In-gel Digestion

After Flag-SHDAg was immunoprecipitated by anti-Flag M2 agarose beads, proteins were separated in SDS-PAGE. The gel was stained with coomassie blue and then destained gel with destain buffer (5% acetic acid, 30% methonal). The band corresponding to the FSHDAg band was cut off and transferred to a new microfuge tube.

The gel band was washed three times with 1ml of 25 mM NH4HCO3-50% acetonitrile for 10 minutes, and then dried in the Speed-Vac for 10 min. The protein was digested in 100µl of 25 mM NH4HCO3 containing 0.2 U of trypsin (Promega) per µg protein at 37℃ overnight. The supernatant was transferred to a new tube, then 100µl of 25 mM NH4HCO3 -50% acetonitril was added and incubated for 30 min. The supernatant was pooled with the previous one. They were completely dried in a Speed-Vac

centrifugation, and stored at -80℃ until analysis.

2.10 Analysis of Flag-SHDAg and SHDAg by Liquid Chromatography-Tandem Mass Spectrometry.

Before analysis, the in-gel digestion dried sample was lyophilized in 30µl solvent A (2% acetonitrile, 0.1% formic acid), and centrifuged at 13000 rpm speed for 5 min.

The supernatant was subjected for sequence analysis using a Q-STAR^TMXL Q-TOF (Applied Biosystems) coupled to an UltiMate^TM Nano LC system (Dionex/LC Packings). The samples were trapped and desalted for 5 min on a precolumn packed with PepMap^TM C18 100A(300μm ID x 5mm, Dionex, Sunnyvale, CA, USA) which was using solvent A at a flow rate of 30μl/min via a Switchos^TM pump (Dionex/LC Packings). The peptide were separated on an LC Nanocolumn packed with PepMap^TM C18 100A (3μm particle size, 75mm IDx150mm, Dionex) at a flow rate of 200nl/min by a gradient elution from 5% to 60% solution B (80% acetonitrile, 0.1% formic acid) in 65 min followed by an isocratic step at 95% solution B for 10 min. The peak lists were uploaded to Mascot MS/MS Ions Search program (Mascot version 2.0) on the Matrix Science public web site and the identification peptide was matched in NCBI nr database. The MH22+ and MH33+ were selected as the precursor peptide charge states in the searching. The error windows for peptide and MS/MS fragment ion mass values were 0.3 and 0.5 Da, respectively. Quantitative data were obtained from peptides by inputting their m/z values and retention times provided with the Analyst QS1.1 software.

The software obtains extracted ion chromatograms for each of the input m/z values, which together with retention time (63, 93). For the part II section on the proteomic study of SHDAg, the purified SHDAg were in-gel digested with Trypsin (Promega),

Asp N and GluC (Roche) and subjected to LC/MS/MS mass-spectrometry. Peptide sequencing experiments utilizing LC-MS/MS were performed on an LCQ Deca mass spectrometer (Thermo Finnigan, San Jose, CA) and Q-STAR^TMXL Q-TOF.

2.11 Fractionation of Nuclear and Cytoplasmic Extracts.

To extract the nuclear fraction from HeLa S3 SHDAg cell line, we culture HeLa S3 SHDAg cell line in 10 % FBS in suspension condition. Cell were washed three times with PBS, re-suspended in 10-fold volume buffer A (10 mM HEPES-KOH [pH 7.9], 1.5 Mm MgCl2, 10 mM KCL, 10 mM NaF, 0.5 mM DTT with added protease inhibitors(Roche)) 30 min at 4℃, and dounce homogenized 60 times using a B type pestle. Dounced nuclei were centrifuged at 228 × g for 5 min at 4°C. Collect the nuclear pellet and storage the supernatant, labeled cytoplasm fraction. The pelletednuclei were collected for nucleoli purification or washed once with buffer A and then used for SHDAg cation chromatography purification. The samples were lysis in binding buffer and briefly sonicated, and centrifuged at 13,000 xg for 15 min at 4 °C. The supernatants were kept as nuclearextracts for cation chromatography purification.

2.12 Nucleoli Purification.

To extract the high purity nucleolus from the nuclear fraction, nucleoli were prepared from HeLa S3 SHDAg cell line nuclei, using a method based on that first described by Muramatsu and coworkers in 1963 (1). The nuclear pellet was resuspended in 5 volume buffer B: 0.25 M sucrose, 10 mM MgCl2, and layered over the same volume buffer C: 0.35 M sucrose, 0.5 mM MgCl2, and centrifuged at 1430 × g for 5 min

at 4°C. The clean, pelleted nuclei were resuspended in 3×10 ml 0.35 M sucrose, 0.5 mM MgCl2, and sonicated for 6 × 10 s using a microtip probe and a Misonix XL 2020 sonicator at power setting 5. The sonicate was checked using phase contrast microscopy, ensuring that there were no intact cells and that the nucleoli were readily observed as dense, refractile bodies. The sonicated sample was then layered over 3×10 ml 0.88 M sucrose, 0.5 mM MgCl2 and centrifuged at 2800 × g for 10 min at 4°C. The pellet contained the nucleoli, while the supernatant consisted of the nucleoplasmic fraction.

The nucleoli were then washed by resuspension in 500 µl of 0.35 M sucrose, 0.5 mM MgCl2, followed by centrifugation at 2000 × g.

2.13 Cation Chromatography.

After the nucleolus, nucleoplasm, and cytoplasm three fractions was fractionated well. The SHDAg could be purified from the nucleoli, nucleoplasm, and cytoplasm fractions. We use the Bio-CAD700E machine for purify the SHDAg. All the fractions should be incubated with binding buffer (HEPES-KOH buffer, pH7.9, with or without the 0.5 M NaCl, 1% NP40 and 1% Trixton X100). The samples were sonicated at 4℃

by the sonicator and filtrated the 0.22 μm filter. It could be loading to HS (or CM) cation chromatography. The beads were packed with POROS HS (or CM) perfusion chromatography medium (PerSeptive Biosystems, Framingham, MA, USA). After binding, we used the binging buffer wash 10 column volume or 20ml buffer, then used the elution buffer HEPES-KOH, pH7.9, with 0.5M or 2 M NaCl gradient buffer to elute SHDAg. The eluted fractions were collected on the cation chromatography. After protein was purified, the protein was digested with trypsin or AspN and Glu C. The digested peptides were analysis by the LC/MS/MS to identify the post-translational

modification patterns.

2.14 Immunofluorescence Microscopy Analysis.

The subcellular distribution of SHDAg proteins was analyzed after 48h transfection. Cells grown on coverslips were washed three times for 5 minutes with PBS, and fixed with 4% paraformaldehyde for 10 minutes at room temperature (RT).

Following PBS washed the coverslips for 5 minutes three times, the cells were permeabilized with 0.2% Triton X-100 in PBS for 5 min, washed three times in PBS for 5 minutes and blocked in 10% BSA for 30 minutes at RT. The coverslips were incubated with the primary antibody which was diluted in 0.1% Tween 20 for 1 hour at RT (anti-SHDAg dilution, 1:200). The coverslips were then washed three times for 5 minutes in 0.1% Tween 20, blocked in 10% fetal serum for 10 minutes at RT and incubated for 1 hour at RT with the secondary antibody (Texus Red-conjugated donkey anti-human to detect SHDAg, dilution 1:200). The coverslips were washed three times for 5 minutes in 0.1% Tween 20 with DAPI (100 pg/ml) and then mounted with Gel/Mount. The cells were then examined by fluorescence microscopy.

3. RESULTS

Part I. ERK1/2-mediated Phosphorylation of SHDAg at Serine-177 enhances HDV Replication from Antigenomic RNA to Genomic RNA Through RNA Polymerase II Regulation

3.1 The Kinase of SHDAg Serine 177 Phosphorylation.

The contributionsof phosphorylation to SHDAg function have been addressed by mutating Ser-177 or examining the effectsof kinase inhibitor in HDV replication (4, 17, 39, 73, 75). However, the kinase of SHDAg at Ser-177 is never fully understood. In order to identify the kinase for SHDAg Ser-177 phosphorylation, we used the Scansite database (http://scansite.mit.edu/) to look for the clues for the likely enzymes. The searching at low stringency gave a short list of Proline-directed kinases, including CDC2 (CDK1), CDK5, Erk1/2, and p38, with decreasing probabilities (Table 1).

Since direct interaction is a required step for the catalysis of protein phosphorylation, we thus examined whether SHDAg might associate with the enzymes in the above list.

It has been reported that cystein (Cys) mutation at the phosphorylation target residue may stabilize the in vivo interaction between kinase and substrate (48, 69). Hence, Flag-tagged SHDAg mutants, S177C and S2C, along with the wild-type protein were engineered and expressed in HEK293T cells. The amino acid in the serine-2 of SHDAg is highly conserved in HDV strains, therefore, we use S2C as the control in this experiment.

Western blotting showed that these mutations did not affect the SHDAg amounts

in transfected cells and they could be effectively immunopurified using Anti-Flag antibody (Fig. 1).The Erk1/2 kinase was associated with Flag-SHDAg^S177C (Fig. 1, lane 8) but that was not detected in the Flag-SHDAg^WT and Flag-SHDAg^S2C immunoprecipitates (Fig. 1, lanes 6 and 7). In contrast, the CDC2 (CDK1), CDK5, and P38 kinases were not identified in the immunoprecipitates containing either wild type or mutant SHDAg protein. This result suggests that ERK1/2 kinase may associate with SHDAg in vivo and Ser-177 played a role in SHDAg-ERK1/2 interaction.

3.2 MEK1–Mediated Activation of ERK1/2 Induces the Phosphorylation of SHDAg at Ser-177.

The Ser177-directed interaction of SHDAg by ERK1/2 prompted a possibility that ERK1/2 might be a kinase that could phosphorylate SHDAg at Ser-177. To address this, a constitutively active form of MEK1, HA-AcMEK1, was expressed in HEK293T cells, which catalyzed ERK1/2 phosphorylation and thus activated their enzyme activities in a dose-dependent fashion (Fig. 2A) (79). The protein expression of Flag-SHDAg was not perturbed by either the MEK1 or ERK1/2 activities. The lysates from cells expressing Flag-SHDAg protein with or without coexpression of HA-AcMEK1 were immunoprecipitated with anti-Flag antibody-conjugated resin and the proteins were run with SDS-PAGE (Fig. 2B).

After the gel was stained with Commassie Blue, the protein bands of interest were excised and subjected to mass spectrometric analysis (see materials and methods). The Ser-177 residue is located in the peptide, 161GAPGGGFVPNLQGVPESPFSR181, which was extracted from the in gel tryptic digestion. The non-phosphorylated isoform of

161GAPGGGFVPNLQGVPESPFSR181 have m/z values of 1035.97 and 690.99 for its

doubly and triply charged ions, respectively. This non-phosphorylated peptide was consistently eluted at ~51.2 min in different experiments with or without HA-AcMEK1 coexpression (Fig. 2C). Meanwhile, the phosphorylated peptide has m/z values of 1075.97 and 717.65 corresponded to the doubly and triply charged ions, was eluted closely with the non-modified counterpart at 49 min (Fig. 2D). This close elution property is consistent with our previous observations on many phosphorylated peptides (93). The collision-induced dissociation (CID) spectra of the phosphorylated peptides were identified (data not shown), and confirmed that Ser-177 was the phosphorylated residue within the m/z 1075.97 peptides recorded in our previous study (17).

We examined the mass spectrometric data to quantitatively estimate the extent of Ser-177 phosphorylation in vivo using the extracted ion chromatography (XIC) analysis.

Without over-expression of HA-AcMEK1, (Fig. 2D, panel a), XIC analysis showed the intensity of the Ser-177 phosphorylated was as low as 24 cps. However, in the presence of 0.2 μg pHA-AcMEK1 transfection, the intensity of Ser-177 phosphorylation was increased to 120 cps (Fig. 2D, panel b) and was even heightened up to 180 cps when the MEK1 protein expression was further augmented (Fig. 2D, panel c). The intensity of Ser-177 phosphorylation was enhanced for 7.5-folds increase, when compared to that in the cells without HA-AcMEK1 expression. Taken altogether, these data suggest that MEK1-mdiated activation of ERK1/2 is sufficient to enhance SHDAg phosphorylation at Ser-177 in vivo in a dose-dependent manner.

3.3 ERK1/2–Modulated Phosphorylation of SHDAg at Ser 177 is Specifically Recognized by a pS177 Antibody.

By LC/MS/MS analysis, we found that the phosphorylation level of SHDAg at

Ser-177 is increased with HA-AcMEK1 activation in a dose-dependent manner. To further demonstrate that ERK1/2 activation enhances phosphorylation of SHDAg at Ser-177. We probed lysates through Western blot analysis using an antibody specifically against phosphorylated Ser-177, αpS177. This antibody only targeted phosphorylated SHDAg since the prior treatment of the lysate proteins with λ-phosphatase could completely remove the blotting signals (Fig. 3, lane 4). This antibody did not react with S177A mutant (Fig. 3, lanes 5 and 6), which implicates that Ser-177 phosphorylation is required for its reactivity against SHDAg. More importantly, this antibody could monitor the dose-dependent increase in phosphorylation state of SHDAg protein in response to in vivo MEK1 activation (Fig. 3, lanes 2 and 3). These data verify the effectiveness of this αpS177 antibody in quantitative assessment of Ser-177 phosphorylation of SHDAg protein.

3.4 ERK1/2 Phosphorylates SHDAg at Ser-177 in the in vitro Kinase Assay.

To examine whether ERK1/2 kinases could catalyze Ser-177 phosphorylation of SHDAg, we employed a cell-free in vitro kinase assay using highly purified components to examine the enzyme activity on purified SHDAg proteins. The kinases, Flag-ERK1 and Flag-ERK2, expressed in the cells were first activated by treatment with the 20%

serum activation for 30 min, and then Flag-tagged kinases were purified with immunoprecipitation under more stringent conditions. The purity of the enzymes and substrates was verified by silver staining of the polypeptides present in these in vitro reactions (Fig. 4, panel b). Without expression of cellular Flag-tagged proteins, no phosphorylation signals could be detected by αpS177 antibody. In contrast, either Flag-ERK1 or Flag-ERK2 kinase purified with immunoprecipitation could specifically mediate the in vitro phosphate transfer onto wild type Flag-SHDAg protein in a

ATP-dependent fashion (Fig.4, lanes 6 and 8, panel a). These data strongly suggest that ERK1/2 is sufficient to directly phosphorylate SHDAg at Ser-177 residue in vitro.

Taken together, our results indicate that ERK1/2 are kinases capable of catalyzing phosphorylation of SHDAg at Ser-177 both in vivo and in vitro.

3.5 The Activity of HDV Replication from Antigenomic RNA to Genomic RNA is Enhanced by ERK1/2 Kinase Activity Induced by HA-AcMEK1.

We have established that ERK1/2 can catalyze the in vitro and in vivo phosphorylation SHDAg at Ser-177. In previous reports, the Ser-177 residue of SHDAg is crucial for the replication of HDV genomic RNA using the antigenomic RNA as the template (Fig. 5) (73). We wondered whether ERK1/2-mediated Ser-177 phosphorylation might stimulate this RNA replication process.

To test this possibility, we over-expressed constitutively active MEK1 proteins to activate ERK1/2 kinases, and to examine how their activation might affect the HDV production of genomic RNA in a cellular replication system. This system requires the presence of both SHDAg protein and dimeric antigenomic RNA template that is initially transcriptionally synthesized using the pCDm2AG plasmid. The RNA-dependent RNA replication can proceed to both antigenome-to-genome (AG→G) and genome-to-antigenome (G→AG) directions in these cells. It is noteworthy that, although antigenomic RNA can be synthesized in both DNA- and RNA-dependent processes, its amplification is only observed when RNA-dependent replication cycle is established. Thus, antigenomic RNA is usually not detectable in cells only bearing pCDm2AG without the co-expression of SHDAg protein (Fig. 5, lane 3).

In the presence of active ERK1/2 proteins, Ser-177 phosphorylation could be

stimulated dramatically. When more phosphorylated ERK1/2 was present, there existed a significant increase in Ser-177 phosphorylation state in a dose-dependent fashion (Fig.

5 D, lanes 5-7). When MEK1 activity increased, there was a concomitant accumulation of genomic RNA using Northern blot analyses. Paradoxically, the antigenomic RNA was not induced correspondingly when the genomic RNA was more accumulation by ERK1/2 activation in cells (Fig. 5A). These data suggests that activated MEK1-ERK1/2 pathway may increase the efficiency of AG→G replication, while reciprocally decreasing the efficiency of G→AG replication (Fig. 5A, lanes 5-7). Meanwhile, genomic RNA was not detectable when S177A SHDAg mutant was introduced. This is consistent with the notion that this mutant cannot provided the functional activity required for AG→G replication (73).

We also examined the differential effects of MEK1-ERK1/2 pathway on the two directions in the replication cycle using the transcriptionally synthesized genomic RNA as the initiator templates. In the cells expressing pCDm2G plasmid, a similar reciprocal change was seen in the expression of antigenomic and genomic RNAs. The genomic RNA accumulated as MEK activity increased, whereas there was a dose-dependent decrease in antigenomic RNA amount (Fig. 6A, lanes 5-7). As expected, little HDV replication was observed in the system complemented with S177A SHDAg protein (Fig.

6A, lanes 8-10). In accordance with the above results, these observations strongly suggest that MEK1-ERK1/2 pathway, probably through SHDAg Ser-177 phosphorylation, has differential effects on AG→G and G→AG replications.

3.6 The U0126 Reduce the Phosphorylation Level of Ser-177 and Inhibit HDV Replication Activity on Antigenomic RNA to Genomic RNA.

To further confirm the role of ERK1/2-dependent pathway in HDV replication from antigenomic RNA to genomic RNA, we had performed DNA-free HDV antigenomic RNA transfection system. To do this, the pHA-AcMEK1 and pCDSHDAg plasmids were co-transfected with DNA-free dimeric AGm RNA into 293T cells. We blocked the ERK1/2 activities by pharmacological inhibition drug and examined the DNA-free HDV antigenomic RNA replication. As shown in our above result, the HA-AcMEK1 induced high level expression of pERK1/2 and pSer-177 (Fig.7, lane 1). We found that HA-AcMEK1-activated phosphorylation of ERK1/2 was abolished by UO126 (U0126, MEK1 inhibitor), which was accompanied by a decrease in the Ser-177 phosphorylation (Fig. 7, lane 2) as compared to the cells without U0126 treatment (Fig. 7, lane 1). Under these experimental condition, the genomic RNA accumulation from AG to G replication in the cells was reduced (Fig.7, lanes 4). Since the role of the MEK1-ERK1/2 pathway was examined under constitutive MEK1 overexpression conditions, it is not clear whether our findings are physiologically relevant. It would be interesting to know whether U0126 has any effect on Ser177 phosphorylation and HDV replication without MEK1 overexpression. The experiment shown in Figure 8 was performed after prior treatment of cells with UO126 reagents that inhibit the MEK1–ERK1/2 pathway in Huh7 cells. U0126 treatment decreased Ser-177 phosphorylation to the lower level in a dose-dependent manner (Fig. 8A, lanes 3 and 4). The low concentration of 1 μM U0126 did not inhibit HDV replication (Fig.

8B, lane 3), although 10 μM U0126 decreased HDV (Fig. 8B, lane 4). The U0126 studies suggest that the MEK1–ERK1/2 MAPK pathway regulates HDV AG→G replication by modulating the phosphorylation level of SHDAg at Ser-177.

3.7 The Serine 177 is a Critical Amino Acid for Interacting with RNA Polymerase

II.

In previous studies, the Ser177 is phosphorylated and important for HDV replication from antigenomic RNA to genomic RNA (73, 75). The AG to G replication mechanism is α–amanitin sensitive (15, 68). The cellular RNA polymerase II is a α–amanitin sensitive DNA-dependent RNA polymerase. However, many papers showed that RNA polymerase II can help HDV RNA replication in vitro without the DNA template (13, 26, 59, 102, 103). The Ser-177 within the C-terminal region of SHDAg is associated with RNAPII (102). It is possible that SHDAg phosphorylation at Ser177 may play a role in HDV replication on genomic RNA synthesis through regulating the RNAPII.

To study this, we first investigated the role of Ser177 residue in SHDAg-RNAPII complex. We genetically generated a series N-terminal Flag-tagged SHDAg serine mutation constructs, termed Flag-SHDAg ^S177A (serine 177 change to alanine for mimic the non-phosphorylation form), Flag-SHDAg ^S177C (serine 177 to cysteine), Flag-SHDAg ^S177D (serine 177 to aspartic acid for minic the phosphorylation form), Flag-SHDAg ^S2A (serine 2 change to alanine), and Flag-SHDAg ^S2C (serine 2 change to cysteine). The amino acid in the Ser-2 of SHDAg is highly conserved in HDV strains, and we used S2C and S2A as the controls in this experiment. The large delta antigen shares the same amino-acid sequences, except for the additional 19-amino acid extension in the C termini. The Flag-LHDAg was also constructed and examined in the experiment. All the pFlag-HDAg plasmids were transiently introduced into 293T cells, and the RNAPII was co-immunoprecipitated with anti-FLAG M2 antibody-bound resin (Sigma) to examine the RNAPII-HDAg associations. The endogenous RNAPII was efficiently recruited by Flag-SHDAg, Flag-SHDAg S2C, and Flag-LHDAg immune-complexes (Fig. 9A, lanes 2, 3 and 5). However, the endogenous RNAPII was

not recruited with Flag-SHDAg S177C as the wild type SHDAg did (Fig 9A, lane 4).

The RNA polymerase I, RPA135, can be recruited with the Flag-SHDAg even the mutagenesis in serine 177 amino acid (Fig 9A, lane 4). This raised an interesting possibility that active turnover of S177 phosphorylation is required for the interaction, since both the mimic for non-phosphorylated and phosphorylated S177 did not associate with RNAPII. The ERK1/2 was also co-immunoprecipitated with Flag-SHDAgS177C (Fig 9A, lane 4). Both of immunoprecipitated RPA135 and ERK1/2 could be the positive control in the experiment.

There are many studies showing that a mutation to aspartic acid can mimic the serine

There are many studies showing that a mutation to aspartic acid can mimic the serine