The issue of whether overexpression of wild-type TDP-43 in rodent brain could be neurotoxic is neatly brought out in this paper from the group of Ronald Klein with senior authors Dennis Dickson and Mike Hutton. Using an adeno-associated virus type 9 (AAV9) vector for human TDP-43 expression by stereotactic injection into the rat substantia nigra (SN), Tatom and colleagues show that overexpression of human wild-type TDP-43 on its own can kill dopaminergic neurons in rats in a dose-dependent manner (Tatom et al., 2009).

This approach is surely welcome at a time when many laboratories are struggling to get a decent TDP-43 expression in transgenic germlines. The reason why this is problematic is made apparent by this paper, where (roughly estimated) threefold wild-type TDP-43 overexpression almost completely wipes out the targeted neurons accompanied by neurodegeneration-related astro- and microgliosis. TDP-43 was selectively expressed in neurons as AAV9 has a natural neurotropism, perhaps due to the virus capsid (Bartlett et al., 1998); and while SN is chosen for sake of...  Read more

The issue of whether overexpression of wild-type TDP-43 in rodent brain could be neurotoxic is neatly brought out in this paper from the group of Ronald Klein with senior authors Dennis Dickson and Mike Hutton. Using an adeno-associated virus type 9 (AAV9) vector for human TDP-43 expression by stereotactic injection into the rat substantia nigra (SN), Tatom and colleagues show that overexpression of human wild-type TDP-43 on its own can kill dopaminergic neurons in rats in a dose-dependent manner (Tatom et al., 2009).

This approach is surely welcome at a time when many laboratories are struggling to get a decent TDP-43 expression in transgenic germlines. The reason why this is problematic is made apparent by this paper, where (roughly estimated) threefold wild-type TDP-43 overexpression almost completely wipes out the targeted neurons accompanied by neurodegeneration-related astro- and microgliosis. TDP-43 was selectively expressed in neurons as AAV9 has a natural neurotropism, perhaps due to the virus capsid (Bartlett et al., 1998); and while SN is chosen for sake of convenience allowing a rapid estimation of neuronal loss and behavior deficit, it is not without relevance. TDP-43 is naturally expressed in this brain region, and TDP-43 pathology is observed in the nigrostriatal pathway in a variety of neurodegenerative diseases.

The descriptive neuropathology is fascinating to read. The transgenic protein predominantly homes to the neuronal nuclei, but in approximately 1 percent of the neurons, diffuse cytoplasmic TDP-43 accumulations are also observed. Occasionally, granular textures are observed that are indicative of pre-inclusion lesions as expected at 4 weeks of disease duration. One of the pathological hallmarks of diseased neurons in ALS and FTLD-TDP patients (FTLD-U in the old terminology) is redistribution of TDP-43 from its normal nuclear localization to the cytoplasm, where it is phosphorylated and ubiquitinated and forms insoluble aggregates.

Interestingly, ubiquitin labeling was also observed in neuronal cytoplasm. It is not clear what proportion of neurons were labeled or whether it co-localized with TDP-43, but given that it was not observed in the contralateral, uninjected side, it is very likely that it is the transgenic TDP-43 protein that is ubiquitinated. It will be interesting to study whether TDP-43 is phosphorylated and/or cleaved into the disease-characteristic ~25-kDa C-terminal fragments. This would further strengthen the justification of trying to develop a rodent model of TDP-43 proteinopathy.

At present the mechanism for TDP-43 dose-related neurodegeneration is unclear. Even so, considering the various important physiological functions of TDP-43 and a tight control on its expression levels, these data are not surprising. Of relevance to the majority of ALS and FTLD-TDP patients is that it is the wild-type TDP-43 protein that is accumulating and causing neurotoxicity. Given the speed at which these studies could be accomplished, it would be very interesting to study whether a similar overexpression of TDP-43 C-terminus recapitulates key features of TDP-43 proteinopathy in rat brain, as has been shown recently in cell cultures (Igaz et al., 2009).

Lastly, having an AAV-based TDP-43 rodent model is not a substitute for a rodent model with stable germline transmission achieved by either a constitutive or inducible expression system. That amongst all would allow us to understand which neurons are more vulnerable to TDP-43 gene dosage. But until these models are developed, expression of TDP-43 by AAV-mediated somatic cell transfer approaches will continue to shed light on TDP-43 mediated disease mechanisms.

References:
Bartlett JS, Samulski RJ, McCown TJ. Selective and rapid uptake of adeno-associated virus type 2 in brain. Hum Gene Ther. 1998 May 20;9(8):1181-6. Abstract

Igaz LM, Kwong LK, Chen-Plotkin A, Winton MJ, Unger TL, Xu Y, Neumann M, Trojanowski JQ, Lee VM. Expression Of TDP-43 C-terminal fragments in vitro recapitulates pathological features of TDP-43 proteinopathies. J Biol Chem. 2009 Jan 21; Abstract

Tatom JB, Wang DB, Dayton RD, Skalli O, Hutton ML, Dickson DW, Klein RL. Mimicking Aspects of Frontotemporal Lobar Degeneration and Lou Gehrig's Disease in Rats via TDP-43 Overexpression. Mol Ther. 2009 Feb 17; Abstract

View all comments by Samir Kumar-Singh

From a clinical perspective, the identification of TDP-43 protein represents a major breakthrough in our understanding of both frontotemporal lobar degeneration (FTLD) and amyotrophic lateral sclerosis (ALS). The TDP-43 is the mystery protein that is associated with the ubiquitin-positive inclusions that are commonly found in many patients with FTLD and in most, if not all, patients with ALS.

This finding is particularly important because several recent papers suggest that patients who have FTLD with ubiquitin inclusions at autopsy (FTLD-U) account for approximately 50 percent of all autopsy-confirmed FTLD cases (1-3). The remaining majority of FTLD cases are associated with the tau protein, but other neuropathological diagnoses exist. The finding that possibly one-half of all FTLD patients may have ubiquitin-positive neuropathology means that any breakthroughs in the biology of this protein could potentially translate into helping a large proportion of FTLD patients.

In addition, the finding that the TDP-43 protein is also found in patients with ALS further supports...  Read more

From a clinical perspective, the identification of TDP-43 protein represents a major breakthrough in our understanding of both frontotemporal lobar degeneration (FTLD) and amyotrophic lateral sclerosis (ALS). The TDP-43 is the mystery protein that is associated with the ubiquitin-positive inclusions that are commonly found in many patients with FTLD and in most, if not all, patients with ALS.

This finding is particularly important because several recent papers suggest that patients who have FTLD with ubiquitin inclusions at autopsy (FTLD-U) account for approximately 50 percent of all autopsy-confirmed FTLD cases (1-3). The remaining majority of FTLD cases are associated with the tau protein, but other neuropathological diagnoses exist. The finding that possibly one-half of all FTLD patients may have ubiquitin-positive neuropathology means that any breakthroughs in the biology of this protein could potentially translate into helping a large proportion of FTLD patients.

In addition, the finding that the TDP-43 protein is also found in patients with ALS further supports the overlap between FTLD and ALS. Future research on the TDP-43 protein will likely also benefit ALS patients and help us understand how these two very different clinical phenotypes are related.

References:
1. Lipton AM, White CL 3rd, Begio EH. Frontotemporal lobar degeneration with motor neuron disease-type inclusions predominates in 76 cases of frontotemporal degeneration. Acta Neuropathol (Berl). 2004 Nov;108(5):379-85. Abstract

2. Johnson JK, Diehl J, Mendez MF, Neuhaus J, Shapira JS, Forman M, Chute DS, Roberson ED, Pace-Savitsky C, Neumann M, Chow TW, Rosen HJ, Forstl H, Kurz A, Miller BL.. Frontotemporal lobar degeneration: demographic characteristics of 353 patients. Archives of Neurology. 2005;62:925-930. Abstract

3. Forman MS, Farmer J, Johnson JK, Clark CM, Arnold SE, Coslett HB, Chatterjee A, Hurtig HI, Karlawish JH, Rosen HJ, Van Deerlin V, Lee V M-Y, Miller BL, Trojanowski JQ, & Grossman M. (2006). Frontotemporal dementia: Clinicopathological correlations. Annals of Neurology. 2006;59:952-962. Abstract

View all comments by Julene K. Johnson

In this paper, Drs. Lee and Trojanowski and colleagues have at long last identified the mystery protein hiding within the ubiquitinated inclusions that characterize certain histological forms of frontotemporal lobar degeneration (FTLD), termed FTLD-U. This task has challenged neuroscientists for well over a decade, with all prior attempts at identification using immunohistochemical or biochemical methods proving fruitless. The culprit protein is a TAR DNA-binding protein, known as TDP-43. This protein is present within all the ubiquitinated structures in FTLD-U, viz., the neuronal cytoplasmic inclusions, the neuronal intranuclear inclusions, and the neuritic changes, though whether this is the sole component of these structures (other than ubiquitin) remains uncertain. Some previous studies reported the presence of p62 protein within neuronal cytoplasmic inclusions, but such findings have been inconsistent. Moreover, Lee and Trojanowski have shown that the ubiquitinated neuronal cytoplasmic inclusions seen within spinal and cranial nerve nuclear motor neurons in motor neuron...  Read more

In this paper, Drs. Lee and Trojanowski and colleagues have at long last identified the mystery protein hiding within the ubiquitinated inclusions that characterize certain histological forms of frontotemporal lobar degeneration (FTLD), termed FTLD-U. This task has challenged neuroscientists for well over a decade, with all prior attempts at identification using immunohistochemical or biochemical methods proving fruitless. The culprit protein is a TAR DNA-binding protein, known as TDP-43. This protein is present within all the ubiquitinated structures in FTLD-U, viz., the neuronal cytoplasmic inclusions, the neuronal intranuclear inclusions, and the neuritic changes, though whether this is the sole component of these structures (other than ubiquitin) remains uncertain. Some previous studies reported the presence of p62 protein within neuronal cytoplasmic inclusions, but such findings have been inconsistent. Moreover, Lee and Trojanowski have shown that the ubiquitinated neuronal cytoplasmic inclusions seen within spinal and cranial nerve nuclear motor neurons in motor neuron disease (amyotrophic lateral sclerosis) also contain TDP-43.

This is an immensely important study with huge implications for neurobiology.

Firstly, it pinpoints a key biochemical constituent in the pathogenesis of FTLD-U and motor neuron disease (MND), and one which previous work would never have regarded as a likely candidate protein. Secondly, although an association between FTLD and MND had long been known on account of some cases showing defined clinical features of both disorders, sharing pathological features of both disorders, and families being known where some members had FTLD, others MND, and others the combined disorder, it was never clear whether this association was coincidental or causal. Now we can see causality, and the implication that FTLD and MND are part and parcel of the same disease spectrum will have major ramifications for understanding pathogenesis, and eventual treatment. Thirdly, the finding of TDP-43 pathological changes in FTLD patients with mutations in the newly identified progranulin (PGRN) gene, who typically show FTLD-U pathological changes, firmly brings together a causal relationship in these two fundamental proteins in driving the pathogenesis of the disorder, and opens up untapped vistas of neurobiological research.

Therefore, in rapid time, two major (protein) pieces in the jigsaw puzzle of FTLD have been identified. The challenge now will be to fit the pieces around these and eventually identify the linking processes that bring these together into the fuller picture. Nonetheless, it is clear that even within FTLD-U there are different histological and clinical phenotypes, and it will be necessary to dissect out biochemical or other factors that might determine where the TDP-43 pathological changes take place in the brain to produce the clinical phenotype. That is, why is it that in some patients the most common clinical manifestation of FTLD-U, frontotemporal dementia, is present in association with bilateral involvement of the frontal and temporal lobes, yet in others only the temporal lobes are affected—producing semantic dementia—and in others the left hemisphere is preferentially affected to give progressive non-fluent aphasia. Also, what determines whether TDP-43 changes will be in the brainstem and spinal cord to give MND, or in the cerebral cortex to give FTLD? Lastly, in all this flurry of excitement, it should not be forgotten that tauopathy is still a major cause of FTLD, and it is not immediately apparent how pathological changes in the expression or function of tau might link in with progranulin and TDP-43. Clearly, changes in all three molecules can produce the same disorder of FTLD either separately or collectively: it is not possible to unequivocally discriminate FTD patients with MAPT mutations from those with PGRN mutations, or others without mutations in either. Interrelationships within this Bermuda triangle of tau, progranulin, and TDP-43 will need to be addressed.

The identification of TDP-43 as a (major/sole) component of the ubiquitinated protein of FTLD and MND, in conjunction with the identification of mutations in PGRN, have opened up huge new fields within the neurobiology of neurodegenerative disease with tentacles that may stretch far wider than these two disorders themselves. Whether there is a role for either or both of these proteins in other disorders like Alzheimer disease and Parkinson disease remains to be seen. The gauntlet has been cast down—it is up to the neuroscience community to pick this up and address these issues. What is certain is that there will be a major change in the focus of neurobiological research as groups worldwide seek to investigate the implications of changes in proteins such as progranulin and TDP-43 in terms of health and disease. We can look forward within the near future to major advances in our understanding of how the brain works in respect of these molecules and why neurodegenerative disease occurs when they fail to function properly. Maybe even a treatment for neurodegenerative disease may come a little closer.

View all comments by David M.A. Mann

Neumann, Sampathu, Kwong, and colleagues have resolved a long-standing issue in the research field of frontotemporal lobar degeneration (FTLD) and amyotrophic lateral sclerosis (ALS). These authors have identified TDP-43 as a major component of ubiquitin-positive inclusions that characterize these disorders. They first extracted a fraction from the patients' brains using monoclonal antibodies and then analyzed it by mass spectrometry. Their findings have greatly facilitated the understanding of the molecular pathogenesis of FTLD and ALS.

Independently, we have also found TDP-43 as a component of the inclusions in FTLD [1]. Following electrophoresis of the sarkosyl-insoluble brain extracts from FTLD, Alzheimer disease (AD) and dementia with Lewy bodies (DLB), we have done exhaustive analyses by mass spectrometry. Following identification of each molecule that is more abundant in FTLD than AD/DLB, we have studied FTLD brain samples immunochemically and immunohistochemically. The antibodies to TDP-43 have immuno-stained neuronal inclusions and dystrophic neurites in the...  Read more

Neumann, Sampathu, Kwong, and colleagues have resolved a long-standing issue in the research field of frontotemporal lobar degeneration (FTLD) and amyotrophic lateral sclerosis (ALS). These authors have identified TDP-43 as a major component of ubiquitin-positive inclusions that characterize these disorders. They first extracted a fraction from the patients' brains using monoclonal antibodies and then analyzed it by mass spectrometry. Their findings have greatly facilitated the understanding of the molecular pathogenesis of FTLD and ALS.

Independently, we have also found TDP-43 as a component of the inclusions in FTLD [1]. Following electrophoresis of the sarkosyl-insoluble brain extracts from FTLD, Alzheimer disease (AD) and dementia with Lewy bodies (DLB), we have done exhaustive analyses by mass spectrometry. Following identification of each molecule that is more abundant in FTLD than AD/DLB, we have studied FTLD brain samples immunochemically and immunohistochemically. The antibodies to TDP-43 have immuno-stained neuronal inclusions and dystrophic neurites in the hippocampus and the temporal cortex in FTLD, and skein-like inclusions in the spinal cord in FTLD and ALS. Immunoblotting of the sarkosyl-insoluble fraction has shown abnormal changes in TDP-43 including hyperphosphorylation, fragment formation, and smear-like staining, all of which are similar to abnormal tau in AD and suggest a central role for the formation of abnormal aggregates. These findings are comparable with those by Lee's group. This is not surprising, since both groups have employed principally the same polyclonal antibody which is the only commercially available rabbit polyclonal.

In addition, however, we have found TDP-43-positive glial inclusions in the spinal cord in FTLD and ALS. These inclusions were also positive for tau. The distribution of glial inclusions was consistent with the degenerating areas, suggesting that glial abnormalities are involved in the pathological processes of ALS and FTLD. A difference between our results and theirs is the TDP-43-positive staining of some, but not all, tau-positive structures including Pick bodies and neurofibrillary tangles. The significance of these findings remains to be established, since immunoblot analysis did not show any abnormality in TDP-43 in Pick disease and Alzheimer disease. Our paper will appear shortly in Biochem Biophys Res Commun [1].

In the case of tau and α-synuclein, detection of abnormally modified molecules has revealed far more extensive pathology than that seen by ubiquitin immunohistochemistry. While lesions immunohistochemically labeled for TDP-43 are a little more numerous than those labeled for ubiquitin, the difference is far less than that we have experienced for tau and α-synuclein immunohistochemistry. This may be a point that remains to be cleared up. Another issue that is open for further investigations is to prove, by protein chemistry, the ubiquitination of TDP-43.

In any event, it has to be emphasized that two different approaches have come to the same conclusion, establishing with certainty that TDP-43 is the major component of the inclusions in FTLD and ALS. This further strengthens the hypothesis that these disorders are part of a clinicopathological spectrum that shares similar pathogenesis, and suggests the possibility that TDP-43 may be a common therapeutic target for these disorders. It is now necessary to investigate the relationship of TDP-43 to other molecules that have been reported to be associated with familial FTD, FTD with motor neuron disease, or ALS. Such molecules include progranulin, charged multivesicular body protein 2B (CHMP2B), valosin-containing protein, dynactin, and an unidentified protein in familial disease linked to chromosome 9.

References:
1. T. Arai, M. Hasegawa, H. Akiyama, K. Ikeda, T. Nonaka, H. Mori, D. Mann, K. Tsuchiya, M. Yoshida, Y. Hashizume, T. Oda, TDP-43 is a component of ubiquitin-positive tau-negative inclusions in frontotemporal lobar degeneration and amyotrophic lateral sclerosis, Biochem Biophys Res Commun, in press

View all comments by Tetsuaki Arai

This study elegantly gives a first insight on a transgenic mouse model of mutant TDP-43 (A315T) identified in familial ALS patients. For those in the field, it is clear that generating these mouse models is a mammoth task on its own. Among the many interesting findings in this paper, the first to catch my attention was that the 25-kDa TDP-43 C-terminal fragments (CTFs) were recovered from detergent-soluble fractions but not from urea fractions as observed in sporadic and familial ALS/FTLD patients. If the TDP-43 25-kDa CTFs would indeed be confirmed as the real culprit, this would yet again emphasize the importance of soluble but not aggregated protein/peptide in cellular toxicity, as has been shown for a number of other proteinopathies including Aβ, α-synuclein, polyglutamine expansion in Huntingtin, and mutant SOD1.

Another important observation made in this paper was that ubiquitin-immunoreactive (ir) inclusions observed in select neurons including motor neurons were not TDP-43-ir. Thus, the mutant TDP-43 (A315T) mice do not completely model ALS, where...  Read more

This study elegantly gives a first insight on a transgenic mouse model of mutant TDP-43 (A315T) identified in familial ALS patients. For those in the field, it is clear that generating these mouse models is a mammoth task on its own. Among the many interesting findings in this paper, the first to catch my attention was that the 25-kDa TDP-43 C-terminal fragments (CTFs) were recovered from detergent-soluble fractions but not from urea fractions as observed in sporadic and familial ALS/FTLD patients. If the TDP-43 25-kDa CTFs would indeed be confirmed as the real culprit, this would yet again emphasize the importance of soluble but not aggregated protein/peptide in cellular toxicity, as has been shown for a number of other proteinopathies including Aβ, α-synuclein, polyglutamine expansion in Huntingtin, and mutant SOD1.

Another important observation made in this paper was that ubiquitin-immunoreactive (ir) inclusions observed in select neurons including motor neurons were not TDP-43-ir. Thus, the mutant TDP-43 (A315T) mice do not completely model ALS, where ubiquitin-ir inclusions are also TDP-43-ir; nevertheless, this work does lead to a very interesting question: what are these inclusions composed of?

Knowing earlier studies (see Tatom et al., 2009 and ARF related news story), I am also not surprised at the glaring omission of wild-type TDP-43 mice as a better control than the non-transgenic mice utilized in this study. So although clearly not all is answered yet, let's see how these and other TDP-43 mouse models currently being developed will unfold the mysteries of TDP-43-led neurodegeneration.

View all comments by Samir Kumar-Singh