LONG-NEXT: A new accurate and efficient NGS-based method for GBA1 analysis in Parkinson disease

Abstract

Introduction

GBA1 variants are the most common genetic risk factor for Parkinson disease (PD). Sequencing of GBA1 on a large scale represents a burdensome task with currently adopted diagnostic techniques, namely Sanger sequencing and conventional short read next generation sequencing (sr-NGS). The high degree of sequence homology between GBA1 and its pseudogene GBA1LP is the major driver behind this complexity, leading to false positive and false negative results.

We designed, optimized and validated LONG-NEXT, a new NGS-based strategy to streamline large scale GBA1 sequencing.

Methods

LONG-NEXT relies on a specific long-range PCR, encompassing the whole GBA1 gene, in one fragment (6.5 kb), followed by short-read NGS and a tailored bioinformatic pipeline masking the GBA1LP sequence on the reference genome.

Results

This protocol was optimized and tested on selected cases suspected of diagnostic mistakes by conventional testing (n = 13) and then validated on consecutively collected PD patients already screened either by Sanger sequencing (n = 101) or conventional sr-NGS (n = 294). LONG-NEXT reanalysis of 13 patients disclosed: 3 false positive cases due to mismapping of pseudogene reads on GBA1, 4 false homozygotes due to PCR-related allele dropout events, and 6 false negative cases, due to misalignment of GBA1 reads against the pseudogene or PCR-related allele dropout events. The validation phase disclosed one additional false homozygote in the Sanger cohort, and one false negative result in the sr-NGS cohort.

Conclusion

LONG-NEXT is a reliable, fast, cost-effective alternative for GBA1 sequencing and may prove strategic in light of current genotype-based tailored therapies specifically targeting GBA1-PD patients.