Oxford Nanopore long-read sequencing with CRISPR/Cas9-mediated target selection for accurate characterization of copy number variants in the LDLR gene

Abstract

Introduction

Familial hypercholesterolemia (FH) affects around 1 in 250 people. Most FH cases are caused by pathogenic LDLR variants, with copy number variations (CNVs) accounting for about 10 %. However, short-read gene panel sequencing and multiplex ligation-dependent probe amplification (MLPA) are limited in the specification of CNV breakpoints and the identification of complex structural variants (SVs).

Materials and methods

We designed crRNAs for Cas9-mediated target selection of LDLR and performed long-read sequencing (LRS) on an Oxford Nanopore MinION device using high-molecular-weight (HMW) DNA or DNA from standard purification. After establishing the LRS approach, we characterized two known LDLR CNVs and tested two individuals with strong clinical evidence of FH but no pathogenic variant in short-read gene panel sequencing.

Results

Complete coverage of LDLR was achieved for both HMW DNA and DNA from standard purification. LRS allowed us to specify CNV breakpoints and showed that the known LDLR deletion is 19.2 kb in size encompassing exons 1–2 and the 5′-untranslated and promoter regions. Furthermore, LRS verified the in tandem localization of a large LDLR duplication covering exons 4–8. Both CNVs were classified as loss-of-function. Moreover, breakpoint information enabled confirmatory analysis by PCR and Sanger sequencing for both CNVs. No SVs were detected in two apparently mutation-negative FH probands using our approach.

Conclusions

Nanopore LRS with CRISPR/Cas9-mediated target selection allows for accurate characterization of CNVs and can therefore serve as a complementary method to short-read sequencing-based FH diagnostics by facilitating variant interpretation and enabling cost-effective PCR-based variant confirmation in subsequent familial analyses.