Characterizing circular and linear non-coding RNA in human aged brains and their association with dementia

Total RNA, depleted of ribosomal RNA, is commonly used for RNA sequencing to capture both polyadenylated and non-polyadenylated transcripts, whereas poly(A) tail-targeting protocols can either enrich (Poly(A)-selection) or deplete (Poly(A)-depletion) the polyadenylated fraction. To ascertain the best library strategy for capturing long non-coding RNAs (lncRNAs, > 200nt) and circular RNAs (circRNAs) in aged brains, we applied three protocols (total ribo-depleted, poly(A)-selected, and poly(A)-depleted) to 128 human hippocampus samples. Our results show that both ribo-depleted and poly(A)-depleted perform well for lncRNA and circRNA profiling, with the poly(A)-depleted protocol generally detecting a slightly higher number of non-coding RNAs. Additionally, we compared poly(A)-selected and -depleted data for predicting the poly(A) status of hippocampal transcripts. The proportion of non-polyadenylated transcripts was higher than previously found. This study demonstrates how the choice of library construction affects outcomes, helping to make informed methodological choices when studying brain non-coding RNAs. In addition, we examined, for the first time, whether the poly(A) status of protein-coding mRNAs is associated with AD and AD-related dementias.