The neonatal PKU card is usually the first interaction between a neonate and the health care system. The samples are drawn at a time where foetal growth and development has ended and just after the child has undergone the trauma of birth. While it represents a convenient and systematic source of DNA for studies of genomic variation, the prospects are much more exciting when studying congenital predispositions/diseases and birth trauma.

2017.06.01 | Annette Bang Rasmussen

Section leader Jonas Bybjerg-Grauholm, Section for Neonatal Genetics, Statens Serum Institut

About the study

Building on our 2015 exploration of RNA array expression we used a simple test system of five males and females to demonstrate the feasibility of doing RNA-seq on PKU samples. To avoid analytical bias we followed the pipeline proposed by Trapnell et al in their 2012 Nature Protocol paper point-by-point.

We find a perfect concordance between sex called on XY-SNPs and at transcription levels. From this, we propose a list of six genes expressed at suitable levels to determine genders in neonates.

It is a well-established fact that the vast majority of reads from total RNA are of a ribosomal/transfer origin. A further issue for transcription studies on whole blood is that the majority of mRNA encodes globin transcripts. For these reasons, it is advisable to do both globin and rRNA depletion when working with blood samples. For neonates this is further complicated as the dominant species of foetal haemoglobin is subtly different from adult globin. The perhaps most important conclusion in this paper is that commercial Globin depletion kits developed for depletion of adult blood are sufficient to ensure the depletion of foetal globin as this omits the need for costly custom procedures.

The article “RNA sequencing of archived neonatal dried blood spots” was published in Mol Genet Metab Rep. 2017;10:33-37.

Facts about the study

• First reported proof-of-principle that RNA-seq is possible on DBS/PKU samples
• Feotal Globin is subtly different from adult globin, yet similar enough to be depleted with the same protocol. 
• Unfiltered we detected 64% (SD = 15%, range 32–81%) of the 26,799 features contained in the mapping reference
• HBA1 is a candidate to distinguish gender in neonates.

The work briefly presented here was featured by RNA-seq blog and later by Videnskab.dk (in Danish) Life lessons learned, when talking to reporters the things they emphasise might not be what you expected. 

Further information

Jonas Bybjerg-Grauholm, Section leader: Section for Neonatal Genetics, Statens Serum Institut. Email: JORG@ssi.dk