New opportunities for treatment of neurodegenerative disease through the modulation of TDP-43
Pasha Apontes, Ph.D1,2
1 Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, NY, USA. 2 Independent Neuroscientist 28 Pine Drive Oyster Bay, NY 11771
Abstract
Therapeutic options remain very limited for many neurodegenerative diseases. The transactive response DNA binding protein, TDP-43, has emerged as an important contributing factor. The TDP-43 protein has important roles in RNA processing (transcription, alternative splicing, exon-skipping, microRNA, mRNA stability), and the co-regulation of translation in conjunction with other regulatory factors. However, in many neurological diseases, TDP-43's cellular functions are abrogated or attenuated and, additionally, deleterious gain-of function changes emerge. There are a multitude of mechanisms underlying changes in the normal functionality of this protein, for example, DNA mutations, mislocalization, or post-translational modifications. Whereas TDP-43 is normally enriched in the nucleus, in disease states, aberrant forms of this critical protein form inclusion bodies and protein macroaggregates in the diverse compartments, notably, the cytoplasm, rough ER, micronuclei, and mitochondria. TDP-43 has been defined as a hallmark in the etiology of neurodegenerative disorders, notably familial and sporadic forms of ALS, even in the absence of a mutation in the TARDBP gene. It has been gaining wide-spread recognition as an important proteinopathy and, accordingly, a promising target for medicinal interventions. This chapter will focus on the therapeutic modulation of TDP-43 as a rational therapeutic approach to treating patients with gravely disabling cognitive and neurodegenerative disorders.
Keywords
amyotrophic lateral sclerosis (ALS) Alzheimer's Disease (AD), TDP-43, randomized clinical trial, cytoplasmic aggregates, mislocalization, and frontotemporal dementia (FTD), Pharmacological interventions and Drug discovery.
We look at the world once, in childhood. The rest is memory.
-Louise Glück
Overview of protein aggregates that occur in neurodegenerative disorders and rationale for targeting of TDP-43
There are a number of well-described pathological proteins which cluster or form aggregates in the nervous system, either a cause or an effect of neurodegenerative diseases. Among these, TDP-43 has been recognized as an important protein in many neurological diseases and was first identified by Neumann et al in 2006 as an important misfolded, ubiquitinated protein found in inclusions of patients with with frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) \cite{Neumann_2006}. It was first characterized as a factor that activates transcription of human immunodeficiency syndrome type 1 at the long terminal repeat sequence called TAR \cite{Ou_1995} and with the help of the RNA Helicase p68 (DDX5) \cite{Marciniak_1990}. The diminishment of the normal cellular functions of TDP-43 and the acquisition of new characteristics when it undergoes conformational changes and forms aggregates has been widely implicated in neurodegenerative diseases \cite{Neumann_2006}\cite{Suk_2020} . Besides TDP-43, there are other well-known proteins that similarly undergo conformational changes and that are likewise implicated in cognitive disorders. For example, Rho Guanine Nucleotide Exchange Factor (RGNEF) in Parkinson’s Disease (PD), α-synuclein, which accumulates in Lewy Bodies Dementia (LBD) and Multiple System Atrophy (MSA) and, β -amyloid and hyper-phosphorylated tau, which form plaques and neurofibrillary tangles, respectively, in Alzheimer's Disease (AD). It is somewhat uncertain as to whether one or more of these protein aggregates may represent compensatory responses to cellular stresses to forestall, or otherwise halt death of neuronal, astrocytes, or microglial cells, as opposed to reflecting causal mechanisms contributing to neurodegeneration \cite{Espay2019}. However, in the case of TDP-43, given its widespread prominence in many neurodegenerative conditions, and the promising effects of modulating the protein in many preclinical and clinical trials to date, as will be focus of this chapter, it appears that this protein plays more of a causal role in dementias and cognitive diseases.
The normal biological role of the transactive response DNA binding protein, TDP-43
The transactive response DNA binding protein, TARDBP gene, encodes the TAR DNA-binding protein of 43 kDa (TDP-43), a ubiquitous nuclear protein that binds to both DNA and RNA \cite{Mitra_2018}. TDP-43 regulates many aspects of RNA processing, pre-mRNA splicing \cite{Jiang_2017} and DNA damage repair \cite{Mitra_2018}. However, in addition to mRNA alternative splicing, TDP-43 is a critical protein with pleiotropic DNA and RNA-processing activities, including transcriptional repression, regulation of non-coding RNA, miRNA biogenesis, RNA stability, nucleocytoplasmic RNA trafficking, and auto-regulation of its own protein production \cite{Mejzini_2019}\cite{Nagano2020}. TDP-43 is transcribed early during embryonic development, and is ubiquitously expressed in a variety of tissues. In addition to its role in a wide range of DNA and RNA-regulatory processes, TDP-43 also has important functions in neurite outgrowth. Neurite projections emanate from the neuron cell bodies, such as axons and dendrites, and TDP-43's role is to transport ribosomal RNA from the nuclei to axons \cite{Nagano2020}. In addition, TDP-43 has a role in maintenance of genomic integrity and is a component of the non-homologous end-joining DNA damage-response pathway, which is activated in response to DNA double-stranded breaks \cite{Mitra_2019}.
Structure of TDP-43
Structurally, TDP-43 is a 414 amino acid protein consisting of four domains: two RNA recognition motifs (RRM1 and RRM2), a C-terminal domain (CTD) and N-terminal domain (NTD). The dynamic structure of TDP-43 NTD has been determined to be solenoid in shape and there is a physical separation between the C and N terminal domains. The CTD of has an intrinsically disordered region (IDR), with a glycine-rich and prion-like region involved in phase-separation and stress granule formation \cite{Colombrita_2009}, and it is in the CTD that is involved in forming cytoplasmic aggregates, and TDP-43's ability to form its correct shape and, interestingly, nearly half of all ALS associated TDP-43 mutations occur here \cite{Pesiridis2009}\cite{Wright_2020}. These structures are important for the normal physiological function of the protein and have roles in disease-associated accumulation \cite{Mitra_2018}\cite{Afroz_2017}.
TDP-43 is a highly conserved nuclear phase-separating protein and exhibits regions of low complexity and intrinsically disordered domains. Phase separation occurs when a homogenous solution of molecules forms a membraneless phase-separated compartment. It is in the C-terminal region of the protein that the IDR and a prion-like motifs are located, which are sequences that have been implicated in the aggregate proteinopathy \cite{Yang_2010}. TDP-43 also has a nuclear localization signal (NLS). Both the NTD and NLS are important regions for the nuclear localization of the protein as deletion of these regions results in its mislocalization, whereby the protein translocates from the nucleus, where it normally is present, to various cytoplasmic regions. In the cytoplasm inclusions are formed. Importantly, using algorithms to predict the ability of peptide sequences to self-aggregate, domains of TDP-43 were analyzed and the highest proclivity to self-aggregate and misfold was found to be the RRM2 and NTD peptide sequences, which should prove very important in efforts to screen for peptide inhibitors of TDP-43 based upon the interaction with these domains \cite{Kumar2019}.
The role of TDP-43 in ALS and other groups of neurodegenerative diseases
TDP-43 has emerged as a critical player in neurodegenerative disease. One such disorder is amyotrophic lateral sclerosis (ALS), a rapidly progressive and fatal group of motor neuron diseases that leads to upper- and lower motor neuron (UMN and LMN, respectively) degeneration and cell death, resulting in loss of the brain's communication with muscles. Ultimately, patients are unable to move voluntarily and, eventually, this leads to the loss of the ability to breath, which accounts for ALS's high fatality rate of between 2-5 years after diagnosis. ALS occurs most often in late middle life. Classic ALS leads to the progressive death of motor neurons in the cerebral motor cortex and spinal cord, and the gradual loss of respiratory ability, muscle paralysis, and atrophy. Treatment consists of supportive care for respiratory difficulties, symptom relief and nutritional needs. There are currently only two FDA-approved medications for ALS, Riluzole, an anti-glutamatergic compound and Edaravone, an antioxidant, both of which only provide modest benefit in a subset of patients\cite{Jaiswal_2018}. ALS often culminates in death within 2-5 years of clinical presentation, although there exist variants that have a slower rate of progression and better prognosis. Approximately 5-10 % of patients with ALS have a familial history (fALS), while the majority of cases, 90-95%, emerge sporadically (sALS), in patients without a family history as a result of indeterminate genetic and environmental etiological causes. Uncovering the genetic variants associated with sALS may be hampered by the methodological approaches employed. For instance, most genetic studies have relied upon short-read and whole exome sequencing platforms, approached more conducive to detecting diseases attributed to one or more single nucleotide polymorphisms. This misses larger structural variants such as deletions, inversions or tandem repeats, which appear likely to play a fundamental role in ALS pathology.
In ALS, TDP-43 forms aberrant aggregates in both neurons and glia \cite{Fernandopulle_2019}. Within the CTD is a glycine rich region that is involved in protein-protein interactions and the binding of TDP-43 to heterogeneous nuclear ribonucleoproteins (hnRNPs). It is in this region that the majority of ALS mutations occur. The TDP-43 NTD was found to form a homodimer in solution and as a homodimer the formation of cytoplasmic aggregates is prevented and the mRNA splicing functions of TDP-43 in the cell nucleus is enhanced \cite{Jiang_2017}. During cell stress, TDP-43 relocates to cytoplasmic stress granules. Although TDP-43 is normally present in the nucleus, during pathology it is cleaved and this cleaved form of TDP-43 enters the cell cytoplasm. Cell stress granules form in the cytoplasm and are composed of mRNA and RNA binding proteins following a stress to the cell, such as in the presence of toxins or high temperature. In normal physiology, cell stress granules are dynamic structures that can be assembled and disassembled when the cell stress trigger subsides. In ALS stress granules are persistent in motor neurons and this culminates in the degeneration of the motor neurons.
In addition to ALS, TDP-43 is also implicated in progressive muscular atrophy (PMA), which affects primarily LMN and, in some patients, also UMN; Primary lateral sclerosis (PLS) is a slowly progressive motor neuron disease that affects upper neurons, and is more readily managed than ALS and also may not decrease overall lifespan; Frontotemporal dementia (FTD), the second most common form of early-onset dementia, following AD, alters personality, behavior, language and mental functions, and has a life expectancy of 6-8 years with the first onset of symptoms. Other neuropathology diseases with TDP-43 involvement include Perry Syndrome, PD, LD, Huntington’s disease, chronic traumatic encephalopathy and limbic predominant age-related TDP-43 encephalopathy (LATE). Based on the foregoing, TDP-43 has become a major focus for therapeutic interventions.
Mutations and epigenetic variations of TDP-43 and other ALS-linked genes in familial and sporadic neurodegenerative diseases
There are at least 60 known mutations in the TDP-43 gene
which have been identified in ALS. Certain mutations in this gene are
responsible for the development of several other neurodegenerative
disorders. These include primary lateral sclerosis, progressive muscular
atrophy and FTD. Yet, TDP-43 mutations only account
for an estimated 5-10 % of familial ALS (fALS) patients \cite{Prasad_2019}, while the remainder of fALS mutations occur in other ALS-linked
genes \cite{Hergesheimer_2019}. The most well-characterized of ALS-linked genes are the superoxide dismutase
(SOD1), chromosome 9 open reading frame 72 (C9ORF72) , and
fused in sarcoma (FUS) genes. For fALS-associated genes, all are inherited in a autosomal dominant fashion.
TDP-43 gain-of-function: sequestering of TDP-43 in cytoplasm and formation of aggregates
In ALS there is a perturbation in TDP-43 trafficking between the nucleus and cytoplasm. The TDP-43 protein is predominantly localized to the nucleus under normal endogenous conditions. However, sequestration of the protein in the cytoplasm, results in loss of endogenous TDP-43 function and to the accumulation of insoluble cytoplasmic aggregates \cite{Barmada_2010}\cite{Suk_2020} . Evidence suggests that the NTD of TDP-43 proteins, a region which contains its nuclear localization signal, homodimerizes with its protein partners and appears important in its function of targeting splicing of RNA. In fact, mutation of the nuclear localization or nuclear export signals results in cytoplasmic or nuclear aggregate formation \cite{Winton_2008}, whereas exogenous accumulation of cytoplasmic TDP-43 has been demonstrated to be specifically cytotoxic in primary rat cortical neurons \cite{Hergesheimer_2019} . TDP-43 has also been reported to be sequestered in the cytoplasm by co-aggregation with other proteins, such as poly(GR) \cite{Nagano2020}. Therefore, homeostatic auto-regulation of TDP-43 is critical for its normal function. Normally, TDP-43 binds to the 3’ UTR of its own pre-mRNA, which leads to its undergoing of nonsense-mediated mRNA decay \cite{Ayala_2010}, thereby decreasing the nuclear to cytoplasmic shuttling of the transcript as well as its corresponding translation in the functional protein \cite{Koyama_2016}. Loss of homeostatic nucleo-cytoplasmic localization resulting in either nuclear or cytoplasmic TDP-43 aggregates appears critical is the pathology of all variants of ALS \cite{Jovi_i__2016} . The TDP-43 inclusions are composed of the aberrant posttranslational modifications of the protein, such as these with their effects ubiquitinated (solubility), phosphorylated (mislocalization and aggregation), acetylation (activity in mitochondria and RNA-binding), PARylation (phase-separation and stress granules) and cysteine oxidation (self dimers) as well as the C-terminus fragments produced by proteolytic enzymes, pathogenic form that may be targeted for removal by autophagosomes \cite{Prasad_2019}.
TDP-43 loss-of-function: loss of normal DNA and RNA processing functions
TDP-43 is an important protein that is normally enriched in the nucleus, where it binds to both DNA and RNA. Some of the normal functions of the protein include the regulation of transcription, mRNA splicing, RNA stability, transport and translation. The attenuation of these critical cell processes is implicated in a variety of neurological disorders, and it has been argued \cite{Vanden2014} that the loss of normally functioning TDP-43 is of primary cause of neuropathology, while that toxic formation of aggregates is a secondary pathological feature, arising from protein conformational changes. The abnormal tertiary structural changes form oligomers of repeating units that result in higher order structures, which accumulate and cluster in specific brain regions and in motor neurons.
TDP-43 represents a promising therapeutic target
TDP-43 remains a
promising therapeutic target even when a patient does not have a
mutation event in the TARDBP gene. For example, the vast majority (~95%) of ALS patients (sporadic and familial), exhibit TDP-43 neuronal
inclusions in their cortical and spinal cord neurons although a mere 5-10% have TDP-43 mutations. TDP-43 was found
to be a major constituent of ubiquitin-positive inclusions in ALS
patients \cite{Arai_2006}, leading to the recognition of this
protein aggregate as a hallmark of ALS \cite{Wolozin_2019}\cite{Neumann_2006}.
Introduction to therapeutic modulators of TDP-43
Therapeutic options remain very limited for the many neurodegenerative diseases, especially the motor neuron disease ALS, a fatal neurodegenerative disorder that occurs most often in late middle life. Classic ALS leads to progressive death of motor neurons in the cerebral motor cortex and spinal cord, and the gradual loss of respiratory ability, muscle paralysis and atrophy. Treatment consists of supportive care for respiratory difficulties, symptom relief and nutritional support. ALS often culminates in death within 2-5 years of clinical presentation, although there exist variants that have a slower progression and better prognosis. There are currently only two FDA-approved medications for ALS, riluzole, an anti-glutamatergic compound and edaravone, an antioxidant, which only provide modest benefit in some patients \cite{Jaiswal_2018}.
New therapeutic options are therefore very urgently needed for the ALS spectrum of disorders as well as for the many other devastating neurological conditions, many of which exhibit TDP-43 pathology.
Pre-clinical and clinical studies based on pharmacological modulation of TDP-43
On June 20, 2020, we conducted a search on Clinical Trials.gov (
https://clinicaltrials.gov) for clinical trials for patients with ALS based upon the modulation of TDP-43 ALS. We searched under the
Other terms field for "TDP-43", which yielded 692 Studies. We then applied the following filters to limit studies to those that were further along in clinical development stages, to phase 2, 3, and 4 trials. Limiting studies to the terms TDP-43, Active, not recruiting, Completed, Unknown status Studies, Studies With Results and Phase 2, 3 or 4 yielded 39 studies. We will describe the results of those studies. In addition, a Pubmed Central Search of TDP-43 with the filters Clinical Trial, Meta-Analysis and Randomized Controlled Trial yielded 19 results.
Web of Science databases
Tamoxifen Treatment in Patients with Motor Neuron Disease
Tamoxifen has been reported to decrease TDP-43 aggregates in animal models of ALS via tamoxifen-induced autophagy \cite{Wang_2013}. In a small clinical trial, eighteen ALS patients without mutations in superoxide dismutase-1 (SOD-1) or fused in sarcoma (FUS) gene, were randomized to receive either 40 mg/day tamoxifen or placebo, with all participants receiving riluzole twice daily (100 mg/day total) for a period of 1, 3 or 12 months (ClinicalTrials.gov number NCT02166944) \cite{Chen_2020}. The primary end-points of time to death or need for mechanical ventilation, was not statistically different between the tamoxifen vs placebo group at any of the time points. However, at a 1, 3 and 6 month evaluation, one of the two secondary end points, ALSFRS-R score, but not forced vital capacity (FVC), decreased less in the tamoxifen group, while there were no statistically significant differences of any end point observed at 12 months. The authors conclude that larger studies would be required to confirm the modest, although transient, improvement in ALSFRS-R scores in the tamoxifen-treated group and to determine whether enhanced autophagy to decrease TDP-43 results in clinical improvement of ALS, particularly if administered at the earliest stages of the disease.
Perampanel (FYCOMPA), a selective non-competitive AMPA receptor antagonist
Downreguation of adenosine deaminase acting on RNA 2 (ADAR2) occurs in sporadic ALS (sALS) \cite{Kawahara_2004} . ADAR2 is an RNA adenosine-to-inosine (A-to-I) mRNA editing enzyme. Modifying mRNA adenosine to inosine (translated into guanosine), results in new protein isoforms, and it is a dynamic process which diversifies the proteome \cite{Rosenthal2012}. The influx of Ca2+ through α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors is regulated by ADAR2 A-to-I editing. However, in sALS, ADAR2 is deficient and the majority of patients with sALS have a GluA2 subunit of the glutamate AMPA receptor that is unedited \cite{Kawahara_2004}. The outcome of this loss of glutamate AMPA editing is that Ca2+ permeability is elevated, and this has been associated with TDP-43 pathology and glutamate-induced excitotoxicity in motor neurons. Perampanel is a non-competitive, selective antagonist of the AMPA receptor and is approved in the treatment of seizures. In mice in which the AMPA receptor is knocked out in motor neurons to model sALS, perampanel greatly attenuates neuropathology, decreased motor neuron death and prevented the TDP-43 mislocalization from the nucleus to the cytoplasm \cite{Akamatsu2016}. Currently, there is an ongoing phase 2 clinical trial of perampanel, Perampanel for Sporadic Amyotrophic Lateral Sclerosis (ALS): A Multicenter, Randomized, Double-blind, Placebo-controlled, Parallel-group Phase 2 Trials, (ClinicalTrials.gov Identifier NCT0301941) consisting of placebo, 4mg and 8mg per day for 48 weeks and the primary endpoint of change in ALS functional rating scale. The results are currently unavailable, but were projected to be available early in 2020 \cite{Aizawa2019}.