Clumps of TDP-43 build up in the cytoplasm of motor neurons in most cases of ALS. But how this aggregation occurs remains unclear. Now, Todd Cohen of the University of North Carolina and colleagues report that the acetylation of TDP-43 may promote aggregation of TDP-43, which may be reversed by an HSF1-dependent mechanism.
The findings suggest that upregulating existing refolding mechanisms in motor neurons may be of benefit in many forms of ALS.
“Activating cellular compensatory responses [including] HSF1 to refold these aggregates can reduce inclusions, at least in cultured cell systems,” explained Cohen.
TDP-43’s Special Ks?
In 2015, Todd Cohen, while working in the laboratories of University of Pennsylvania’s Virginia Lee and John Trojanowski, discovered that acetylation regulates the function of TDP-43 and promotes its aggregation. This acetylation, which occurs at lysines 145 and 192 within its RNA-binding domain, reduces its ability to bind RNAs according to CLIP analysis and triggers the formation of insoluble TDP-43 inclusions, at least in cultured cells (Cohen et al., 2015).
Acetylated TDP-43 could be detected in aggregates in the spinal cord of at least some people with ALS according to a post-mortem tissue analysis. But the extent of acetylation, and whether it contributes to the disease, remains unknown according to Cohen.
To explore how acetylation may play a role in ALS, Cohen’s team expressed an acetylated TDP-43 mimic in cultured mammalian cells including neurons. The transfected protein, generated by switching the lysine at residue 145 to a glutamine, led to formation of robust hyper-phosphorylated, ubiquitinated aggregates of TDP-43 in the cytoplasm of these cells and mitochondrial dysfunction. Subsequent immunofluorescence analysis indicated that these aggregates also contained p62, ubiquilin-2 and optineurin – key proteins also detected in motor neuronal inclusions in people with ALS.
The approach is one of the first to rapidly recreate TDP-43 aggregates at high frequency in cellular models of ALS in the laboratory, enabling the study of underlying disease mechanisms.
“It has been very challenging to reproduce TDP-43 pathology using model systems,” according to Cohen, “and it’s not clear why.”
One possibility, he said, is that acetylation, which has not been carefully reproduced in experimental models, is a critical part of the aggregation process.
“We are causing TDP-43 to detach from its targets, and in so doing, it becomes conformationally unstable, and forms really robust pathology similar to what one sees in an ALS patient.”
The reason that TDP-43 becomes acetylated in ALS remains unclear. But it may be a normal part of TDP-43 regulation, controlling its ability to bind and thereby regulate the processing of RNAs, Cohen speculated (Cohen et al., 2015). There is much left to be learned according to Cohen.
“We don’t yet know what enzymes are doing this, and we don’t know where they are localized.” However, if it can be shown that deacetylation of TDP-43 helps restore its normal function in models of ALS, these enzymes could be important targets of the disease.
The Heat Is On
In the meantime, Cohen’s team is investigating motor neurons’ own protein refolding mechanisms in hopes to identify key enzymes that might break up TDP-43 acetylated aggregates. On the top of their list: heat shock transcription factor 1 (HSF1), a central coordinator of the chaperone response, which helps cells manage in part, proteins that are misfolded or aggregated. HSF1, which is activated during times of stress, switches on multiple genes, including the gene encoding the disaggregase Hsp40 (see Torrente and Shorter, 2013).
To investigate whether HSF1 may facilitate the clearance of TDP-43 inclusions, Cohen’s team co-transfected wild-type, constitutively active or inactive HSF1 in cultured cells and determined the number of TDP-43 aggregates. They found that wild-type HSF1 cleared about 80%, and constitutively active HSF1 about 90% of inclusions, while inactive HSF1 had no significant effect.
What’s more, the clearance of TDP-43 aggregates appears to be mediated by key proteins including the disaggregase Hsp40. Hsp40 partially co-localized to TDP-43 aggregates in motor neurons in the spinal cord of people with ALS according to post-mortem tissue analysis. In addition, the overexpression of Hsp40 in cultured cells blocked aggregation of the acetylated TDP-43 mimic whereas siRNA-mediated knockdown of Hsp40 (and Hsp27 to a much lesser extent) resulted in no clearance.
To confirm their results, Cohen’s team also treated cells with compounds known to upregulate HSF1 activity. Treatment of cells with a small-molecule HSF1 activator (called HSF1A) cleared about 60% of TDP-43 aggregates. Riluzole, which has been reported to increase HSF1 activity, had no significant effect.
The results suggest that aggregation of acetylated TDP-43 may be reversed by upregulating key components of the chaperone system. “We think HSF1 is pushing the equilibrium toward a chaperone-rich environment,” Cohen said. “We can reduce inclusions, either genetically or pharmacologically, at least in cultured cell systems,” he said.
Whether TDP-43 is refolded and reactivated or is degraded, however, is still to be determined according to Cohen.
Now, Cohen’s team is using his cellular assay to recreate TDP-43 aggregates in cultured cells and screen for potential therapies that re-activate these folding mechanisms to dissolve them.
There have been many attempts at promoting chaperone-mediated protection in neurodegenerative diseases. But according to Cohen, most strategies developed to date have focused on broad chaperone upregulators, rather than individual chaperones. “There are a lot of reasons to think that increasing the general chaperone response is not a way forward,” said Cohen. “Possibly upregulating specific ones may be more effective.”
Cancer cells rely on increased levels of HSF1 to help them survive, proliferate and invade surrounding tissues according to studies led by late Susan Lindquist at the Whitehead Institute in Cambridge, MA (Dai et al., 2007; Scherz-Shouval et al., 2014). By targeting key chaperones downstream of HSF1, such as Hsp40, we may mitigate at least some of these risks, according to Cohen.
Hsp40, a key component of the motor neuron’s own disaggregase machinery, is present in TDP43 inclusions, he added, strengthening the hypothesis that it may be specifically involved in salvage of misfolded TDP-43.
“What is most interesting in this study is that HSF1 seems to be able to reverse aggregation,” commented James Shorter of the University of Pennsylvania. “That would be a very interesting approach to take in ALS patients.”
But it may not be enough to just break up TDP-43 aggregates, said Shorter, since oligomeric species may be even more toxic (see Fang et al., 2014). And, TDP-43 needs to re-localize back into the nucleus.
Shorter is also developing potential therapies for ALS that target aggregates in motor neurons (see August 2014, May 2017 conference news). One strategy, which involves re-engineering the yeast disaggregase Hsp104, breaks up inclusions – including those that contain TDP-43 (Jackrel et al., 2014; Sweeney et al., 2015). The approach, developed in part by Washington University-St Louis’ Meredith Jackrel while working in the Shorter laboratory, results also in the re-localization of TDP-43 into the nucleus, mitigating cytotoxicity – at least in cultured mammalian cells.
“It may be interesting to explore synergistic approaches to boosting chaperone activity,” Jackrel said.
Developing such a strategy, however, is complicated. In ALS, TDP-43 is just one of dozens of proteins that likely aggregate in motor neurons and astrocytes, noted Justin Yerbury of the University of Wollongong, Australia. “We need to think outside our favorite pathological protein,” said Yerbury. “We need to understand the impact [of the disease] across the entire proteome.”
Many therapeutic hopes have been pinned on restoring proteostasis, and much work has gone into trying to make it work, so far with limited success.
“The proteostasis network is extraordinarily complex with many different interacting components, all of which must be tightly regulated,” said Jackrel.
Justin Yerbury agrees. “Proteostasis is a very complex process to maintain, because there are millions of individual protein molecules that need to be expressed, folded, transported, and degraded to maintain proper function,” explained Yerbury. “Changing a single protein may alter the network in unexpected ways.”
What’s more, according to recent studies led by Yerbury, many proteins are metastable and therefore at high risk of misfolding in motor neurons in ALS, making them vulnerable to the disease (Ciryam et al. 2017).
But, according to Shorter, that does not mean that proteostasis of proteins cannot be modulated successfully to treat neurodegenerative disease.
Progress is being made in one disease, notes Shorter. The small molecule tafamidis slows the progression of transthyretin familial amyloid polyneuropathy (TTR-FAP) in at least some people with the disease according to recent studies (Keohane et al., 2017). The drug, developed by Pfizer, stabilizes transthyretin, reducing aggregation and thereby, amyloidosis. The therapy is now approved for the treatment of TTR-FAP in many places including Europe, Japan and South Korea.
“We’ve known about misfolding and aggregation for a long time [in ALS]. It is frustrating there is so little progress in therapy, said Shorter. “But this does give us some hope that protein misfolding [in ALS] can be addressed therapeutically.”
Wang P, Wander CM, Yuan CX, Bereman MS, Cohen TJ. Acetylation-induced TDP-43 pathology is suppressed by an HSF1-dependent chaperone program. Nat Commun. 2017 Jul 19;8(1):82. [PubMed].
Ciryam P, Lambert-Smith IA, Bean DM, Freer R, Cid F, Tartaglia GG, Saunders DN, Wilson MR, Oliver SG, Morimoto RI, Dobson CM, Vendruscolo M, Favrin G, Yerbury JJ. Spinal motor neuron protein supersaturation patterns are associated with inclusion body formation in ALS. Proc Natl Acad Sci U S A. 2017 May 16;114(20):E3935-E3943. [PubMed].
Jackrel ME, DeSantis ME, Martinez BA, Castellano LM, Stewart RM, Caldwell KA, Caldwell GA, Shorter J. Potentiated Hsp104 variants antagonize diverse proteotoxic misfolding events. Cell. 2014 Jan 16;156(1-2):170-82. [PubMed].
Chen HJ, Mitchell JC, Novoselov S, Miller J, Nishimura AL, Scotter EL, Vance CA, Cheetham ME, Shaw CE. The heat shock response plays an important role in TDP-43 clearance: evidence for dysfunction in amyotrophic lateral sclerosis. Brain. 2016 May;139(Pt 5):1417-32. [PubMed].
Cohen TJ, Hwang AW, Restrepo CR, Yuan CX, Trojanowski JQ, Lee VM. An acetylation switch controls TDP-43 function and aggregation propensity.Nat Commun. 2015 Jan 5;6:5845. [PubMed].
Sweeny EA, Jackrel ME, Go MS, Sochor MA, Razzo BM, DeSantis ME, Gupta K, Shorter J. The Hsp104 N-terminal domain enables disaggregase plasticity and potentiation. Mol Cell. 2015 Mar 5;57(5):836-49. [PubMed].
Fang YS, Tsai KJ, Chang YJ, Kao P, Woods R, Kuo PH, Wu CC, Liao JY, Chou SC, Lin V, Jin LW, Yuan HS, Cheng IH, Tu PH, Chen YR. Full-length TDP-43 forms toxic amyloid oligomers that are present in frontotemporal lobar dementia-TDP patients. Nat Commun. 2014 Sep 12;5:4824. [PubMed].
Dai C, Whitesell L, Rogers AB, Lindquist S. Heat shock factor 1 is a powerful multifaceted modifier of carcinogenesis. Cell. 2007 Sep 21;130(6):1005-18. [PubMed].
Scherz-Shouval R, Santagata S, Mendillo ML, Sholl LM, Ben-Aharon I, Beck AH, Dias-Santagata D, Koeva M, Stemmer SM, Whitesell L, Lindquist S. The reprogramming of tumor stroma by HSF1 is a potent enabler of malignancy. Cell. 2014 Jul 31;158(3):564-78. [PubMed].
Webster CP, Smith EF, Shaw PJ, De Vos KJ. Protein Homeostasis in Amyotrophic Lateral Sclerosis: Therapeutic Opportunities? Front Mol Neurosci. 2017 May 2;10:123. [PubMed].
Shorter J. Designer protein disaggregases to counter neurodegenerative disease. Curr Opin Genet Dev. 2017 Jun;44:1-8. [PubMed].