VIEWPOINT
Preclinical Models: Needed in Translation? A Pro/Con Debate Thomas Philips PhD,1 Jeffrey D. Rothstein MD, PhD,1 and Mahmoud A. Pouladi PhD2* 1 Johns Hopkins University, Brain Science Institute, Baltimore, Maryland, USA Translational Laboratory in Genetic Medicine (TLGM), Agency for Science Technology and Research (A*STAR), and Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore (NUS), Singapore
2
ABSTRACT:
The discovery of the causative mutations and many of the predisposing risk factors for neurodegenerative disorders such as Amyotrophic Lateral Sclerosis, Alzheimer’s, Parkinson’s, and Huntington’s disease (HD), has led to the development of a large number of genetic animal models of disease. In the case of HD, for example, over 20 different transgenic rodent models have been generated. These models have been of immense value in providing novel insights into mechanisms of disease, with the promise of accelerating the development of therapies that can
For many neurodegenerative diseases, including Amyotrophic Lateral Sclerosis (ALS), Alzheimer’s, Parkinson’s, and Huntington’s disease (HD), the causative mutations and many of the predisposing genetic risk factors have been known for many years. In the case of HD, for example, the mutation in HTT was identified more than 20 years ago, enabling the development of genetic animal models and leading to a wealth of knowledge and novel insights into mechanisms of disease. Yet, despite the availability and intensive use of such preclinical models, a treatment that can delay the onset of disease or slow its rate of progression is yet to be developed for HD or any of the other neurodegenerative diseases. This leads one to raise the question: Are preclinical models needed in translational research?
------------------------------------------------------------
*Correspondence to: Mahmoud A. Pouladi, Translational Laboratory in Genetic Medicine (TLGM), Agency for Science Technology and Research and National University of Singapore, 8A Biomedical Grove, Immunos L5, Singapore 138648. E-mail: [email protected] Funding agencies: M.A.P. is supported by grants from the Agency for Science Technology and Research (A*STAR), and the National University of Singapore (NUS). Relevant conflicts of interest/financial disclosures: Nothing to report. Full financial disclosures and author roles may be found in the online version of this article. Received: 5 August 2014; Accepted: 11 August 2014 Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/mds.26010
delay the onset or slow the progression of the disease. Yet, despite extensive use of such models, no effective treatment for HD has been developed. Here, we discuss the value of animal models, highlighting their strengths and shortcomings in the context of translaC 2014 International Parkinson tional research for HD. V and Movement Disorder Society
K e y W o r d s : Animal models; Huntington’s; HTT; transgenic; translation
CON: The Case Against Preclinical Models in HD Many different animal models for studying HD have been generated, ranging from nonhuman primates, mice, Caenorhabditis elegans, and Drosophila melanogaster and zebrafish. Many of these animal models develop a phenotype reminiscent of some aspects of disease as seen in humans, including the loss of gaminobutyric acid–ergic medium spiny neurons (MSN neurons) in the striatum (as reviewed in Pouladi et al.1). In fact, interference with specific disease pathways in these models has led to improvement of the disease phenotype and an improvement of our understanding of some basic biological aspects of neurodegeneration in HD. None of these models, especially the mice models of HD, have so far led to the development of an effective treatment that can halt or slow the disease from progressing in HD patients. We describe several characteristics of preclinical models of HD that often limit their translational value for studying HD in humans.
Caveats of Small Animal Models to Study HD Small animal models such as zebrafish, Drosophila, and C. elegans do have the advantage of being relatively cheap to maintain and being a great tool for
Movement Disorders, Vol. 29, No. 11, 2014
1391
P H I L I P S
E T
A L
high-throughput genetic or drug screenings. However, small animal models have the disadvantage of not being very biologically related to humans in terms of their phylogenetic differences and complexity of their nervous system, particularly of relevance for HD, in which patients’ cognitive and psychiatric functions are highly affected. In addition, some of the small animal models lack a corresponding disease-causing genetic homolog, for example, the C. elegans models for HD, in which human mutant Huntingtin protein (mHTT) is inserted in the worm genome to model disease.2 In both these small but also the larger animal models discussed later, many caveats are associated with insertion of exogenous genes in a host genome (reviewed in O’Sullivan et al.3). Other caveats are associated with the design of small animal models. For example, HD is caused by a dominant mutation in a single gene encoding huntingtin (HTT). Because the HTT gene is a particularly large gene (approximately 170 kbp), developing HD animal models has been very difficult. Therefore, many HD models only express a (small) part of the mHTT gene, with excluded but potentially pathological regions of the mHTT gene not represented in the phenotype being observed.2,4 C. elegans and Drosophila models of HD usually express either truncated or full-length mHTT and often express mHTT from complementary DNA constructs, potentially failing to recapitulate pathological events related to other regions of the mHTT gene such as mHTT introns.2,5,6 Many caveats are associated with the phenotype that small animal models of neurodegeneration develop. These often do not recapitulate the unique spectrum of the disease phenotype as seen in patients with neurodegenerative disease. In HD, specific hallmarks of the disease such as dysphagia and dysarthria are clearly very difficult to model in such small animal models. Drosophila models clearly recapitulate motor dysfunction as seen in reduced climbing behavior (negative geotaxis),6 but this is phenotypically very similar to what is seen in Drosophila models of other, but very different, neurodegenerative disease such as ALS.7 Distinctive motor symptoms such as the rigidity and chorea seen in HD patients, as well as the cognitive and psychiatric features characterizing the HD phenotype are very hard to model using small animal models. Therefore, these studies using small animal models might rather lead to enhanced insight into less subtle disease mechanisms that are relevant for a broader range of neurodegenerative diseases.
Caveats of Rodent and Large Animal Models to Study HD Similarly to the small HD models, one has to be very careful about the biological differences between rodent models and humans. These might confound proper translation to the human clinic. For example, given the
1392
Movement Disorders, Vol. 29, No. 11, 2014
differences in brain mass (3,0003 larger in humans as compared with mice) and number of brain neurons (86 billion in humans versus 90 million in mice), and the much larger complexity of brain structure in humans, one should not be surprised that accurately modeling human disease in rodent HD models is very difficult. Therefore, although particular aspects of human HD are well modeled in rodents (see later discussion), other aspects of human HD such as facial movements, oculomotor deficiencies, dysarthria, and dysphagia are very difficult to model in rodent models of HD. More than 20 different transgenic rodent models expressing the human or mouse HTT gene have been generated (reviewed in Pouladi et al.1). These rodents express either truncated or full-length fragments of mHTT with varying CAG repeat length. Although these rodent models develop symptoms of cognitive loss and motor neuron dysfunction, some major caveats are associated with the development of these rodent models. Adequate gene dosage is essential during the development of animal models of neurodegeneration. High gene dosage in transgenic animals, or long repeat lengths in models of expansion disorder diseases such as HD, often lead to a fast progressive disease phenotype because of the ubiquitous expression of a high copy number or long repeat length of a transgene. This could be motivated by the financial aspects of animal research, which is usually a significant financial burden for a research facility, as well as the publication pressure in a usually highly competitive environment. This fast progressive disease phenotype often precludes the analysis of specific pathology related to human HD. Conversely, using these models, one might try to explore whether alterations in expression of genes affecting aging might modify the phenotype. In HD, mice overexpressing truncated mHTT often develop very-early-onset, fast progressive fatal disease that is more reminiscent of a generalized degenerative phenotype rather than specifically recapitulating HD symptoms as seen in patients.4 Therefore, in HD rodent models, the generalized expression of a CAG repeat expansion in cell types only mildly involved in human HD could have significant impact on the presentation of the phenotype in these rodents, potentiating fast progression and early demise. For example, mHTT expression in mouse motor cortex or other brain regions, whose corresponding brain regions in human HD are more mildly affected, might contribute to a very fast progressive motor phenotype, which is usually not seen in HD patients. Transgenic animal models expressing more modest levels of the transgene, similar to the levels of the endogenous gene, or mouse models of repeat expansion disorders with limited repeat length, might circumvent this issue but often lead to a phenotype that is only partial or very mild1,8 and might take a long time to develop. This might be more relevant for
P R E C L I N I C A L
modeling disease as seen in patients, because HD onset usually occurs in mid-adult life and might take many years or decades to progress until significant burden in the patient is noticed. Very often in such instances the animal model might not live long enough to develop the whole plethora of the neurodegenerative phenotype as seen in patients. These rodent versions of HD often show very simplistic clinical representations of the disease and lack the many relevant nuances (clinical and pathological) seen in human HD. For example, HD patients develop significant loss of MSN neurons in the striatum. Conversely, a rat model of HD, the BACHD rat, develops many alterations in the striatal compartment but shows no loss of striatal neurons.9 In fact, many rodent models for HD only present certain aspects of disease, and effective therapeutic drug trials using HD rodents are of low translational value because their implications for many other disease-related mechanisms could not be explored in these models. When only partial aspects of disease are being targeted, that a clinical trial will fail in the earlier stages might not be surprising, because drug effects on other disease-relevant pathways have not been taken into account in the preclinical model. Nonhuman primate models might give the best correlation to human disease. The earliest generated nonhuman primate models for HD were generated by administration of neurotoxins that selectively ablated striatal neuronal subpopulations.10 These models do not recapitulate the phenotype seen in HD patients, who develop progressive widespread cortical, subcortical, and even peripheral nervous system degeneration, in contrast to the acute localized lesions seen in the neurotoxin models. Transgenic large animal models for HD have also been developed. Unfortunately, none of these models is a true HD animal model. Transgenic mHTT macaque monkeys develop a very aggressive phenotype, with early death and unusual symptoms that are not seen in early stages of human HD.11 Transgenic sheep and minipigs did not develop overt disease symptoms, however, although this could be because of the limited mHTT CAG repeat size these models expressed.12,13 Even if one is successful in developing a large animal model to study HD, still many inherent caveats will be associated with these studies. Large animal studies, especially those using nonhuman primate models most relevant to study human disease, are extremely expensive to maintain, and the neurodegenerative disease in these animals takes many years to develop. In fact, HD onset in humans is usually midadult life, and so disease progression might need to be studied during several decades. Even if they were to develop disease, given the genetic variability of primates (like humans), this would likely necessitate a large number of animals—similar to human trials—to see a positive and reproducible clinical response. This makes
M O D E L S :
N E E D E D
I N
T R A N S L A T I O N ?
their use for drug screening suboptimal, and studies using nonhuman primate models usually take many years if not decades to complete.
Alternatives to Preclinical Animal Models: Value of iPS Cells in Studying HD Rather than using animal models to study HD disease pathogenesis and treatment opportunities, one could consider the use of patient-derived induced pluripotent stem (iPS) cells. The development of iPS cells with a patient-specific disease signature could be explored both for understanding disease pathogenesis as well as for testing drugs for therapeutical efficacy. The iPS cells could be differentiated into different central nervous system cell types such as neurons, astrocytes, and oligodendrocytes. Protocols have recently been established for specifically generating MSN neurons from iPS cells.14 The concept that neuronal death is non–cell autonomous, that is, not only neurons but also specific glial cell types and other neuronal cell types contribute to the disease pathogenesis, is well established.15 For example, in HD mouse models, astrocyte GLT-1 glutamate transporters are significantly reduced, which contributes to glutamate-mediated excitotoxic cell death of the nearby g-aminobutyric acid–ergic neurons most susceptible in HD.16 The ability to model all contributing cell types in a dish, by analyzing a patient-specific disease signature with effects of co-morbidity and disease modifier genes incorporated as well, is most relevant for the specific HD patient under study. Of particular interest is that in all HD patients, disease is caused in a monogenetic fashion by a CAG repeat expansion of varying length. Unlike in other neurodegenerative diseases in which most cases are sporadic with the underlying cause still at large, specific targeting of the HD disease–causing gene using RNAse H active antisense oligonucleotides or small interfering RNAs targeting single nucleotide polymorphisms associated with the expanded allele is of interest to all HD patients. A recent study has shown that transient infusion of antisense oligonucleotides into the cerebrospinal fluid effectively targeted the mHTT gene and reversed the phenotype in HD models.17 In addition, infused nonhuman primate models showed a clear reduction in wild-type HTT protein expression in many brain regions.17 The development of iPSderived neuronal and non-neuronal cells from individual HD patients could at least in part bypass the need for these preclinical therapeutic trials in HD animal models by providing us a great tool to assess the feasibility and therapeutic dosage of antisense treatment specifically designed for the patient under study. In addition, this antisense therapy also could be directed against many potential disease modifier genes in highthroughput screenings. In fact, many different genes have been reported to aggravate or alleviate HD
Movement Disorders, Vol. 29, No. 11, 2014
1393
P H I L I P S
E T
A L
disease symptoms.18 Conversely, studying these cells in a dish, extracted from their usual microenvironment, might lead to significant loss of their translational value. Therefore, whether specific HD disease– related phenotypes can be effectively mimicked in the iPS-derived cell culture system remains to be seen.
sion.22 That animal models allow the examination of the neuroprotective potential of cellular and molecular targets implicated in HD provides a degree of reassurance before the initiation of potentially complicated and likely costly clinical trials, further emphasizing the importance of animal models in translational studies.
PRO: The Case for Preclinical Models in HD
High Construct and Face Validity of Animal Models of HD
In spite of the various caveats about the HD models, preclinical animal, and in particular rodent, models of HD have played an indisputably important role in identifying disease processes and elucidating cellular pathways disrupted in Huntington’s disease.1 Indeed, much of our current understanding of the pathogenic mechanisms implicated in HD has been shaped by studies of animal models of HD.19 In addition to their utility in discovering disrupted cellular processes, animal models have been used to evaluate the impact of normalizing the identified cellular and molecular deficits on diseaserelated phenotypes,20,21 shedding light on their potential as targets for therapeutic intervention. Despite developments that have led to the value of animal models in translational research being debated, animal models of HD will remain important in bridging the translational gap from basic discoveries to trials in patients with HD for a number of reasons. Because rare disorders such as HD have limited patient populations that are available to participate in the trial of potential therapies, approaches that help build confidence and provide evidence, even if only in proof-of-concept form, that supports a given therapeutic strategy are of immense importance. Such data provide an additional criterion by which to prioritize experimental therapies and may help in the nomination of promising candidates for clinical evaluation. In this regard, animal models provide an opportunity to test the therapeutic potential of candidate compounds before initiating clinical trials in the limited pool of patients with HD that may compromise their ability to participate in subsequent trials.
Value of Proof-of-Concept Studies of Neuroprotection Furthermore, although treatments that target symptoms are generally the first line of therapies developed for many neurodegenerative disorders, including HD, the ultimate goal of translational research is not only to provide temporary relief of symptoms but also to delay the onset and slow the progression of disease.22 For these slow progressing diseases, the development of such neuroprotective therapies is challenging not only from a clinical trial design point of view, but also economically because of the longer duration of trials needed to ascertain treatment effects on disease progres-
1394
Movement Disorders, Vol. 29, No. 11, 2014
Unlike many forms of neurodegenerative diseases, HD is extraordinary in its monogenic etiology and its dominant and highly penetrant presentation.23 These features entail not only a clearly defined patient population, but also animal models of high apparent validity. Three broad criteria are often used to judge a given model’s relevance: construct validity, face validity, and predictive validity.1 Many animal models of HD are considered to be of high construct validity given that they closely reproduce the single genetic mutation that underlies HD in humans, namely CAG trinucleotide repeat expanded HTT allele(s).1 Similarly, animal models of HD exhibit high face validity, recapitulating many phenotypic features of HD, including progressive motor, cognitive, and psychiatric-like disturbances coupled with HD-like neuropathological and molecular changes.1 In terms of predictive validity, which is gauged based on how closely response to treatment in the animal model parallel, or predict, improvements in patients, this aspect is yet to be proven, because no disease-modifying therapies have been developed. Indeed, failure of a therapy shown to be effective in humans to improve disease phenotypes in animal models would be cause for pause. With respect to the only treatment currently approved for HD,24,25 which is mostly of symptomatic benefit, evidence indicates that tetrabenazine improves disease phenotypes in an animal model of HD.26 Furthermore, the outcome of a given clinical trial is influenced by a number of factors, which include not only the role of the target being engaged in the disease process, but also issues of dosage, frequency, disease stage, and specific outcomes being assessed. This can be readily appreciated when considering that for many clinically approved therapies, multiple clinical trials of a given treatment were often necessary to arrive at the right combination of dose, disease stage, and outcome measure. Thus, for treatment approaches showing robust and reproducible benefits in animal models, caution is warranted when interpreting discordance between animal and human trials in which the disease stage or outcome measures may have been different.
Availability of Versatile Animal-Based Experimental Tools to Dissect Disease Processes and Mechanism of Action of Therapeutic Candidates Safe, specific, and brain-permeable pharmacological compounds are available for a relatively limited
P R E C L I N I C A L
number of molecular targets, and in most cases, screening for and development of novel compounds to engage molecular targets of interest, a nontrivial undertaking, is required. The ability to assess the impact of specifically enhancing or decreasing the activity of a given target before the initiation of potentially costly and lengthy drug screening and development efforts is of great value. Since the generation of the first transgenic rodent model over three decades ago,27 thousands of genetically modified animal lines have been established that make such studies possible. By cross-breeding to animal models of HD, these lines have allowed for the interrogation of the effect of increased or decreased expression, and hence activity, of a given gene of interest on HD-related phenotypes. Furthermore, the availability of transgenic animal technologies, such as inducible and tissue-specific Cre recombinase lines,28 allows the dissection of timedependent and tissue-specific cellular processes involved in HD. These versatile transgenic animalbased tools have enabled a better understanding of the pathogenesis of HD and will continue to play an important role in providing novel insights into HD.2931 Transgenic animal lines also may be of translational value in examining the specificity and the mechanism of action of therapeutic candidates of interest.
Alternative Approaches Are Not as Well Developed An exciting recent development that promises to revolutionize the study of human disease, and in particular brain disorders, is the discovery that somatic and terminally differentiated cells may be induced to a pluripotent stem cell state that may subsequently be differentiated into any cell type of interest.32 For brain disorders such as HD, a major challenge has been the inability to directly study the cell type most affected in the disease, namely striatal medium spiny neurons, in a human context. The ability to generate iPS cells from readily accessible patient tissues (eg, fibroblasts or peripheral blood monocytes), coupled with the development of genome editing tools, such as the transcription activator–like effector nucleases and the RNA-based clustered regularly interspaced short palindromic repeats–Cas9 system,33 which enable the correction of the HD mutation, provides an opportunity to study pathogenic processes and to evaluate potential therapies in isogenic stem cell–derived MSNs of human origin. This potential has indeed been demonstrated in a number of recent studies.34-37 The value of stem cell–derived human neurons in the study of HD cannot be overestimated; however, a number of factors may limit their utility, at least in the short-term. First, whereas in animals isolating and studying virtually any cell type of interest is possible, the ability to do the same using human stem cells
M O D E L S :
N E E D E D
I N
T R A N S L A T I O N ?
depends largely on the availability of the appropriate directed differentiation protocols necessary to derive the cell type of interest. Furthermore, with respect to MSNs, the most highly validated protocols available to derive them from human stem cells are relatively inefficient, with only a small fraction (10-15%) of the differentiated neurons exhibiting the desired MSN identity at the conclusion of the differentiation procedure.14,38 Second, in their simplest and most widely adopted configuration, studies of human stem cell– derived neurons neglect to recreate the complex milieu that includes other central nervous system cell types, such as astrocytes, microglia, and oligodendrocytes, which may be key to recapitulating disease phenotypes. The importance of non–cell autonomous effects in HD disease processes is indeed supported by a number of recent animal studies.29,31,39-41 By failing to model many of the environmental cues in stem cell– based neuronal models, important non–cell autonomous disease processes may be missed. Finally, agedependent progression of whole system, complex phenotypes, in particular cognitive, motor, and psychiatric features, which can be readily assessed in animal models of HD, simply cannot be examined in stem cell– derived neuronal systems. Thus, much room for improvement in stem cell–based modeling of HD remains, and the exciting advances in our ability to study HD disease processes in neurons and cell types of human origin will serve to complement studies in animal models of HD.
Conclusion Numerous factors argue for the continued use of preclinical animal models of HD in translational research, including the fact that the patient population available to participate in clinical trials is limited; the costs of failure for trials seeking to document neuroprotection are high; the available animal models show a high level of validity because of HD’s dominant, monogenic, and highly penetrant nature; and finally, the available alternative experimental systems are less developed.
References 1.
Pouladi MA, Morton AJ, Hayden MR. Choosing an animal model for the study of Huntington’s disease. Nat Rev Neurosci 2013;14: 708-721.
2.
Faber PW, Alter JR, MacDonald ME, Hart AC. Polyglutaminemediated dysfunction and apoptotic death of a Caenorhabditis elegans sensory neuron. Proc Natl Acad Sci USA 1999;96:179-84.
3.
O’Sullivan GJ, O’Tuathaigh CM, Clifford JJ, O’Meara GF, Croke DT, Waddington JL. Potential and limitations of genetic manipulation in animals. Drug Discov Today Technol 2006;3:173-180.
4.
Mangiarini L, Sathasivam K, Seller M, et al. Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice. Cell 1996;87:493506.
Movement Disorders, Vol. 29, No. 11, 2014
1395
P H I L I P S
5.
6.
E T
A L
Jackson GR, Salecker I, Dong X, et al. Polyglutamine-expanded human huntingtin transgenes induce degeneration of Drosophila photoreceptor neurons. Neuron 1998;21:633-642. Romero E, Cha G-H, Verstreken P, et al. Suppression of neurodegeneration and increased neurotransmission caused by expanded full-length huntingtin accumulating in the cytoplasm. Neuron 2008;57:27-40.
7.
Li Y, Ray P, Rao EJ, et al. A Drosophila model for TDP-43 proteinopathy. Proc Natl Acad Sci U S A 2010;107:3169-3174.
8.
Tieu K. A guide to neurotoxic animal models of Parkinson’s disease. Cold Spring Harb Perspect Med 2011;1:a009316.
9.
Yu-Taeger L, Petrasch-Parwez E, Osmand AP, et al. A novel BACHD transgenic rat exhibits characteristic neuropathological features of Huntington disease. J Neurosci 2012;32:15426-15438.
10.
Beal MF, Kowall NW, Ellison DW, Mazurek MF, Swartz KJ, Martin JB. Replication of the neurochemical characteristics of Huntington’s disease by quinolinic acid. Nature 1986;321:168-171.
11.
Yang SH, Cheng PH, Banta H, Piotrowska-Nitsche K, Yang JJ, Cheng EC, Snyder B, Larkin K, Liu J, Orkin J, Fang ZH, Smith Y, Bachevalier J, Zola SM, Li SH, Li XJ, Chan AW. Towards a transgenic model of Huntington’s disease in a non-human primate. Nature 2008;453:921-924.
12.
Jacobsen JC, Bawden CS, Rudiger SR, et al. An ovine transgenic Huntington’s disease model. Hum Mol Genet 2010;19:1873-1882.
13.
Yang D, Wang C-E, Zhao B, et al. Expression of Huntington’s disease protein results in apoptotic neurons in the brains of cloned transgenic pigs. Hum Mol Genet 2010;19:3983-3994.
14.
Delli Carri A, Onorati M, Castiglioni V, et al. Human pluripotent stem cell differentiation into authentic striatal projection neurons. Stem Cell Rev 2013;9:461-474.
15.
Ilieva H, Polymenidou M, Cleveland DW. Non-cell autonomous toxicity in neurodegenerative disorders: ALS and beyond. J Cell Biol 2009;187:761-772.
16.
Behrens PF, Franz P, Woodman B, Lindenberg KS, Landwehrmeyer GB. Impaired glutamate transport and glutamateglutamine cycling: downstream effects of the Huntington mutation. Brain 2002;125:1908-1922.
23.
Myers RH. Huntington’s disease genetics. NeuroRx 2004;1:255-262.
24.
Huntington Study Group. Tetrabenazine as antichorea therapy in Huntington disease: a randomized controlled trial. Neurology 2006;66:366-372.
25.
Hayden MR, Leavitt BR, Yasothan U, Kirkpatrick P. Tetrabenazine. Nature Rev Drug Discovery 2009; 17-18.
26.
Wang H, Chen X, Li Y, Tang T-S, Bezprozvanny I. Tetrabenazine is neuroprotective in Huntington’s disease mice. Mol Neurodegener 2010;5:18.
27.
Hanahan D, Wagner EF, Palmiter RD. The origins of oncomice: a history of the first transgenic mice genetically engineered to develop cancer. Genes Dev 2007;21:2258-2270.
28.
Nagy A. Cre recombinase: the universal reagent for genome tailoring. Genesis 2000;26:99-109.
29.
Wang N, Gray M, Lu X-H, et al. Neuronal targets for reducing mutant huntingtin expression to ameliorate disease in a mouse model of Huntington’s disease. Nat Med 2014;20:536-541.
30.
Hult S, Soylu R, Bj€ orklund T, et al. Mutant huntingtin causes metabolic imbalance by disruption of hypothalamic neurocircuits. Cell Metab 2011;13:428-439.
31.
Crotti A, Benner C, Kerman BE, et al. Mutant Huntingtin promotes autonomous microglia activation via myeloid lineagedetermining factors. Nat Neurosci 2014;17:513-521.
32.
Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 2006;126:663-676.
33.
Kim H, Kim J-S. A guide to genome engineering with programmable nucleases. Nat Rev Genet 2014;15:321-334.
34.
HD iPSC Consortium. Induced pluripotent stem cells from patients with Huntington’s disease show CAG-repeat-expansion-associated phenotypes. Cell Stem Cell 2012;11:264-278.
35.
An MC, Zhang N, Scott G, Montoro D, Wittkop T, Mooney S, et al. Genetic correction of Huntington’s disease phenotypes in induced pluripotent stem cells. Cell Stem Cell 2012;11:253-263.
36.
Kaye JA, Finkbeiner S. Modeling Huntington’s disease with induced pluripotent stem cells. Mol Cell Neurosci 2013;56:50-64.
17.
Kordasiewicz HB, Stanek LM, Wancewicz EV, et al. Sustained therapeutic reversal of Huntington’s disease by transient repression of huntingtin synthesis. Neuron 2012;74:1031-1044.
37.
Ross CA, Akimov SS. Human-induced pluripotent stem cells: potential for neurodegenerative diseases. Hum Mol Genet 2014; May 13.
18.
Clabough EBD. Huntington’s disease: the past, present, and future search for disease modifiers. Yale J Biol Med 2013;86:217.
38.
19.
Ross CA, Tabrizi SJ. Huntington’s disease: from molecular pathogenesis to clinical treatment. Lancet Neurol 2011;10:83-98.
Carri AD, Onorati M, Lelos MJ, et al. Developmentally coordinated extrinsic signals drive human pluripotent stem cell differentiation toward authentic DARPP-321 medium-sized spiny neurons. Development 2013;140:301-312.
20.
Gil-Mohapel JM. Screening of therapeutic strategies for Huntington’s disease in YAC128 transgenic mice. CNS Neurosci Ther 2012;18:77-86.
39.
Gu X, Li C, Wei W, et al. Pathological cell-cell interactions elicited by a neuropathogenic form of mutant Huntingtin contribute to cortical pathogenesis in HD mice. Neuron 2005;46:433-444.
21.
Gil JM, Rego AC. The R6 lines of transgenic mice: a model for screening new therapies for Huntington’s disease. Brain Res Rev 2009;59:410-431.
40.
Gu X, Andre VM, Cepeda C, et al. Pathological cell-cell interactions are necessary for striatal pathogenesis in a conditional mouse model of Huntington’s disease. Mol Neurodegen 2007;2:8.
22.
Katsuno M, Tanaka F, Sobue G. Perspectives on molecular targeted therapies and clinical trials for neurodegenerative diseases. J Neurol Neurosurg Psychiatry 2012;83:329-335.
41.
Tong X, Ao Y, Faas GC, et al. Astrocyte Kir4.1 ion channel deficits contribute to neuronal dysfunction in Huntington’s disease model mice. Nat Neurosci 2014;17:694-703.
1396
Movement Disorders, Vol. 29, No. 11, 2014
Preclinical Models: Needed in Translation? A Pro/Con Debate Thomas Philips PhD,1 Jeffrey D. Rothstein MD, PhD,1 and Mahmoud A. Pouladi PhD2* 1 Johns Hopkins University, Brain Science Institute, Baltimore, Maryland, USA Translational Laboratory in Genetic Medicine (TLGM), Agency for Science Technology and Research (A*STAR), and Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore (NUS), Singapore
2
ABSTRACT:
The discovery of the causative mutations and many of the predisposing risk factors for neurodegenerative disorders such as Amyotrophic Lateral Sclerosis, Alzheimer’s, Parkinson’s, and Huntington’s disease (HD), has led to the development of a large number of genetic animal models of disease. In the case of HD, for example, over 20 different transgenic rodent models have been generated. These models have been of immense value in providing novel insights into mechanisms of disease, with the promise of accelerating the development of therapies that can
For many neurodegenerative diseases, including Amyotrophic Lateral Sclerosis (ALS), Alzheimer’s, Parkinson’s, and Huntington’s disease (HD), the causative mutations and many of the predisposing genetic risk factors have been known for many years. In the case of HD, for example, the mutation in HTT was identified more than 20 years ago, enabling the development of genetic animal models and leading to a wealth of knowledge and novel insights into mechanisms of disease. Yet, despite the availability and intensive use of such preclinical models, a treatment that can delay the onset of disease or slow its rate of progression is yet to be developed for HD or any of the other neurodegenerative diseases. This leads one to raise the question: Are preclinical models needed in translational research?
------------------------------------------------------------
*Correspondence to: Mahmoud A. Pouladi, Translational Laboratory in Genetic Medicine (TLGM), Agency for Science Technology and Research and National University of Singapore, 8A Biomedical Grove, Immunos L5, Singapore 138648. E-mail: [email protected] Funding agencies: M.A.P. is supported by grants from the Agency for Science Technology and Research (A*STAR), and the National University of Singapore (NUS). Relevant conflicts of interest/financial disclosures: Nothing to report. Full financial disclosures and author roles may be found in the online version of this article. Received: 5 August 2014; Accepted: 11 August 2014 Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/mds.26010
delay the onset or slow the progression of the disease. Yet, despite extensive use of such models, no effective treatment for HD has been developed. Here, we discuss the value of animal models, highlighting their strengths and shortcomings in the context of translaC 2014 International Parkinson tional research for HD. V and Movement Disorder Society
K e y W o r d s : Animal models; Huntington’s; HTT; transgenic; translation
CON: The Case Against Preclinical Models in HD Many different animal models for studying HD have been generated, ranging from nonhuman primates, mice, Caenorhabditis elegans, and Drosophila melanogaster and zebrafish. Many of these animal models develop a phenotype reminiscent of some aspects of disease as seen in humans, including the loss of gaminobutyric acid–ergic medium spiny neurons (MSN neurons) in the striatum (as reviewed in Pouladi et al.1). In fact, interference with specific disease pathways in these models has led to improvement of the disease phenotype and an improvement of our understanding of some basic biological aspects of neurodegeneration in HD. None of these models, especially the mice models of HD, have so far led to the development of an effective treatment that can halt or slow the disease from progressing in HD patients. We describe several characteristics of preclinical models of HD that often limit their translational value for studying HD in humans.
Caveats of Small Animal Models to Study HD Small animal models such as zebrafish, Drosophila, and C. elegans do have the advantage of being relatively cheap to maintain and being a great tool for
Movement Disorders, Vol. 29, No. 11, 2014
1391
P H I L I P S
E T
A L
high-throughput genetic or drug screenings. However, small animal models have the disadvantage of not being very biologically related to humans in terms of their phylogenetic differences and complexity of their nervous system, particularly of relevance for HD, in which patients’ cognitive and psychiatric functions are highly affected. In addition, some of the small animal models lack a corresponding disease-causing genetic homolog, for example, the C. elegans models for HD, in which human mutant Huntingtin protein (mHTT) is inserted in the worm genome to model disease.2 In both these small but also the larger animal models discussed later, many caveats are associated with insertion of exogenous genes in a host genome (reviewed in O’Sullivan et al.3). Other caveats are associated with the design of small animal models. For example, HD is caused by a dominant mutation in a single gene encoding huntingtin (HTT). Because the HTT gene is a particularly large gene (approximately 170 kbp), developing HD animal models has been very difficult. Therefore, many HD models only express a (small) part of the mHTT gene, with excluded but potentially pathological regions of the mHTT gene not represented in the phenotype being observed.2,4 C. elegans and Drosophila models of HD usually express either truncated or full-length mHTT and often express mHTT from complementary DNA constructs, potentially failing to recapitulate pathological events related to other regions of the mHTT gene such as mHTT introns.2,5,6 Many caveats are associated with the phenotype that small animal models of neurodegeneration develop. These often do not recapitulate the unique spectrum of the disease phenotype as seen in patients with neurodegenerative disease. In HD, specific hallmarks of the disease such as dysphagia and dysarthria are clearly very difficult to model in such small animal models. Drosophila models clearly recapitulate motor dysfunction as seen in reduced climbing behavior (negative geotaxis),6 but this is phenotypically very similar to what is seen in Drosophila models of other, but very different, neurodegenerative disease such as ALS.7 Distinctive motor symptoms such as the rigidity and chorea seen in HD patients, as well as the cognitive and psychiatric features characterizing the HD phenotype are very hard to model using small animal models. Therefore, these studies using small animal models might rather lead to enhanced insight into less subtle disease mechanisms that are relevant for a broader range of neurodegenerative diseases.
Caveats of Rodent and Large Animal Models to Study HD Similarly to the small HD models, one has to be very careful about the biological differences between rodent models and humans. These might confound proper translation to the human clinic. For example, given the
1392
Movement Disorders, Vol. 29, No. 11, 2014
differences in brain mass (3,0003 larger in humans as compared with mice) and number of brain neurons (86 billion in humans versus 90 million in mice), and the much larger complexity of brain structure in humans, one should not be surprised that accurately modeling human disease in rodent HD models is very difficult. Therefore, although particular aspects of human HD are well modeled in rodents (see later discussion), other aspects of human HD such as facial movements, oculomotor deficiencies, dysarthria, and dysphagia are very difficult to model in rodent models of HD. More than 20 different transgenic rodent models expressing the human or mouse HTT gene have been generated (reviewed in Pouladi et al.1). These rodents express either truncated or full-length fragments of mHTT with varying CAG repeat length. Although these rodent models develop symptoms of cognitive loss and motor neuron dysfunction, some major caveats are associated with the development of these rodent models. Adequate gene dosage is essential during the development of animal models of neurodegeneration. High gene dosage in transgenic animals, or long repeat lengths in models of expansion disorder diseases such as HD, often lead to a fast progressive disease phenotype because of the ubiquitous expression of a high copy number or long repeat length of a transgene. This could be motivated by the financial aspects of animal research, which is usually a significant financial burden for a research facility, as well as the publication pressure in a usually highly competitive environment. This fast progressive disease phenotype often precludes the analysis of specific pathology related to human HD. Conversely, using these models, one might try to explore whether alterations in expression of genes affecting aging might modify the phenotype. In HD, mice overexpressing truncated mHTT often develop very-early-onset, fast progressive fatal disease that is more reminiscent of a generalized degenerative phenotype rather than specifically recapitulating HD symptoms as seen in patients.4 Therefore, in HD rodent models, the generalized expression of a CAG repeat expansion in cell types only mildly involved in human HD could have significant impact on the presentation of the phenotype in these rodents, potentiating fast progression and early demise. For example, mHTT expression in mouse motor cortex or other brain regions, whose corresponding brain regions in human HD are more mildly affected, might contribute to a very fast progressive motor phenotype, which is usually not seen in HD patients. Transgenic animal models expressing more modest levels of the transgene, similar to the levels of the endogenous gene, or mouse models of repeat expansion disorders with limited repeat length, might circumvent this issue but often lead to a phenotype that is only partial or very mild1,8 and might take a long time to develop. This might be more relevant for
P R E C L I N I C A L
modeling disease as seen in patients, because HD onset usually occurs in mid-adult life and might take many years or decades to progress until significant burden in the patient is noticed. Very often in such instances the animal model might not live long enough to develop the whole plethora of the neurodegenerative phenotype as seen in patients. These rodent versions of HD often show very simplistic clinical representations of the disease and lack the many relevant nuances (clinical and pathological) seen in human HD. For example, HD patients develop significant loss of MSN neurons in the striatum. Conversely, a rat model of HD, the BACHD rat, develops many alterations in the striatal compartment but shows no loss of striatal neurons.9 In fact, many rodent models for HD only present certain aspects of disease, and effective therapeutic drug trials using HD rodents are of low translational value because their implications for many other disease-related mechanisms could not be explored in these models. When only partial aspects of disease are being targeted, that a clinical trial will fail in the earlier stages might not be surprising, because drug effects on other disease-relevant pathways have not been taken into account in the preclinical model. Nonhuman primate models might give the best correlation to human disease. The earliest generated nonhuman primate models for HD were generated by administration of neurotoxins that selectively ablated striatal neuronal subpopulations.10 These models do not recapitulate the phenotype seen in HD patients, who develop progressive widespread cortical, subcortical, and even peripheral nervous system degeneration, in contrast to the acute localized lesions seen in the neurotoxin models. Transgenic large animal models for HD have also been developed. Unfortunately, none of these models is a true HD animal model. Transgenic mHTT macaque monkeys develop a very aggressive phenotype, with early death and unusual symptoms that are not seen in early stages of human HD.11 Transgenic sheep and minipigs did not develop overt disease symptoms, however, although this could be because of the limited mHTT CAG repeat size these models expressed.12,13 Even if one is successful in developing a large animal model to study HD, still many inherent caveats will be associated with these studies. Large animal studies, especially those using nonhuman primate models most relevant to study human disease, are extremely expensive to maintain, and the neurodegenerative disease in these animals takes many years to develop. In fact, HD onset in humans is usually midadult life, and so disease progression might need to be studied during several decades. Even if they were to develop disease, given the genetic variability of primates (like humans), this would likely necessitate a large number of animals—similar to human trials—to see a positive and reproducible clinical response. This makes
M O D E L S :
N E E D E D
I N
T R A N S L A T I O N ?
their use for drug screening suboptimal, and studies using nonhuman primate models usually take many years if not decades to complete.
Alternatives to Preclinical Animal Models: Value of iPS Cells in Studying HD Rather than using animal models to study HD disease pathogenesis and treatment opportunities, one could consider the use of patient-derived induced pluripotent stem (iPS) cells. The development of iPS cells with a patient-specific disease signature could be explored both for understanding disease pathogenesis as well as for testing drugs for therapeutical efficacy. The iPS cells could be differentiated into different central nervous system cell types such as neurons, astrocytes, and oligodendrocytes. Protocols have recently been established for specifically generating MSN neurons from iPS cells.14 The concept that neuronal death is non–cell autonomous, that is, not only neurons but also specific glial cell types and other neuronal cell types contribute to the disease pathogenesis, is well established.15 For example, in HD mouse models, astrocyte GLT-1 glutamate transporters are significantly reduced, which contributes to glutamate-mediated excitotoxic cell death of the nearby g-aminobutyric acid–ergic neurons most susceptible in HD.16 The ability to model all contributing cell types in a dish, by analyzing a patient-specific disease signature with effects of co-morbidity and disease modifier genes incorporated as well, is most relevant for the specific HD patient under study. Of particular interest is that in all HD patients, disease is caused in a monogenetic fashion by a CAG repeat expansion of varying length. Unlike in other neurodegenerative diseases in which most cases are sporadic with the underlying cause still at large, specific targeting of the HD disease–causing gene using RNAse H active antisense oligonucleotides or small interfering RNAs targeting single nucleotide polymorphisms associated with the expanded allele is of interest to all HD patients. A recent study has shown that transient infusion of antisense oligonucleotides into the cerebrospinal fluid effectively targeted the mHTT gene and reversed the phenotype in HD models.17 In addition, infused nonhuman primate models showed a clear reduction in wild-type HTT protein expression in many brain regions.17 The development of iPSderived neuronal and non-neuronal cells from individual HD patients could at least in part bypass the need for these preclinical therapeutic trials in HD animal models by providing us a great tool to assess the feasibility and therapeutic dosage of antisense treatment specifically designed for the patient under study. In addition, this antisense therapy also could be directed against many potential disease modifier genes in highthroughput screenings. In fact, many different genes have been reported to aggravate or alleviate HD
Movement Disorders, Vol. 29, No. 11, 2014
1393
P H I L I P S
E T
A L
disease symptoms.18 Conversely, studying these cells in a dish, extracted from their usual microenvironment, might lead to significant loss of their translational value. Therefore, whether specific HD disease– related phenotypes can be effectively mimicked in the iPS-derived cell culture system remains to be seen.
sion.22 That animal models allow the examination of the neuroprotective potential of cellular and molecular targets implicated in HD provides a degree of reassurance before the initiation of potentially complicated and likely costly clinical trials, further emphasizing the importance of animal models in translational studies.
PRO: The Case for Preclinical Models in HD
High Construct and Face Validity of Animal Models of HD
In spite of the various caveats about the HD models, preclinical animal, and in particular rodent, models of HD have played an indisputably important role in identifying disease processes and elucidating cellular pathways disrupted in Huntington’s disease.1 Indeed, much of our current understanding of the pathogenic mechanisms implicated in HD has been shaped by studies of animal models of HD.19 In addition to their utility in discovering disrupted cellular processes, animal models have been used to evaluate the impact of normalizing the identified cellular and molecular deficits on diseaserelated phenotypes,20,21 shedding light on their potential as targets for therapeutic intervention. Despite developments that have led to the value of animal models in translational research being debated, animal models of HD will remain important in bridging the translational gap from basic discoveries to trials in patients with HD for a number of reasons. Because rare disorders such as HD have limited patient populations that are available to participate in the trial of potential therapies, approaches that help build confidence and provide evidence, even if only in proof-of-concept form, that supports a given therapeutic strategy are of immense importance. Such data provide an additional criterion by which to prioritize experimental therapies and may help in the nomination of promising candidates for clinical evaluation. In this regard, animal models provide an opportunity to test the therapeutic potential of candidate compounds before initiating clinical trials in the limited pool of patients with HD that may compromise their ability to participate in subsequent trials.
Value of Proof-of-Concept Studies of Neuroprotection Furthermore, although treatments that target symptoms are generally the first line of therapies developed for many neurodegenerative disorders, including HD, the ultimate goal of translational research is not only to provide temporary relief of symptoms but also to delay the onset and slow the progression of disease.22 For these slow progressing diseases, the development of such neuroprotective therapies is challenging not only from a clinical trial design point of view, but also economically because of the longer duration of trials needed to ascertain treatment effects on disease progres-
1394
Movement Disorders, Vol. 29, No. 11, 2014
Unlike many forms of neurodegenerative diseases, HD is extraordinary in its monogenic etiology and its dominant and highly penetrant presentation.23 These features entail not only a clearly defined patient population, but also animal models of high apparent validity. Three broad criteria are often used to judge a given model’s relevance: construct validity, face validity, and predictive validity.1 Many animal models of HD are considered to be of high construct validity given that they closely reproduce the single genetic mutation that underlies HD in humans, namely CAG trinucleotide repeat expanded HTT allele(s).1 Similarly, animal models of HD exhibit high face validity, recapitulating many phenotypic features of HD, including progressive motor, cognitive, and psychiatric-like disturbances coupled with HD-like neuropathological and molecular changes.1 In terms of predictive validity, which is gauged based on how closely response to treatment in the animal model parallel, or predict, improvements in patients, this aspect is yet to be proven, because no disease-modifying therapies have been developed. Indeed, failure of a therapy shown to be effective in humans to improve disease phenotypes in animal models would be cause for pause. With respect to the only treatment currently approved for HD,24,25 which is mostly of symptomatic benefit, evidence indicates that tetrabenazine improves disease phenotypes in an animal model of HD.26 Furthermore, the outcome of a given clinical trial is influenced by a number of factors, which include not only the role of the target being engaged in the disease process, but also issues of dosage, frequency, disease stage, and specific outcomes being assessed. This can be readily appreciated when considering that for many clinically approved therapies, multiple clinical trials of a given treatment were often necessary to arrive at the right combination of dose, disease stage, and outcome measure. Thus, for treatment approaches showing robust and reproducible benefits in animal models, caution is warranted when interpreting discordance between animal and human trials in which the disease stage or outcome measures may have been different.
Availability of Versatile Animal-Based Experimental Tools to Dissect Disease Processes and Mechanism of Action of Therapeutic Candidates Safe, specific, and brain-permeable pharmacological compounds are available for a relatively limited
P R E C L I N I C A L
number of molecular targets, and in most cases, screening for and development of novel compounds to engage molecular targets of interest, a nontrivial undertaking, is required. The ability to assess the impact of specifically enhancing or decreasing the activity of a given target before the initiation of potentially costly and lengthy drug screening and development efforts is of great value. Since the generation of the first transgenic rodent model over three decades ago,27 thousands of genetically modified animal lines have been established that make such studies possible. By cross-breeding to animal models of HD, these lines have allowed for the interrogation of the effect of increased or decreased expression, and hence activity, of a given gene of interest on HD-related phenotypes. Furthermore, the availability of transgenic animal technologies, such as inducible and tissue-specific Cre recombinase lines,28 allows the dissection of timedependent and tissue-specific cellular processes involved in HD. These versatile transgenic animalbased tools have enabled a better understanding of the pathogenesis of HD and will continue to play an important role in providing novel insights into HD.2931 Transgenic animal lines also may be of translational value in examining the specificity and the mechanism of action of therapeutic candidates of interest.
Alternative Approaches Are Not as Well Developed An exciting recent development that promises to revolutionize the study of human disease, and in particular brain disorders, is the discovery that somatic and terminally differentiated cells may be induced to a pluripotent stem cell state that may subsequently be differentiated into any cell type of interest.32 For brain disorders such as HD, a major challenge has been the inability to directly study the cell type most affected in the disease, namely striatal medium spiny neurons, in a human context. The ability to generate iPS cells from readily accessible patient tissues (eg, fibroblasts or peripheral blood monocytes), coupled with the development of genome editing tools, such as the transcription activator–like effector nucleases and the RNA-based clustered regularly interspaced short palindromic repeats–Cas9 system,33 which enable the correction of the HD mutation, provides an opportunity to study pathogenic processes and to evaluate potential therapies in isogenic stem cell–derived MSNs of human origin. This potential has indeed been demonstrated in a number of recent studies.34-37 The value of stem cell–derived human neurons in the study of HD cannot be overestimated; however, a number of factors may limit their utility, at least in the short-term. First, whereas in animals isolating and studying virtually any cell type of interest is possible, the ability to do the same using human stem cells
M O D E L S :
N E E D E D
I N
T R A N S L A T I O N ?
depends largely on the availability of the appropriate directed differentiation protocols necessary to derive the cell type of interest. Furthermore, with respect to MSNs, the most highly validated protocols available to derive them from human stem cells are relatively inefficient, with only a small fraction (10-15%) of the differentiated neurons exhibiting the desired MSN identity at the conclusion of the differentiation procedure.14,38 Second, in their simplest and most widely adopted configuration, studies of human stem cell– derived neurons neglect to recreate the complex milieu that includes other central nervous system cell types, such as astrocytes, microglia, and oligodendrocytes, which may be key to recapitulating disease phenotypes. The importance of non–cell autonomous effects in HD disease processes is indeed supported by a number of recent animal studies.29,31,39-41 By failing to model many of the environmental cues in stem cell– based neuronal models, important non–cell autonomous disease processes may be missed. Finally, agedependent progression of whole system, complex phenotypes, in particular cognitive, motor, and psychiatric features, which can be readily assessed in animal models of HD, simply cannot be examined in stem cell– derived neuronal systems. Thus, much room for improvement in stem cell–based modeling of HD remains, and the exciting advances in our ability to study HD disease processes in neurons and cell types of human origin will serve to complement studies in animal models of HD.
Conclusion Numerous factors argue for the continued use of preclinical animal models of HD in translational research, including the fact that the patient population available to participate in clinical trials is limited; the costs of failure for trials seeking to document neuroprotection are high; the available animal models show a high level of validity because of HD’s dominant, monogenic, and highly penetrant nature; and finally, the available alternative experimental systems are less developed.
References 1.
Pouladi MA, Morton AJ, Hayden MR. Choosing an animal model for the study of Huntington’s disease. Nat Rev Neurosci 2013;14: 708-721.
2.
Faber PW, Alter JR, MacDonald ME, Hart AC. Polyglutaminemediated dysfunction and apoptotic death of a Caenorhabditis elegans sensory neuron. Proc Natl Acad Sci USA 1999;96:179-84.
3.
O’Sullivan GJ, O’Tuathaigh CM, Clifford JJ, O’Meara GF, Croke DT, Waddington JL. Potential and limitations of genetic manipulation in animals. Drug Discov Today Technol 2006;3:173-180.
4.
Mangiarini L, Sathasivam K, Seller M, et al. Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice. Cell 1996;87:493506.
Movement Disorders, Vol. 29, No. 11, 2014
1395
P H I L I P S
5.
6.
E T
A L
Jackson GR, Salecker I, Dong X, et al. Polyglutamine-expanded human huntingtin transgenes induce degeneration of Drosophila photoreceptor neurons. Neuron 1998;21:633-642. Romero E, Cha G-H, Verstreken P, et al. Suppression of neurodegeneration and increased neurotransmission caused by expanded full-length huntingtin accumulating in the cytoplasm. Neuron 2008;57:27-40.
7.
Li Y, Ray P, Rao EJ, et al. A Drosophila model for TDP-43 proteinopathy. Proc Natl Acad Sci U S A 2010;107:3169-3174.
8.
Tieu K. A guide to neurotoxic animal models of Parkinson’s disease. Cold Spring Harb Perspect Med 2011;1:a009316.
9.
Yu-Taeger L, Petrasch-Parwez E, Osmand AP, et al. A novel BACHD transgenic rat exhibits characteristic neuropathological features of Huntington disease. J Neurosci 2012;32:15426-15438.
10.
Beal MF, Kowall NW, Ellison DW, Mazurek MF, Swartz KJ, Martin JB. Replication of the neurochemical characteristics of Huntington’s disease by quinolinic acid. Nature 1986;321:168-171.
11.
Yang SH, Cheng PH, Banta H, Piotrowska-Nitsche K, Yang JJ, Cheng EC, Snyder B, Larkin K, Liu J, Orkin J, Fang ZH, Smith Y, Bachevalier J, Zola SM, Li SH, Li XJ, Chan AW. Towards a transgenic model of Huntington’s disease in a non-human primate. Nature 2008;453:921-924.
12.
Jacobsen JC, Bawden CS, Rudiger SR, et al. An ovine transgenic Huntington’s disease model. Hum Mol Genet 2010;19:1873-1882.
13.
Yang D, Wang C-E, Zhao B, et al. Expression of Huntington’s disease protein results in apoptotic neurons in the brains of cloned transgenic pigs. Hum Mol Genet 2010;19:3983-3994.
14.
Delli Carri A, Onorati M, Castiglioni V, et al. Human pluripotent stem cell differentiation into authentic striatal projection neurons. Stem Cell Rev 2013;9:461-474.
15.
Ilieva H, Polymenidou M, Cleveland DW. Non-cell autonomous toxicity in neurodegenerative disorders: ALS and beyond. J Cell Biol 2009;187:761-772.
16.
Behrens PF, Franz P, Woodman B, Lindenberg KS, Landwehrmeyer GB. Impaired glutamate transport and glutamateglutamine cycling: downstream effects of the Huntington mutation. Brain 2002;125:1908-1922.
23.
Myers RH. Huntington’s disease genetics. NeuroRx 2004;1:255-262.
24.
Huntington Study Group. Tetrabenazine as antichorea therapy in Huntington disease: a randomized controlled trial. Neurology 2006;66:366-372.
25.
Hayden MR, Leavitt BR, Yasothan U, Kirkpatrick P. Tetrabenazine. Nature Rev Drug Discovery 2009; 17-18.
26.
Wang H, Chen X, Li Y, Tang T-S, Bezprozvanny I. Tetrabenazine is neuroprotective in Huntington’s disease mice. Mol Neurodegener 2010;5:18.
27.
Hanahan D, Wagner EF, Palmiter RD. The origins of oncomice: a history of the first transgenic mice genetically engineered to develop cancer. Genes Dev 2007;21:2258-2270.
28.
Nagy A. Cre recombinase: the universal reagent for genome tailoring. Genesis 2000;26:99-109.
29.
Wang N, Gray M, Lu X-H, et al. Neuronal targets for reducing mutant huntingtin expression to ameliorate disease in a mouse model of Huntington’s disease. Nat Med 2014;20:536-541.
30.
Hult S, Soylu R, Bj€ orklund T, et al. Mutant huntingtin causes metabolic imbalance by disruption of hypothalamic neurocircuits. Cell Metab 2011;13:428-439.
31.
Crotti A, Benner C, Kerman BE, et al. Mutant Huntingtin promotes autonomous microglia activation via myeloid lineagedetermining factors. Nat Neurosci 2014;17:513-521.
32.
Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 2006;126:663-676.
33.
Kim H, Kim J-S. A guide to genome engineering with programmable nucleases. Nat Rev Genet 2014;15:321-334.
34.
HD iPSC Consortium. Induced pluripotent stem cells from patients with Huntington’s disease show CAG-repeat-expansion-associated phenotypes. Cell Stem Cell 2012;11:264-278.
35.
An MC, Zhang N, Scott G, Montoro D, Wittkop T, Mooney S, et al. Genetic correction of Huntington’s disease phenotypes in induced pluripotent stem cells. Cell Stem Cell 2012;11:253-263.
36.
Kaye JA, Finkbeiner S. Modeling Huntington’s disease with induced pluripotent stem cells. Mol Cell Neurosci 2013;56:50-64.
17.
Kordasiewicz HB, Stanek LM, Wancewicz EV, et al. Sustained therapeutic reversal of Huntington’s disease by transient repression of huntingtin synthesis. Neuron 2012;74:1031-1044.
37.
Ross CA, Akimov SS. Human-induced pluripotent stem cells: potential for neurodegenerative diseases. Hum Mol Genet 2014; May 13.
18.
Clabough EBD. Huntington’s disease: the past, present, and future search for disease modifiers. Yale J Biol Med 2013;86:217.
38.
19.
Ross CA, Tabrizi SJ. Huntington’s disease: from molecular pathogenesis to clinical treatment. Lancet Neurol 2011;10:83-98.
Carri AD, Onorati M, Lelos MJ, et al. Developmentally coordinated extrinsic signals drive human pluripotent stem cell differentiation toward authentic DARPP-321 medium-sized spiny neurons. Development 2013;140:301-312.
20.
Gil-Mohapel JM. Screening of therapeutic strategies for Huntington’s disease in YAC128 transgenic mice. CNS Neurosci Ther 2012;18:77-86.
39.
Gu X, Li C, Wei W, et al. Pathological cell-cell interactions elicited by a neuropathogenic form of mutant Huntingtin contribute to cortical pathogenesis in HD mice. Neuron 2005;46:433-444.
21.
Gil JM, Rego AC. The R6 lines of transgenic mice: a model for screening new therapies for Huntington’s disease. Brain Res Rev 2009;59:410-431.
40.
Gu X, Andre VM, Cepeda C, et al. Pathological cell-cell interactions are necessary for striatal pathogenesis in a conditional mouse model of Huntington’s disease. Mol Neurodegen 2007;2:8.
22.
Katsuno M, Tanaka F, Sobue G. Perspectives on molecular targeted therapies and clinical trials for neurodegenerative diseases. J Neurol Neurosurg Psychiatry 2012;83:329-335.
41.
Tong X, Ao Y, Faas GC, et al. Astrocyte Kir4.1 ion channel deficits contribute to neuronal dysfunction in Huntington’s disease model mice. Nat Neurosci 2014;17:694-703.
1396
Movement Disorders, Vol. 29, No. 11, 2014
