenon whereby instead of having the usual two active copies of a gene, one inherited from each parent, only one of the inherited copies is active. The other copy is silenced by epigenetic marks. The absolute risk of imprinting disorders in children conceived by ART is very small but epigenetic errors might also influence susceptibility to cancer and other common diseases and understanding the origins of these errors is of vital importance.
We inherit one set of genes from each parent and so we have two copies of each gene (except the genes on sex chromosomes). Imprinted genes are unusual as they are only active from one copy. Epigenetic marks, which are inherited from either the father or the mother, switch off the other copy. 'Imprinted loci' are regions which contain these special genes and the epigenetic marks that switch them on and off. If these marks are lost or if a region has the wrong mark, this can result in either both copies of the gene being incorrectly switched off or both copies being switched on - that means either no gene activity or twice as much activity than required. Both types of change can cause disease.
We know that epigenetic errors can occur in response to diet or on exposure to certain chemicals and this can 'turn on' or 'turn off' certain genes and cause disease. The most vulnerable periods seem to be during sperm and egg formation and also very early in development, before the embryo implants into the mother's womb. The identification of epigenetic changes at imprinted loci in ART infants, both IVF and ICSI, raises the possibility that some aspect of the fertility treatment causes these epigenetic errors. Indeed, several researchers groups have shown that the in vitro culture of embryos can cause epigenetic errors at imprinted loci.
Recent work suggests a further source of epigenetic defects, the father's sperm (1). In 2007, Dr Takahiro Arima and his colleagues had noticed that men with oligospermia carried epigenetic errors in their sperm. Rather that inducing epigenetic errors by the procedure of ART, perhaps the men were passing on these errors? In order to test this theory, DNA was isolated from a number of ART conceptuses from unsuccessful pregnancies to identify samples that carried imprint errors. Next, the sperm DNA from the fathers was examined to see if any carried the same errors. There were seven of 17 samples where exactly the same errors were present in both the sperm and the unsuccessful ART pregnancy. This showed that the errors were most likely inherited directly from the father and not induced by the ART technique.
It was also remarkable that within the group of men with imprinting errors, there was generally a low sperm count (oligospermia). This observation was intriguing and suggested a possible link between male infertility and imprinting errors. In male mice, complete loss of a gene called DNMT3L causes infertility and abnormal imprints (2, 3, 4). This suggested that perhaps the infertile men carried mutations in their DNMT3L gene. In fact, two of the fathers who passed on their imprint errors had small changes in DNMT3L. Several other infertile men also carried changes to this gene. The changes in the gene were predicted to have only a mild effect on the function of the protein. This might explain why these infertile men were oligospermic (low sperm count) rather that azoospermic (no sperm), like the mice. This might also explain why there was only a partial change at some of the imprinted loci and not a complete disruption of imprinting.
This work demonstrates that the quality of sperm is critical both for fertilisation and also for a healthy pregnancy. ART is perhaps too effective at over riding nature's way of eliminating poor quality sperm. However, we now have the means to test sperm for the presence of imprinting errors and for alterations in DNMT3L. This means that infertile men can be pre screened before undergoing ART. Just as with genetic mutations, ART embryos can also be screened for imprinting errors so that the best possible embryos are used. This should reduce the number of failed pregnancies and the number of cases of BWS and AS. The next step forward will be to develop techniques to recover the 'good' sperm from within the oligospermic samples.