Modulation of west nile virus capsid protein and viral RNA interaction through phosphorylation 4

4-6] plasmids encoding wild type or mutant myc-C proteins were transfected into 293FT cells and BHK cells and myc-C proteins were immuno-purified with anti-myc antibody.. The immuno-puri

5.0 PHOSPHORYLATION OF WEST NILE VIRUS (WNV) CAPSID (C)

PROTEIN AND RNA INTERACTION

5.1 Introduction

Having established that phosphorylation on C protein peptides could attenuate its RNA binding ability the questions to address are if WNV C protein is indeed phosphorylated and how does modification of the phosphorylation status affect C protein functions Known characteristics of C protein like nuclear localisation and oligomerization could be affected by phosphorylation

5.2 Validation of anti-phosphoserine antibodies

Prior to using the anti-phosphoserine antibodies for the detection of phosphorylation, it had to be validated since such reagents are known to be unspecific Several anti-phosphoserine antibodies were tested but only one (Clone 4A4, Milipore) produced satisfactory results (The other antibodies either resulted in overexposed blots

or no signal was observed.) The antibody was first tested by Western blot on the positive control Calyculin A/Okadaic-treated human A431 carcinoma cell lysate supplied by Milipore The cell lysate was also treated with λ-phosphatase as a negative control (Fig 5-1) The antibody was able to differentiate between cell lysate that was either treated or untreated with λ-phosphatase and the results were reproducible As expected, His-C protein that was expressed in bacteria was unphosphorylated

Figure 5-1 Validation of anti-phosphoserine antibodies Human A431 carcinoma cell

lysate (positive control supplied by Upstate) were either treated (Lanes 2 and 7) or untreated (Lanes 1 and 6) with λ-phosphatase and subjected to both Western blot and SDS-PAGE analysis At the same time bacteria expressed histidine-tagged West Nile Virus (WNV) capsid (C) protein is also subjected to Western blot and SDS-PAGE analyses (Lanes 3 and 8) Markers are in Lanes 4 and 5 on the Western blot and SDS-PAGE, respectively λ-phosphatase is indicated by a black arrowhead (Lane 7) The anti-phosphoserine antibody is able to detect phosphorylated proteins in the A431 carcinoma cell lysate (Lane 1) but not in the λ-phosphatase-treated one (Lane 2) The antibody is not able to detect any serine phosphorylation on the histidine-tagged C (His-C) protein (Lane 3)

5.3 Phosphorylation of the West Nile virus (WNV) capsid (C) protein

5.3.1 West Nile virus (WNV) capsid (C) protein is a phosphoprotein

In order to determine if WNV C protein is a phospho-protein, C protein was immuno-purified from infected BHK cells using anti-C antibody The immuno-purified protein was blotted onto a nitrocellulose membrane and probed with anti-phosphoserine and anti-C antibody (Fig 5-2) Immunoblot (Western blot) analysis showed that WNV C protein is indeed a phospho-protein since anti-phosphorine antibodies detected immunopurified WNV C protein

Bioinformatics analysis revealed 5 putative phosphorylation sites; mutations to myc-C protein were performed to target serine 26, 36, 83, 99 and threoine 100 to abolish phosphorylation These residues were mutated to alanine Because serine 26 and 36 were close to each other a myc-C mutant carrying double mutations was constructed (Fig 5-3A) Similarly, since serine 83, 99 and theronine 100 were close together a triple mutant was constructed (Fig 5-3A) In addition a mutant where all 5 putative phosphorylation sites were mutated to alanine was also constructed (Fig 5-3A)

In order to resolve whether myc-C protein is phosphorylated, [and hence explain its inability to bind RNA (Fig 4-6)] plasmids encoding wild type or mutant myc-C proteins were transfected into 293FT cells and BHK cells and myc-C proteins were immuno-purified with anti-myc antibody The 293FT cell line was used because it is a human cell line, therefore phosphorylation of C protein in this cell line would be more relevant Nonetheless, myc-C protein should be phosphorylated in both eukaryotic cell lines The immuno-purified wild type myc-C protein was subjected to Western blot analysis and was found to be phosphorylated in both cell lines (Fig 5-3B) In contrast,

the mutant myc-C proteins were found to be hypophosphorylated although phosphorylation was not completely abolished, as was the case when myc-C protein was treated with λ-phosphatase (Fig 5-3B)

Mutagenesis of putative phosphorylation sites could have perturbed the native confirmation of C protein, homology modeling of the S26/36/83/99/T100A mutant was performed to ensure that the mutations did not drastically alter the confirmation of C protein Using SWISS-MODEL (http://swissmodel.expasy.org/), it was found that the mutations did not make any drastic conformational changes to C protein (Fig 5-4) Hence the different phenotypes observed most likely were due to hypophosphorylation

Figure 5-2 West Nile virus C protein is a phosphoprotein BHK cells were infected with

WNV and cell lysates were immuno-precipitated using anti-WNV C protein antibody (Ab) and immunoblotted with anti-phosphoserine Ab (i) or anti-WNV C Ab (ii) The appearance of band corresponding to the size of C protein in WNV-infected samples (i and ii) confirms that WNV C protein is a phospho-protein (iii and iv) Isotype and actin loading controls are shown

72

55

26

IP : Anti-WNVC Ab WB: Anti-phosphoserine

B

Figure 5-3 Mutagenesis of putative phosphorylation sites (A) Schematic diagram of

constructed mutant myc-C proteins The putative phosphorylated sites are indicated by a thick black line and the names of each of these mutants are indicated on the right (B) Phosphorylation of myc-tagged capsid protein in 293FT and BHK cells Plasmids encoding wild type capsid (Lanes 2 and 8) or mutants (Lanes 4, 5, 6, 10, 11 and 12) were transfected into 293FT or BHK cells The cell lysates was harvested and immunopurified with anti-myc antibody Cell lysates from 293FT and BHK cells were either treated (Lanes 3 and 9) or untreated with λ-phosphatase and subsequently subjected to Western blot analysis with anti-c-myc antibody and anti-phosphoserine antibody Mock-transfected cell lysate are Lanes 1 and 7 This shows that the anti-phosphoserine antibody

S26/36/83/99/T100

A

4 99-100

Mutant Capsid protein constructed

Amino acid position

Amino acid

position

Amino acid position

Figure 5-4 Homology modeling of mutant C protein The mutant S26/36/83/99/T100A

C protein (Blue) was modeled using SWISS-MODEL (http://swissmodel.expasy.org/)

and superimposed on the crystal structure of WNV (Kunjin strain) C protein (Dokland et al., 2004) (Red) The point mutations are indicated in yellow Root mean square

deviation value = 0.1

5.3.2 Protein kinase C phosphorylates myc-capsid (C) protein

Bioinformatics analysis revealed that protein kinase C as the most probable kinase that could phosphorylate the C protein at serine 83, 99 and theronine 100 Experiments were performed to confirm if protein kinase C was involved in phosphorylating WNV C protein To address this, kinase-specific activator or inhibitor was added to BHK cells transfected with wild type myc-C proteins or mutant myc-C proteins carrying mutations S83/99/100A Protein kinase A activator and inhibitors were used as a negative control

Western blot analysis showed that protein kinase C was involved phosphorylating myc-C protein since protein kinase C inhibitor (BIS) brought the phosphorylation level of the wild type myc-C protein to the same level as the S83/99/100A mutant [Fig 5-5A (i), Lanes 8 and 9] In addition, wild type myc-C protein harvested from protein kinase C activator-treated (PMA) cells had a higher level of phosphorylation than myc-C protein harvested from DMSO-treated cells [Fig5-5B (i) Lanes, 2 and 8]

In contrast, myc-C protein from protein kinase A inhibitor- and activator-treated cells (H89 and Forskolin) did not exhibit a reduced nor an increased level of phosphorylation [Fig 5-5A (i) Lane 5, Fig 5-5B (i) Lane 5] when compared to myc-C protein harvested from DMSO-treated cells [Fig 5-5A (i) Lane 2, Fig 5-5B (i) Lane 8 ]

In order to confirm that protein kinase C can indeed phosphorylate C protein, in vitro

phosphorylation of C protein was performed using His-C protein since His-C protein was shown to be unphosphorylated (Fig 5-1) When protein kinase C was added to and

incubated with His-C protein in vitro, His-C protein was found to be phosphorylated (Fig

5-5C, Lane 1) but hypophosphorylated in the presence of protein kinase inhibitor (BIS)

Figure 5-5 Phosphorylation of C protein by protein kinase C (A and B) BHK cells

treated with PKA/PKC inhibitors (A) or activators (B) were transfected with plasmid encoding wild type myc-C or mutant myc-C S83/99/T100A protein Both wild type and mutant proteins were subjected to co-immunoprecipitation using anti-myc Ab followed

by immunoblotting with anti-phosphoserine Ab The phosphorylation signal of C protein

is reduced [Lane 8, A(i)]/increased [Lane 2, B(i)] following treatment with PKC inhibitor/activator, respectively The numbers at the left of the immunoblot represent the molecular weight of the protein ladder (ii) Immunoprecipitation control (iii) Actin

loading control (C) (i) In vitro phosphorylation of His-tagged C protein is performed

with PKC in the absence (Lane 1) or presence (Lane 2) of kinase inhibitor The presence

of band in Lane 1 indicates that PKC phosphorylates C protein Phosphorylation of tagged C protein by PKC is reduced by PKC inhibitor (Lane 2) (ii) Input controls showing equal loading of His-tagged C protein

His-PKC(-)BIS PKA(-)H89

A PKA(+)forskolin

5.4 Effects of phosphorylation on myc-capsid (C) protein

5.4.1 RNA binding of myc-capsid (C) protein is attenuated by phosphorylation

Since it was shown that myc-C protein as well as the West Nile virus C protein from infected cells are phosphorylated the next question to ask is if phosphorylation can attenuate myc-C protein and RNA interaction as demonstrated with C peptides (Fig 4-10) Hence wild type myc-C protein, mutant myc-C protein as well as λ-phosphatase treated myc-C protein were subjected to Northwestern blot analysis Similarly to Figure 4-6, wild type myc-C protein was unable to bind to any viral RNA (Fig 5-6A, Lane 2) but the binding activity was restored through λ-phosphatase treatment (Fig 5-6A, Lane 3) Moreover, the hypophosphorylated mutant myc-C proteins were able to bind to viral RNA (Fig 5-6A, Lanes 4 to 6)

In order to complement the above assay, an RNA pull down assay was developed Both biotinylated-RNA and wild type, mutant or λ-phosphatase treated myc-C protein were mixed in RNA binding buffer overnight to allow C-RNA complex to form The following day, the RNA was pulled down with streptavidin beads conjugated to magnetic beads The beads were then washed and the proteins captured on the beads were released

by boiling The relative amount of proteins captured was visualized by Western blotting The results showed that the RNA pulled down most of λ-phosphatase treated myc-C protein (Fig 5-6B, Lane 2) and the S26/36/83/99/T100A mutant (Fig 5-6B, Lane 6) while only a trace amount of the wild type myc-C protein was pulled down (Fig 5-6B, Lane 1) Hence it can be concluded that phosphorylation of C protein attenuates its ability

to bind viral RNA and this ability was restored either through phosphatase treatment or

disruption of phosphorylation through mutation Hence, this result explains why initially myc-C in Figure 4-6 could not bind to viral RNA

Figure 5-6 Phosphorylation attenuates RNA binding (A) Plasmid encoding wild type

(WT) (Lanes 2 and 3) or mutant (Lanes 4, 5 and 6) myc-C proteins were transfected into 293FT cells and myc-C proteins were harvested with anti-myc antibody from the resulting cell lysates Some cell lysate were treated with λ-phosphatase (Lanes 3 - A and B) The cell lysates (Lanes 1 - 6) with purified His-C protein (Lane 7) were subjected to Northwestern (i) and Western blot (ii and iii) analyses with biotinylated 3’UTR viral RNA and anti-phosphoserine, anti-myc and anti-His antibodies respectively Mock-transfected cell lysate is in Lane 1 RNA binding activity is only observed with the λ-phosphatase-treated myc-C, mutant myc-C and His-C proteins (B) Plasmid encoding WT (Lanes 1, 2 and 3) or mutant (Lanes 4, 5 and 6) myc-C proteins were transfected into 293FT cells The resulting cell lysates (Lanes 1 - 6) and purified C protein (Lane 7) were mixed with biotinylated 3’UTR viral RNA (Lanes 1 - 2, 4 - 7) or unlabelled 3’UTR viral RNA (Lane 3) while some were treated with λ-phosphatase (Lanes 2 and 3) prior to mixing with biotinylated RNA The biotinylated RNA was then pulled down with streptavidin-conjugated magnetic beads The samples were then subjected to Western blot analysis and probed with anti-myc antibody (ii) and anti-phosphoserine antibody (iii) The RNA is only able to pull down λ-phosphatase-treated myc-C (Lane 2), mutant myc-C (Lanes 4 to 6) and His-C (Lane 7) proteins Panel (i) shows the equal amount of myc-C protein added to each tube for pull-down

5.4.2 Phosphorylation of myc-capsid (C) protein is needed for efficient nuclear

translocation

Although the initial aim was to elucidate the effect of phosphorylation on C-RNA interaction, it was observed that mutations at S26, S36, S83, S99 and T100 cause other effects on the functions of C protein When wild type myc-C proteins were expressed in BHK cells, they were predominantly localised in the nucleus [Fig 5-7A (iv)] These

corroborated with what other studies have observed (Bhuvanakantham et al., 2009; Oh et al., 2006) However, when mutant myc-C proteins were expressed in BHK cells, they

were found in the cytoplasm in greater proportion of transfected cells [Fig 5-7A (ix, xii, xv) – arrowheads]

For quantification the cells were divided into two categories – cells with myc-C protein localised exclusively in the nuclei and cells with myc-C protein localised in both nuclei and cytoplasm One-hundred cells were randomly selected and assigned into these two categories accordingly It was found that all the hypophosphorylated mutants had 80% or less cells with with myc-C protein found in the nucleus only In contrast, wild type myc-C was found in the nucleus only in about 90% of the cells (Fig 5-7B) Similar results were observed when cells expressing wild type myc-C proteins were treated with protein kinase C inhibitor (Fig 5-8A and B) In both instances (Fig 5-7 and 5-8), nuclear localisation of myc-C protein was disrupted by its inability to phosphorylate

Figure 5-7 Nuclear localisation of myc-C protein affected by mutation (A) BHK cells

were transfected with empty myc vector (i - iii), plasmid encoding WT (iv - vi) or mutant

C proteins (vii - xv) and probed with anti-myc antibodies Bound anti-myc antibody was detected with anti-mouse antibody conjugated to Alexafluor 488 (Green) The cells’ nuclei were stained with DAPI (Blue) Cells with myc-C protein localized exclusively in the nuclei are indicated with white arrows and cells where the C protein is localized both

in the nuclei and cytoplasm are indicated with white arrowheads Wild type (WT) myc-C protein is observed to localize mainly in the nuclei while the localisation pattern of the mutant myc-C proteins are more diffused (B) For each transfection, cells expressing myc-C protein were divided into two categories: cells with myc-C protein found exclusively in the nuclei and cells with myc-C protein found in both nuclei and cytoplasm A total of 100 cells were counted at random and assigned to its appropriate category The proportion of cells with myc-C protein found exclusively in the nucleus is observed to be greater for cells expressing the WT myc-C protein than for cells expressing mutant myc-C proteins *, p < 0.05 vs WT myc-C protein in nucleus

exclusively using student’s t-test

*

PKC(-)BIS DMSO PKA(-H89) PKA(+)forskolin PKC(+)PMA

5.4.3 Phosphorylation of myc-capsid (C) protein and oligomerization

Another effect that was observed to be affected by the mutations was oligomerization of C protein Capsid protein has been known to spontaneously dimerize and this was observed with purified His-C proteins (Fig 3-12, Lane 1) Oligomers of C protein were observed in transfected cell lysate as early as 24 hr post-transfection (Fig 5-

9, Lanes 3 - 5 and 13 - 15) The question to address was how oligomerization might be affected if C protein was hypophosphorylated In this experiment, the observation was that the oligomerization patterns between wild type and hypophosphorylated mutant myc-

C proteins expressed in 293FT or BHK cells, showed little difference Phosphorylation had little consequence on the quantity of each oligomer formed at 48 hr post-transfection However, the rate of oligomerization was affected by phosphorylation

The results showed that at 24 hr post-transfection, the S26/36/83/99/T100A mutant myc-C protein had already formed obvious oligomers (Fig 5-9, Lanes 5 and 15)

In contrast, oligomers were not observed with WT myc-C protein and only dimers were observed with the protein expressed in BHK and 293FT cells (Fig 5-9, Lanes 2 and 12) This observation was more apparent in BHK cells than in 293FT cells However, at 48 hr post-transfection this difference was no longer apparent (Fig 5-9, Lanes 6 and 9 and 16 and 19) In order to eliminate the possibility that the assay conditions induced self-oligomerization, a green-fluorescence protein (GFP) encoding plasmid was transfected into either 293FT or BHK cells and asked if oligomerization was induced The results showed that oligomerization of the GFP did not occur indicating that the assay conditions did not induce oligomerization of C protein (Fig 5-9 Lanes 10, 11, 21 and 22)

Figure 5-9 Mutation of putative phosphorylation sites and oligomerization of C protein

Plasmid encoding wild type (WT) or mutant C protein were transfected into 293FT cells (Lanes 2 - 9) or BHK cells (Lanes 12 - 19) and their cell lysates were harvested at 24 hr (Lanes 2 - 5 and 12 - 15) or 48 hr (Lanes 6 - 9 and 16 - 19) post-transfection Whole cell lysate was subjected to Western blot analysis with anti-myc antibodies The membranes were stripped and probed with anti-actin antibodies Mock-transfected 293FT and BHK cell lysates are in Lanes 1 and 20, respectively Plasmid encoding GFP (control) was also transfected into 293FT and BHK cells and the cell lysate was harvested at 24 hr (Lanes

10 and 21) and 48 hr (Lanes 11 and 22) post-transfection Oligomerization of myc-C protein is not affected by the mutations however, the rate of oligomerization is observed

to be faster for the mutant myc-C proteins since higher ordered oligomers are already present at 24 hr post-transfection (Lanes 3 - 5 and 13 - 15) while such oligomers are not observed for the WT myc-C protein (Lanes 2 and 12)

5.5 Phosphorylation of West Nile virus (WNV) capsid (C) protein diminishes

over time

If phosphorylation plays a regulatory role in modulating the functions of C protein then the phosphorylation status of C protein should be dynamic If the role of phosphorylation is to prevent premature association of viral RNA or even cellular RNA with C protein during the early stages of infection then C protein can be expected to be phosphorylated Conversely, at later stages of infection when most of the C protein would interact with viral RNA to form the nucleocapsid, phosphorylation is expected to dimes Hence, BHK cells were infected with WNV and C protein was harvested from the infected cell lysate at 6, 12, 18 and 24 hr post infection The relative level of phosphorylation of C protein at these time points was analysed by Western blot There was relatively no difference between the level of phosphorylation on C protein between 6 and 12 hr post infection (Fig 5-10), however, it was observed that the level of C protein phosphorylation was significantly lower at 24 hr post infection compared to all other time points (Fig 5-10) Hence, phosphorylation of C protein in infected cells is dynamic and it

is dephosphorylated at the later stages of infection