- Email: [email protected]
FTD-causing FUS
Biochemical and Biophysical Research Communications xxx (2018) 1e7
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS Jian Kang, Liangzhong Lim, Jianxing Song* Department of Biological Sciences, Faculty of Science, National University of Singapore, 10 Kent Ridge Crescent, 119260, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 August 2018 Accepted 3 September 2018 Available online xxx
ATP is the universal energy currency but mysteriously its cellular concentration is much higher than that needed for providing energy. Recently ATP was decoded to act as a hydrotrope to dissolve liquid-liquid phase separation (LLPS) of FUS whose aggregation leads to ALS/FTD. By DIC microscopy and NMR, here we characterized the effect of ATP on LLPS of FUS and its N-/C-terminal domains. Very unexpectedly, we found that like nucleic acids, ATP enhances LLPS of FUS at low but dissolves at high concentrations. Intriguingly, ATP monotonically dissolves LLPS of NTD, while it induces LLPS of CTD at low but dissolves at high concentrations. Our study reveals for the first time that ATP can enhance LLPS most likely by behaving as a bivalent binder. Most importantly, our results imply that age-dependent reduction of ATP concentrations may not only result in decreasing its capacity in preventing protein aggregation, but also in enhancing aggregation. © 2018 Elsevier Inc. All rights reserved.
Keywords: Liquid-liquid phase separation (LLPS) Fused in sarcoma (FUS) Amyotrophic lateral sclerosis (ALS) Frontotemporal dementia (FTD) Adenosine triphosphate (ATP) NMR spectroscopy
1. Introduction Living cells need subcompartments to achieve spatiotemporal regulation of various biological reactions. It has been increasingly found that in addition to classic membrane-bound organelles, there exist many membrane-less organelles capable of compartmentalizing and concentrating specific sets of molecules, which include nucleolus, Cajal bodies and nuclear speckles in the nucleoplasm, as well as stress granules, P-bodies and germ granules in the cytoplasm [1,2]. Recently it has been revealed that these membrane-less organelles are not structurally defined complexes such as the ribosome, but are dynamic macromolecular assemblies formed by weak and multivalent interactions among components. Amazingly, these structures behave as liquid droplets, which are round, dynamic and able to coalesce into a larger assembly upon contacting with one another [1,2]. Very recently, the formation of these liquid-like droplets of biomolecules within the cytoplasm or nucleoplasm has been characterized to arise from the self-assembly through a physicochemical process known as liquid-liquid phase separation (LLPS), which has been now recognized to represent a general mechanism for forming membrane-less intracellular organelles [1,2].
* Corresponding author. E-mail address: [email protected] (J. Song).
One group of proteins involved in driving LLPS is constituted by RNA-binding proteins (RBPs) composed of RNA-binding motif (RRM) and low-complexity (LC) domains, which include FUS and TDP-43 [1e4]. FUS is involved in forming cellular granules in cytoplasm including stress granules (SGs) composed of both RBPs and nucleic acids in response to environmental stresses. Intriguingly, these dynamic liquid droplets can further become “aged” or exaggerated into amyloid fibrils or inclusions that lead to a large spectrum of neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) [1e10]. FUS consists of 526-residues which is composed of three major domains (Fig. 1A): the N-terminal domain (NTD) over residues 1e267 composed of the QGSY-rich prion-like domain (PLD) and an RG-rich region (RGG1), an RRM and C-terminal domain (CTD) over 371e526 containing RGG2, a zinc finger (ZnF) and RGG3. Recently, cellular factors mediating LLPS are beginning to be identified. These factors critically mediate LLPS and also protein aggregation. Consequently they represent key targets for developing therapeutic strategies/molecules to treat a large variety of human diseases. Very strikingly, ATP, the universal energy currency in the cell, has been decoded to dissolve LLPS and aggregates of several RBPs including FUS [11]. It was proposed that ATP dissolves LLPS and aggregated proteins by acting as a hydrotropic molecule (Fig. 1B): while its aromatic adenine ring is clustered over the
https://doi.org/10.1016/j.bbrc.2018.09.014 0006-291X/© 2018 Elsevier Inc. All rights reserved.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
2
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
Fig. 1. Two-stage effect of ATP on LLPS of FUS. (A) 526-residue FUS is composed of: N-terminal low-sequence complexity (LC) region (1e267) including a QGSY-rich prion-like domain (PLD) and an RG/RGG-rich region (RGG1); RNA-recognition motif (RRM: 285e370); and C-terminal domain CTD (371e526) which contains an RG/RGG-rich region (RGG2), a zinc finger (ZnF), another RG/RGG-rich region (RGG3) carrying a nuclear localization signal (NLS). (B) Chemical structure of ATP. Inlet: a cartoon model of ATP showing its triphosphate chain, which is highly polar and negatively-charged, and relatively hydrophobic aromatic ring of adenine base linked by a ribose. (C) DIC microscopy images of liquid droplets formed by FUS in the presence of ATP at different concentrations. The videos used to output these images are provided in Supplementary Data.
aromatic or hydrophobic patches of proteins by p-p or hydrophobic interactions, the polar triphosphate chain strongly interacts with water [12,13]. Very recently, nucleic acids including RNA [14] and single-stranded DNA (ssDNA) [15] have been also found to have a two-stage effect on LLPS of several RBPs including FUS and TDP-43: enhancement of LLPS at low concentrations but dissolution at high concentrations. So far, no biophysical characterization of the effect of ATP on LLPS has been reported. Here, we attempted to understand the biophysical basis of the effect of ATP on LLPS of FUS as monitored by differential interference contrast (DIC) microscopy and NMR spectroscopy. Very unexpectedly, in addition to the dissolution of LLPS of FUS at high ATP concentrations which was previously reported [11], we found for the first time that like nucleic acids, ATP could in fact enhance LLPS of the full-length FUS at low concentrations. To understand this observation, we further assessed the effect of ATP on LLPS of the dissected FUS NTD and CTD. Remarkably, ATP monotonically dissolves LLPS of the FUS NTD. On the other hand, ATP is able to induce LLPS of the FUS CTD at low concentrations but dissolves at high concentrations. The results together allow the proposal of a speculative model for ATP to mediate LLPS of FUS, in which ATP acts more than just as a hydrotrope, but also behaves as a bivalent binder. Most importantly, our results imply that agedependent reduction of ATP concentrations may not only just result in decreasing its capacity in preventing protein aggregation, but also in enhancing aggregation at low ATP concentrations. This thus holds immediate implications in understanding the molecular mechanisms as well as further design of therapeutic molecules for various human diseases caused by protein aggregation such as neurodegenerative diseases and cardiovascular diseases [13]. 2. Materials and methods 2.1. Preparation of recombinant FUS proteins Previously, we have cloned DNA fragments encoding FUS and its differentially-dissected fragments into a modified vector pET28a
with a C-terminal His-tag [16]. However, although the presence of 6xHis-tag had no detectable effect on their solution conformations, we found that the His-tag has in general weakened the capacity in LLPS, consistent with the recent report [18]. Therefore, in the present study, we removed the His-tag by adding a stop codon immediately after the DNA sequences encoding the full-length FUS, FUS NTD over 1e267 and FUS CTD over 371e526 (Fig. 1A). The expression and purification of these recombinant FUS proteins followed the previous protocols [16]. The purity of the recombinant proteins was checked by SDS-PAGE gels, and the molecular weights were verified by a Voyager STR matrix-assisted laser desorption ionization time-of-flight-mass spectrometer (Applied Biosystems). The same procedures were used to generate isotope-labeled proteins for NMR studies except that the bacteria were grown in M9 medium with addition of (15NH4)2SO4 for 15N-labeling. The protein concentration was determined by the UV spectroscopic method in the presence of 8 M urea, under which the extinct coefficient at 280 nm of a protein can be calculated by adding up the contribution of Trp, Tyr, and Cys residues [10,15,16]. ATP was purchased from SigmaAldrich with the same catalog number as previously reported [11]. MgCl2 was added into ATP for stabilization by forming the ATP-Mg complex [11]. The protein and ATP samples were all prepared in 5 mM sodium phosphate buffer with the final solution pH at 6.0. For all samples studied here by DIC and NMR, ZnCl2 was also added to the samples of the full-length FUS and CTD containing the zinc finger (ZnF) to reach a final concentration of 4 mM. 2.2. Differential interference contrast (DIC) microscopy The formation of liquid droplets was imaged at 25 C on 50 ml of different FUS samples at a protein concentration of 20 mM in 5 mM sodium phosphate buffer at pH 6.0 in the presence of ATP at different molar concentrations: 0, 0.2, 0.6, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 mM by DIC microscopy (OLYMPUS IX73 Inverted Microscope System with OLYMPUS DP74 Color Camera) as previously described [15].
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
2.3. NMR characterizations All NMR experiments were acquired at 25 C on an 800 MHz Bruker Avance spectrometer equipped with pulse field gradient units and a shielded cryoprobe as described previously [10,15,16]. To have the enhancing effect of the cryoprobe for NMR signal sensitivity, which is essential for NMR HSQC titration experiments at a very low protein concentration (20 mM), NMR samples had to be prepared in a low-salt buffer: 5 mM sodium phosphate buffer, while pH value was optimized to 6.0 as many HSQC peaks of highly disordered NTD and CTD disappeared at higher pH values [16]. For NMR titration studies of the interactions between FUS/domains and ATP, two dimensional 1H-15N NMR HSQC spectra were collected on 15N-labeled FUS samples at a protein concentration of 20 mM in 5 mM sodium phosphate buffer (pH 6.0) at 25 C in the presence of ATP at different molar concentrations: 0, 0.2, 0.6, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 mM.E. 3. Results 3.1. ATP has a two-stage effect on LLPS of FUS To gain insights into the biophysical basis for ATP-induced dissolution of LLPS of FUS previously reported [11], here we characterized the effect of ATP on LLPS of FUS by DIC microscopy. FUS without ATP could undergo LLPS to form dynamic and spherical liquid droplets. However, only a small number of droplets were observed with the maximal diameter of ~3.0 mm (I of Fig. 1C and Video S1CI). Interestingly, addition of ATP to 0.2 mM slightly enhanced LLPS by mainly increasing the size of droplets with the maximal diameter of ~3.8 mm (II of Fig. 1C and Video S1CII). At 1.0 mM, ATP significantly enhanced LLPS by increasing both number and size of droplets with the maximal diameter reaching up to ~12.1 mm (III of Fig. 1C and Video S1CIII). Further addition of ATP at concentrations >2 mM started to reduce the number and size of droplets, and at 4 mM, ATP triggered the reduction of the size of droplets with the maximal diameter of ~9.1 mm (IV of Fig. 1C and Video S1CIV). At ATP concentrations 8 mM, the droplets were completely dissolved. Supplementary video related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. Our current results revealed that ATP at high concentrations could indeed dissolve LLPS of FUS exactly as previously reported [11]. Nevertheless, for the first time our study decoded a previously-unknown phenomenon that like nucleic acids [14,15], ATP can also significantly enhance LLPS of FUS at low concentrations (1e2 mM). Therefore, we decided to better understand this observation by further characterizing the effect of ATP on LLPS of the FUS NTD and CTD respectively. 3.2. ATP monotonically dissolves LLPS of the FUS NTD Here we used the entire FUS NTD over residues 1e267 also containing RRG1 additional to PLD whose LLPS has been extensively characterized by NMR [18] (Fig. 1A). The FUS NTD could undergo LLPS to form many droplets with the maximal diameter up to ~7.2 mm (I of Fig. 2A and Video S2AI). Interestingly, addition of ATP led to monotonic dissolution of LLPS of NTD. Even at 0.2 mM, ATP could significantly reduce both number and size of droplets with the maximal diameter reduced to only ~3.8 mm (II of Fig. 2A and Video S2AII). At 1.0 mM, ATP further dissolved LLPS with the maximal diameter of droplets to ~3.2 mm (III of Fig. 2A and Video S2AIII). At 4.0 mM, ATP largely dissolved LLPS with the maximal diameter less than ~1.0 mm for all droplets (IV of Fig. 2A and Video
3
S2AIV). Upon further increasing ATP concentrations 6.0 mM, all the droplets were dissolved. Supplementary video related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. To gain more detailed view of the interaction between the FUS NTD and ATP, we assessed the dissolution of its liquid droplets by NMR HSQC titrations under exactly the same condition as used for DIC imaging (Fig. 2A). Due to being intrinsically disordered, the FUS NTD has an HSQC spectrum of narrow dispersions on both 1H and 15 N dimensions (Fig. 2B) [16]. Interestingly, at a molar concentration of 0.2 mM, ATP induced no significant shift of HSQC peaks, but only the intensity reduction of almost all HSQC peaks (I of Fig. 2B). Further addition of ATP also only led to more significant reduction of the peak intensity and at an ATP concentration of 10 mM, the intensity of most HSQC peaks is too weak to be detected (IV of Fig. 2B). Therefore, the obtained NMR results clearly indicate that ATP appears to be able to interact with the majority, if not all, of the NTD residues, thus implying that no significant specificity exists for the interaction. Furthermore, at high ATP concentrations, the FUS NTD appears to interact with a large number of ATP molecules, which thus provokes significant ms-ms dynamics or/and results in large molecular weight for the ATP-NTD complexes, thus leading to significant broadening of NMR resonance peaks. Indeed, the FUS NTD contains 29 Tyr and 2 Phe residues, as well as 9 Arg and 1 Lys residues. Therefore, the aromatic ring of adenine of ATP may interact with aromatic rings of Tyr and Phe of the FUS NTD by p-p interaction, or/and interact with the side chains of Arg and Lys by pcation interaction. Furthermore, the triphosphate chain of ATP may also interact the side chains of Arg and Lys by electrostatic interaction [12,13,17]. 3.3. ATP induces LLPS of the FUS CTD at low concentrations but dissolves at high concentrations Recently we have assessed LLPS of the FUS CTD by DIC imaging under a variety of conditions. The results indicate that the FUS CTD alone was incapable of undergoing LLPS at protein concentrations even up to 500 mM in different buffers and at various pH values, with NaCl or phosphate concentrations up to 150 mM. Strikingly, however, even at 0.2 mM, ATP was able to induce LLPS of the FUS CTD by forming some small droplets with the maximal diameter of ~2.2 mm (I of Fig. 3A and Video S3AI). At 1.0 mM, ATP induced the formation of many large droplets with the maximal diameter of ~8.1 mm (II of Fig. 3A and Video S3AII). However, at 4.0 mM, ATP started to dissolve LLPS of the FUS CTD by reducing both number and size of droplets with the maximal diameter reduced to ~6.0 mm (III of Fig. 3A and Video S3AIII). At ATP concentrations 8 mM, all droplets were dissolved. Supplementary video related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. We also assessed the ATP-mediated induction and dissolution of LLPS of the FUS CTD by HSQC titrations. As shown in Fig. 3B, in the HSQC spectrum of the FUS CTD in the free state, some welldispersed HSQC peaks are from the residues of the well-folded ZnF, while most narrowly-dispersed peaks are from the intrinsically disordered RGG2 and RGG3 regions [16]. At 0.2 mM, ATP induced no significant shift of HSQC peaks, but only the intensity reduction of most HSQC peaks (I of Fig. 3B). Further addition of ATP to 1.0 mM, at which LLPS was significantly induced (II of Fig. 3A), led to more significant reduction of the peak intensity as well as shifts of some HSQC peaks (II of Fig. 3B). Intriguingly, addition of ATP at high concentrations triggered further reduction of the peak intensity and at an ATP concentration of 10 mM, the intensity of most HSQC peaks became too weak to be detected (IV of Fig. 3B). The NMR results here suggest that ATP can also extensively interact
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
4
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
Fig. 2. DIC and NMR characterization of the effect of ATP on LLPS of the FUS NTD. (A) DIC images of liquid droplets formed by the FUS NTD in the presence of ATP at different concentrations. The videos used to output these images are provided in Supplementary Data. (B) 1H-15N HSQC spectra of the 15N-labeled FUS NTD in the free state (blue) and in the presence of ATP at different concentrations (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
with the CTD residues, and these interactions lead to both induction of LLPS of the FUS CTD at low ATP concentrations and dissolution at high ATP concentrations. Examination of the CTD sequence revealed that it contains 5 Tyr, 5 Phe and 2 Trp residues as well as 25 Arg and 4 Lys residues. Therefore, likely due to the relatively small number of aromatic residues as compared to that of the FUS NTD, the FUS CTD alone is not able to undergo LLPS and thus exists as a homogeneous solution (I of Fig. 3C). However, as ATP is composed of an aromatic ring of adenine and a polar and negatively-charged triphosphate chain linked by a ribose. It not only can behave as a hydrotropic molecule to dissolve LLPS and aggregated proteins [11e13], it can also act as a bivalent binder for the FUS CTD. On the one hand, the adenine ring may bind to aromatic sidechains of CTD by p-p interaction, or/and interact with the side chains of Arg and Lys by p-cation interaction. On the other hand, its triphosphate chain may bind the side chains of Arg and Lys by electrostatic interaction [12,13,17]. As a result, at ATP concentration of ~ 1e2 mM, the bivalent binding of ATP to the multiple sites of the FUS CTD is expected to form a large but dynamic ATP-CTD complex, which thus manifesting as the formation of liquid droplets (II of Fig. 3C). However, in the exceeding presence of ATP, further binding of ATP to the FUS CTD will lead to the disruption of the dynamic structure, thus manifesting as dissolution of LLPS by ATP at high concentrations.
3.4. NMR characterization of ATP-induced enhancement and dissolution of LLPS of FUS In parallel to DIC imaging (Fig. 1), we also assessed the ATPinduced enhancement and dissolution of LLPS of the full-length FUS by NMR HSQC titrations. As we previously characterized [16], in the HSQC spectra, the well-dispersed HSQC peaks are from both well-folded RRM and ZnF, while the narrowly-dispersed peaks are from PLD, RGG1, RGG2 and RGG3 (Fig. 4A). Strikingly, upon gradual increase of ATP concentrations, the intensity of HSQC peaks of the full-length FUS also gradually decreased, similar to what was observed on the FUS CTD (Fig. 3B). At 10.0 mM, most HSQC peaks became too broad to be detected (IV of Fig. 4A). In light of the recent mutagenesis study to define the driving force for LLPS of FUS [17], our current results together lead to the proposal of a speculative model to rationalize the ATP-induced enhancement and dissolution of LLPS of FUS (Fig. 4B). Briefly, FUS molecules without ATP can undergo LLPS mainly driven by dynamic and multivalent inter-molecular p-cation interactions between aromatic residues within PLD and Arg/Lys residues within RGG regions (I of Fig. 4B). Upon addition of ATP into FUS, in addition to behaving as a hydrotrope, ATP can also act as a bivalent binder: its aromatic ring of adenine is able to dynamically interact with aromatic, or/and Arg/Ly residues, while its triphosphate chain can also dynamically interact with Arg/Ly residues of FUS. As a
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
5
Fig. 3. DIC and NMR characterization of the effect of ATP on LLPS of the FUS CTD. (A) DIC images of liquid droplets formed by the FUS CTD in the presence of ATP at different concentrations. The videos used to output these images are provided in Supplementary Data. (B) 1H-15N HSQC spectra of the 15N-labeled FUS CTD in the free state (blue) and in the presence of ATP at different concentrations (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
consequence, a large but dynamic ATP-FUS complex is formed, which manifests as enhancement of LLPS of FUS (II of Fig. 4B). However, upon further addition of ATP, the exceeding binding of ATP to FUS will lead to the disruption of the large and dynamic ATPFUS complex (III of Fig. 4B). Consequently LLPS of FUS will be dissolved at high ATP concentrations. As one FUS molecule is bound with a large number of ATP molecules, most HSQC peaks of FUS becomes too broad to be detected, due to the shortening of the transverse relaxation time T2, which most likely results from ms-ms dynamics or/and large molecular weight of the ATP-bound FUS.
4. Discussion LLPS is now recognized not only to underlie a variety of fundamental physiological processes, but its mis-regulation also leads to an increasing spectrum of human diseases such as
neurodegenerative diseases. Very recently, cellular factors including ATP and nucleic acids have been identified to critically regulate LLPS. In particular, identification of their general and key roles in LLPS and protein aggregation has explained two longstanding mysteries: 1) why cells need to maintain very high concentrations of ATP; and 2) RBPs such as FUS and TDP-43 are largely soluble in the nucleus only with a small portion undergoing LLPS, but become severely aggregated upon mislocalization to the cytoplasm, thus leading to ALS. However, so far their biophysical basis remains to be explored and particularly one critical question needs to be unanswered: whether the mediation by ATP and nucleic acids results from the buffering effect, or from specific binding. Such knowledge is not only essential for our understanding of the fundamental cell physiology, but also provides key foundations for developing therapeutic strategies/molecules to treat human diseases characteristic of protein aggregation.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
6
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
Fig. 4. NMR characterization of the effect of ATP on LLPS of FUS. (A) 1H-15N HSQC spectra of the 15N-labeled FUS in the free state (blue) and in the presence of ATP at different concentrations (red). (B) A speculative model for ATP-induced enhancement and dissolution of LLPS of FUS. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Here we aimed to gain the biophysical basis for ATP to dissolve LLPS by DIC and NMR characterization of LLPS of FUS in the presence of ATP at concentrations ranging from 0 to 10 mM. To our surprise, in addition to the dissolution of LLPS by ATP at high concentrations (6 mM) exactly as previously reported [11], we found for the first time that ATP can also significantly enhance LLPS of FUS at low concentrations around 1e2 mM. Further assessment of the effect of ATP on LLPS of the FUS NTD and CTD reveals that ATP monotonically dissolves LLPS of the FUS NTD rich in aromatic residues, likely by mainly acting as a hydrotropic molecule as previously proposed [11]. By a sharp contrast, ATP can completely induce LLPS of the FUS CTD rich in Arg/Lys residues at low concentrations (1e2 mM), followed by further dissolution of LLPS at high ATP concentrations (6 mM). Consequently, the induction of LLPS by ATP cannot be rationalized by the hydrotropic properties of ATP. Here we propose that ATP enhances LLPS of FUS most likely by acting as a bivalent binder: its aromatic ring of adenine can bind aromatic residues of FUS by p-p interaction, or Arg/Lys by p-cation interaction; while its triphosphate chain also bind Arg/Lys by electrostatic interaction. Consequently in the presence of ATP at relatively low concentrations, a large and dynamic ATP-FUS can be formed which manifests as liquid droplets (Fig. 4B). Therefore ATP appears to own the dual ability to mediate LLPS of FUS by acting as a hydrotrope and a bivalent binder. As such the effect of ATP on LLPS is not only dependent on ATP concentrations, but also on the protein sequence. Furthermore, our NMR characterization reveals that ATP is able to extensively interact with most, if not all, residues of FUS. This suggests that the interaction between ATP and FUS is not as specific as other specific protein-ligand interactions. On the other hand, the NMR results also reveal that a large number of ATP molecules are dynamically bound/clustered over FUS molecules (III of Fig. 4B). Therefore, the two-stage effect of ATP on LLPS of FUS appears not to result from simple buffering effects as buffer ions are not supposed to extensively interact with proteins. Our finding that at low concentrations, ATP can enhance LLPS of FUS may bear critical implications in understanding age-dependent aggregation of proteins. Although at present the data of ATP concentrations in cells of different type and age are largely incomplete, it is usually thought that in most cells, ATP concentrations can
range from 6 to 10 mM [11e13]. However, it has been also reported that upon aging, ATP concentrations will keep reducing in cells. It is thus possible that in some aged cells, its ATP concentration becomes very low, at which ATP acts to enhance LLPS of proteins. Therefore, our finding implies that the age-dependent reduction of ATP concentrations not only can decrease its capacity in preventing protein aggregation, but may also act to enhance aggregation. Acknowledgement This study is supported by Ministry of Education of Singapore (MOE) Tier 2 Grant MOE2015-T2-1-111 to Jianxing Song. Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.bbrc.2018.09.014. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. References [1] S.F. Banani, H.O. Lee, A.A. Hyman, et al., Biomolecular condensates: organizers of cellular biochemistry, Nat. Rev. Mol. Cell Biol. 18 (2017) 285e298. [2] Y. Shin, C.P. Brangwynne, Liquid phase condensation in cell physiology and disease, Science 357 (2017) 6357. [3] S.C. Ling, M. Polymenidou, D.W. Cleveland, Converging mechanisms in ALS and FTD: disrupted RNA and protein homeostasis, Neuron 79 (2013) 416e438. [4] Y.R. Li, O.D. King, J. Shorter, et al., Stress granules as crucibles of ALS pathogenesis, J. Cell Biol. 201 (2013) 361e372. [5] M. Neumann, S. Roeber, H.A. Kretzschmar, et al., Abundant FUSimmunoreactive pathology in neuronal intermediate filament inclusion disease, Acta Neuropathol. 118 (2009) 605e616. [6] T.W. Han, M. Kato, S. Xie, et al., Cell-free formation of RNA granules: bound RNAs identify features and components of cellular assemblies, Cell 149 (2012) 768e779. [7] A. Patel, H.O. Lee, L. Jawerth, et al., A liquid-to-solid phase transition of the ALS protein FUS accelerated by disease mutation, Cell 162 (2015) 1066e1077. [8] T. Murakami, S. Qamar, J.Q. Lin, et al., ALS/FTD mutation-induced phase transition of FUS liquid droplets and reversible hydrogels into irreversible hydrogels impairs RNP granule function, Neuron 88 (2015) 678e690.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7 [9] A. Molliex, J. Temirov, J. Lee, et al., Phase separation by low complexity domains promotes stress granule assembly and drives pathological fibrillization, Cell 163 (2015) 123e133. [10] L. Lim, Y. Wei, Y. Lu, et al., ALS-causing mutations significantly perturb the self-assembly and interaction with nucleic acid of the intrinsically disordered prion-like domain of TDP-43, PLoS Biol. 14 (2016), e1002338. [11] A. Patel, L. Malinovska, S. Saha, et al., ATP as a biological hydrotrope, Science 356 (2017) 753e756. [12] A.M. Rice, M.K. Rosen, ATP controls the crowd, Science 356 (2017) 701e702. [13] T. Meyer, N. Voigt, In search for novel functions of adenosine 5'-triphosphate (ATP) in the heart, Cardiovasc. Res. 113 (2017) e59ee60. [14] S. Maharana, J. Wang, D.K. Papadopoulos, et al., RNA buffers the phase separation behavior of prion-like RNA binding proteins, Science (2018), https://
7
doi.org/10.1126/science.aar7366. Apr 12. pii: eaar7366. [15] L. Wang, J. Kang, L. Lim, et al., TDP-43 NTD can be induced while CTD is significantly enhanced by ssDNA to undergo liquid-liquid phase separation, Biochem. Biophys. Res. Commun. 499 (2018) 189e195. [16] Y. Lu, L. Lim, J. Song, RRM domain of ALS/FTD-causing FUS characteristic of irreversible unfolding spontaneously self-assembles into amyloid fibrils, Sci. Rep. 7 (2017) 1043. [17] J. Wang, J.M. Choi, A.S. Holehouse, et al., A Molecular Grammar Underlying the Driving Forces for Phase Separation of Prion-like RNA Binding Proteins, Cell, 2018 accepted (online publication date June 28). [18] K.A. Burke, A.M. Janke, C.L. Rhine, et al., Residue-by-Residue view of in vitro FUS granules that bind the C-terminal domain of RNA polymerase II, Mol. Cell 60 (2015) 231e241.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS Jian Kang, Liangzhong Lim, Jianxing Song* Department of Biological Sciences, Faculty of Science, National University of Singapore, 10 Kent Ridge Crescent, 119260, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 August 2018 Accepted 3 September 2018 Available online xxx
ATP is the universal energy currency but mysteriously its cellular concentration is much higher than that needed for providing energy. Recently ATP was decoded to act as a hydrotrope to dissolve liquid-liquid phase separation (LLPS) of FUS whose aggregation leads to ALS/FTD. By DIC microscopy and NMR, here we characterized the effect of ATP on LLPS of FUS and its N-/C-terminal domains. Very unexpectedly, we found that like nucleic acids, ATP enhances LLPS of FUS at low but dissolves at high concentrations. Intriguingly, ATP monotonically dissolves LLPS of NTD, while it induces LLPS of CTD at low but dissolves at high concentrations. Our study reveals for the first time that ATP can enhance LLPS most likely by behaving as a bivalent binder. Most importantly, our results imply that age-dependent reduction of ATP concentrations may not only result in decreasing its capacity in preventing protein aggregation, but also in enhancing aggregation. © 2018 Elsevier Inc. All rights reserved.
Keywords: Liquid-liquid phase separation (LLPS) Fused in sarcoma (FUS) Amyotrophic lateral sclerosis (ALS) Frontotemporal dementia (FTD) Adenosine triphosphate (ATP) NMR spectroscopy
1. Introduction Living cells need subcompartments to achieve spatiotemporal regulation of various biological reactions. It has been increasingly found that in addition to classic membrane-bound organelles, there exist many membrane-less organelles capable of compartmentalizing and concentrating specific sets of molecules, which include nucleolus, Cajal bodies and nuclear speckles in the nucleoplasm, as well as stress granules, P-bodies and germ granules in the cytoplasm [1,2]. Recently it has been revealed that these membrane-less organelles are not structurally defined complexes such as the ribosome, but are dynamic macromolecular assemblies formed by weak and multivalent interactions among components. Amazingly, these structures behave as liquid droplets, which are round, dynamic and able to coalesce into a larger assembly upon contacting with one another [1,2]. Very recently, the formation of these liquid-like droplets of biomolecules within the cytoplasm or nucleoplasm has been characterized to arise from the self-assembly through a physicochemical process known as liquid-liquid phase separation (LLPS), which has been now recognized to represent a general mechanism for forming membrane-less intracellular organelles [1,2].
* Corresponding author. E-mail address: [email protected] (J. Song).
One group of proteins involved in driving LLPS is constituted by RNA-binding proteins (RBPs) composed of RNA-binding motif (RRM) and low-complexity (LC) domains, which include FUS and TDP-43 [1e4]. FUS is involved in forming cellular granules in cytoplasm including stress granules (SGs) composed of both RBPs and nucleic acids in response to environmental stresses. Intriguingly, these dynamic liquid droplets can further become “aged” or exaggerated into amyloid fibrils or inclusions that lead to a large spectrum of neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) [1e10]. FUS consists of 526-residues which is composed of three major domains (Fig. 1A): the N-terminal domain (NTD) over residues 1e267 composed of the QGSY-rich prion-like domain (PLD) and an RG-rich region (RGG1), an RRM and C-terminal domain (CTD) over 371e526 containing RGG2, a zinc finger (ZnF) and RGG3. Recently, cellular factors mediating LLPS are beginning to be identified. These factors critically mediate LLPS and also protein aggregation. Consequently they represent key targets for developing therapeutic strategies/molecules to treat a large variety of human diseases. Very strikingly, ATP, the universal energy currency in the cell, has been decoded to dissolve LLPS and aggregates of several RBPs including FUS [11]. It was proposed that ATP dissolves LLPS and aggregated proteins by acting as a hydrotropic molecule (Fig. 1B): while its aromatic adenine ring is clustered over the
https://doi.org/10.1016/j.bbrc.2018.09.014 0006-291X/© 2018 Elsevier Inc. All rights reserved.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
2
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
Fig. 1. Two-stage effect of ATP on LLPS of FUS. (A) 526-residue FUS is composed of: N-terminal low-sequence complexity (LC) region (1e267) including a QGSY-rich prion-like domain (PLD) and an RG/RGG-rich region (RGG1); RNA-recognition motif (RRM: 285e370); and C-terminal domain CTD (371e526) which contains an RG/RGG-rich region (RGG2), a zinc finger (ZnF), another RG/RGG-rich region (RGG3) carrying a nuclear localization signal (NLS). (B) Chemical structure of ATP. Inlet: a cartoon model of ATP showing its triphosphate chain, which is highly polar and negatively-charged, and relatively hydrophobic aromatic ring of adenine base linked by a ribose. (C) DIC microscopy images of liquid droplets formed by FUS in the presence of ATP at different concentrations. The videos used to output these images are provided in Supplementary Data.
aromatic or hydrophobic patches of proteins by p-p or hydrophobic interactions, the polar triphosphate chain strongly interacts with water [12,13]. Very recently, nucleic acids including RNA [14] and single-stranded DNA (ssDNA) [15] have been also found to have a two-stage effect on LLPS of several RBPs including FUS and TDP-43: enhancement of LLPS at low concentrations but dissolution at high concentrations. So far, no biophysical characterization of the effect of ATP on LLPS has been reported. Here, we attempted to understand the biophysical basis of the effect of ATP on LLPS of FUS as monitored by differential interference contrast (DIC) microscopy and NMR spectroscopy. Very unexpectedly, in addition to the dissolution of LLPS of FUS at high ATP concentrations which was previously reported [11], we found for the first time that like nucleic acids, ATP could in fact enhance LLPS of the full-length FUS at low concentrations. To understand this observation, we further assessed the effect of ATP on LLPS of the dissected FUS NTD and CTD. Remarkably, ATP monotonically dissolves LLPS of the FUS NTD. On the other hand, ATP is able to induce LLPS of the FUS CTD at low concentrations but dissolves at high concentrations. The results together allow the proposal of a speculative model for ATP to mediate LLPS of FUS, in which ATP acts more than just as a hydrotrope, but also behaves as a bivalent binder. Most importantly, our results imply that agedependent reduction of ATP concentrations may not only just result in decreasing its capacity in preventing protein aggregation, but also in enhancing aggregation at low ATP concentrations. This thus holds immediate implications in understanding the molecular mechanisms as well as further design of therapeutic molecules for various human diseases caused by protein aggregation such as neurodegenerative diseases and cardiovascular diseases [13]. 2. Materials and methods 2.1. Preparation of recombinant FUS proteins Previously, we have cloned DNA fragments encoding FUS and its differentially-dissected fragments into a modified vector pET28a
with a C-terminal His-tag [16]. However, although the presence of 6xHis-tag had no detectable effect on their solution conformations, we found that the His-tag has in general weakened the capacity in LLPS, consistent with the recent report [18]. Therefore, in the present study, we removed the His-tag by adding a stop codon immediately after the DNA sequences encoding the full-length FUS, FUS NTD over 1e267 and FUS CTD over 371e526 (Fig. 1A). The expression and purification of these recombinant FUS proteins followed the previous protocols [16]. The purity of the recombinant proteins was checked by SDS-PAGE gels, and the molecular weights were verified by a Voyager STR matrix-assisted laser desorption ionization time-of-flight-mass spectrometer (Applied Biosystems). The same procedures were used to generate isotope-labeled proteins for NMR studies except that the bacteria were grown in M9 medium with addition of (15NH4)2SO4 for 15N-labeling. The protein concentration was determined by the UV spectroscopic method in the presence of 8 M urea, under which the extinct coefficient at 280 nm of a protein can be calculated by adding up the contribution of Trp, Tyr, and Cys residues [10,15,16]. ATP was purchased from SigmaAldrich with the same catalog number as previously reported [11]. MgCl2 was added into ATP for stabilization by forming the ATP-Mg complex [11]. The protein and ATP samples were all prepared in 5 mM sodium phosphate buffer with the final solution pH at 6.0. For all samples studied here by DIC and NMR, ZnCl2 was also added to the samples of the full-length FUS and CTD containing the zinc finger (ZnF) to reach a final concentration of 4 mM. 2.2. Differential interference contrast (DIC) microscopy The formation of liquid droplets was imaged at 25 C on 50 ml of different FUS samples at a protein concentration of 20 mM in 5 mM sodium phosphate buffer at pH 6.0 in the presence of ATP at different molar concentrations: 0, 0.2, 0.6, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 mM by DIC microscopy (OLYMPUS IX73 Inverted Microscope System with OLYMPUS DP74 Color Camera) as previously described [15].
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
2.3. NMR characterizations All NMR experiments were acquired at 25 C on an 800 MHz Bruker Avance spectrometer equipped with pulse field gradient units and a shielded cryoprobe as described previously [10,15,16]. To have the enhancing effect of the cryoprobe for NMR signal sensitivity, which is essential for NMR HSQC titration experiments at a very low protein concentration (20 mM), NMR samples had to be prepared in a low-salt buffer: 5 mM sodium phosphate buffer, while pH value was optimized to 6.0 as many HSQC peaks of highly disordered NTD and CTD disappeared at higher pH values [16]. For NMR titration studies of the interactions between FUS/domains and ATP, two dimensional 1H-15N NMR HSQC spectra were collected on 15N-labeled FUS samples at a protein concentration of 20 mM in 5 mM sodium phosphate buffer (pH 6.0) at 25 C in the presence of ATP at different molar concentrations: 0, 0.2, 0.6, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 mM.E. 3. Results 3.1. ATP has a two-stage effect on LLPS of FUS To gain insights into the biophysical basis for ATP-induced dissolution of LLPS of FUS previously reported [11], here we characterized the effect of ATP on LLPS of FUS by DIC microscopy. FUS without ATP could undergo LLPS to form dynamic and spherical liquid droplets. However, only a small number of droplets were observed with the maximal diameter of ~3.0 mm (I of Fig. 1C and Video S1CI). Interestingly, addition of ATP to 0.2 mM slightly enhanced LLPS by mainly increasing the size of droplets with the maximal diameter of ~3.8 mm (II of Fig. 1C and Video S1CII). At 1.0 mM, ATP significantly enhanced LLPS by increasing both number and size of droplets with the maximal diameter reaching up to ~12.1 mm (III of Fig. 1C and Video S1CIII). Further addition of ATP at concentrations >2 mM started to reduce the number and size of droplets, and at 4 mM, ATP triggered the reduction of the size of droplets with the maximal diameter of ~9.1 mm (IV of Fig. 1C and Video S1CIV). At ATP concentrations 8 mM, the droplets were completely dissolved. Supplementary video related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. Our current results revealed that ATP at high concentrations could indeed dissolve LLPS of FUS exactly as previously reported [11]. Nevertheless, for the first time our study decoded a previously-unknown phenomenon that like nucleic acids [14,15], ATP can also significantly enhance LLPS of FUS at low concentrations (1e2 mM). Therefore, we decided to better understand this observation by further characterizing the effect of ATP on LLPS of the FUS NTD and CTD respectively. 3.2. ATP monotonically dissolves LLPS of the FUS NTD Here we used the entire FUS NTD over residues 1e267 also containing RRG1 additional to PLD whose LLPS has been extensively characterized by NMR [18] (Fig. 1A). The FUS NTD could undergo LLPS to form many droplets with the maximal diameter up to ~7.2 mm (I of Fig. 2A and Video S2AI). Interestingly, addition of ATP led to monotonic dissolution of LLPS of NTD. Even at 0.2 mM, ATP could significantly reduce both number and size of droplets with the maximal diameter reduced to only ~3.8 mm (II of Fig. 2A and Video S2AII). At 1.0 mM, ATP further dissolved LLPS with the maximal diameter of droplets to ~3.2 mm (III of Fig. 2A and Video S2AIII). At 4.0 mM, ATP largely dissolved LLPS with the maximal diameter less than ~1.0 mm for all droplets (IV of Fig. 2A and Video
3
S2AIV). Upon further increasing ATP concentrations 6.0 mM, all the droplets were dissolved. Supplementary video related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. To gain more detailed view of the interaction between the FUS NTD and ATP, we assessed the dissolution of its liquid droplets by NMR HSQC titrations under exactly the same condition as used for DIC imaging (Fig. 2A). Due to being intrinsically disordered, the FUS NTD has an HSQC spectrum of narrow dispersions on both 1H and 15 N dimensions (Fig. 2B) [16]. Interestingly, at a molar concentration of 0.2 mM, ATP induced no significant shift of HSQC peaks, but only the intensity reduction of almost all HSQC peaks (I of Fig. 2B). Further addition of ATP also only led to more significant reduction of the peak intensity and at an ATP concentration of 10 mM, the intensity of most HSQC peaks is too weak to be detected (IV of Fig. 2B). Therefore, the obtained NMR results clearly indicate that ATP appears to be able to interact with the majority, if not all, of the NTD residues, thus implying that no significant specificity exists for the interaction. Furthermore, at high ATP concentrations, the FUS NTD appears to interact with a large number of ATP molecules, which thus provokes significant ms-ms dynamics or/and results in large molecular weight for the ATP-NTD complexes, thus leading to significant broadening of NMR resonance peaks. Indeed, the FUS NTD contains 29 Tyr and 2 Phe residues, as well as 9 Arg and 1 Lys residues. Therefore, the aromatic ring of adenine of ATP may interact with aromatic rings of Tyr and Phe of the FUS NTD by p-p interaction, or/and interact with the side chains of Arg and Lys by pcation interaction. Furthermore, the triphosphate chain of ATP may also interact the side chains of Arg and Lys by electrostatic interaction [12,13,17]. 3.3. ATP induces LLPS of the FUS CTD at low concentrations but dissolves at high concentrations Recently we have assessed LLPS of the FUS CTD by DIC imaging under a variety of conditions. The results indicate that the FUS CTD alone was incapable of undergoing LLPS at protein concentrations even up to 500 mM in different buffers and at various pH values, with NaCl or phosphate concentrations up to 150 mM. Strikingly, however, even at 0.2 mM, ATP was able to induce LLPS of the FUS CTD by forming some small droplets with the maximal diameter of ~2.2 mm (I of Fig. 3A and Video S3AI). At 1.0 mM, ATP induced the formation of many large droplets with the maximal diameter of ~8.1 mm (II of Fig. 3A and Video S3AII). However, at 4.0 mM, ATP started to dissolve LLPS of the FUS CTD by reducing both number and size of droplets with the maximal diameter reduced to ~6.0 mm (III of Fig. 3A and Video S3AIII). At ATP concentrations 8 mM, all droplets were dissolved. Supplementary video related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. We also assessed the ATP-mediated induction and dissolution of LLPS of the FUS CTD by HSQC titrations. As shown in Fig. 3B, in the HSQC spectrum of the FUS CTD in the free state, some welldispersed HSQC peaks are from the residues of the well-folded ZnF, while most narrowly-dispersed peaks are from the intrinsically disordered RGG2 and RGG3 regions [16]. At 0.2 mM, ATP induced no significant shift of HSQC peaks, but only the intensity reduction of most HSQC peaks (I of Fig. 3B). Further addition of ATP to 1.0 mM, at which LLPS was significantly induced (II of Fig. 3A), led to more significant reduction of the peak intensity as well as shifts of some HSQC peaks (II of Fig. 3B). Intriguingly, addition of ATP at high concentrations triggered further reduction of the peak intensity and at an ATP concentration of 10 mM, the intensity of most HSQC peaks became too weak to be detected (IV of Fig. 3B). The NMR results here suggest that ATP can also extensively interact
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
4
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
Fig. 2. DIC and NMR characterization of the effect of ATP on LLPS of the FUS NTD. (A) DIC images of liquid droplets formed by the FUS NTD in the presence of ATP at different concentrations. The videos used to output these images are provided in Supplementary Data. (B) 1H-15N HSQC spectra of the 15N-labeled FUS NTD in the free state (blue) and in the presence of ATP at different concentrations (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
with the CTD residues, and these interactions lead to both induction of LLPS of the FUS CTD at low ATP concentrations and dissolution at high ATP concentrations. Examination of the CTD sequence revealed that it contains 5 Tyr, 5 Phe and 2 Trp residues as well as 25 Arg and 4 Lys residues. Therefore, likely due to the relatively small number of aromatic residues as compared to that of the FUS NTD, the FUS CTD alone is not able to undergo LLPS and thus exists as a homogeneous solution (I of Fig. 3C). However, as ATP is composed of an aromatic ring of adenine and a polar and negatively-charged triphosphate chain linked by a ribose. It not only can behave as a hydrotropic molecule to dissolve LLPS and aggregated proteins [11e13], it can also act as a bivalent binder for the FUS CTD. On the one hand, the adenine ring may bind to aromatic sidechains of CTD by p-p interaction, or/and interact with the side chains of Arg and Lys by p-cation interaction. On the other hand, its triphosphate chain may bind the side chains of Arg and Lys by electrostatic interaction [12,13,17]. As a result, at ATP concentration of ~ 1e2 mM, the bivalent binding of ATP to the multiple sites of the FUS CTD is expected to form a large but dynamic ATP-CTD complex, which thus manifesting as the formation of liquid droplets (II of Fig. 3C). However, in the exceeding presence of ATP, further binding of ATP to the FUS CTD will lead to the disruption of the dynamic structure, thus manifesting as dissolution of LLPS by ATP at high concentrations.
3.4. NMR characterization of ATP-induced enhancement and dissolution of LLPS of FUS In parallel to DIC imaging (Fig. 1), we also assessed the ATPinduced enhancement and dissolution of LLPS of the full-length FUS by NMR HSQC titrations. As we previously characterized [16], in the HSQC spectra, the well-dispersed HSQC peaks are from both well-folded RRM and ZnF, while the narrowly-dispersed peaks are from PLD, RGG1, RGG2 and RGG3 (Fig. 4A). Strikingly, upon gradual increase of ATP concentrations, the intensity of HSQC peaks of the full-length FUS also gradually decreased, similar to what was observed on the FUS CTD (Fig. 3B). At 10.0 mM, most HSQC peaks became too broad to be detected (IV of Fig. 4A). In light of the recent mutagenesis study to define the driving force for LLPS of FUS [17], our current results together lead to the proposal of a speculative model to rationalize the ATP-induced enhancement and dissolution of LLPS of FUS (Fig. 4B). Briefly, FUS molecules without ATP can undergo LLPS mainly driven by dynamic and multivalent inter-molecular p-cation interactions between aromatic residues within PLD and Arg/Lys residues within RGG regions (I of Fig. 4B). Upon addition of ATP into FUS, in addition to behaving as a hydrotrope, ATP can also act as a bivalent binder: its aromatic ring of adenine is able to dynamically interact with aromatic, or/and Arg/Ly residues, while its triphosphate chain can also dynamically interact with Arg/Ly residues of FUS. As a
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
5
Fig. 3. DIC and NMR characterization of the effect of ATP on LLPS of the FUS CTD. (A) DIC images of liquid droplets formed by the FUS CTD in the presence of ATP at different concentrations. The videos used to output these images are provided in Supplementary Data. (B) 1H-15N HSQC spectra of the 15N-labeled FUS CTD in the free state (blue) and in the presence of ATP at different concentrations (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
consequence, a large but dynamic ATP-FUS complex is formed, which manifests as enhancement of LLPS of FUS (II of Fig. 4B). However, upon further addition of ATP, the exceeding binding of ATP to FUS will lead to the disruption of the large and dynamic ATPFUS complex (III of Fig. 4B). Consequently LLPS of FUS will be dissolved at high ATP concentrations. As one FUS molecule is bound with a large number of ATP molecules, most HSQC peaks of FUS becomes too broad to be detected, due to the shortening of the transverse relaxation time T2, which most likely results from ms-ms dynamics or/and large molecular weight of the ATP-bound FUS.
4. Discussion LLPS is now recognized not only to underlie a variety of fundamental physiological processes, but its mis-regulation also leads to an increasing spectrum of human diseases such as
neurodegenerative diseases. Very recently, cellular factors including ATP and nucleic acids have been identified to critically regulate LLPS. In particular, identification of their general and key roles in LLPS and protein aggregation has explained two longstanding mysteries: 1) why cells need to maintain very high concentrations of ATP; and 2) RBPs such as FUS and TDP-43 are largely soluble in the nucleus only with a small portion undergoing LLPS, but become severely aggregated upon mislocalization to the cytoplasm, thus leading to ALS. However, so far their biophysical basis remains to be explored and particularly one critical question needs to be unanswered: whether the mediation by ATP and nucleic acids results from the buffering effect, or from specific binding. Such knowledge is not only essential for our understanding of the fundamental cell physiology, but also provides key foundations for developing therapeutic strategies/molecules to treat human diseases characteristic of protein aggregation.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
6
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7
Fig. 4. NMR characterization of the effect of ATP on LLPS of FUS. (A) 1H-15N HSQC spectra of the 15N-labeled FUS in the free state (blue) and in the presence of ATP at different concentrations (red). (B) A speculative model for ATP-induced enhancement and dissolution of LLPS of FUS. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Here we aimed to gain the biophysical basis for ATP to dissolve LLPS by DIC and NMR characterization of LLPS of FUS in the presence of ATP at concentrations ranging from 0 to 10 mM. To our surprise, in addition to the dissolution of LLPS by ATP at high concentrations (6 mM) exactly as previously reported [11], we found for the first time that ATP can also significantly enhance LLPS of FUS at low concentrations around 1e2 mM. Further assessment of the effect of ATP on LLPS of the FUS NTD and CTD reveals that ATP monotonically dissolves LLPS of the FUS NTD rich in aromatic residues, likely by mainly acting as a hydrotropic molecule as previously proposed [11]. By a sharp contrast, ATP can completely induce LLPS of the FUS CTD rich in Arg/Lys residues at low concentrations (1e2 mM), followed by further dissolution of LLPS at high ATP concentrations (6 mM). Consequently, the induction of LLPS by ATP cannot be rationalized by the hydrotropic properties of ATP. Here we propose that ATP enhances LLPS of FUS most likely by acting as a bivalent binder: its aromatic ring of adenine can bind aromatic residues of FUS by p-p interaction, or Arg/Lys by p-cation interaction; while its triphosphate chain also bind Arg/Lys by electrostatic interaction. Consequently in the presence of ATP at relatively low concentrations, a large and dynamic ATP-FUS can be formed which manifests as liquid droplets (Fig. 4B). Therefore ATP appears to own the dual ability to mediate LLPS of FUS by acting as a hydrotrope and a bivalent binder. As such the effect of ATP on LLPS is not only dependent on ATP concentrations, but also on the protein sequence. Furthermore, our NMR characterization reveals that ATP is able to extensively interact with most, if not all, residues of FUS. This suggests that the interaction between ATP and FUS is not as specific as other specific protein-ligand interactions. On the other hand, the NMR results also reveal that a large number of ATP molecules are dynamically bound/clustered over FUS molecules (III of Fig. 4B). Therefore, the two-stage effect of ATP on LLPS of FUS appears not to result from simple buffering effects as buffer ions are not supposed to extensively interact with proteins. Our finding that at low concentrations, ATP can enhance LLPS of FUS may bear critical implications in understanding age-dependent aggregation of proteins. Although at present the data of ATP concentrations in cells of different type and age are largely incomplete, it is usually thought that in most cells, ATP concentrations can
range from 6 to 10 mM [11e13]. However, it has been also reported that upon aging, ATP concentrations will keep reducing in cells. It is thus possible that in some aged cells, its ATP concentration becomes very low, at which ATP acts to enhance LLPS of proteins. Therefore, our finding implies that the age-dependent reduction of ATP concentrations not only can decrease its capacity in preventing protein aggregation, but may also act to enhance aggregation. Acknowledgement This study is supported by Ministry of Education of Singapore (MOE) Tier 2 Grant MOE2015-T2-1-111 to Jianxing Song. Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.bbrc.2018.09.014. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.09.014. References [1] S.F. Banani, H.O. Lee, A.A. Hyman, et al., Biomolecular condensates: organizers of cellular biochemistry, Nat. Rev. Mol. Cell Biol. 18 (2017) 285e298. [2] Y. Shin, C.P. Brangwynne, Liquid phase condensation in cell physiology and disease, Science 357 (2017) 6357. [3] S.C. Ling, M. Polymenidou, D.W. Cleveland, Converging mechanisms in ALS and FTD: disrupted RNA and protein homeostasis, Neuron 79 (2013) 416e438. [4] Y.R. Li, O.D. King, J. Shorter, et al., Stress granules as crucibles of ALS pathogenesis, J. Cell Biol. 201 (2013) 361e372. [5] M. Neumann, S. Roeber, H.A. Kretzschmar, et al., Abundant FUSimmunoreactive pathology in neuronal intermediate filament inclusion disease, Acta Neuropathol. 118 (2009) 605e616. [6] T.W. Han, M. Kato, S. Xie, et al., Cell-free formation of RNA granules: bound RNAs identify features and components of cellular assemblies, Cell 149 (2012) 768e779. [7] A. Patel, H.O. Lee, L. Jawerth, et al., A liquid-to-solid phase transition of the ALS protein FUS accelerated by disease mutation, Cell 162 (2015) 1066e1077. [8] T. Murakami, S. Qamar, J.Q. Lin, et al., ALS/FTD mutation-induced phase transition of FUS liquid droplets and reversible hydrogels into irreversible hydrogels impairs RNP granule function, Neuron 88 (2015) 678e690.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014
J. Kang et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e7 [9] A. Molliex, J. Temirov, J. Lee, et al., Phase separation by low complexity domains promotes stress granule assembly and drives pathological fibrillization, Cell 163 (2015) 123e133. [10] L. Lim, Y. Wei, Y. Lu, et al., ALS-causing mutations significantly perturb the self-assembly and interaction with nucleic acid of the intrinsically disordered prion-like domain of TDP-43, PLoS Biol. 14 (2016), e1002338. [11] A. Patel, L. Malinovska, S. Saha, et al., ATP as a biological hydrotrope, Science 356 (2017) 753e756. [12] A.M. Rice, M.K. Rosen, ATP controls the crowd, Science 356 (2017) 701e702. [13] T. Meyer, N. Voigt, In search for novel functions of adenosine 5'-triphosphate (ATP) in the heart, Cardiovasc. Res. 113 (2017) e59ee60. [14] S. Maharana, J. Wang, D.K. Papadopoulos, et al., RNA buffers the phase separation behavior of prion-like RNA binding proteins, Science (2018), https://
7
doi.org/10.1126/science.aar7366. Apr 12. pii: eaar7366. [15] L. Wang, J. Kang, L. Lim, et al., TDP-43 NTD can be induced while CTD is significantly enhanced by ssDNA to undergo liquid-liquid phase separation, Biochem. Biophys. Res. Commun. 499 (2018) 189e195. [16] Y. Lu, L. Lim, J. Song, RRM domain of ALS/FTD-causing FUS characteristic of irreversible unfolding spontaneously self-assembles into amyloid fibrils, Sci. Rep. 7 (2017) 1043. [17] J. Wang, J.M. Choi, A.S. Holehouse, et al., A Molecular Grammar Underlying the Driving Forces for Phase Separation of Prion-like RNA Binding Proteins, Cell, 2018 accepted (online publication date June 28). [18] K.A. Burke, A.M. Janke, C.L. Rhine, et al., Residue-by-Residue view of in vitro FUS granules that bind the C-terminal domain of RNA polymerase II, Mol. Cell 60 (2015) 231e241.
Please cite this article in press as: J. Kang, et al., ATP enhances at low concentrations but dissolves at high concentrations liquid-liquid phase separation (LLPS) of ALS/FTD-causing FUS, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/ j.bbrc.2018.09.014