Ongoing research in Dr. Petrucelli's Neurodegeneative Diseases Lab centers around studying cellular mechanisms that cause abnormal protein aggregation.

Assay development for the measurement of pathological TDP-43

The refinement of biomarkers is a key strategy for elucidating the pathogenesis of neurodegenerative diseases and for developing sensitive measures of disease activity and progression. In addition to their diagnostic value, biomarker assays provide a reliable and sensitive means to test drug efficacy and, ultimately, help determine the best course of action for patients once treatments become available.

We hypothesize that the presence of TDP-43-positive inclusions in ALS, FTLD-U and other TDP-43 proteinopathies such as Alzheimer's disease, suggests that TDP-43 levels in cerebrospinal fluid (CSF), and potentially in plasma or serum, may parallel TDP-43 brain pathology. Indeed, TDP-43 can be detected in human plasma [1, 2] and CSF [3, 4] and levels are reportedly elevated in cases of ALS, FTLD and AD [1-4], though wide variations among individuals are observed. Nonetheless, we believe that the sensitivity and usefulness of TDP-43 as a biomarker may be improved by developing Enzyme-Linked Immunosorbent Assays (ELISAs) that detect pathological forms of TDP-43, such as phosphorylated TDP-43 and C-terminal TDP-43 fragments.

At present, data on longitudinal levels of TDP-43 in biological fluids of ALS or FTLD-U patients is not available. Thus, we aim to develop sensitive ELISAs for the measurement of phosphorylated and truncated TDP-43 to determine if pathologically modified TDP-43 in CSF, plasma or serum is a suitable surrogate marker of disease activity and progression.

Cell and animal models of TDP-43 proteinopathies

TDP-43 is a principal component of ubiquitin-positive inclusions in amyotrophic lateral sclerosis (ALS) and frontotemporal dementia with ubiquitin-positive inclusions (FTLD-U). TDP-43 pathology is also observed, to varying degrees, in other neurodegenerative disorders, including Alzheimer's disease, hippocampal sclerosis, Lewy body disease, parkinsonism-dementia complex of Guam, corticobasal degeneration, Pick's disease and Perry syndrome. In order to elucidate both the normal functions of TDP-43 as well as how TDP-43 dysfunction causes neuronal death, we have developed cell culture, Caenorhabditis elegans (C. elegans) and mouse models of TDP-43 proteinopathies.

  • Modeling TDP-43 proteinopathies in cells. We and others have shown that TDP-43 is cleaved by caspases to produce fragments of approximately 35 and 25 kDa. We have also shown that the overexpression and aggregation of the 25 kDa C-terminal fragment generated by caspase cleavage of TDP-43 (GFP-TDP) is detrimental to neuroblastoma cells [5]. The aggregation of GFP-TDP is associated with increased cytotoxicity which likely results from a toxic gain of function since aggregate formation neither alters the nuclear distribution or function of endogenous full-length TDP-43. We thus believe that compounds that decrease the aggregation of TDP-43 will provide neuroprotection to patients suffering from TDP-43 proteinopathies. To identify such compounds, we have generated a human neuroblastoma cell line (M17D3 cells) that inducibly expresses GFP-TDPin the absence of doxycycline (Dox). Removal of Dox from the culture media leads to the time-dependent expression of GFP-TDP and the formation of cytoplasmic GFP-TDP inclusions. We are currently using this model to screen a library of ~58,000 small-molecules with increased probability of oral bioavailability and blood brain barrier penetration to identify compounds that diminish both TDP-43 aggregation and toxicity. Compounds of interest will be validated using:
    1. In vitro aggregation models of recombinant TDP-43;
    2. Secondary cell model systems (for example, primary neurons expressing aggregation-prone TDP-43 products);
    3. Our transgenic mouse models of TDP-43 proteinopathies (see below)
  • Modeling TDP-43 proteinopathies in the nematode C. elegans. To model function and neurotoxicity of TDP-43 in vivo, we have engineered C. elegans with neuronal expression of either wild-type or mutant human TDP-43. Transgenic worms with neuronal expression of human TDP-43 exhibit "uncoordinated" movement and have abnormal motor neuron synapses. Just as overexpression of human TDP-43 results in the worms having an uncoordinated phenotype, so does the neuronal overexpression of TDP-1, the C.elegans ortholog of TDP-43. By using the uncoordinated phenotype as an indicator of TDP-43-induced neurotoxicity, we have investigated the contribution of specific TDP-43 domains and its subcellular localization in promoting this phenotype. Furthermore, we have shown that the orthologous C. elegans TDP-1 is functionally conserved in cell culture-based splicing assays of a CFTR minigene. By investigating the conserved functions between TDP-43 and TDP-1, we intend to determine the roles of TDP-43 in RNA homeostasis in neurons under basal or stressful conditions.
  • Mouse models of TDP-43 proteinopathies. We are in the process of generating and characterizing transgenic mouse models of TDP-43 proteinopathies to allow us to determine the behavioral, biochemical, and neuropathological impact of wild-type and mutant TDP-43 expression. Given that the majority of TDP-43 proteinopathies are not associated with mutations in the gene encoding TDP-43, it is essential to develop model systems that can be used to elucidate the normal function of wild-type TDP-43 in the central nervous system and to determine if wild-type TDP-43 can directly cause neurodegeneration. Therefore, we have generated transgenic (TDP-43PrP) mice expressing full-length human TDP-43 (hTDP-43) driven by the mouse prion protein (Prp) promoter. To understand how ALS-specific mutations contribute to disease, we are also generating a conditional mouse model expressing M337V TDP-43, as well as a control line conditionally expressing wild-type TDP-43. The ability to control the expression TDP-43 will allow us to determine if TDP-43 pathology can be prevented, halted or reversed by suppression of TDP-43 expression. Overall, these transgenic TDP-43 mice will be valuable tools in understanding the normal role of TDP-43 and will provide an essential resource to dissect the pathogenic mechanisms between wild-type and mutant TDP-43.

Identification of pharmacological chaperones for the treatment of hereditary protein conformational disorders

A number of genetic disorders are caused by inheritance of mutations that alter the conformation of proteins translated from a mutated gene. These disorders are collectively known as protein conformational disorders (PCDs). PCDs range from congenital conditions such as alpha-1 antitrypsin (AAT) deficiency and the early-onset juvenile form of primary open angle glaucoma (POAG), to neurodegenerative diseases such as Huntington's disease. One common pathological hallmark among PCDs is the accumulation of aggregated proteins caused by mutation-induced changes in protein conformation. In many cases, these aggregated proteins are toxic to the host cells due to mechanisms such as overloading of protein degradation machinery and the subsequent apoptotic response. In addition, the sequestration of proteins into aggregates inevitably depletes the supply of functional protein, causing a loss-of-function phenotype.

Using the myocilin related POAG and AAT deficiency as disease models, our laboratory is endeavoring to identify and explore the potential use of pharmacological chaperones to correct abnormal protein folding in an effort to prevent protein aggregate formation, alleviate cytoxicity and restore normal protein function. To detect specific protein-chemical binding and to perform high-throughput screening of chemical libraries, we have applied the Corning Epic system which is a new, unique platform designed to sensitively measure surface plasmon resonance changes during micro-binding events. From one of our FDA approved drug libraries, we have identified potential leads that interact with recombinant myocilin and, more importantly, reduce aggregation and increase secretion of disease-related myocilin mutants in cultured cell models. Currently, we are validating these lead compounds in an in vivo model of POAG caused by mutant myocilin. The identification of pharmacological chaperones for myocilin not only will increase our understanding on the disease mechanism, but also potentially reveal a new class of therapeutics for myocilin-related POAG and other PCDs.

Progranulin, TDP-43 and toxicity

TDP-43 is a principal component of ubiquitin-positive inclusions in amyotrophic lateral sclerosis (ALS) and frontotemporal dementia with ubiquitin-positive inclusions (FTLD-U). Under normal conditions, TDP-43 is a nuclear protein but neurons with cytoplasmic inclusions have a substantial loss of nuclear TDP-43. TDP-43 also exhibits a disease-specific biochemical signature; pathologically altered TDP-43 is hyperphosphorylated and cleaved to generate C-terminal fragments of 24-26 kDa in affected brain and spinal cord regions. The molecular basis for TDP-43 proteolysis and the generation of these C-terminal fragments remains largely unknown, although it is reasonable to believe that many TDP-43 truncation products are generated in TDP-43 proteinopathies. We have shown that the in vitro incubation of recombinant TDP-43 with caspase-3 or caspase-7 produces distinct fragments of ~42, 35 and 25 kDa [1]. Furthermore, the activation of caspases in cells by staurosporine treatment leads to the redistribution of TDP-43 from the nucleus to the cytoplasm [1]. Of interest, null mutations in the gene encoding progranulin (PGRN) are a cause of FTLD-U with TDP-43-positive inclusions [2-5]. When we modeled progranulin haploinsufficiency in cultured cells using progranulin siRNA, activated caspase-3 levels were increased as was the cleavage of TDP-43 into ~25 and ~35 kDa fragments similar in molecular weight to the fragments observed in FTLD-U brain tissue [1]. These findings provide a mechanistic link between decreased progranulin expression and abnormal TDP-43 processing which may contribute to TDP-43-mediated toxicity. Indeed, we have shown that, upon overexpression of the 25 kDa C-terminal TDP-43 caspase-cleavage product in mammalian cells, cytoplasmic TDP-43 aggregates are formed that are both ubiquitinated and phosphorylated and which confer cytotoxicity [6]. Currently, we are undertaking a multi-tiered approach to further investigate the involvement of abnormal progranulin and TDP-43 in neurotoxicity.

  • Progranulin mutations and neurotoxicity. As mentioned, null mutations in PGRN lead to reduced levels of PGRN and cause FTLD-U with TDP-43 pathology. Missense mutations in PGRN of unknown pathogenicity are also observed in cases of FTLD-U and ALS [2, 3, 7, 8]. PGRN has wide-ranging functions in the periphery as well as in the central nervous system where its expression is limited to microglia and certain neuronal populations. In the adult brain and spinal cord, PGRN may function in neuronal repair and growth. The normal function of PGRN, however, is complex; on the one hand, full-length PGRN has anti-inflammatory and trophic activity but, on the other hand, the proteolytic cleavage products of PGRN, termed granulins, promote inflammatory activity [9]. At present, the mechanisms by which PGRN haploinsufficiency leads to neurodegeneration, and how it relates to TDP-43-mediated toxicity, remains unclear. To gain further insight on this issue, we are using PGRN knockout mice and cell culture techniques. We are also examining the effect of missense PGRN mutations to determine if they, like null mutations, are pathogenic.

Targeting tau as a treatment for tauopathies

The accumulation of hyperphosphorylated tau causes the formation of neurofibrillary tangles, a pathological hallmark of tauopathies, a group of diseases which includes Alzheimer's disease, frontal temporal dementia with Parkinsonism linked to chromosome 17 (FTDP-17), progressive supranuclear palsy, Pick's disease and corticobasal degeneration. Therapeutics aimed at eliminating hyperphosphorylated tau are thus of considerable interest. One promising approach to promote the clearance of hyperphosphorylated tau is via the manipulation of molecular chaperones, like Hsp90. Molecular chaperones function to refold misfolded proteins or to target them for proteasomal degradation.

  • Degradation of abnormal tau by Hsp90 inhibitors. We have shown that hyperphosphorylated tau is a client of the heat shock protein, Hsp90 [1, 2]. Hsp90 regulates the folding or degradation of its client proteins by forming multi-component complexes with other chaperones. The binding of nucleotides to Hsp90 alters its conformation and defines the subset of chaperones with which it interacts. When bound by ADP, Hsp90 associates with Hsp70/Hsp40 complexes bound to client proteins. This complex recruits ubiquitin ligases, like CHIP, to direct the client protein to proteasomes for degradation. The replacement of ADP with ATP alters Hsp90 conformation, releasing Hsp70/Hsp40 and allowing the recruitment of another set of cochaperones, including p23 and cdc37. This complex folds and stabilizes the client protein now bound by Hsp90. Hsp90 can thus cycle between different multi-chaperone complexes: one complex favoring the refolding of abnormal proteins and the other targeting proteins for degradation. Hsp90 inhibitors (HSP90i) convert the Hsp90 complex from a catalyst for protein folding into one that induces protein degradation. Thus, HSP90i have two functional consequences: 1) the degradation of Hsp90 and its bound client protein; and 2) the activation of heat shock factor 1 (HSF1), which is normally suppressed by Hsp90. HSF1 activation results in the transcription of stress-induced chaperones, like Hsp70.

    Our findings indicated that HSP90i cause the preferential degradation of hyperphosphorylated tau in cultured cells and mice [1, 2], thereby highlighting the therapeutic potential of HSP90i for Alzheimer's disease and other tauopathies. Our studies are now geared towards determining if HSP90i, by enhancing the degradation of abnormally phosphorylated tau, provide protection against tau-induced neuronal loss and behavior deficits in rTg4510 mice, a model of 4R tauopathy. It is noteworthy that we have found that Hsp90 in affected areas of Alzheimer's disease brain is markedly more sensitive to Hsp90 inhibition than Hsp90 in unaffected areas, suggesting that HSP90i possess an exploitable therapeutic index [1]. Currently, we are studying whether pathological tau underlies the heightened sensitivity of Hsp90 for HSP90i. Given that the binding affinity of Hsp90 for HSP90i is increased when Hsp90 is acetylated [3, 4], and given that tau has been reported to inhibit histone deacetylase 6 (HDAC6) [5], a cytosolic Hsp90 deacetylase, tau could indirectly increase the affinity of Hsp90 for HSP90i. Therefore, we are studying the interplay between tau and HDAC6 and whether HDAC6 regulates Hsp90-dependent degradation of tau.

  • Targeting heat shock protein complexes for the treatment of tauopathies. The fact that there exists cellular machinery capable of recognizing abnormally modified tau provides a powerful means for targeting only the tau that is harmful to neuronal health. That heat shock proteins recognize abnormal tau is particularly impressive, but heat shock proteins are also actively involved in antigen presentation through major histocompatibility complex molecules leading to a robust immune response. In fact, complexes formed between heat shock proteins and abnormal peptides that have been purified from affected individuals have a proven track record for providing immunity against a variety of viruses and cancers; most critically, the abnormal peptides do not have to be identified for this type of immunization. This, coupled with the finding that active immunization against phosphorylated tau peptides inhibits tau aggregation and slows the behavioral phenotype in at least one model of human 4R tauopathy [6], provides substantial impetus for these techniques to be combined in order to determine if heat shock protein-peptide complexes purified from rTg4510 mice are able to prevent or inhibit tau-induced neurodegeneration. Should this approach prove feasible, it could lead to a novel therapy for Alzheimer's disease and other tauopathies, as well as provide a means of treating the large variety of neurodegenerative diseases characterized by abnormal protein folding. Finally, the binding of heat shock proteins to abnormal tau species could expedite the identification of harmful tau products which, in turn, would help increase our understanding of the detrimental consequences of aberrant tau products as well as how such products are generated. As our knowledge regarding tau-induced toxicity expands, so too will the likelihood of developing effective therapies for tauopathies.