The years of our fathers: paternal age and the rate of de novo mutations

Aging fathers. An increase in risk of aneuploidies, i.e. chromosomal aberrations such as Trisomy 21, is well established with maternal age.  Whether the paternal age also increases the risk for disorders in the offspring had long been disputed. However, a connection between paternal age and autism has been found in recent years. Now a recent study in Nature finds a surprisingly strong correlation on the genetic level…

Whole genome sequencing in trios. Kong and colleagues performed full genome sequencing in 78 trios with autism and schizophrenia. This data allowed them to determine the genome-wide rate of de novo mutations. They found an average of 70 new mutations per individual on a genome-wide level. From exome studies, we already know that between three and five of these variants will be within exons, leading to base pair changes, which might influence protein function. While the exomic de novo rate is too small to establish sufficiently powered correlations with factor that influence the rate of de novos, this can be performed with the genome-wide data. When Kong and colleagues plotted the rate per individual versus the paternal age, the found and amazingly linear correlation.

Paternal age impacts on the de novo rate. The correlation between paternal age and the frequency of de novo mutations is as strong as a correlation might biologically be. Predictions by Kong et al. suggest that there might be little influence except the paternal age to influence the rate of de novo mutations. The rate of de novo mutations per genome virtually doubles when the paternal age increases from 20 to 40. With an age of 80, the rate is 10x as high. And most of these mutations are transmitted from the paternal germline.

The correlation between paternal age and rate of de novo variants is quite strong. The resulting de novo mutations might be the cause of the disease, particularly in CNS disorders.

The impact on human disease. Kong and colleagues find some interesting mutations as possible candidates for autism and schizophrenia in their sample. This finding highlights that an increase in mutations rate also increases the risk for disease. In addition, as the human brain is far more susceptible to de novo mutations than other organ systems, the impact on CNS disorders might be particularly relevant. Therefore, diseases like autism or schizophrenia might be ideal candidates.

The impact for EuroEPINOMICS. In the same way that de novo mutations might lead to autism, de novo mutations may also occur in epilepsy candidate genes. Therefore, the analysis of trio exome data for de novo variants has a top priority in the trio analysis of the EuroEPINOMICS project, and we have already informed the partners about the first results. On a technical exome level, the analysis of de novo variants is both interesting and quite difficult. Exome data is notoriously noisy and filtering for variants present in the child but absent in the parents would automatically results in hundreds, if not thousands of false positive findings, i.e. genomic noise. Modern algorithms for de novo calling such as the Sanger inhouse program DeNovoGear therefore need to take into account various quality measures. Whichever programs are used by researchers to identify de novo variants, I find the mere existence of these technologies breathtaking.

Ingo Helbig

Child Neurology Fellow and epilepsy genetics researcher at the Children’s Hospital of Philadelphia (CHOP), USA and Department of Neuropediatrics, Kiel, Germany

Facebook Twitter