Nanopore sequencing of the glucocerebrosidase (GBA) gene in a New Zealand Parkinson's disease cohort

https://doi.org/10.1016/j.parkreldis.2019.11.022Get rights and content

Highlights

  • Validated nanopore sequencing of the GBA gene for mutation detection.

  • 9.2% of patients from a New Zealand Parkinson's disease cohort carry GBA mutations.

  • Significantly higher prevalence of dementia in patients carrying GBA mutations.

  • Two novel mutations were detected including a putative splicing variant.

Abstract

Introduction

Bi-allelic mutations in the gene for glucocerebrosidase (GBA) cause Gaucher disease, an autosomal recessive lysosomal storage disorder. Gaucher disease causing GBA mutations in the heterozygous state are also high risk factors for Parkinson's disease (PD). GBA analysis is challenging due to a related pseudogene and structural variations (SVs) that can occur at this locus. We have applied and refined a recently developed nanopore DNA sequencing method to analyze GBA variants in a clinically assessed New Zealand longitudinal cohort of PD.

Method

We examined amplicons encompassing the coding region of GBA (8.9 kb) from 229 PD cases and 50 healthy controls using the GridION nanopore sequencing platform, and Sanger validation.

Results

We detected 23 variants in 21 PD cases (9.2% of patients). We detected modest PD risk variant p.N409S (rs76763715) in one case, p.E365K (rs2230288) in 12 cases, and p.T408 M (rs75548401) in seven cases, one of whom also had p.E365K. We additionally detected the possible risk variants p.R78C (rs146774384) in one case, p.D179H (rs147138516) in one case which occurred on the same haplotype as p.E365K, and one novel variant c.335C > T or p.(L335 = ), that potentially impacts splicing of GBA transcripts. Additionally, we found a higher prevalence of dementia among patients with GBA variants.

Conclusion

This work confirmed the utility of nanopore sequencing as a high-throughput method to identify known and novel GBA variants, and to assign precise haplotypes. Our observations may contribute to improved understanding of the effects of variants on disease pathogenesis, and to the development of more targeted treatments.

Introduction

Several monogenic forms of Parkinson's disease (PD) and many genetic risk factors, which predispose to development of PD, are now known. One of the strongest genetic risk factors for PD is mutation of the gene known as GBA, which encodes the lysosomal enzyme glucocerebrosidase (GCase) [1]. GCase is an enzyme involved in glycolipid metabolism, and mutations in GBA lead to a marked decrease in GCase activity [2]. Bi-allelic mutations in GBA result in excessive accumulation of the GCase substrate, GlcCer in macrophages, causing the rare lysosomal storage disorder, Gaucher disease (GD) [3,4]. Heterozygous GBA mutations were initially believed to be benign; however, Parkinsonism was frequently reported in GD patients and their non-GD family members [5]. It is now clear that individuals with these GBA mutations associated with GD in the heterozygous state have a substantially elevated risk of developing PD compared with non-carriers [3,6]. Furthermore, PD patients with GBA mutations have a noticeably earlier age of PD onset and a slightly increased risk of cognitive effects such as dementia, as well as a higher incidence of psychosis and delirium [7,8].

Over 300 different GBA variants have been described, including single nucleotide polymorphisms (SNPs) (the majority of which are missense variants), frameshift mutations, splice-site alterations, and a range of structural variants (SVs) (i.e. variations >50bp) [4]. However, analysis of GBA is not straightforward. A nearby pseudogene called GBAP1 shares 96% sequence homology with the coding region of the GBA gene, and care is required to prevent pseudogene contamination when analyzing GBA [9,10].

Although application of a long-PCR step targeted at GBA can effectively solve many (but not all) of the issues arising from structural variants and the presence of a pseudogene [10], challenges remain using short read sequencing with such amplicons, including the inability to precisely assign haplotypes. Long read sequencing approaches such as Oxford Nanopore Sequencing (Oxford Nanopore Technologies; ONT) and PacBio (Pacific Biosciences) offer new opportunities for analyzing complex genes such as GBA. Indeed a recent report by Leija-Salazar et al. described a procedure for carrying out genotyping and haplotyping of GBA on the MinION sequencer (ONT) [11]. We have further refined the method of Leija-Salazar [11], and applied it for the discovery of GBA variants in a New Zealand PD longitudinal cohort. Our refined protocol utilizes updated hardware and software, thereby reducing the computational workload and improving sequencing accuracy.

Section snippets

Cohort description, recruitment and ethics

DNA for GBA genotyping was from a convenience sample (n = 229) of Parkinson's disease patients from the New Zealand Brain Research Institute (NZBRI) PD cohort recruited through our movement disorders clinic for a range of PD-related research studies. We selected this convenience sample as our lab had previously collected blood samples from these patients for another study. Patients met the UK Brain Bank criteria for PD. Exclusion criteria included an indication of an atypical parkinsonian

Nanopore sequencing of GBA in NZBRI cohort

We applied the nanopore sequencing method to examine GBA amplicons derived from 229 PD and 50 healthy control DNA samples from the NZBRI cohort. We sequenced the amplicons over five libraries containing multiplexed samples, on separate flow cells. All of the nanopore runs resulted in similar yields (~10–15 Gb with QC ≥ 7) (Supplementary Fig. 1). The majority of these were the 8.9 kb fragment of interest (Supplementary Fig. 2). For the first run, ~60% of reads mapped to the GBA reference genome,

Nanopore sequencing as a tool for GBA mutation discovery

We successfully validated the protocol of Leija-Salazar et al. [11] as an effective method for genotyping of GBA. Additionally, we have updated the software pipeline utilized in the method of Leija-Salazar et al. [11], improving the accuracy and reducing the computational workload. More specifically our pipeline utilizes Guppy (flip-flop) instead of the now discontinued base-calling tool Albacore. Guppy runs orders of magnitude faster (1,500,000 bp/s vs 120,000 bps/s) due to its use of GPU

Conclusions

We have successfully validated ONT sequencing of GBA as an efficient and accurate method for detecting a wide range of variants in multiple patient samples, as well as resolving the haplotype of these variants. Applying this method to a sizeable New Zealand cohort of PD revealed the presence of multiple known GBA variants, and a range of additional variants of unclear clinical significance.

These findings may contribute to a better understanding of the effects of these variants on disease

Funding

Support for this work came from the Jim and Mary Carney Charitable Trust (Whangarei, New Zealand), The McGee Fellowship (University of Otago, Christchurch), and The Helen Poole and Ian McDonald Memorial Summer Studentship (Canterbury Medical Research Foundation, Christchurch, New Zealand).

Acknowledgements

The Health Research Council and Lotteries Health provided funding for sample collection and storage. We would also like to acknowledge NZBRI Assistant Research Fellows Leslie Livingston, Sophie Grenfell, and Bob Young for cognitive and clinical assessments, as well as NZBRI Research Fellows Dr Kyla Horne for cognitive testing, and Dr Michael MacAskill and Dr Daniel Myall for cohort database management.

References (30)

  • S.Y. Jeong

    Identification of a novel recombinant mutation in Korean patients with Gaucher disease using a long-range PCR approach

    J. Hum. Genet.

    (2011)
  • M. Leija-Salazar

    Evaluation of the detection of GBA missense mutations and other variants using the Oxford Nanopore MinION

    Mol. Genet. Genom. Med.

    (2019)
  • K.-L. Wood

    Different PD-MCI criteria and risk of dementia in Parkinson's disease: 4-year longitudinal study

    Npj Parkinson's Dis.

    (2016)
  • S.A. Miller et al.

    A simple salting out procedure for extracting DNA from human nucleated cells

    Nucleic Acids Res.

    (1988)
  • R. Nagar et al.

    DNA Size Selection (>3-4kb) and Purification of DNA Using an Improved Homemade SPRI Beads Solution

    (2018)
  • Cited by (14)

    • Evaluation of nanopore sequencing technology to identify Salmonella enterica Choleraesuis var. Kunzendorf and Orion var. 15<sup>+</sup>, 34<sup>+</sup>

      2021, International Journal of Food Microbiology
      Citation Excerpt :

      It was shown in our previous study that the workflow including the R9 flow cell, rapid library construction kit and guppy with the transducer basecalling model was sufficient for in silico serotyping prediction using ONT sequences (Xu et al., 2020). A high accuracy (HAC) basecalling model (flip-flop), which was trained on a corpus of sequence contexts that cover a variety of pro- and eukaryotic organisms, was recently released and expected to generate data with higher accuracy than using the transducer model (Graham et al., 2020). Therefore, in workflow 1, the combination of the R9 flow cell, rapid kit and guppy with the HAC basecalling model was applied.

    • The New Zealand Parkinson’s progression programme

      2023, Journal of the Royal Society of New Zealand
    View all citing articles on Scopus
    View full text