Alysha S. Taylor

Expanded CAG/CTG repeats cause more than 15 different diseases that all remain without a disease-modifying treatment. Because repeat length accounts for most of the variation in disease severity, contracting CAG/CTG repeats presents an attractive therapeutic avenue. Here, we show that the CRISPR-Cas9 nickase targeted to CAG/CTG repeats leads to efficient repeat contractions in Huntington’s disease (HD) patient–derived neurons and astrocytes and in myotonic dystrophy type 1 patient–derived neurons. The approach was allele selective and free of detectable off-target mutations. Injection of the Cas9 nickase into the striatum of the R6/1 mouse model of HD using adeno-associated viral vectors led to repeat contractions in more than half of striatal neurons. Upon injection, we observed a reduction in the number of inclusion bodies, an improved gene expression profile, and amelioration of locomotion deficits in the open field test. The effects were greater than expected from the contractions induced, suggesting that non–cell autonomous mechanisms may be involved. Our results provide the proof of concept that correction of CAG/CTG repeats can improve HD phenotypes in a mouse disease model.

Expanded repeat disorders remain without a disease modifying treatment. Some of the most important modifiers of disease severity include inherited repeat size, their somatic mosaicism, and whether they contain interruptions. Targeted sequencing approaches are becoming the gold standard on how these parameters are measured despite repeats being notoriously challenging to sequence. Here we developed CHARLIE (Comprehensive High-throughput Analysis of Repeat Length, Interruptions and Expansions) to identify and position interruptions within expanded repeats. We used samples from Huntington's disease and myotonic dystrophy type 1 sequenced with Illumina MiSeq and PacBio's Single-Molecule Real-Time Sequencing. We validated the pipeline against previous methods for identifying germline-inherited interruptions. One challenge that can potentially mask the identification of true interruptions and sequence variation is the accuracy of sequencing platforms, which, when applied to expanded repeats, is largely unknown. Our results suggest that each sequencing platform produces a distinct error profile and we show that PCR-free library preparation for SMRT sequencing improves sequencing accuracy. CHARLIE, therefore, provides a method that can help differentiate between trivial sequencing errors and those interruptions that modify disease presentation, which can be inherited or somatic in origin.

Expanded CAG/CTG repeats cause over 15 different diseases that all remain without a disease-modifying treatment. Because repeat length accounts for most of the variation in disease severity, contracting them presents an attractive therapeutic avenue. Here, we show that the CRISPR-Cas9 nickase targeted to CAG/CTG repeats leads to efficient contractions in Huntington’s disease patient-derived neurons and astrocytes, and in myotonic dystrophy type 1 patient-derived neurons. The approach is allele-selective and free of detectable off-target mutations. Striatal injection of the Cas9 nickase in a mouse model for Huntington’s disease using adeno-associated viral vectors led to contractions in over half the infected cells. Upon injection, we observed a reduction in the number of inclusion bodies, improved transcriptome, and ameliorated locomotion. The effects were greater than expected from the contractions induced and suggest that non-cell autonomous mechanisms may be involved. Our results provide the proof-of-concept that correction of CAG/CTG repeats can improve Huntington’s disease phenotypes in vivo . One sentence summary: The Cas9 nickase contracts CAG/CTG repeats at multiple disease loci in patient-derived cells and improves molecular and behavioral phenotypes in HD.

Targeted DNA sequencing approaches will improve how the size of short tandem repeats is measured for diagnostic tests and preclinical studies. The expansion of these sequences causes dozens of disorders, with longer tracts generally leading to a more severe disease. Interrupted alleles are sometimes present within repeats and can alter disease manifestation. Determining repeat size mosaicism and identifying interruptions in targeted sequencing datasets remains a major challenge. This is in part because standard alignment tools are ill-suited for repetitive and unstable sequences. To address this, we have developed Repeat Detector (RD), a deterministic profile weighting algorithm for counting repeats in targeted sequencing data. We tested RD using blood-derived DNA samples from Huntington’s disease and Fuchs endothelial corneal dystrophy patients sequenced using either Illumina MiSeq or Pacific Biosciences single-molecule, real-time sequencing platforms. RD was highly accurate in determining repeat sizes of 609 blood-derived samples from Huntington’s disease individuals and did not require prior knowledge of the flanking sequences. Furthermore, RD can be used to identify alleles with interruptions and provide a measure of repeat instability within an individual. RD is therefore highly versatile and may find applications in the diagnosis of expanded repeat disorders and in the development of novel therapies.

Targeted DNA sequencing approaches will improve how the size of short tandem repeats is measured for diagnostic tests and pre-clinical studies. The expansion of these sequences causes dozens of disorders, with longer tracts generally leading to a more severe disease. Interrupted alleles are sometimes present within repeats and can alter disease manifestation. Determining repeat size mosaicism and identifying interruptions in targeted sequencing datasets remains a major challenge. This is in part because standard alignment tools are ill-suited for repetitive and unstable sequences. To address this, we have developed Repeat Detector (RD), a deterministic profile weighting algorithm for counting repeats in targeted sequencing data. We tested RD using blood-derived DNA samples from Huntington’s disease and Fuchs endothelial corneal dystrophy patients sequenced using either Illumina MiSeq or Pacific Biosciences single-molecule, real-time sequencing platforms. RD was highly accurate in determining repeat sizes of 609 blood-derived samples from Huntington’s disease individuals and did not require prior knowledge of the flanking sequences. Furthermore, RD can be used to identify alleles with interruptions and provide a measure of repeat instability within an individual. RD is therefore highly versatile and may find applications in the diagnosis of expanded repeat disorders and the development of novel therapies.

Strigolactones (SLs) are an important class of carotenoid-derived signalling molecule in plants, which function both as exogenous signals in the rhizosphere and as endogenous plant hormones. In flowering plants, SLs are synthesized by a core pathway of four enzymes and are perceived by the DWARF14 (D14) receptor, leading to degradation of SMAX1-LIKE7 (SMXL7) target proteins in a manner dependent on the SCFMAX2 ubiquitin ligase. The evolutionary history of SLs is poorly understood, and it is not clear whether SL synthesis and signalling are present in all land plant lineages, nor when these traits evolved. We have utilized recently-generated genomic and transcriptomic sequences from across the land plant clade to resolve the origin of each known component of SL synthesis and signalling. We show that all enzymes in the core SL synthesis pathway originated at or before the base of land plants, consistent with the previously observed distribution of SLs themselves in land plant lineages. We also show that the late-acting enzyme LATERAL BRANCHING OXIDOREDUCTASE (LBO) may be considerably more ancient than previously thought. We perform a detailed phylogenetic analysis of SMXL proteins and show that specific SL target proteins only arose in flowering plants. We also assess diversity and protein structure in the SMXL family, identifying several previously unknown clades. Overall, our results suggest that SL synthesis is much more ancient than canonical SL signalling, consistent with the idea that SLs first evolved as rhizosphere signals and were only recruited much later as hormonal signals.