Hydrophobic segment of dengue virus C protein. Interaction with model membranes

Abstract

Dengue virus (DENV) C protein is essential for viral assembly. DENV C protein associates with intracellular membranes through a conserved hydrophobic domain and accumulates around endoplasmic reticulum-derived lipid droplets which could provide a platform for capsid formation during assembly. In a previous work we described a region in DENV C protein which induced a nearly complete membrane rupture of several membrane model systems, which was coincident with the theoretically predicted highly hydrophobic region of the protein. In this work we have carried out a study of the binding to and interaction with model biomembranes of a peptide corresponding to this DENV C region, DENV2_C6. We show that DENV2_C6 partitions into phospholipid membranes, is capable of rupturing membranes even at very low peptide-to-lipid ratios and its membrane-activity is modulated by lipid composition. These results identify an important region in the DENV C protein which might be directly implicated in the DENV life cycle through the modulation of membrane structure.

Keywords::

• Membrane protein
• viral protein
• viral peptide

Introduction

There are three genera in the Flaviviridae family: Flavivirus, Hepacivirus and Pestivirus. Dengue virus (DENV) is sorted into the Flavivirus genus and it is the leading cause of arboviral diseases in the tropical and subtropical regions, with an estimation of 390 million dengue infections per year (Bhatt et al. 2013). DENV comprises four serologically and genetically related viruses, DENV viruses 1–4, which possess 69–78% identity at the amino acid level (Urcuqui-Inchima et al. 2010). DENV infections might be either asymptomatic or result in what is known as dengue fever; some individuals develop a severe and potentially life-threatening disease known as dengue haemorrhagic fever or dengue shock syndrome, leading to more than 25,000 deaths per annum. Despite the urgent medical need and considerable efforts, no antivirals or vaccines against DENV virus are currently available, so that more than 2 billion people, mainly in less developed countries, are at risk in the world (Pastorino et al. 2010). Furthermore, due to the increased global temperature and travelling, there is a real risk of dispersion of the mosquito vector to previously unaffected zones. Although several compounds have been identified to inhibit DENV replication (Noble et al. 2010), there is actually no clinical treatment for DENV infection.

DENV is a positive-sense, single-stranded RNA virus with approximately 10.7 kb. It contains untranslated regions both at the 5′ and 3′ ends, flanking a single open reading frame (ORF) encoding a polyprotein of over 3000 amino acids, which is subsequently cleaved by cellular and viral proteases into three structural proteins, C, prM and E, and seven non-structural proteins, NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5 (Perera and Kuhn 2008). Similarly to other enveloped viruses, the DENV virus enters the cells through receptor mediated endocytosis (Bressanelli et al. 2004, Mukhopadhyay et al. 2005, Kielian and Rey 2006, Perera and Kuhn 2008) and rearranges internal cell membranes to establish specific sites of replication (Miller et al. 2007, Miller and Krijnse-Locker 2008, Welsch et al. 2009). Details about DENV replication process remain largely unclear, but most if not all of the DENV proteins are involved and function in a complex web of protein-protein and protein-lipid interactions. The mature DENV virus has a capsid (C) protein core complexed with the RNA genome (forming the nucleocapsid), surrounded by a host-derived lipid bilayer in which multiple copies of the viral envelope (E) and membrane (M) proteins are embedded.

The C proteins of Flaviviridae are dimeric, basic, have an overall helical fold and are responsible for genome packaging. These proteins are essential for viral assembly in order to ensure specific encapsidation of the viral genome. Nonetheless, little is known about the recognition of the viral RNA by protein C or RNA packaging. Furthermore, protein C seems to associate with intracellular membranes through a conserved hydrophobic domain (Markoff et al. 1997). Recently, it has been found that protein C accumulates around endoplasmic reticulum (ER) derived lipid droplets (Samsa et al. 2009). Similarly to other enveloped viruses, DENV replicates its genome in a membrane-associated replication complex, and morphogenesis and virion budding has been suggested to take place in the ER or modified ER membranes. These modified membranes could provide a platform for capsid formation during viral assembly (Samsa et al. 2009). Although Flaviviridae C proteins are shorter than the Hepacivirus core proteins, their roles should be similar as well as their capacity to bind to phospholipid membranes (Boulant et al. 2005, Perez-Berna et al. 2008, Ivanyi-Nagy and Darlix 2010). From the analysis of structural parameters (Boulant et al. 2005), dimers of protein C contain a highly hydrophobic region comprising the α2 helices of each of the monomers. This region also forms a concave groove which would be the responsible of binding to the viral membrane. On the other hand, this region could be also involved in lipid droplet association (Martins et al. 2012). Since no specific treatment is currently available for treating DENV infection, it is therefore essential to understand the DENV life cycle as well as protein/protein and protein/membrane interactions.

In a previous work (Nemesio et al. 2011), we described a region that induced a nearly complete membrane rupture of several membrane model systems, which was coincident with a theoretically predicted highly hydrophobic region of protein C. The main objective of this work has been the characterization of this region in the presence of different membrane model systems, in short DENV2_C6 (Figure 1A), using fluorescence spectroscopy techniques to assess membrane rupture, alteration of the fluorescence signal of FPE-labelled membranes in the presence of this peptide as well as steady state-fluorescence anisotropy. Calorimetric studies using differential scanning calorimetry (DSC) and Fourier transform infrared spectroscopy (FTIR) were also performed, using several model membrane systems. This characterization was motivated by the fundamental role this protein might play in the viral infection and the importance of this specific region for the interaction with biological membranes, an essential step of enveloped viruses' infection. Moreover, an understanding of protein-membrane molecular interactions during the DENV replication cycle might allow the identification of new targets for the treatment of Dengue virus infection (Table I).

Figure 1. (A) Scheme indicating the position of the domain corresponding to DENV2_C6 peptide on the DENVC structure (grey shading). The dimer of DENVC has been represented (Ma et al. 2004). (B) Fluorescence signal amplitude of FPE versus peptide concentration in μM to determine DENV2_C6 peptide binding to membrane model systems with different lipid compositions. The lipid compositions used were EPC (▪), EPC/CHOL at a molar ratio of 5:1 (○), EPC/ESM/CHOL at a molar ratio of 5:2:1 (□), EPC/EPA at a molar ratio of 5:2 (•), EPC/BPS at a molar ratio of 5:2 (▴) and synthetic ER membranes (Δ).The lipid concentration was 200 μM. Vertical bars indicate standard deviations of the mean of triplicate samples.

Materials and methods

Materials and reagents

The peptide DENV2_C6 corresponding to the sequence ³⁹GRGPLKLFMALVAFLRFL⁵⁶ from Dengue Virus Type 2 NGC (New Guinea C) C protein (with N-terminal acetylation and C-terminal acetylation) was obtained from Genemed Synthesis (San Antonio, TX, USA). This peptide was purified by reverse-phase HPLC (Kromasil C18, 250 × 4.6 mm, with a flow rate of 1 ml/min, solvent A, 0.1% trifluoroacetic acid, solvent B, 99.9% acetonitrile and 0.1% trifluoroacetic acid) to > 95% purity and its composition and molecular mass were confirmed by amino acid analysis and mass spectroscopy. Considering that trifluoroacetate has a strong infrared absorbance at approximately 1673 cm^-1, that can interfere significantly with the peptide Amide I band (Surewicz et al. 1993), residual trifluoroacetic acid, used both in peptide synthesis and in the HPLC mobile phase, was removed after several lyophilization/solubilization cycles in 10 mM HCl (Zhang et al. 1992). Bovine brain phosphatidylserine (BPS), bovine liver L-α-phosphatidylinositol (BPI), cholesterol (Chol), egg phosphatidic acid (EPA), egg L-α-phosphatidylcholine (EPC), egg sphingomyelin (ESM), egg trans-phosphatidylated L-α-phosphatidylethanolamine (TPE), tetramyristoyl cardiolipin (CL), bovine liver lipid extract, bis(monomyristoylglycero)phosphate (14BMP), bis(monooleoylglycero)phosphate (18BMP), 1,2-dimyristoilphosphatidylcholine (DMPC), 1,2-dimyristoylphosphatidylglycerol (DMPG), 1,2-dimyristoylphosphatidylserine (DMPS), 1,2-dimyristoylphosphatidic acid (DMPA), N-palmitoyl-D-erythro-sphingosylphosphorylcholine (PSM) and 1,2-dielaidoyl-sn-glycero-3-phosphatidylethanolamine (DEPE) were obtained from Avanti Polar Lipids (Alabaster, AL, USA). The lipid composition of the synthetic ER membrane was EPC/CL/BPI/TPE/BPS/EPA/ESM/Chol at a molar ratio of 59:0.37:7.4:18:3.1:1.2:3.4:7.8 (Krainev et al. 1995). 5-Carboxyfluorescein (CF, > 95% by HPLC), fluorescein isothiocyanate-labeled dextrans FD20 and FD70 (with respective average molecular weights of 20,000 and 70,000), deuterium oxide (99.9% by atom), Triton X-100, EDTA and HEPES were purchased from Sigma-Aldrich (Madrid, ES). 1,6-diphenyl-1,3,5-hexatriene (DPH) and 1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene (TMA-DPH) and N-(Fluorescein-5-thiocarbamyl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (FPE) were obtained from Molecular Probes (Eugene, OR, USA). All other chemicals were commercial samples of the highest purity available (Sigma-Aldrich, Madrid, Spain). Water was deionized, twice-distilled and passed through a Milli-Q equipment (Millipore Ibérica, Madrid, Spain) to a resistivity higher than 18 MΩ·cm.

Vesicle preparation

Aliquots containing the appropriate amount of lipid in chloroform/methanol (2:1 vol/vol) were placed in a test tube, the solvents were removed by evaporation under a stream of O₂-free N₂, and finally, traces of solvents were eliminated under vacuum in the dark for > 3 h. The lipid films were re-suspended in an appropriate buffer and incubated either at 25°C or 10°C above the phase transition temperature (T_m) with intermittent vortexing for 30 min to hydrate the samples and obtain multilamellar vesicles (MLV). The samples were frozen and thawed five times to ensure complete homogenization and maximization of peptide/lipid contacts with occasional vortexing. Large unilamellar vesicles (LUV) with mean diameters of 0.1 μm and 0.2 μm were prepared from MLV by the extrusion method (Mayer et al. 1986) using polycarbonate filters with a pore size of 0.1 μm and 0.2 μm (Nuclepore Corp., Cambridge, CA, USA). For infrared spectroscopy, aliquots containing the appropriate amount of lipid in chloroform/methanol (2:1, v/v) were placed in a test tube containing 200 μg of dried lyophilized peptide. After vortexing, the solvents were removed by evaporation under a stream of O₂-free N₂, and finally, traces of solvents were eliminated under vacuum in the dark for more than three hours. The samples were hydrated in 100 µl of D₂O buffer containing 20 mM HEPES, 100 mM NaCl, 0.1 mM EDTA, pH 7.4 and incubated at 10°C above the T_m of the phospholipid mixture with intermittent vortexing for 45 min to hydrate the samples and obtain MLV. The samples were frozen and thawed as above. Finally the suspensions were centrifuged at 14,000 rpm at 25°C for 10 min to remove the peptide unbound to the membranes. The pellet was re-suspended in 25 μl of D₂O buffer and incubated for 45 min at 10°C above the T_m of the lipid mixture, unless stated otherwise. The phospholipid and peptide concentration were measured by methods described previously (Böttcher et al 1961, Edelhoch 1967).

Membrane leakage measurement

LUVs with a mean diameter of 0.1 μm (for CF) and 0.2 μm (for FD20/FD70) were prepared as indicated above in buffer containing 10 mM Tris, 20 mM NaCl, pH 7.4 (at 25°C), and either CF at a concentration of 40 mM or FD20/FD70 at a concentration of 100 mg/ml. Non-encapsulated CF or FD20/FD70 were separated from the vesicle suspension through a filtration column containing Sephadex G-50 or Sephadex S500HR Sephacryl, respectively (GE Healthcare), eluted with buffer containing 10 mM Tris, 10 mM NaCl and 0.1 mM EDTA at pH 7.4. Membrane rupture (leakage) of intraliposomal CF was assayed by treating the probe-loaded liposomes (final lipid concentration, 0.125 mM) with the appropriate amount of peptide (peptide-to-lipid molar ratio of 1:25) on microtiter plates using a microplate reader (FLUOstar; BMG Labtech, Offenburg, Germany), stabilized at 25°C, each well containing 170 μl. The medium in the microtiter plates was continuously stirred to allow the rapid mixing of peptide and vesicles. Membrane leakage of intraliposomal FD20 or FD70 was carried out using 5 × 5 mm quartz cuvettes stabilized at 25°C in a final volume of 400 μl (100 μM lipid concentration). Leakage was assayed until no more change in fluorescence was obtained. The fluorescence was measured using a Varian Cary Eclipse spectrofluorimeter. Changes in fluorescence intensity were recorded with excitation and emission wavelengths set at 492 and 517 nm, respectively. Excitation and emission slits were set at 5 nm. One hundred per cent release was achieved by adding Triton X-100 to either the microtiter plate or the cuvette to a final concentration of 0.5% (w/w). Fluorescence measurements were made initially with probe-loaded liposomes, afterwards by adding peptide solution and finally adding Triton X-100 to obtain 100% leakage. Leakage was quantified on a percentage basis according to the equation, %Release = 100(F_f-F₀)/ (F₁₀₀-F₀), F_f being the equilibrium value of fluorescence after peptide addition, F₀ the initial fluorescence of the vesicle suspension and F₁₀₀ the fluorescence value after addition of Triton X-100.

Steady-state fluorescence anisotropy

DPH and its derivatives represent popular membrane fluorescent probes for monitoring the organization and dynamics of membranes; whereas DPH is known to partition mainly into the hydrophobic core of the membrane, TMA-DPH is oriented at the membrane bilayer with its charge localized at the lipid-water interface (Lentz 1993). MLVs were formed in a buffer composed of 20 mM HEPES, 100 mM NaCl and 0.1 mM EDTA at pH 7.4 (at 25°C). Aliquots of TMA-DPH or DPH in N, N′-dimethylformamide (0.2 mM) were directly added to the lipid dispersion to obtain a probe/lipid molar ratio of 1:500. Samples were incubated for 15 and 60 min, respectively, when TMA-DPH and DPH were used, 10°C above the T_m of each lipid for 1 h, with occasional vortexing. All fluorescence studies were carried out using 5 × 5 mm quartz cuvettes in a final volume of 400 μl (315 μM lipid concentration). All the data were corrected for background intensities and progressive dilution. The steady state fluorescence anisotropy, <r>, was measured with an automated polarization accessory using a Varian Cary Eclipse fluorescence spectrometer, coupled to a Peltier device (Varian) for automatic temperature change. The vertically and horizontally polarized emission intensities, elicited by vertically polarized excitation, were corrected for background scattering by subtracting the corresponding polarized intensities of a phospholipid preparation lacking probes. The G-factor, accounting for differential polarization sensitivity, was determined by measuring the polarized components of the fluorescence of the probe with horizontally polarized excitation (G = I_HV/I_HH). Samples were excited at 360 nm and emission was recorded at 430 nm, with excitation and emission slits of 5 nm. Anisotropy values were calculated using the formula <r> = (I_VV-GI_VH)/ (I_VV+2GI_VH), where I_VV and I_VH are the measured fluorescence intensities (after appropriate background subtraction) with the excitation polarizer vertically oriented and the emission polarizer vertically and horizontally oriented, respectively.

Fluorescence measurements using FPE-labelled membranes

LUVs with a mean diameter of 0.1 μM were prepared in buffer containing 10 mM Tris-HCl at pH 7.4 (at 25°C). The vesicles were labelled exclusively in the outer bilayer leaflet with FPE as described previously (Wall et al. 1995). LUVs were incubated with 0.1 mol% FPE dissolved in ethanol (never more than 0.1% of the total aqueous volume) at 37°C for 1 h in the dark. Any remaining unincorporated FPE was removed by gel filtration on Sephadex G-25 column equilibrated with the appropriate buffer. FPE-vesicles were stored at 4°C until use in an O₂-free atmosphere. Fluorescence time courses of FPE-labelled vesicles were measured after the desired amount of peptide was added into 400 μl of lipid suspensions (200 μM lipid) using a Varian Cary Eclipse fluorescence spectrometer. Excitation and emission wavelengths were set at 490 and 520 nm, respectively, using excitation and emission slits set at 5 nm. Temperature was controlled with a thermostatic bath at 25°C. The contribution of light scattering to the fluorescence signals was measured in experiments without the dye and was subtracted from the fluorescence traces. Data were fitted to a hyperbolic binding model (Golding et al. 1996) using the equations F = F_max[P]/ (K_d+[P]) or F = F_max[P]ⁿ/ (K_d+[P]ⁿ), where F is the fluorescence variation, F_max the maximum fluorescence variation, [P] the peptide concentration, K_d the dissociation constant of the membrane binding process and n, the Hill coefficient.

Differential scanning calorimetry

MLVs were formed as stated above in 20 mM HEPES, 100 mM NaCl, 0.1 mM EDTA, pH 7.4. The peptide was added to obtain a peptide/lipid molar ratio of 1:15. The final volume was 0.8 ml (0.6 mM lipid concentration), and incubated 10°C above the T_m of each phospholipid for 1 h with occasional vortexing. Differential scanning calorimetry (DSC) experiments were performed in a VP-DSC differential scanning calorimeter (MicroCal LLC, MA, USA) under a constant external pressure of 30 psi in order to avoid bubble formation and samples were heated at a constant scan rate of 60°C/h. Experimental data were corrected from small mismatches between the two cells by subtracting a buffer baseline prior to data analysis. The excess heat capacity functions were analysed using Origin 7.0 (Microcal Software). The thermograms were defined by the onset and completion temperatures of the transition peaks obtained from heating scans. In order to avoid artefacts due to the thermal history of the sample, the first scan was never considered; second and further scans were carried out until a reproducible and reversible pattern was obtained.

Infrared spectroscopy

Approximately 25 μl of a pelleted sample in D₂O were placed between two CaF₂ windows separated by 56-μm thick Teflon spacers in a liquid demountable cell (Harrick, Ossining, NY, USA). The spectra were obtained in a Bruker IF66S spectrometer using a deuterated triglycine sulphate detector. Each spectrum was obtained by collecting 250 interferograms with a nominal resolution of 2 cm^-1, transformed using triangular apodization and, in order to average background spectra between sample spectra over the same time period, a sample shuttle accessory was used to obtain sample and background spectra. The spectrometer was continuously purged with dry air at a dew point of −40°C in order to remove atmospheric water vapour from the bands of interest. All samples were equilibrated at the lowest temperature for 20 min before acquisition. An external bath circulator, connected to the infrared spectrometer, controlled the sample temperature. For temperature studies, samples were scanned using 2°C intervals and a 2-min delay between each consecutive scan. The data were analyzed as previously described (Guillen et al. 2010, Palomares-Jerez and Villalain 2011).

Results

Since peptide DENV2_C6 lacks a Trp residue and therefore intrinsic fluorescence, we have used the electrostatic surface potential probe FPE (Moreno et al. 2007) to monitor its ability to bind to model membranes composed of different lipid compositions at different lipid/peptide ratios (Figure 1B). DENV2_C6 had a higher affinity for model membranes composed of zwitterionic, i.e., EPC, and negatively charged phospholipids, i.e., EPC/EPA and EPC/BPS, as well as the complex membrane simulating ER membranes. Interestingly, lower affinity was observed for liposomes composed of EPC/Chol and EPC/SM/Chol (Figure 1B). Except for DENV2_C6 in the presence of ER-like model membranes, all data could be adjusted to a binding profile having either a sigmoidal (Hill coefficient of approximately 1) or a hyperbolic dependence, which might suggest that the interaction of the peptide with the membrane was monomeric. For DENV2_C6 in the presence of ER-like membranes, a sigmoidal dependence with a Hill coefficient of approximately 2.9 could be observed, suggesting that the interaction of the peptide with this complex membrane could be through a trimeric form of the peptide.

The effect of the DENV2_C6 peptide on the release of encapsulated fluorophores trapped inside model membranes was studied to explore the interaction of the peptide with phospholipid membranes (Palomares-Jerez and Villalain 2011). DENV2_C6 was able to induce the release of encapsulated CF in a dose-dependent manner and the effect was significantly different for different lipid compositions (Figure 2A). Liposomes composed of EPC elicited significant leakage effects, since DENV2_C6 induced leakage values of about 83% at a very high lipid/peptide ratio of 800:1. Since it is known that late endosome membranes are the place where membrane fusion entry takes place and have high concentrations of negatively charged phospholipids, namely PS and BMP (Zaitseva et al. 2010), we have tested membranes containing different contents of negatively-charged phospholipids. However, liposomes composed of negatively-charged phospholipids presented lower leakage values than zwitterionic membranes. At the lipid/peptide ratio of 800:1, liposomes composed of EPC/EPA and EPC/BPS at molar ratios of 5:2 as well as the ER-like complex mixture presented leakage values of 30–20% (Figure 2A). Liposomes composed of EPC/BMP at a molar ratio of 5:2 presented similar leakage values at similar lipid/peptide ratios (Figure 2A).

Figure 2. Effect of the DENV2_C6 peptide on the release (membrane rupture) of CF (A, B, C and D), FD20 (E) and FD70 (F) for general different lipid compositions (A, B, E and F), lipid compositions containing different molar ratios of EPC, ESM and Chol (C) and the ER^58:6 complex lipid mixture and its variations (D). In (A), model membranes with the following lipid compositions are shown: EPC (▪), EPC/EPA in a molar proportion of 5:2 (•), EPC BPS at a molar ratio of 5:2 (▴), total lipid liver extract (▾), ER-like membranes (◂), and EPC/BMP at a molar ratio of 5:2 (⧫). In (B), the following model membranes are shown: EPC/CHOL at a molar ratio of 5:1 (▪), EPC/CHOL at a molar ratio of 5:2 (▴), EPC/CHOL at a molar ratio of 5:3 (▴), EPC/EPA/CHOL at a molar ratio of 5:2:1 (▾) and EPC/BPS/CHOL at a molar ratio of 5:2:1 (◂). In (C), the following model membranes are shown: EPC/ESM/CHOL at a molar ratio of 5:1:1 (▪), EPC/ESM/CHOL at a molar ratio of 5:2:1 (▴), EPC/ESM/CHOL at a molar ratio of 5:3:1 (▴), EPC/ESM/CHOL at a molar ratio of 5:2:2 (▾) and EPC/ESM at a molar ratio of 5:2 (♦). In (D), the following model membranes are shown: ER ^58:6 (x), ER^58:6 minus Chol (▴), ER^58:6 minus BPI (▾), ER^58:6 minus TPE (♦), ER^58:6 minus ESM (▴), ER^58:6 minus BPS (◂), ER^58:6 minus EPA (◊) and ER^58:6 minus CL (▸). The lipid compositions of the model membranes used to study the release of FD20 (E) and FD70 (F) were: EPC (▪), EPC/CHOL in a molar proportion 5:1 (▴), EPC/ESM/CHOL in a molar proportion 5:2:1 (▴), EPC/EPA in a molar proportion 5:2 (▾), EPC/BPS in a molar proportion 5:2 (◂) and ER-like membranes (▸). Vertical bars indicate standard deviations of the mean of triplicate samples.

Significantly lower values were found for liposomes composed of liver lipid extract where leakage values of about 10% were observed at a lipid/peptide ratio of 300:1 (Figure 2A). Since Chol seems to be required in the fusion process (Carro and Damonte 2013), we have tested the presence of Chol in negatively-charged membranes. However, addition of Chol to negatively charged lipid compositions, i.e., EPC/EPA/Chol and EPC/BPS/Chol at a molar ratio of 5:2.1, did not change significantly the leakage values, since at a lipid/peptide ratio of 800:1 leakage values of 10 and 35% were observed, respectively (Figure 2B). In the same way, Chol did not change leakage values when it was added to EPC liposomes at diverse molar ratios: EPC/Chol at molar ratios of 5:1, 5:2 and 5:3 presented leakage values of about 83%, 89% and 93%, respectively (Figure 2B). We have also used membranes containing both SM and Chol lipids, since their presence has been related to the occurrence of laterally segregated membrane microdomains or ‘lipid rafts'. Interestingly, it has been shown that viral active replication complexes and/or non-structural proteins can be associated to lipid rafts (Noisakran et al. 2008). Different EPC, ESM and Chol compositions were studied to observe any specific interaction between the peptide and the phospholipids (Figure 2C). DENV2_C6 elicited a leakage value of about 85% for liposomes containing EPC/ESM at a molar ratio of 5:2, similar to the leakage values found for EPC/ESM/Chol at a molar ratio of 5:2:1 (Figure 2C). Likewise, EPC/ESM/Chol at molar ratios of 5:1:1, 5:2:1 and 5:3:1 presented leakage values of about 95%. From all these data it could be concluded that neither Chol nor ESM affect significantly the interaction of DENV2_C6 with lipid membranes, yet leakage was noticeably lower when negatively charged phospholipids were used.

We have studied the effect of the DENV2_C6 peptide on membrane rupture using a complex lipid composition resembling the ER-like membrane to assess the effect of each component of the complex mixture (Figure 2D). The synthetic ER complex membrane used above was composed of EPC, CL, BPI, TPE, BPS, EPA, ESM and Chol at a molar ratio of 59:0.37:7.4:18:3.1:1.2:3.4:7.8 (Krainev et al. 1995). Therefore we have designed an additional ER synthetic membrane composed of the same types of lipids, i.e., EPC/CL/BPI/TPE/BPS/EPA/ESM/Chol, but at a molar ratio of 58:6:6:6.6:6:6:6 (ER^58:6).

As shown in Figure 2D, DENV2_C6 was capable of rupturing the ER^58:6 complex membrane effectively: at a lipid/peptide ratio of 800:1, about 87% leakage was observed. In contrast, a leakage value of about 28% was obtained for the ER-like complex mixture at a lipid/peptide ratio of 800:1 (Figure 2A). When BPS was removed, leakage values slightly increased, whereas when either one of the other lipid molecules was removed, i.e., CL, EPA, ESM or Chol, leakage values decreased to a lesser extent. However, a significant decrease in leakage was observed when either TPE or BPI were removed from the membrane (Figure 2D). Since CF is a relatively small molecule (Stokes radius about 6 Å) we have studied the release of membrane-encapsulated molecules with a larger size (FITC-dextrans FD20 and FD70 having a Stokes radius of about 33 Å and 60 Å, respectively) (Laurent and Granath 1967, Fisher and Cash-Clark 2000). As observed in Figures 2E and 2F, DENV2_C6 was capable of inducing a significant percentage of leakage for both FD20 and FD70 fluorophores, although at slightly slower different values than CF at comparable lipid/peptide ratios. The capacity of the DENV2_C6 peptide to induce leakage of FITC-dextrans although to a lower extent than the small CF molecule demonstrates that DENV2_C6 is capable of forming relatively large pores in the membrane (Yang et al. 2001).

Phospholipids can undergo a cooperative melting reaction linked to the loss of conformational order of the lipid chains; this melting process can be influenced by many types of molecules including peptides and proteins. The effect of DENV2_C6 on the structural and thermotropic properties of phospholipid membranes was investigated by measuring the steady-state fluorescence anisotropy of the fluorescent probes DPH and TMA-DPH incorporated into model membranes as a function of temperature (Figure 3).

Figure 3. Steady-state anisotropy, <r>, of TMA-DPH and DPH (left and right columns, respectively) incorporated into MLVs composed of DMPC (A and B), DMPG (C and D), DMPS (E and F), DMPA (G and H), DPPC (I and J), DSPC (K and L), and 14BMP (M and N) model membranes as a function of temperature. Data correspond to vesicles in the absence (▪) and presence of the DENV2_C6 peptide (○). The peptide to phospholipid molar ratio was 1:15.

DMPC, in the presence of the peptide, presented a slight decrease in the cooperativity of the thermal transition, but no change in the T_m of the phospholipid was observed when compared to the pure lipid; additionally, the anisotropy of DPH but not TMA-DPH was increased above the T_m (Figures 3A and B). When the negatively charged phospholipids DMPG and DMPS were studied, DENV2_C6 induced a similar behavior to that found for DMPC, i.e., a slight decrease in the cooperativity of the thermal transition, no change in the T_m of the phospholipid and the anisotropy of DPH but not TMA-DPH was increased above the T_m (Figures 3C and 3F). The case of DMPA was different, since an increase in T_m of about 2°C was found in the presence of DENV2_C6 (Figures 3G and 3H). However, the anisotropy of DPH was increased above the T_m but no significant change in cooperativity was found. We also studied the phospholipids DPPC and DSPC to compare with DMPC, since DPPC is longer that DMPC by two methylene groups whereas DSPC is larger by four. Additionally, the T_m of DPPC and DSPC appear at about 43°C and 54°C, whereas the T_m of DMPC appears at about 23°C. As observed in Figures 3I and 3L, DENV2_C6 induces in both phospholipid a similar behavior as DMPA, since it induces an increase of about 1–2°C in the T_m of both phospholipids as well as an increase in anisotropy above T_m. However, a decrease in cooperativity was observed for DPH but not for TMA-DPH (Figures 3I and 3L). The effect of DENV2_C6 on another phospholipid, namely 14BMP, was also studied (Figures 4M and 4N). DENV2_C6 induced a significant decrease in cooperativity in both DPH and TMA-DPH probes, as well as an increase in anisotropy above the T_m for DPH and below the T_m for TMA-DPH. DENV2_C6 is therefore capable of affecting the thermal transition T_m of these phospholipids; moreover, the peptide changed the anisotropy of DPH to a greater extent than TMA-DPH, so it should be located slightly deeper in the membrane, influencing the fluidity of the phospholipids (Contreras et al. 2001).

Figure 4. Differential scanning calorimetry heating thermograms corresponding to model membranes composed of (A) DMPC, (B) DEPE, (C) DMPS, (D) DMPA, (E) DMPG, (F) 14BMP and (G) PSM in the absence (top curves) and presence of DENV2_C6 peptide (bottom curves) at a phospholipid/peptide molar ratio of 15:1. All the thermograms were normalized to the same amount of lipid.

The effect of DENV2_C6 on the thermotropic phase behavior of phospholipid multilamellar vesicles was also studied using differential scanning calorimetry, DSC (Figure 4). Incorporation of DENV2_C6 in DMPC at a lipid/peptide ratio of 15:1 significantly altered the thermotropic behavior of the phospholipid, since the peptide completely abolished the pre-transition, and decreased the T_m of the lipid concomitantly with a significant broadening of the main transition (Figure 5A). The main transition of DMPC in the presence of DENV2_C6 is apparently composed of at least two different peaks, which should be due to mixed phases. In the case of DEPE, DENV2_C6 induced a slight decrease in the T_m of the phospholipid but no significant broadening was observed, neither on the gel to liquid-crystalline phase transition nor on the lamellar liquid-crystalline to hexagonal-H_II phase transition (Figure 4B). For both DMPS and DMPA, DENV2_C6 slightly decreased the cooperativity of the transitions but no change in T_m was apparent (Figures 4C and 4D). Incorporation of DENV2_C6 into DMPG membranes at a lipid/peptide ratio of 15:1 (Figure 4E) did not significantly alter the thermotropic behavior of the phospholipid, since the peptide did not completely abolish the pre-transition; however, the main transition of the phospholipid was broadened and shifted to lower temperatures than the T_m. The incorporation of DENV2_C6 into 14BMP membranes did not alter the thermotropic behavior of the lower temperature endothermic peak (L_c1-L_c2) but broadened and shifted to lower temperatures the higher temperature peak (L_c2-L_α) (Figure 4F). Interestingly, a small relatively broad peak appeared at slightly lower temperatures than that of the main transition (Figure 4F).The incorporation of DENV2_C6 into PSM membranes altered the thermotropic behavior of the phospholipid, inducing a broadening of the main transition peak without altering T_m (Figure 4G). These results are in accordance with those commented above, since DENV2_C6 is capable of affecting the thermal transition T_m of all phospholipids studied here; furthermore, in some cases, it is capable of inducing the presence of mixed phases.

Figure 5. Stacked infrared spectra of the C = O and Amide I' regions of DENV2_C6 (A) in solution and in the presence of (B) DMPC, (C) DMPG, (D) DMPG and (E) 14BMP at different temperatures as indicated. The pure phospholipid is shown on the left whereas the phospholipid/peptide mixture is shown on the right. The temperature dependence of the frequency at the maximum of the Amide I' region of DENV2_C6 in solution (□) and the intensity ratio of the 1684 cm^-1 and 1694 cm^-1 bands (▪) are shown in the upper right figure. The phospholipid/peptide molar ratio was 15:1.

The infrared Amide I' region of fully hydrated DENV2_C6 in buffer is shown in Figure 5A. The Amide I' band presented different bands with varying intensities depending on temperature. At low temperatures, two narrow bands at about 1624 and 1693 cm^-1, the former with higher intensity that the later, were apparent as well as a broad one with a maximum at about 1651 cm^-1 (Figure 5A). At higher temperatures, the band at 1624 cm^-1 shifted to about 1621 cm^-1 whereas the band at about 1693 cm^-1 diminished steadily in intensity. Increasing the temperature, a new band at 1683 cm^-1 appeared and increased gradually in intensity; however, the broad band at about 1651 cm^-1 did not change significantly in frequency upon increasing the temperature although its intensity decreased. The band at about 1624–1621 cm^-1 would indicate the existence of aggregated β structures, whereas the broad band with the intensity maxima at about 1651 cm^-1 would correspond to a mixture of unordered and helical structures (Byler and Susi 1986, Arrondo and Goni 1999). The conjoint appearance of bands at about 1693 cm^-1 and narrow intense bands at about 1624 cm^-1 would correspond to aggregated β structures. The band at about 1683 cm^-1 would correspond to β-turn structures. The global spectrum envelope would suggest that, although helical and unordered structures might be present, the main secondary structure component of DENV2_C6 consists of aggregated structures (Figure 5A). In the presence of DMPC, the Amide I' envelope of the DENV2_C6 peptide presented a broad band with a maximum at about 1650 cm^-1 and a shoulder at about 1625 cm^-1 at all temperatures (Figure 5B). When DENV2_C6 was combined with either DMPG (Figure 5C) or 14BMP (Figure 5D), the Amide I' band was reasonably similar to the peptide in solution (Figure 5A), except that the intensity of the different peaks varied. These data would suggest that the main secondary structure component of the peptide should be aggregated structures with some helical and disordered structures in the presence of all lipids, except DMPC. In the presence of DMPC, the main secondary structure component of the peptide should be composed of helical and disordered structures and a minor content of aggregated structures.

We have also analyzed the hydrocarbon CH₂ and ester C = O stretching bands of the phospholipids in the presence of DENV2_C6 (Figure 6). The frequency maximum of the ester C = O band of pure DMPC displayed two transitions, one at about 12°C, coincident with the pre-transition, and another one at about 24°C, coincident with the main gel to liquid crystalline phase transition (Figure 6A). The hydrocarbon CH₂ frequency displayed only one transition coincident with the main gel to liquid crystalline phase transition at about 24°C (Figure 6D). In the presence of DENV2_C6, the frequencies of the hydrocarbon CH₂ and ester C = O stretching bands of DMPC displayed only one transition at about 20°C, in accordance with the DSC data (Figure 6A). The absolute frequency of the C=O band was higher in the presence of the peptide than in its absence, suggesting that the peptide increased the intensity of the 1743 cm^-1 component relative to the 1727 cm^-1 one, i.e., the amount of non-hydrogen bonded C=O ester bands increased in the presence of the peptide (Blume et al. 1988, Lewis et al. 1994). The same behavior was observed for the CH₂ symmetric stretching band, but in this case the increase in frequency indicates that the peptide increased the mobility of the hydrocarbon chains at all studied temperatures (Figure 6D). The pattern of the thermotropic phase behavior exhibited by pure DMPG was similar to that of pure DMPC, since a transition at about 24°C was observed; however, no change in the frequency of the C=O band of the phospholipid coincidental with the pre-transition of the lipid was observed (Figures 6B and 6E). However, in the presence of the peptide and at low temperatures, the C=O carbonyl band of DMPG presented a main peak and a shoulder, which could suggest the formation of a quasi-crystalline lamellar phase (Figures 6B and 5C) (Epand et al. 1992, Garidel et al. 2000). At higher temperatures the two peaks coalesced into a broad C=O band similar in appearance to the one observed for pure DMPG, suggesting the formation of a normal liquid crystalline phase (Figures 6B and 5D). The ester C=O bands of pure 14BMP displayed a sharp band at low temperatures, with a maximum at about 1735 cm^-1, and a broad one at high temperatures, with a maximum at about 1732 cm^-1 (Figure 5D). The frequency of both CH₂ and ester C=O bands defined two transition temperatures in accordance with the DSC results commented above (Figures 6G and 6H). In the presence of DENV2_C6, 14BMP displayed two transitions, a broad one at approx. 25°C observed through the ester C=O band frequency (Figure 6C) and a cooperative one at approx. 41°C, observed through the hydrocarbon CH₂ frequency (Figure 6F).

Figure 6. Temperature dependence of the frequencies of the (A, B, C) C = O carbonyl and (D, E, F) CH₂ symmetric stretching bands for samples of (A, D) DMPC, (B, E) DMPG and (C, F) 14BMP in the pure form (▪) and in the presence of the DENV2_C6 peptide (○). The phospholipid/peptide molar ratio was 15:1.

Table I. Alignment (ClustalW2) of DENV2_C6 peptide from reference strains representing the major genotypes of Dengue virus. The DENV2_C6 and consensus sequences are shown beneath the alignment.

DENV	Strain	Sequence	DENV	Strain	Sequence
Type 2	MD917	GRGPLKLFMALVAFLRFL	Type 3	C036094-C	GQGPMKLVMAFIAFLRFL
	BIDV687	GRGPLKLFMALVAFLRFL		ThD31283_98-C	GQGPMKLVMAFIAFLRFL
	MD903	GRGPLKLFMALVTFLRFL		BIDV1831VN2007-C	GQGPMKLVMAFIAFLRFL
	BIDV633	GHGPLKLFMALVAFLRFL		07CHLS001-C	GQGPMKLVMAFIAFLRFL
	CSF381-C	GRGPLKLFMALVAFLRFL		BIDV1874VN2007-C	GQGPMKLVMAFIAFLRFL
	MD1504	GRGPLKLFMALVAFLRFL		05K797DK1-C	GQGPMKLVMAFIAFLRFL
	CSF63-C	GRGPLKLFMALVAFLRFL		98TWmosq-C	GQGPMKLVMAFIAFLRFL
	DakArD20761	GRGPLKLFMALVAFLRFL		BR29002-C	GQGPMKLVMAFIAFLRFL
	DF755-C	GRGPLKLFMALVAFLRFL		98-C	GQGPMKLVMAFIAFLRFL
	DF707	GRGPLKLFMALVAFLRFL		TB55i-C	GQGPMKLVMAFIAFLRFL
Type 1	02_20-C	GQGPMKLVMAFIAFLRFL	Type 4	2A-C	GKGPLRMVLAFITFLRVL
	ThD1004901-C	GQGPMKLVMAFIAFLRFL		BIDV2165VE1998-C	GKGPLRMVLAFITFLRVL
	BIDV1841-C	GQGPMKLVMAFIAFLRFL		BIDV2170VE1999-C	GKGPLRMVLAFITFLRVL
	BIDV1926VN2008	GQGPMKLVMAFIAFLRFL		Vp4-C	GKGPLRMVLAFITFLRVL
	BIDV1800	GQGPMKLVMAFIAFLRFL		Yama-C	GKGPLRMVLAFITFLRVL
	BIDV1323	GQGPMKLVMAFIAFLRFL		Sin897695-C	GKGPLRMVLAFITFLRVL
	BIDV2243VE2007	GQGPMKLVMAFIAFLRFL		rDEN4del30-C	GKGPLRMVLAFITFLRVL
	05K4147DK1	GQGPMKLVMAFIAFLRFL		H241-C	GKGPLRMVLAFITFLRVL
	297arg00	GQGPMKLVMAFIAFLRFL		ThD4048501-C	GKGPLRMVLAFITFLRVL
	BIDV2143-C	GQGPMKLVMAFIAFLRFL		ThD4047697-C	GKGPLRMVLAFITFLRVL
	Consensus sequence	*:**:::.:*:::***.*
	PEPTIDE DENV2C6	GRGPLKLFMALVAFLRFL

Discussion

DENV C protein is a very basic protein which is essential for virus assembly since it is responsible for genome packaging. DENV C protein has a hydrophobic segment, approximately spanning residues 45–60, which seems to be required for maturation, lipid droplet binding and assembly of the viral particles (Markoff et al. 1997, Samsa et al. 2009). In a previous work (Nemesio et al. 2011), we found a region in DENV C protein that induced membrane rupture of several membrane model systems, which was coincident with the predicted highly hydrophobic region of the protein. Since the biological role/roles of DENV C protein is/are intrinsically related to its interaction with the RNA and the membrane, we have extended our previous work to investigate the binding and interaction of this highly conserved membranotropic region of DENV C, i.e., peptide DENV2_C6, with different membrane model systems. We have carried out an in-depth biophysical study aimed at the elucidation of the capacity of this region to interact and disrupt membranes, as well as to study the structural and dynamic features which might be relevant for that disruption.

Peptide DENV2_C6 binds with high affinity to phospholipid model membranes, as it has been previously found for other peptides (Pascual et al. 2005). We have also shown that the DENV2_C6 peptide is capable of affecting the steady state fluorescence anisotropy of fluorescent probes located in the palisade structure of the membrane. Calorimetry experiments further corroborated these results, and additionally indicated that the DENV2_C6 peptide induced the formation of mixed lipid phases, enriched and impoverished in peptide. Similar results were obtained by studying the hydrocarbon CH₂ and ester C=O stretching bands of the phospholipids in the presence of DENV2_C6. The DENV2_C6 peptide was capable of altering membrane stability causing the release of fluorescent probes, this effect being dependent on the lipid/peptide molar ratio and on lipid composition. The highest CF release was observed for liposomes containing the phospholipid EPC, although lower yet still significant leakage values were observed for liposomes containing negatively-charged phospholipids. It should be recalled that leakage was observed even at a very high lipid/peptide ratio as 800:1. Interestingly, Chol addition did not lower the leakage elicited by the peptide, since its addition to different membrane compositions induced neither higher nor lower leakage values. Apart from that, DENV2_C6 was able to form relatively large pores in the membrane as demonstrated by the release of relatively large size molecules such as FITC-dextrans. Since DENV2_C6 affected more significantly membranes containing BMP, which is known to be enriched in the late endosome membrane (Zaitseva et al. 2010), a concerted action of DENV C and E proteins and late endosome lipids should not be ruled out to be essential for viral fusion. Taking into account all these data, the specific disrupting effect elicited by DENV2_C6 should be primarily due to hydrophobic interactions within the bilayer, although the specific charge of the phospholipid head-groups would influence the extent of membrane leakage. The infrared spectra of the Amide I' region of the fully hydrated DENV2_C6 peptide in solution displayed a coexistence of mainly aggregated structures although unordered and helical structures were also present. This overall structure did not change in the presence of either DMPG or 14BMP; however, in the presence of DMPC the main components were helical and disordered structures with a minor content of aggregated structures. These results imply that the secondary structure of the DENV2_C6 peptide was affected by its binding to specific lipid in the membrane, so that membrane binding modulates the secondary structure of the peptide as it has been suggested for other peptides (Guillen et al. 2010, Palomares-Jerez et al. 2010).

Recently, Carvalho et al. (2012) have found that DENVC is capable of binding to several proteins on the surface of intracellular lipid droplets. It could be envisaged that DENVC, specifically interacting with lipid droplet proteins, would initiate a conformational arrangement of the proteins; this conformational changes would allow the direct interaction of the hydrophobic segment where the DENV2_C6 resides with lipid droplet lipids. The binding of DENV2_C6 to the membrane and the modulation of the phospholipid biophysical properties that takes place as its consequence could be related to the conformational changes that might occur during the biological activity of the DENV C protein. The conservation of its sequence and hence its structure and physico-chemical properties among different DENV strains should be essential to its function. Our results provide new insight as to how this segment can contribute to the interaction with the membrane. Although the peptide is not deeply buried in the membrane, its interaction with the membrane depends on its composition and it is able to affect the lipid milieu from the membrane surface down to the hydrophobic core. These results identify an important region in the DENV C protein which might be directly implicated in the DENV life cycle. Moreover, an understanding of DENV2_C6 membrane molecular interactions during the DENV replication cycle might allow the identification of new targets for the treatment of Dengue virus infection.

Acknowledgements

This work was partially supported by grant BFU2008-02617-BMC (Ministerio de Ciencia y Tecnología, Spain) to J.V. H.N. is supported by a “Santiago Grisolía” fellowship from Generalitat Valenciana Autonomous Government.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.