Article

Acetylation within the First 17 Residues of Huntingtin Exon 1 Alters Aggregation and Lipid Binding Maxmore Chaibva,1 Sudi Jawahery,2 Albert W. Pilkington IV,1 James R. Arndt,1 Olivia Sarver,1 Stephen Valentine,1 Silvina Matysiak,2,5,* and Justin Legleiter1,3,4,* 1 The C. Eugene Bennett Department of Chemistry, West Virginia University, Morgantown, West Virginia; 2Fischell Department of Bioengineering, University of Maryland, College Park, Maryland; 3NanoSAFE and 4Center for Neurosciences, West Virginia University, Morgantown, West Virginia; and 5Biophysics Program, Institute for Physical Chemistry and Technology, University of Maryland, College Park, Maryland

ABSTRACT Huntington’s disease (HD) is a genetic neurodegenerative disorder caused by an expanded polyglutamine (polyQ) domain near the N-terminus of the huntingtin (htt) protein. Expanded polyQ leads to htt aggregation. The first 17 amino acids (Nt17) in htt comprise a lipid-binding domain that undergoes a number of posttranslational modifications that can modulate htt toxicity and subcellular localization. As there are three lysines within Nt17, we evaluated the impact of lysine acetylation on htt aggregation in solution and on model lipid bilayers. Acetylation of htt-exon1(51Q) and synthetic truncated htt-exon 1 mimicking peptides (Nt17-Q35-P10-KK) was achieved using a selective covalent label, sulfo-N-hydroxysuccinimide (NHSA). With this treatment, all three lysine residues (K6, K9, and K15) in Nt17 were significantly acetylated. N-terminal htt acetylation retarded fibril formation in solution and promoted the formation of larger globular aggregates. Acetylated htt also bound lipid membranes and disrupted the lipid bilayer morphology less aggressively compared with the wild-type. Computational studies provided mechanistic insights into how acetylation alters the interaction of Nt17 with lipid membranes. Our results highlight that N-terminal acetylation influences the aggregation of htt and its interaction with lipid bilayers.

INTRODUCTION Huntington’s disease (HD) is a neurodegenerative disorder caused by an abnormal stretch of more than 35–40 polyglutamine (polyQ) repeats within the first exon of the huntingtin (htt) protein (1). In HD, the age of onset and severity of disease are correlated with the polyQ length (2,3). Expanded polyQ is directly implicated in the aggregation of htt into fibrils and a variety of other structures (4–6). Substantial evidence suggests that N-terminal fragments of htt comparable to exon 1 are involved in HD (7–12). In HD brains, mutant htt is detected predominantly in microscopic inclusion bodies in the cytoplasm and nucleus (10), but it also associates with membranous organelles such as mitochondria, the endoplasmic reticulum (ER), tubulovesicles, endosomes, lysosomes, and synaptic vesicles (13–16). Approximately half of endogenous htt partitions with membranes after subcellular fractionation of neuron-

Submitted August 27, 2015, and accepted for publication June 15, 2016. *Correspondence: [email protected] or [email protected] Editor: Ka Yee Lee. http://dx.doi.org/10.1016/j.bpj.2016.06.018

like clonal striatal cells (17). Furthermore, membranous structures are assimilated onto the surfaces of htt inclusion bodies in cellular models (17,18), and htt aggregates accumulate brain lipids in mouse models (19,20). In N-terminal fragments of htt, the first 17 amino-terminal residues of htt that directly precede the polyQ domain, collectively referred to as Nt17, are required for membrane localization (21,22) and regulate lipid binding (23). Nt17 also promotes oligomer formation (24,25), and targeting Nt17 directly is an effective way to inhibit aggregation (24,26). In solution, monomeric Nt17 is predominantly disordered, as demonstrated by NMR (25), but circular dichroism studies suggest that Nt17 contains 10–50% a-helical content depending on the buffer conditions used (21,25,27,28), suggesting that Nt17 adopts multiple configurations. This is supported by computational studies, some of which predict that Nt17 forms an a-helix in aqueous solution (29,30), whereas others suggest that Nt17 is predominantly unstructured in solution (28,31–34). However, Nt17 is a-helical in at least some htt aggregates (26,35,36). Nt17 transitions to an a-helical structure upon self-association (24), when it interacts

Ó 2016 Biophysical Society.

Biophysical Journal 111, 349–362, July 26, 2016 349

Chaibva et al.

with phospholipids, detergents, or other apolar compounds (21,25,27,37). Furthermore, ion-mobility spectrometry/ mass spectrometry (MS) experiments suggest the presence of coexisting solution conformers for Nt17 peptides (38). Several posttranslational modifications (PTMs) are associated with htt pathogenesis, including proteolytic cleavage, acetylation, phosphorylation, ubiquitination, and SUMOylation (39–41). Dysregulation of PTMs triggered by expanded polyQ tracts may lead to aberrant htt interactions (42). A number of PTM sites are located in Nt17, and some of these sites are associated with multiple modifications (43). Specifically, proteomic mapping by MS identified acetylation sites at K6, K9, and K15 (44). This N-terminal acetylation preferentially occurred in htt with polyQ lengths below the disease threshold, suggesting that decreased acetylation of expanded htt could play a role in pathogenic mechanisms associated with HD. Membrane-related changes are a clear biochemical feature of HD (45–47). Htt’s strong association with lipids may play a prominent role in its normal functions (45,48). Amino-terminal mutant htt fragments aggregate on and damage a variety of phospholipid bilayers, suggesting that phospholipid interactions may play a role in htt-related toxicity (19,23,49). Perinuclear inclusions of htt disrupt the nuclear envelope in a mouse model of HD (50), and expanded polyQ embedded in the ER is associated with membrane distortion (51). As a result, determining the factors that regulate the affinity of htt for membranes is critical not only for understanding the normal functions of htt but also for identifying ways to modify htt-lipid interactions for potential therapeutic strategies. It is clear that Nt17 influences the aggregation of N-terminal fragments of htt (24,26,35) and its direct interactions with lipid membranes (23,27,37,52,53). Furthermore, PTMs within Nt17 have a profound effect not only on htt function and translocation (54–56) but also on the toxicity associated with mutant htt containing expanded polyQ domains (42,43,57–59). However, the impact of Nt17 acetylation on modulating the aggregation of htt and its interactions with lipid bilayers is poorly understood. Here, we aimed to elucidate the influence of acetylation of Nt17 on these phenomena. MATERIALS AND METHODS

Fusion RP-80 2.0  100 mm column (particle size of 4 mm; Phenomenex, Torrance, CA) with water 0.1% formic acid as solvent A, and acetonitrile with 0.1% formic acid as solvent B. A 60 min gradient was used (5–100% B over 60 min) with a 10 min column reequilibration between successive runs, at a flow rate of 0.200 mL/min. A 20 mL sample was used for all injections. The mass spectrometer was operated in a datadependent mode, and the five most intense ions in each scan were submitted for MS/MS high-energy collisional dissociation fragmentation. Data analysis was performed with Thermo Proteome Discoverer (Thermo Scientific). A database containing only the peptide of interest was generated manually and used as search criteria. The program searched for both acetylated and unmodified peptide ions within each LC-MS/MS data set. Peak areas for ions that contained the specific lysine residue were recorded for all labeled and unlabeled ions. The percent modification for each lysine residue was reported as a percent of the total ion signal:

P

%Mod ¼ P

Modified ion peak area P : Modified peak areaþ Unmodified peak area (1)

Thioflavin T assay A thioflavin T (ThT) assay was performed using a Molecular Devices M5 spectrophotometer (Molecular Devices, Sunnyvale, CA). Stock ThT (Sigma Aldrich, St. Louis, MO) was prepared in nanopure water at 1.0 mg/mL. Sulfo-N-hydroxysuccinimide acetate (NHSA) labeling was performed immediately after reconstitution. Then, 2.0 mL of the stock ThT was added to the sample (100 mL total sample volume/well). Fluorescence was monitored using 440-nm excitation and 480-nm emission wavelengths. The assay was performed at 25 C. Emission data were acquired every minute for 6 h. Background fluorescence was obtained from a blank sample containing no peptide and used for background correction. Each sample was run in triplicate and averaged.

Purification of glutathione S-transferase-httexon1 fusion proteins Glutathione S-transferase (GST)-htt-exon1 51Q fusion proteins were purified as described elsewhere (20,61). For details, see Supporting Materials and Methods.

Acetylation of htt Htt was acetylated using NHSA (62). Stock solutions of NHSA were prepared in water at 1.71 mg/mL. Labeling was carried out at NHSA/htt molar ratios to achieve 1, 2, and 3 NHSA for 1 min, and the reaction was quenched by adding 10 mM Tris. After acetylation, factor Xa protease was added to each protein sample and allowed to sit in ice for 1 h before bilayer experiments were conducted.

Peptide preparation Synthetic Nt17-Q35P10KK was purchased from the Keck Institute (Yale University) and stored at
20 C until later use. Disaggregation and solubilization of peptides were achieved with the use of published protocols (60). For details, see Supporting Materials and Methods in the Supporting Material.

Preparation of lipid bilayers

Liquid chromatography-tandem MS

Atomic force microscopy imaging conditions

Liquid chromatography-tandem MS (LC-MS/MS) was performed with an Accela binary pump LC coupled to a Q Exactive mass spectrometer (Thermo Scientific, San Jose, CA). Separation was facilitated by a Synergi

Atomic force microscopy (AFM) experiments were performed with a Nanoscope V Multimode atomic force microscope (Veeco, Santa Barbara, CA) equipped with a vertical engage J-scanner. For ex situ AFM, all images

350 Biophysical Journal 111, 349–362, July 26, 2016

Supported bilayers of total brain lipid extract (TBLE; Avanti Polar Lipids, Alabaster, AL) were prepared as previously described (23,49,63). For details, see Supporting Materials and Methods.

Htt Acetylation Alters Lipid Binding were acquired with diving-board-shaped silicon-oxide cantilevers with a spring constant of ~40 N/m and resonance frequency of ~300 kHz. For in situ AFM, a tapping-mode fluid cell with an O ring was equipped with a rectangular-shaped silicon nitride cantilever (Vista Probes, Phoenix, AZ) with a spring constant of ~0.1 N/m. For details on AFM image processing, see Supporting Materials and Methods.

Preparation of TBLE/polydiacetylene vesicles A polydiacetylene (PDA) assay was prepared as previously described (64–67). For details, see Supporting Materials and Methods.

Cell culture GT1-7 hypothalamic neurons were grown according to protocols described previously (68–72). Nonacetylated or acetylated htt was translocated into cells using the BioPORTER reagent (Genlantis, San Diego, CA). Cell viability was determined by MTT analysis. For details regarding the cell culture conditions, htt translocation, and viability assays, see Supporting Materials and Methods.

Computational methods A helical Nt17 peptide was built using the UCSF Chimera molecular modeling software package (73). Acetylation modifications to Nt17 were made using the PyMOL molecular graphics software package (74). All simulations were performed using the Gromacs molecular dynamics package (75) with the AMBER99SB force field (76–78). The parameters used for the acetylated lysine residue (79) and dioleoyl phosphatidyl choline (DOPC) lipids (80–82) were suitable for use with AMBER force fields.

amines (62,83). The model htt peptide contained Nt17, 35 repeat glutamines, a 10-residue polyproline domain, and two C-terminal lysines to aid solubility (this peptide will be referred to as Nt17-Q35-P10-KK). Nt17-Q35-P10-KK was used in place of the GST-fusion construct to reduce spectral complexity and eliminate ions corresponding to missed GST-htt cleavages. When MS/MS data were insufficient or not available, the exact monoisotopic mass of a labeled ion was used to identify an ion of interest. If the ion had been produced from a peptide of a missed GST cleavage product, the method would have led to a false positive for a labeling event, skewing the results. As there are three lysines in Nt17, labeling was carried out at molar ratios of 1, 2, and 3 NHSA to htt. We verified which lysines were acetylated by using chymotrypsin and MS to measure the relative amount of labeled and unlabeled fragments that contained single lysine residues (Fig. 1 a), and ran ThT assays to determine the impact of acetylation on Nt17-Q35-P10KK aggregation (Fig. 1 b). K9 was the most readily labeled, followed by K6 and K15. At 1 NHSA, all three lysines in Nt17 were labeled to some extent, but no more than 30%. At 2 NHSA, K9 was acetylated to a significantly larger extent (69%) than the other two lysines. At 3 NHSA, K9 acetylation increased to >80%, and >60% of K6 was labeled.

Peptide-membrane simulation Completely helical peptides were placed at least 0.5 nm away from one leaflet of an equilibrated DOPC lipid bilayer (128 lipids) with explicit simple point charge water molecules. At the beginning of each simulation, the helix vector of the peptide was aligned perpendicular to the bilayer normal. The simulation box size for all systems was ~6.5 nm in the x and y directions, and ~12 nm in the z direction. The systems consisted of ~55,000 total atoms and ~12,000 water molecules (36,000 water atoms). A 1 ns NVT simulation at 330 K was followed by a 1 ns NPT simulation at 330 K and 1 bar with semi-isotropic pressure coupling and position restraints on the lipid and peptide. The 200 ns production simulation was performed under the same conditions as the preceding NPT equilibration run, without position restraints. Three replicate simulations of each peptide-membrane system were performed. Each replica was assigned different initial atom velocities using a different set of random seeds. In addition, the peptide in each replica simulation was rotated differently around its own helix vector such that different residues were initially in contact with the membrane. The variation between replicas was used to verify that patterns observed in the association of the peptide with the membrane were not caused by a bias in the initial conformation. For additional details, see Supporting Materials and Methods.

RESULTS Chemically induced acetylation occurs at all lysines in Nt17 and inhibits aggregation In htt-exon1, three lysine residues are available for acetylation within Nt17 (K6, K9, and K15). To determine their availability for acetylation, model htt peptides were exposed to NHSA, an established agent for acetylating primary

FIGURE 1 Lysines within Nt17 are accessible for acetylation and impact fibril formation. (a) MS analysis of the extent of lysine acetylation by NHSA in Nt17 for a model peptide Nt17-Q35-P10-KK. Error bars represent the standard deviation. (b) ThT aggregation assays demonstrating the impact of lysine acetylation by NHSA on Nt17-Q35-P10-KK fibril formation. Markers represent every 12th data point.

Biophysical Journal 111, 349–362, July 26, 2016 351

Chaibva et al.

Acetylation of K15 increased with larger doses of NHSA, but it was never >30%. The two flanking lysines at the C-terminus were not appreciably acetylated at any dose of NHSA. As acetylation increased, aggregation was inhibited in a dose-dependent manner. At 1 NHSA, there was a slightly slower rate of aggregation (Fig. 1). At 2 NHSA, the lag phase was increased, with an overall reduced rate of aggregation. At 3 NHSA, aggregation was significantly decreased. This inhibition of aggregation suggests that K6 and K9 play important roles in aggregation, as reduced aggregation was observed once these residues were appreciably labeled. Acetylation of Nt17 inhibits htt-exon1 fibril formation To determine the impact of acetylation on full-length httexon1 aggregation, a mutant htt fragment that expresses exon1 with 51Q (htt-exon1(51Q)) was acetylated with NHSA and the aggregation was tracked. Htt-exon1(51Q) was purified from Escherichia coli as a soluble fusion with GST (61). The GST moiety was cleaved with Factor Xa, releasing htt-exon1(51Q) and initiating aggregation (84). Htt-exon1(51Q) was exposed to NHSA at molar ratios of 1, 2, and 3 NHSA/htt. To verify that addition of the NHSA would not interfere with removal of the GST tag, fusion protein that had been exposed to the different molar ratios of NHSA was incubated with Factor Xa on ice for 1 h and analyzed using SDS-PAGE (Fig. S1). A densitometry analysis of the resulting gels demonstrated that there was no appreciable difference in the efficiency of cleaving the GST tag upon addition of NHSA. Incubation experiments were performed with freshly prepared 20 mM solutions of htt-exon1(51Q) that had been acetylated by NHSA at the different molar ratios. Aliquots were removed from the solutions after 1, 3, 5, 8, and 24 h of incubation for AFM analysis (Fig. 2). To quantify the effect of acetylation on htt aggregation, we analyzed the AFM images from all incubations by counting the number of oligomers or fibrils per mm2 at each time point (Fig. 2, b and c). For this analysis, aggregates were defined as objects in the image taller than 2 nm. Fibrils were distinguished from oligomers by a length/width (aspect) ratio filter. Fibrils had an aspect ratio of >3 and oligomers had an aspect ratio of 1 mm in length at later time points. When httexon1(51Q) was labeled with 1 NSHA, there was no significant difference in the number of oligomers compared

352 Biophysical Journal 111, 349–362, July 26, 2016

FIGURE 2 Nt17 acetylation in htt-exon1(51Q) suppresses aggregation. (a) Representative AFM images taken in air after 5 and 24 h of incubation of htt-exon1(51Q) alone or labeled with 1, 2, or 3 NHSA. Blue arrows indicate oligomers; black arrows indicate fibrils. (b and c) Number of (b) oligomers and (c) fibrils per unit area observed in AFM images of aggregates formed by htt-exon1(51Q) alone or labeled with 1, 2, or 3 NHSA as a function of time. The error bars represent the standard deviation.

with nonacetylated htt-exon1(51Q); however, after 8 h, there was a significant decrease in the number of fibrils. Increasing the NHSA treatment to 2 again resulted in no significant change in the number of oligomers, but there was a clear inhibition of fibril formation as early as after 1 h of incubation. Treatment with 3 NHSA resulted in a significant decrease in the number of oligomers and fibrils as a function of time. Overall, increasing the ratio of NHSA to htt decreased the number of fibrils that formed in a dose-dependent manner, but only at 3 NHSA was the number of oligomers significantly affected. Next, we compared the morphology of aggregates that formed under the different acetylating conditions (Fig. S2). Fibrils of htt-exon1(51Q) with increasing acetylation tended to be shorter, presumably due to a decreased time for fibril expansion as a result of the longer lag phase. However, the height along the contour of these fibrils was indistinguishable (~5–6 nm; Fig. S2 a), suggesting a similar fibril structure. The size distribution of oligomers varied greatly with increasing acetylation (Fig. S2 b). Without any NHSA, the distribution of oligomer size remained relatively tight, with a slight increase in the mode height from 5–6 nm to 6–7 nm after 24 h. The oligomer size distribution was similar for up to 5 h of incubation, when 1 NHSA was added, but at 8 and 24 h a pronounced population of larger oligomers (~9–11 nm) began to form. This second, larger population of oligomers became more pronounced at earlier

Htt Acetylation Alters Lipid Binding

time points with increasing amounts of NHSA. With 3 NHSA, the larger population of oligomers eventually dominated the size distribution. The increased size of the oligomers formed with 3 NHSA may explain the significant reduction in the total number of oligomers, as more peptide was incorporated into fewer aggregates.

Nt17 acetylation inhibits htt-lipid interactions To determine whether acetylation of Nt17 would alter the interaction of htt-exon1 with lipid membranes, we performed a TBLE/PDA vesicle lipid interaction assay (Fig. 3 a). This assay uses mixed vesicles comprised of phospholipids and polymerized PDA. A rapid, visible color change from blue to red is induced by direct interaction between the peptides and the phospholipid within the vesicles. The colorimetric response can be used to track peptide-membrane interactions, and PDA-based vesicles have been used to study a variety of protein-lipid interactions, including those involving amyloid proteins (65,67,86,87). A quantitative value representative of the blue-to-red color

transition is obtained by determining the % colorimetric response (%CR) (67,87): %CR ¼ ½ðPB0 -PBI Þ=PB0   100;

(2)

where PB is the red/blue ratio of absorbance (A), defined as Ablue/(Ablue þ Ared); Ablue is the absorbance at the blue component in the UV-vis spectrum (z640 nm); Ared is the absorbance at the red component (z500 nm); PB0 is the red/blue ratio of the control sample (before induction of color change); and PBI is the value obtained for the vesicle solution after addition of peptides. TBLE/PDA vesicles were exposed to 20 mM httexon1(51Q) or htt-exon1(51Q) that had been acetylated by NHSA (1, 2, or 3), and %CR was measured for 16 h (Fig. 3 a). Upon exposure to htt-exon1(51Q), there was an initial steady increase in %CR as peptide bound and aggregated on the vesicle. Eventually, %CR leveled off to a quasisteady-state value. Initially, there was no significant change in the %CR comparing exposures of unaltered httexon1(51Q) with those of htt-exon1(51Q) þ 1 NHSA. After ~14 h, there was a small decrease in the %CR between

FIGURE 3 Nt17 acetylation modulates the interaction between htt-exon1(51Q) and lipid membranes. (a) TBLE/PDA vesicle assay demonstrating the impact of lysine acetylation by NHSA on the ability of htt-exon1 (51Q) to bind and aggregate on lipid membranes. (b) AFM image of an unperturbed TBLE bilayer. (c–f) AFM images of TBLE bilayers taken after 1 and 5 h of exposure to (c) htt-exon1(51Q), (d) htt-exon1(51Q) þ 1 NHSA, (e) httexon1(51Q) þ 2 NHSA, (f) htt-exon1(51Q) þ 3 NHSA. Insets are zoomed-in images of the indicated regions in the larger images. Blue arrows indicate oligomers. (g) Quantification of the RMS roughness associated with disrupted regions of the bilayer and the percent area of the bilayer that was disrupted for bilayers exposed to htt-exon1(51Q) alone or labeled with 1, 2, or 3 NHSA as a function of time.

Biophysical Journal 111, 349–362, July 26, 2016 353

Chaibva et al.

these two samples that was significant. As the amount of NHSA was increased to 2 or 3, there was a clear dosedependent reduction in the %CR, indicating that acetylation reduced the interaction between htt-exon1 and the lipid membrane. However, there was still an appreciable interaction of htt-exon1(51Q) with the lipid vesicles under all conditions. To further investigate the impact of acetylation on the interaction of htt-exon1 with lipid bilayers, we prepared supported model lipid membranes (Fig. 3 b). Supported lipid bilayers preserve many important properties of their free membrane counterparts (e.g., lateral fluidity) (88,89), and one can continuously track the interaction of amyloidforming proteins with these bilayers by using in situ AFM. Smooth bilayers (as determined by AFM) were exposed to unmodified htt-exon1(51Q) proteins or httexon1(51Q) that had been acetylated with varying amounts of NHSA (1, 2, or 3 NHSA). To cleave the GST moiety, Factor Xa was incubated with the GST-exon1 fusion protein on ice for 1 h before the protein was exposed to a supported bilayer. The resulting htt-exon1 preparations predominantly contained monomers and small oligomers (4,85). Upon initial exposure of the TBLE bilayer to 20 mM htt-exon1(51Q), there was a noticeable deposition of globular protein aggregates on the bilayer and regions of perturbed bilayer morphology (evidenced by an increase in surface roughness), consistent with previous reports (Fig. 3 c) (49,63). The globular aggregates were stable on the bilayer surface, as they were identifiable in consecutive images. The extent (or area) of bilayer that was perturbed by htt-exon1(51Q) increased with time. Although most of the increased roughness appeared as an accumulation of protein above the initial bilayer surface, height profiles across these perturbed regions demonstrated that gaps and cavities within the bilayer formed (Fig. S3), verifying bilayer disruption. This is consistent with previous results based on calcein dye leakage assays (19) and mechanical changes in bilayers (23,49,63) indicating that htt-exon1 damages membranes. When TBLE bilayers were exposed to 20 mM aliquots of htt-exon1(51Q) that had been acetylated, morphological changes and globular aggregates similar to those associated with nonmodified htt were observed (Fig. 3, d–f). However, the extent of bilayer perturbation appeared to be reduced as a function of increasing exposure to NHSA. Interestingly, no fibril aggregates of any of the modified or nonmodified httexon1(51Q) proteins were observed on the bilayer surface. In an effort to quantify the extent of interaction between the TBLE bilayer and the different acetylated or nonacetylated htt-exon1(51Q) proteins, we performed an AFM image analysis. First, the root mean-squared (RMS) roughness of the perturbed regions of the bilayer was measured (Fig. 3 g). A freshly formed TBLE bilayer has an RMS surface roughness of 0.3 5 0.05 nm. To prevent error associated with the varying extent of perturbation of the bilayer

354 Biophysical Journal 111, 349–362, July 26, 2016

structure, RMS roughness measurements of areas that had been destabilized by modified or nonmodified httexon1(51Q) were restricted only to portions of the images that showed an altered morphology. This allowed for a direct comparison of the extents of altered bilayer morphology within the perturbed regions. Within the first hour of exposure to the different acetylated or nonacetylated htt-exon1(51Q) constructs, regions of the bilayer that were significantly rougher (p < 0.01) than the unperturbed bilayer were observed; however, the roughness decreased with increasing acetylation. The roughness associated with perturbed bilayer regions remained relatively constant for 4 h; however, there was a pronounced increase in roughness after 5 h of exposure to nonacetylated htt-exon1(51Q) or htt-exon1(51Q) with 1 NHSA. We next wanted to determine the extent of bilayer perturbation as a function of protein acetylation. To accomplish this, we determined the percentage of the surface that displayed any increased roughness (Fig. 3 g). Whereas the portion of the bilayer that was perturbed by acetylated or nonacetylated httexon1(51Q) steadily increased with time, the percent of the bilayer that was morphologically altered decreased as a function of increased protein acetylation. This trend corresponds well with the results from the first 5 h of the TBLE/PDA lipid binding assays. Regardless of the acetylation state of the protein, similar species of aggregates formed on the bilayer surface. The sizes of the globular aggregates of acetylated and nonacetylated htt-exon1(51Q) on the bilayer surface were compared (Fig. S4). The mode height of the oligomers observed on the bilayer was 2–3 nm for all experimental conditions and all time points, although larger aggregates increased in population over 5 h. It cannot be determined whether these globular aggregates were comprised solely of peptide or also contained lipid components. Because all of the potential acetylation sites in htt-exon1 are contained within Nt17, and Nt17 has known lipid-binding properties, acetylation may change the affinity of this domain for lipid membranes. To address this issue, we made a series of Nt17 peptides that contained lysine-toglutamine mutations to mimic acetylation. We performed TBLE/PDA lipid binding assays to assess the relative affinity of these Nt17 peptides for lipid membranes (Fig. S5). By excluding the polyQ domain, we could remove the change in the %CR in the PDA assay associated with aggregation and directly assess the ability of these peptides to bind. Although htt-exon1 and synthetic peptides containing the polyQ domains have been shown to disrupt and aggregate on lipid membranes (19,23,49,63), pure Nt17 peptides do not permeabilize lipid membranes (27). All of the peptides that contained a single mutation reduced the affinity of Nt17 for the vesicles, with the K9Q and K15Q mutants having a significantly lower affinity for the membrane compared with the K6Q mutant. This suggests that K9 and K15 may play a greater role in the interaction of

Htt Acetylation Alters Lipid Binding

Nt17 with lipid membranes. Double and triple mutants also reduced the affinity of Nt17 for the lipid membrane, albeit only to a similar extent as the K6Q mutation alone. As the overall %CR is considerably smaller for the Nt17 peptides compared with htt-exon1(51Q) (Fig. 3 a), aggregation on the vesicle surface appears to be the major contributor to the %CR associated with exposure to the full htt-exon1. Nt17 acetylation reduces htt toxicity Next, we examined the influence of Nt17 acetylation on htt toxicity. Htt-exon1(51Q) was acetylated by exposure to NHSA at various ratios (0x, 1, 2, or 3 molar ratios) and translocated into GT1-7 mouse neuronal cells at a final concentration of 2 mM. MTT reduction assays were used to assess cell viability 3 h after translocation of htt (Fig. S6). The viability of cells exposed to htt-exon1(51Q) (no acetylation) was only 69% 5 5% compared with control cells that were not exposed to htt. Preacetylation of htt with NHSA increased cell viability in a dose-dependent manner. Treatment of htt with 1 NHSA increased cell viability to 77% 5 5%; however, this was not statistically significant. Treatment with 2 and 3 NHSA resulted in increased cell viability, to 93% 5 4% and 100 5 4%, respectively, suggesting that acetylation of Nt17 reduces the toxicity associated with htt. Acetylation alters the interaction of Nt17 with lipid membranes Since our experimental results demonstrated that acetylation of Nt17 modulates the interaction between htt and lipid membranes, we next wanted to explore the mechanisms underlying those observations. Therefore, we characterized the conformational dynamics of individual Nt17 domains with different acetylated lysines in the presence of DOPC lipid bilayers via a computational approach using atomistic molecular-dynamics simulations. Htt interacts with phosphatidylcholines in general and DOPC specifically (19,90). The aim of these simulations was to examine early events in the surface binding of Nt17. Along with the wild-type Nt17 sequence, we studied three acetylation states: Nt17with K9 acetylated (Nt17-ACK9); Nt17with K6 and K9 acetylated (Nt17-ACK6/9); and Nt17 with K6, K9, and K15 acetylated (Nt17-ACK6/9/15). Three replicates of each peptide interacting with a DOPC bilayer were run for 200 ns. To obtain a mechanistic understanding of the impact of acetylation on the interaction of Nt17 with lipid bilayers, we determined the insertion depth and hydrogen-bond formation between Nt17 and a DOPC bilayer in a residue-specific manner as a function of time (Fig. 4). With unmodified Nt17, phenylalanine17 (F17) tended to insert into the membrane first (two of the three simulations), and this occurred within the first 50 ns of the simulation. Once F17 inserted into the membrane, hydrogen bonds with the membrane

were formed by nearby K15 and serine16 (S16). In the peptides containing various amounts of acetylation, F17 sometimes associated with the membrane (replicate 1 for Nt17-ACK9, and replicates 1 and 2 for Nt17-ACK6/9/15); however, in contrast to the nonacetylated peptide, this association tended to be temporary. With acetylation, an interaction between the peptide and membrane would sometimes be initiated by residues 1–3, and this interaction was accompanied by hydrogen-bond formation between threonine3 (T3) and the membrane. The stability of this interaction mediated by residues 1–3 was variable, as sometimes the peptide would remain bound to the membrane and at other times it would detach. As interactions between the different Nt17 peptides and the DOPC bilayer did not lead, in some simulation runs, to permanent peptide-membrane association, we determined the interaction energetics and compared them across simulations to better understand the stability requirement for complete association (Fig. S7). Attractive electrostatic interactions of significant magnitude (
300 to
600þ kJ/mol) were often observed, but these were usually insufficient to cause permanent peptide association with the membrane (replicate 3 for Nt17, replicates 2 and 3 for Nt17-ACK9, replicate 2 for Nt17-ACK6/9, and replicates 1 and 3 for Nt17-ACK6/9/15). For permanent peptide association with the membrane, the magnitude of van der Waals interactions typically had to exceed the magnitude of electrostatic interactions between the peptide and membrane (replicates 1 and 2 for Nt17, replicate 1 for Nt17-ACK9, replicates 1 and 3 for Nt17-ACK6/9, and replicate 2 for Nt17-ACK6/9/15). The largest contribution to van der Waals interactions between the peptide and membrane came from hydrophobic residues. Consistent with previous work demonstrating that peptide association can cause local membrane thinning (34), we observed pronounced thin patches in membranes to which peptides were able to bind permanently (Fig. S8). In addition, changes in the lateral distribution of lipids around the site of peptide insertion could be observed when the peptide was embedded deep enough in the membrane (Fig. S9). Peptides close to the inserted lipid were less closely packed with respect to neighboring lipids. This result is consistent with previous observations of fibril disruption of membranes (91). As the binding of Nt17 to membranes is associated with a transition from a disordered structure to an a-helical one (27,92), the helical content in modified and unmodified Nt17peptides bound to a DOPC bilayer was analyzed (Fig. 5). The helical content was not significantly modified by addition of acetylated lysine residues when a binding event occurred. Stable a-helices were associated with a 90 orientation to the lipid bilayer (the helix is oriented perpendicular to the bilayer normal), indicating that specific interactions with the membrane have a stabilizing effect on the helix, in agreement with solid-state NMR structures of Nt17 bound to lipid membranes (37). Although the helical

Biophysical Journal 111, 349–362, July 26, 2016 355

Chaibva et al.

FIGURE 4 Acetylation alters the interaction of Nt17 with DOPC bilayers. Computational studies were conducted to examine the interaction between DOPC bilayers and Nt17 with no acetylation, K9 acetylated, K6-K9 acetylated, or K6-K9-K15 acetylated. Three replicates of each interaction were performed for 200 ns. (a) Distance between each residue and the phosphorus atoms of the DOPC bilayer around that residue plotted as a function of time. The color code represents the distance in nanometers. Deeper insertion into the membrane is shown in darker shades of black. All residues that are >0.5 nm above the membrane are shown in white. (b) Number of hydrogen bonds that formed between each residue and the membrane as a function of time. The color code represents the number of hydrogen bonds.

contents and orientations of bound modified and unmodified Nt17 peptides were similar, there were subtle differences among the replicates. In the first replicate of wild-type Nt17, the peptide fully embedded into the bilayer and formed a stable a-helix that spanned almost the entire length of the domain (approximately residues 3–17; Fig. S10). The charged residues (K, glutamic acid (E)) in the bound a-helix pointed toward the bilayer headgroups, whereas other slightly polar and hydrophobic residues pointed toward the hydrophobic core. K9 associated with E5 and E12 similarly to what was observed in previous reports (52). In the second replicate for wild-type Nt17, the peptide did not fully embed into the membrane during the 200 ns, and the a-helical content was constrained to approximately residues 3–8 (Fig. S10). Despite the formation of an extensive a-helix, the peptide remained anchored to the membrane for most of the simulation. Although the central region of the wild-type Nt17 did not insert into the membrane during this replicate, this region was structurally stabilized by interactions between K9, E5, and E12. The second replicate of Nt17-ACK6/9/15 inserted into the membrane via initial interactions associated with F17 and formed a stable a-helix that spanned approximately residues 3–17 (Fig. S11). How-

356 Biophysical Journal 111, 349–362, July 26, 2016

ever, several polar and hydrophobic residues (methionine (M1), M8, and F17) now pointed toward the phosphorus lipid headgroups. As an acetylated K is less attracted to polar groups than a charged, unmodified K, the peptide orientation was altered in the a-helix to accommodate these modified residues, and the oxygen (OG1) and nitrogen (N) atoms of T3 preferentially formed a hydrogen bond with an oxygen atom of the phosphate group (O14) from DOPC. Several acetylated peptides initially interacted with the membrane using T3 as an anchor instead of F17. In replicate 1 of Nt17-ACK9, residues near T3 and the unmodified K15 inserted into the membrane. Instead of associating with E5 and E12 (as K9 typically would in its unmodified form), ACK9 pointed into the water and away from the membrane (Fig. S12). Instead, K6 associated with E5. Interaction with the membrane via T3 was also observed in replicates 1 and 3 of Nt17-ACK6/9 (Fig. S13). In replicate 1, the initial anchoring via T3 was followed by insertion of more residues and the formation of a relatively stable a-helical structure spanning approximately residues 3–12. As would be typically expected, hydrophobic residues pointed toward the hydrophobic core in the a-helical form. ACK6 and ACK9 pointed toward the headgroups but

Htt Acetylation Alters Lipid Binding

FIGURE 5 Acetylation does not significantly alter the helical content of Nt17 on DOPC bilayers. (a–d) Secondary structure per residue as a function of time and alignment angle of (a) Nt17 with no acetylation, (b) K9 acetylated, (c) K6-K9 acetylated, or (d) K6-K9-K15 acetylated is presented as a function of time.

were now clustered together, with M1 also pointing toward them. In replicate 3, K15 was still able to interact directly with the membrane. However, ACK6 and ACK9 now pointed away from the membrane and toward each other and M8. The OG1 and N atoms of T3 again formed a hydrogen bond with atom O14 from DOPC to anchor the peptide to the membrane. An energetic analysis of the entire peptide with the membrane yielded insight into the stability criteria for permanent association. Because we have observed differences in residue orientation within membrane-embedded peptides with varying degrees of acetylation, we compared the interaction energies between different residues and membrane headgroups (Figs. S14 and S15). The structural changes between native Nt17 and acetylated peptides appeared to cause hydrophobic and polar residues to point toward membrane headgroups, as described above. This observation is supported by the fact that, as expected, although acetylation decreased the favorable interactions of charged residues (K, E) with negatively charged phosphate headgroups, it also increased the interaction of polar residues with phosphate headgroups (Fig. S15). Acetylation also increased the interaction of charged residues with choline headgroups. A combined effect of increased polar residue interactions with phosphates and increased charged residue interactions with cholines may therefore stabilize acetylated Nt17 inside the membrane

in the absence of favorable native interactions between charged residues and phosphate headgroups. DISCUSSION While polyQ expansion plays a key role in htt aggregation and toxicity, flanking sequences adjacent to the polyQ domain influence aggregation, aggregate stability, important biochemical and functional properties of the protein, and ultimately htt’s role in pathogenesis. In particular, the Nt17 domain of htt influences localization, aggregation, and degradation of htt in a cellular context (21,22,43,59,93– 96). In this study, we investigated the impact of acetylation of lysine residues within the Nt17 domain of htt on its aggregation and interactions with lipid membranes. Acetylation of Nt17 impeded the formation of htt fibrils and promoted the formation of larger, globular oligomers. Acetylation also lowered the affinity of Nt17 for lipid membranes, reducing the membrane damage associated with exposure to htt. Acetylating htt with NHSA reduced htt toxicity in a dose-dependent manner in Gt1-7 cells. Computational studies suggested that the reduced affinity of Nt17 for lipid membranes caused by acetylation may be due to a number of factors associated with removal of the positive charge, including altered hydrogen bonding (both intrapeptide and peptide-phospholipid) and electrostatic interactions. As

Biophysical Journal 111, 349–362, July 26, 2016 357

Chaibva et al.

aberrant htt-lipid interactions have been implicated in a variety of toxic mechanisms (50,51), understanding how acetylation of Nt17 impacts htt aggregation and its interaction with membranous surfaces could lead to therapeutic interventions for HD. For example, acetylation in Nt17 is reduced in htt, which contains expanded polyQ domains in comparison with the wild-type (44), and our data indicate that acetylation reduces the affinity of htt for lipid membranes and has a protective effect against htt toxicity in cell cultures. Nt17 has been implicated in the initial stages of htt-exon1 aggregate formation (21,26,30,36,97). As such, investigators have proposed a variety of aggregation mechanisms mediated by Nt17, with most supporting the notion that Nt17 self-association plays a key role in the formation of small, a-helix-rich oligomers (25,28,98), suggesting that Nt17 promotes fibril formation through an oligomer-mediated pathway (36,53,99). Given that oligomers are widely considered to be important toxic species in HD, Nt17-directed oligomerization could play a vital role in htt-related toxicity. Indeed, self-association of Nt17 represents the rate-determining step in the formation of htt-exon1 oligomers in cellular models (24,100). Here, we demonstrated that acetylation of Nt17 slows htt aggregation and promotes the formation of larger oligomeric species. There are several plausible mechanisms that could explain this observation. One scenario is that acetylation promotes the formation of distinct htt oligomers that are off the pathway to fibril formation. Alternatively, it is possible that acetylation stabilizes the a-helix-rich structure of the oligomers, delaying the transition to fibril structures and giving the oligomers time to increase in size by accumulating more monomers into the aggregate. We demonstrate that chemically acetylating lysines within Nt17 slows aggregation into fibrils without altering the fibril structure and reduces htt toxicity. Our finding further supports the notion that Nt17 plays an important role in htt aggregation and, more specifically, that targeting the charged lysines within Nt17 may be a viable strategy for modulating aggregation. Full-length htt is a multidomain protein associated with numerous, distinct cellular functions. With regard to the potential specificity of Nt17 for membranes of different composition, many membrane-associated functions are attributable to htt, such as cellular adhesion (101), motility (102,103), cholesterol and energy homeostasis (104), molecular scaffolding for coordination of membrane and cytoskeletal communication (47), and facilitating microtubule-dependent vesicle transport (48). Acetylation of Nt17 may play a role in regulating these activities. The trafficking of htt-exon1 between membranes associated with the ER, autophagic vacuoles, mitochondria, and Golgi appear to be facilitated by Nt17 (21,22). Trafficking of htt to specific subcellular compartments, such as the nucleus (105,106), can have a strong influence in htt-related toxicity. The Nt17 domain forms an amphipathic a-helix (AH) (26,35,36), and a number of biophysical functions are asso-

358 Biophysical Journal 111, 349–362, July 26, 2016

ciated with AHs (107), including the ability to detect and bind highly curved membranes (108–110). As AHs tend to bend membranes weakly, their interaction with lipids is easily regulated. Given the large number of potential sites for PTMs contained within Nt17, including acetylation, regulation of htt-lipid interactions via Nt17 may play a role in the trafficking of htt within the cellular environment. As such, the altered affinity upon acetylation of Nt17 for lipid membranes could play a role in trafficking htt within cells, with potential implications for htt’s normal function and toxicity. Such a scenario has been demonstrated for phosphorylation of Nt17, which affects the nuclear export of htt (21,111). Acetylation of K9 plays a critical role in clearance of htt via the proteasome and lysosome (95). As a result, the acetylation of Nt17 and its modulating effect on membrane affinity may play a role in a variety of normal functions of htt (27,92,112). In a theoretical AH, lysines 6 and 15 would be positioned at the interface between the hydrophilic and hydrophobic faces. It would appear that residues located in this interface play a critical role in the ability of Nt17 to bind membranes. Simulations of htt peptide fragments associated with lipid membranes indicate that K6 and K15 form many hydrogen bonds with bilayer components (34). All-atom simulations demonstrated that K6 and K15 often form salt-bridges with phospholipids; however, K9 was predominantly involved in intramolecular salt-bridges with glutamic acid residues within Nt17 (52), suggesting that perturbing these residues would likely impact Nt17’s ability to bind membranes. A combination of experiments and simulations indicates that the structure and orientation of Nt17 on membranes are mainly influenced by the sequestration of nonpolar residues into the hydrophobic core of a membrane; however, membrane composition influences the affinity of Nt17 for phospholipid bilayers via specific electrostatic interactions (34,52). Collectively, this suggests a potential role for charged lysines in Nt17 selectively targeting membranes of specific phospholipid compositions, and removing their charge via acetylation could represent a mechanism to regulate this specificity. In this regard, mutating charged residues in Nt17, including the lysines, to alanines alters the localization of htt within cells (21). In a study investigating the role of SUMOylation in htt toxicity, mutating the lysine residues in Nt17 to arginine was shown to reduce the cytotoxicity of Drosophila models of HD (43). This reduction in cytotoxicity was attributed to the lack of lysine residues available for SUMOylation. As both acetylation and SUMOYlation modify lysines, acetylation is a known mechanism for regulating the SUMOylation of proteins (113). Furthermore, proteomic mapping by MS verified that K9 is appreciably acetylated in mammalian cell lysates (44). The role of K9 acetylation has not yet been fully evaluated; however, acetylation of Nt17 may function to regulate SUMOylation, reducing the toxicity associated with SUMOylation of htt (114).

Htt Acetylation Alters Lipid Binding

As the function, localization, and aggregation of htt have all been linked to Nt17, there is a need to identify at a mechanistic level how Nt17 influences the formation of toxic aggregates and their interaction with membranes. Here, we demonstrated that lysine residues play an important role in both aggregation and membrane binding, and this role can be affected by the removal of lysine’s positive charge via acetylation. As acetylation in Nt17 is known to occur naturally (44,114), the biophysical impact of this PTM could be important for htt’s normal function and have a modulating effect on the toxicity related to mutant htt. Such knowledge is important because it could lead to better-defined therapeutic targets (in addition to the polyQ tract) and strategies to treat HD. SUPPORTING MATERIAL

7. Kim, Y. J., Y. Yi, ., M. DiFiglia. 2001. Caspase 3-cleaved N-terminal fragments of wild-type and mutant huntingtin are present in normal and Huntington’s disease brains, associate with membranes, and undergo calpain-dependent proteolysis. Proc. Natl. Acad. Sci. USA. 98:12784–12789. 8. Cooper, J. K., G. Schilling, ., C. A. Ross. 1998. Truncated N-terminal fragments of huntingtin with expanded glutamine repeats form nuclear and cytoplasmic aggregates in cell culture. Hum. Mol. Genet. 7:783–790. 9. Hoffner, G., M.-L. Island, and P. Djian. 2005. Purification of neuronal inclusions of patients with Huntington’s disease reveals a broad range of N-terminal fragments of expanded huntingtin and insoluble polymers. J. Neurochem. 95:125–136. 10. Davies, S. W., M. Turmaine, ., G. P. Bates. 1997. Formation of neuronal intranuclear inclusions underlies the neurological dysfunction in mice transgenic for the HD mutation. Cell. 90:537–548. 11. Landles, C., K. Sathasivam, ., G. P. Bates. 2010. Proteolysis of mutant huntingtin produces an exon 1 fragment that accumulates as an aggregated protein in neuronal nuclei in Huntington disease. J. Biol. Chem. 285:8808–8823.

Supporting Materials and Methods and 15 figures are available at http:// www.biophysj.org/biophysj/supplemental/S0006-3495(16)30464-7.

12. Sathasivam, K., A. Neueder, ., G. P. Bates. 2013. Aberrant splicing of HTT generates the pathogenic exon 1 protein in Huntington disease. Proc. Natl. Acad. Sci. USA. 110:2366–2370.

AUTHOR CONTRIBUTIONS

13. DiFiglia, M., E. Sapp, ., N. Aronin. 1997. Aggregation of huntingtin in neuronal intranuclear inclusions and dystrophic neurites in brain. Science. 277:1990–1993.

M.C., J.R.A., S.V., S.M., and J.L. designed the research. M.C. performed AFM experiments. S.J. performed computational modeling. A.W.P. performed cell culture assays. J.R.A. performed MS and ThT assays. M.C., O.S., and J.R.A. performed lipid binding assays. M.C., S.J., A.W.P., J.R.A., S.M., S.V., and J.L. analyzed the data. M.C., J.R.A., S.M., S.V., and J.L. wrote the manuscript.

ACKNOWLEDGMENTS This research was supported in part by National Institutes of Health grant R15NS090380 (J.L.), National Science Foundation grant CHE-1454948 (S.M.), and XSEDE resources grant TG-MCB120045 (S.M.).

SUPPORTING CITATIONS

14. Gutekunst, C. A., A. I. Levey, ., S. M. Hersch. 1995. Identification and localization of huntingtin in brain and human lymphoblastoid cell lines with anti-fusion protein antibodies. Proc. Natl. Acad. Sci. USA. 92:8710–8714. 15. Gutekunst, C. A., S. H. Li, ., X. J. Li. 1999. Nuclear and neuropil aggregates in Huntington’s disease: relationship to neuropathology. J. Neurosci. 19:2522–2534. 16. Kodali, R., A. D. Williams, ., R. Wetzel. 2010. Abeta(1-40) forms five distinct amyloid structures whose beta-sheet contents and fibril stabilities are correlated. J. Mol. Biol. 401:503–517. 17. Kegel, K. B., M. Kim, ., M. DiFiglia. 2000. Huntingtin expression stimulates endosomal-lysosomal activity, endosome tubulation, and autophagy. J. Neurosci. 20:7268–7278. 18. Qin, Z. H., Y. Wang, ., M. DiFiglia. 2004. Huntingtin bodies sequester vesicle-associated proteins by a polyproline-dependent interaction. J. Neurosci. 24:269–281.

Reference (115) appears in the Supporting Material.

19. Suopanki, J., C. Go¨tz, ., E. E. Wanker. 2006. Interaction of huntingtin fragments with brain membranes–clues to early dysfunction in Huntington’s disease. J. Neurochem. 96:870–884.

REFERENCES

20. Valencia, A., P. B. Reeves, ., M. DiFiglia. 2010. Mutant huntingtin and glycogen synthase kinase 3-beta accumulate in neuronal lipid rafts of a presymptomatic knock-in mouse model of Huntington’s disease. J. Neurosci. Res. 88:179–190.

1. The Huntington’s Disease Collaborative Research Group. 1993. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes. Cell. 72:971–983. 2. Snell, R. G., J. C. MacMillan, ., D. J. Shaw. 1993. Relationship between trinucleotide repeat expansion and phenotypic variation in Huntington’s disease. Nat. Genet. 4:393–397. 3. Tobin, A. J., and E. R. Signer. 2000. Huntington’s disease: the challenge for cell biologists. Trends Cell Biol. 10:531–536.

21. Atwal, R. S., J. Xia, ., R. Truant. 2007. Huntingtin has a membrane association signal that can modulate huntingtin aggregation, nuclear entry and toxicity. Hum. Mol. Genet. 16:2600–2615. 22. Rockabrand, E., N. Slepko, ., L. M. Thompson. 2007. The first 17 amino acids of Huntingtin modulate its sub-cellular localization, aggregation and effects on calcium homeostasis. Hum. Mol. Genet. 16:61–77.

4. Legleiter, J., E. Mitchell, ., P. J. Muchowski. 2010. Mutant huntingtin fragments form oligomers in a polyglutamine length-dependent manner in vitro and in vivo. J. Biol. Chem. 285:14777–14790.

23. Burke, K. A., K. J. Kauffman, ., J. Legleiter. 2013. The interaction of polyglutamine peptides with lipid membranes is regulated by flanking sequences associated with huntingtin. J. Biol. Chem. 288:14993– 15005.

5. Vonsattel, J. P., R. H. Myers, ., E. P. Richardson, Jr. 1985. Neuropathological classification of Huntington’s disease. J. Neuropathol. Exp. Neurol. 44:559–577.

24. Jayaraman, M., R. Kodali, ., R. Wetzel. 2012. Slow amyloid nucleation via a-helix-rich oligomeric intermediates in short polyglutamine-containing huntingtin fragments. J. Mol. Biol. 415:881–899.

6. Poirier, M. A., H. Li, ., C. A. Ross. 2002. Huntingtin spheroids and protofibrils as precursors in polyglutamine fibrilization. J. Biol. Chem. 277:41032–41037.

25. Thakur, A. K., M. Jayaraman, ., R. Wetzel. 2009. Polyglutamine disruption of the huntingtin exon 1 N terminus triggers a complex aggregation mechanism. Nat. Struct. Mol. Biol. 16:380–389.

Biophysical Journal 111, 349–362, July 26, 2016 359

Chaibva et al. 26. Mishra, R., M. Jayaraman, ., R. Wetzel. 2012. Inhibiting the nucleation of amyloid structure in a huntingtin fragment by targeting a-helix-rich oligomeric intermediates. J. Mol. Biol. 415:900–917.

47. Kegel, K. B., E. Sapp, ., M. DiFiglia. 2005. Huntingtin associates with acidic phospholipids at the plasma membrane. J. Biol. Chem. 280:36464–36473.

27. Michalek, M., E. S. Salnikov, ., B. Bechinger. 2013. Membrane interactions of the amphipathic amino terminus of huntingtin. Biochemistry. 52:847–858.

48. Caviston, J. P., and E. L. F. Holzbaur. 2009. Huntingtin as an essential integrator of intracellular vesicular trafficking. Trends Cell Biol. 19:147–155.

28. Williamson, T. E., A. Vitalis, ., R. V. Pappu. 2010. Modulation of polyglutamine conformations and dimer formation by the N-terminus of huntingtin. J. Mol. Biol. 396:1295–1309.

49. Burke, K. A., K. M. Hensal, ., J. Legleiter. 2013. Huntingtin disrupts lipid bilayers in a polyQ-length dependent manner. Biochim. Biophys. Acta. 1828:1953–1961.

29. D1ugosz, M., and J. Trylska. 2011. Secondary structures of native and pathogenic huntingtin N-terminal fragments. J. Phys. Chem. B. 115:11597–11608.

50. Liu, K.-Y., Y.-C. Shyu, ., J. L. Marsh. 2015. Disruption of the nuclear membrane by perinuclear inclusions of mutant huntingtin causes cell-cycle re-entry and striatal cell death in mouse and cell models of Huntington’s disease. Hum. Mol. Genet. 24:1602– 1616.

30. Kelley, N. W., X. Huang, ., V. S. Pande. 2009. The predicted structure of the headpiece of the Huntingtin protein and its implications on Huntingtin aggregation. J. Mol. Biol. 388:919–927. 31. Coˆte´, S., G. Wei, and N. Mousseau. 2012. All-atom stability and oligomerization simulations of polyglutamine nanotubes with and without the 17-amino-acid N-terminal fragment of the Huntingtin protein. J. Phys. Chem. B. 116:12168–12179. 32. Coˆte´, S., G. Wei, and N. Mousseau. 2014. Atomistic mechanisms of huntingtin N-terminal fragment insertion on a phospholipid bilayer revealed by molecular dynamics simulations. Proteins. 82:1409– 1427. 33. Lakhani, V. V., F. Ding, and N. V. Dokholyan. 2010. Polyglutamine induced misfolding of huntingtin exon1 is modulated by the flanking sequences. PLOS Comput. Biol. 6:e1000772. 34. Nagarajan, A., S. Jawahery, and S. Matysiak. 2014. The effects of flanking sequences in the interaction of polyglutamine peptides with a membrane bilayer. J. Phys. Chem. B. 118:6368–6379. 35. Sahoo, B., D. Singer, ., R. Wetzel. 2014. Aggregation behavior of chemically synthesized, full-length huntingtin exon1. Biochemistry. 53:3897–3907. 36. Sivanandam, V. N., M. Jayaraman, ., P. C. A. van der Wel. 2011. The aggregation-enhancing huntingtin N-terminus is helical in amyloid fibrils. J. Am. Chem. Soc. 133:4558–4566. 37. Michalek, M., E. S. Salnikov, and B. Bechinger. 2013. Structure and topology of the huntingtin 1-17 membrane anchor by a combined solution and solid-state NMR approach. Biophys. J. 105:699–710. 38. Arndt, J. R., S. G. Kondalaji, ., S. J. Valentine. 2015. Huntingtin N-terminal monomeric and multimeric structures destabilized by covalent modification of heteroatomic residues. Biochemistry. 54:4285–4296. 39. Ehrnhoefer, D. E., L. Sutton, and M. R. Hayden. 2011. Small changes, big impact: posttranslational modifications and function of huntingtin in Huntington disease. Neuroscientist. 17:475–492. 40. Wang, J., C. M. Pfleger, ., G. M. Pasinetti. 2010. Potential application of grape derived polyphenols in Huntington’s Disease. Transl. Neurosci. 1:95–100.

51. Ueda, M., S. Li, ., T. Nakagawa. 2016. Expanded polyglutamine embedded in the endoplasmic reticulum causes membrane distortion and coincides with Bax insertion. Biochem. Biophys. Res. Commun. 474:259–263. 52. Coˆte´, S., V. Binette, ., N. Mousseau. 2015. Probing the Huntingtin 1-17 membrane anchor on a phospholipid bilayer by using all-atom simulations. Biophys. J. 108:1187–1198. 53. Arndt, J. R., M. Chaibva, and J. Legleiter. 2015. The emerging role of the first 17 amino acids of huntingtin in Huntington’s disease. Biomol. Concepts. 6:33–46. 54. Dorval, V., and P. E. Fraser. 2007. SUMO on the road to neurodegeneration. Biochim. Biophys. Acta. 1773:694–706. 55. Dı´az-Herna´ndez, M., A. G. Valera, ., J. J. Lucas. 2006. Inhibition of 26S proteasome activity by huntingtin filaments but not inclusion bodies isolated from mouse and human brain. J. Neurochem. 98:1585–1596. 56. Borrell-Page`s, M., D. Zala, ., F. Saudou. 2006. Huntington’s disease: from huntingtin function and dysfunction to therapeutic strategies. Cell. Mol. Life Sci. 63:2642–2660. 57. O’Rourke, J. G., J. R. Gareau, ., L. M. Thompson. 2013. SUMO-2 and PIAS1 modulate insoluble mutant huntingtin protein accumulation. Cell Reports. 4:362–375. 58. Hannoun, Z., S. Greenhough, ., D. C. Hay. 2010. Post-translational modification by SUMO. Toxicology. 278:288–293. 59. Aiken, C. T., J. S. Steffan, ., J. L. Marsh. 2009. Phosphorylation of threonine 3: implications for Huntingtin aggregation and neurotoxicity. J. Biol. Chem. 284:29427–29436. 60. Chen, S., and R. Wetzel. 2001. Solubilization and disaggregation of polyglutamine peptides. Protein Sci. 10:887–891. 61. Muchowski, P. J., G. Schaffar, ., F. U. Hartl. 2000. Hsp70 and hsp40 chaperones can inhibit self-assembly of polyglutamine proteins into amyloid-like fibrils. Proc. Natl. Acad. Sci. USA. 97:7841–7846.

41. Young, F. B., S. L. Butland, ., M. R. Hayden. 2012. Putting proteins in their place: palmitoylation in Huntington disease and other neuropsychiatric diseases. Prog. Neurobiol. 97:220–238.

62. Mendoza, V. L., M. A. Baro´n-Rodrı´guez, ., R. W. Vachet. 2011. Structural insights into the pre-amyloid tetramer of b-2-microglobulin from covalent labeling and mass spectrometry. Biochemistry. 50:6711–6722.

42. Pennuto, M., I. Palazzolo, and A. Poletti. 2009. Post-translational modifications of expanded polyglutamine proteins: impact on neurotoxicity. Hum. Mol. Genet. 18 (R1):R40–R47.

63. Burke, K. A., E. A. Yates, and J. Legleiter. 2013. Amyloid-forming proteins alter the local mechanical properties of lipid membranes. Biochemistry. 52:808–817.

43. Steffan, J. S., N. Agrawal, ., J. L. Marsh. 2004. SUMO modification of Huntingtin and Huntington’s disease pathology. Science. 304:100–104.

64. Jelinek, R., and T. Sheynis. 2010. Amyloid-membrane interactions: experimental approaches and techniques. Curr. Protein Pept. Sci. 11:372–384.

44. Cong, X., J. M. Held, ., L. M. Ellerby. 2011. Mass spectrometric identification of novel lysine acetylation sites in huntingtin. Mol. Cell. Proteomics. 10, M111.009829.

65. Kolusheva, S., L. Boyer, and R. Jelinek. 2000. A colorimetric assay for rapid screening of antimicrobial peptides. Nat. Biotechnol. 18:225–227.

45. Kegel-Gleason, K. B. 2013. Huntingtin interactions with membrane phospholipids: strategic targets for therapeutic intervention? J. Huntingtons Dis. 2:239–250.

66. Porat, Y., S. Kolusheva, ., E. Gazit. 2003. The human islet amyloid polypeptide forms transient membrane-active prefibrillar assemblies. Biochemistry. 42:10971–10977.

46. Leoni, V., and C. Caccia. 2015. The impairment of cholesterol metabolism in Huntington disease. Biochim. Biophys. Acta. 1851:1095– 1105.

67. Sokolovski, M., T. Sheynis, S. Kolusheva, and R. Jelinek. 2008. Membrane interactions and lipid binding of casein oligomers and early aggregates. Biochim. Biophys. Acta. 1778:2341–2349.

360 Biophysical Journal 111, 349–362, July 26, 2016

Htt Acetylation Alters Lipid Binding 68. Liposits, Z., I. Merchenthaler, ., A. Negro-Vilar. 1991. Morphological characterization of immortalized hypothalamic neurons synthesizing luteinizing hormone-releasing hormone. Endocrinology. 129:1575–1583. 69. Mellon, P. L., J. J. Windle, ., R. I. Weiner. 1990. Immortalization of hypothalamic GnRH neurons by genetically targeted tumorigenesis. Neuron. 5:1–10. 70. Wetsel, W. C., P. L. Mellon, ., A. Negro-Vilar. 1991. Metabolism of pro-luteinizing hormone-releasing hormone in immortalized hypothalamic neurons. Endocrinology. 129:1584–1595. 71. Wetsel, W. C., M. M. Valenc¸a, ., A. Negro-Vilar. 1992. Intrinsic pulsatile secretory activity of immortalized luteinizing hormonereleasing hormone-secreting neurons. Proc. Natl. Acad. Sci. USA. 89:4149–4153. 72. Whyte, D. B., M. A. Lawson, ., P. L. Mellon. 1995. A neuron-specific enhancer targets expression of the gonadotropin-releasing hormone gene to hypothalamic neurosecretory neurons. Mol. Endocrinol. 9:467–477. 73. Pettersen, E. F., T. D. Goddard, ., T. E. Ferrin. 2004. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 25:1605–1612. 74. DeLano, W., and S. Bromberg. 2004. PyMOL User’s Guide. DeLano Scientific LLC, San Carlos, CA. 75. Van Der Spoel, D., E. Lindahl, ., H. J. C. Berendsen. 2005. GROMACS: fast, flexible, and free. J. Comput. Chem. 26:1701–1718. 76. Cornell, W. D., P. Cieplak, ., P. A. Kollman. 1995. A second generation force field for the simulation of proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 117:5179–5197. 77. Hornak, V., R. Abel, ., C. Simmerling. 2006. Comparison of multiple Amber force fields and development of improved protein backbone parameters. Proteins. 65:712–725. 78. Sorin, E. J., and V. S. Pande. 2005. Exploring the helix-coil transition via all-atom equilibrium ensemble simulations. Biophys. J. 88:2472– 2493. 79. Papamokos, G. V., G. Tziatzos, ., E. Kaxiras. 2012. Structural role of RKS motifs in chromatin interactions: a molecular dynamics study of HP1 bound to a variably modified histone tail. Biophys. J. 102:1926– 1933. 80. Ja¨mbeck, J. P. M., and A. P. Lyubartsev. 2012. Derivation and systematic validation of a refined all-atom force field for phosphatidylcholine lipids. J. Phys. Chem. B. 116:3164–3179. 81. Ja¨mbeck, J. P. M., and A. P. Lyubartsev. 2012. An extension and further validation of an all-atomistic force field for biological membranes. J. Chem. Theory Comput. 8:2938–2948. 82. Ja¨mbeck, J. P. M., and A. P. Lyubartsev. 2013. Another piece of the membrane puzzle: extending slipids further. J. Chem. Theory Comput. 9:774–784. 83. Markoutsa, S., U. Bahr, ., B. L. Sorg. 2014. Sulfo-NHS-SS-biotin derivatization: a versatile tool for MALDI mass analysis of PTMs in lysine-rich proteins. Proteomics. 14:659–667.

89. Jass, J., T. Tja¨rnhage, and G. Puu. 2000. From liposomes to supported, planar bilayer structures on hydrophilic and hydrophobic surfaces: an atomic force microscopy study. Biophys. J. 79:3153–3163. 90. Pieri, L., K. Madiona, ., R. Melki. 2012. Fibrillar a-synuclein and huntingtin exon 1 assemblies are toxic to the cells. Biophys. J. 102:2894–2905. 91. Yip, C. M., and J. McLaurin. 2001. Amyloid-beta peptide assembly: a critical step in fibrillogenesis and membrane disruption. Biophys. J. 80:1359–1371. 92. Michalek, M., C. Aisenbrey, and B. Bechinger. 2014. Investigation of membrane penetration depth and interactions of the amino-terminal domain of huntingtin: refined analysis by tryptophan fluorescence measurement. Eur. Biophys. J. 43:347–360. 93. Atwal, R. S., C. R. Desmond, ., R. Truant. 2011. Kinase inhibitors modulate huntingtin cell localization and toxicity. Nat. Chem. Biol. 7:453–460. 94. Gu, X., E. R. Greiner, ., X. W. Yang. 2009. Serines 13 and 16 are critical determinants of full-length human mutant huntingtin induced disease pathogenesis in HD mice. Neuron. 64:828–840. 95. Thompson, L. M., C. T. Aiken, ., J. S. Steffan. 2009. IKK phosphorylates Huntingtin and targets it for degradation by the proteasome and lysosome. J. Cell Biol. 187:1083–1099. 96. Zheng, Z., A. Li, ., M. I. Diamond. 2013. An N-terminal nuclear export signal regulates trafficking and aggregation of Huntingtin (Htt) protein exon 1. J. Biol. Chem. 288:6063–6071. 97. Crick, S. L., K. M. Ruff, ., R. V. Pappu. 2013. Unmasking the roles of N- and C-terminal flanking sequences from exon 1 of huntingtin as modulators of polyglutamine aggregation. Proc. Natl. Acad. Sci. USA. 110:20075–20080. 98. Vitalis, A., and R. V. Pappu. 2011. Assessing the contribution of heterogeneous distributions of oligomers to aggregation mechanisms of polyglutamine peptides. Biophys. Chem. 159:14–23. 99. Bugg, C. W., J. M. Isas, ., R. Langen. 2012. Structural features and domain organization of huntingtin fibrils. J. Biol. Chem. 287:31739– 31746. 100. Olshina, M. A., L. M. Angley, ., D. M. Hatters. 2010. Tracking mutant huntingtin aggregation kinetics in cells reveals three major populations that include an invariant oligomer pool. J. Biol. Chem. 285:21807–21816. 101. Strehlow, A. N. T., J. Z. Li, and R. M. Myers. 2007. Wild-type huntingtin participates in protein trafficking between the Golgi and the extracellular space. Hum. Mol. Genet. 16:391–409. 102. Myre, M. A., A. L. Lumsden, ., J. F. Gusella. 2011. Deficiency of huntingtin has pleiotropic effects in the social amoeba Dictyostelium discoideum. PLoS Genet. 7:e1002052. 103. Ritch, J. J., A. Valencia, ., K. B. Kegel. 2012. Multiple phenotypes in Huntington disease mouse neural stem cells. Mol. Cell. Neurosci. 50:70–81. 104. Jacobsen, J. C., G. C. Gregory, ., J.-M. Lee. 2011. HD CAG-correlated gene expression changes support a simple dominant gain of function. Hum. Mol. Genet. 20:2846–2860.

84. Scherzinger, E., A. Sittler, ., E. E. Wanker. 1999. Self-assembly of polyglutamine-containing huntingtin fragments into amyloid-like fibrils: implications for Huntington’s disease pathology. Proc. Natl. Acad. Sci. USA. 96:4604–4609.

105. Saudou, F., S. Finkbeiner, ., M. E. Greenberg. 1998. Huntingtin acts in the nucleus to induce apoptosis but death does not correlate with the formation of intranuclear inclusions. Cell. 95:55–66.

85. Burke, K. A., J. Godbey, and J. Legleiter. 2011. Assessing mutant huntingtin fragment and polyglutamine aggregation by atomic force microscopy. Methods. 53:275–284.

106. Schilling, G., A. V. Savonenko, ., D. R. Borchelt. 2004. Nuclear-targeting of mutant huntingtin fragments produces Huntington’s diseaselike phenotypes in transgenic mice. Hum. Mol. Genet. 13:1599–1610.

86. Yates, E. A., S. L. Owens, ., J. Legleiter. 2013. Specific domains of Ab facilitate aggregation on and association with lipid bilayers. J. Mol. Biol. 425:1915–1933.

107. Drin, G., and B. Antonny. 2010. Amphipathic helices and membrane curvature. FEBS Lett. 584:1840–1847.

87. Zheng, F., Z. Wu, and Y. Chen. 2012. A quantitative method for the measurement of membrane affinity by polydiacetylene-based colorimetric assay. Anal. Biochem. 420:171–176. 88. Groves, J. T., N. Ulman, and S. G. Boxer. 1997. Micropatterning fluid lipid bilayers on solid supports. Science. 275:651–653.

108. Brass, V., E. Bieck, ., D. Moradpour. 2002. An amino-terminal amphipathic a-helix mediates membrane association of the hepatitis C virus nonstructural protein 5A. J. Biol. Chem. 277:8130–8139. 109. Georgieva, E. R., S. Xiao, ., D. Eliezer. 2014. Tau binds to lipid membrane surfaces via short amphipathic helices located in its microtubule-binding repeats. Biophys. J. 107:1441–1452.

Biophysical Journal 111, 349–362, July 26, 2016 361

Chaibva et al. 110. Cui, H., E. Lyman, and G. A. Voth. 2011. Mechanism of membrane curvature sensing by amphipathic helix containing proteins. Biophys. J. 100:1271–1279. 111. Maiuri, T., T. Woloshansky, ., R. Truant. 2013. The huntingtin N17 domain is a multifunctional CRM1 and Ran-dependent nuclear and cilial export signal. Hum. Mol. Genet. 22:1383–1394. 112. Chaibva, M., K. A. Burke, and J. Legleiter. 2014. Curvature enhances binding and aggregation of huntingtin at lipid membranes. Biochemistry. 53:2355–2365.

362 Biophysical Journal 111, 349–362, July 26, 2016

113. Ullmann, R., C. D. Chien, ., S. Muller. 2012. An acetylation switch regulates SUMO-dependent protein interaction networks. Mol. Cell. 46:759–770. 114. Jeong, H., F. Then, ., D. Krainc. 2009. Acetylation targets mutant huntingtin to autophagosomes for degradation. Cell. 137:60–72. 115. Kabsch, W., and C. Sander. 1983. Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers. 22:2577–2637.