Article pubs.acs.org/biochemistry
The Disulfide Bonds within BST‑2 Enhance Tensile Strength during Viral Tethering Kelly E. Du Pont, Aidan M. McKenzie, Oleksandr Kokhan, Isaiah Sumner, and Christopher E. Berndsen* Department of Chemistry and Biochemistry, James Madison University, Harrisonburg, Virginia 22807, United States S Supporting Information *
ABSTRACT: Human BST-2/tetherin is a host factor that inhibits the release of enveloped viruses, including HIV-1, HIV-2, and SIV, from the cell surface by tethering viruses to the host cell membrane. BST-2 has an α-helical ectodomain that forms disulfide-linked dimers between two monomers forming a coiled coil. The ectodomain contains three cysteine residues that can participate in disulfide bond formation and are critical for viral tethering. The role of the disulfides in viral tethering is unknown but proposed to be for maintaining the dimer. We explored the role of the disulfides in the structure of BST-2 using experimental, biophysical methods. To understand the role of the disulfides in viral tethering, we used a new approach in viral tethering, steered molecular dynamics. We find that the disulfides coordinate the unfolding of the BST-2 monomers, which adds tensile strength to the coiled coil. Structural differences between oxidized and reduced BST-2 are apparent during unfolding, showing the monomers slide past each other in the absence of the disulfides. We found no evidence to support dissociation of the dimer upon reduction of the disulfide bonds. Moreover, the structure of BST-2 in the absence of the disulfides is similar to that of the oxidized form of BST-2, supporting previous X-ray crystallography and cellular work that showed the disulfides are not required for expression of BST-2. These data provide new insights into viral tethering by using novel techniques in the analysis of BST-2 to give amino acid level insight into functions of BST-2.
suggesting a structural role for the disulfides beyond dimer maintenance.2 Several mechanisms of BST-2 action have been proposed; each of them requires a distinct spatial arrangement of the BST2 monomers.11,12 For a mechanism in which one monomer is in the host cell and one is in the budding virus membrane, there is an obvious role for the disulfides in keeping the dimer together as the virus diffuses away from the host cell. However, a recent study of the mechanism of BST-2 viral tethering suggests the budding viruses pull BST-2 axially.11 In this mechanism, the N- and C-termini of the BST-2 dimer are in the host cell and viral membrane, respectively, and are pulled in opposite directions.11 A role for the disulfides is not apparent in this mode of tethering. Therefore, we sought to identify a functional role for the disulfides of BST-2 in viral tethering. We find that the disulfides do not alter the structure of the coiled coil or the dimeric state. Using molecular dynamics (MD) simulations of viral tethering, we find that the disulfides increase the resistance of BST-2 to pulling by a virus. Moreover, the disulfides alter the orientation of the BST-2 monomers, which affects the unfolding pathway during viral pulling.
BST-2/tetherin/CD317 is a cellular host factor that inhibits the release of HIV-1 and other viruses from infected cells.1,2 Structurally, BST-2 consists of two transmembrane anchors linked together by an extracellular ectodomain that forms a coiled coil.3−6 Although the coiled-coil structure is required for efficient viral tethering, it has been shown that the length and composition of BST-2 can vary.1,3,7 Separately, Swiecki and coworkers, Hinz and co-workers, and Andrew and co-workers proposed that the coiled coil of BST-2 is not a rigid structure but contains a “hinge” region that permits independent motion of the N- and C-terminal ends of the ectodomain.3,6,7 The reason for this flexibility is not clear but may be to accommodate membrane curvature during viral budding.2,7 Within the N-terminus of the ectodomain (residues 50−99), three cysteines (C53, C63, and C91) form covalently linked interchain disulfide bonds.8,9 Cellular studies have previously shown that at least one disulfide bond must be present for BST2 to have viral tethering function.8−10 Hinz and co-workers suggested a dynamic interaction between BST-2 monomers that is stabilized by the presence of the disulfides.3 However, evidence of dimer exchange or a significant increase in the population of monomeric BST-2 in the absence of the disulfide bonds is lacking. Moreover, cysteine scanning mutagenesis showed that simply forming a disulfide linked dimer is insufficient for viral tethering.2 Instead, the cysteines need to be at specific positions in the N-terminal end of the ectodomain for BST-2 to be able to block viruslike particle release, © 2016 American Chemical Society
Received: December 18, 2015 Published: January 20, 2016 940
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry
■
EXPERIMENTAL PROCEDURES Protein Purification. The ectodomain of BST-2 (residues 47−166) was cloned into vector pET-15b (Genscript), which provided the open reading frame start site and added a TEV cleavable six-His tag onto the N-terminus of BST-2. BST-2 was transformed into Rosetta competent cells for expression in 2xYT broth medium containing ampicillin (60 μg/mL) and chloramphenicol (37 μg/mL). The bacterial cultures were grown at 37 °C until the OD600 reached 0.6−0.8. The cultures were induced with IPTG (0.25 g/L) for 3 h at 37 °C. Cells were harvested by centrifugation at 4000 rpm and 4 °C. The pellet was resuspended in buffer A [20 mM HEPES (pH 7.5), 100 mM NaCl, and 0.2 mM TCEP]. The cells were lysed via sonication on a FB-120 sonicator with 120 W of power (Fisher Scientific) using a 3 mm probe for two 2.5 min cycles consisting of pulses of 2 s on and 4 s off at an amplification of 75%. Lysate was clarified by centrifugation at 17500 rpm with a JA-25.50 rotor for 20 min. The supernatant was allowed to incubate for 30 min at 4 °C with HisPur Ni-NTA Resin (Thermo Scientific). The supernatant was discarded; the Ni beads were washed with buffer A, and BST-2 was eluted stepwise in buffer A with 50 to 300 mM imidazole. An 8 to 16% Tris-glycine sodium dodecyl sulfate−polyacrylamide gel electrophoresis protein gel (Thermo Scientific) was run to determine which fractions contained purified BST-2 (around 48 kDa). The protein concentration was determined by Beer’s law using an extinction coefficient of 3105 M−1 cm−1 at 280 nm in a 10 mm cell.13 Circular Dichroism. BST-2 (30 μM) in 10 mM CaCl2, 20 mM phosphate (pH 7.5), and 150 mM NaCl was placed in a 2 mm quartz cuvette, and spectra were recorded using a JASCO J-810 spectropolarimeter. Spectra of reduced BST-2 were recorded in the presence of 1 mM DTT or 1 mM TCEP. Each spectrum was an average of three accumulations from 190 to 290 nm. Melting experiments recorded the spectrum at temperatures from 10 to 85 °C in 2.5 °C increments with a 2 min equilibration time. The turbidity of each melting experiment was obtained by taking the voltage applied to the detector needed to maintain a constant signal output at 285 nm.14 The unfolding point for each condition was determined from the average of three replicates using the equation described by Kemmer and Keller.15 Molecular Dynamics Simulations. The human BST-2 structure (PDB entry 3MQB) corresponding to residues 50− 158 of the ectodomain was used to create several models in YASARA 12.7.16.5,16 Each simulation was equilibrated in explicit solvent at 310 K, 0.9% NaCl, and pH 7.4, using the AMBER ff03 force field in a simulation cell with periodic boundaries.17 Simulations were run with a time step of 1.25 fs with the temperature adjusted using a Berendsen thermostat as described by Krieger et al.16 For these simulations, the Nterminal nitrogen and C-terminal carbon were fixed to constrain the drift of the chain ends, as they would normally be attached to membrane anchors.18 Viral tethering was simulated using steered molecular dynamics (SMD) with the model created from YASARA.19,20 All SMD simulations were run with the PMEMD module of the AMBER 12 MD software package, generalized Born implicit solvent, and the AMBER ff12SB force field.21−24 Implicit solvent was used because of the decrease in the computational expense over explicit solvent. An explicit solvent calculation would require a large simulation box considering the fact that BST-2 completely unravels during the
simulations. Similar considerations are also why the lipid membrane was not included. Each simulation was run with a 2 fs time step, a nonbonded cutoff of 999.0 Å, and a temperature held at 310.0 K using a Langevin thermostat with a relaxation of 1 ps−1, and the covalent bonds to hydrogens were held fixed via SHAKE.25 During equilibration, harmonic restraints (2.4 kcal mol−1 Å−2) were placed on the α-carbon of residues S50 and Q158 on chain A and S50 and P155 on chain B to mimic the transmembrane anchors, which would reduce the level of flexing of the ectodomains.26 Two types of SMD simulations were run: constant force and constant velocity. In constant force SMD, forces of 50, 100, 150, and 300 pN were applied along a reaction coordinate defined as the distance between the centers of mass of α-carbons of residues S50 and Q158 on chain A and S50 and P155 on chain B. In constant velocity SMD, the protein is pulled along the same reaction coordinate at a constant rate via a time-dependent, harmonic potential term added to the protein force field [VFF(R)] VTot(R, t ) = VFF(R) +
1 k[x(R) − (xo − vt )]2 2
where R is the atomic coordinates, k is the harmonic force constant, x(R) is the reaction coordinate, and v is the velocity. The simulations were performed at a rate of 6 Å/ns and a force constant of 20 kcal mol−1 Å−2. All SMD trajectories were analyzed with CPPTRAJ.27 Small-Angle X-ray Scattering Analysis (SAXS). Oxidized and DTT-reduced BST-2 ectodomain SAXS data were obtained with 2−360 μM protein samples in 20 mM sodium phosphate (pH 7.2) and 150 mM sodium chloride buffer at beamline 12-ID-B of the Advanced Photon Source of Argonne National Laboratory (Argonne, IL). A Pilatus 2M detector was used. It provided a coverage range of momentum transfer q from 0.007 to 0.7 Å−1. Twenty sequential images with a 2 s exposure time per image were collected for each sample, and the scattering contribution from buffer was subtracted. To minimize X-ray damage, samples were continuously refreshed in a flow cell during data collection. After Guinier fitting in Primus,28 the distance distribution function for BST-2 was obtained via GNOM.29 Radius of gyration (Rg) measurements were taken for both oxidized and reduced BST-2 ectodomain and plotted in Microsoft Excel.
■
RESULTS Disulfide Bonds Are Not Required for Maintenance of the BST-2 Dimer. We investigated the oligomeric state of BST-2 under reducing conditions by small-angle X-ray scattering. SAXS can show the shape, conformation, and oligomeric state of a protein in solution.30 Given the general dimerization mechanism of coiled coils, where preformed helical structures within a larger unfolded region trigger the two subunits to “zip up”, we would expect a partial or total loss of helical structure if reducing agents alter the association state which would be seen as a change in the radius of gyration.31−34 The oxidized BST-2 and DTT-reduced BST-2 had similar Rg values over a concentration range of 5−380 μM (Figure 1A). The large errors bars at low concentrations of BST-2 are a result of a low signal-to-noise ratio due to a much stronger relative contribution from the solvent. LC−MS of BST-2 samples after SAXS collection confirmed the reduced samples contained mostly reduced BST-2 (data not shown). These data indicate that the disulfides were reduced in our SAXS experiments, and the absence of changes in the Rg is not due 941
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry
Disulfide Bonds Are Not Necessary for the α-Helical Structure of BST-2. To investigate the role of the disulfides in maintaining the structure of BST-2, we determined the secondary structure via circular dichroism and observed changes with heating. Our data match the previous circular dichroism data on shorter constructs of BST-2 and show a predominantly α-helical structure that is not affected by the presence of DTT or TCEP (Supplemental Figure 1).3,4,6 The ratio of ellipticities at 222 nm and 208 nm is approximately 1.1 for all conditions, which is consistent with a coiled-coil structure, and this value is independent of whether the disulfides are reduced.35 We then determined the stability of the ectodomain by observing the change in ellipticity at 222 nm over a temperature range from 10 to 85 °C (Table 1). The Table 1. Summary of Melting Temperature (Tm) and Turbidity50 Values form of BST-2
Tm (°C)
turbidity50 (°C)
oxidized reduced with DTT reduced with TCEP
61.5 ± 0.2 45.8 ± 0.4 47.4 ± 0.6
66.3 67.6 67.9
oxidized BST-2 unfolded at 61.5 ± 0.2 °C, while the unfolding point of BST-2 in the presence of DTT was at 45.8 ± 0.4 °C and with TCEP was at 47.4 ± 0.6 °C (Table 1). The 222 nm/ 208 nm ratio is ≥1 until ∼50 °C, when the measurement error increases and the average value of the ratio begins to deviate, suggesting a change in the folded state (Figure 2A). Our observed melting temperature for the reduced form of BST-2 is higher than previously published values, while that of the oxidized form matches previous measurements. We attribute this difference to using the complete ectodomain (residues 50− 166) rather than truncated protein as truncated coiled coils are known to be less stable.3,6,33,34 However, the conclusion that the disulfides stabilize the folded state of BST-2 is consistent between our studies and previous work.3,6 The consistency of the analysis presented above of far-UV CD data with a coiled-coil structure and previous melting temperature analysis gave us confidence to further analyze our CD data in the near-UV range and to monitor changes in turbidity. The change in voltage at 285 nm is a measure of sample turbidity and is indicative of aggregation of unfolded protein (Figure 2B).14,36 Oxidized BST-2 shows an increase in turbidity at 66 °C, while the reduced forms show increases in the turbidity value between 67 and 68 °C (Table 1). These data further support the SAXS and simulation data, indicating that the disulfides do not mediate dimerization of the BST-2 monomers, as we would expect unfolded BST-2 monomers to aggregate. Changes in near-UV ellipticity are related to exposure of aromatic amino acid side chains to the solvent and other side chains that can alter electronic properties of the aromatic ring.37,38 BST-2 has two regions with aromatic amino acids (Figure 1C). Between C63 and C91 in the ectodomain is F81, and at the C-terminal end of the ectodomain coiled coil are Y153 and Y154. Comparison of the change in ellipticity at 260 nm, reflecting the environment around F81, showed a decrease in ellipticity for reduced and oxidized BST-2 from 25 to 60 °C, followed by an increase in the value between 65 and 85 °C (Figure 2C). These data suggest unfolding of the coiled coil near F81 regardless of the presence of the disulfides and at a temperature lower than that of global unfolding of the α-helices
Figure 1. Disulfides are not necessary for dimerization. (A) Rg vs concentration plot for oxidized BST-2 (black) and BST-2 reduced with 1 mM DTT (gray). (B) Distance vs time plot from simulations of reduced BST-2 at cysteines 53 (red), 63 (blue), and 91 (black). (C) Structure of the BST-2 coiled coil showing the location of the disulfides (yellow), phenylalanine 81 (red), and tyrosines 153 and 154 (green). Structure taken from PDB entry 3MQB.5
to the nonreduced disulfides that persist in the presence of DTT. The absence of a significant change in Rg value indicates that there are not large changes in the molecular volume and suggests that the reduced BST-2 dimer does not dissociate into monomers. BST-2 Dimers Are Stable Oligomers. To further investigate if the disulfides stabilized the dimeric structure of BST-2, we simulated the dynamics of reduced BST-2. We then measured the distance between the cysteines (C53, C63, and C91) over a 200 ns simulation (Figure 1B). We observed no consistent increase in the distance between the two BST-2 chains, indicating the dimer does not drift apart in the absence of interchain disulfide bonds. The spikes in distance for the C53 data (e.g., at 125 and 138 ns) are indicative of the N-terminus unwinding during the simulation, allowing the residues to drift apart. However, the data at C63 do not follow the pattern of increased distance observed with C53, suggesting this is a localized separation of the N-termini but not a separation of the two monomers (Figure 1B). These data support our SAXS data that show that the disulfides are not changing the oligomeric state of BST-2 but serve another role in viral tethering. 942
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry
data between 65 and 85 °C suggests a burial or change in structure, which given our turbidity data in Figure 2A suggests this change in ellipticity is due to BST-2 aggregation. Taken together, our spectroscopic data suggest that the disulfides stabilize the helical structure of the coiled coil, and given the observed changes in structure around F81, the disulfides may reduce the spread of helix unfolding from this region during thermal denaturation. Disulfides Increase the Resistance to Unfolding. As a viral particle buds from the cell, BST-2 would have to resist unfolding to slow the release of the virus from the cell membrane. Thus, we hypothesized a possible function for the disulfides is to increase the stiffness of the coiled coil. To address whether the disulfides of BST-2 strengthen the coiled coil during viral tethering, we simulated viral tethering by using SMD to pull on the ends of the BST-2 dimer according to the mechanism proposed by Venkatesh and Bieniasz.11 In this mechanism, the N-terminal transmembrane anchor of BST-2 is in the host cell membrane while the C-terminal membrane anchor is in the virus membrane. The ectodomain bridges the gap between the two membranes that has been directly observed by cryo-electron tomography.10 SMD is an ideal tool to use, because monitoring the unfolding process at the atomic level is difficult by other means. We first compared our experimental CD data on oxidized BST-2 to a predicted CD spectrum of 93% α-helix, 5% coil, and 2% other based on the A and B monomers of the BST-2 structure 3MQB (Supplemental Figure 2A).5,39 The correlation for the two spectra is 0.99 for the 200−240 nm range, where the α-helical structure predominates. These data indicate the starting structure for the simulations is very similar to the structure of BST-2 in solution and allows for comparison between solution and computational methods. We simulated viral tethering initially using constant force simulations to determine the resistance of BST-2 to pulling (Figure 3A). Oxidized and reduced BST-2 were pulled at constant forces of 50, 100, 150, and 300 pN to obtain information about the resistance of BST-2 being pulled by a tethered virus (Figure 3B−E). These forces are typical for unfolding simulations of other coiled-coil structures.40 Oxidized and reduced BST-2 showed similar resistance to the constant 300 pN force (Figure 3B). However, we note that 300 pN is above the threshold to pull an α-helix out of the membrane and to unravel an α-helix.32,41 The predicted CD spectrum for simulated unfolded BST-2 and the experimental CD spectrum of BST-2 at 80 °C have a correlation of 0.98, indicating the simulated and unfolded BST-2 have similar secondary structure content (Supplemental Figure 2B). At 150 pN of force, the reduced BST-2 unfolded at a rate higher than that of the oxidized BST-2 as shown by the difference in the normalized distance between 0 and 15 ns (Figure 3C). The rapid increase in distance for the oxidized data at 15 ns is due to the unfolding of residues 106−120 in the second and third heptad of the coiled coil. We repeated the 150 pN force simulation using a different starting structure and observed a similar trend, where the oxidized form of BST-2 is more stable than the reduced form (Supplemental Figure 3). At 100 pN, the oxidized form of BST-2 reached a stable conformation after simulation for 70 ns, and after 130 ns, the reduced form of BST-2 extended 60 Å longer than the oxidized form (Figure 3D). At 50 pN of pulling force, both forms of BST-2 extend ∼20 Å with the oxidized form initially trending below the reduced form (Figure 3E). However, these differences observed at this force are small.
Figure 2. (A) Plot of the 222 nm/208 nm ratio for oxidized (blue) and DTT-reduced (red) BST-2. Data for TCEP are not shown for the sake of simplicity. (B) Turbidity plots for the oxidized BST-2 () and reduced BST-2 with either DTT (···) or TCEP (---). (C) Circular dichroism of BST-2 showing ellipticity at 260 nm (red lines) and 280 nm (blue lines). Data for oxidized BST-2 are shown as a solid line, while data for BST-2 reduced with TCEP or DTT are shown as dashed lines. Data shown for all experiments are averages of two or three experiments. Error bars are not shown for the sake of simplicity.
(Table 1). Analysis of the ellipticity at 280 nm reflecting the environment around the tyrosine residues shows little change in the value between 25 and 60 °C, followed by a rapid increase between 65 and 75 °C and then a plateau between 75 and 85 °C (Figure 2C). The trend of the tyrosine and phenylalanine 943
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry
Figure 3. (A) Diagram showing the experimental setup for pulling experiments and the direction of force. Distance vs time plots for pulling of BST2. (B−E) Normalized distance vs time plots at 300, 150, 100, and 50 pN, respectively, for the oxidized (black) and reduced (gray) forms of BST-2.
were undergoing significant dissociation.33,34 Hinz and coworkers proposed the presence of the interchain disulfide bonds reduces the rate of BST-2 dimer dissociation.3 However, we found that the dimer remains intact regardless of the presence or absence of the disulfide bonds (Figures 1 and 2A). We do not see differences in the Rg values of reduced BST-2 representing the solution ensemble of monomeric and dimeric proteins, suggesting a stable dimer. Furthermore, we do not see significant changes in the turbidity, Θ222/Θ208, and temperature-induced aggregation, suggesting the dimer is not dissociating (Figure 2B and Table 1). If F81 were solventexposed (such as would be expected in a dissociated dimer), we would expect no change in the ellipticity until the protein had aggregated. However, we observe an overall decrease in the ellipticity at 260 nm regardless of the reduced state of the disulfides, suggesting an unfolding of the helix around this site followed by aggregation (Figure 2C). Thus, we conclude the critical role of the disulfides in viral tethering is not to maintain the dimeric structure of BST-2 but another function in viral tethering. These data support previous X-ray crystal structures of BST-2 with a C3A mutation which is dimeric and cellular studies showing the disulfides are required for function but not expression.2,5 Under both oxidizing and reducing conditions, the α-helical structure is maintained, but when the disulfides are reduced, we observe a 15 °C reduction in the unfolding temperature (Table 1). Similar observations have been reported with shorter constructs of the BST-2 ectodomain, suggesting the length of the ectodomain plays a lesser role in the stability of the coiled coil, which is supported by cellular data.3,6,7 We further observed that during thermal denaturation the residues around F81 unfold followed by helix unwinding leading to aggregation of the protein. Only helix unfolding was affected by the presence of the disulfides, suggesting that the disulfides limit the spreading of unfolding that occurs in the middle of the ectodomain.
The differences in the unfolding rate and distance at 100 pN led us to examine the dihedral angles during the simulation at this force. The ψ dihedral angle for an α-helix is typically −40° to −60°, and ψ angles outside of this range indicate a conversion to a β-strand structure.42 Therefore, observation of the number of changes in the dihedral angles with time can show when and where the α-helix is unwinding. By 130 ns with 100 pN of force, 75% of the ψ angles in the reduced form are greater than 90° while this number is 45% in oxidized BST-2 (Figure 4A). Per residue comparison of the ψ angles shows that unwinding of the helix starts around residue 80 and spreads toward the C-terminus in the reduced form (Figure 4B,C). This spreading of helix unwinding, however, is not present in oxidized BST-2. We observe a similar trend upon comparison of the hydrogen bonding between the reduced and oxidized forms in constant velocity simulations (Supplemental Figure 4). Comparing the per residue ψ angles between the monomers of the BST-2 dimer in each form can show whether each subunit unfolds similarly. We find that the ψ angles evolve similarly in each subunit in oxidized BST-2 (Figure 4B), while there are distinct differences between the subunits in the reduced form, suggesting that the monomers are “sliding past” each other as one monomer unfolds followed by the other (Figure 4C). Therefore, our simulations and spectroscopy suggest the disulfides coordinate the unfolding of the BST-2 dimer during tethering, resulting in increased tensile strength.
■
DISCUSSION The three disulfide bonds within BST-2 are key for viral tethering, and at least one of the three disulfides must be present for function.8 However, the role of the disulfides in viral tethering is still unknown. Previous work on coiled coils and truncations of helical proteins suggest the isolated helices or subunits of the coiled coils are unstable; thus, we would expect a reduction in the helical structure content if reduced BST-2 944
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry
before to characterize viral tethering. In our simulations we assumed that (1) the tethering mechanism proposed by Venkatesh and Bieniasz is accurate, (2) BST-2 functions as a single dimer,10 (3) the cysteines in BST-2 are either all in disulfides or all reduced, and (4) only protein−water interactions occur during tethering; thus, BST-2 is not interacting with a lipid membrane in the ectodomain (Figure 3A).11 These assumptions create a simple model that is wellfounded in the current literature on viral tethering.10,11,43 Moreover, the distance between the host cell and viral particle did not appear to be influenced by cell type or the type of viral particle, suggesting the lipid membrane plays no role in tethering.10 The N−C length of the fully unfolded ectodomain in the simulations was ∼360 Å, which corresponds well with a recent in vivo study showing the distance between tethered viruses and the host cell to be between 7 and 37 nm.10 The mean distance between tethered membranes was between 15.3 and 19.9 nm with a standard deviation of 4.19−7.24 depending on the cell type and virion particle.10 In simulations with 100 pN of force, we observed the maximal length of the oxidized form of BST-2 was extended to ∼240 Å or 24 nm, which is within the error of the mean values of the cellular study (Figure 3D). The reduced form of BST-2 extended to nearly 300 Å or 30 nm, which is outside the error of the mean values of Strauss and co-workers (Figure 3D).10 The correlation among our simulations, our biophysical characterization of the BST-2 ectodomain, and the cellular studies shows the utility of computational simulations for understanding viral tethering and suggests the conclusions we draw regarding the unfolding of BST-2 during tethering are likely to reflect the cellular structure and function of BST-2. In simulations of reduced and oxidized BST-2 applying greater than 150 pN of force, the coiled coil was unfolded by the end of the simulation. In the oxidized BST-2 simulations at or below 100 pN, the coiled coil is stable, suggesting this is the upper limit of strength in the coiled-coil region with the disulfides present (Figure 3). In the absence of the disulfides, the coiled coil unfolds at forces between 50 and 100 pN as shown by the minute change in distance in the 50 pN simulations while at 100 pN there is a large change in distance indicative of unfolding (Figure 3D). Regardless of the force applied, reduced BST-2 stretched faster than oxidized BST-2 and the unfolding of the subunits of BST-2 was uncoordinated as shown by the differences in ψ angles that indicate one monomer of BST-2 is unfolding faster than the other (Figures 3 and 4). These data show that the disulfides prevent the individual subunits of the BST-2 dimer from slipping past each other, which increases the stability of the protein structure. Our analysis of the ψ angles shows rapid unfolding in the Nterminus of BST-2 between residues 80 and 110, suggesting an area of flexibility, which is supported by our spectroscopic observations of the local environment around F81 during unfolding (Figures 2B and 4B,C). Unfolding in this region in the reduced BST-2 spreads toward the C-terminus, suggesting local unfolding in the hinge region leads to destabilization of the coiled coil (Figure 4C).7 This spreading of local unfolding is not observed in the oxidized BST-2, supporting the idea that the disulfides limit the spreading of local unfolding (Figure 4B). Our unfolding data are similar to what has been observed for the β-myosin S2 domain coiled coil, where global unfolding of both subunits occurs only when local unfolding in one subunit spreads to several other residues in the same chain.40 This effect, as with BST-2 in this study, is dependent on the amount
Figure 4. Dihedral analysis of a 100 pN pulling simulation. (A) Percent of ψ angles exceeding 90° for reduced (gray) and oxidized (black) BST-2. Heat map showing the per residue dihedral angle for ψ changes with time for the simulation in panel A of oxidized (B) and reduced (C) BST-2. The ψ angle for a typical α-helix is −40° to −60° (light blue) with a value of >90° corresponding to β-strand structures that are colored red.42 The different length X-axes are due to different length chains in the BST-2 coiled coil from the X-ray crystal structure.
We simulated viral tethering through steered molecular dynamics (SMD) to observe virus-induced unfolding of BST-2 at the atomic level and to determine how the disulfides affect the tensile strength of BST-2. This approach has not been used 945
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry
operated for the DOE Office of Science by Argonne National Laboratory under Contract DE-AC02-06CH11357.
and duration of the force. Thus, when the subunits begin to slip past each other, global unfolding is likely to happen. In BST-2, the disulfides prevent sliding of the monomers past each other, resulting in limited unfolding. This effect is especially prevalent in the region to the C-terminal side of F81. While this region unfolds similarly regardless of the oxidized or reduced state of BST-2 in simulation and in thermal melts, there are clear differences in melting temperature and the spreading of the unfolding down the coiled coil (Figure 4, Table 1, and Supplemental Figure 4). In oxidized BST-2, this spreading is limited by the intersubunit disulfides until the force (or temperature) exceeds a critical load, and then unfolding occurs. Much of the unfolding in the reduced ectodomain of BST-2 centers on residues 80−110 (Figure 4 and Supplemental Figure 4). Our previous analysis of the B factors in the BST-2 ectodomain crystal structures shows that some of the crystal structures have higher than average flexibility in this region.2 Moreover, independent alanine scanning and cysteine scanning mutagenesis show little or mixed effects of mutations in this region on structure and function, while some deletions of this region do show functional effects.1−3,7,11 Thus, this region of BST-2 is clearly flexible, but the biological role of the flexibility is not clear. However, our results and those of others show a clear need for the disulfides in tethering that we attribute to limiting the unfolding of BST-2.1,2,8 BST-2 must adjust rapidly to the fluid nature of the cell membrane and undergo the transition from being parallel to the cell surface to being perpendicular to the membrane during viral budding and tethering, indicating that a rigid structure may not be favorable for function. Whether the disulfides affect the dynamics of BST-2 and the structural transition is not clear. However, our studies show that during tethering, when a more rigid structure is helpful, limiting the unfolding of the two chains of the BST-2 dimer linked by the interchain disulfides limits the rate of unfolding of the protein.
■
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank Dr. Klaus Strebel for helpful comments during the drafting of this manuscript and Dr. Chrisi Hughey for help with the MS of BST-2. We gratefully acknowledge the help of Dr. Xiaobing Zuo and the staff of sector 12 of the Advanced Photon Source at Argonne National Laboratory.
■
ABBREVIATIONS SMD, steered molecular dynamics; TEV, tobacco etch virus; IPTG, isopropyl β-D-1-thiogalactopyranoside; PDB, Protein Data Bank; BST-2, bone marrow stromal cell antigen 2.
■
(1) Hammonds, J., Ding, L., Chu, H., Geller, K., Robbins, A., Wang, J.-J., Yi, H., and Spearman, P. (2012) The tetherin/BST-2 coiled-coil ectodomain mediates plasma membrane microdomain localization and restriction of particle release. J. Virol. 86, 2259−2272. (2) Welbourn, S., Kao, S., Du Pont, K. E., Andrew, A. J., Berndsen, C. E., and Strebel, K. (2015) Positioning of cysteine residues within the N-terminal portion of the BST-2/Tetherin ectodomain is important for functional dimerization of BST-2. J. Biol. Chem. 290, 3740−3751. (3) Hinz, A., Miguet, N., Natrajan, G., Usami, Y., Yamanaka, H., Renesto, P., Hartlieb, B., McCarthy, A. a., Simorre, J. P., Göttlinger, H., and Weissenhorn, W. (2010) Structural basis of HIV-1 tethering to membranes by the BST-2/tetherin ectodomain. Cell Host Microbe 7, 314−323. (4) Schubert, H. L., Zhai, Q., Sandrin, V., Eckert, D. M., Garcia-Maya, M., Saul, L., Sundquist, W. I., Steiner, R. a, and Hill, C. P. (2010) Structural and functional studies on the extracellular domain of BST2/ tetherin in reduced and oxidized conformations. Proc. Natl. Acad. Sci. U. S. A. 107, 17951−17956. (5) Yang, H., Wang, J., Jia, X., McNatt, M. W., Zang, T., Pan, B., Meng, W., Wang, H.-W., Bieniasz, P. D., and Xiong, Y. (2010) Structural insight into the mechanisms of enveloped virus tethering by tetherin. Proc. Natl. Acad. Sci. U. S. A. 107, 18428−18432. (6) Swiecki, M., Scheaffer, S. M., Allaire, M., Fremont, D. H., Colonna, M., and Brett, T. J. (2011) Structural and biophysical analysis of BST-2/tetherin ectodomains reveals an evolutionary conserved design to inhibit virus release. J. Biol. Chem. 286, 2987−2997. (7) Andrew, A. J., Berndsen, C. E., Kao, S., and Strebel, K. (2012) The size and conservation of a coiled-coil structure in the ectodomain of human BST-2/tetherin is dispensable for inhibition of HIV-1 virion release. J. Biol. Chem. 287, 44278−44288. (8) Andrew, A. J., Miyagi, E., Kao, S., and Strebel, K. (2009) The formation of cysteine-linked dimers of BST-2/tetherin is important for inhibition of HIV-1 virus release but not for sensitivity to Vpu. Retrovirology 6, 80. (9) Perez-Caballero, D., Zang, T., Ebrahimi, A., McNatt, M. W., Gregory, D. a, Johnson, M. C., and Bieniasz, P. D. (2009) Tetherin inhibits HIV-1 release by directly tethering virions to cells. Cell 139, 499−511. (10) Strauss, J. D., Hammonds, J. E., Yi, H., Ding, L., Spearman, P., and Wright, E. R. (2016) Three-Dimensional Structural Characterization of HIV-1 Tethered to Human Cells. J. Virol. 90, 1507. (11) Venkatesh, S., and Bieniasz, P. D. (2013) Mechanism of HIV-1 Virion Entrapment by Tetherin. PLoS Pathog. 9, e1003483. (12) Neil, S. (2013) The Antiviral Activities of Tetherin. In Intrinsic Immunity (Cullen, B. R., Ed.) Vol. 371, pp 67−104. (13) Grimsley, G. R., and Pace, C. N. (2004) Spectrophotometric determination of protein concentration. Current Protocols in Protein Science, Chapter 3, Unit 3.1, Wiley, New York.10.1002/ 0471140864.ps0301s33
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biochem.5b01362. Melting curves for oxidized and reduced BST-2, comparison of experimental and simulated CD spectra for folded and unfolded BST-2, comparison of two normalized distance versus time plots for oxidized and reduced BST-2 unfolded at 150 pN, and comparison of hydrogen bonding for oxidized and reduced BST-2 constant velocity SMD (PDF)
■
REFERENCES
AUTHOR INFORMATION
Corresponding Author
*Department of Chemistry and Biochemistry, James Madison University, Harrisonburg, VA 22807. E-mail: berndsce@jmu. edu. Telephone: (540) 568-2355. Funding
This work was supported by a NSF-REU grant (CHE1461175), a Research Corporation Department Development Award (7957), the James Madison University Program of Grants for Faculty Assistance, and the 4-VA organization. This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility 946
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
Article
Biochemistry (14) Metrick, M. a., Temple, J. E., and Macdonald, G. (2013) The effects of buffers and pH on the thermal stability, unfolding and substrate binding of RecA. Biophys. Chem. 184, 29−36. (15) Kemmer, G., and Keller, S. (2010) Nonlinear least-squares data fitting in Excel spreadsheets. Nat. Protoc. 5, 267−281. (16) Krieger, E., Darden, T., Nabuurs, S. B., Finkelstein, A., and Vriend, G. (2004) Making optimal use of empirical energy functions: force-field parameterization in crystal space. Proteins: Struct., Funct., Genet. 57, 678−683. (17) Duan, Y., Wu, C., Chowdhury, S., Lee, M. C., Xiong, G., Zhang, W., Yang, R., Cieplak, P., Luo, R., Lee, T., Caldwell, J. W., Wang, J., and Kollman, P. A. (2003) A point-charge force field for molecular mechanics simulations of proteins based on condensed-phase quantum mechanical calculations. J. Comput. Chem. 24, 1999−2012. (18) Andrew, A. J., Kao, S., and Strebel, K. (2011) C-terminal hydrophobic region in human bone marrow stromal cell antigen 2 (BST-2)/tetherin protein functions as second transmembrane motif. J. Biol. Chem. 286, 39967−39981. (19) Lu, H., Isralewitz, B., Krammer, a, Vogel, V., and Schulten, K. (1998) Unfolding of titin immunoglobulin domains by steered molecular dynamics simulation. Biophys. J. 75, 662−671. (20) Isralewitz, B., Gao, M., and Schulten, K. (2001) Steered molecular dynamics and mechanical functions of proteins. Curr. Opin. Struct. Biol. 11, 224−230. (21) Goetz, a W., Williamson, M. J., Xu, D., Poole, D., Le Grand, S., and Walker, R. C. (2012) Routine microsecond molecular dynamics simulations with amber - part i: Generalized born. J. Chem. Theory Comput. 8, 1542−1555. (22) Hornak, V., Abel, R., Okur, A., Strockbine, B., Roitberg, A., and Simmerling, C. (2006) Comparison of multiple amber force fields and development of improved protein backbone parameters. Proteins: Struct., Funct., Genet. 65, 712. (23) Case, D. A., Darden, T. A., Cheatham, T. E., III, Simmerling, C. L., Wang, J., Duke, R. E., Luo, R., Walker, R. C., Zhang, W., Merz, K. M., Roberts, B., Hayik, S., Roitberg, A., Seabra, G., Swails, J., Goetz, A. W., Kolossváry, I., Wong, K. F., Paesani, F., Vanicek, J., Wolf, R. M., and Kollman, P. A. (2012) AMBER 12, University of California, San Francisco. (24) Nguyen, H., Roe, D. R., and Simmerling, C. (2013) Improved generalized born solvent model parameters for protein simulations. J. Chem. Theory Comput. 9, 2020−2034. (25) Ryckaert, J.-P., Ciccotti, G., and Berendsen, H. J. C. (1977) Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput. Phys. 23, 327−341. (26) Ilan, B., Tajkhorshid, E., Schulten, K., and Voth, G. a. (2004) The mechanism of proton exclusion in aquaporin channels. Proteins: Struct., Funct., Genet. 55, 223−228. (27) Roe, D. R., and Cheatham, T. E., III (2013) PTRAJ and CPPTRAJ: software for processing and analysis of molecular synamics trajectory data. J. Chem. Theory Comput. 9, 3084−3095. (28) Konarev, P. V., Volkov, V. V., Sokolova, A. V., Koch, M. H. J., Svergun, D. I., and Koch, H. J. (2003) PRIMUS: a Windows PC-based system for small-angle scattering data analysis. J. Appl. Crystallogr. 36, 1277−1282. (29) Semenyuk, A. V., and Svergun, D. I. (1991) Gnom - a Program Package for Small-Angle Scattering Data-Processing. J. Appl. Crystallogr. 24, 537−540. (30) Putnam, C. D., Hammel, M., Hura, G. L., and Tainer, J. a. (2007) X-ray solution scattering (SAXS) combined with crystallography and computation: defining accurate macromolecular structures, conformations and assemblies in solution. Q. Rev. Biophys. 40, 191− 285. (31) Steinmetz, M. O., Jelesarov, I., Matousek, W. M., Honnappa, S., Jahnke, W., Missimer, J. H., Frank, S., Alexandrescu, A. T., and Kammerer, R. A. (2007) Molecular basis of coiled-coil formation. Proc. Natl. Acad. Sci. U. S. A. 104, 7062−7067.
(32) Hamed, E., and Keten, S. (2014) Hierarchical cascades of instability govern the mechanics of coiled coils: Helix unfolding precedes coil unzipping. Biophys. J. 107, 477−484. (33) Reymond, M. T., Merutka, G., Dyson, H. J., and Wright, P. E. (1997) Folding propensities of peptide fragments of myoglobin. Protein Sci. 6, 706−716. (34) Saudek, V., Pasley, H. S., Gibson, T., Gausepohl, H., Frank, R., and Pastore, a. (1991) Solution structure of the basic region from the transcriptional activator GCN4. Biochemistry 30, 1310−1317. (35) Lau, S. Y. M., Taneja, A. K., and Hodges, R. S. (1984) Synthesis of a Model Protein of Defined Secondary and Quaternary Structure. J. Biol. Chem. 259, 13253−13261. (36) Benjwal, S., Verma, S., Röhm, K.-H., and Gursky, O. (2006) Monitoring protein aggregation during thermal unfolding in circular dichroism experiments. Protein Sci. 15, 635−639. (37) Kelly, S. M., and Price, N. C. (2000) The use of circular dichroism in the investigation of protein structure and function. Curr. Protein Pept. Sci. 1, 349−384. (38) Kelly, S. M., Jess, T. J., and Price, N. C. (2005) How to study proteins by circular dichroism. Biochim. Biophys. Acta, Proteins Proteomics 1751, 119−139. (39) Abriata, L. a. (2011) A simple spreadsheet program to simulate and analyze the far-UV circular dichroism spectra of proteins. J. Chem. Educ. 88, 1268−1273. (40) Kreuzer, S. M., and Elber, R. (2013) Coiled-coil response to mechanical force: Global stability and local cracking. Biophys. J. 105, 951−961. (41) Ganchev, D. N., Rijkers, D. T. S., Snel, M. M. E., Killian, J. A., and de Kruijff, B. (2004) Strength of Intergration of Transmembrane α-helical Peptides in Lipid Bilayers As Determined by Atomic Force Spectroscopy. Biochemistry 43, 14987−14993. (42) Carugo, O., and Djinovic-Carugo, K. (2013) Half a century of Ramachandran plots. Acta Crystallogr., Sect. D: Biol. Crystallogr. 69, 1333−1341. (43) Hammonds, J., Wang, J. J., Yi, H., and Spearman, P. (2010) Immunoelectron microscopic evidence for tetherin/BST2 as the physical bridge between HIV-1 virions and the plasma membrane. PLoS Pathog. 6, e1000749.
947
DOI: 10.1021/acs.biochem.5b01362 Biochemistry 2016, 55, 940−947
