The MAPKs are activated by a variety of cellular stimuli to participate in a series of signaling cascades and mediate diverse intracellular responses. One potential target of the MAPKs is Op18/stathmin, a molecule that acts as an integrator of diverse cell signaling pathways and regulates the dynamics of microtubules, which are involved in modulating a variety of cellular processes, including cell cycle progression and cell growth. Our study focused on the regulation of the MAPK-mediated Op18/stathmin signaling pathway, which is triggered by the Epstein–Barr virus-encoded latent membrane protein 1 (LMP1) oncogene in nasopharyngeal carcinoma cells. The results showed that the activity of MAPK, which was induced by LMP1, varied with cell cycle progression; LMP1 upregulated phosphorylation of ERK during the G1/S phase, but negatively regulated phosphorylation of ERK during the G2/M phase. We found that the regulation of Op18/stathmin signaling by LMP1 was mainly mediated through ERK. The inhibition of LMP1 expression attenuated the interaction of ERK with Op18/stathmin and promoted microtubule depolymerization. These findings indicate the existence of a new cell cycle-associated signaling pathway in which LMP1 regulates ERK-mediated Op18/stathmin signaling. (Cancer Sci 2012; 103: 993–999)
The Epstein–Barr virus (EBV) is a prototype gamma herpes virus. Infection with EBV is mainly characterized by the expression of type II latent proteins, including latent membrane protein (LMP) 1 and LMP2 in nasopharyngeal carcinoma (NPC).[1-3] LMP1, an EBV-encoded oncogene, abnormally activates the nuclear factor-κB, activator protein-1, and signal transducer and activator of transcription signaling pathways by phosphorylating epidermal growth factor receptor, Jun N-terminal kinase (JNK), Janus-activated kinase, and other signaling molecules in NPC cells.[1-4] However, the full complement of LMP1-regulated proteins has not been thoroughly elucidated. Recently, we identified phosphorylation sites on 25 new components of the LMP1 signaling pathway, including oncoprotein 18 (Op18)/stathmin, heat shock protein 27, annexin I, and several kinases.
A novel target of LMP1, Op18/stathmin is a highly conserved small cytosolic phosphoprotein. Op18/stathmin that is overexpressed in tumors, including leukemia and breast carcinomas, regulates microtubule dynamics. During the cell cycle, Op18/stathmin integrates different signals to regulate microtubule polymerization and depolymerization, and its activation adapts to the phase of the cell cycle. Recently, we showed that LMP1 accelerates the cell cycle progression by cdc2-mediated Op18/stathmin phosphorylation during the G2/M phase. A dynamic microtubule equilibrium is crucial to a series of biological features, including cell morphology stabilization, substance transportation, cell division and proliferation, and cell migration and invasion. The level of Op18/stathmin expression also correlates with pathologic features and outcomes in the clinic.
Numerous kinases, such as MAPKs, which are associated with cell proliferation, regulate Op18/stathmin signaling. The MAPK family includes three main subfamilies: ERK; JNK; and p38. A typical Ser/Thr kinase, MAPK can interact with wild-type or mutant Op18/stathmin. Activated p38, ERK, and JNK can phosphorylate Op18/stathmin at Ser25 and Ser38; p38 showed the strongest activity, and JNK showed the weakest activity.[11, 12] The clustering of prions, which induces ERK1/2 and stathmin phosphorylation at the surface of GT1-7 cells, suggests that MAPK might mediate the regulation of Op18/stathmin signaling by LMP1.
In this study, we showed that LMP1 promoted the phosphorylation, but not the expression, of Op18/stathmin. We also determined that the LMP1-induced MAPK activity was not constant but rather changed with cell cycle progression. LMP1 upregulated the phosphorylation of MAPK mainly during the G1/S phase, but the activity of MAPK was negatively regulated by LMP1 during the G2/M phase. The main pathway regulated by LMP1 is the ERK/MAPK pathway. The inhibition of LMP1 expression significantly attenuated the interaction of ERK with Op18/stathmin. The mechanism by which LMP1 controls MAPK to regulate Op18/stathmin varies with the cell cycle phase.
Materials and Methods
Cell lines and cell culture
The CNE-LMP1 cell line, which was established in our laboratory, stably expresses LMP1. The parental cell line CNE1, which was derived from poorly differentiated NPC cells, does not express LMP1 and was used as a control in this study. Tet-on LMP1 HNE2 is an NPC cell line in which the expression of LMP1 is turned on by doxycycline. The NP69, NP69-pLNSX, and NP69-LMP1 cells were propagated in defined keratinocyte-SFM (Gibco, Life Technologies, Basel, Switzerland) supplemented with growth factors and maintained at 37°C with 5% CO2. The parent cell line NP69 is an SV40-transformed line with previously described characteristics. All of the cells were cultured as previously described[14-16] and were harvested during the logarithmic growth phase for experimental purposes.
Western blot analysis, antibodies, and chemical reagents
Western blotting was carried out as previously described.[8, 14] The primary antibodies used included mouse monoclonal anti-tubulin (B-7; Santa Cruz Biotechnology, Santa Cruz, CA, USA), mouse monoclonal anti-EBV LMP1 (Dako, Glostrup, Denmark), mouse monoclonal anti-β-actin (Sigma, St Louis, MO, USA), rabbit anti-phosphoserine (Zymed Laboratories, San Francisco, CA, USA), rabbit polyclonal anti-ERK1 (C-16; Santa Cruz Biotechnology), mouse monoclonal anti-phospho-ERK1/2 (sc-7383; Santa Cruz), rabbit anti-phospho-SAPK/JNK (Thr183/Thr185; Cell Signaling Technology, Danvers, MA, USA), rabbit anti-JNK (Cell Signaling Technology), rabbit anti-phospho-p38 (sc-535; Santa Cruz Biotechnology), and rabbit anti-p38 (sc-17825-R; Santa Cruz Biotechnology). Doxycycline was obtained from Sigma and PD98059, a specific MEK1/2 inhibitor, was obtained from Cell Signaling Technology (cat. #9900). DZ1, an LMP1-targeting DNAzyme that effectively inhibits LMP1 expression, was used as previously described. Paclitaxel and colcemid were kindly provided by the Xiangya Hospital of Central South University (Changsha, China).
Cell synchronization and cell cycle analysis
Cell synchronization was carried out with the MESACUP cdc2 kinase assay kit (code No. 5235; MBL, Nagoya, Japan). When the cells neared confluence, 20 ng/mL colcemid was added to the culture medium. The mitotic cells were harvested 14 h later.
Cell cycle distributions were analyzed by flow cytometry. The harvested cells were washed with ice-cold PBS, suspended in 75% ethanol, and stored at −20°C. The cells were stained with 50 mg/mL phosphatidylinositol and 1% RNase A in 400 mL PBS in an opaque tube at 25°C for 30 min. The stained cells were assayed using the FACS, and the cell cycle parameters were determined using the CellQuest software program (Becton Dickinson, Franklin Lakes, NJ, USA).
Construction and transfection of a plasmid that targets Op18/stathmin with siRNA
We constructed the plasmid pGC-silencer-U6/Neo/GFP-RNAi, which targeted the coding region of Op18/stathmin at 5′-AGAGAAACTGACCCACAAA-3′ (GenBank no. 53305, sequence 374-393). The siRNA sense strand was 5′-GAAGAGAAACTGACCCACAAA-3′, and the antisense strand was 5′-TTTGTGGGTCAGTTTCTCTTC-3′. Both of the siRNA strands were inserted into the BamH1/HindIII restriction sites of pGC-silencer-U6/Neo/GFP. pGC-silencer-U6/Neo/GFP-non contains a non-coding sequence that was inserted into the plasmid; pGC-silencer-U6/Neo/GFP is an empty plasmid. The shRNA plasmids were transfected into the cells with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) following the instructions in the manual.
Extraction of solubilized and polymerized tubulin
Cells were harvested and washed with ice-cold PBS, lysed in microtubule stabilization buffer (0.1 M PIPES, pH 6.9, 2 M glycerol, 1.5 mM MgCl2, 2 mM EGTA, 0.5% Triton X-100, and protease inhibitors) supplemented with 4 μM paclitaxel to maintain microtubule stability for 30 min on ice, and centrifuged. The supernatant containing the solubilized tubulin was clarified with centrifugation (20 000g for 45 min) and separated from the pellet fraction containing the polymerized form of tubulin. The pellet was washed in microtubule stabilization buffer then denatured in Laemmli buffer.
LMP1 does not affect expression of Op18/stathmin
To determine the effect of LMP1 on Op18/stathmin signaling, we first investigated the levels of Op18/stathmin expression after induction by LMP1. Western blot analysis showed that LMP1 did not change the expression of Op18/stathmin, as the CNE1-LMP1 cells contained a similar amount of Op18/stathmin as the CNE1 cells (Fig. 1A). The Op18/stathmin expression in NP69-LMP1 cells was the same as in the NP69 and NP69-pLNSX cells, indicating that LMP1 does not affect Op18/stathmin in NP69 cells (Fig. 1B). We also explored the expression of Op18/stathmin in the Tet-on LMP1 HNE2 cell line, in which the expression of LMP1 is turned on by tetracycline in a dose-dependent manner. Western blot analysis showed that Op18/stathmin expression was not affected by an increase in LMP1 expression induced by doxycycline at different dosages (Fig. 1C). To further confirm the effect of LMP1 on the expression of Op18/stathmin, we examined the expression of Op18/stathmin induced by 0.6 μg/mL doxycycline for 4, 8, 12, and 24 h. The results showed that the expression of Op18/stathmin remained unchanged at the different time points (Fig. 1D). Collectively, these experiments showed that LMP1 does not change the expression of Op18/stathmin, which is consistent with our previous results.
LMP1 promotes the phosphorylation of Op18/stathmin
Phosphorylated Op18/stathmin promotes microtubule polymerization, in contrast to unphosphorylated Op18/stathmin, which causes microtubule depolymerization. To explore whether LMP1 regulates Op18/stathmin signaling, we investigated the change in Op18/stathmin phosphorylation induced by LMP1. As no appropriate or specific anti-phospho-stathmin antibodies were available, the anti-phosphoserine antibody was used to study the phosphorylation status of Op18/stathmin in different cells by immunoprecipitation (IP) analysis. Op18/stathmin was pulled down with the anti-stathmin antibody, and an IP/Western blot assay was carried out to analyze the phosphorylation of Op18/stathmin. As shown in Figure 2(A), LMP1 enhanced the phosphorylation of Op18/stathmin in CNE1-LMP1 cells when compared with control CNE1 cells lacking LMP1. The level of Op18/stathmin phosphorylation was markedly upregulated when Tet-on LMP1 HNE2 cells were treated with 0.6 μg/mL doxycycline (Fig. 2B). The same blots were stripped and reprobed to assess the quantity of Op18/stathmin, which was equal. Furthermore, we confirmed that the expression of LMP1 in NP69-LMP1 cells caused an increase in the phosphorylation of Op18/stathmin when compared with NP69-pLNSX cells (Fig. 2C).
LMP1-induced upregulation of ERK phosphorylation mainly occurs during G1/S phase
The FACS analysis showed that most cells treated with colcemid (61.7% of CNE1-LMP1 cells; 89.1% of CNE1 cells) were blocked at the G2/M phase, whereas the majority of untreated cells (83.9% of CNE1-LMP1 cells; 74.8% of CNE1 cells) remained in the G1/S phase (Fig. S1, Table S1). These data showed that colcemid treatment successfully generated cells arrested at the G2/M stage, which were used in the next experiment. Because the majority of untreated cells accumulated in the G1/S phase (Table S1), we analyzed the expression and phosphorylation status of MAPK in cells in the G1/S phase. These experiments showed that the changes of the phosphorylation of p38 and JNK were weak after LMP1 induction; however, the levels of phosphorylated ERK were clearly elevated in LMP1-induced cells (Fig. 3B). Figure 3(A) shows that after the same blots were stripped and reprobed, the expression levels of p38, ERK, and JNK were not markedly changed in CNE1-LMP1 cells.
These data indicated that LMP1 promoted the phosphorylation of ERK/MAPK more strongly than the other MAPKs, and that ERK/MAPK signaling was the most important pathway regulated by LMP1. LMP1 upregulated the phosphorylation of ERK/MAPK primarily during the G1/S phase. Of the untreated cells, a higher percentage of CNE1-LMP1 cells remained in the G1/S phase when compared with CNE1 cells. This result is consistent with our findings from a previous study in which LMP1 caused an accelerated entry into the G1/S phase and promoted cell proliferation.
LMP1 negatively regulates ERK phosphorylation during G2/M phase
After showing that LMP1 mainly promoted the phosphorylation of ERK/MAPK activity in G1/S phase cells, we asked whether MAPK activity could be induced by LMP1 in G2/M phase cells. We arrested the majority of the cells at the G2/M phase with colcemid, as described above. Western blot analysis revealed that LMP1 expression did not affect the levels of phosphorylation of p38 or JNK in the G2/M phase, but there was a marked reduction in the level of phosphorylation of ERK in CNE1-LMP1 cells during the G2/M phase when compared with control CNE-1 cells (Fig. 4B). When the same blots were stripped and reprobed, no obvious differences in the expression levels of p38, ERK, or JNK were observed when comparing the CNE1 and CNE1-LMP1 cells in the G2/M phase (Fig. 4A). These findings indicated that the regulation of MAPK activation by LMP1 was closely associated with cell cycle progression (i.e., LMP1 negatively regulated ERK activity during the G2/M phase but upregulated ERK activity during the G1/S phase) and that the regulation of MAPK mediation occurred mainly through the ERK/MAPK signaling pathway.
Op18/stathmin siRNA effectively inhibits Op18/stathmin to depolymerize microtubules
To further confirm the existence of the LMP1-ERK-Op18/stathmin signaling pathway, we explored the change in the ability of Op18/stathmin to destabilize microtubules. We constructed a recombinant plasmid carrying an siRNA that targets the coding region of Op18/stathmin. The results show that the siRNA effectively inhibited mRNA transcription of Op18/stathmin in the pGCsi.U6/neo/GFP-RNAi-transfected CNE1-LMP1 cells. Analysis using RT-PCR showed that the 167-bp PCR product of the Op18/stathmin gene was decreased after pGCsi.U6/neo/GFP-RNAi transfection compared with pGCsi.U6/neo/GFP and pGCsi.U6/neo/GFP-NON transfections. No obvious difference in the mRNA transcription of Op18/stathmin was detected in the pGCsi.U6/neo/GFP- and pGCsi.U6/neo/GFP-NON-transfected CNE1-LMP1 cells. The 250-bp PCR product for the β-actin gene did not change following transfection with these plasmids (Fig. 5A). Western blot analysis also showed that siRNA expression markedly inhibited the expression of Op18/stathmin in pGCsi.U6/neo/GFP-RNAi-transfected CNE1-LMP1 cells; however, pGCsi.U6/neo/GFP and pGCsi.U6/neo/GFP-NON transfection did not affect the expression of Op18/stathmin (Fig. 5B). The Western blot analysis of solubilized tubulin showed that the level of solubilized tubulin decreased following pGCsi.U6/neo/GFP-RNAi transfection and that the level of solubilized tubulin did not change following the transfection of the same cells with pGCsi.U6/neo/GFP or pGCsi.U6/neo/GFP-NON (Fig. 5C).
These findings indicate that siRNA effectively inhibited the expression of Op18/stathmin and downregulated the activity of Op18/stathmin, resulting in a decrease in the level of solubilized tubulin. Op18/stathmin is an active molecule that regulates microtubule dynamics and promotes microtubule depolymerization in NPC, thus providing indirect evidence that the regulation of MAPK-mediated Op18/stathmin signaling induced by LMP1 causes a change in microtubule dynamics.
Inhibition LMP1 expression attenuates interaction between ERK and Op18/stathmin and promotes microtubule depolymerization
DZ1, an LMP1-specific DNAzyme, blocks LMP1 expression. Analysis by co-IP and Western blotting showed that the presence of DZ1 resulted in the downregulation of the phosphorylation of Op18/stathmin; the same blots were stripped and reprobed to show a decrease in the interaction of Op18/stathmin with ERK (Fig. 6A). To further determine whether LMP1 repression affects ERK-mediated Op18/stathmin signaling, the status of solubilized and polymerized tubulin was analyzed. Western blot analysis showed that the inhibition of LMP1 expression greatly downregulated the level of polymerized tubulin, which was consistent with the change in phosphorylation of Op18/stathmin (Fig. 6B). DZ1 effectively inhibited the expression of LMP1 (Fig. 6C). These results confirmed that LMP1 regulated ERK-mediated Op18/stathmin signaling and further affected microtubule dynamics.
Selective downregulation of ERK signaling inhibits the phosphorylation of Op18/stathmin and microtubule polymerization
Because ERK1/2 is the main MAPK pathway regulated by LMP1, we selectively blocked ERK/MAPK signaling in CNE-LMP1 cells using PD98059, a MEK1 inhibitor that specifically inhibits the activation of ERK/MAPK signaling (Fig. 7A). As shown in Figure 7(B), the phosphorylation of Op18/stathmin was reduced after treatment with increasing doses of PD98059. After treatment with increasing doses of PD98059, the changes in the levels of solubilized and polymerized tubulin in CNE1-LMP1 cells were also analyzed, revealing a marked increase in solubilized tubulin at 50 μM PD98059 (Fig. 7C).
The regulation of Op18/stathmin signaling is closely associated with multiple kinases, such as cyclin-dependent kinase, MAPK, and Ca2+/calmodulin-dependent protein kinase type IV/Gr,[21, 22] and affects cell cycle progression and cell proliferation. High levels of Op18/stathmin expression have been associated with poorly differentiated, highly proliferative ovarian cancer and breast carcinomas.[23, 24] Similarly, levels of Op18/stathmin phosphorylation are also markedly increased in poorly differentiated lung adenocarcinomas compared with moderately or highly differentiated lung adenocarcinomas. The transfection of wild-type Op18/stathmin in leukemia cells has been shown to lead to cell growth repression.
LMP1 can induce cell proliferation. In Burkitt's lymphoma cells, EBV-encoded LMP1 has been shown to induce interleukin-10 expression by the p38/SAPK2 pathway to regulate cell growth.[27, 28] Our previous study showed that LMP1, encoded by EBV, can trigger the expression of Survivin, an apoptosis inhibitor. These findings suggest that LMP1 regulation of Op18/stathmin signaling may be linked to cell proliferation.
Our results show that LMP1 accelerated the accumulation of CNE1-LMP1 cells in the G1/S phase, which is consistent with our previous observation.[20, 29] The accumulation of G1/S-phase cells aids in accelerating cell proliferation. DNA replication requires the production of three types of RNA (mRNA, rRNA, and tRNA), which are synthesized during the G1/S phase. These types of RNA also provide the energy stores and building blocks for subsequent cell division. It appears that the regulation of ERK-mediated Op18/stathmin signaling by LMP1 is closely associated with cell proliferation. These experiments also indicate that LMP1 upregulated ERK activity and that the interaction of ERK with Op18/stathmin occurred during the G1/S phase. These activities help maintain the stability of the interphase microtubular arrays, which is considered a requirement for protein and nucleotide synthesis for cell proliferation. This study also shows that LMP1 negatively regulated the activation of ERK during the G2/M phase, and a previous study showed that G2/M-phase cells accumulate on LMP1 expression. These findings support the hypothesis that the change in ERK activity induced by LMP1 is closely associated with cell cycle progression. Interestingly, the factors directly involved in LMP1-induced ERK phosphorylation during the G1/S and G2/M phases remain unidentified.
The negative regulation of ERK activity during the G2/M phase may involve multiple factors. The MAPKs can directly or indirectly phosphorylate cyclin B1 to promote its nuclear translocation and cdc2 activity, providing a mechanism by which MAPK can speed up the transition from G2 to M phase.[30, 31] Cdc2, which is mainly activated during the mitotic phase, is a typical cell cycle-dependent kinase and a crucial initiator of the transition to the G2/M phase. Studies have found that cdc2 leads to global transcription repression by directly phosphorylating RNA polymerase II and indirectly inhibiting signal transduction by targeting multiple kinases including receptor tyrosine kinase, Grb2, and Sos-1/Raf-1/MEK during mitosis.[32-35] Our previous study showed that LMP1 upregulated cdc2 activity during the G2/M phase. Despite a general inhibition of cdc2 activity, the downregulation of ERK activity during the G2/M phase may be linked to p53 transactivity,[29, 36] whereas MAPK activation can affect p53 activity under different stress stimuli.[37-41] Similarly, cdc2 can also phosphorylate p53 to enhance its transcription, which eventually leads to the inhibition of MAPK activation. Our studies indicate that LMP1 can downregulate ERK activity at the G2/M phase through multiple mechanisms.
LMP1 may regulate Op18/stathmin signaling through multiple pathways, and the mechanism by which LMP1 regulates Op18/stathmin signaling varies according to the cell cycle phase. It has been confirmed that cdc2 can directly phosphorylate Op18/stathmin Ser 25 and Ser 38. An EBV-encoded kinase, BGLF4, can phosphorylate Op18/stathmin and decrease its ability to regulate microtube dynamics during the lytic stage of EBV. It remains unknown whether BGLF4 contributes the ability of EBV-encoded proteins such as LMP1 to regulate Op18/stathmin during the latent stage of EBV. Moreover, EBNA1, a viral nuclear protein encoded by EBV, increases Op18/stathmin expression, further indicating the role of EBV in the regulation of Op18/stathmin.
We speculate that LMP1 upregulates ERK activity and promotes the interaction of ERK with Op18/stathmin during the G1/S phase. These activities help maintain the stability of interphase microtubules, thus ensuring adequate levels of protein and nucleotide synthesis that are required for cell proliferation. During the G2/M phase, LMP1 negatively regulates ERK activity in multiple ways, such as through the activities of cdc2 and p53. Moreover, LMP1 promotes G2/M phase by cdc2-mediated Op18/stathmin. Therefore, both cdc2 kinase-mediated and ERK-mediated Op18/starthmin activation is involved in cell proliferation in different cell cycle phases in NPC. The inhibition of ERK activity may be helpful in preserving the energy needs for the dramatic changes in cellular structure that occur during the mitotic phase.
Our results reveal that LMP1 regulated MAPK-mediated Op18/stathmin signaling, mainly through the ERK/MAPK signaling pathway. We also determined that the change in ERK activity induced by LMP1 was associated with cell cycle progression. These studies not only add to the known LMP1 regulation network but also provide new insights into elucidating the molecular mechanisms of LMP1 that lead to tumorigenesis.
We would like to thank the members of our laboratory for critical discussions of this manuscript. This work was supported by the National Basic Research Program of China (Grant No. 2011CB504300), the National Natural Science Foundation of China (Grant Nos 30930101, 30873010, and 81028012), and the Fundamental Research Funds for the Central Universities (Grant No. 2011JQ019).