Neurodegenerative diseases are fast becoming a major health issue in aging societies. According to the World Health Organization, neurodegenerative diseases will surpass cancer to become the second leading cause of death worldwide by the year 2040 [1]. Treatments for neurodegenerative diseases already constitute a market valued at over US$ 20 billion, and this still expected to grow. A major clinical hallmark of such diseases is the appearance of protein aggregates in neurons. The major constituent protein in these aggregates varies with disease type, although there may be considerable cross-talk between protein pathologies [2]. The 43 kDa transactive response DNA-binding protein (TDP-43) was originally identified more than 20 years ago as a cellular protein that interacts with the transactive response DNA sequence of HIV type 1 [3]. TDP-43 is a member of the heterogeneous nuclear ribonucleoproteins (hnRNP) family, and is implicated in various processes, including transcription, regulation of splicing, mRNA stability, microRNA processing, mRNA transport and translation [4, 5]. Its link to neurodegeneration was not established until 2006 [6]. TDP-43 is directly responsible for certain types of amyotrophic lateral sclerosis, a major debilitating disorder in which the patient gradually loses the ability to move. It is also the major culprit behind fronto-temporal lobar degeneration with ubiquitinated inclusions, one of the most common forms of dementia among patients under the age of 65. TDP-43 pathology has also been detected in patients with tauopathy and Alzheimer's disease, suggesting that TDP-43 may be associated with other types of neurodegenerative diseases [7]. All of these diseases are characterized by the presence of TDP-43 inclusions in the cell. Considerable effort has been devoted to understanding the mechanism behind TDP-43 aggregation, with the hope of finding ways to disrupt or slow down the process [8]. Recent reports on the structure of TDP-43 have the potential to open new avenues for therapeutic intervention in neurodegenerative diseases, although these studies also raise additional questions about the overall structure and function of the N-terminal domain [9-11]. TDP-43 may be roughly divided into four regions (Fig. 1). The middle two regions (residues 106–265) fold into two classic RNA recognition motifs and are responsible for the nucleic acid-binding activity of the protein. The C-terminal glycine-rich fragment (residues 266–414) is intrinsically disordered in the native state, and is the major aggregation site. Whilst the function of the two RNA recognition motifs and the C-terminal fragment have been relatively well studied, only recently have the functional implications of the N-terminal domain (NTD, residues 1–105) come to light; it is essential for splicing and aggregation of the protein [12, 13]. Given the new-found importance of the NTD, elucidation of its structure is paramount to understand its role in both normal and pathological states. Chang et al. first characterized the structure and physical properties of TDP-43 NTD, and suggested that the NTD contains a well-structured domain, with the full TDP-43 protein adopting a 'beads-on-a-string' topology [9]. They further demonstrated the importance of the NTD in promoting TDP oligomerization and enhancing DNA binding. More recently, Qin et al. determined the structure of a construct consisting of the first 80 residues of the protein (NTD1-80) based on sparse experimental NMR data using CS-ROSETTA[14], and found that NTD1-80 encodes a novel ubiquitin-like fold in equilibrium with its unfolded form [10]. They speculated that this unique property may be the basis for the functional dichotomy of the TDP-43 N-terminus, as well as the ability of TDP-43 to form both reversible high-order structures and irreversible inclusions that are relevant to physiological and pathological processes, respectively. In this issue of The FEBS Journal, Mompeán et al. describe the use of NMR and proton/deuterium exchange methods to elucidate the structure, stability and dynamics of the NTD of TDP-43. They found that a construct consisting of residues 1–77 (NTD1-77) adopts a stable fold that is reminiscent of the axin-1 family, leading to the speculation that the NTD of TDP-43 may act as an interaction site with other proteins [11]. They also identified a disordered fragment consisting of residues 78–102 (NTD78-102) that binds to DNA with sub-millimolar affinity. On comparing the NTD structure reported by Qin et al. with the one described by Mompeán et al., two issues come to light. First, although the two structures are quite similar overall, there are a number of inconsistencies between them, including significant differences in the overall fold and a missing β-hairpin. It is possible that the different experimental approaches to solving the NTD structure may account for some of the observed conflicts. Mompeán et al. used over 1100 experimental restraints to obtain the structure for the NTD, with a backbone RMSD of 0.67 Å. In contrast, Qin et al. used 78 experimental restraints as input into CS-ROSETTA, which applies known correlations between protein sequence, chemical shifts and protein structure to model the probable structure, and thus the reported structure is of lower resolution. As CS-ROSETTA is essentially 'trained' using proteins of known structure, its effectiveness at determining the conformations of new or unexpected structural folds, such as that of TDP-43 NTD, may not be optimal. ROSETTA-based approaches are probably better used for cases where there is at least one closely related system with a known structure that may serve as a positive control [14]. Second, Qin et al. found two sets of NMR signals in their work, characteristic of the co-existence of folded and unfolded conformers in equilibrium. The folded conformer is favored at low concentration (< 40 μm) and stabilized by nucleic acid binding. Mompeán et al., however, found that NTD1-77 adopted a single stable conformation, and that ssDNA bound to a cation-rich unstructured region in NTD78-102. Their results challenge the proposed role of the TDP-43 NTD in conformational equilibrium. A reduction in HSQC NMR resonances characteristic of a folded conformer at higher concentration (> 25 μm) for NTD1-105 was also observed by Chang et al. [9], and was explained by the potential of NTD1-105 to reversibly oligomerize. It will be interesting to examine whether the results obtained by Qin et al. may also be interpreted as due to oligomerization at higher concentration. Chang et al. also performed DNA binding studies of NTD1-105, and observed relatively weak binding (KD > 40 mm) between the NTD and ssDNA at physiological pH. This is quantitatively much lower than the KD values reported by both Qin et al. and Mompeán et al. A probable explanation for the discrepancy may lie in the pH at which the binding studies were performed. The DNA binding studies of both Qin et al. and Mompeán et al. were performed at low pH, at which the protein is more positively charged, which may enhance non-specific binding between the protein and ssDNA. Furthermore, the experimental conditions were very different between the studies performed by Mompeán et al. and Qin et al. The polyhistidine tag used for purification is located on the C-terminus of the plasmid construct in Qin et al., and the NMR experiments were performed using pure water as the sample buffer. Qin et al. claimed that use of pure water as the solvent ameliorated aggregation problems encountered during the experiments. By contrast, the construct by Mompeán et al. includes a polyhistidine tag at the N-terminus, and the NMR experiments were performed in sodium acetate buffer solution. It is possible that the location of the polyhistidine tag may have changed the aggregation properties of the protein, as reported previously [15]. The use of water versus buffer as the sample solvent may also affect the behavior and/or conformation of the protein, especially if the protein is subject to ionic effects. It is also possible that both structures are valid under different experimental conditions, and a simple check of the original data from both groups would provide clues into the reason behind the structural discrepancy. However, Qin et al. have not yet made structural coordinates or NMR data publicly available, complicating efforts to compare the results. The conflicting results reported by Mompeán et al. and Qin et al. are perplexing and significant. More work is needed to resolve the discrepancy, as it significantly complicates efforts to understand the molecular roles of the TDP-43 NTD and TDP-43 as a whole. If NTD1-77 exists in a single conformation, it remains to be shown how the structure and its topology contribute to physiological and pathological protein function. As noted by the authors, the high sequence conservation of the TDP-43 NTD throughout the animal kingdom strongly suggests that it has an important physiological role. Proteomic studies performed on TDP-43 in several cell lines have detected large number of potential TDP-43-interacting partners [16]. Dysregulation of many of these interactions may precede or follow aberrant aggregation and modification of TDP-43 in affected cells. The results of the present study will help to evaluate these potential pathways and mechanisms that ultimately control disease pathogenesis. C-k. C. and T.-h.H. wrote the commentary.
No takes yet. Share an insight, caveat, or question.
Chang et al. (2016) studied this question.