Genomic Repeats Link Cancer and Neurodegeneration

By correcting errors arising during DNA replication, DNA mismatch repair (MMR) is crucial for cancer avoidance. Inherited defects in the MMR genes MLH1, MSH2, MSH6, and PMS2 cause susceptibility to Lynch syndrome, a common form of cancer predisposition with an estimated population prevalence of up to 1:100. Apart from antimutator functions, MMR genes can act as mutators, e.g., by promoting trinucleotide repeat expansion.

Genomic repeats are linked to Lynch syndrome in at least four ways. First, 15-30% of Lynch syndromeassociated mutations are large rearrangements facilitated by Alu-Alu recombination. Second, failure of MMR function manifests itself as contractions or expansions of short repetitive sequences, microsatellites, in tumor tissues when compared to paired normal tissues (microsatellite instability). Third, selective involvement of coding microsatellite repeats within tumor suppressor genes may serve as organ-specific drivers of tumorigenesis, contributing to the Lynch syndrome tumor spectrum. Fourth, frameshift mutations of coding repeats give rise to neoantigens and elicit immune responses with prognostic and therapeutic relevance.

In Lynch syndrome, one allele of a given MMR gene is defective from birth, and there are no signs of disease until the remaining wild-type allele is inactivated in a target organ. Recent findings suggest that biallelic constitutional inactivation of MMR genes causes susceptibility to intestinal polyposis.

Acute myeloid leukaemia (AML) is a highly aggressive haematopoietic malignancy, defined by a series of genetic and epigenetic alterations. These alterations result in dysregulation of transcriptional networks in AML. One understudied but important source of transcriptional regulators are transposons. Making up around half of the human genome, transposons are a rich source of tissue-specific cis-regulatory DNA sequences and have the potential to control cellular processes. Aberrant activity of these sequences could therefore contribute to oncogenic transcriptional circuits. Despite extensive functional genomic analyses of AML, crucially the contribution of transposons to AML pathogenesis is currently unknown. Using cell line models and publicly available AML patient data, we demonstrated that certain transposon families are epigenetically activated and act as novel enhancers in AML. Using both locus-specific genetic deletion and simultaneous epigenetic silencing of multiple transposons, we showed that transposon deregulation directly alters the expression of adjacent genes in AML. Strikingly, deletion or epigenetic silencing of a transposon-derived enhancer suppresses cell growth in leukaemia cell lines through modulating expression of apolipoprotein genes. Our work reveals that transposons are a previously unappreciated source of AML enhancers that have the potential to play key roles in leukemogenesis.
Currently, we are investigating the regulatory roles of transposons and their contribution to cellular phenotype in a poor-risk group of AML, which harbors concurrent mutations in NPM1, DNMT3A and FLT3 genes (‘NDFmut’). NPM1, DNMT3A and FLT3 represent the most frequent three-gene co-occurrence in AML. Concurrent mutations in these genes have distinct clinical relevance, with a very poor clinical outcome in patients. With a combination of computational and experimental approaches, we have identified transposons that are epigenetically activated in this poor-risk group. Using our comprehensive epigenetic and transcriptional analyses, we are currently assessing the contribution of transposons to the NDFmut-driven gene expression changes and characterizing transcription factors that play a driver role in stimulating transposon regulatory function in NDFmut AML.

Friedreich’s ataxia (FRDA) is a repeat expansion disease (RED) caused by the expansion of GAA repeats in the first intron of the frataxin gene. REDs are characterized by stable secondary structure-forming repetitive sequences that expand both between generations and within somatic cells. Somatic expansions that accumulate throughout a patient’s lifetime can explain two characteristics of REDs: the tissue-specificity and the delayed age of onset. In FRDA, affected tissues are the pancreas, heart, and the nervous system. Notably, these tissues consist of largely post-mitotic cells. It is important, therefore, to elucidate the mechanisms by which repeats can expand in nondividing cells and evaluate the entire mutational spectrum at the repeats. We use the chronological aging of quiescent yeast containing GAA repeats to investigate these mechanisms. Although expansion, deletion, and recombination events occur in both dividing and nondividing yeast, the predominant event shifts from expansions in dividing cells to deletions in nondividing cells. Delving into the genetic control of these events revealed that mismatch repair facilitates a repeat-mediated double-strand break. This double-strand break is then healed by non-homologous end-joining to yield deletions or by double-strand break repair to yield recombination events. Expansions, in turn, occur in the absence of mismatch repair machinery via gap repair executed by Pol delta. We believe that investigation of GAA-mediated genome instability in nondividing yeast may shed light on the mechanisms crucial for disease onset in FRDA patients.

Somatic expansion of the CAG repeat tract that causes Huntington’s disease (HD) is thought to contribute to the rate of disease pathogenesis. Therefore, factors influencing repeat expansion are potential therapeutic targets. Genes in the DNA mismatch repair pathway are critical drivers of somatic expansion in HD mouse models. Here, we have tested, using genetic and pharmacological approaches, the role of the endonuclease domain of the mismatch repair protein MLH3 in somatic CAG expansion in HD mice and patient cells. A point mutation in the MLH3 endonuclease domain completely eliminated CAG expansion in the brain and peripheral tissues of a HD knock-in mouse model (HttQ111). To test whether the MLH3 endonuclease could be manipulated pharmacologically, we delivered splice switching oligonucleotides in mice to redirect Mlh3 splicing to exclude the endonuclease domain. Splice redirection to an isoform lacking the endonuclease domain was associated with reduced CAG expansion. Finally, CAG expansion in HD patient-derived primary fibroblasts was also significantly reduced by redirecting MLH3 splicing to the endogenous endonuclease domain-lacking isoform. These data indicate the potential of targeting the MLH3 endonuclease domain to slow somatic CAG repeat expansion in HD, a therapeutic strategy that may be applicable across multiple repeat expansion disorders.

Many cancers undergo epigenetic changes that permit expression of otherwise silenced transposable elements. In this talk, I will focus on Long INterspersed Element-1 (LINE-1, L1), the only protein-coding retrotransposon active in modern humans. L1-encoded open reading frame 1 protein (ORF1p) expression is a hallmark of human cancers. We know that ORF1p is an indicator of L1 activity as a mobile genetic element, i.e., cancers that express ORF1p have somatically-acquired insertions of L1 sequences that distinguish tumor genomes from a patient’s germline DNA. I will review data showing somatically-acquired L1 insertions in pancreatic and ovarian cancers, as well as a pan-cancer International Cancer Genome Consortium project. Together, these studies have shown that L1 expression is commonplace in cancers, and that it contributes to genome instability.

I will also describe our ongoing efforts to understand implications of L1 expression for cancer cell biology. We have found that non-transformed cells undergo a p53-dependent growth arrest when L1 expression is forced, and that L1 can induce interferon responses like those elicited by viral infection. In models where these anti-proliferative effects are overcome, L1(+) cells appear especially vulnerable to loss of replication-coupled DNA repair pathways, and DNA-damaging chemotherapies. These findings indicate that L1 retrotransposition may conflict with DNA replication in a manner that can be exploited for cancer therapeutics.

DNA damage response (DDR) genes are modifiers of onset and severity of DNA Repeat Expansion Disorders (REDs) including Huntington’s disease (HD), myotonic dystrophy, and spinocerebellar ataxias. DDR genes modulate somatic repeat expansion at disease causing loci (e.g. HTT, DMPK, ATXN1/2/3/7). Triplet Therapeutics is developing oligonucleotides (ASOs and siRNAs) to knockdown (KD) specific DDR genes selected for somatic instability phenotypes and loss of function tolerance. Oligonucleotides afford specificity, rapid preclinical development, and clinical validation in central nervous system (CNS) disorders. Triplet’s approach operates upstream of disease genes and addresses a unifying pathway driving multiple REDs. Data generated by Triplet and others show that lowering certain DDR genes by 50% halts somatic expansion in both HD patient-derived cells and HD model mice. Based on safety and efficacy (e.g. selectivity, distribution, durability), Triplet selected TTX-3360 as our first clinical candidate. Single and repeat doses of TTX-3360 administered via multiple routes to NHPs were well tolerated and drove significant target KD. Specifically, a single injection of TTX-3360 in NHPs lowered target mRNA by >50% in caudate and cortex. IND-enabling studies with TTX-3360 are underway. To support its Phase 1/2a, Triplet is conducting SHIELD HD, a natural history study in pre/early manifest HD patients to assess DDR gene expression, somatic expansion, and disease progression over time. SHIELD HD completed enrollment in 2020 and interim data will inform the Phase 1/2a design. This trial will support initiation of trials with TTX-3360 seeking to halt somatic expansion and progression in multiple indications.