I.INTRODUCTION

Cisplatin (cis-diamminedichloroplatinum(II)) is one of the most widely used cytotoxic anticancer drugs, but its clinical efficacy is often limited by primary resistance or the development of secondary resistance^1,2). Cisplatin is primarily considered a DNA-damaging anticancer drug that forms different types of bifunctional adducts with cellular DNA³⁾. Cisplatin modulates various signaling pathways that prevent or promote cell death, and its effects include DNA damage, DNA repair, cell proliferation, transcription, and extrinsic and intrinsic apoptosis^4,5). Cellular mechanisms of resistance to cisplatin are multifactorial and include reduced drug accumulation, deactivation of drugs by thiols, and increased DNA repair^3,6). The dominant expression of RAS oncogenes produces resistance to cisplatin by reducing drug uptake and increasing DNA repair. Furthermore, c-fos/AP-1 complex, a master switch for turning on other genes in response to DNA-damaging agents, has been shown to play a major role in cisplatin resistance. AP-2 transcription factors, modulated by protein kinase A, have also been implicated in cisplatin resistance via the regulations of genes encoding for DNA polymerase beta and metallothioneins⁷⁾. In addition, alterations in the expressions of oncogenes, such as, c-fos, c-myc, H-ras, c-jun, and c-abl, and in tumor suppressor genes, like p53, have also been implicated in cellular resistance to cisplatin^5,8). Changes in the expressions of regulatory proteins involved in signal transduction pathways that control the apoptotic pathway can also affect sensitivity⁸⁾, and thus, protein expressional alterations induced by cisplatin can provide important information regarding anticancer efficacy at the molecular level^6,9).

Although cisplatin is known to have anticancer effects in different cell lines, such as, lung cancer, breast cancer, renal cancer, and oral cancer cells, this study was performed to examine protein expression changes in the KB cell line (an oral squamous cell carcinoma cell line), which provides a useful experimental system for studies on different types of anticancer therapies^10-12). In a previous study of DNA damage induced by irradiating KB cells, cells underwent apoptosis in a time and dose dependent manner after irradiation and showed G2 arrest accompanied by the upregulations of p53, ubiquitous CDK-Is, and S and G2 accelerator genes11). Furthermore, KB cells are highly sensitive to hexadecylphosphocholine (HePC, half-maximal growth inhibitory concentration (IC50) 1.2 uM; half lethal concentration (LC50) 2.8 uM) and slowly adapt to increasing concentrations of HePC¹²⁾, and thus, it is presumed that KB cells were appropriate for studying molecular signaling in the context of cisplatin-induced anticancer effect.

Immunoprecipitation-based high performance liquid chromatography (IP-HPLC) was developed to explore molecular signaling in biological samples^13,14). The technique is similar to the enzyme-linked immunosorbent assay (ELISA), but it uses antibody-attached protein A/G beads instead of antibody-conjugated wells. However, the immunoassay mechanism involved is the same, though IP-HPLC enables more multiple reactions to be conducted simultaneously than ELISA. In addition, because of the HPLC component, IP-HPLC can produce more precise data than ELISA. Nevertheless, the efficiency of IP-HPLC has been not evaluated in the context of molecular or cell biological experimentation.

This study was performed to determine the fidelity of IP-HPLC by examining protein expressional changes in KB cells treated with cisplatin. In order to confirm the accuracy of IP-HPLC data, KB in vitro experiments were performed repeatedly and results were analyzed statistically. Using ELISA as a control was not available because 83 target proteins were included in this study, nevertheless results obtained demonstrate relative protein expression changes(%) between experiments and controls. Therefore, the present study was undertaken to devise an IP-HPLC method and to analyze the analytical data produced in the context of the cisplatin treatment of KB cells (an oral squamous cell carcinoma cell line).

II.MATERIALS AND METHODS

KB cells (ATCC, USA) were cultured in Dulbecco’s modified Eagle’s medium (WelGene Inc. Korea) supplemented with 10% (vol/vol) heat-inactivated fetal bovine serum (WelGene Inc. Korea), 100 units/mL penicillin, 100 ug/mL streptomycin, and 250ng/mL amphotericin B (WelGene Inc. Korea) in a 5% CO₂/95% air atmosphere at 37.5°C. The cells were tested for mycoplasma on a regular basis to ensure that only mycoplasma-free cell lines were submitted for assay.

During the active growth period cells in the experimental group was treated with 10 ug/mL cisplatin (Sigma, USA; IC50 value for cisplatin treatment in human lung adenocarcinoma SLC-89 cells is 18.47 ug/mL)¹⁵⁾ for 12 or 24 hours, while controls were not treated. About 1x10¹⁰ KB cells were harvested per experimental condition and immediately lysed with protein extraction buffer (PRO-PREP^TM, iNtRON Biotech., USA) on ice. The protein extracts from experimental and control groups were centrifuged at 1600g for 30 min, and the supernatants were collected and stored in a deep freezer at -70°C until required.

The protein concentrations of supernatants were measured using a protein assay solution (Bio-Rad Laboratories, Inc., USA) at 592 nm using a UV spectrophotometer, and about 1 mg of protein was used for each immunoprecipitation procedure, which was conducted using a protein A/G agarose column (Amicogen, Korea). Protein A/G agarose columns were separately pre-incubated with about 1 ug of each of 83 different antisera, which included growth factor-related proteins, proliferation-related proteins, transcription signaling proteins, apoptosis-related proteins, cell survival-related proteins, tumor suppressor proteins, oncoproteins, protection proteins, proinflammatory proteins, WNT/β-catenin pathway proteins, matrix proteolysis proteins, angiogenesis-related proteins, and cytoskeletal proteins(Table 1).

Briefly, protein samples were mixed with 5 mL of binding buffer (150 mM NaCl, 10 mM Tris pH 7.4, 1 mM EDTA, 1 mM EGTA, 0.2 mM sodium vanadate, 0.2 mM PMSF, 0.5% NP-40), and incubated in the protein A/G agarose columns at room temperature for 1 hour. The columns were placed on a rotating stirrer. After washing each column with a sufficient amount of PBS solution (pH 7.3, 137 mM NaCl, 2.7 mM KCl, 43 mM Na₂HPO₄-7H₂O, 1.4 mM KH₂PO₄), target proteins were eluted with 250 uL of IgG elution buffer (Pierce, USA). Immunoprecipitated proteins were analyzed by HPLC (Hewlett Packard, Agilent, USA) using a reverse phase column (YMC-Pack SIL, SL12S05-2506WT, USA) and 0.05M NaCl/5% acetonitrile solution as eluent at 0.6 mL/min for 30 min^13,14). IP-HPLC was performed on control and experimental groups simultaneously. Protein peak areas obtained by HPLC were calculated after subtracting control antibody peak areas, and protein expression changes (%) between experiments and controls were calculated. Protein peak areas (mAU*s) and protein expression changes (%) were plotted for different molecular signaling pathways.Table 2

III.RESULTS

The anticancer effect of cisplatin in KB cells by IP-HPLC was conducted using protein expression changes (%), as described above. Changes in the protein expressions of different biomarkers were examined after 12 or 24 hours of cisplatin treatment.

For cytoskeletal proteins, the expressions of α-tubulin (-27.7%), α-actin (-45.4%), and β-actin (-90.3%) were reduced after 12 hours of cisplatin treatment, and after 24 hours of treatment (-15.9%, -30.6%, and -57.9%, respectively).

For epigenetic factors, histone H1 (13%) and demethylase (61.4%) expressions increased but deacetylase (-12.2%) expression decreased after 12 hours, whereas the expressions of histone H1 (-57.8%) and demethylase (-49.7%) decreased and the expression of deacetylase (84.3%) greatly increased after 24 hours.

For growth factor-related proteins, the expressions of EGFR (1.7%), c-erbB2 (68.9%), and bFGF (39.2%) increased but TGF-β1 (-3.6%) and HGF (-31.3%) decreased after 12 hours, while EGFR (7.1%), TGF-β1 (4.8%),and HGF (40.6%) increased but c-erbB2 (-56.8%) and bFGF (-12%) decreased at 24 hours.

For proliferation-related proteins, DHS (94.9%), PCNA (31.4%), MPM-2 (10.3%), hTERT (20.9%), and lamin A (105.4%) increased but eIF5A (-0.8%), DOHH (-24.2%), CDK4 (-54.8%), cMyc (-60.2%), and MAX (-36.4%) decreased after 12 hours of cisplatin treatment, while MPM-2 (21.6%), hTERT (27.1%), lamin A (2.6%), DOHH (31%), cMyc (40%), and MAX (21%) increased but DHS (-27.8%), PCNA (-35.5%), eIF5A (-40%), and CDK4 (-21.8%) decreased after 24 hours of treatment.

For transcription signaling proteins, NFkB (7%), p38 (20.5%), and E2F-1 (30.5%) increased after 12 hours, while NFkB (203.7 %) greatly increased but p38 (-36.3%) and E2F-1 (-22.5%) decreased after 24 hours.

For the apoptosis-related proteins NOXA (64.4%), PUMA (44.2%), BAX (56.3%), BAD (117.3%), BAK (64.6%), caspase 3 (51.3%), caspase 9 (246.5%), and FADD (14.4%) increased but FAS (-5.9%), FASL (-85.2%), PARP (-72.8%), BID (-55.4%), caspase 8 (-49.2%), and FLIP (-66.9%) decreased at 12 hours, while PUMA (37.8%), BAX (39.1%), BAD (125.5%), BAK (157.1%), caspase 3 (12.3%), FADD (6.9%), FAS (117%), FASL (27.2%), PARP (42%), BID (2.1%), and FLIP (5.7 %) increased but NOXA (-54.5%), caspase 8 (-30.7%), and caspase 9 (-39.4%) decreased after 24 hours of cisplatin treatment.

For cell survival-related proteins, the expressions of pAKT (74.8%) and PKC (21.1%) increased after 12 hours of treatment but those of MDM2 (-20.6%) and BCL-2 (-66.6%) decreased, and after 24 hours, MDM2 (21%) increased but BCL-2 (-34%), pAKT (-15.5%) and PKC (-49.1%) decreased.

The expressions of the tumor suppressor proteins p16 (21.7%), p53 (138.4%), RB1 (63.7%), and NF-1 (146.7%) increased but p21 (-10%), p63 (-51.7%), PTEN (-23.6%), and PTCH (-42.5%) decreased after 12 hours of treatment, while p16 (110.7%), p21 (37.7%), and PTEN (131.1%) increased but p53 (-12.5%), RB1 (-48.7%), NF-1 (-57.5%), p63 (-67.3%), and PTCH (-42.8%) decreased after 24 hours of treatment.

For oncoproteins, 14-3-3 (179.6%), CEA (87.3%), and DMBT1 (35.8%) increased but STAT3 (-53.2%), survivin (-45.6%), maspin (-64.4%), snail (-31.5%) and K-RAS (-29.2%) decreased after 12 hours of treatment with cisplatin, while 14-3-3 (35.8%), CEA (148.7%), DMBT1 (69.5%), maspin (64.5%), snail (31.4%), and K-RAS (14.9%) increased but STAT3 (-57.9%) and survivin (-4.5%) decreased after 24 hours.

For differentiation factors, pan-K (16.4%), CK-7 (36.6%), vimentin (34.8%), E-cadherin (2.7%), TGase-K (30.6%) increased after 12 hours of treatment, but CK-14 (-69.5%) and TGase-E (-47.7%) decreased, while after 24 hours, pan-K (72.9%), CK-7 (17.2%), TGase-E (16.5%) increased but vimentin (-32.8%), E-cadherin (-45.1%), TGase-K (-51.5%), and CK-14 (-44.4%) decreased.

For protection proteins, HO-1 (16.1%), caveolin (72.1%), and FAK (77.1%) increased but HSP-70 (-61.6%) decreased after 12 hours of treatment, while FAK (35%) and HSP-70 (124.1%) increased but HO-1 (-26%) and caveolin (-44.4%) decreased after 24 hours.

For the proinflammatory proteins, TNFα expression was decreased after 12 (-82.6%) and 24 hours (-31.9%).

For WNT/β-catenin pathway proteins, APC (189.3%) expression greatly increased but the expressions of SHH (-48.2%), β-catenin (-59.1%), and WNT1 (-23.5%) decreased after 12 hours, while APC (7.2%) slightly increased but SHH (-15.3%), β-catenin (-47.3%), and WNT1 (-48.5%) decreased after 24 hours.

For the matrix proteolysis proteins, at 12 hours, the expressions of MMP-1 (20.3%), MMP-3 (31.8%), and MMP-10 (162.8%) were increased, but the expressions of MMP-2 (-48.8%) and MMP-9 (-27.3%) were decreased, while after 24 hours, the expression of MMP-9 (95.2%) was greatly increased but the expressions of MMP-1 (-52%), MMP-2 (-20.5%), MMP-3 (-60.2%), and MMP-10 (-21.5%) were decreased.Fig .1,2,3,4

For angiogenesis-related proteins, at 12 hours, HIF-1 (6.6%) and VEGF (4.1%) were increased but vWF (-27.8%) and angiogenin (-24.4%) were decreased, while after 24 hours VEGF (28.5%) was increased but HIF-1 (-35.2%), vWF (-10.7%), and angiogenin (-38.1%) were decreased.

IV.DISCUSSION

Cisplatin, a representative anticancer agent, can interact with genomic DNA by direct incorporation and depress the expressions of many genes in cells. However, gene regulation by cells may act to overcome the inhibitory effects of cisplatin¹⁶⁾. The present study was conducted to explorer genome-wide gene expression changes in KB cells caused by cisplatin treatment

The expressional changes caused by cisplatin depressed many essential proteins, including cytoskeletal home keeping proteins, proliferation-related proteins, transcription signaling proteins, and growth factor-related proteins, while other protein expressions were elevated to compensate for the genetic damage caused by cisplatin. In addition, the present study shows changes in the expressions of essential proteins in KB cells were also influenced by exposure time, that is, 12 or 24 hours.

The expressions of cytoskeletal proteins, α-tubulin, α -actin, and β-actin, were greatly reduced by cisplatin at 12 hours, but recovered slightly at 24 hours, indicating cellular atrophy and degeneration consistently progressed in the presence of cisplatin, possibly resulting in cellular necrosis and apoptosis.

Besides the theoretical property of cisplatin, that is, direct Pt-DNA adduct formation⁶⁾, the present study shows that cisplatin depresses the epigenetic environment. In particular, with regard to the expressions of the epigenetic factors, demethylase, and deacetylase, it is considered that the involved DNA was initially activated by increased demethylase and decreased deacetylase after 12 hours of cisplatin treatment, but this primary DNA activation was reversed after 24 hours.

For the growth factor-related proteins, EGFR, c-erbB2, bFGF, TGF-β1, and HGF, cisplatin variably affected protein expressions depending on treatment time. In particular, EGFR (a member of the ErbB family of receptors) showed only a slight increase in protein level, that is, 1.7% and 7.1% after 12 and 24 hours, respectively. Because EGFR can permit the stimulations of various cell signaling pathways that mediate gene transcription and cell cycle progression¹⁷⁾, and higher levels of EGFR expression have been shown to be correlated with malignancy in a variety of cancers, cisplatin may affect KB cells resistant to cetuximab (a chimeric human-murine monoclonal antibody against EGFR)¹⁸⁾. It has also been shown tumors that secrete greater amounts of TGF-β are more therapeutically resistant in vivo^19,20). In the present study, a slight elevation in TGF-β1 protein expression was observed after cisplatin treatment (-3.6% and 4.8% at 12 and 24 hours, respectively), which suggests cisplatin may induce KB cells to be resistant to the various types of anticancer therapy^21,22).

Of the proliferation-related proteins, the expressions of CDK4, cMyc, MAX, eIF5A, and DOHH were reduced after 12 hours of cisplatin treatment, but the expressions of some supportive proteins, that is, MPM2, hTERT, lamin A, DOHH, cMyc, and MAX, were much increased after 24 hours of cisplatin treatment. In particular, cMyc and MAX were down-regulated after 12 hours of cisplatin treatment, by -60.2% and -36.4%, respectively, but up-regulated after 24 hours, by 40% and 21%, respectively. These time-dependent changes in polyamine metabolism and c-Myc expression have been previously reported in the kidneys of mice treated with cisplatin, and it has been shown cisplatin significantly induces the expressions of two enzymes, namely, ornithine decarboxylase (ODC) and spermidine/spermine N1-acetyltransferase (SSAT), which are required for the homeostasis of cellular polyamines²³⁾. Therefore, it appears the genetic damage caused by cisplatin DNA adduct formation can stimulate DNA repair programs via cell cycle progressing and protective signaling.

The expressions of the transcription factors p38 and E2F-1 were elevated at 12 hours, but depressed at 24 hours. Furthermore, p38 (the stress-activated protein kinase) was up-regulated by 20.5% at 12 hours, but down-regulated by -36.3% at 24 hours, presumably this was also associated with the induction of a drug-resistant phenotype by cisplatin²⁴⁾. On the other hand, NFkB (203.7%) was intensely overexpressed at 24 hours, which might be related to cisplatin resistance via negative modulation by the farnesoid X receptor (FXR) pathway^25,26), and the inhibition of the mitochondria-mediated activation of caspases²⁷⁾.

Cisplatin is a potent DNA-damaging anticancer agent, and its cytotoxic action is exerted by the induction of apoptosis. The expressions of proteins involved in endogenous apoptosis signaling via the mitochondria-related apoptotic pathway (NOXA, PUMA, BAX, BAD, BAK, caspase 9, and caspase 3,) were activated after 12 hours of cisplatin treatment but inactivated at 24 hours, while the exogenous apoptosis signaling pathway, which includes FASL, FAS, FADD, FLIP, PARP, BID, and caspase 8, became activated at 24 hours.

The cell survival-related proteins, pAKT and PKC, were initially up-regulated at 12 hours, but down-regulated at 24 hours. In particular, MDM2 (degrades p53) was up-regulated at 24 hours, which would have resulted in a rapid reduction in mitochondria-dependent apoptosis via caspase 9²⁸⁾.

Of the tumor suppressor proteins examined, p53, RB1, and NF-1 were overexpressed at 12 hours, but reversed at 24 hours, whereas p21 and PTEN were down-regulated at 12 hours and up-regulated at 24 hours. In particular, the overexpression of p53 at 12 hours would be relevant to DNA damage by cisplatin, and would cause mitochondria-mediated apoptosis^29,30). However, this overexpression was rapidly reduced at 24 hours, and thus, it appears in KB cells the apoptotic effect of cisplatin is evident only immediately after treatment (12 hours), but gradually disappears of the next 12 hours.

It has been proposed that the modes of cellular apoptosis caused by diverse anticancer drugs and radiation therapy may be under the control of several oncogenes³¹⁾. Many oncoproteins including 14-3-3, CEA, DMBT1, maspin, snail, and hTERT, were up-regulated after 24 hours of cisplatin treatment. These markers might predict response to anticancer therapy, and be associated with resistance to cisplatin^32,33). On the other hand, most of the WNT/β-catenin pathway proteins, including SHH, β -catenin, and WNT1 were found to be consistently down-regulated, although APC was up-regulated by 189.3% at 12 hours and by only 7.2% at 24 hours.

The expressions of the differentiation factors pan-K (16.4%), CK-7 (36.6%), vimentin (34.8%), E-cadherin (2.7%), TGase-K (30.6%) increased at 12 hours, whereas those of CK-14 (-69.5%) and TGase-E (-47.7%) were reduced, while at 24 hours, the expressions of pan-K (72.9%), CK-7 (17.2%), and TGase-E (16.5%) were increased but those of vimentin (-32.8%), E-cadherin (-45.1%), TGase-K (-51.5%), and CK-14 (-44.4%) were decreased.

Regarding the protection proteins, the expressions of HO-1, caveolin, and FAK were up-regulated, by 16.1%, 72.1%, and 77.1%, respectively, at 12 hours, but down-regulated by -26%, -44.4%, and 35%, respectively, at 24 hours, whereas HSP-70 was down-regulated by -61.6% at 12 hours and up-regulated by 124.1% at 24 hours. These findings suggest that the cytotoxic effect of cisplatin greatly and differentially influences the signaling of protective gene pathways.

Contrary to the long held belief that chemotherapy is immunosuppressive, emerging evidence indicates that the activity of cisplatin is not limited to its ability to inhibit cellular mitosis, but that cisplatin also has important immunomodulatory effects³⁴⁾. In the present study, the expression of the proinflammatory protein TNFα was down-regulated at 12 hours (-82.6%) and at 24 hours (-31.9%). The expressions of MMP-1, MMP-3, and MMP-10 were up-regulated at 12 hours (20.3%, 31.8%, and 162.8%, respectively), but down-regulated at 24 hours (-52%, -60.2%, and -21.5%, respectively), whereas MMP-9 was initially down-regulated (-27.3%) at 12 hours and up-regulated (95.2%) at 24 hours. Making it evident that cisplatin affects the immunomodulatory environment in a time-dependent manner during therapy.

Anti-angiogenic therapy represents one of the more promising treatment modalities for cancer. For example thalidomide (alpha-N-phthalimidoglutarimide) is a potent inhibitor of angiogenesis, and has been reported to overcome classical drug resistance in human multiple myeloma cells^35.) In the present study, the majority of angiogenesis-related proteins, HIF-1, vWF, and angiogenin, were down-regulated at 24 hours, whereas VEGF was up-regulated (by 28.5%). Increased levels of VEGF suggest a poor prognosis in cisplatin based cancer therapy³⁶⁾, and is viewed as a potential marker of tumor cell survival.

In summary, the author used IP-HPLC to perform a preliminary analysis of the anticancer effect of cisplatin in KB cells. The results demonstrate expressional changes of proteins with widely different functions after 12 and 24 hours of cisplatin treatment. In order to address the fidelity of IP-HPLC protein expression changes (%) of essential proteins were compared between experiment and untreated groups. Cisplatin was found to dramatically down-regulate the expressions of cytoskeletal proteins, proliferation-related proteins, matrix proteins and differentiating factors, and to affect genome wide signaling, including, apoptosis, cell survival, protection, and angiogenesis, by up- or down-regulating proteins in a variable treatment time-dependent manner. Based on the results obtained, it appears that cisplatin produces different cellular responses that include cellular protection and survival, proinflammatory and immunological modulation, and de novo angiogenesis in addition its anticancer effect. Furthermore, it was found that IP-HPLC can produce accurate measures of relative protein expression changes (%), and that it is useful for determining genome-wide protein expression changes in cell and molecular biological experiments.