Results and discussion

Subcellular localization of the NS5 protein during DENV infection in Vero cells

To determine the subcellular location of the NS5 protein during dengue virus infection, Vero cells infected with DENV-2 NGC at an MOI of approximately 10 were examined at specific stages p.i., by indirect immunofluorescence using a poly-clonal anti-NS5 and an Alexa Fluor® 488 green fluorescent dye-conjugated secondary antibody (Fig. 2A). Prior to 12 h p.i. (data not shown), NS5 was unable to be detected in either the cytoplasm or the nucleus of infected Vero cells, supporting the long latent period of DENV-2 infections described by Cleaves et al (1981). At 12 h p.i., a small percentage of cells showed staining for NS5 in both the cytoplasm and the nucleus of infected cells although the staining was predominantly nuclear. After 20 h p.i., the number of cells staining NS5 significantly increased, with the predominantly nuclear localization of NS5 and the diffuse cytoplasmic staining continuing throughout the virus infection. The observed localization of NS5 in the nucleus early in infection is surprising in view of the fact that NS5 has a major role in cytoplasmic replication of the virus, although it is known that only small quantities of NS5 are required for replicative activity, as shown for several flaviviruses (Chu & Westaway 1992, Grun & Brinton 1987, Uchil & Satchidanan-dam 2003). NS5 has also been detected in the nucleus of yellow fever virus infected cells (Buckley et al 1992), but not in the case of Kunjin and Japanese encephalitis virus (Edward & Takegami 1993, Westaway et al 1997). Several other flavivirus proteins have been detected in the nucleus during infection; including the capsid protein of dengue (Bulich & Aaskov 1992, Makino et al 1989, Wang et al 2002),

12 1B 24 4O

hours p.i.

12 1B 24 4O

hours p.i.

and Japanese encephalitis (Mori et al 2005) and both the capsid and NS4B proteins of Kunjin (Westaway et al 1997).

LMB treatment increases nuclear localization of DENV-2 NS5 in infected cells

The presence of DENV-2 NS5 in the nucleus of infected cells very early in infection prompted us to investigate if NS5 may shuttle between the nucleus and cytoplasm during infection. To assess the possibility that NS5 may possess a CRM1 dependent nuclear export pathway, Vero cells infected with DENV-2 were treated with LMB at 0 h p.i., and NS5 nuclear accumulation compared to that of non-treated samples at specific stages during infection (Fig. 2A). Image analysis of digitized CLSM images (see Materials and Methods) was performed to determine the nuclear to cytoplasmic ratio (Fn/c) for the infected cells treated without or with LMB (Fig. 2B). All samples had a mean Fn/c of greater than 2, indicative of nuclear accumulation, although the Fn/ c was significantly higher at each time point in the case of the LMB-treated cells. Further, nuclear accumulation peaked more rapidly (at c. 24 h) in the presence of LMB, compared to c. 30 h in its absence. These results suggest that DENV-2 NS5 possesses the ability to be exported from the nucleus by CRM1, which presumably represents a mechanism to control NS5 subcellular localization during infection. This would appear to be of great importance since all of the known functions of DENV-2 NS5 occur in the cytoplasm of infected cells.

LMB treatment of DENV-2 infected cells was also found to increase the percentage of cells showing NS5 nuclear localization during the early stages of infection (Fig. 2C). At 12 and 18 h p.i., the number of cells showing nuclear staining of NS5 was significantly higher than in untreated cells. This difference was insignificant at times later than 18 h p.i, when the majority of cells treated without or with LMB show NS5 nuclear localization. These results show that blocking the CRM1 nuclear export pathway can alter the kinetics of NS5 nuclear accumulation during infection.

FIG. 2. Nuclear accumulation of NS5 in DENV-2-infected Vero cells is increased by treatment with the CRM1 inhibitor LMB. (A) CLSM images of the subcellular localization of NS5 in DENV-2 infected Vero cells, treated without and with LMB (5 ng/ml), as visualized by indirect immunofluorescence. Mock (left panels) or DENV-2-infected Vero cells were fixed at the indicated times p.i. and stained using polyclonal anti-NS5 antiserum and a goat anti-rabbit Alexa Fluor® 488 green fluorescent dye-conjugate. (B) Quantitative analysis of the levels of NS5 nuclear accumulation in DENV-2 infected Vero cells ± 06 LMB. Mean Fn/c values (n = > 70) ± SEM were calculated from CLSM images such as those in (A) using the Image J public domain software as previously (Harley et al 2003). (C) The percentage of DENV-2-infected Vero cells (±LMB) showing NS5 nuclear localization. NS5 nuclear localization was determined and expressed as a percentage of the total cells (n > 110, except LMB treated virus at 60 h p.i., where n = 37).

LMB treatment increases NS5 nuclear accumulation in transfected cells

To study the subcellular localization of DENV-2 NS5 independently of the other DENV-2 proteins, plasmid pEPI-GFP-NS5, encoding GFP fused N-terminally in frame with the full-length DENV-2 NS5, was transfected into Vero cells treated with and without LMB treatment for 5 h prior to live cell imaging by CLSM at 20-24 h post transfection (p.t.) (Fig. 3A). In the absence of LMB, GPF-NS5 was

FIG. 3. Nuclear accumulation of GFP-NS5 fusion protein in transfected Vero cells is increased by treatment with the CRM1 inhibitor LMB. (A) CLSM images of the subcellular localization of GFP-NS5 in transfected Vero cells treated without and with LMB. Cells were treated with LMB (2.8 ng/ml) 5 h prior to live cell imaging by CLSM 20-24 h p.t. (B) Quantitative analysis of the levels of GFP-NS5 nuclear accumulation in Vero cells ± LMB. Mean Fn/c values (n = > 80) ± SEM were calculated from CLSM images such as those in (A) using the Image J public domain software.

GFP-NS5 GFP GFP-Rev

FIG. 3. Nuclear accumulation of GFP-NS5 fusion protein in transfected Vero cells is increased by treatment with the CRM1 inhibitor LMB. (A) CLSM images of the subcellular localization of GFP-NS5 in transfected Vero cells treated without and with LMB. Cells were treated with LMB (2.8 ng/ml) 5 h prior to live cell imaging by CLSM 20-24 h p.t. (B) Quantitative analysis of the levels of GFP-NS5 nuclear accumulation in Vero cells ± LMB. Mean Fn/c values (n = > 80) ± SEM were calculated from CLSM images such as those in (A) using the Image J public domain software.

found predominantly in the nucleus, in contrast to the GFP control that showed a diffuse localization throughout the nucleus and the cytoplasm. That GFP-NS5 was nuclear in the absence of the other DENV-2 proteins supported the previous observation that NS5 contains a functional NLS (Brooks et al 2002). Analogous to the results in DENV-2 infected cells, LMB effected a significant (P < 0.0001) increase in the Fn/c compared to untreated samples (Fig. 3B). These results confirm that NS5 possess a CRM1-dependent pathway, and thereby is capable of shuttling between the nucleus and cytoplasm in the absence of the other DENV-2 proteins. A plasmid expressing the CRM1-recognised HIV-1 Rev protein fused to GFP (GFP-Rev) was also transfected into Vero cells as a control. In the presence of LMB, GFP-Rev is localized in the nucleus and nucleolus of cells, a dramatic change from the largely cytoplasmic staining in untreated cells (Fig. 3A, 3B).

CRM1-recognized NESs are typically short peptide sequences with regularly spaced leucine or hydrophobic amino acids. Examination of the DENV-2 NS5 protein sequence has revealed several candidate NESs; these sequences are currently being evaluated by site-directed mutagenic approaches.

Effect of LMB on the DENV-2 NGC virus production

The observation that LMB treatment of DENV-2-infected Vero cells altered the kinetics of NS5 nuclear localization prompted us to test if LMB treatment affects virus production. Viral titres were determined from culture supernatants harvested at various time points in DENV-2-infected Vero cells treated in the absence without or with LMB (Fig. 4). In the presence of LMB, a marked increase in virus production was detected for all the time points excluding 72 h p.i. These results support our earlier findings that blocking the CRM1 nuclear export pathway altered the kinetics of NS5 nuclear accumulation early in infection (Fig. 2C). Taken together, these results suggest that LMB treatment of DENV-2-infected Vero cells results in increased accumulation of NS5 within the nucleus, which correlates with the acceleration of the rate of virus production.

Was this article helpful?

0 0
Swine Influenza

Swine Influenza

SWINE INFLUENZA frightening you? CONCERNED about the health implications? Coughs and Sneezes Spread Diseases! Stop The Swine Flu from Spreading. Follow the advice to keep your family and friends safe from this virus and not become another victim. These simple cost free guidelines will help you to protect yourself from the swine flu.

Get My Free Ebook


Post a comment