SAN FRANCISCO (GenomeWeb) – Researchers from the Parkinson's Institute and Clinical Center in Sunnyvale, California have shown that a CRISPR/Cas9-based capture enrichment technique, combined with sequencing on Pacific Biosciences' platform, can identify pathogenic repeat expansions, including slight differences in the expansions that can cause different phenotypes.
The team used the technology to sequence through long repeat expansions in the ATXN10 gene of six members of the same family, identifying a likely molecular explanation for the family's unusual history of Parkinson's disease and spinocerebellar ataxia.
The work, described last week in the journal NPJ Parkinson's Disease, is one of the first examples of using the CRISPR/Cas9 technology to design a targeted sequencing assay. A Stanford group previously described using CRISPR/Cas9 to target short tandem repeats.
PacBio is now working to develop the method into a commercial product. Kevin Corcoran, senior vice president of market development at PacBio, said that the company is working to refine its protocol at several early-access sites and expects to offer broader commercial access during the first part of 2018.
Birgitt Schüle, co-lead author of the study and an associate professor at the Parkinson's Institute and Clinical Center, said that the center initially saw one of the patients included in the study about 10 years ago. Although the Parkinson patient had been responding to therapy, there were two unusual things about his disease that the team wanted to study further: he was 37 years old when he first started having symptoms, which is young, and he had an "unusual family history" with four siblings with a different movement disorder, spinal cerebellar ataxia. "That was very puzzling, and we wanted to find out why," Schüle said.
Standard genetic testing involved repeat prime PCR, which identified a repeat expansion in the ATXN10 gene in each affected family member. Repeat expansions in the ATXN10 gene have been implicated in SCA, but have not previously been linked to Parkinson's disease, so the researchers wanted to look deeper. Because the PCR-based technology could not span the entire repeat, the team turned to PacBio for its long reads, and the firm suggested using the CRISPR/Cas9 technology it was developing as a target enrichment method.
Such a target enrichment scheme is advantageous because it does not require amplification, "which is especially important in error-prone regions with repeat expansions," Schüle said. It enables regions that are difficult to amplify to be targeted and sequenced.
In order to use CRISPR/Cas9 for target enrichment, genomic DNA is first fragmented and a PacBio adaptor known as a SMRTbell is attached, per the standard PacBio library prep protocol, Jonas Korlach, PacBio's chief scientific officer, explained. Then, guide RNAs target and selectively cut just outside the region of interest. That generates target molecules that have one end with the original adaptor and a second end containing the sequence cut by Cas9. Next, a capture adaptor is ligated onto the cut end, so that only molecules containing the targeted region have those capture adaptors. And finally, magnetic beads and probes are used to capture just the target molecules, while the remaining DNA is washed away. The target molecules are then sequenced.
In this case, using the CRISPR/Cas9-based target enrichment and PacBio sequencing for the patient and his siblings, the researchers found that although all six siblings did indeed have repeat expansions in the ATXN10 gene, there were slight differences. The four siblings who had a form of spinocerebellar ataxia known as SCA10 had interruptions in the repeats, Schüle said, which was not an unusual finding, since such interruptions have been previously documented and are thought to be genetic modifiers that can contribute to phenotypic differences. The patient with Parkinson's disease had no such interruption and instead had a 5-nucleotide sequence, ATTCT, that repeated more than 1,300 times, Schüle said. For his siblings, the first 40 percent of the expansion consisted of ATTCT repeats, while the remainder switched to ATTCC.
Schüle noted that the researchers also analyzed 188 genes that have been previously associated with Parkinson's disease, Alzheimer's disease, and amyotrophic lateral sclerosis, in order to rule out other explanations, but they were all negative.
She said that the researchers think the differences in the repeat expansions could explain the different phenotypes, but added that her team plans to do functional studies as well. Typically, she said, Parkinson's disease and SCA10 impact different brain regions, so her team is interested in designing pluripotent stem cell models from the patients and studying functional differences and the different proteins that bind to the different repeats.
In addition, she noted, there was one unaffected sister who was also found to have the same "pure" repeat expansion as the individual with Parkinson's disease. The sister, who is already 53 years old, has "one of the longest repeat expansions I've seen for someone who is unaffected," Schüle said. Perhaps she will develop symptoms later in life, but all of the other family members did so before the age of 48.
Schüle said her team now plans to screen around 700 additional Parkinson's patients for the ATXN10 repeat expansion. She said they would do the initial screen using a PCR assay to look for a certain haplotype and then follow up with the same PacBio/CRISPR/Cas9 targeted sequencing strategy for the entire expansion. In addition, she hopes to continue to work with PacBio to study other large structural variants.
For this study, the Parkinson's Institute team collaborated directly with PacBio, which captured and sequenced the repeats. Korlach said that the CRISPR/Cas9 system is an advantage for target enrichment because it does not need amplification. As such, it can work on regions that are typically difficult to amplify — such as repetitive regions, regions of high homology, and pseudogenes. PacBio sequencing provides the extra advantage of long reads, which enables such regions to be sequenced completely, he added.
The main hurdle of using CRISPR/Cas9 for target enrichment has been figuring out the off-target effects, Korlach said. For instance, in the firm's early development of the technology, the guide RNAs sometimes landed on a different region of the genome than the target, which he said the company has fixed by designing more specific guide RNAs.