2016 Conference | CHDI Foundation

CHDI’s 11^th Annual HD Therapeutics Conference took place February 22 – 25, 2016, in Palm Springs, California. This unique conference series focuses on drug discovery and development for Huntington’s disease, and draws participants and speakers from the biotech and pharmaceutical sectors as well as academia and research institutions. The conference is intended as a forum where all participants can share ideas, learn about new disciplines, network with colleagues and build new collaborative partnerships. We are indebted to all of the conference speakers, and especially grateful to those who are able to make their presentations available here for a wider audience.

An update on CHDI’s preclinical activities: biology

Thomas F. Vogt, PhD – CHDI

Top

An update on CHDI’s preclinical activities: chemistry

Celia Dominguez, PhD – CHDI

Top

An update on CHDI’s clinical activities

Cristina Sampaio, MD, PhD – CHDI

Top

Effect of polyQ expansions on the solution structure and misfolding of huntingtin

Ralf Langen, PhD – University of Southern California

Huntington’s disease is caused by polyQ expansions in the huntingtin (htt) protein. While the mechanisms of toxicity remain under investigation, it is likely that the polyQ expansions alter huntingtin structure and cause the formation of toxic states. Such toxic states could include monomeric protein with altered functional properties or oligomeric/brillar misfolded forms of htt. Restoring the conformational properties of normal htt may therefore be a potential therapeutic avenue. Proof-of-principle for this notion comes from transthyretin-related hereditary amyloidosis (also called familial amyloid neuropathy, FAP), where a small molecule drug, tafamidis, maintains the native fold and prevents the conversion into toxic misfolded forms. Structural information was essential for the design of tafamidis and will likely be essential for applying analogous approaches to HD.

Structural studies on htt or its exon1 (httex1) with polyQ expansions are complicated by the aggregation prone nature of the proteins and technical difficulties of many high-resolution structural methods. We therefore first developed methods for preparing stable monomeric, oligomeric or brillar species of httex1 and investigated their structures using a combination of biophysical methods including EPR, solution NMR, solid-state NMR, circular dichroism, fluorescence and electron microscopy. We found that monomeric httex1 is largely unfolded, but that it has residual α-helical structure in its N-terminal region and strong polyproline II helical structure in its C-terminal region. Interestingly, the α-helical structure increased with increasing polyQ length leading to increased rigidity that effectively freezes out the motion of httex1. This polyQ-dependent enhancement of rigidity could contribute to the toxicity by altering the functional or protein-protein interaction properties of httex1. By monitoring misfolding over time, we found that httex1 misfolding is a highly seeded process. While monomeric httex1 with expanded Q-lengths was highly stable for many hours even at tens of micromolar concentrations, the addition of seeds significantly sped up aggregation. Monitoring this seeded reaction has provided us with insights regarding the mechanism of misfolding and allowed us to investigate molecules that promote or inhibit misfolding. Finally, we have generated different types of misfolded httex1 fibrils and oligomers. These forms include oligomers and fibrils previously suggested to be toxic or only weakly toxic. In all cases, the central ß-sheet core region of the misfolded species was located in the polyQ region surrounded by extended polyproline II helix bristles. The central difference among the structures was that these bristles are solvent exposed in the putatively toxic structures while they are bundled up in the less toxic forms. These data suggest the possibility that interaction with the polyproline II bristles may additionally contribute to pathogenesis.

Top

Elucidating the sequence and structural determinants of huntingtin aggregation and toxicity using protein semisynthesis and single molecule techniques

Presentation not available

Hilal A. Lashuel, PhD – Ecole Polytechnique Fédérale de Lausanne

Increasing evidence suggest that post-translational modi cations (PTMs), such as phosphorylation, ubiquitination and SUMOylation, within the N-terminal 17 residues of the huntingtin protein play important roles in modulating the aggregation, subcellular localization and toxicity of full-length and N-terminal huntingtin protein fragments (Htt) in cellular and animal models of Huntington’s disease. However, the lack of knowledge about the natural enzymes involved in regulating these PTMs has hampered studies aimed at elucidating the molecular mechanisms by which these modifications influence the function of Htt in health and disease. To address this knowledge gap, our group has developed and optimized several protein expression and protein semisynthetic strategies and protocols that enable 1) the efficient and rapid production of recombinant expression of tag-free exon1 proteins with polyQ repeat lengths ranging from 6-49Q; 2) site-specific introduction of single or multiple PTMs within the Nt17 region of exon1. In collaboration with the chemistry group at IRBM and Evotec, we have further optimized these protocols to scale up the production of these proteins.

With these highly pure and homogeneous preparations of unmodified and modified forms of exon1 in hand, we applied interdisciplinary approaches that combine standard and advanced biophysical techniques, including circular dichroism (CD), transmission electron microscopy (TEM), atomic force microscopy (AFM), nano-IR spectroscopy and single molecule FRET (In collaboration with Edward Lemke, EMBL) to investigate the effect of polyQ repeat length, the Nt17 domain and its PTMs in regulating the structure of monomeric and aggregated forms of exon1 at the single molecule and nanometric scale. Together, these advances allowed, for the first time, dissection of the role of individual PTMs, including phosphorylation (at T3, S13 and S16), acetylation (K9), ubiquitination and SUMOylation (K6, K9, and K15), in regulating the structure and aggregation of exon1 in vitro and led to the identification of novel PTM-dependent structural switch elements involving cross-talk between different PTMs. These findings and their implication in understanding the sequence and structural determinants of Htt aggregation, toxicity and function in health and disease will be presented and discussed. Finally, preliminary studies conducted in collaboration with IRBM (Andrea Caricasole) demonstrate the great potential of using these highly pure and homogeneous preparations of exon1 as powerful tools to facilitate the development of novel sensitive assays for the detection and quantification of modified forms of Htt and to discover enzymes involved in regulating Htt phosphorylation.

Top

A structure/function analysis in knock–in mice of the huntingtin N–terminal domains encoded by exon 1

Scott Zeitlin, PhD – University of Virginia

To understand the role of the huntingtin (Htt) N-terminal domains in normal Htt function and in modulating HD mouse model pathogenesis, we have generated and are characterizing knock-in mouse models that express versions of normal Htt lacking the N17 domain (ΔN17), the polyQ stretch (ΔQ), the proline-rich region-or PRR (ΔP), a combination of both the polyQ and PRR domains (ΔQP), and all three N-terminal domains encoded by Htt exon 1 (ΔNQP). In homozygosity, all the deletion mutants are born at normal Mendelian frequency, suggesting that the exon-1 encoded domains of Htt are not required for normal embryonic development. As adults, the homozygous deletion mutants exhibit only subtle behavioral phenotypes, and lack obvious neuropathology. However, when the Htt N-terminal domain deletion mutants are expressed together with the CAG140 allele, the individual domain deletions have differential effects on HD mouse model phenotypes. While the Q deletion can rescue some aspects of pathogenesis, the N17 and QP deletions exacerbate some CAG140 phenotypes. To better understand the effect of these deletions on CAG140 pathogenesis, we are in the process of characterizing autophagy in primary striatal and cortical neuronal cultures obtained from postnatal day 5 Htt140Q/ΔN17, Htt140Q/ΔQP, wild-type, Htt140Q/ΔQ and Htt140Q/+ pups. Preliminary results suggest that while autophagic flux is enhanced in the Htt140Q/ΔQ cultures in comparison to controls, Htt140Q/ΔN17 cultures exhibit a significant deficit in autophagy.

We are also studying the effect of deleting the Htt PRR in cis to the expanded polyQ stretch in Htt140QΔP/+ mice, and the effect of replacing the murine PRR with the human PRR in the normal allele of Htt140Q/7QhuP mice. Consistent with the results obtained from analyzing Htt PRR deletions in other model systems, deletion of the PRR in cis to the 140Q stretch affects mutant Htt aggregation. Our preliminary data indicate that substituting a human PRR for the murine PRR in the normal allele of Htt140Q/7QhuP mice exacerbates HD model phenotypes in comparison to Htt140Q/+ controls. These results suggest that the Htt PRR can modulate the toxic effects of 140Q-Htt expression both in cis and in trans.

Top

Huntingtin and the control of intracellular dynamics: Role of proteolysis and non–polyQ fragments

Frédéric Saudou, PhD – Grenoble Institute of Neurosciences

Huntington’s disease is caused by the abnormal polyglutamine expansion in the N-ter part of huntingtin (HTT), a large protein of 350kDa. Over the past years, we proposed that HTT acts a scaffold for the molecular motors and, through this function, regulates the efficiency and directionality of vesicular transport along microtubules in neurons. This function is conserved in Drosophila. In particular, HTT controls the microtubule- based fast axonal transport (FAT) of neurotrophic factors such as BDNF. PolyQ expansion in HTT alters this function, leading to a decrease in neurotrophic support and death of striatal neurons. Interestingly, the defect in transport might not be restricted to axons but could also involve defects in the retrograde transport of TrkB in striatal dendrites.

In addition to the role of HTT in scaffolding the molecular motors both in cortical and striatal neurons, we found that HTT scaffolds GAPDH on vesicles and that vesicular GAPDH is necessary to propel vesicles in GAPDH deficient neurons. Here we will extend these findings and discuss how HTT, by specifically localizing the glycolytic machinery on vesicles, may supply constant energy for the transport of vesicles over long distances in axons. We will also discuss how this machinery is altered in disease.

Finally, we will extend the function of HTT as a scaffold for dynamin1 for regulating intracellular dynamics in health and disease. Indeed, we found, using time- and site-specific control of HTT proteolysis, that specific cleavages are required to disrupt intramolecular interactions within HTT and to cause toxicity in cells and flies. Surprisingly, in addition to the canonical pathogenic N-ter fragments, the C-ter fragments generated, that do not contain the polyQ stretch, induced toxicity via dilation of the endoplasmic reticulum (ER). C-ter HTT bound to dynamin 1 and subsequently impaired its activity at ER membranes. Our findings thus identify a new pathogenic mechanism in HD and suggest that non-polyQ fragments of HTT generated by proteolysis contribute to disease by altering ER dynamics and homeostasis.

Top

Biochemical–genetic investigation of the influence of the polyglutamine tract on huntingtin structure and function

Ihn Sik Seong, PhD – Massachusetts General Hospital

The CAG trinucleotide repeat expansion mutation that is the genetic cause of Huntington’s disease (HD) elongates a normally variable segment of polyglutamine located sixteen amino acids from the start methionine of the 350kD huntingtin protein to more than about forty residues. To understand this mutation at the level of the full-length huntingtin HEAT/HEAT-like domain molecule itself, we are using a classical biochemical-genetic strategy that couples the physical analyses of a series of highly purified full-length huntingtins with different polyglutamine tracts to the functional activity of these proteins in a validated cell-free assay (Seong et al. HMG (2010) 19:573-583): enhancement of polycomb repressive complex 2-dependent nucleosome deposition of the histone H3K27me3 mark (PRC2-assay). Our results reveal that full-length huntingtin, which is bisected by a protease sensitive hinge site at residues 1184-1254 into two large arms, assumes a spherical α-helical solenoid shape with a large internal cavity, that involves short-, medium- and long-range intramolecular contacts, with the latter mediating the interaction of the amino terminal arm with the carboxyl terminal arm. This 3D shape is globally similar regardless of polyglutamine size. However, polyglutamine length does alter the pattern of full-length huntingtin intramolecular contacts and the pattern of posttranslational modifications, including ‘crosstalk’ between distant phosphosites, such that huntingtin’s polyglutamine tract size and phosphorylation status is associated with the protein’s function activity. The 3D closed α-helical shape assumed by full-length huntingtin’s HEAT/HEAT-like rich domains, therefore, provides an elegant solution to the conundrum of full-length huntingtin as the polyglutamine tract size dependent disease-initiator. Our results pave the way for high-resolution studies that will be required to determine in detail the influence of its polyglutamine segment on the structure of full-length huntingtin, while our biochemical-genetic approach provides a route to identify factors that can selectively manipulate full-length mutant huntingtin without affecting normal full-length huntingtin activity.

Top

DNA methylation age analysis of human Huntington’s disease

Steve Horvath, PhD, ScD – University of California, Los Angeles

Biological tissue age seems to play an important role in Huntington’s disease (HD), e.g. the age of disease onset is strongly related to the number of CAG trinucleotide repeats in the HD gene. It is a plausible hypothesis that HD is associated with an increased biological age of affected tissues, but molecular evidence for this hypothesis has been sparse.

Here, we utilize a quantitative molecular marker of aging (known as the epigenetic clock) to analyze the tissue age of multiple postmortem brain samples from 26 HD patients and 39 controls. The epigenetic clock is a DNA methylation based biomarker of aging that can be used to measure the DNA methylation (DNAm) age of any human tissue, cell type, or fluid that contains DNA with the exception of sperm. Recent results demonstrate that DNAm age captures aspects of organismal age and the biological age of brain tissue. Here we demonstrate that HD is associated with a significant increase in the epigenetic age of specific brain regions (on average 3.3 years) and that the observed age acceleration effect is independent of underlying changes in cell composition.

Our systems biologic analysis of the brain methylation data identified 9 co-methylation modules with a significant association with HD status in several brain regions. Overall, we demonstrate that HD is associated with an increased epigenetic age of brain tissue and more broadly with substantial changes in brain methylation levels.

Top

Striatal super–enhancer signature is altered in Huntington’s disease: Enhancer regions, a therapeutic target?

Karine Merienne, PhD – University of Strasbourg

Epigenetic alterations are documented in several models of Huntington’s disease (HD). However, whether epigenetic modifications play a causal role in the mechanism leading to transcriptional dysregulation and contribute to disease pathogenesis is unclear. Using the striatum of HD R6/1 mice, we generated RNA-seq and ChIP-seq data, focusing on RNA polymerase II (RNAPII) and histone modifications associated with enhancers, such as H3K27 acetylation (H3K27ac). Integrated analysis of RNA-seq and ChIP-seq data revealed that down-regulated genes in HD mouse striatum display a super-enhancer signature. Super-enhancers are a class of broad enhancers, regulating genes that de ne cell-type-specific identity and function (e.g. neuronal identity and function genes in neurons). In addition, we show that this super-enhancer signature is altered in HD mouse striatum, since H3K27ac levels and enhancer RNA (eRNA) transcription are selectively decreased at HD striatal super-enhancers, and this correlates with reduced levels of RNAPII at super-enhancer-regulated genes. Finally, H3K27ac ChIP-seq data generated from the striatum of HD patients and control individuals indicate that down-regulated genes associate with broad enhancer regions and H3K27ac signal is decreased at those regions, suggesting that striatal super-enhancer signature is also altered in HD patients. Thus, our data support a mechanism where impaired super-enhancer activity in HD striatum contributes to altered expression of neuronal identity and function genes, defining a class of genes essential to regulation of neuronal activity such as synaptic plasticity and adaptive behavior. We suggest that targeting enhancer activity may be of therapeutic interest.

Top

Instability of the Huntington’s disease CAG repeat mutation: Genetic approaches towards dissecting mechanism and the relationship to pathogenesis

Vanessa Wheeler, PhD – Massachusetts General Hospital/Harvard Medical School

The expanded CAG repeat in the Huntington’s disease gene, HTT, is the major contributor to disease onset and severity. The HTT CAG repeat expands somatically in a time-dependent and cell-type/tissue-dependent manner, with medium-spiny neurons of the striatum exhibiting particularly dramatic expansions. We have shown in genetic studies in HttQ111 knock-in mice that genes in the mismatch repair (MMR) pathway are absolutely required for somatic CAG expansion and that the same genes enhance the pathogenic process. The relevance of these findings to human disease is highlighted by a recent genome-wide association study that identified MMR gene MLH1 as a candidate modifier of disease onset and MMR as a disease onset- associated pathway. While these findings are consistent with somatic expansion as a disease modifier, they do not rule out an alternative hypothesis that MLH1 and other MMR genes modify disease pathogenesis via mechanisms unrelated to the instability of the CAG repeat. Understanding the pathways by which MMR genes modify pathogenesis has important implications for the development of disease-modifying therapies.

To better understand the role of somatic expansion and MMR genes as disease modifiers we are performing genetic studies in HD patients and in mice. In mice, we are utilizing novel Htt CAG knock-in lines that carry either a pure CAG repeat or a CAG repeat tract that is interrupted with CAA residues, with pairs of pure
and interrupted repeat-containing mice expressing huntingtin with approximately matching glutamine tract lengths (45, 80, 105). The pure CAG knock-in mice exhibit somatic expansion that increases with mouse age and constitutive CAG repeat length and occurs in a tissue-specific manner. In contrast, the interrupted repeat configuration results in complete repeat stabilization. These models thus allow us to investigate the potential impact of somatic CAG expansion on disease expression and provide tools to test whether genetic modifiers such as MLH1 impact HD pathogenesis via mechanisms that are dependent on or independent of somatic expansion. Molecular, histological and behavioral phenotyping of the pure and interrupted mice is ongoing and results from these analyses will be presented.

Top

The role of discovery biochemical profiling in metabolic individuality and its application to Huntington’s disease

Laura P.W. Ranum, PhD – University of Florida

Huntington disease (HD) is caused by a CAG•CTG expansion in the huntingtin (HTT) gene. While most research has focused on the HTT polyGln-expansion protein, we demonstrate that four additional, novel, homopolymeric expansion proteins (polyAla, polySer, polyLeu, polyCys) accumulate in HD human brains. These sense and antisense repeat-associated non-ATG (RAN) translation proteins accumulate most abundantly in brain regions with neuronal loss, microglial activation and apoptosis, including caudate/ putamen, white matter, and in juvenile-onset cases, also the cerebellum. RAN protein accumulation and aggregation are length-dependent and individual RAN proteins are toxic to neural cells independent of RNA effects. These data suggest RAN proteins contribute to HD, and that therapeutic strategies targeting both sense and antisense genes may be required for efficacy in HD patients. This is the first demonstration that RAN proteins are expressed across an expansion located in an open-reading-frame and suggests RAN translation may also contribute to other polyglutamine diseases.

Top

Huntingtin functions in response to oxidative DNA damage via a novel signaling pathway that co–regulates huntingtin and P53 activity in ROS DNA–damaged cells

Ray Truant, PhD – McMaster University

Huntingtin is a scaffolding protein involved in ER stress response, vesicle trafficking, DNA association in protein complexes, and is transcriptionally regulated by p53. The master intracellular locator and regulator of huntingtin is N17, a multifunctional domain identified as a critical sub-target within huntingtin for drug discovery in HD.

Using High Content Analysis of a natural compounds library with multi-parametric, unbiased, non-supervised machine scoring, and chemical biology studies, we have now defined an additional function of N17 as a ROS sensor. Via a single amino acid oxidation in N17, huntingtin is quickly translocated to the nucleus to sites of DNA oxidative damage. At these sites, a complex of DNA repair factors that include ATM and p53 are scaffolded to restore and monitor DNA integrity. One by-product of this DNA repair signals feedback regulation of huntingtin and p53, in a novel kinase signaling mechanism that uses an atypical triphosphate. N17 hypo-phosphorylation in mutant huntingtin is completely restored by adding back this triphosphate precursor, absent in HD due to lack of DNA repair. Restoration of this phosphorylation restores huntingtin functions, and triggers a positive cascade of events restoring DNA repair. Upon completion of DNA repair, both huntingtin and p53 protein levels are reduced. This work defines a new role of huntingtin in a DNA damage response that is affected by age-onset ROS, affects somatic CAG expansion, huntingtin protein levels, and suggests that HD is a metabolic disorder that may be corrected by dosing with the missing metabolite. These pathways outlined are consistent with HD GWAS data and reinforce or de ne targets for HD along one pathway of ROS damage that is consistent with the human disease, yet highlights caveats of classic mouse and cell models of HD. We present a new allelic series of human HD cell lines that have intact p53 function, and new chromobodies that allow visualization of endogenous human huntingtin in live cells during DNA repair. This natural product of oxidized DNA is a potent, non-toxic and brain penetrant therapeutic lead for HD that restores the functions of mutant huntingtin to normal.

Top

Beacons along the way – Biomarkers to monitor HTT lowering in the brain

Ignacio Munoz-Sanjuan, PhD – CHDI

As a monogenic neurodegenerative indication, HD can theoretically be effectively treated through the suppression of mutant huntingtin expression. Toward this goal, CHDI has been actively involved in the development of molecular therapies designed to repress the HTT locus. One key limitation of all these approaches at present is the restricted brain distribution of the candidate therapeutic agents. This might limit the therapeutic efficacy of existing clinical development programs, and highlights the need for early markers that are temporally and regionally responsive to mHTT lowering. Over the last few years we have developed a set of experimental paradigms to identify proximal pharmacodynamic biomarkers to detect either mHTT itself or molecules/physiological signals that are potentially dysregulated in HD and that respond to HTT- lowering therapeutics after administration to phenotypic animals. A subset of these potential biomarkers have progressed to the level of being evaluated in human subjects. The presentation will focus on describing our strategy for the identification and development of target engagement and mechanism of action biomarkers, its current status, and future plans.

Top

Refining experimental gene therapies for Huntington’s and other diseases

Pavlina Konstantinova, PhD – uniQure

Gene therapy is one of the most advanced approaches for the treatment of neurodegenerative diseases like Huntington’s disease (HD) and amyloid lateral sclerosis (ALS). For HD therapy, the mutated huntingtin (HTT) is silenced with therapeutic miRNAs (miHTT) delivered with adeno-associated viral vector (AAV). Two major approaches have been undertaken for the development of RNAi-based gene therapy of HD: total HTT silencing by targeting exon 1 and allele-specific inhibition by targeting heterozygous SNPs linked to the mutant HTT. SNP rs362331 in exon 50 and SNP rs362307 in exon 67 were selected as they have the highest prevalence of heterozygosity in HD. The most efficient miHTT candidates were incorporated in AAV5 vectors and produced using the established uniQure baculovirus-based manufacturing platform. Proof of concept studies have shown efficacy of AAV5-miHTT in a HD rat model and the humanized Hu128/21 mouse model. In both models AAV5-miHTT delivery resulted in a lower concentration of the disease-inducing HTT protein associated with a delay of neurodegeneration and in reduction of mutant HTT aggregates. For ALS gene therapy artificial miALS targeting c9orf72 were generated and evaluated for silencing efficacy. On-going work aims to evaluate the therapeutic efficacy of the miALS candidates in rodents.

Direct intrastriatal delivery of AAV5-GFP by convection-enhanced diffusion (CED) injection or cerebrospinal fluid (CSF) delivery were evaluated in non-human primates (NHP) and minipigs to identify the best biodistribution profile for HD and ALS therapy. CED injection resulted in almost complete transduction of the NHP striatum and different areas of the cortex. Similarly, intrastriatal transduction of neuronal and glial cells of AAV5-GFP was observed in minipig putamen and caudate nucleus. Intrathecal delivery of the vector in the minipig resulted in predominant transduction of frontal cortex, cerebellum and lumbal spine. Further studies in HD minipig model will aim to determine the level of HTT silencing in large-brained animals, the safety of the AAV5-miHTT approach and the long-term viral persistence. The miHTT processing and o -target potential will be determined to support the clinical development of the therapeutic candidate. AAV5-miHTT provides a huge therapeutic benefit for the HD patients as it will allow for life-long HTT suppression upon single vector administration.

Top

Caveats and challenges in gene silencing for Huntington’s disease: Lessons from mice and monkeys

Presentation not yet available

Jodi McBride, PhD – Oregon Health and Science University

Silencing strategies that target the disease-causing gene in Huntington’s disease (HD) have made remarkable headway over the past decade. A therapeutic strategy that once read more like biological fiction than reality, reduction of mutant HTT has emerged as one of the most promising therapies for this disease, with the first safety trial in human patients already underway. We, along with several other research groups from around the world, have been on a steady trajectory towards moving HTT-specific constructs from the bench to the bedside. There are myriad considerations in moving HTT-lowering strategies forward that will be discussed here, including 1) the type of silencing construct that will be employed, of which there are now several options, 2) the choice of target brain region(s) and delivery route, 3) the small and large animal models in which to evaluate efficacy of lead candidate sequences and 4) the safety of the silencing approach, which is of increased importance when partially silencing both the mutant and normal HTT allele. Here, I will discuss the lessons learned by our laboratory, in collaboration with The University of Iowa, The Children’s Hospital of Philadelphia and Spark Therapeutics, as we continue to move our AAV-mediated microRNA platform forward using studies conducted in HD mouse models as well as ongoing studies in the rhesus macaque.

Top

Development and applications of genome editing technologies

Presentation not available

Feng Zhang, PhD – The Broad Institute of MIT and Harvard

Advances in genome sequencing technology have accelerated the rate at which we can identify genetic variants associated with phenotypes related to human health and disease, but functionally interrogating these variants remains time intensive. Being able to quickly find the causative variants in a sea of natural variation is essential to the goal of personalized medicine. To this end, new genome editing tools adapted from the microbial CRISPR-Cas system can be employed to rapidly screen through variants for functional effects as well as to model diseases based on patient-specific mutations. I discuss here how the CRISPR-Cas system can be deployed as a powerful discovery platform, highlighting recent findings from CRISPR screens, and describe therapeutic applications for this powerful tool. Finally, I present recent work exploring the next generation of genome editing technologies beyond Cas9, and how these new tools will further expand our ability to connect genotype to phenotype and, ultimately, treat human disease.

Top

Autophagy and other pathways that protect against neurodegeneration

David Rubinsztein, PhD – University of Cambridge

Intracellular protein aggregation is a feature of many late-onset neurodegenerative diseases, including Parkinson’s disease, tauopathies, and polyglutamine expansion diseases (like Huntington’s disease (HD)). Many of these mutant proteins, like that causing HD, cause disease via toxic gain-of-function mechanisms. Therefore, the factors regulating their clearance are crucial for understanding disease pathogenesis and for developing rational therapeutic strategies.

The two major intracellular protein degradation pathways are the ubiquitin-proteasome system and (macro) autophagy. Autophagy is initiated by double-membraned structures, which engulf portions of cytoplasm. The resulting autophagosomes ultimately fuse with lysosomes, where their contents are degraded.

I will briefly describe the basic biology of autophagy before outlining its roles in neurodegeneration. We showed that the autophagy inducer, rapamycin, reduced the levels of mutant huntingtin and attenuated its toxicity in cells, and in Drosophila and mouse HD models. We have extended the range of intracellular proteinopathy substrates that are cleared by autophagy to other related neurodegenerative disease targets, like Parkinson’s disease. While autophagy induction is protective in models of various neurodegenerative diseases, certain other conditions are associated with compromised autophagy. I will discuss how two genetic variants in Parkinson’s disease and Alzheimer’s disease impact on autophagosome biogenesis.

Finally, I will describe two new pathways that protect against neurodegenerative disease models. One of these acts via autophagy and the other is autophagy-independent.

Top

Selective clearance of mHtt through exploitation of the ubiquitin–proteasome pathway

Eric Reits, PhD – University of Amsterdam

The ubiquitin-proteasome system (UPS) is one of the two main degradation machineries in the cell, but its role in relation to HD is controversial. Earlier reports indicated that proteasomes were irreversibly sequestered into aggregates of polyQ-expanded mutant Htt fragments (mHtt), and that proteasomes are not able to cleave within polyQ repeats. Furthermore, it was suggested that proteasomes even become clogged by the polyQ expanded mHtt fragments.

Recently we showed that ubiquitinated mHtt is efficiently and entirely degraded by the proteasome. We also demonstrated that proteasomes remain accessible for substrates while recruited into mHtt accumulations, and are dynamically recruited into aggregates but not irreversibly sequestered. Together this indicates that proteasomes remain active and accessible for other substrates during mHtt aggregation. So why is mHtt normally not efficiently degraded?

My hypothesis is that poor recognition and/or ubiquitination of mHtt is the underlying basis for its poor breakdown. The polyQ expansion indeed affects ubiquitination of mHtt, as will be shown. This leads to the next question: which (de)ubiquitinating enzymes factors are involved in its recognition? And is there a link with the observed changes in proteasome complex formation in neuronal cells expressing mHtt?

Top

SRX246 Trial

Neal G. Simon, PhD – Azevan Pharmaceuticals

Top

PDE10A ICM Trial

Marielle Delnomdedieu, PhD – Pfizer

Top

SIGNAL Trial

John Leonard, PhD – Vaccinex

Top

HTT Rx CS1 Trial

Sarah Tabrizi, PhD, FRCP – University College London (for Ionis Pharmaceuticals)

Top

Reproducibility: Crisis or opportunity?

Marcus Munafò, PhD – University of Bristol

There have been several high-profile cases of academic fraud recently, but a more insidious threat to the integrity of science is the extent to which distortions of scientific best practice increase the likelihood that published research findings are false. There is growing evidence for a range of systemic problems within science, such as flexibility in the analysis of data, selective reporting of study results, publication bias against null results, influence of vested (e.g., financial) interests, and distorted incentive structures. These serve to reduce the likelihood that published research findings are true.

One critical problem is low statistical power. A study with low statistical power has a reduced chance of detecting a true effect, but it is less well appreciated that low power also reduces the likelihood that a statistically significant result reflects a true effect. There is growing evidence that the average statistical power of studies across a range of biomedical disciplines is very low. The consequences of this include overestimates of effect size and low reproducibility of results. There are also ethical dimensions to this problem, as unreliable research is inefficient and wasteful. Under plausible and quite conservative assumptions, it is likely that the majority of published findings are in fact false.

Improving reproducibility in neuroscience is a key priority and requires attention to well-established but often ignored methodological principles. Several opportunities exist for improving the situation, including adjusting the incentive structures within which scientists operate, improving statistical and methodological training of scientists, encouraging collaboration and independent review of study designs, and promoting open science through the pre-registration of study protocols, open data and sharing of materials. Ultimately, this will serve to maximize the scientific value of research.

Top

Phenotypic signature of HD in isogenic human embryonic stem cells

Ali Brivanlou, PhD – The Rockefeller University

We have developed a micropattern technology that allows high-resolution, quantitative and dynamic analysis of human embryonic stem cells (hESCs). When grown on micropatterns hESCs display a surprising self- organizing ability and, in response to BMP4, induce a specific signature of radially symmetrical germ layers. Using CRISPR/Cas9 genome-editing technology, we have generated a collection of isogenic Huntington’s disease (HD)-hESC lines that are genetically identical to their normal counterpart except for the length of the polyQ expansion. Unexpectedly, we find that single mutations of the HTT locus change the micropattern signature in response to the same ligand. We have used this technology to screen for therapeutic compounds that revert the aberrant HD signature back to normal. The use of this phenotypic screen has led to the identification of two positive hits thus validating our approach.

Top

A role of the protein kinase JNK3 on mutant huntingtin–induced neuropathology

Gerardo Morfini, PhD – University of Illinois at Chicago

Cumulative evidence from numerous animal models indicate that the devastating motor and cognitive symptoms characteristic of Huntington’s disease (HD) result from progressive deficits in neuronal connectivity that long preceded neuronal cell death. Accordingly, recent work from our group and others documented axonal pathology as an important pathogenic event induced by mutant huntingtin (mhtt) in vivo. Despite this knowledge, specific mechanisms and molecular components mediating axonal degeneration in HD remain unknown.

Maintenance of appropriate axonal connectivity depends upon intracellular trafficking events collectively referred to as axonal transport (AT). The unique dependence of axons on AT prompted us to evaluate whether mhtt affects this critical cellular process. Interestingly, results from these studies indicated that mhtt inhibits AT though a mechanism involving activation of the protein kinase JNK3 and phosphorylation of the major motor protein conventional kinesin. However, whether JNK3 contributes to mhtt-induced pathology in vivo remained unknown.

R6/2 mice, a widely used HD animal model, feature well-established behavioral, locomotor, and neuropathological phenotypes. Remarkably, results from studies presented at this meeting revealed a dramatic amelioration of these phenotypes upon JNK3 deletion. Moreover, a significant increase in survival was observed in R6/2-JNK3-/- mice, compared to R6/2-JNK3+/+ mice. Providing a mechanism for these beneficial effects, quantitative fluorescence microscopic analysis studies documented a marked reduction in axonal pathology in YFP-R6/2-JNK3-/- reporter mice.

Collectively, these exciting results demonstrate a major role of JNK3 on mhtt-induced neuropathology, providing a basis for the development of therapeutic strategies aimed to preserve neuronal connectivity in this disease.

Top

Reprogramming neurons and circuits in vivo

Paola Arlotta, PhD – Harvard University

Once programmed to acquire a specific identity and function, cells rarely change in vivo. Neurons of
the mammalian central nervous system (CNS) in particular are a classic example of a stable, terminally differentiated cell type. With the exception of the adult neurogenic niches, where a limited set of neuronal subtypes continue to be generated throughout life, CNS neurons are only born during embryonic and early postnatal development. Once generated, neurons become permanently postmitotic and do not change their identity for the life span of the organism. We have investigated whether neurons can be instructed to directly reprogram their identity postmitotically from one subtype into another, in vivo. We find that within a defined temporal window of “nuclear” plasticity neurons can change their class-defining traits to those of another class. Notably, afferent circuit onto the reprogrammed neurons also changes in response to direct neuronal reprogramming, adapting to the final “induced” neuronal identity. The data suggest that mammalian neurons and circuits may retain enough plasticity to change in response to appropriate stimuli and paves the way for mechanistic studies aimed at enhancing the process of neuronal conversion by extending the period of neuronal reprogramming to the mature brain. The work informs the development of new strategies to replace defined types of neurons lost to disease via the lineage reprogramming of a small number of neighboring, disease-resistant neurons of a different class.

Top

Analyses of mood–related neural circuits in relation to Huntington’s disease

Ann Graybiel, PhD – Massachusetts Institute of Technology

Affective and cognitive symptoms are early and often debilitating signs of disease progression in HD that can predominate over motor symptoms for extended periods. Our laboratory is focused on delineating, at multiple levels of analysis, mood-related circuits that likely are affected in HD patients. We are using methods in rodents and non-human primates to define circuits leading from medial prefrontal and orbitofrontal cortex to the striatum, especially to the striosomal compartment of the striatum, which is thought to be the main source of striatal projections to dopamine-containing substantia nigra neurons. Clinicopathologic evidence has suggested that striosomes may be preferentially vulnerable, relative to the surrounding matrix, in some mood-disordered HD patients.

In functional studies, our behavioral and electrophysiological work in rats and monkeys suggests that cortico-striosomal circuits could be crucial components of mood stabilizing networks engaged in a basic process underlying mood and affect: weighing the relative costs and benefits of particular actions under conditions of conflict. We are collaborating to optimize these tests for human subjects. In mice, our on-going work suggests that striosomes may have a specialized input to subsets of dopamine-containing neurons, and in monkeys, our work supports a striosome output path to the lateral habenula, a controller of dopamine and serotonin systems. These results point to striosome-based circuits as potentially critical components of networks responsible for at least some of the symptoms of anxiety, apathy, perseverativeness and related mood difficulties suffered by patients with HD and related disorders.

In Q175 work, we are generating Q175 mice with genetic tags for striosome or matrix SPNs, with the aim of developing models for therapeutic testing and assessment of compartmental function and dysfunction. To date in Q175 mice, our evidence suggests that striosomes can be progressively and regionally affected, supporting evidence in other HD rodent models. Compartmentally selective CalDAG-GEFI and CalDAG-GEFII are down-regulated in Q175 mice, as in R6/2 mice, in which we earlier found that knock-down of matrix-enriched CalDAG-GEFI promotes neuronal survival in a brain slice model of HD. Thus we hope to provide and to use cell-molecular strategies to pursue analysis of mood-related circuitry in relation to HD.

Top

Glia as therapeutic reagents in HD

Steve Goldman, MD, PhD – University of Rochester

Glial pathology contributes to a range of neurodegenerative and neuropsychiatric diseases traditionally considered disorders of neuronal dysfunction. To explore the contribution of glial pathology to Huntington’s disease (HD), we established human HD glial chimeras by neonatally engrafting the striata of immunodeficient mice with mutant huntingtin (mHTT)-expressing glial progenitor cells (GPCs), derived from either huntingtin mutant human embryonic stem cells (hESC), or from mHTT-transduced human forebrain hGPCs. Mice engrafted with mHTT hESC GPCs exhibited worse motor performance than controls chimerized with normal hESC GPCs, while striatal medium spiny neurons (MSNs) in mHTT glial chimeras were hyperexcitable and manifested high input resistance. Genomic analysis revealed that mHTT hESC GPCs exhibited the transcriptional suppression of a number of potassium channels, associated with the down-regulation of both astroglial and oligodendrocytic differentiation-linked gene expression, relative to stage-matched normal hESC GPC controls. These data suggested that glial pathology, including both astrocytic and oligodendrocytic dysfunction, might contribute significantly to HD phenotype. On that basis, we asked if the replacement of diseased glia by normal glia might slow disease progression in HD. To this end, we transplanted immunodeficient rag2-/- x R6/2 HD (120 CAG) mice with normal, human fetal tissue-derived and CD44-isolated astrocyte-biased glia, and assessed the effects of glial engraftment on their disease progression. We found that striatal engraftment with normal glia slowed motor deterioration, rescued aspects of behavioral phenotype, and extended survival in these R6/2 mice. Electrophysiological analysis of striatal slices derived from the transplanted mice revealed that their MSNs exhibited lower input resistance and were less excitable than untreated R6/2s. Whereas R6/2 striata exhibited high interstitial K+, chimerization with wild-type glia normalized K+ levels. Together, these findings suggest a causal role for glia in HD, and indicate that the colonization of HD striata with wild-type glia can rescue aspects of neuronal function while substantially slowing disease progression, thus suggesting a new cell therapeutic strategy for disease amelioration in HD.

Top

Huntington’s disease measurements and models: Quality and relevance

Douglas R. Langbehn, MD, PhD – University of Iowa

In HD, an extraordinarily diverse range of biomedical and psychological measurements have been used or proposed. One can contemplate employment of these measures in a staggering range of data models. However, the fact that one can measure some aspect of HD does not mean that one can necessarily measure it well. Further, even if measured well, it may not be very useful. Somewhat similarly, although some modeling techniques ideally may be able to unlock new insights into the complicated relationships among measurements, one often does not have the quantity or quality of data needed to exploit these potential advantages. And if appropriate data is available, a beautiful model may still only provide the right answers to the wrong questions. As with the measurements themselves, a model may or may not be useful. This may all seem rather self-evident, but experience has often reminded me how easily these considerations are under- appreciated in practice.

In the above, the meanings of “well” and “useful” depend heavily on context. For example, measurements capable of testing the presence or absence of a disease phenomenon may not have the sensitivity or stability needed to accurately quantify it. Regarding utility, precise, reliable measurements of a well-established aspect of disease are invaluable for clinical trials, though they provide no new insight into HD pathogenesis. Conversely, techniques that may shed new light on the pathophysiology of HD are often inappropriate for clinical trials.

In today’s talk, I will review and distinguish the various goals that commonly drive HD studies, and I will discuss measurement and modeling criteria needed to meet these different goals. I will emphasize the (sometimes mutually exclusive) requirements for scientific discovery versus clinical trials or clinical prediction. I will also emphasize the increasing stringency requirements needed to establish the presence of an aspect of HD, quantify it, and quantify its change over a relatively short time. Throughout, I will illustrate with examples—both successes and failures—largely taken from the TRACK-HD and TRACK-ON observational HD studies.

Top

Clinical rating scales in Huntington’s disease: Validation, validation, validation

Tiago A. Mestre, MD, MSc – University of Ottawa

Clinical rating scales are fundamental tools in clinical research and clinical practice. The ability to measure clinical phenomena in Huntington’s disease (HD) is fundamental for the development of any treatment in HD and an accurate characterization of its natural history. Clinical rating scales need to be well-designed and appropriate for a given purpose of measurement. There are standards to assist in the development of quality clinical rating scales and determine the degree of validation and appropriateness of existing clinical rating scales. Currently, it is not known how well validated and appropriate are the clinical rating scales used in HD populations for its various purposes of measurement.

In this session, the process of development and validation of a clinical rating scale will be introduced as well as the concepts of validity, reliability and responsiveness that are core clinimetric properties assessed in a clinical rating scale. A joint effort of CHDI and the International Parkinson and Movement Disorder Society to critically assess all clinical rating scales used in HD will be presented. This project aims at characterizing how well validated and appropriate are clinical rating scales used in HD studies focusing on the core domains of motor symptoms, behavior, cognition, as well as functional ability and quality of life. The conclusions of this project will help set the priorities for clinical rating scale development and/or validation in HD, addressing the more emerging needs of a particular symptom area or purpose.

Top

Validation of a prognostic index for Huntington’s disease

Jeffrey D. Long, PhD – University of Iowa

A prognostic index (PI) for Huntington’s disease (HD) is developed from variables measured at study entry to predict the occurrence of motor diagnosis, de ned as the highest rating on the UHDRS Diagnostic Confidence Level. The recent availability of large HD observational databases (PREDICT, TRACK, COHORT, REGISTRY) provides an unprecedented opportunity to develop a PI and assess its generalizability in an external validation analysis. Using survival modeling for time-to-event data, a PI model was developed with the PREDICT dataset. To maximize reproducibility, different models were evaluated using leave-one-site-out cross-validation with the 32 sites of PREDICT. Results of the model development showed that a PI based on the weighted sum of the UHDRS Total Motor Score (TMS), Symbol Digit Modalities Test (SDMT), and the CAG-Age Product (CAP), had the best predictive performance among the PREDICT sites. External validation analysis consisted of using the PREDICT weights to compute the PI in three other datasets: TRACK, COHORT, and REGISTRY. External validation performance was assessed by site using Harrell’s C, which indexes the extent to which the PI score predicts the timing of motor diagnosis after study entry. Using effect size standards from other fields like oncology, the results showed excellent validation performance for TRACK and COHORT, but relatively poor performance for REGISTRY. The poor performance for REGISTRY is explained by the participants being much more progressed than in the other studies. The PREDICT, TRACK, and COHORT datasets were combined (N = 1421) to determine the final PI, computed as PI = (52 × TMS) + (-34 × SDMT) + (7 × Age × (CAG – 34)). The results suggest the PI for HD is reproducible, having similar performance among sites in diverse geographical settings and among different research studies. The poor performance for the most progressed cohort (REGISTRY) suggests the PI predictive ability for future motor diagnosis may be limited to individuals currently in the prodromal HD stage, though the PI will still reflect advanced progression. The PI is useful for characterizing progression, forming risk groups, and selecting participants for clinical trials.

Top

Understanding Huntington’s disease progression: A multi–level probabilistic modeling approach

Jianying Hu, PhD – IBM T.J. Watson Research Center

Huntington’s disease (HD) is an autosomal dominant neuro-degenerative disease. The progression of HD is manifested in complex symptoms and signs both before and after clinical diagnosis, and is not well understood. Data from observational studies such as Enroll-HD, Registry and Track-HD provide an unprecedented opportunity to understand the natural history of HD. However, significant challenges remain, including how to combine heterogeneous sources of data while accounting for noises, biases, and inherent variability, and how to uncover salient patterns from heterogeneous longitudinal observations. Multi-level Probabilistic Disease Progression Modeling is a new machine learning approach for inferring disease states from longitudinal observational data. It has been applied to the modeling of chronic conditions such as COPD. In this talk I will describe our proposal to apply and expand this approach to the modeling of HD progression, with the goal of providing a comprehensive view of HD disease states, transitions, areas of manifestation and progression pathways. I will also share some early results from our efforts to combine data from Enroll-HD, Registry and Track-HD, as the first step of this comprehensive modeling exercise.