A multimodal approach to identify clinically relevant biomarkers to comprehensively monitor disease progression in a mouse model of pediatric neurodegenerative disease

https://doi.org/10.1016/j.pneurobio.2020.101789Get rights and content

Highlights

  • Longitudinal neuroimaging reveals changes in Cln6nclf mice that correlate with changes reported in patients with Batten disease.

  • Brain region-specific degeneration in the cortex, striatum, hippocampus, and cerebellum results from CLN6 dysfunction in mice.

  • CLN6 disease progression in mice leads to widespread brain metabolism defects and white matter loss as measured by FDG-PET and DTI imaging, respectively.

  • Kinematic gait analysis combined with principal component analysis identifies sex-dependent gait variations that change dramatically with disease progression.

  • Contrastive principal component analysis of longitudinal data shows progressive changes in Cln6nclf mice that exceed those of any single phenotype with traditional metrics.

Abstract

While research has accelerated the development of new treatments for pediatric neurodegenerative disorders, the ability to demonstrate the long-term efficacy of these therapies has been hindered by the lack of convincing, noninvasive methods for tracking disease progression both in animal models and in human clinical trials. Here, we unveil a new translational platform for tracking disease progression in an animal model of a pediatric neurodegenerative disorder, CLN6-Batten disease. Instead of looking at a handful of parameters or a single “needle in a haystack”, we embrace the idea that disease progression, in mice and patients alike, is a diverse phenomenon best characterized by a combination of relevant biomarkers. Thus, we employed a multi-modal quantitative approach where 144 parameters were longitudinally monitored to allow for individual variability. We use a range of noninvasive neuroimaging modalities and kinematic gait analysis, all methods that parallel those commonly used in the clinic, followed by a powerful statistical platform to identify key progressive anatomical and metabolic changes that correlate strongly with the progression of pathological and behavioral deficits. This innovative, highly sensitive platform can be used as a powerful tool for preclinical studies on neurodegenerative diseases, and provides proof-of-principle for use as a potentially translatable tool for clinicians in the future.

Introduction

Rare diseases, conditions that affect fewer than 200,000 patients in the U.S. or less than 1 in 2000 people in the EU (NIH, 2017), represent a particular challenge for medical diagnosis as clinical features are often complex and enigmatic. While very few rare diseases have effective treatments, resulting from the limited information that is typically available for many of these conditions, access to improved animal models and state-of-the art medical diagnostic capabilities are helping to accelerate the number of clinical trials and treatments available to patients. Due to their rarity, access to patients is particularly limited, so researchers and clinicians must rely on comprehensive natural history studies that provide a snapshot of where a typical patient would exist in time. Moreover, because many of these diseases are pediatric and ultimately fatal, the 21st Century Cures Act made it possible to accelerate the clinical trial design process, with one concession being that the trial need not include untreated controls, thus making the natural history data even more essential. Although much attention has been focused on collecting this information, much of what is captured can be subjective and qualitative. Thus, quantitative biomarkers that can be monitored longitudinally and are minimally invasive are greatly needed in order to monitor treatment responses in both preclinical animal models and human clinical trials. Translational utilization of animal models of human disease benefits greatly from relevant phenotypic characterization. Unfortunately, the techniques most commonly used in animal models often suffer from a lack of translatability. Behavioral assays for mice most often focus on murine-relevant behaviors that are not necessarily applicable to the clinic. Similarly, mouse pathology is typically focused on invasive post-mortem analysis of tissues, processes that are not practiced in human patients. This lack of translatability renders many preclinical phenotypic characterizations difficult to translate to the clinic.

Comprehensive noninvasive biomarker panels are not available for many neurodegenerative disorders, and these are of special interest for pediatric disorders, as determining disease progression early on is critical to identifying proper clinical interventions and monitoring responses to potential corrective therapies. Batten disease (i.e., neuronal ceroid lipofuscinoses), a family of lysosomal storage disorders resulting from mutations in one of 13 genes, collectively represents the most common neurodegenerative disease in children (Santavuori et al., 2000; Rider and Rider, 1988). Although the functions of many of these genes are unknown, much work in the past decade has been dedicated to developing and testing therapies. With the recent advancements in research tools, including high-throughput and high-content screening methods, the Batten disease scientific community has been progressing toward potential therapies at an unprecedented pace, and as a result, treatments are moving from the preclinical phase to clinical trials more quickly and efficiently than ever before. Also, with such heterogeneous disease states, resulting from different mutations that lead to more aggressive or protracted forms of the disease, clinical research teams are forming large, international collaborations to ensure that comprehensive natural history studies are completed and in place as a resource for all clinical trials. These interdisciplinary groups have paved the way for these natural history studies, led by the DEM-CHILD NCL Patient Database Consortium and the University of Rochester Batten Center (Schulz et al., 2015; Adams et al., 2013). Additionally, these groups have developed clinical rating scales to assess cognitive, motor, and behavioral function of patients with Batten disease (Marshall et al., 2005; Steinfeld et al., 2002; Worgall et al., 2007). With a growing number of clinical trials for Batten disease therapies, there is an increasing need for noninvasive, clinically-relevant biomarkers to track therapeutic efficacy.

To address these needs, we used a mouse model of CLN6 Batten disease to perform an exhaustive multi-factorial characterization of biomarkers, moving away from mouse behavioral assays to clinical outcomes that are congruent to those used in human patients. CLN6 disease is caused by autosomal recessive mutations in CLN6, which results in the reduction or complete absence of the CLN6 protein. This disease is characterized by the accumulation of autofluorescent storage material in lysosomes, progressive neurodegeneration throughout cortex and thalamus, as well as massive gliosis throughout the central nervous system (CNS). Patients often present with language deficits, cognitive impairment and progressive motor decline. As the disease progresses, patients lose vision, develop seizures and ultimately succumb to the disease around 10–12 years of life. The spontaneously occurring Cln6nclf mouse model of CLN6 disease has been shown to faithfully recapitulate many of the hallmarks of the human disease both behaviorally and pathologically (Morgan et al., 2013a; Bronson et al., 1998), but noninvasive, clinically relevant assays have not yet been employed to characterize longitudinal changes in this model. Various scientists, our lab included, have performed a very comprehensive pathological assessment of Batten disease rodent and large animal models to reveal how different brain regions change over time, however, these experiments were all conducted on post-mortem brain samples (Thelen et al., 2012; Kielar et al., 2009; Oswald et al., 2008; Cooper et al., 2006; Kay et al., 2006; Oswald et al., 2005; Weimer et al., 2006, 2009; Kuhl et al., 2013; Parviainen et al., 2017). These studies have shown that brain pathology is present months before any noted behavioral changes, similar to what has been noted in human patients. Thus, cellular changes and degeneration are occurring in the the brain long before one notices any behavior or cognitive changes, so developing more sensitive tools for detecting disease states prior to the onset of behavioral symptoms would be of immense value in the clinic. Our previous work, as well as the work of others, has largely been focused on finding a single biomarker – the elusive “needle in a haystack” - associated with Batten disease and its progression (Chattopadhyay et al., 2002a, b; Hersrud et al., 2016; Timm et al., 2018a), but this approach has failed to yield any reliable metrics.

In this study, rather than focusing on one or a few metrics, we used multiple imaging modalities as well as a comprehensive gait assessment to track hundreds of parameters over time, and performed a combinatorial analysis to identify a biomarker signature for CLN6-disease. Cohorts of wild type and Cln6nclf mice of both sexes were monitored longitudinally using noninvasive imaging, including T2-weighted magnetic resonance imaging (T2-MRI), diffusion tensor MRI (DTI), 1H magnetic resonance spectroscopy (MRS), and positron emission tomography (PET) was performed periodically between 3–12 months of age while kinematic gait analysis (KGA), which captures a large number of metrics describing gait, was assayed from 6 to 12 months of age. Variations of all of these techniques are widely used in the clinic and have shown correlations between mouse models and human subjects (Broom et al., 2017; Waerzeggers et al., 2010). Once we had characterized these parameters in the CLN6 disease mouse model, we used a recently developed form of principal component analysis (PCA), contrastive PCA (cPCA), to cluster the four imaging modalities and gait analysis in order to derive new variables that best capture and define the progressive nature of the disease. Together, this approach provides a robust and translatable platform for longitudinal monitoring of disease progression that can have profound utility, not only for Batten disease, but in a variety of animal models of rare neuropediatric diseases.

Section snippets

Progressive changes in brain volume and anatomy in a model of CLN6 disease

The selective vulnerability of various populations of neurons is a key feature of many neurodegenerative diseases, including Batten disease (Braak and Goebel, 1978; Cooper, 2010). Prior studies have demonstrated that thinning of select anatomical regions and different cortical layers is present in Cln6nclf mice (Morgan et al., 2013a; Kielar et al., 2009; Cooper et al., 2006; Kuhl et al., 2013), however, all of these measurements are based on invasive, histopathological analysis of post-mortem

Discussion

Pediatric neurodegenerative disorders are typically diagnosed by a combination of clinical assessment, neuroradiologic imaging, cellular pathology, and genetic testing (Kohan et al., 2009, 2015; Mole and Cotman, 2015). Unfortunately, even with advancements in genomic diagnostics, seldom are children definitively diagnosed following initial clinical assessment, and more typically, numerous misdiagnoses are offered before a correct diagnosis (Fietz et al., 2016). The clinical features of Batten

Study design

Neuroimaging, gait analysis, and principal component analyses were conducted on aged-matched wildtype and Cln6 mutant mice (description of animals in Ethics Statement/Animals) to determine a longitudinal biomarker scoring system in a preclinical model of neuropediatric disease. We hypothesized that by looking at various minimally-invasive disease markers as a system, rather than individually, we could provide a highly sensitive tool that may be translatable to the clinic. Sample size,

Funding

This work was supported by funding to J.M.W. (NIHR01NS082283) as well as institutional support from Sanford Research and Charles River Laboratories.

Author contributions

K.K.L, A.N and J.M.W conceptualized and designed the studies. D.T. bred and shipped the mice. K.K.L J.R., and T.B. conducted the experiments and generated data. T.B.J., J.J.B., K.K.L, J.T.C, K.A.W., T.B., J.R. T.B., A.N., and J.M.W. critically analyzed and plotted data. T.B.J and J.J.B wrote the manuscript. T.B.J, J.J.B, K.K.L, J.T.C, K.AW, T.B, J.R., T.H., M.V., J.T.P., A.N, and J.M.W. reviewed and edited the manuscript.

Data and materials availability

All data generated in this study can be found in the paper or supplemental materials or provided upon request.

Declaration of Competing Interest

The authors declare no competing interests

References (85)

  • R. Kohan

    The neuronal ceroid lipofuscinoses program: a translational research experience in Argentina

    Biochim. Biophys. Acta

    (2015)
  • J.R. Moffett et al.

    N-Acetylaspartate in the CNS: from neurodiagnostics to neurobiology

    Prog. Neurobiol.

    (2007)
  • S.E. Mole et al.

    Genetics of the neuronal ceroid lipofuscinoses (Batten disease)

    Biochim. Biophys. Acta

    (2015)
  • L.J. O’Donnell et al.

    An introduction to diffusion tensor image analysis

    Neurosurg. Clin. N. Am.

    (2011)
  • M.J. Oswald

    Glial activation spreads from specific cerebral foci and precedes neurodegeneration in presymptomatic ovine neuronal ceroid lipofuscinosis (CLN6)

    Neurobiol. Dis.

    (2005)
  • M.J. Oswald et al.

    Location and connectivity determine GABAergic interneuron survival in the brains of South Hampshire sheep with CLN6 neuronal ceroid lipofuscinosis

    Neurobiol. Dis.

    (2008)
  • J. Radke et al.

    Human NCL neuropathology

    Biochim. Biophys. Acta

    (2015)
  • P. Santavuori

    Neuronal ceroid-lipofuscinoses in childhood

    Brain Dev.

    (1988)
  • P. Santavuori et al.

    Neuronal ceroid lipofuscinoses in childhood

    Suppl. Clin. Neurophysiol.

    (2000)
  • P. Santavuori et al.

    Clinical and neuroradiological diagnostic aspects of neuronal ceroid lipofuscinoses disorders

    Eur. J. Paediatr. Neurol.

    (2001)
  • A. Schulz

    OP48 – 2893: the DEM-CHILD NCL patient database: a tool for the evaluation of therapies in neuronal ceroid lipofuscinoses (NCL)

    Eur. J. Paediatr. Neurol.

    (2015)
  • L.M. Schutte

    An index for quantifying deviations from normal gait

    Gait Posture

    (2000)
  • Y. Waerzeggers et al.

    Mouse models in neurological disorders: applications of non-invasive imaging

    Biochim. Biophys. Acta

    (2010)
  • J.M. Weimer

    Visual deficits in a mouse model of Batten disease are the result of optic nerve degeneration and loss of dorsal lateral geniculate thalamic neurons

    Neurobiol. Dis.

    (2006)
  • J.M. Weimer

    Cerebellar defects in a mouse model of juvenile neuronal ceroid lipofuscinosis

    Brain Res.

    (2009)
  • R.E. Williams

    Diagnosis of the neuronal ceroid lipofuscinoses: an update

    Biochim. Biophys. Acta

    (2006)
  • A. Abid et al.

    Exploring patterns enriched in a dataset with contrastive principal component analysis

    Nat. Commun.

    (2018)
  • H.R. Adams et al.

    Neurobehavioral features and natural history of juvenile neuronal ceroid lipofuscinosis (Batten disease)

    J. Child Neurol.

    (2013)
  • T. Autti et al.

    Magnetic resonance techniques in neuronal ceroid lipofuscinoses and some other lysosomal diseases affecting the brain

    Curr. Opin. Neurol.

    (1997)
  • E.H. Baker et al.

    MRI brain volume measurements in infantile neuronal ceroid lipofuscinosis

    AJNR Am. J. Neuroradiol.

    (2017)
  • M.F. Beal

    Aging, energy, and oxidative stress in neurodegenerative diseases

    Ann. Neurol.

    (1995)
  • H. Braak et al.

    Loss of pigment-laden stellate cells: a severe alteration of the isocortex in juvenile neuronal ceroid-lipofuscinosis

    Acta Neuropathol. (Berl.)

    (1978)
  • R.T. Bronson

    Neuronal ceroid lipofuscinosis (nclf), a new disorder of the mouse linked to chromosome 9

    Am. J. Med. Genet.

    (1998)
  • L. Broom

    A translational approach to capture gait signatures of neurological disorders in mice and humans

    Sci. Rep.

    (2017)
  • S. Chattopadhyay

    An autoantibody inhibitory to glutamic acid decarboxylase in the neurodegenerative disorder Batten disease

    Hum. Mol. Genet.

    (2002)
  • S. Chattopadhyay et al.

    An autoantibody to GAD65 in sera of patients with juvenile neuronal ceroid lipofuscinoses

    Neurology

    (2002)
  • J.D. Cooper

    The neuronal ceroid lipofuscinoses: the same, but different?

    Biochem. Soc. Trans.

    (2010)
  • S.B. Dolisca et al.

    Batten disease: clinical aspects, molecular mechanisms, translational science, and future directions

    J. Child Neurol.

    (2013)
  • H.H. Goebel et al.

    Current state of clinical and morphological features in human NCL

    Brain Pathol.

    (2004)
  • H.H. Goebel et al.

    Ultrastructural investigations of peripheral nerves in neuronal ceroid-lipofuscinoses (NCL)

    J. Neurol.

    (1976)
  • S.L. Hersrud et al.

    Plasma biomarkers for neuronal ceroid lipofuscinosis

    FEBS J.

    (2016)
  • A. Hervé et al.

    Principal component analysis

    Wiley Interdiscip. Rev. Comput. Stat.

    (2010)
  • Cited by (0)

    1

    Authors contributed equally.

    View full text