Role of tryptophan 135 of Chandipura virus phosphoprotein P in dimerization and complex formation with leader RNA: structural aspect using time resolved anisotropy and simulation

Abstract

The aggregation of phosphoprotein P of Chandipura virus (CHPV) and its interaction with viral leader (le) RNA in aqueous buffer and 40% ethylene glycol (EG)-buffer have been characterised using two single tryptophan (Trp/W) mutants W105F and W135F. The longer rotational correlation time [(θ_C)_T] originating from overall motion observed at 300 nM concentration conforms to a globular structure of the monomer of WT and both the mutants. The (θ_C)_T values also indicate that W135F does not form a dimer at 1500 nM concentration; while a dimer with disordered structure is predicted for both WT and W105F. The complexes of WT and W105F with le RNA at monomeric and dimeric conditions are indicated to have tight core packing. Dimerization and complex formation at both the concentrations enhance the correlation time arising from the localised motion of Trp side chain [(θ_C)_S] for WT and W105F predicting hindered localized rotation of Trp 135. Comparative protein modelling and molecular dynamics (MD) simulations using the amino acids domain ranging from 105–168 of the full length CHPVP based on the vesicular stomatitis virus phosphoprotein (VSVP) also indicate that formation of dimers are more feasible for WT and W105F compared to W135F.

---

1. Introduction

Phosphoprotein P, of vesiculoviruses are multifunctional proteins involved in key regulatory steps of the viral life cycle. Structurally, P protein is a component of the viral RNA dependent RNA polymerase (RdRp) and acts as a cofactor modulating the RNA synthetic activities of the large protein (L). Though a part of the polymerase, the P protein is the least conserved structural protein among vesiculoviruses, with about 30% overall identity.¹ However, in spite of the sequence diversity, P proteins of various vesiculoviruses possess very similar distribution of hydropathy, suggesting that the overall conformation of the protein may be similar and is important for its function.² Among the most important functions of the P protein is its role in template recognition, which it facilitates by binding to both the nucleocapsid protein (N) and the large protein (L).² The P protein has also been identified as a chaperone functional in maintaining the N protein in an encapsidation competent monomeric form termed N⁰.³ Importantly, the P protein has been shown to be indispensible for productive virion assembly.⁴ However, the most intriguing and the least understood function of the P protein appears to be its ability to bind specifically to the small, untranslated viral leader RNA (le RNA) both in vitro as well as in vivo.^5,6 The importance of this interaction has been increased by the finding that the expression of the le RNA and thus its interaction with the P protein is augmented during the replication phase of the viral cycle. This suggests that this interaction possibly plays an important role in CHPV replication.

Studies on the human pathogen, Chandipura virus (CHPV) a vesiculovirus belonging to the Rhabdovirus family have revealed that the unphosphorylated form of the P protein (P₀) binds to the le RNA, while phosphorylation at serine 62 (P₁) impairs this interaction.^6,7 Based on this differential interaction with the le RNA, Basak et al. proposed that this interaction plays an important role in the regulation of replication in this class of viruses.^5,7 Stoichiometrically P forms two different complexes with the le RNA. A monomer–le RNA complex, termed complex I and a dimer–le RNA complex, termed complex II have been elucidated in electrophoretic mobility shift assays.^5,6 Oligomerization of P was found to be concentration dependent. Sub-micromolar concentration (300 nM) favoured monomeric populations whereas, concentration above 1.5 μM resulted in dimers.^5,6

Based on the limited proteolytic digestion observations made by Ding et al., the P protein can be divided into three domains – the N-terminal domain, the central domain, and the C-terminal domain (Fig. 1).⁸ For CHPV P, the N-terminal domain contains the Ser 62 residue that is phosphorylated by host casein kinase II and is indispensable for the transcriptional activity of the P protein.⁹ The C-terminal domain is instrumental in binding to both the N and the L proteins.¹⁰ Das et al. showed that mutations in this domain abrogates the replication activity of the P protein.⁴ The central domain has been established to be the oligomerization domain in VSV P protein.^8,11 Ding et al. solved the crystal structure of the central domain of the VSV P protein¹¹ and showed that the fold of this domain consists of a β hairpin followed by an α helix and then another β hairpin. According to this structure, the α helix provides the stabilizing force for forming a homodimer, while the two β hairpins add additional stabilization by forming a four-stranded β sheet through domain swapping between two P molecules.¹¹

Fig. 1 Schematic representation of the predicted domains and their amino acid sequences of the P-protein of Chandipura virus (CHPV). The domains were identified on the basis of global alignment between the amino acid sequences of the P protein from CHPV with that of VSV. The colors of the amino acid sequences correspond to those of the respective domains.

VSVP has also been shown to have a modular organization with alternating structured and unstructured domains. Two autonomous folding domains (P_CED, 107–177 and P_CTD, 195–265) and two unstructured domains have been reported. The first unstructured domain is present as an N-terminal intrinsically disordered region (IDR) (IDR_NT; amino acids 1–106) and the other, a disordered linker, separates P_CED from P_CTD.¹² The N-terminal region (IDR_NT) constitutes a molecular recognition element (MoRE) which undergoes a disorder-to-order transition when it binds to its physiological partner, monomeric nucleocapsid protein (N⁰).¹³

The central domain of CHPV P has been recognised as important for its self association and mutation of a particular tryptophan (Trp/W) residue (W135) located in the α helix was shown to adversely affect dimerization.⁶ Moreover, it has been reported that binding of the le RNA induces quenching of the intrinsic fluorescence of the P protein.⁶ To facilitate the study of this le RNA binding phenomenon, two tryptophan mutants were employed: W105F and W135F.⁶ It was found that mutation of W135 to phenylalanine (F) resulted in reduced binding of the le RNA under both monomeric (300 nM) and dimeric (1500 nM) conditions. The apparent binding constant of W135F was calculated to be approximately two orders of magnitude weaker than that of the wild type protein WT.⁶ It has also been reported that binding to the le RNA induces a conformational change in the P protein that may have an implication in its proposed role in replicational control in CHPV.^5,6,14

In this work we presented time resolved anisotropy for the wild type at both 300 nM and 1500 nM concentrations in aqueous buffer and also in buffer containing 40% ethylene glycol (EG) in order to find the rotational correlation time corresponding to total motion of the free protein and that of the complex with le RNA. The segmental motion of Trp residue/residues in the free protein and in the complex with le-RNA has been also addressed. We also carried out similar experiments for both the single Trp mutant viz., W105F and W135F under similar conditions of concentration and medium. In order to interpret the data in 40% EG-buffer and pure buffer medium we also presented the Trp fluorescence decay of CHPV and the two single Trp mutants and their complexes with le RNA at both 300 nM and 1500 nM concentrations of protein in 40% EG-buffer medium. The anisotropy data in both the media have been exploited to find the nature of the structure of the aggregated form of the protein and that of the complexes with le RNA. The segmental motion of Trp in the free mutants and in the complexes with le RNA has been employed to find the perturbation of Trp environment due to aggregation as well as due to complexation with le RNA. A molecular dynamics simulation study has been carried out to help understand the effect of mutation of Trp by Phe on the dimerization of CHPV.

2. Materials and methods

2.1 Production and purification of WT and its tryptophan mutants

For the present study WT, W105F, W135F were produced and purified using the same procedure as described previously.⁶

2.2 Synthesis of leader RNA

The 49 nt long positive sense leader RNA (le RNA) was synthesized in vitro as described previously by Basak et al.⁵

2.3 Time resolved emission measurement

Singlet state lifetime (τ) was measured by Time Master Fluorimeter from Photon Technology International (PTI, USA). The system consists of a pulsed laser driver of a PDL series i.e., PDL-800-B (from Picoquant, Germany) with interchangeable sub nano second pulsed LEDs and pico-diode lasers (Picoquant, Germany) with a TCSPC set up (PTI, USA). The software Felix 32 controls all acquisition modes and data analysis of the Time Master system.¹⁵ All the samples were excited using PLS-290 of Picoquant, Germany (pulse width < 500 ps) at a repetition frequency 10 MHz. Instrument response function (IRF) were measured at the respective excitation wavelengths, namely, 290 nm using slits with a band pass of 3 nm using Ludox as the scatterer. The decay of sample was analyzed by nonlinear iterative fitting procedure based on the Marquardt algorithm. Deconvolution technique (PTI) used can determine the lifetime up to 120 ps with sub nano second pulsed LEDs. The quality of fit has been assessed over the entire decay, including the rising edge, and tested with a plot of weighted residuals and other statistical parameters, for example, the reduced χ² ratio and the Durbin–Watson (DW) parameters.¹⁵

Intensity decay curves were fitted as a sum of exponential terms:¹⁶

F(t) = ∑α_i exp(−t/τ_i) (1)

where α_i represents the pre-exponential factor to the time resolved decay of the component with a lifetime τ_i. Amplitude average lifetime 〈τ〉 was calculated using the equation:¹⁶

〈τ〉 = ∑(α_iτ_i)/∑α_i (2)

Anisotropy decay measurement was also carried out in Time Master Fluorimeter (PTI, USA) using PLS-290 and motorized Glen Thompson polarizer. The anisotropy, r(t) is defined as

r(t) = [I_VV(t) − G × I_VH(t)]/[I_VV(t) + 2 × G × I_VH(t)] (3)

where I(t) terms are defined as intensity decay of emission of protein with excitation polarizer orientated vertically and the emission polarizer oriented vertically and horizontally, respectively:

G = I_HV(t)/I_HH(t) (4)

where G is the correction term for the relative throughput of each polarization through the emission optics. The entire data analysis was done with the software Felix 32 which analyses the raw data I_VV and I_VH simultaneously by global multi-exponential program and then the deconvolved curves (ID_VV and ID_VH) are used to construct r(t).¹⁵ The anisotropy decay r(t) was considered to be a sum of discrete exponential functions:where β_i and θ_i were the numerical parameters recovered to ensure the best fit of the r(t) function and r₀ is the anisotropy at t = 0. From the fitted curve the correlation time (θ_C) can be recovered.

2.4 Modelling of the dimerization domain of the CHPV phosphoprotein

2.4.1 Modelling of the monomer and its mutants. Comparative protein modeling was carried out using the program MODELLER (version 9.11).¹⁷ Based on the maximum sequence identity (∼29%) with the dimerization domain of the vesicular stomatitis virus phosphoprotein (VSVP), amino acids ranging from 105–168 of the full length Chandipura virus phosphoprotein (CHPVP) (UniProt ID P16380) was selected and modeled using the chain A of the same domain from VSV (PDB ID 2FQM).¹¹ The lowest scoring model out of 40 was chosen as wild-type and the two mutants W105F and W135F were generated using suitable utilities of the molecular visualization software Pymol (version 1.2r.2).¹⁸

2.4.2 Modelling of the homodimer and its mutants. The steps were identical to the modelling of the monomer, apart from the fact that in this case chains C and D of the VSV phosphoprotein were used as template. As in the case of the monomer, two variants, one containing mutation W135F in both the chains and another having mutation W105F in both the chains were generated using Pymol. Since the variant containing the mutation W135F had a significantly reduced preference for forming a dimer,⁶ we built this model to study if this mutation had a destabilizing effect on the dimer.

2.5 Simulation methodology

2.5.1 Simulation of the monomers. All molecular dynamics (MD) simulations were performed with the simulation package GROMACS 4.5.5 ¹⁹ using the all atom OPLS force field.²⁰ All the three monomers, namely, the wild-type, the W105F mutant and the W135F mutant, were protonated corresponding to pH 7.0 and subjected to a short energy minimization in vacuo using the steepest descent algorithm to remove any short contact. The systems were then solvated with SPC water²¹ in an octahedral box with periodic boundary condition. The dimensions of the box were set such that all the protein atoms were at least 1.0 nm away from the edge of the box. After neutralizing the system it was subjected to a steepest descent energy minimization with a maximum force tolerance of 1000 kJ mol⁻¹ nm⁻¹ while restraining the protein's heavy atoms. Next an unrestrained energy minimization was carried out using the steepest descent followed by conjugate gradient algorithm with maximum force tolerances of 100 and 10 kJ mol⁻¹ nm⁻¹ respectively. To attain the desired temperature of 300 K and 1 atm pressure, a 300 ps of NVT simulation followed by a 1.4 ns of NPT equilibration were carried out restraining the heavy atoms of the protein. Finally, a 600 ps of unrestrained MD simulation was carried out to relax the solute within its environment and the resulting structure was subjected to 57 ns production simulation carried out in the isothermal–isobaric ensemble (NPT) at 300 K. Leap-frog integrator was used to solve the Newtonian equation of motion. LINCS algorithm²² was employed to constrain bond lengths involving hydrogen. Electrostatic interactions were calculated using the Particle Mesh Ewald (PME) summation method.²³ van der Waals and Coulomb interactions were truncated at 1.0 nm. The protein and the solvent were independently coupled to a thermal bath by a modified Berendsen thermostat²⁴ at 300 K and coupling period of 0.1 ps. Isotropic Berendsen pressure coupling was used to maintain the system pressure within 1 atm.

2.5.2 Simulation of the dimers. The simulation steps for the dimer were the same as the monomer, except that energy minimization in vacuum was not performed for this case. NVT equilibration was continued up to 600 ps, position restrained NPT equilibration was extend up to 2.6 ns and an unrestrained NPT simulation was performed for 1 ns prior to the production simulation. Production simulations were carried out for 50 ns.

2.5.3 Analysis. Secondary structure assignments were carried out with the DSSP²⁵ module integrated with GROMACS.

3. Results and discussion

3.1 Time resolved emission of WT and single tryptophan mutants and of the complexes with le-RNA

The fluorescence maxima of WT, W105F and W135F at 300 nM and 1500 nM concentration in 40% EG-buffer remain same as observed in aqueous buffer with λ_exc = 295 nm (Fig. S1A and B†). The fluorescence quenching behaviour with le RNA is also not altered in 40% EG-buffer. Fig. 2 shows the decay of WT, W105F and W135F in 40% EG-buffer at 300 nM and 1500 nM concentrations monitoring the λ_max of fluorescence in each case. A representative decay for W105F at 1500 nM concentration in the presence and in the absence of le RNA in 40% EG-buffer is shown in Fig. 3. The decay in each case is found to fit satisfactorily to a sum of three components (Table 1). Fitting with single component or two components resulted a χ² value greater than unity. The goodness of the fit has been established by randomness of the residual distribution and the χ² is found to be close to unity in each case. The life time together with χ² values used to judge the quality of the fit are summarized in Table 1. The average lifetime values of all the species in 40% EG-buffer are found to be longer than that observed in aqueous buffer medium.⁶ WT and W105F show almost 50% enhanced average lifetime at 1500 nM concentration compared to that observed at 300 nM, whereas W135F exhibits similar average lifetime at both the concentrations (Table 1). The data clearly indicates that W135F does not form dimer at 1500 nM concentration in 40% EG-buffer as observed in pure aqueous buffer medium.⁶ It is also observed that the WT CHPV and the mutant W105F show practically unchanged τ_av in the absence and in the presence of le RNA indicating the formation of ground state complexes in each case (Table 1). The quenching experiments carried out (data not shown) in 40% EG-buffer provide a similar binding trend for the proteins as observed in the pure aqueous buffer medium.

Fig. 2 Fluorescence decay at 298 K monitoring at λ_max for (A) WT, (B) W105F and (C) W135F in 40% EG-buffer; λ_exc = 290 nm, excitation and emission band pass 10 nm each.

Fig. 3 Fluorescence decay in 40% EG-buffer at 298 K monitoring λ_max for 1500 nM W105F with 0, 750 nM le RNA, λ_exc = 290 nm, excitation and emission band passes are 10 nm in each case.

Table 1 Time resolved data of WT, W105F, W135F and its complexes with RNA in 40% EG-buffer at 298 K

System	Singlet state life time λexc = 290 nm
	τ1 (ns)	τ2 (ns)	τ3 (ns)	τav (ns)	χ^2
300 nM WT	8.20 (10.07%)	2.42 (25.66%)	0.71 (64.27%)	1.90	1.07
300 nM WT + 100 nM RNA	8.12 (8.59%)	2.88 (26.29%)	0.70 (65.12%)	1.91	0.98
300 nM W105F	8.84 (5.329%)	3.21 (37.07%)	1.11 (57.6%)	2.30	0.97
300 nM W105F + 100 nM RNA	8.82 (5.607%)	3.23 (40.95%)	1.02 (53.44%)	2.36	1.09
300 nM W135F	6.95 (17.23%)	2.91 (34.46%)	0.88 (48.31%)	2.62	1.02
300 nM W135F + 100 nM RNA	6.73 (17.33%)	2.70 (35.07%)	0.91 (52.60%)	2.59	0.98
1500 nM WT	7.14 (18.45%)	2.98 (38.07%)	0.76 (43.37%)	2.78	1.15
1500 nM WT + 750 nM RNA	6.17 (15.2%)	3.37 (39.39%)	0.81 (45.4%)	2.63	1.13
1500 nM W105F	5.05 (40.91%)	4.83 (17.68%)	1.42 (41.41%)	3.51	1.08
1500 nM W105F + 750 nM RNA	5.17 (42.27%)	3.53 (16.11%)	1.09 (41.62%)	3.21	0.97
1500 nM W135F	6.73 (14.72%)	2.793 (41.17%)	0.69 (44.11%)	2.44	1.025
1500 nM W135F + 750 nM RNA	6.80 (14.37%)	2.70 (40.36%)	0.72 (45.27%)	2.39	1.017

---

3.2 Time resolved anisotropy decay of WT and single tryptophan mutants and their complexes with le-RNA

In order to explore the local environmental rigidity of Trp residues and the perturbation of the environments upon dimerization and binding with le RNA as well as the structural aspects of the dimers and the complexes, we studied solvation dynamics of the WT, W105F, W135F at both 300 nM and 1500 nM concentrations in the absence and in the presence of leader RNA with λ_exc = 290 nm both in aqueous buffer and in 40% EG-buffer.

Fluorescence anisotropy decays in each case monitoring Trp emission have been measured at 298 K in aqueous buffer and in 40% EG-buffer. Fig. 4 shows representative fluorescence anisotropy decays in 40% EG-buffer for both WT and WT-RNA complex using two different concentrations of the protein viz., 300 nM and 1500 nM. All the time resolved fluorescence anisotropy decays of free WT and its single Trp mutants and their complexes with le RNA exhibit a shorter pico-second resolved rotational correlation time [(θ_C)_S] and a longer nano second resolved rotational correlation time [(θ_C)_T] in all the media at both the concentrations used (Table 2). The goodness of fit is obtained by χ² values in each case. The χ² values are found to close to 1 for each analysis.

Fig. 4 Fluorescence anisotropy decays in 40% EG-buffer at 298 K monitoring λ_max for (A) 300 nM WT, (B) 300 nM WT + 100 nM RNA complex, (C) 1500 nM WT, (D) 1500 nM WT + 750 nM RNA complex.

Table 2 Time resolved anisotropy data of WT, W105F, W135F and their complexes with RNA in aqueous buffer and in 40% EG-buffer at 298 K

System	Medium	Rotational correlation time using 300 nM protein concentration^a		Rotational correlation (θC) using 1500 nM protein concentration^b
		Total rotational motion (θC)T (ns)	Segmental motion^c (θC)S (ps)	Total rotational motion (θC)T (ns)	Segmental motion^c (θC)S (ps)
a RNA concentration = 100 nM.b RNA concentration = 750 nM.c Error ± 10 ps.
WT	Buffer (pH 8)	9.49 (85%)	130 (15%)	24.87 (77%)	180 (23%)
WT	40% EG-buffer	10.61 (93%)	150 (7%)	28.02 (88%)	220 (12%)
WT + RNA	Buffer (pH 8)	14.18 (83%)	160 (17%)	29.43 (75%)	225 (25%)
WT + RNA	40% EG-buffer	16.09 (92%)	200 (8%)	32.63 (77%)	250 (17%)
W105F	Buffer (pH 8)	8.88 (90%)	180 (10%)	23.95 (80%)	200 (20%)
W105F	40% EG-buffer	11.51 (97%)	200 (3%)	29.08 (88%)	210 (12%)
W105F + RNA	Buffer (pH 8)	14.88 (91%)	190 (9%)	28.67 (83%)	530 (17%)
W105F + RNA	40% EG-buffer	18.10 (94%)	300 (6%)	30.96 (86%)	550 (14%)
W135F	Buffer (pH 8)	8.63 (89%)	120 (11%)	9.36 (86%)	130 (14%)
W135F	40% EG-buffer	9.74 (92%)	190 (8%)	10.70 (88%)	190 (12%)
W135F + RNA	Buffer (pH 8)	10.92 (96%)	130 (4%)	11.97 (91%)	140 (9%)
W135F + RNA	40% EG-buffer	11.80 (87%)	180 (13%)	12.28 (84%)	180 (16%)

---

The (θ_C)_S indicates the segmental motion of Trp and the larger component arises from the rotation of the whole system. The observed changes in rotational correlation time due to total motion (θ_C)_T and segmental motion (θ_C)_S in aqueous buffer and in 40% EG-buffer for various systems (Table 2) are shown as bar diagram (Fig. 5 and 6). Initial anisotropy values (r₀) for each analysis are also provided in Table 3.

Fig. 5 Change in rotational correlation time (θ_C)_T due to total motion at 298 K monitoring λ_max for WT, W105F, W135F. Error = ±5%.

Fig. 6 Change in rotational correlation time (θ_C)_S due to segmental motion in buffer at 298 K monitoring λ_max for WT, W105F, W135F. Error = ±10%.

Table 3 Initial anisotropy data of WT, W105F, W135F and their complexes with RNA in aqueous buffer and in 40% EG-buffer at 298 K

System	Medium	Initial anisotropy (r0)
		Using 300 nM protein concentration^a	Using 1500 nM protein concentration^b
a RNA concentration = 100 nM.b RNA concentration = 750 nM.
WT	Buffer (pH 8)	0.20	0.25
WT	40% EG-buffer	0.17	0.18
WT + RNA	Buffer (pH 8)	0.26	0.27
WT + RNA	40% EG-buffer	0.21	0.26
W105F	Buffer (pH 8)	0.24	0.26
W105F	40% EG-buffer	0.25	0.28
W105F + RNA	Buffer (pH 8)	0.26	0.29
W105F + RNA	40% EG-buffer	0.27	0.32
W135F	Buffer (pH 8)	0.18	0.18
W135F	40% EG-buffer	0.20	0.21
W135F + RNA	Buffer (pH 8)	0.19	0.20
W135F + RNA	40% EG-buffer	0.19	0.21

---

3.3 Total and segmental rotational correlation time

In addition to being affected by overall dimension, the decay of fluorescence anisotropy in proteins is affected by shape and segmental flexibility.^26–28 Rotational correlation times that are too short to be approximated by overall rotational diffusion of a protein molecule have been attributed to independent motions of tryptophan residues within the protein. Short rotational correlation times provide invaluable information about the internal dynamics of proteins. It will be interesting to relate sub nanosecond segmental motions and it's fluctuation in proteins as it may be involved with the elementary steps in conformational transitions that are essential for dimerization and RNA binding.

WT protein in buffer showing total correlation time 9.49 ns (Table 2) is comparable to the value of 9.0 ns calculated from the eqn (6) for a 32.53 kDa (molecular weight of CHPV P protein) rigid hydrated sphere at 298 K, assuming ν = 0.72 ml g⁻¹ and h = 0.0 cm³ g⁻¹.¹⁶

θ_C = [ηM(ν + h)][1/RT] (6)

The calculated value becomes 11.70 ns using h = 0.23 cm³ g⁻¹. The single Trp mutants W105F and W135F also exhibit the total correlation time similar to WT (8.88 ns and 8.63 ns) implying site directed mutation by Phe does not induce any significant change in the size and shape of the overall protein.

We have previously reported that the Stokes radius (R_S) of monomeric P is close to that of a globular protein of similar molecular weight with a R_S (measured)/R_S (globular) ratio close to 1.⁶ The (θ_C)_T values support approximately globular shape of the monomer CHPV. When the protein concentration changes from 300 nM to 1500 nM, the ratio [(θ_C)_T]_{1500 nM}/[(θ_C)_T]_{300 nM} becomes 2.62 for WT and 2.70 for W105F in the aqueous buffer medium, whereas the ratio for W135F is found to be 1.1 (Table 2) due to change of concentration (Table 2 and Fig. 5). The increased (θ_C)_T for WT and W105F in 1500 nM concentration are thus essentially due to dimeric state. These results thus conclusively prove that W135F is defective in dimerization and Trp 135 plays an important role in dimerization.

We also observed that upon dimerization the R_S value increases and the R_S (measured)/R_S (globular) ratio increases to 1.7.⁶ Thus the present results of (θ_C)_T support the hypothesis that the CHPVP attains a more disordered structure upon dimerization which is reflected by its increased hydrodynamic radius. This observation is in agreement with the fact that the P protein of VSV and other vesiculoviruses have also been shown to partially unstructured under dimeric conditions.^12,29,30 However, not much information is available on the structure of the monomeric protein in these viruses.

It is expected that the total motion of a protein are relatively free in the absence of RNA, but becomes more hindered as le-RNA of larger size binds with it. The ratio of the [(θ_C)_T]_complex/[(θ_C)_T]_{free protein} at 300 nM protein concentration (monomeric state) is found to be 1.49 for WT and 1.66 for W105F and 1.26 for W135F in aqueous buffer (Table 2). The ratio obtained for WT and W105F in the presence of le RNA is consistent with the formation of a stable complex having a tightly associated structure considering the molecular weight of le RNA (19.5 kDa). However, the (θ_C)_T values increase less significantly for W135F in the presence of le RNA compared to free W135F. These finding of low degree of rotational freedom for the total motion in the WT-RNA complex and W105F-RNA complex definitely indicates that Trp 135 and possibly its adjacent amino acids are involved in binding to one site of the le RNA.

At 1500 nM concentration, we observed that the ratio [(θ_C)_T]_complex/[(θ_C)_T]_{free protein} is 1.28 for W135F in aqueous buffer. This value is practically same as observed at 300 nM concentration. This indicates further that dimer formation is prevented when Trp 135 is mutated with Phe. The same ratio is observed to be 1.2 for WT and W105F even though both of them exist as dimer at this concentration. Our steady state results, however, indicate complex formation for the dimers also.⁶ It is to be noted that we measured the anisotropy decay with [RNA] = 750 nM. Use of larger concentration of RNA quenches the protein fluorescence to an extent where measurement of anisotropy decay could not be analyzed properly due to very weak signal. However, the ratio of [(θ_C)_T] values for the complex at 300 nM and the complex at 1500 nM [using [RNA] = 750 nM] is found to be close to 2 for WT as well as for W105F (Table 3). This ratio, on the other hand, is 1.08 for W135F.

The calculation of accessible surface area (ASA) of Trp residues and the location of immediate neighbouring (within 5 Å) residues around Trp residues in the generated CHPV P structure using the crystal structure of VSV protein, indicate that Trp 135 is buried with lower ASA (ASA 3.09 Ų) and surrounded by lesser number of polar residues than that of Trp 105 (ASA 86.17 Ų).⁶ This is nicely supported by the observed (θ_C)_S in all the systems. Trp 135 in W105F shows larger segmental correlation time due to hindered rotation than that of Trp 105 in W135F (Table 2). Dimerization and complexation with le RNA also enhances the (θ_C)_S for WT and W105F indicating buried nature of Trp 135 in the dimer and in the complexes (Table 2). The enhancement of (θ_C)_S is found to be quite considerable for W105F (Table 2).

The time resolved fluorescence anisotropy decays (using λ_exc = 290 nm) for WT, W105F, W135F at both 300 nM and 1500 nM concentrations by using the Trp as probe in the absence and the presence of leader RNA in 40% EG-buffer mixtures were also analysed (Table 2). The increase in (θ_C)_T and (θ_C)_S for each system in 40% EG-buffer compared to that in aqueous buffer could be attributed to the low degree of rotational freedom in viscous 40% EG-buffer mixture (Table 2). It is expected that hydration of protein could be somewhat different in such medium. The values of (θ_C)_T, however, do not vary linearly with viscosity coefficient (η) of the medium according to eqn (3). This could be due to the local viscosity surrounding the protein being smaller than the bulk viscosity.³¹ The changes observed in (θ_C)_T and (θ_C)_S for WT and it's single Trp mutants with and without RNA in 40% EG-buffer are found to follow similar trend as obtained in the case of aqueous buffer medium. WT and W105F responded significantly to concentration and the le RNA induced time resolved anisotropy, confirming the disordered structure of the dimer at 1500 nM concentration. W135F, on the other hand, did not aggregate at 1500 nM concentration as observed in aqueous buffer medium. These results clearly imply that the perturbation of the microenvironment of the Trp 135 during dimerization and le-RNA interaction was responsible for the observed change in WT as well as in W105F.

Time resolved anisotropy data thus indicate that CHPV P mutated at 135 position does not self associate nor forms stable complex with the le RNA. This observation is in agreement with the crystal structures of the oligomerization domains of the phosphoproteins of VSV and the rabies virus (RV), both of which are members of the Rhabdovirus family.^11,32 Though both the structures are substantially different from each other, one common element between them is the prominent role of a tryptophan residue (W138 for VSV P and W118 for RV P) situated in the dimer interface in the dimerization process. Both these residues align with W135 of CHPV P. According to the structure described by Ivanov et al., the surface of the RV P oligomerization domain consists of a hydrophobic cavity and a complementary bulge next to the hydrophobic cavity.³² This bulge is formed by W118. When the two monomers form a dimer, both tryptophans insert into the opposite cavities. Further, they observed a large blue shift of the fluorescence emission spectrum of the dimerization domain which they attributed to the W118 being deeply buried within the hydrophobic dimer interface. On the other hand, in the structure of the VSV P dimerization domain, two α-helices in the dimer interface interact primarily through hydrophobic cluster at their respective N-terminal ends.¹¹ W138 is at the core of this hydrophobic interaction. The side chain of this Trp residue is packed with the side chains of L126, P127, and L130 on the loop leading to the α-helix. Aligning the sequence of the P proteins of CHPV and VSV revealed that P127 and L130 are conserved (as P124 and L127) in CHPV P protein (alignment not shown). Although L126 is not conserved, leucine at the adjacent 121 position may serve its function. Moreover, this Trp residue has been shown to be involved in further hydrogen bonding, and hydrophobic interactions that stabilize the dimeric interactions. It is therefore evident that our observations using time resolved fluorescence emission and anisotropy corroborate well with these homologous structures.

Examples of specific mutation preventing homoaffinity as observed in CHPV are known. It has been shown that Cyan Fluorescent Protein (CFP) exhibits much weaker homo affinity than other fluorescent protein e.g. Green Fluorescent Protein (GFP) or Yellow Fluorescence Protein (YFP).³³ This has been predicted to be due to the constitutive N146I mutation originally introduced into CFP to improve it's brightness.³³

3.4 Structural effect of W105F and W135F mutations on the monomer of the CHPVP dimerization domain

The modeled structures of the wild-type CHPVP dimerization domain in its monomeric and homodimeric forms are given in Fig. 7. Relative positions of the two tryptophan residues, W105 and W135, are indicated in Fig. 7A.

Fig. 7 (A) The generated structure of the monomeric CHPVP dimerization domain. The residues W105 and W135 are depicted as sticks. (B) Homodimeric structure of the CHPVP dimerization domain. (C) Top view of the CHPVP dimer.

A previous study of intramolecular energy transfer between Trp 105 and Trp 135 in CHPV³⁴ indicated that the distance between Trp 105 and Trp 135 was of the order of 10 Å. In Fig. 8, we have plotted the time evolution of the distance between the centers of mass of these two residues from a 57 ns simulation of the wild-type CHPVP dimerization domain. We observed that the distance between these two residues in the model for the wild-type never exceeded 10 Å indicating a good correlation with the experimental result. Similar analysis of the distance between the same two residues for the W105F and W135F mutants revealed that in the case of the W135F mutant this distance increased sharply after 30 ns of simulation and diverged to 2 nm. Such sharp increase was indicative of structural rearrangements occurring after 30 ns in the present simulation of the W135F mutant.

Fig. 8 Time evolution of the center of mass distance between residues W105 and W135 for the wild-type (in black), between residues F105 and W135 for the W105F mutant (in green) and between residues W105 and F135 for the W135F mutant (in red).

Secondary structure analysis of the trajectory (Fig. 9) revealed that in the W135F mutant, the helix, spanning residues 129–145, in the dimerization interface was extensively disrupted after 30 ns. The final structure of the W135F mutant also showed the presence of the severely distorted helix (Fig. 10).

Fig. 9 Time evolution of the secondary structures of the dimerization domain during 57 ns MD simulation. Residues 105–168 are numbered as 1–64 here. Colors indicate different types of secondary structural elements at a given time point according to DSSP classification: white: coil, red: beta-sheet, black: beta-bridge, green: bend, yellow: turn, blue: alpha-helix, magenta: 5-helix and grey: 3–10 helix. (A) Wild type, (B) W105F mutant, (C) W135F mutant.

Fig. 10 Initial and final structures of the wild-type and the mutated monomers of the CHPVP dimerization domain as obtained from MD simulation. Initial energy minimized structures of the wild-type and its mutated variants are given in the left panel and the corresponding final structures after 57 ns MD simulation are given in the right panel. (A) wild-type (B) W105F and (C) W135F. The red arrow indicates the region of helix distortion in the W135F mutant.

3.5 Structural effect of W105F and W135F mutations on the homodimer of the CHPVP dimerization domain

Comparison of the root mean square deviations (RMSD) of the backbone atoms of the respective dimers indicated that the dimer containing the W135F double mutation underwent a large structural change with time (Fig. 11A) while the dimers of the wild-type domain and the W105F mutant behaved similarly and showed a much less structural change. Secondary structural analysis did not reveal any large distortion in the interface helices in the dimer of the W135F mutant. To further investigate the origin of the structural deviation we measured the time evolution of the inter-helical distance from 50 ns simulations of the wild-type, W105F and W135F mutants (Fig. 11B). It was evident from this analysis that the inter-helical distance in the dimerization interface became significantly larger for the W135F mutant compared to those of the wild-type and the W105F mutant.

Fig. 11 (A) Time evolution of backbone RMSD of the wild-type (green), W105F (black) and W135F (red) variants. (B) Time evolution of the interhelical distance for the three dimers. The inset shows the time evolution of the same during unrestrained equilibration, indicating that the helices in the W135F mutant start to move away from each other right from the beginning of unrestrained dynamics.

From the simulation results it is tempting to infer that the W135F mutation may become destabilizing for the protein as it was found to distort the inter-facial helix in the monomer simulations and generate unfavourable helix–helix interactions leading to increase in the inter-helical distance in the dimer simulations. These results are largely consistent with the fact that formation of dimers were less probable for the W135F mutant compared to the wild-type or the W105F mutant. However, one should keep in mind that the experimental dimer formation studies were carried out with the entire protein and there may be other contributing factors from the portion of the protein which could not be used for modelling.

4. Conclusion

The present manuscript explores the nanosecond and picosecond dynamics of two tryptophan residues, Trp105 and Trp135, of the phosphoprotein (P), a multifunctional protein involved in key regulatory steps of the life cycle of Chandipura virus (CHPV), a prototype Rhabdoviridae. The wild type (WT) protein and its two single Trp mutants (W105F and W135F) have been used to study structural change in the protein as a function of (i) protein concentration and (ii) binding of a short non-coding RNA, the leader (le) RNA, by time resolved fluorescence and anisotropy. The results also demonstrate the role of Trp 135 in dimerization of the protein and in the complex formation with le RNA. The overall motion of the protein as revealed by the anisotropy decay conforms to a disordered structure of the dimer WT and W105F. The segmental motion of Trp residues indicates the buried nature of Trp 135 in the dimeric state and also in the complexes with le RNA. This interaction between the P protein and the le RNA is deemed indispensable for the viral life cycle. Molecular dynamics simulation has been employed to understand the formation of dimer of wild type and W105F mutant. W135F is shown to be unable to form dimers due to unfavourable helix–helix interaction. Our observations enhance the present understanding of the oligomerization and the le-RNA interacting phenomena of the phosphoprotein, both of which are poorly understood and are potential targets for future development of antiviral strategies against this important emerging human pathogen.

Acknowledgements

This work is financially supported by the Department of Science and Technology (DST), Government of India (Grant No. SR/S1/PC-57/2008, SB/S1/PC-003/2013) to SG, and (Grant No. SB/FT/CS-141/2012) to PSS and the Department of Science and Technology (DST), Government of India (Grant No. SR/SO/BB-67-2009) and the Council of Scientific and Industrial Research, Government of India (Grant No. 21(0871)/11/EMR-II) to SG and (Grant No. 08/155(0046)/2013-EMR-I) to MM; UGC-DSA, Government of India; DST-FIST, Government of India; UGC-RFSMS, Government of India; Dr D. S. Kothari fellowship.

References and Notes

1. P. S. Masters and A. K. Banerjee, Virology, 1987, 157, 298.
2. S. Basak, A. Mondal, S. Polley, S. Mukhopadhyay and D. Chattopadhyay, Biosci. Rep., 2007, 27, 275.
3. A. Majumdar, R. Bhattacharya, S. Basak, M. S. Shaila, D. Chattopadhyay and S. Roy, Biochemistry, 2004, 43, 2863.
4. S. C. Das and A. K. Pattnaik, J. Virol., 2005, 79, 8101.
5. S. Basak, S. Polley, M. Basu, D. Chattopadhyay and S. Roy, J. Mol. Biol., 2004, 339, 1089.
6. A. Roy, M. Mukherjee, S. Mukhopadhyay, S. S. Maity, S. Ghosh and D. Chattopadhyay, Biochimie, 2013, 95, 180.
7. S. Basak, T. Raha, D. Chattopadhyay, A. Majumder, M. S. Shaila and D. Chattopadhyay, Virology, 2003, 307, 372.
8. H. Ding, T. J. Green and M. Luo, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2004, 60, 2087.
9. T. Raha, R. Kaushik and M. S. Shaila, Virus Res., 2004, 104, 191.
10. P. R. Paul, D. Chattopadhyay and A. K. Banerjee, Virology, 1988, 166, 350.
11. H. Ding, T. J. Green, S. Lu and M. Luo, J. Virol., 2006, 80, 2808.
12. C. Leyrat, R. Schneider, E. A. Ribeiro Jr, F. Yabukarski and M. Yao, et al., J. Mol. Biol., 2012, 423, 182.
13. C. Leyrat, F. Yabukarski, N. Tarbouriech, E. A. Ribeiro, M. R. Jensen Jr and M. Blackledge, et al., PLoS Pathog., 2011, 7, e1002248.
14. A. Roy, P. Chakraborty, S. Polley, D. Chattopadhyay and S. Roy, Antiviral Res., 2013, 100, 346.
15. FELIX 32, Operation Manual, Version 1.1, Photon Technology International, 1009 Lenox Drive, Suite 104, Lawrenceville, NJ 08645, 2003.
16. J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluwer Academic/Plenum Publishers, 2006.
17. A. Sali and T. L. Blundell, J. Mol. Biol., 1993, 234, 779.
18. The PyMol Molecular Graphics System, Version 1.2r.2, Schrodinger, LLC.
19. S. Pronk, S. Pall, R. Schulz, P. Larsson, P. Bjelkmar, R. Apostolov, M. R. Shirts, J. C. Smith, P. M. Kasson, S. D. van der, B. Hess and E. Lindahl, GROMACS 4.5: Bioinformatics, 2013, vol. 29, p. 845.
20. W. L. Jorgensen, D. S. Maxwell and J. Tirado-Rives, J. Am. Chem. Soc., 1996, 118, 11225.
21. H. J. C. Berendsen, J. P. M. Postma, W. F. van Gunsteren and J. Hermans, Intermolecular Forces, ed. B. Pullman, 1981, vol. 331.
22. B. Hess, H. Bekker, H. J. C. Berendsen and J. G. E. M. Fraaije, J. Comput. Chem., 1997, 18, 1463.
23. T. Darden, L. Perera, L. Li and L. Pedersen, Structure, 1999, 7, R55.
24. H. J. C. Berendsen, J. P. M. Postma, W. F. van Gunsteren, A. DiNola and J. R. Haak, J. Chem. Phys., 1984, 81, 3684.
25. W. Kabsch and C. Sander, Biopolymers, 1983, 22, 2577.
26. C. Rischel, P. Thyberg, R. Rigler and F. M. Poulsen, J. Mol. Biol., 1996, 257, 877.
27. J. B. A. Ross, K. W. Rousslang and L. Brand, Biochemistry, 1981, 20, 4361.
28. G. Krishnamoorthy, Curr. Sci., 2012, 102, 266.
29. D. Karlin, F. François, C. Bruno and S. Longhi, J. Gen. Virol., 2003, 84, 3239.
30. F. C. A. Gérard, E. A. Ribeiro, C. Leyrat, I. Ivanov and D. Blondel, et al., J. Mol. Biol., 2009, 388, 978.
31. D. Lavalette, C. Tétreau, M. Tourbez and Y. Blouquit, Biophys. J., 1999, 76, 2744.
32. I. Ivanov, T. Crépin, M. Jamin and R. W. H. Ruigrok, J. Virol., 2010, 84, 3707.
33. A. Espagne, M. Erard, K. Madiona, V. Derrien, G. Jonasson, B. Levy, H. Pasquier, M. Ronald and F. Merola, Biochemistry, 2011, 50, 437.
34. S. Mukhopadhyay, S. S. Maity, A. Roy, D. Chattopadhyay, K. S. Ghosh, S. Dasgupta and S. Ghosh, Biochimie, 2010, 92, 136.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra20863g

‡ The present address: The Department of Chemistry, The Bhawanipur Education Society College, Kolkata-700020, India.

---