
==== Front
J Biol Chem
J Biol Chem
The Journal of Biological Chemistry
0021-9258
1083-351X
American Society for Biochemistry and Molecular Biology

S0021-9258(24)02161-6
10.1016/j.jbc.2024.107660
107660
Research Article
TDP43 and huntingtin Exon-1 undergo a conformationally specific interaction that strongly alters the fibril formation of both proteins
George Gincy
Ajayan Anakha
Varkey Jobin
Pandey Nitin K.
Chen Jeannie
Langen Ralf langen@usc.edu
∗
Physiology and Neuroscience, Zilkha Neurogenetic Institute, Keck School of Medicine, University of Southern California, Los Angeles, California, USA
∗ For correspondence: Ralf Langen langen@usc.edu
13 8 2024
9 2024
13 8 2024
300 9 10766011 6 2024
25 7 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by/4.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Protein aggregation is a common feature of many neurodegenerative diseases. In Huntington’s disease, mutant huntingtin is the primary aggregating protein, but the aggregation of other proteins, such as TDP43, is likely to further contribute to toxicity. Moreover, mutant huntingtin is also a risk factor for TDP pathology in ALS. Despite this co-pathology of huntingtin and TDP43, it remains unknown whether these amyloidogenic proteins directly interact with each other. Using a combination of biophysical methods, we show that the aggregation-prone regions of both proteins, huntingtin exon-1 (Httex1) and the TDP43 low complexity domain (TDP43-LCD), interact in a conformationally specific manner. This interaction significantly slows Httex1 aggregation, while it accelerates TDP43-LCD aggregation. A key intermediate responsible for both effects is a complex formed by liquid TDP43-LCD condensates and Httex1 fibrils. This complex shields seeding competent surfaces of Httex1 fibrils from Httex1 monomers, which are excluded from the condensates. In contrast, TDP43-LCD condensates undergo an accelerated liquid-to-solid transition upon exposure to Httex1 fibrils. Cellular studies show co-aggregation of untagged Httex1 with TDP43. This interaction causes mislocalization of TDP43, which has been linked to TDP43 toxicity. The protection from Httex1 aggregation in lieu of TDP43-LCD aggregation is interesting, as it mirrors what has been found in disease models, namely that TDP43 can protect from huntingtin toxicity, while mutant huntingtin can promote TDP43 pathology. These results suggest that direct protein interaction could, at least in part, be responsible for the linked pathologies of both proteins.

Keywords

Huntington’s disease
ALS
huntingtin exon-1
TDP43
protein aggregation
coaggregation
electron paramagnetic resonance
condensates
Abbreviations

HD Huntington’s disease

Httex1 huntingtin exon-1

TDP43 Tar-DNA binding protein

Reviewed by members of the JBC Editorial Board. Edited by Ursula Jakob
==== Body
pmcHuntington’s disease (HD) is a neurodegenerative disorder caused by abnormal CAG expansions in the exon1 (Httex1) of huntingtin, giving rise to a protein with a stretch of 36 or more consecutive Gln (polyQ) residues (1, 2, 3, 4). Mutant Httex1 is generated by aberrant splicing of huntingtin with abnormal CAG expansions, and it is sufficient to induce disease in HD animal models (5). In disease, it is found to misfold into various oligomeric species and amyloid-like fibrils, which deposit inside affected neurons. Some of these misfolded species are thought to exert cytotoxicity, potentially by disrupting cellular processes, including transcription (6, 7, 8), protein homeostasis (9, 10, 11), and RNA metabolism (12). Mutant huntingtin aggregates can sequester and interfere with the function of proteins such as RNA-binding proteins (13), and transcription factors (14, 15, 16). Among the proteins found in Httex1 aggregates is the Tar-DNA binding protein (TDP43). TDP43 is a nuclear RNA-binding protein involved in amyotrophic lateral sclerosis (ALS), limbic-predominant age-related TDP-43 encephalopathy (LATE), and frontotemporal lobar degeneration (FTLD) (17, 18, 19, 20). In these diseases, TDP43 and its C-terminal fragments containing the low complexity domain (LCD) mislocalize into cytoplasmic inclusions (21, 22). PolyQ expansions are amyloidogenic and they are increasingly recognized as a risk factor for ALS (23, 24, 25, 26). Moreover, some HD patients show ALS-like clinical symptoms (23, 27, 28, 29, 30, 31). Interestingly, TDP43 also mislocalizes nuclear and cytoplasmic huntingtin-containing inclusions in human HD and HD R6/2 transgenic mouse brains (32, 33, 34, 35). However, it remains unclear how the TDP43-huntingtin co-localization occurs and whether it is driven by a direct interaction between the two aggregation-prone proteins.

TDP43 and Httex1 share many similarities in terms of their misfolding properties. Httex1 monomers with expanded polyQ can form oligomers, protofibrils, and fibrils (36, 37, 38, 39, 40, 41, 42, 43, 44). The early aggregation intermediates, such as oligomers, protofibrils, and unbundled fibrils are believed to be more toxic than the bundled fibrils (44, 45, 46, 47, 48, 49). Like Httex1, the low-complexity C-terminal fragment of TDP43 (TDP43-LCD) can also form oligomers and protofibrils prior to the formation of fibrils (50, 51). The similarities between the two proteins only partially extend to liquid-liquid phase separation (LLPS), a process that significantly affects fibrillization (52, 53, 54, 55, 56, 57). While TDP43 readily undergoes LLPS (55, 56, 57), Httex1 requires crowding agents as well as a C-terminal fluorescent protein tag, such as GFP or RFP, for LLPS formation (54).

Here we sought to examine whether mutant Httex1 and TDP43 can physically interact, which of the respective conformers are involved, and how these interactions impact the misfolding of the respective proteins. The answer to these questions could shed light on the origin of huntingtin and TDP43 colocalization, as well as the linked pathologies of these proteins. We, therefore, evaluated the interaction between different Httex1 and TDP43-LCD conformers and the resulting effects on their respective misfolding propensities through a combination of biophysical, biochemical, and cell biological techniques.

Results

Httex1 aggregates sequester cytoplasmic TDP43

Cellular interactions between Httex1 and TDP43 are often examined using fluorescently tagged proteins (58, 59, 60). A caveat of this approach is that tagging with genetically encoded fluorescent proteins may alter the biophysical characteristics and/or misfolding steps of the protein under investigation. For example, we observed that the addition of an RFP tag significantly promoted liquid-liquid phase separation followed by rapid fibrillization of Httex1-RFP whereas the untagged protein did not undergo a liquid phase during the aggregation process (54). When expressed in cells, the tagged Httex1 aggregates exhibit a different morphology and antibody reactivity as compared to the untagged Httex1 aggregates (54, 61), suggesting that they may be structurally distinct. To investigate whether untagged Httex1 recruits TDP43, we expressed Httex1(Q72) in HEK293T cells and visualized Httex1 and endogenous TDP43 using immunofluorescence microscopy (Fig. 1). Httex1(Q72) readily formed round aggregates in some transfected cells which were labeled by PHP1, an antibody that recognizes the proline-rich domain of Httex1 (62) (Fig. 1, arrowheads). As previously shown, the PHP1 fluorescent signal was restricted to the periphery of the aggregates, presumably due to limited access to the interior of the dense structures (54). Some transfected cells showed diffuse cytoplasmic labeling, likely reflecting monomeric Httex1(Q72). In cells with round aggregates, TDP43 reactivity largely co-localized with Httex1(Q72), while its normal nuclear signal was strongly reduced (Fig. 1, arrowheads). In contrast, cells with diffuse cytoplasmic Httex1(Q72) did not exhibit nuclear depletion of TDP43 and thus TDP43 did not co-localize with Httex1 (Fig. 1, arrows). These results show that misfolded Httex1 without any fluorescent protein tags recruits TDP43, thereby reducing freely available nuclear TDP43.Figure 1 Misfolded Httex1 causes nuclear depletion and sequestration of endogenous TDP43 to Httex1 aggregates. Representative images of HEK293T cells 24 h after transfection with Httex1(Q72) expression construct. Httex1 is visualized by PHP1 antibody (magenta), endogenous TDP43 is visualized by TDP43 antibody (green) and cell nuclei are visualized by DAPI counterstain (blue). The arrows highlight cells with cytoplasmic Httex1(Q72) and normal nuclear TDP43 labeling. Arrowheads point to cells with aggregates where Httex1(Q72) and TDP43 colocalize and TDP43 signal appears depleted from the nucleus. Scale bar is 20 μm for all the panels.

Httex1 monomers do not interact with TDP43-LCD

We used purified recombinant proteins to investigate direct physical interactions between Httex1 and TDP43. It has previously been shown that the low complexity domain of TDP43 (TDP43-LCD) is sufficient for phase separation. We therefore used TDP43-LCD, which can be readily purified from recombinant sources, for studying the Httex1-TDP43 interaction. First, we investigated the interactions of the freshly prepared proteins prior to aggregation into fibrils. To avoid label-induced artifacts, a small amount of Alexa Fluor-labeled protein (2%) was spiked into the respective unlabeled proteins. Fluorescence microscopy of TDP43-LCD sparsely labeled with Alexa Fluor 488 showed that this protein gave rise to rounded structures (Fig. 2, A1 and A2), consistent with reports that TDP43-LCD undergoes LLPS at physiological pH without the need for any crowding agents (55). In contrast, no visible phase separation was observed for freshly prepared, largely monomeric Httex1(Q46) in the absence of TDP43-LCD (Fig. 2 B1 and B2). This is also consistent with prior studies, which indicated that Httex1 phase separation requires crowding agents and C-terminal RFP or GFP tags (54). When the two proteins were mixed, TDP43-LCD retained its ability to undergo phase separation (Fig. 2C1). Interestingly, TDP43-LCD not only failed to recruit Httex1(Q46) into its condensates, but Httex1(Q46) was actively excluded from phase separated TDP43-LCD (Fig. 2 C1, C2, C3). This can be seen from the dark spots for Httex1(Q46) (Fig. 2, C2, and C3) at locations where TDP43-LCD condensates were present (for quantification see Fig. 2C4).Figure 2 Httex1(Q46) monomers are excluded from TDP43-LCD condensates. Fluorescence (A1) and DIC (A2) images of 2% Alexa Fluor 488 labeled TDP43-LCD alone at 50 μM reveal rounded puncta that are typical for TDP43 condensates. Fluorescence (B1) and DIC imaging (B2) show a homogeneous distribution of 2% Alexa Fluor 594-labeled Httex1(Q46) monomer (magenta) alone, suggesting it does not undergo detectable phase separation at 10 μM. When 50 μM of Alexa Fluor 488 labeled TDP43-LCD are mixed with 10 μM of freshly prepared Alexa Fluor 594 labeled Httex1(Q46), TDP-43-LCD retains its ability to phase separate (see green puncta in C1). However, Httex1 (magenta) is largely excluded from TDP43-LCD condensates (C2), as illustrated with the merged images (C3). C4 shows a quantification of the fluorescence intensities of TDP43-LCD (green) and Httex1(Q46) (magenta) along the line shown in C3. Note that the Httex1(Q46) signal does not approach 0 at the location of the condensates. This would have not been expected, as Httex1(Q46)-containing bulk solution was present above and below the TDP43-LCD condensates, contributing to the signal. A more detailed quantification of the amounts of monomeric Httex1(Q46) in condensates is given in Figure 4. The scale bars are 10 μm. Images are representative of at least three independent experiments.

Httex1 fibrils interact with TDP43-LCD

Having shown that the Httex1(Q46) monomers are largely excluded from TDP43-LCD condensates, we next examined whether Httex1(Q46) fibrils might behave differently. Toward this end, we incubated 50 μM freshly prepared TDP43-LCD with 5 μM Httex1(Q46) fibrils and tested for colocalization using fluorescence microscopy. Under these conditions (ten-fold excess of TDP43-LCD), nearly all Httex1(Q46) fibrils co-localized with TDP43-LCD condensates, but many of the TDP43-LCD condensates did not contain detectable Httex1 fibrils (Fig. 3, A1-3 and C). When the Httex1(Q46)/TDP43-LCD ratio was raised to 1:2, almost complete colocalization was observed (Fig. 3 B1-3 and D). Thus, in contrast to Httex1(Q46) monomers, fibrils bind to TDP43-LCD condensates in a dose-dependent manner.Figure 3 Httex1 fibrils colocalize with TDP43-LCD condensates. (A1-3) Fluorescence microscopy images of a freshly prepared mixture of 5 μM Alexa Fluor 594 labeled Httex1(Q46) fibrils (magenta) in the presence of 10-fold excess (50 μM) of Alexa Fluor 488 labeled TDP43-LCD condensates (green). Most Httex1-Q46 fibrils colocalize with TDP43-LCD (A2-3). The colocalization is further quantified in (C) using a bar graph showing the percentage of colocalization between Httex1 fibrils and TDP43-LCD and vice versa. While nearly 100% of the Httex1 fibrils colocalized with TDP43-LCD condensates, about 60% of the TDP43-LCD condensates contained fibrils. (B) The experiments in (A) were repeated with 2-fold excess of TDP43-LCD by mixing 10 μM Alexa Fluor 594 labeled Httex1(Q46) fibrils (B2, magenta) with 20 μM Alexa Fluor 488 labeled TDP43-LCD (B1, green). The colocalization can be seen in the merged image (B3). The quantification in (D) reveals nearly 100% colocalization of Httex1 fibrils and TDP43-LCD condensates. Scale bars are 10 μm. Images in 3A and 3B are representative of at least three independent experiments, all of which were used for the bar graphs and error bars in 3C and 3D.

TDP43-LCD condensates pull down Httex1 fibrils, but not Httex1 monomers

To further validate the imaging results, we tested whether Httex1(Q46) monomers or fibrils can co-pellet with TDP43-LCD condensates. This readout is based on the observation that TDP43-LCD condensates readily form a transparent pellet at 123,000g. The amounts of Httex1(Q46) monomer and fibril in the pellet and supernatant were quantified using EPR by virtue of a spin label at position 35 in the polyQ, referred to as 35R1. As a control, we tested whether Httex1(Q46)-35R1 monomers pelleted in the absence of TDP43-LCD. As expected for the soluble, monomeric protein, nearly identical concentrations were obtained for freshly prepared Httex1(Q46)-35R1 in the top half and bottom half volume of the test tubes after centrifugation (Fig. 4A). When Httex1(Q46)-35R1 monomer was centrifuged in the presence of TDP43-LCD condensates, a relatively low concentration of Httex1(Q46) was detected in the pellet (<0.3 μM), while that of the supernatant was much higher (16 μM) (Fig. 4A). Thus, Httex1(Q46) monomer was largely excluded from the pellet volume, which was almost entirely taken up by TDP43-LCD condensates. This result is consistent with the imaging experiments, which showed that monomeric Httex1(Q46) was unable to partition into TDP43-LCD condensates in any appreciable manner.Figure 4 Httex1(Q46) fibrils, but not monomers co-precipitate with TDP43-LCD condensates. The co-precipitation of 35R1 spin-labeled Httex1 (A) monomers and (B) fibrils with TDP43-LCD condensates was quantified by EPR spectroscopy. (A) In the absence of TDP43-LCD condensates, no pellet was observed and an equal concentration of Httex1(Q46)-35R1 monomer was in the top and bottom regions of the centrifuged tubes (solid bars). In the presence of TDP43-LCD condensates, Httex1 monomers were almost exclusively in the supernatant and not in the TDP43-LCD condensate-containing pellet (striped bars). Panel (B) shows the results from a pull-down of Httex1(Q46)-35R1 fibrils with TDP43-LCD condensates (striped bars). Fibrils were strongly enriched in the pellet fraction. The control experiment in the absence of TDP43-LCD condensates (solid bars) show that the concentration of Httex1(Q46)-35R1 fibrils in the top volume and bottom volume are nearly identical. The concentration of TDP43-LCD was 50 μM and the starting concentration of Httex1(Q46) fibrils or monomers was 20 μM. Data points in the bar graph are from three independent experiments.

Next, the experiments were repeated with Httex1(Q46)-35R1 fibrils that had been sonicated and unbundled (see Methods). Due to their small size, these fibrils did not pellet significantly on their own. In fact, they showed nearly identical concentrations in the top volume and bottom volume, when spun in the absence of TDP43-LCD condensates (Fig. 4B). However, in the presence of TDP43-LCD condensates, the vast majority of the Httex1(Q46)-35R1 fibrils were in the pellet fraction (Fig. 4B). These data further support the notion that Httex1(Q46) fibrils, but not monomers bind to TDP43-LCD condensates. These results are furthermore consistent with our cell data, which show that fibrillar aggregates, but not diffusive, presumably monomeric Httex1 sequester TDP43 (Fig. 1).

TDP43-LCD inhibits Httex1(Q46) aggregation

To test whether the mixing of TDP43-LCD with Httex1(Q46) fibrils altered Httex1(Q46) aggregation, we employed an EPR-based aggregation readout (39). This readout leverages the characteristic spectral changes that occur as spin-labeled sites in the polyQ, such as Httex1(Q46)-35R1, change from their relatively mobile state in the monomer to an immobile state with spin-spin coupling in fibril (39). This spectral change is accompanied by a pronounced reduction in intensity, which can be conveniently monitored to follow the aggregation kinetics (39). In the absence of TDP-43, Httex1(Q46)-35R1 at 20 μM concentration underwent significant aggregation, as evidenced by the strong reduction of the EPR amplitude after 24 h (Fig. 5-top left). In contrast, the addition of TDP43-LCD slowed aggregation and reduced the concomitant amplitude loss in a dose-dependent manner (Fig. 5). While 1 μM TDP43-LCD already led to a noticeable reduction in Httex1(Q46)-35R1 aggregation, close to complete inhibition was observed in the presence of 5 μM TDP43-LCD. This is a remarkably strong inhibition, especially considering that TDP43-LCD is sub-stoichiometric relative to the 4-fold excess of Httex1(Q46).Figure 5 TDP43-LCD inhibits Httex1(Q46) aggregation in sub-stoichiometric manner. EPR spectra of 20 μM Httex1(Q46)-35R1 in the presence of the indicated concentrations of TDP43-LCD. The drop in spectral amplitude when comparing the spectra at 0 h (blue) and 24 h (orange) is a measure of Httex1(Q46)-35R1 aggregation over time, which is reduced by increasing amounts of TDP43-LCD.

Since TDP-43-LCD interacted with Httex1(Q46) fibrils, but not monomers, it appeared likely that TDP-43-LCD inhibited Httex1(Q46) aggregation by preventing access to seeding-competent surfaces on Httex1(Q46) fibrils. To further investigate whether TDP43-LCD interfered with seeded aggregation of Httex1(Q46), we repeated the EPR-based aggregation measurements in the presence of 5% Httex1(Q46) sonicated fibrils, which were used as seeds. In the absence of TDP43-LCD, the Httex1(Q46) aggregation was very rapid and largely complete after 3 h (Fig. 6A). In the presence of TDP43-LCD condensates, however, only minor spectral changes were observed (Fig. 6B). These data indicated that freshly prepared TDP43-LCD potently inhibited the seeding process, consistent with the ability of TDP43-LCD condensates to bind and sequester fibrils. As a control, we also tested whether pre-aggregated, fibrillar TDP43-LCD can inhibit Httex1 aggregation. However, the addition of TDP43-LCD fibrils did not reveal an inhibitory effect (Fig. 6C).Figure 6 Inhibition of seeded Httex1(Q46) aggregation by TDP43-LCD condensates, but not fibrils. Aggregation in the presence of 5% Httex1(Q46) fibrils was monitored by EPR using the Httex1(Q46)-35R1 derivative. EPR spectra were obtained at 0 h (blue), and 3 h (orange) in (A) the absence or presence of (B) of 20 μM TDP43-LCD condensates or (C) 20 μM TDP43-LCD fibrils. The presence of TDP43-LCD condensates almost completely prevented seeded aggregation of Httex1(Q46)-35R1, while preformed fibrils of TDP43-LCD had no such effect. The Httex1(Q46) concentration was 20 μM.

Httex1 fibrils promote the aggregation of TDP43-LCD

Having shown that the Httex1-TDP43 interaction strongly inhibits Httex1(Q46) aggregation, we next explored the reverse scenario and asked whether TDP-43 aggregation might be altered by the presence of Httex1(Q46) fibrils. Towards this end, we employed ThT fluorescence, where TDP43-LCD was incubated with increasing concentrations of Httex1(Q46) fibrils. TDP43-LCD aggregation in the absence of Httex1(Q46) fibrils gave rise to a t1/2 of ∼7 h (Fig. 7, blue). In the presence of increasing concentrations of Httex1(Q46) fibrils, the normalized ThT fluorescence increased, resulting in t1/2 values of ∼ 4 h and ∼1.5 h for 5 μM and 20 μM Httex1(Q46) fibrils, respectively (Fig. 7, orange and grey). Overall, these data indicated that TDP43-LCD aggregation was enhanced by Httex1(Q46) fibrils.Figure 7 Httex1(Q46) fibrils promote the aggregation of TDP43-LCD. ThT fluorescence of 50 μM TDP43-LCD alone (blue), 50 μM TDP43-LCD in the presence of 5 μM Httex1 fibrils (orange), and 50 μM TDP43-LCD in the presence of 20 μM Httex1 fibrils (grey). ThT curves were normalized by setting their starting values to 0 and the final value at 35 h to 1. Error bars represent the standard deviation of the mean obtained from at least three independent measurements.

To investigate the morphological changes caused by the interaction of TDP43-LCD with Httex1(Q46) fibrils, we performed fluorescence microscopy imaging after 4 h of co-incubation. When TDP43-LCD was aged for 4 h in the absence of Httex1(Q46) fibrils, fluorescence microscopy revealed that only ∼10% of condensates became irregularly shaped (Fig. S1). In contrast, ∼60% of TDP43-LCD condensates became more irregularly shaped in the presence of Httex1(Q46) fibrils. Moreover, the formation of much larger aggregates was commonly observed in the presence of Httex1(Q46) fibrils (Fig. S1, white arrows). Interestingly, several of the irregularly shaped structures exhibited elongated protrusions (see zoomed-in insets in Fig. S1B). These structures were reminiscent of fibrils that have been observed to emanate from condensates in other studies on phase-separating proteins (53, 54, 55, 63, 64, 65).

To further validate the notion that Httex1 fibrils facilitate the liquid-to-solid transition of TDP43-LCD, we performed fluorescence recovery after photobleaching (FRAP). As shown in Figure S2, TDP43-LCD alone exhibited significant and rapid recovery. In contrast, the addition of Httex1(Q46) fibrils strongly reduced the amplitude and rate of recovery, consistent with a reduced liquid-like character of the TDP43-LCD condensates in the presence of Httex1(Q46).

Discussion

Motivated by linked pathologies and colocalization of TDP43 and huntingtin in aggregates, we investigated how these proteins interact and how these interactions affect aggregation. We observed conformation-dependent interactions, which had very different effects on the aggregation of the respective proteins. While Httex1(Q46) aggregation was inhibited by TDP43-LCD, the TDP43-LCD liquid-to-solid transition was promoted by Httex1(Q46).

One might have expected that TDP43-LCD and Httex1(Q46) form hetero-coacervates that alter the respective aggregation propensities, considering prior reports showing that TDP43 and fluorescent protein-tagged Httex1 can undergo LLPS (51, 54, 55, 66, 67). However, monomeric Httex1(Q46) was largely excluded from TDP43-LCD condensates (Figs. 2 and 8) and cells with diffuse Httex1 signal showed little to no co-localization with TDP43 (Fig. 1). In contrast to Httex1(Q46) monomers, Httex1(Q46) fibrils readily interacted with TDP43-LCD condensates (Fig. 8 A, B, and E), and this interaction altered the aggregation of both Httex1 and TDP43-LCD. TDP43-LCD strongly inhibited the misfolding of Httex1(Q46) at concentrations that are sub-stoichiometric relative to the free Httex1(Q46) monomer concentrations. These data further illustrate that the inhibition was not mediated by stoichiometric binding to bulk monomers. Rather, our EPR data suggest that TDP43-LCD condensates interfere with the seeding process (Fig. 6A). In the presence of TDP43-LCD condensates, only a few Httex1 seeds are freely available to promote further aggregation (transition from B to C in Fig. 8). Instead, Httex1 fibrils or seeds are readily engulfed by TDP43-LCD condensates (B to E in Fig. 8). This interaction is likely to have two consequences. First, TDP43-LCD condensates make Httex1 seeds much less accessible to Httex1 monomers, as those are largely excluded from the condensates. This scarcity of unaggregated protein needed for fibril growth and secondary seeding is expected to slow down aggregation. Second, given the favorable interaction between Httex1 fibrils and TDP43-LCD condensates, it seems reasonable to assume that favorable interactions exist between the Httex1 fibril surfaces and TDP43-LCD condensates. This raises the possibility that seeding active surfaces on Httex1 fibrils are covered up by TDP43-LCD.Figure 8 Schematic diagram outlining the conformationally specific interactions of Httex1 and TDP-LCD. The normal aggregation path for Httex1 (from (A) to (B) to (C)) and for TDP43-LCD is altered when both proteins are present together. The key intermediate that changes the aggregation pathways is the complex made up of the LLPS state of TPD43-LCD and Httex1 fibrils (E). This intermediate speeds up the liquid-to-solid transition (D–F) of TDP43-LCD while slowing down the aggregation of Httex1.

In contrast to the inhibitory effect of TDP43-LCD on Httex1(Q46) fibril formation, the opposite effect was observed for TDP43-LCD, as Httex1(Q46) fibrils accelerated the formation of thioflavinT positive structures of TDP43-LCD. The presence of Httex1(Q46) fibrils in TDP43-LCD condensates also significantly slowed the FRAP response of TDP43-LCD and caused TDP43-LCD condensates to become more irregularly shaped with rod-like, fibrillar protrusions visible in fluorescence microscopy. Together these data indicate that Httex1 fibrils promote a liquid-to-solid transition of TDP43 condensates. Critical for this liquid-to-solid transition appears to be the complex of TDP43-LCD condensates with Httex1 fibrils (Fig. 8E), which reduced the aggregation propensity of Httex1 (transition from E to C in Fig. 8), but which has the opposite effect on TDP43-LCD (Fig. 8, transition from E to F). At this point, it is unclear exactly how Httex1 fibrils promote the liquid-to-solid transition in TDP43 condensates. However, it is interesting to note that TDP43-LCD has regions that are Gln and Asn rich. In fact, polyQ binders used to detect Httex1, such as QBP1, also bind to TDP43 aggregates and alter TDP43 aggregation (68). Thus, the surface of TDP43 fibrils might have local regions that are structurally related to Httex1 fibrils. It may also be possible that TDP43 similarly interacts with other aggregate-forming polyQ-containing proteins, such as ataxins implicated in spinocerebellar ataxia which exhibit shared pathological features with TDP43 (24, 26, 69, 70). Whether the Gln and Asn-rich regions of TDP43 are responsible for the observed effects, and whether the present findings might also apply to other polyQ proteins still needs to be determined.

From a disease perspective, the reduction in Httex1 aggregation in lieu of an increased TDP43 aggregation is interesting, as it is consistent with the ability of TDP43 to reduce toxicity in HD models (58, 71). Moreover, the Httex1-dependent increase in TDP-43 aggregation might contribute to increased TDP43 toxicity. One additional aspect that is likely to contribute to TDP43 toxicity is its mislocalization from the nucleus to Httex1 aggregates (32, 33, 34, 35). Prior studies have indicated that passive diffusion plays an important role in the localization and nuclear depletion of TDP43 (72). Thus, it is likely that the affinity of TDP43 for Httex1 fibrils contributes to the TDP43 mislocalization seen in vivo and in our cellular studies using untagged Httex1. This mislocalization is generally thought to be a pathogenic event that contributes to TDP43 pathology (73). In addition, TDP43 mislocalization may also directly impact huntingtin toxicity. That is, a recent study used an HD mouse model and found that a reduction in nuclear TDP43 function increased polyQ expansions (74), a process that is thought to significantly contribute to huntingtin pathogenesis. Future work will be required to address these mechanisms in more detail.

In addition to the role of direct protein–protein interactions studied here, RNA interactions can further modulate TDP43 localization, LLPS, as well as aggregation into solids (75, 76). While RNA interactions typically promote TDP-43 nuclear localization (75) and liquid properties (77), some RNA interactions appear to play important roles in pathogenesis. A recent study showed that TDP43 undergoes aberrant interactions with N1-methyladenosine (m1A) in CAG repeat RNA, leading to mislocalization and the formation of gel-like aggregates (76). Moreover, TDP43-RNA interactions can further impact interactions with some amyloidogenic proteins (78, 79, 80, 81, 82). Future studies will have to show the extent to which various TDP43-RNA contacts might be able to further modulate the Httex1 interactions described here, especially in the context of full-length TDP43.

In summary, our study provides insights into how TDP43 and Httex1 interact and how this interaction affects the aggregation of both proteins. In addition to the potential impacts of these findings on the linked pathologies of the two proteins, the potent Httex1 aggregation inhibition observed here might point to potential new means for inhibiting Httex1 aggregation in vivo. Understanding the precise molecular mechanism by which TDP43 inhibits Httex1 aggregation could facilitate the development of new inhibitors that prevent aggregation of huntingtin and perhaps other polyQ proteins.

Experimental procedures

Expression, purification, and labeling of Httex1(Q46)

Httex1(Q46) and its single Cys mutants were expressed using the pET32a vector as described (44). Briefly, the expression plasmids encoding a Httex1(Q46)-thioredoxin fusion protein (Trx-Httex1(Q46)) were transformed into BL21(DE3) competent cells and grown overnight at 37 ˚C, followed by 50-fold dilution into LB media. Induction was done by adding 1 mM IPTG followed by incubation at 16 °C for 48 h. Cells were harvested by centrifugation at 2300g for 30 min. Bacteria were lysed in 20 mM Tris pH 7.4 containing 20 mM imidazole, 300 mM NaCl, and 1% Triton X-100 (Sigma) using a tip sonicator. For EPR and fluorescence experiments, cys mutants were used for covalent linkage of the respective reporter groups. For these cys mutants, 1 mM DTT was added to the lysis buffer. Lysates were separated from cell debris by centrifugation at 43,000g for 30 min at 4 ˚C and then incubated with NiHis60 (Takara Bio) for 30 min to bind the His-tag containing Trx-Httex1(Q46) protein. The resin was washed with 4 to 5-column volumes of wash buffer (20 mM Tris pH 7.4, 300 mM NaCl, 50 mM imidazole) for cysteine-free Httex1. For cysteine mutants Httex1(Q46)-35C (in polyQ) and Httex1(Q46)-111C (in PRD), the wash buffer contained an additional 1 mM DTT. Further, excess DTT was washed away prior to the elution using 20 mM Tris pH 7.4, 300 mM NaCl, 300 mM imidazole, 4M urea from NiHis60 beads. The same elution buffer was also used for the unlabeled Httex1(Q46). The cys mutants were labeled immediately after elution using a 5-fold molar excess of (1-oxyl-2,2,5,5-tetramethyl-Δ3-pyrroline-3-methyl) methanethiosulfonate (MTSL;Toronto Research Chemicals) to give rise to side chain R1 or Alexa-Fluor 594 (Thermo Fisher Scientific). Excess MTSL spin label or Alexa Fluor was removed using HiTrap Q XL anion-exchange column (Cytiva Life Sciences). The thioredoxin fusion tag was then cleaved by digesting the fusion protein with enterokinase (EKMax) (Thermo Fisher Scientific) for 30˚min at 32 °C. Subsequent purification was performed using the reverse-phase Phenomenex C4 column as described (39). Httex1 fractions were collected and lyophilized until further use. Httex1(Q46) was filtered using a 100 kD Amicon filter before each experiment.

Expression and purification of TDP43-LCD

The TDP43 low complexity domain (LCD) (aa. 267–414) wild-type constructs cloned in a pRSET- B vector was a gift from Dr Witold Surewicz (55). The construct encoded the LCD protein with an N-terminal histidine tag which can be cleaved off using thrombin. The plasmid was transformed into BL21(DE3) competent cells and starter cultures were grown at 37 ˚C for 16 h followed by 50-fold dilution into LB media. For induction 1 mM IPTG was added, followed by incubation at 16 °C for 48 h. Cells were harvested by centrifugation at 2300g for 30 min. Bacterial cells were lysed in 8 M urea, 20 mM Tris pH 7.4, 25 mM imidazole, 300 mM NaCl containing 1% Triton ×-100 (Sigma) using a tip sonicator. For cys mutants of TDP43-LCD, 1 mM DTT was added to the lysis buffer. Lysates were separated from cell debris by centrifugation at 43,000g for 30 min at 4 ˚C and then incubated with NiHis60 (Takara Bio) for 30 min to bind the TDP43-LCD protein which contained an N-terminal His tag. After 30 min, the resin was washed with 4 to 5 column volumes of wash buffer (20 mM Tris pH 7.4, 300 mM NaCl, 50 mM imidazole) for TDP43-LCD. For Alexa Fluor 488 (Thermo Fisher Scientific) labeling, a cysteine was introduced at residue 311. For this cys mutant, the wash buffer contained an additional 1 mM DTT. An additional washing step using DTT-free wash buffer was added to remove excess DTT prior to the elution of the protein using 20 mM Tris pH 7.4, 300 mM NaCl, 300 mM imidazole, 8 M urea from NiHis60 beads. The cys mutant was labeled immediately with either MTSL or Alexa Fluor 488 as described for Httex1(Q46). The final purification step involved a C4 reverse column as in the case of Httex1. The TDP43-LCD-containing fraction was collected and lyophilized. For unlabeled TDP43-LCD, absorption at 280 nm was used for concentration determination using an extinction coefficient of 17,990 M-1 cm-1.

Preparation of Httex1(Q46) fibrils

Httex1 fibrils were grown in 20 mM Tris, 150 mM NaCl pH 7.4 from a solution of 20 μM Httex1 monomers that were incubated without agitation at 4 ˚C. A previously described protocol was used to disaggregate the fibrils (39). After 20 h of incubation, fibrils were harvested at 123,000g for 60 min and later redissolved in a 1:4000 mixture of TFA and H2O. Fibrils were further disaggregated using tip sonication and concentrations were determined by CD.

Fluorescence microscopy of recombinantly generated proteins and generation of TDP43-LCD condensates

Lyophilized TDP43-LCD was taken up in Milli-Q water and diluted into the buffer to reach a final buffer concentration of 50 mM Tris-HCl, 150 mM NaCl pH 7.4. This methodology has previously been described to be sufficient for the formation of TDP43-LCD condensates (55). Lyophilized Httex1(Q46) was taken up in 0.5% TFA/methanol for disaggregation followed by evaporation of the TFA/methanol mixture. The film was then dissolved in 50 mM Tris-HCl, and 150 mM NaCl pH 7.4. Samples were imaged using a Zeiss LSM 800 upright confocal microscope at 63x. For TDP43-LCD droplets, we mixed 2% of Alexa Fluor 488 labeled TDP43 (labeled at position 311) with unlabeled TDP43-LCD. Similarly, for Httex1 monomer, we mixed 2% of Alexa Fluor 594 labeled Httex1 (labeled at position 111) with an unlabeled Httex1 monomer. Fiji was used to quantify the fluorescence intensities in Figure 1C4. The intensity profiles were generated using the Line tool option and analyzed with Plot Profile (Fig. 1C4). The quantification in Fig. S1 was done using Fiji software. A total of 370 structures of TDP43-LCD condensates and 403 structures of TDP43-LCD in the presence of Httex1 fibrils were analyzed. The roundness of condensates was measured using a circularity value ≥ 0.9, while condensates with a value below 0.9 were considered irregular.

Co-sedimentation of Httex1(Q46) with TDP43-LCD condensates

TDP43-LCD condensates were mixed with either Httex1 monomers (pre-filtered using 100 kD Amicon filter) or Httex1 fibrils (sonicated and unbundled). After mixing, TDP43-LCD condensates were pelleted by centrifugation at 123,000g at 4 ˚C. We used spin-labeled Httex1 at position 35 to measure the amount of Httex1(Q46) that co-pelleted with TDP43-LCD droplets using EPR. The control was done by measuring the concentration of the spin-labeled Httex1(Q46)-35R1 protein in the top 50% volume and bottom 50% volume of the Httex1(Q46) monomer and Httex1(Q46) sonicated fibril. Concentration determination was performed using EPR (see below).

EPR spectroscopy

Samples were loaded in borosilicate capillaries with a 0.6 mm inner diameter. The CW-EPR spectra of freshly prepared Httex1(Q46) spin-labeled at residue 35 (Httex1(Q46)-35R1) were recorded at 20 μM on an X-band Bruker EMX EPR spectrometer fitted with a Bruker ER4119HS resonator at room temperature. An incident power of 12.7 mW along with a scan width of 100 G was used. The concentration of spin-labeled protein was measured by double integration of the EPR spectra, and the use of double integrals from concentration standards. All measurements were performed at room temperature.

Thioflavin T(ThT) fluorescence

The aggregation kinetics of purified TDP43-LCD in the presence of Httex1(Q46) fibrils were measured by ThT fluorescence using an Eppendorf Plate Reader AF2200 equipped with appropriate filters in a Falcon 96-well (clear flat bottom) plate, over a period of 35 h at 25 °C. The excitation and emission wavelengths were 440 and 484 nm, respectively, with a slit width of 20 nm each. The concentration of TDP43-LCD was 50 μM while the concentration of Httex1(Q46) fibrils was variable as indicated in the legend. EPR was used for concentration determination of TDP43-LCD by virtue of a spin label at position 307. The fluorescence of all samples was recorded at least in triplicate. The ThT concentration was kept at 50 μM in all cases. The fluorescence intensity values were normalized to the final measurement taken at 35 h, which was set to 1.

FRAP measurements

FRAP of TDP43-LCD was carried out using a Leica SP8 DIVE multiphoton confocal fluorescence imaging system at 63x oil-immersion objective powered by a Chameleon Discovery laser at 1050 nm (Coherent) and DMi8 inverted microscope’s external Leica 4Tune spectral hybrid detectors (emission at 519 nm for Alexa Fluor 488 fluorescence). A region of interest was selected followed by photobleaching for 4.5 s and then capturing images every 0.5 s for ∼45 s after photobleaching. The analysis of FRAP data was performed in Fiji using Stowers ImageJ plugins (https://research.stowers.org/imagejplugins/zipped_plugins.html). The intensity obtained from Fiji was normalized to the prebleach image intensity, which was set to 1.

Cell culture and immunocytochemistry

HEK293T cells were split and seeded onto 24-well plates with poly-D-lysine coated coverslips and grown to 70% confluency. Cells were transfected with 500 ng of Httex1(Q72) plasmid DNA (Genscript) using Lipofectamine LTX with Plus Reagent transfection kit (ThermoFisher), according to the manufacturer’s protocol. Cells were allowed to express constructs for 24 h, after which the cells were fixed with 3.7% formaldehyde, permeabilized with 0.1% Triton ×, treated with 70% formic acid for 1 min, and blocked with 1% bovine serum albumin (BSA) for 1 h at room temperature. To label the Httex1 aggregates and endogenous TDP43, cells were incubated with mouse PHP1 antibody (EMD Millipore) and rabbit TDP43 polyclonal antibody (Proteintech) respectively, both at 1:500 in 0.1% BSA blocking buffer overnight at 4°C. This was followed by incubation with Alexa Fluor 488 donkey anti-rabbit and Alexa Fluor 594 donkey anti-mouse secondary antibodies (Invitrogen) at 1:200 in the same block for 1 h at room temperature. Coverslips with cells were mounted onto glass micro slides with Vectashield antifade mounting medium with DAPI (Vector Laboratories) and imaged with a Zeiss LSM 800 confocal microscope.

Data availability

All data are contained within the manuscript.

Supporting information

This article has supporting information.

Conflict of interests

The authors declare that they have no conflict of interest with the contents of this article.

Supporting information

Supplementary Figures

Author contributions

N. K. P., J. C., R. L., A. A., J. V., and G. G. writing–review & editing; N. K. P., A. A., J. V., and G. G. visualization; N. K. P., A. A., J. V., and G. G. validation; N. K. P., J. C., R. L., A. A., J. V., and G. G. methodology; N. K. P., A. A., J. V., and G. G. investigation; N. K. P., J. C., R. L., A. A., J. V., and G. G. formal analysis; N. K. P., A. A., J. V., and G. G. data curation. J. C., R. L., and G. G. writing–original draft; J. C. and R. L. supervision; J. C. and R. L. resources; J. C. and R. L. project administration, J. C. and R. L. funding acquisition; J. C., R. L., and G. G. conceptualization.

Funding and additional information

This study was supported by funding from the 10.13039/100000052 NIH (NS120704 to RL and JC, NS125769 to R. L. and J. C., and NS118859 to R. L.).
==== Refs
References

1 Walker F.O. Huntington's disease The Lancet 369 2007 218 228
2 Lee J.-M. Ramos E.M. Lee J.H. Gillis T. Mysore J.S. Hayden M.R. CAG repeat expansion in Huntington disease determines age at onset in a fully dominant fashion Neurology 78 2012 690 695 22323755
3 Penney Jr, J.B. Vonsattel J.P. MacDonald M.E. Gusella J.F. Myers R.H. CAG repeat number governs the development rate of pathology in Huntington's disease Ann. Neurol. Official J. Am. Neurol. Assoc. Child Neurol. Soc. 41 1997 689 692
4 Kremer B. Goldberg P. Andrew S.E. Theilmann J. Telenius H. Zeisler J. A worldwide study of the Huntington's disease mutation: the sensitivity and specificity of measuring CAG repeats New Engl. J. Med. 330 1994 1401 1406 8159192
5 Mangiarini L. Sathasivam K. Seller M. Cozens B. Harper A. Hetherington C. Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice Cell 87 1996 493 506 8898202
6 Cha J.-H.J. Transcriptional dysregulation in Huntington’s disease Trends Neurosciences 23 2000 387 392
7 Menalled L. Zanjani H. MacKenzie L. Koppel A. Carpenter E. Zeitlin S. Decrease in striatal enkephalin mRNA in mouse models of Huntington’s disease Exp. Neurol. 162 2000 328 342 10739639
8 Kotliarova S. Jana N.R. Sakamoto N. Kurosawa M. Miyazaki H. Nekooki M. Decreased expression of hypothalamic neuropeptides in Huntington disease transgenic mice with expanded polyglutamine-EGFP fluorescent aggregates J. Neurochem. 93 2005 641 653 15836623
9 Wang J. Wang C.E. Orr A. Tydlacka S. Li S.H. Li X.J. Impaired ubiquitin–proteasome system activity in the synapses of Huntington's disease mice J. Cel. Biol. 180 2008 1177 1189
10 Jana N.R. Zemskov E.A. Wang Gh Nukina N. Altered proteasomal function due to the expression of polyglutamine-expanded truncated N-terminal huntingtin induces apoptosis by caspase activation through mitochondrial cytochrome c release Hum. Mol. Genet. 10 2001 1049 1059 11331615
11 Waelter S. Boeddrich A. Lurz R. Scherzinger E. Lueder G. Lehrach H. Accumulation of mutant huntingtin fragments in aggresome-like inclusion bodies as a result of insufficient protein degradation Mol. Biol. Cel. 12 2001 1393 1407
12 Tano V. Utami K.H. Yusof N.A.B.M. Bégin J. Tan W.W.L. Pouladi M.A. Widespread dysregulation of mRNA splicing implicates RNA processing in the development and progression of Huntington's disease EBioMedicine 94 2023
13 Doi H. Koyano S. Suzuki Y. Nukina N. Kuroiwa Y. The RNA-binding protein FUS/TLS is a common aggregate-interacting protein in polyglutamine diseases Neurosci. Res. 66 2010 131 133 19833157
14 McCampbell A. Taylor J.P. Taye A.A. Robitschek J. Li M. Walcott J. CREB-binding protein sequestration by expanded polyglutamine Hum. Mol. Genet. 9 2000 2197 2202 10958659
15 Nucifora Jr, F.C. Sasaki M. Peters M.F. Huang H. Cooper J.K. Yamada M. Interference by huntingtin and atrophin-1 with cbp-mediated transcription leading to cellular toxicity Science 291 2001 2423 2428 11264541
16 Steffan J.S. Kazantsev A. Spasic-Boskovic O. Greenwald M. Zhu Y.Z. Gohler H. The Huntington's disease protein interacts with p53 and CREB-binding protein and represses transcription Proc. Natl. Acad. Sci. 97 2000 6763 6768 10823891
17 Jo M. Lee S. Jeon Y.M. Kim S. Kwon Y. Kim H.J. The role of TDP-43 propagation in neurodegenerative diseases: integrating insights from clinical and experimental studies Exp. Mol. Med. 52 2020 1652 1662 33051572
18 Arai T. Hasegawa M. Akiyama H. Ikeda K. Nonaka T. Mori H. TDP-43 is a component of ubiquitin-positive tau-negative inclusions in frontotemporal lobar degeneration and amyotrophic lateral sclerosis Biochem. biophysical Res. Commun. 351 2006 602 611
19 Neumann M. Sampathu D.M. Kwong L.K. Truax A.C. Micsenyi M.C. Chou T.T. Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis Science 314 2006 130 133 17023659
20 Nelson P.T. Dickson D.W. Trojanowski J.Q. Jack C.R. Boyle P.A. Arfanakis K. Limbic-predominant age-related TDP-43 encephalopathy (LATE): consensus working group report Brain 142 2019 1503 1527 31039256
21 Igaz L.M. Kwong L.K. Xu Y. Truax A.C. Uryu K. Neumann M. Enrichment of C-terminal fragments in TAR DNA-binding protein-43 cytoplasmic inclusions in brain but not in spinal cord of frontotemporal lobar degeneration and amyotrophic lateral sclerosis Am. J. Pathol. 173 2008 182 194 18535185
22 Barmada S.J. Skibinski G. Korb E. Rao E.J. Wu J.Y. Finkbeiner S. Cytoplasmic mislocalization of TDP-43 is toxic to neurons and enhanced by a mutation associated with familial amyotrophic lateral sclerosis J. Neurosci. 30 2010 639 649 20071528
23 Dewan R. Chia R. Ding J. Hickman R.A. Stein T.D. Abramzon Y. Pathogenic huntingtin repeat expansions in patients with frontotemporal dementia and amyotrophic lateral sclerosis Neuron 109 2021 448 460 33242422
24 Elden A.C. Kim H.J. Hart M.P. Chen-Plotkin A.S. Johnson B.S. Fang X. Ataxin-2 intermediate-length polyglutamine expansions are associated with increased risk for ALS Nature 466 2010 1069 1075 20740007
25 Lee T. Li Y.R. Ingre C. Weber M. Grehl T. Gredal O. Ataxin-2 intermediate-length polyglutamine expansions in European ALS patients Hum. Mol. Genet. 20 2011 1697 1700 21292779
26 Tazelaar G.H. Boeynaems S. De Decker M. van Vugt J.J.F.A. Kool L. Goedee H.S. ATXN1 repeat expansions confer risk for amyotrophic lateral sclerosis and contribute to TDP-43 mislocalization Brain Commun. 2 2020 fcaa064 32954321
27 Kanai K. Kuwabara S. Sawai S. Nakata M. Misawa S. Isose S. Genetically confirmed Huntington's disease masquerading as motor neuron disease Movement Disord. 23 2008 748 751 18186118
28 Papageorgiou S.G. Antelli A. Bonakis A. Vassos E. Zalonis I. Kalfakis N. Association of genetically proven Huntington’s disease and sporadic amyotrophic lateral sclerosis in a 72-year-old woman J. Neurol. 253 2006 1649 1650 17219039
29 Phukan J. Ali E. Pender N.P. Molloy F. Hennessy M. Walsh R.J. Huntington's disease presenting as amyotrophic lateral sclerosis Amyotroph. Lateral Scler. 11 2010 405 407 19530012
30 Rubio A. Steinberg K. Figlewicz D.A. MacDonald M.E. Greenamyre T. Hamill R. Coexistence of Huntington’s disease and familial amyotrophic lateral sclerosis: case presentation Acta neuropathologica 92 1996 421 427 8891076
31 Sadeghian H. O'Suilleabhain P.E. Battiste J. Elliott J.L. Trivedi J.R. Huntington chorea presenting with motor neuron disease Arch. Neurol. 68 2011 650 652 21555641
32 Sanchez I.I. Nguyen T.B. England W.E. Lim R.G. Vu A.Q. Miramontes R. Huntington’s disease mice and human brain tissue exhibit increased G3BP1 granules and TDP43 mislocalization J. Clin. Invest. 131 2021
33 Schwab C. Arai T. Hasegawa M. Yu S. McGeer P.L. Colocalization of transactivation-responsive DNA-binding protein 43 and huntingtin in inclusions of Huntington disease J. Neuropathol. Exp. Neurol. 67 2008 1159 1165 19018245
34 Tada M. Coon E.A. Osmand A.P. Kirby P.A. Martin W. Wieler M. Coexistence of Huntington’s disease and amyotrophic lateral sclerosis: a clinicopathologic study Acta neuropathologica 124 2012 749 760 22735976
35 Nguyen T.B. Miramontes R. Chillon-Marinas C. Maimon R. Vazquez-Sanchez S. Lau A.L. Aberrant splicing in Huntington's disease via disrupted TDP-43 activity accompanied by altered m6A RNA modification bioRxiv 2 2023 565004
36 Miller J. Arrasate M. Brooks E. Libeu C.P. Legleiter J. Hatters D. Identifying polyglutamine protein species in situ that best predict neurodegeneration Nat. Chem. Biol. 7 2011 925 934 22037470
37 Bugg C.W. Isas J.M. Fischer T. Patterson P.H. Langen R. Structural features and domain organization of huntingtin fibrils J. Biol. Chem. 287 2012 31739 31746 22801429
38 Bravo-Arredondo J.M. Kegulian N.C. Schmidt T. Pandey N.K. Situ A.J. Ulmer T.S. The folding equilibrium of huntingtin exon 1 monomer depends on its polyglutamine tract J. Biol. Chem. 293 2018 19613 19623 30315108
39 Pandey N.K. Isas J.M. Rawat A. Lee R.V. Langen J. Pandey P. The 17-residue-long N terminus in huntingtin controls stepwise aggregation in solution and on membranes via different mechanisms J. Biol. Chem. 293 2018 2597 2605 29282287
40 Newcombe E.A. Ruff K.M. Sethi A. Ormsby A.R. Ramdzan Y.M. Fox A. Tadpole-like conformations of huntingtin exon 1 are characterized by conformational heterogeneity that persists regardless of polyglutamine length J. Mol. Biol. 430 2018 1442 1458 29627459
41 Hoop C.L. Lin H.K. Kar K. Magyarfalvi G. Lamley J.M. Boatz J.C. Huntingtin exon 1 fibrils feature an interdigitated β-hairpin–based polyglutamine core Proc. Natl. Acad. Sci. 113 2016 1546 1551 26831073
42 Lin H.-K. Boatz J.C. Krabbendam I.E. Kodali R. Hou Z. Wetzel R. Fibril polymorphism affects immobilized non-amyloid flanking domains of huntingtin exon1 rather than its polyglutamine core Nat. Commun. 8 2017 15462
43 Sivanandam V. Jayaraman M. Hoop C.L. Kodali R. Wetzel R. van der Wel P.C.A. The aggregation-enhancing huntingtin N-terminus is helical in amyloid fibrils J. Am. Chem. Soc. 133 2011 4558 4566 21381744
44 Mario Isas J. Pandey N.K. Xu H. Teranishi K. Okada A.K. Fultz E.K. Huntingtin fibrils with different toxicity, structure, and seeding potential can be interconverted Nat. Commun. 12 2021 4272 34257293
45 Schaffar G. Breuer P. Boteva R. Behrends C. Tzvetkov N. Strippel N. Cellular toxicity of polyglutamine expansion proteins: mechanism of transcription factor deactivation Mol. Cel. 15 2004 95 105
46 Kim Y.E. Hosp F. Frottin F. Ge H. Mann M. Hayer-Hartl M. Soluble oligomers of PolyQ-expanded huntingtin target a multiplicity of key cellular factors Mol. Cel. 63 2016 951 964
47 Leitman J. Ulrich Hartl F. Lederkremer G.Z. Soluble forms of polyQ-expanded huntingtin rather than large aggregates cause endoplasmic reticulum stress Nat. Commun. 4 2013 2753 24217578
48 Arrasate M. Mitra S. Schweitzer E.S. Segal M.R. Finkbeiner S. Inclusion body formation reduces levels of mutant huntingtin and the risk of neuronal death Nature 431 2004 805 810 15483602
49 Nekooki-Machida Y. Kurosawa M. Nukina N. Ito K. Oda T. Tanaka M. Distinct conformations of in vitro and in vivo amyloids of huntingtin-exon1 show different cytotoxicity Proc. Natl. Acad. Sci. 106 2009 9679 9684 19487684
50 Mompeán M. Buratti E. Guarnaccia C. Brito R.M.M. Chakrabartty A. Baralle F.E. Structural characterization of the minimal segment of TDP-43 competent for aggregation Arch. Biochem. Biophys. 545 2014 53 62 24440310
51 Conicella A.E. Zerze G.H. Mittal J. Fawzi N.L. ALS mutations disrupt phase separation mediated by α-helical structure in the TDP-43 low-complexity C-terminal domain Structure 24 2016 1537 1549 27545621
52 Aktar F. Burudpakdee C. Polanco M. Pei S. Swayne T.C. Lipke P.N. The huntingtin inclusion is a dynamic phase-separated compartment Life Sci. alliance 2 2019
53 Peskett T.R. Rau F. O'Driscoll J. Patani R. Lowe A.R. Saibil H.R. A liquid to solid phase transition underlying pathological huntingtin exon1 aggregation Mol. Cel. 7070 2018 588 601
54 Pandey N.K. Varkey J. Ajayan A. George G. Chen J. Langen R. Fluorescent protein tagging promotes phase separation and alters the aggregation pathway of huntingtin exon-1 J. Biol. Chem. 300 2024
55 Babinchak W.M. Haider R. Dumm B.K. Sarkar P. Surewicz K. Choi J.K. The role of liquid–liquid phase separation in aggregation of the TDP-43 low-complexity domain J. Biol. Chem. 294 2019 6306 6317 30814253
56 Li H.-R. Chen T.C. Hsiao C.L. Shi L. Chou C.Y. Huang J.R. The physical forces mediating self-association and phase-separation in the C-terminal domain of TDP-43 Biochim. Biophys. Acta (BBA)-Proteins Proteomics 1866 2018 214 223 28988034
57 Gu J. Wang C. Hu R. Li Y. Zhang S. Sun Y. Hsp70 chaperones TDP-43 in dynamic, liquid-like phase and prevents it from amyloid aggregation Cell Res. 31 2021 1024 1027 34239072
58 Fuentealba R.A. Udan M. Bell S. Wegorzewska I. Shao J. Diamond M.I. Interaction with polyglutamine aggregates reveals a Q/N-rich domain in TDP-43 J. Biol. Chem. 285 2010 26304 26314 20554523
59 Marte L. Boronat S. Barrios R. Barcons-Simon A. Bolognesi B. Cabrera M. Expression of Huntingtin and TDP-43 derivatives in fission yeast can cause both beneficial and toxic effects Int. J. Mol. Sci. 23 2022 3950 35409310
60 Coudert L. Nonaka T. Bernard E. Hasegawa M. Schaeffer L. Leblanc P. Phosphorylated and aggregated TDP-43 with seeding properties are induced upon mutant Huntingtin (mHtt) polyglutamine expression in human cellular models Cell Mol. Life Sci. 76 2019 2615 2632 30863908
61 Riguet N. Mahul-Mellier A.L. Maharjan N. Burtscher J. Croisier M. Knott G. Nuclear and cytoplasmic huntingtin inclusions exhibit distinct biochemical composition, interactome and ultrastructural properties Nat. Commun. 12 2021 6579 34772920
62 Ko J. Isas J.M. Sabbaugh A. Yoo J.H. Pandey N.K. Chongtham A. Identification of distinct conformations associated with monomers and fibril assemblies of mutant huntingtin Hum. Mol. Genet. 27 2018 2330 2343 29912367
63 Patel A. Lee H.O. Jawerth L. Maharana S. Jahnel M. Hein M.Y. A liquid-to-solid phase transition of the ALS protein FUS accelerated by disease mutation Cell 162 2015 1066 1077 26317470
64 Pytowski L. Lee C.F. Foley A.C. Vaux D.J. Jean L. Liquid–liquid phase separation of type II diabetes-associated IAPP initiates hydrogelation and aggregation Proc. Natl. Acad. Sci. 117 2020 12050 12061 32414928
65 Wegmann S. Eftekharzadeh B. Tepper K. Zoltowska K.M. Bennett R.E. Dujardin S. Tau protein liquid–liquid phase separation can initiate tau aggregation EMBO J. 37 2018 e98049
66 Sun Y. Chakrabartty A. Phase to phase with TDP-43 Biochemistry 56 2017 809 823 28112502
67 Chien H.-M. Lee C.-C. Huang J.J.-T. The different faces of the TDP-43 Low-complexity domain: the formation of liquid droplets and amyloid fibrils Int. J. Mol. Sci. 22 2021 8213 34360978
68 Mompean M. Ramírez de Mingo D. Hervás R. Fernández-Ramírez M.D.C. Carrión-Vázquez M. Laurents D.V. Molecular mechanism of the inhibition of TDP-43 amyloidogenesis by QBP1 Arch. Biochem. Biophys. 675 2019 108113
69 Tan C.-F. Yamada M. Toyoshima Y. Yokoseki A. Miki Y. Hoshi Y. Selective occurrence of TDP-43-immunoreactive inclusions in the lower motor neurons in Machado–Joseph disease Acta neuropathologica 118 2009 553 560 19526244
70 Toyoshima Y. Takahashi H. TDP-43 pathology in polyglutamine diseases: with reference to amyotrphic lateral sclerosis Neuropathology 34 2014 77 82 23889603
71 Tauffenberger A. Chitramuthu B.P. Bateman A. Bennett H.P.J. Parker J.A. Reduction of polyglutamine toxicity by TDP-43, FUS and progranulin in Huntington's disease models Hum. Mol. Genet. 22 2013 782 794 23172908
72 Duan L. Zaepfel B.L. Aksenova V. Dasso M. Rothstein J.D. Kalab P. Nuclear RNA binding regulates TDP-43 nuclear localization and passive nuclear export Cell Rep. 40 2022
73 Suk T.R. Rousseaux M.W. The role of TDP-43 mislocalization in amyotrophic lateral sclerosis Mol. neurodegeneration 15 2020 45
74 Bai D. Zhu L. Jia Q. Duan X. Chen L. Wang X. Loss of TDP-43 promotes somatic CAG repeat expansion in Huntington’s disease knock-in mice Prog. Neurobiol. 227 2023 102484
75 Dos Passos P.M. Hemamali E.H. Mamede L.D. Hayes L.R. Ayala Y.M. RNA-mediated ribonucleoprotein assembly controls TDP-43 nuclear retention PLoS Biol. 22 2024 e3002527 38422113
76 Sun Y. Dai H. Dai X. Yin J. Cui Y. Liu X. m1A in CAG repeat RNA binds to TDP-43 and induces neurodegeneration Nature 623 2023 580 587 37938769
77 Grese Z.R. Bastos A.C. Mamede L.D. French R.L. Miller T.M. Ayala Y.M. Specific RNA interactions promote TDP-43 multivalent phase separation and maintain liquid properties EMBO Rep. 22 2021 e53632
78 Dhakal S. Mondal M. Mirzazadeh A. Banerjee S. Ghosh A. Rangachari V. α-Synuclein emulsifies TDP-43 prion-like domain—RNA liquid droplets to promote heterotypic amyloid fibrils Commun. Biol. 6 2023 1227 38052886
79 Tian T. Huang C. Tong J. Yang M. Zhou H. Xia X.G. TDP-43 potentiates alpha-synuclein toxicity to dopaminergic neurons in transgenic mice Int. J. Biol. Sci. 7 2011 234 21448284
80 Wijegunawardana D. Vishal S.S. Venkatesh N. Gopal P.P. Ataxin-2 polyglutamine expansions aberrantly sequester TDP-43, drive ribonucleoprotein condensate transport dysfunction and suppress local translation bioRxiv 15 2023 526372
81 Nihei Y. Ito D. Suzuki N. Roles of ataxin-2 in pathological cascades mediated by TAR DNA-binding protein 43 (TDP-43) and Fused in Sarcoma (FUS) J. Biol. Chem. 287 2012 41310 41323 23048034
82 Josephs K.A. Murray M.E. Whitwell J.L. Parisi J.E. Petrucelli L. Jack C.R. Staging TDP-43 pathology in Alzheimer’s disease Acta neuropathologica 127 2014 441 450 24240737
